Keywords: data-driven approach, model-based approach, latent variables
Abstract: Machine learning and control engineering have evolved over years, mutually influencing each other. Through the history of their interaction, there has been a debate over whether to prioritize models or data. Especially in the last decade, influenced by the success of deep learning, data-driven approaches have gained significant importance in control engineering, as in many other fields. On the other hand, it is interesting to note that in machine learning, there has been a growing interest in embedding physics into machine learning models instead of relying on expensive big data. These trends imply that it is more important to integrate models and data, rather than choosing between them. The speaker believes that the key to achieving the integration of both is extracting low-dimensional latent spaces from the high-dimensional observations of the real-world systems and capturing the dynamics of the systems in the latent spaces. Therefore, in this talk, I will focus on the concept of 'latent space' and review the studies on learning dynamical systems, a common topic in both fields. Additionally, I will present case studies on simultaneous estimation of state/models, reinforcement learning, residual life prediction, and other related research based on this concept.

Keywords: Modeling, System Identification and Estimation, Identification and Estimation, Sensors and Transducers
Abstract: Localization is one of the critical techniques for self-driving, for which strict estimation is essential to maintain safety. Sensor fusion is the method that uses multiple sensor information to improve accuracy and robustness. However, more discussion is needed to measure the delay between sensors in integrating information obtained from different sensors. Therefore, we propose a periodic disturbance observer-based method to calculate a delay. This paper explains how to estimate a delay using an observer and then practically measure a delay with an electric wheelchair. The proposed method measured the delay between a tracking camera and a rotary encoder equipped in the wheelchair with an accuracy of 1 ms.

Keywords: Modeling, System Identification and Estimation
Abstract: This paper addresses a nonlinear system identification problem and proposes a model that combines neural networks with FIR (Finite Impulse Response) model. By highlighting the similarity between FIR model and temporal neural networks, the combined model can be regarded as a single DNN (deep neural network). This interpretation enables efficient training using deep learning frameworks. Furthermore, the paper introduces the application of Kernel-based regularization, a technique developed in the linear system identification field, to the FIR part of the proposed model. This method enhances model accuracy by leveraging prior knowledge that the impulse response of a stable linear system decays exponentially. The effectiveness of the proposed approach is validated through numerical simulations.

Keywords: Modeling, System Identification and Estimation
Abstract: In recent years, artificial intelligence has been increasingly utilized for the automation of waste treatment plant operations. In order to automate the operation of the targeted waste treatment plant, it is necessary to predict the future operating conditions. Therefore, in this paper, an AI model for predicting the operating conditions is constructed. As a modeling approach, a nonlinear system identification approach using deep learning is employed. In particular, the block-oriented model, which is a model structure commonly used for nonlinear system identification, is regarded as a deep neural network, and its parameters are trained collectively using deep learning frameworks.The model is built using actual operating data, and the effectiveness of the proposed method is evaluated.

Keywords: Modeling, System Identification and Estimation, Identification and Estimation, Process Control
Abstract: This study introduces a framework leveraging transfer learning within Gaussian Process Nonlinear Autoregressive with Exogenous Inputs (GP-NARX) models to address system identification challenges in dynamic, data- scarce dynamical systems, with a focus on Sim-to-Real applications. It adeptly employs transfer learning to navigate the complexities of transferring knowledge from extensively studied source environments to novel, less understood target contexts, particularly emphasizing cross-environment adaptation. The framework showcases significant improvement in predictive accuracy, demonstrating considerable reductions in mean-squared error and achieving marked improvements in log predictive density and mean standardized log loss across different environmental settings. Rigorous thermal control system lab evaluations and detailed illustrative examples validate the framework's effectiveness in bridging knowledge gaps between disparate environments, positioning it as potentially scalable solution for Sim-to-Real challenges. This method improves identification and modelling of dynamical systems across different environments, advancing our understanding of them.

Keywords: Modeling, System Identification and Estimation, Identification and Estimation, Signal and/or Image Processing
Abstract: In this study, we develop a sensor calibration system that measures the position of a diagonally thrown color ball using an RGBD sensor D455 and estimates the gravitational acceleration using a Kalman filter. Based on the estimated gravitational acceleration, we show that the system can correct the point cloud obtained by the depth sensor to the horizontal direction even when the sensor is set at an arbitrary orientation.

Keywords: Modeling, System Identification and Estimation, Nonlinear Control, Mechatronics Systems
Abstract: The displacement of piezoelectric actuators is often controlled by varying the applied voltage though the hysteresis of the displacement to the applied voltage exists. In this paper, the mechanical compliance is represented by a Cauer circuit in Martin's model, which is a model of stacked piezoelectric actuators. The charge amount of each capacitance is assumed to vary based on a hyperbolic tangent function, and their parameters are estimated from a hysteresis loop between the displacement and applied voltage. The displacement with hysteresis can be represented as well as the relationship between the charge amount and the displacement.

Keywords: Modeling, System Identification and Estimation, Robust Control, Micro and Nano Devices
Abstract: This paper develops a model estimation mechanism and integrates it with the multiple switching control for a precision piezoelectric transducer (PZT) stage to improve its positioning accuracy. The model estimation mechanism can update the model in real time. At the same time, the multiple-switching control can accurately forecast system responses based on the estimated model and select the optimal control sequence to improve the positioning performance. Finally, we implement the control architecture to the PZT stage to demonstrate the effectiveness of the proposed model estimation method in improving precision positioning.

Keywords: Modeling, System Identification and Estimation, Social Systems, Multivariable Control
Abstract: A pedestrian flow model on the four-way roundabout is considered by employing the mixed logical dynamical (MLD) system representation with model predictive control. Introducing an acceptance rate of directional guidance for pedestrians, the efficiency of crowd flow is investigated from the viewpoints of the number of pedestrians and the acceptance rate of following. The features of the proposed model are discussed based on simulation, and it is shown that adjusting the acceptance rate along with the number of crowd plays a key role in improving the flow efficiency.

Keywords: Signal and/or Image Processing
Abstract: This paper introduces a multi-channel A/D converter for multi-sensor systems. In conventional Delta-Sigma ADCs, there is a problem of latency in the digital filter when switching inputs during multi-channel A/D conversion. A novel interleaved Delta-Sigma ADC with incorporated analog memories is successfully implemented for multiple-sensor interfacing circuits without being affected by latency.

Keywords: Signal and/or Image Processing
Abstract: The aim of this study was to investigate the feasibility of using a 3D point cloud estimated by NeRF. For this purpose, we evaluated the quality of point clouds reconstructed with NeRF and multi-view stereo. The point cloud reconstructed by NeRF contains substantial noise and shows a wavy surface even on a flat surface. Therefore, a point cloud generated by NeRF is not appropriate to use as a highly accurate 3D model with the size and shape of an actual object. However, it is considered to be sufficiently usable for applications that do not require highly accurate point clouds, such as the detection and 3D modeling of objects with few feature-point surfaces or optically transparent objects, which have been difficult to achieve with conventional methods.

Keywords: Signal and/or Image Processing, Analytical Measurement, Sensors and Transducers
Abstract: Ultrasonic testing is one of the most successful nondestructive evaluation (NDE) techniques for the inspection of carbon-fiber-reinforced polymer (CFRP) laminates. Though, the anisotropy of elasticity in CFRP shows the prominent directivity of wave propagation governed by the fiber directions. Due to this anisotropic behavior, it is difficult to define the location of defects by using the arriving time of reflected signals. This article proposes the super-resolution imaging method which utilize the point source constrained partial differential equation on Lp space. It can improve the image quality and realizes more detailed inspection of the CFRP laminates with higher accuracy.

This paper shows the geometrical relation between the coordinates in Euclidean space and the coordinates in Lp space, and after the mathematical formulation of the wave propagation on Lp space, the spatio-temporal behavior of the divergent wave front of point-source is analyzed. and several numerical experiments evaluate the reconstructed silhouette of subsurface point-like defect in CFRP plates obtained with using the point source constrained partial differential equation on Lp space.

Keywords: Signal and/or Image Processing, Computational Intelligence
Abstract: In the medical field, datasets of annotated lung sounds for training machine learning models are limited in scale, and the interpretation of lung sound data requires specialized knowledge and experience. Consequently, large-scale models lead to over-fitting and lack of expertise, and even small-scale models necessitate inventive approaches to feature extraction that is different from conventional methods. This paper leverages existing research and domain knowledge in medical diagnosis to propose an effective feature extraction method tailored to lung sounds. The proposed method is indicated to achieve higher classification performance than traditional feature extraction methods.

Keywords: Signal and/or Image Processing, Identification and Estimation, Remote Sensing
Abstract: Coastal change monitoring is essential for obtaining information to mitigate issues stemming from coastal erosion. Changes in sand beach areas reflect sediment transportation, directly correlating with erosion. LiDAR technology has emerged as a new tool for monitoring beach areas, capturing the coastal area in 3D point cloud format. We can monitor coastal changes by comparing two 3D point cloud data. However, we face a challenge distinguishing between changes caused by sediment transportation and caused by noisy point cloud data in rocky areas. To address the problem, we propose a new technique that distinguishes between sandy and rocky regions using a classification tree with two statistical-based features: altitude variance and gradient orientation. 3D point cloud data were collected from a beach in Thailand that is encountering coastal erosion to validate our proposed scheme. Experimental results demonstrate that our method achieved high scores of more than 99.69% across all evaluation metrics, including precision, recall, 𝐹1 score, accuracy, and balanced accuracy, enabling the improvement of the accuracy of further subsequent change detection analysis.

Keywords: Signal and/or Image Processing, Intelligent Systems
Abstract: Particle filters have difficulty continuing tracking when objects move differently from the transition model. Therefore, a method was proposed to incorporate the amount of movement calculated by the convolution filter into the particle filter's state transition model. However, since the accuracy of this method depends on the accuracy of the movement estimation, it is necessary to improve the accuracy of the movement estimation. Therefore, in this paper, in order to improve the accuracy of movement estimation, we perform pre-training using a large amount of data generated by computer graphics, and verify how much the estimation accuracy can be improved.

Keywords: Networked Sensor System, Remote Sensing, Signal and/or Image Processing
Abstract: In recent years, WSN (Wireless Sensor Network) have been attracting attention as society becomes more information-oriented. In WSN, localization of each wireless node is essential, and the location information system must have low power consumption and be usable both indoors and outdoors. Therefore, in this study, we calculated intersection points of the Apollonius circles for each combination of three anchor nodes selected from four anchor nodes, and extracted intersection points using a mesh to improve the accuracy of localization outside the anchor area. By using this method, the accuracy of localization outside the anchor area for a given anchor placement was significantly improved. However, depending on the arrangement of the nodes, there may remain intersection points of false intersection groups, and the localization accuracy will drop significantly, so further consideration is required.

Keywords: Mechanical Systems Control, Computer Aided Design, Modeling, System Identification and Estimation
Abstract: The control technology of a segmented mirror is indispensable to realize an extremely large telescope. In this paper, we improve Distributed Control System (DCS) for controlling the segmented main mirror of the Seimei telescope by allowing new combinations of sensors and by proposing a new time-domain cost function. The effectiveness of the proposed method is demonstrated by numerical simulations

Keywords: Mechanical Systems Control, Identification and Estimation, Modeling, System Identification and Estimation
Abstract: Operability is the one of the important qualities for products such as vehicles. In this study, we investigated ''Sweeping operation'', which is one of the operation patterns that evaluate the operability of hydraulic excavators. We created a hydraulic excavator simulator using Unity and AGX Dynamics, and investigated the correlation between the image information the operator obtaining during ''Sweeping operation'' and the boom operation. In addition, we created the operator’s model during ''Sweeping operation'' based on this data. This paper revealed the relation between the images 　 obtained by operator and boom operation and represented it is possible to model those correlation.

Keywords: Mechanical Systems Control, Process Control, Nonlinear Control
Abstract: In stochastic systems, the analysis of the stability of ``former equilibrium points,'' which have lost their equilibrium state due to the addition of disturbances, is important. In this paper, we conduct the following as a first step to develop a quantitative stability analysis method for ``former equilibrium points'' of stochastic systems. First, we treat smooth stochastic Lyapunov functions to simplify the discussion, but extend them from ``former equilibrium points'' to subsets so as not to limit the class of systems that can be analyzed. Then, we propose a new metric for quantitatively evaluating stability with respect to the target subset.

Keywords: Mechanical Systems Control, Robotic and Automation Systems, Mechatronics Systems
Abstract: In this paper, we propose a construction method for an energy saving robot manipulator using passive storage elements. We propose a joint configuration and control method using Karakuri that eliminates the need for power consumption associated with electronic control of motors. We build a prototype machine based on the proposed method and confirm the effectiveness of the proposed method through experiments.

Keywords: Nonlinear Control, Mechanical Systems Control
Abstract: This study presents a passive control strategy that improves autoparametric vibration absorbers' performance. The cubic nonlinear stiffness of the absorber is eliminated by using the nonlinear characteristic of the magnetic force. Then, the performance of the autoparametric vibration absorber is kept even if the amplitude of absorber is large under the large amplitude excitation to the main system. In contrast, traditional active control methods require external energy sources and complex feedback mechanisms to adjust the absorber’s natural frequency to match external frequencies. The proposed approach involves the strategic placement of permanent magnets around a pendulum-type vibration absorber, configuring its nonlinear characteristics. Theoretical and experimental analyzes demonstrate the varigity of the proposed method to enhance vibration suppression in dynamic systems.

Keywords: Mechanical Systems Control, Modeling, System Identification and Estimation
Abstract: In this study we discuss an overshoot suppression method of a desired step response of a force servo for hydraulic arms. There are many conventional (model based) force servo. However, it will overshoot at the desired step response. First, we review hydraulic cylinder dynamics with dominant nonlinear pressure and the representations based on coordinate and energy transformations using the structural properties. Second, we propose the overshoot suppression method of the desired step response. Third, we verify the effectiveness of the proposed method by a simulation.

Keywords: Factory Automation, Manufacturing Systems, Laboratory Automation
Abstract: The paper presents a model following approach for the plant input mapping (PIM) discretization of the stepping motor current regulator. The PIM method can derive a digital form of the underlying feedback system while preserving its stability for any non-pathological sampling interval. However, the method cannot take the local block property into account, such as an integrator, which is the fatal issue in the motor current regulation problem. We solve this issue through incorporating the model following structure to the prescribed PIM digital controller. Through an experiment using a commercial stepping motor and driver, the proposed model following scheme of the PIM discretization is verified.

Keywords: Factory Automation, Manufacturing Systems, Robotic and Automation Systems
Abstract: Although there is a need to minimize labor for work at heights in shipbuilding and shipping, labor-saving devices are not widely available because of cost issues. Therefore, there is an urgent need to develop inexpensive and reliable wall-moving robots. Unlike conventional wall-climbing robots, this study used an inexpensive, lightweight, robust air cylinder, which has attracted increased attention in recent years, as the main power source. Permanent magnets are used to adhere to a wall surface, but they require enormous forces to pull them off the steel plate and may damage the wall surface owing to impact during adsorption. In this study, we developed an improved internally balanced (IB) magnetic unit using cams. This unit was pneumatically attached and detached. The developed IB magnet mechanism could operate stably even at pressures as low as 0.15 MPa. A wall-climbing device equipped with a tool that moves only in one direction was also fabricated and tested on a laboratory wall. The device weighed 4.0 kg and had a size of 215 mm × 400 mm × 170 mm (W × D × H). This device could climb the wall at a 1.7 m/min speed without carrying a 4.0 kg load, although its stability and forward speed decreased with increasing load.

Keywords: Innovative Systems Approach for Realizing Smarter World, Robotic and Automation Systems, Signal and/or Image Processing
Abstract: We have been developing a new docking method with an extension mechanism and magnetic field control to realize autonomous docking for small spacecraft, to avoid shocks during contact in environments with insufficient hardware resources. The method needs an algorithm to estimate the permanent magnet’s position on the probe’s tip. In this paper, we produce a prototype of a Hall sensor array for position estimation. And evaluate the accuracy, and investigate the characteristics of the algorithm. In the algorithm, we use two functions: a nonlinear function to estimate the distance from each sensor to the permanent magnet and a matrix operation of geometric least squares with sensor mounting positions and the gaps. As a result, we found that we can estimate accurately estimate the positions in the specific area with the algorithm. However, the accuracy drops rapidly outside of the area.

Keywords: Manufacturing Systems, Flow Measurement and Control, Intelligent Systems
Abstract: A vast number of mass flow controllers (MFCs) are used in semiconductor industry. An efficient production method of MFC is required. The gain tuning of the proportional integral (PI) control to realize a setting flow rate is essential for efficient mass production. The gains are tuned to meet the specifications required for evaluation indices of response time and overshoot amount in a step response waveform. In this paper, we propose a simple method for the PI gain tuning using the Gaussian Mixture Model (GMM) and the direct inverse analysis applicable to the pressure based MFCs’ production. The relationship between the gains and evaluation indices for a standard unit of the MFC is modeled as the GMM. The direct inverse analysis calculates the difference between the standard and a test unit. Under the assumption that the difference can be compensated by a simple shift, gains likely to meet the specifications for the test unit are searched. We applied the method to seven test units. The result showed that the gains of all the test units were tuned within only a few iterations.

Keywords: Process Automation
Abstract: In the field of process automation, there has been a lot of research and development focused on utilizing historical process data to enhance production equipment and product quality. One of the outcomes is Multivariate Time series Shape Analysis (MTSA), which is developed for early anomaly detection in batch processes. MTSA has already been adopted in many plants due to its strong capability to detect anomalies in each variable independently without detailed process knowledge. However, as the adoption increases, it has become apparent that the accuracy of anomaly detection needs to be improved for process data that deviates along the time axis. To address this issue, we propose an online alignment method using Dynamic Time Warping (DTW) with model based global constraints. The effectiveness of our proposed method is demonstrated through a numerical experiment.

Keywords: Process Automation, Identification and Estimation, Transportation Systems
Abstract: Currently, skilled technicians primarily use manual tools to maintain railcars’ condition during inspection. However, improving inspection efficiency requires changing the maintenance basis from time-based to condition-based. The development of an automated inspection method is therefore necessary for this change. This study presents the investigation results of an automated railcar number recognition method for railcar condition monitoring. We focus on recognizing a railcar nameplate, including the manufacturer’s name and serial number. Several nameplate variations, including colors, background patterns, and characters are prepared. Experiments are conducted in to obtain automatic measurements and confirm their feasibility. Our aim is to transfer the inspection of railcar from time-based to condition-based maintenance to reduce maintenance costs. A railcar number recognition method using machine vision to identify a unique railcar character and number on the railcar nameplate is integrated for tracking purposes. The experiments are conducted under both static and dynamic conditions to achieve these objectives.

Keywords: SCADA and Network Systems, Information Management Systems, Process Automation
Abstract: This paper tackles the growing issue of cyber-attacks on control systems, mainly through Advanced Persistent Threats (APTs), which are sophisticated, targeted attacks that can seriously damage systems. With control systems increasingly adopting the OPC UA for secure communication, introduces a new system designed to improve the availability of these control systems without compromising security.

Keywords: Nonlinear Control
Abstract: In this note, we investigate the stability of feedback systems with idempotent nonlinearities. For the analysis of nonlinear feedback systems, the powerful framework of integral quadratic constraint has been frequently employed, where the core of this framework is capturing the behavior of nonlinearities by multipliers. Among such multipliers, O'Shea-Zames-Falb (OZF) multipliers are known to be effective for slope-restricted nonlinearities. However, OZF multipliers only grasp the rough slope properties and hence cannot distinguish nonlinearities with the same slope property. To address this issue, we focus on the fact that some nonlinearities that are important in control engineering satisfy idempotence. By actively using the idempotence property, we first show that we can enlarge the set of the standard OZF multipliers for slope-restricted nonlinearities. In particular, for slope-restricted and idempotent nonlinearities with slope [0,1], we can provide a clear understanding on how the set of the standard OZF multipliers is enlarged. We finally illustrate the effectiveness of the newly proposed multipliers by numerical examples.

Keywords: Nonlinear Control, Discrete Event Systems
Abstract: This paper considers obstacle avoidance of moving objects with model predictive control (MPC) based on the control barrier function (CBF). Since the CBF-based constraints for obstacle avoidance are generally non-convex, one suffers from computational complexity and optimality when solving non-convex MPC problems, even if the constraints are quadratic. The paper proposes the Lagrangian relaxation for the MPC problems with the CBF-based obstacle avoidance constraints for moving objects, which are discrete-time linear systems. The paper discusses rewriting the MPC problem as a Lagrangian dual problem, a linear matrix inequality problem. The solution to the MPC problem consists of the solution to the dual problem. Finally, the paper illustrates numerical examples of obstacle avoidance for objects, objects, including an integrator, a unicycle, and a bicycle.

Keywords: Nonlinear Control, Discrete Event Systems, Adaptive and Optimal Control
Abstract: This paper discusses obstacle avoidance for mobile robots, continuous-time linear time-ivariant model, using Lagrangian relaxation of model predictive control (MPC) with a continuous-time cost function. Since obstacles for moving objects such as mobile robots are often convex, the robots must move in complementary non-convex sets. Commonly, MPC problems under non-convex constraints require a high computational complexity. This paper introduces the Lagrangian relaxation of the MPC problem. It obtains an optimal solution to the MPC problem from an optimal solution to the dual problem, a linear matrix inequality problem, with the Kruch-Kuhn-Tuker condition. However, the mobile robot might break the avoidance constraint intersample. Thus, this paper introduces a continuous-time cost function to improve the intersample performance. Numerical examples of mobile robots, such as a unicycle and a bicycle, illustrate the effectiveness of improving intersample performance.

Keywords: Nonlinear Control, Modeling, System Identification and Estimation
Abstract: This paper presents a new controller design methodology tailored for sampled-data systems structured in canonical forms. Our proposed approach builds upon existing techniques by employing higher-accuracy discretization methods, leading to notable improvements in control performance and system stability. Through comprehensive analysis, we elucidate the intricacies of the new discretization technique and validate its superiority via simulations against conventional methods. The results indicate that our method offers a more precise approximation of continuous-time system behaviors, positioning it as a promising solution for enhancing control in certain sampled-data systems.

Keywords: Nonlinear Control, Power Systems Control
Abstract: In this paper, we present an exploration of output-differential passivity. We establish formal definitions with a Krasovskii's storage function and a sufficient condition for output-differential passivity, emphasizing its preservation under negative feedback interconnections. Aiming to utilize the method of Krasovskii's passivity-based control to address the application of output-differential supply rate, we develop an output-differential passivity-based controller for extended systems with output-differential supply rate.

Keywords: Nonlinear Control, Robust Control, Mechanical Systems Control
Abstract: In this work, we propose a passivity-based super-twisting controller for mechanical port-Hamiltonian systems. The authors have proposed a passivity-based sliding mode controller based on kinetic-potential energy shaping (KPES) which allows us to embed a subsystem whose dimension is the same as that of the input into the closed-loop system. This paper extends KPES to incorporate a higher-order subsystem in the closed-loop system, which enables us to obtain the subsystem that has the same structure as the super-twisting algorithm (STA). Passivity-based control provides a variety of Lyapunov function candidates for interconnected physical systems, while the STA achieves second-order sliding mode control. Therefore, the proposed method is expected to offer a way to integrate the two methods, passivity-based control and second-order sliding mode control, and to allow us to design sliding mode control systems with a variety of energy-based Lyapunov functions. The effectiveness of the proposed method is verified through a numerical example.

Keywords: Nonlinear Control, Transportation Systems, Mechanical Systems Control
Abstract: Systems with non-contractible manifolds as configuration spaces cannot be globally asymptotically stabilized by continuous feedback. To avoid the problem, this paper discusses discontinuous feedback control and a discontinuous feedback control method using locally semiconcave practical control Lyapunov functions is proposed. A method for designing locally semiconcave practical control Lyapunov functions is the étale virtual input assignment, a method for designing systems driven by states on differential manifolds via systems lifted by étale bundles. One problem is that the étale virtual input arrangement method cannot cope with trajectory tracking. This study considers a trajectory following control based on the étale virtual input configuration, and actual experiments confirm its effectiveness.

Keywords: Robust Control
Abstract: H-infinity control is one of the effective design methods for robust control systems and is significant in that it allows frequency shaping in the control system, but the design results are conservative when dealing with parameter as multiplicative uncertainties. The scaled H-infinity control problem has been proposed as an approach to handle this problem, and an effective numerical solution method is the dual iteration method. While the dual iteration method efficiently improves the control performance, it also leads to numerical instability. Accordingly, a numerical method for the scaled H-infinity problem that has numerical stability and high efficiency is needed. This paper proposes a numerical method for the scaled H-infinity control problem based on the approximation of inverse matrices. This paper also shows the effectiveness of the proposed method through a numerical example.

Keywords: Robust Control
Abstract: This paper proposes an H-infitniy controller design method which realizes a two-degree-of-freedom (2-DOF) control system with lower sensitivity characteristics in the low-frequency range. A control system with a disturbance observer is a 2-DOF control system that can achieve low sensitivity characteristics in the specific low-frequency range. However, the design of a controller with a disturbance observer structure is a non-convex optimization problem if it is considered as an observer gain design problem. In addition, it has not yet been revealed whether a 2-DOF control system with the low sensitivity characteristics of the disturbance observer can be achieved by the H-infinity controller design. %Therefore, this paper proposes a design method for H-infitniy controllers that achieves type 2 sensitivity characteristics, maintaining the type 1 servo characteristics without specifying the disturbance observer structure. Therefore, this paper proposes a design method for H-infinity controllers that achieves the low-sensitivity characteristics of a type 2 servo system while maintaining the type 1 servo characteristics without specifying the disturbance observer structure. Numerical examples demonstrate the effectiveness of the proposed method.

Keywords: Robust Control
Abstract: This paper proposes a method for estimating scheduling parameters using a disturbance observer. The input and output values of the scheduling parameters are estimated by a disturbance observer, and the scheduling parameters are estimated by calculating the ratio of these values. The effectiveness of the gain-scheduled control using the estimated values is verified by comparing the gain-scheduled control with actual scheduling parameters and the conventional robust control designed by the scaled H-infinity control.

Keywords: Robust Control, Adaptive and Optimal Control
Abstract: For a class of uncertain large-scale interconnected systems, we propose a design method of decentralized adaptive gain robust controllers based on piecewise Lyapunov functions(PLFs). The decentralized adaptive gain robust controller proposed in this paper consists of fixed gains and compensation inputs with adaptive gains tuned by updating laws. Furthermore, this paper shows LMI-based sufficient conditions for the existence of the proposed decentralized adaptive gain robust controller.

Keywords: Robust Control, Adaptive and Optimal Control, Power Systems Control
Abstract: As a power system operation, a generation schedule indicating optimal on/off and power generation of generators is created so that the power needed by consumers can be generated and transmitted on transmission lines under some uncertainties. To create the economic schedule, we are required to solve a large-scale combinatorial optimization, which takes a long computation time to solve. Therefore, to reduce the computation time, the number of scenarios modeling the uncertainties is reduced based on dual variables which show the importance of the operational constraint violations in the linear relaxed optimization problem. To evaluate the performance of the proposed method, it is applied to the IEEE-118 bus power system model. The results show that the proposed method can reduce the computation time by up to 11% while maintaining the accuracy of the optimization solution.

Keywords: Robust Control, Mechanical Systems Control
Abstract: This paper denotes an approach to discrete–time cascaded LADRC design that theoretically guarantee the stability for the inverted pendulum on a cart. System stability is ensured by determining the LADRC parameters such that all poles are located within the unit circle on the complex domain, and the motor voltage constraint is considered by the objective evaluation index with respect to an integral of time absolute angle. The effectiveness of this approach can be confirmed by performing numerical simulations. It is found that the derivation of discrete–time cascaded LADRC parameters is similar to that for continuous–time ones.

Keywords: Robust Control, Nonlinear Control
Abstract: This work considers the problem of time-varying formation tracking control of second-order multi-agent systems under disturbances. A distributed robust time-varying formation control law is proposed including distributed finite-time estimators of the leader’s states and sliding mode time-varying formation controllers. Firstly, each agent can quickly estimate the leader’s states through the communication network in the distributed finite-time estimators based on the sliding mode estimation. Secondly, utilizing the estimates of the leader’s states, a sliding mode time-varying formation controller is designed using the prescribed time modification function. Unlike traditional distributed control law relying on the exchanges of the agents’ states, the proposed control design achieves great robustness and stability of the overall system by exchanging the estimates of the leader’s states. According to Lyapunov stability analysis, we prove that the proposed approach enables the sliding variables to achieve fast finite-time convergence. Furthermore, simulation examples are provided to illustrate the performance.

Keywords: Robust Control, Simulation of Large Systems
Abstract: In recent years, there has been much discussion about tackling energy utilization with the progress of global warming and the spread of electric vehicles. This study is about HEMS (Home Energy Management System). This system realizes optimal energy use by exchanging electricity between an EV and households based on understanding energy use in the home, which is one of the solutions to such discussions. In operating HEMS, it is very important to predict vehicle usage. Therefore, we proposed a scenario-based MPC (Model Predictive Control) that appropriately performs charging and discharging between the home and an EV that can cope with various vehicle usage scenarios. This shows that the proposed method is more useful than those failing to predict vehicle usage.

Keywords: Human Interfaces, Human-Machine Systems
Abstract: As technology develops and robots gradually become a part of everyday life, devices that provide physical assistance to humans will become increasingly common. Such robots must often physically interact with the user for extended periods of time and must provide auxiliary forces that complement the user’s strength to accomplish certain goals. This requires close cooperation and close coordination between the human user and the assistive robot. The purpose of this study is to understand how two humans cooperate and coordinate in order to gain insight into how to improve human-robot coordination. In this study, a tracking task involving a pair of humans was conducted. The task used “predictable” and “unpredictable” sine waves. In the experiment, the roles of the two humans were divided into a “user” focusing on the experimental task and an “assistant” focusing on support, and the amount of information given to the “assistant” was divided into three levels (“Full knowledge”, “Short-term knowledge”, and “No knowledge”) to try to understand the coordination and cooperation tendencies of the two humans. The experimental results showed that in some conditions, the rate of increase in error and load decreased as the amount of task knowledge given to the “assistant” increased, compared to when the task was performed “solo”.

Keywords: Human Interfaces, Virtual Reality Systems, Human-Machine Systems
Abstract: This paper addresses issues in realizing a tactile display system that reproduces tactile sensation on the whole body. Wiring each actuator with individual cable would require many wires and involves dilemma between durability and flexibility. Wireless transmission using radio waves can eliminate these wires, however, practical wireless power supply to drive actuators is not available. Conductive textile-based two-dimensional (2-D) communications enable DC power transfer and data transfer with low interference. The conductive textile, a highly redundant current conducting path, achieves higher durability of current path and better wearability. Previously, wearable tactile display based on multi- channel frequency modulation (FM) waveform transmission on a conductive textile has been proposed. However, practical implementation of simultaneous multi-channel FM wave generation has not been clarified. In this paper, we propose a practical implementation based on software-defined radio (SDR). Using a single-board computer and a palm-sized SDR, the transmitter can be implemented within a small form factor suitable for practical wearable system.

Keywords: Human-Machine Systems, Human Interfaces
Abstract: Human motion may or may not show preliminary signs based on factors like mass, speed, and body balance. In this paper, we measured an arm swinging motion while changing its mass by attaching weights. Analysis of 3D skeletal coordinates revealed that although changing the arm’s mass did not significantly alter preliminary motions, it did change the movement in other body parts during the arm swinging motion. Additionally, experiments estimating the mass based on movements during the action demonstrated an accuracy of 80-90% in classifying whether the arm was weighted or not. This research offers insights into human motion prediction and the generation of natural movements.

Keywords: Human-Machine Systems, Human Interfaces, Identification and Estimation
Abstract: This paper presents a Bayesian network for estimating the burden factors of order pickers in a robotic mobile fulfillment system (RMFS). Recently, many logistics companies have introduced RMFS to improve the order-picking efficiency. However, they have ignored the fact that RMFS is a human–robot coordinated system, and thus have not been able to design human-friendly operations. Therefore, we replicate the order-picking operation in RMFS and measure the picker workload. The workload is assessed subjectively (via the NASA task load index(NASA-TLX)) and objectively (via heart rate variability (HRV) analysis). In addition, a Bayesian network is used to identify the elements of order picking that cause stress and HRV indices that could estimate subjective workload. The results revealed the conditions under which the picker workload increased. The Mean NNI (the mean of RR intervals on electrocardiogram) was also observed to be a valid HRV index for estimating physical demands, a subscale of the NASA-TLX.

Keywords: Human-Machine Systems, Mechanical Systems Control, Human Interfaces
Abstract: Existing prosthetic knees used by transfemoral amputees for running lack the functionality to prevent knee buckling, which poses a significant barrier to participating in sports and exercise owing to the risk of falls. In our previous study, a passive prosthetic knee that restricts flexion in response to the load on the prosthetic leg during the stance phase was developed. However, the mechanism depended on specific conditions of the load and was unable to fully prevent knee buckling. In the present study, we focused on the duration of the swing phase, which is generally constant regardless of running speed. The function of the proposed prosthetic knee allows flexion only during the first half of the swing phase and restricts flexion for the remaining duration. The mechanism involved the spur gear attached to the knee rotation axis, which restricts knee flexion by engaging with the rack. The rack was controlled by a spring and damper to separate from the spur gear for a certain period. To evaluate the mechanism, an able-bodied participant fitted with the simulated thigh socket and the prototype prosthetic knee ran at various speeds. The results demonstrated that the proposed mechanism restricted knee flexion before the stance phase at various speeds and prevented knee buckling. The proposed prosthetic knee is expected to promote sports participation among transfemoral amputees and lead to improvements in health.

Keywords: Virtual Reality Systems, Human Interfaces, Human-Machine Systems
Abstract: In this study, it is proposed the model of human behavior in the crowd under a simple task and has been working on building the VR experimental system based on a crowd behavior simulation system. However, there is the problem in that if the examinee contacts with the virtual human, the examinee is not able to feel the contact. Therefore, in the study, it is aimed to add the function of tactile when examinee contact with virtual human, to make a strongness of the sense of reality in VR experiments. It is proposed and designed the hardware and software configuration of the tactile system and make the prototyping and realize the basic functions.

Keywords: Virtual Reality Systems, Human Interfaces
Abstract: Tactile representation through vibrotactile sensation is widely used. For example, it is used to display click sensations in smartphones and PCs, and for tactile representations via controllers in video games and VR devices. To improve the quality of such vibrotactile sensations, it is desirable to be able to design the desired vibration more easily. However, compared to intuitively adjustable parameters such as hue, saturation, and value in vision, the parameters to be adjusted in vibrotactile sensation are unclear. The number of necessary and sufficient parameters is also unknown and is one of the most important issues to be clarified. In this study, the number of necessary and sufficient parameters was investigated by perceptual experiments using autoencoder compression. A comparison of the vibration reconstructed from the compressed information with the original vibration before compression showed that several hundred dimensions are sufficient to achieve the necessary compression.

Keywords: Human Interfaces, Virtual Reality Systems, Sensors and Transducers
Abstract: This study shows that in mid-air ultrasound haptics, when a linear pressure sensation stimulus is presented to the fingerpad, the movement of shift of the stimulus can be perceived with a sensitivity of 1.3 mm. It was also confirmed that conventionally used vibrotactile stimuli require a shift of 3.5 mm or more to perceive this shift.

Keywords: Computational Intelligence, Innovative Systems Approach for Realizing Smarter World, Advanced Computics
Abstract: In this study, we make a building up the model and simulation system for foraging and role sharing, according that we use the total behavior model covering all the activities of ants and simulation the swarming intelligence through the ant’s survival behavior. We can get the possibility of its engineering applications from our obtained result. Furthermore, we have a construction of the framework how to make a systematic treatment about the behavior model. In this study, we try to be applicability of the total behavior model for ant colony algorithm and develop 3D simulation system in which it is effective tools for other industrial programs. We show the obtained results that the 3D models and its actions such as ant individual, egg, its nest and their mutual interaction are going well to make a representation of the total behavior model for ant colony system.

Keywords: Computational Intelligence, Innovative Systems Approach for Realizing Smarter World, Signal and/or Image Processing
Abstract: Wind power generation is being rapidly adopted around the world, although generation power is unstable because wind speed changes irregularly time to time. Therefore, in order to achieve a stable supply of electricity, electric utilities must predict the amount of wind power generation based on wind speed predictions. In this paper, we propose a wind speed prediction model for the next day using deep neural network (DNN) based on frequency analysis of observed wind speed data. The input and output of the prediction model are amplitude and phase of wind speed data on frequency domain. The time domain and frequency domain of the wind speed data are transformed into each other using Fourier transform and Inverse Fourier transform. The prediction start time is 9:00 a.m., and the proposed model predicts wind speed for the next day ahead, i.e., from 16 to 39 hours ahead from the prediction start time. The proposed model is compared with DNN with time-series inputs (DNN time) and meso scale model grid point value (MSM-GPV) using four evaluation methods. The results of the predicted wind speed suggest that the proposed prediction model using DNN can learn features on the frequency domain of wind speed and can predict wind speed variations.

Keywords: Computational Intelligence, Intelligent Systems
Abstract: This paper proposes a prompt-based method utilizing a large language model (LLM) to predict changes in foreign exchange rates based on limit order information. While traditional deep learning models for prediction utilize numerical values as input and output, LLMs use sentences and prompts. To address this, we design prompts that incorporate the numerical values. GPT-2, a widely adopted LLM, is employed and fine-tuned using a training dataset. The effectiveness of our proposed method is demonstrated through empirical analysis using actual time series data.

Keywords: Computational Intelligence, Intelligent Systems
Abstract: Personalized news recommendations can alleviate the information overload problem. News recommendation models harness the interactive information between users and news to personalize recommendations. Most existing approaches to news recommendation use advanced natural language processing techniques to extract semantic information from historical news. However, real-world datasets frequently suffer from considerable noise, which limits the effective modeling of user preferences and item representation. We propose an innovative data augmentation technique that can identify potential data omissions, thereby enhancing the overall integrity of the dataset. In addition, the proposed method can be applied to various neural network–based recommendation frameworks. Experimental results obtained on the MIND dataset reveal that integrating the proposed method with a standard recommendation model yields a notable performance enhancement, as evidenced by a 6.79% increase in the hit ratio and a 7.67% improvement in the normalized discounted cumulative gain. These results highlight a substantial improvement in the accuracy and relevance of the news recommendations generated by the model.

Keywords: Computational Intelligence, Intelligent Systems, Computer Aided Design
Abstract: With the advancement of AI technology, generative AI has made remarkable progress in its ability to process multiple languages and adapt to creative tasks. It has also demonstrated its strength in academic fields, such as passing the national medical examinations. In this study, we tested the extent to which ChatGPT (GPT-4) can accurately answer classical control theory questions offered in undergraduate courses. The experimental results showed that GPT-4 showed a correct response rate of under 70% for quiz exercises in classical control theory, and that the correct response rate was lower for problems whose solutions were specific or required step-by-step thinking. In addition, since GPT-4 is a Transformer-based model, and the answers are based on mere prediction, it may give incorrect answers for problems that require complex calculations. In this study, we proposed a method to improve the response accuracy by developing a customized GPT specialized for classical control theory and using prompt engineering. The proposed method was applied to a university undergraduate final exam in undergraduate course, and the results showed that the correct response rate was improved and a passing score (60% or higher) was obtained.

Keywords: Computational Intelligence, Intelligent Systems, Computer Aided Design
Abstract: In real-world applications, constrained multi-modal multi-objective optimization problems (CMMOPs) are prevalent across various science and engineering domains. Multi-objective evolutionary algorithms (MOEAs) are well-suited for solving such challenging problems, owing to their versatility and efficient search capabilities. This paper proposes the switching constraint handling multi-modal multi-objective evolutionary algorithm (SCHMMEA), a MOEA for CMMOPs that switches the constraint-handling strategies during the search process. Specifically, SCHMMEA initially explores high-quality solutions without considering constraints, then the late stage considers the constraints. We compare the proposed method with CMMOCEA, a conventional method, on 14 CMMOP benchmarks. In addition, we examine different parameters switching the constraint handling strategies in the proposed method. The experimental results show that SCHMMEA with the constraint handling at 2/3 of the maximum generations outperformed the conventional method by 13, 12, and 11 problems in IGDX, IGD, and HV, respectively.

Keywords: Computational Intelligence, Signal and/or Image Processing, Intelligent Control
Abstract: Technologies of received signal strength indicator (RSSI) are becoming widely used in recent indoor and outdoor positioning scenarios, particularly in healthcare facilities, with their critical need for precise location data to track their patients. Trilateration (TRI) is one of the conventional localization methods in RSSI-based Positioning systems. However, most of the previous estimation methods of trilateration have some defects in certain aspects. The least-squares (LSQ) method has relatively high accuracy but it takes up lots of time and resources to complete the calculation. The weighted average (WAV) method can process the estimation efficiently but with very low accuracy and the searching area is restricted. The method proposed by this paper introduces the concept of radical axis and radical center into the trilateration process. Tremendous number of simulations results indicate, an optimized balance between efficiency and accuracy can be achieved, and an unrestricted search area can be ensured.

Keywords: Computational Intelligence, Signal and/or Image Processing, Intelligent Systems
Abstract: Robustness to changes in illumination conditions is one of the requirements for real-world vision-based object recognition. In particular, changes in illumination color can alter the apparent color of the target object and lead to misidentification. In this study, we applied color constancy (CC) algorithms to images used as input to convolutional neural networks (CNNs), and investigated the contribution of the algorithms to the classification accuracy of the CNNs under various illumination color. The CC algorithms used were the gray world algorithm and the center/surround (C/S) retinex model. We created a dataset which consists of images of 12 different objects illuminated by five different lighting colors. With this dataset, the classification accuracy of well-known CNN architectures (VGG16, ResNet18, and ResNet50) that use the CC algorithms were compared to those of the same CNNs that use data augmentation techniques for color modulation. The results showed that the C/S retinex model contributes the most to improving the classification accuracy of the CNNs that works under various illumination color.

Keywords: Adaptive and Optimal Control, Autonomous Decentralized Systems, Factory Automation
Abstract: In this paper, we address optimal task assignment for equipment inspection robots in a factory. In view of the complex work environment and variety of the equipments to be inspected, coordination of heterogeneous robots, including drones and ground vehicles, is inevitable for such a system. In our previous study, we proposed a computationally efficient task assignment algorithm for the problem by decomposing the problem through appropriate layering. However, this algorithm did not take into account the work efficiency of the individual robots. In this paper, we present an assignment algorithm that reflects heterogeneity of the work efficiency. To this end, we present novel logical constraints for the upper-layer problem, which is responsible for the area assignment, that guarantee feasibility of the problem in the lower layer. Finally, we demonstrate the effectiveness of the present solution through simulation that duplicate operations in an actual factory.

Keywords: Adaptive and Optimal Control, Mechanical Systems Control, Computer Aided Design
Abstract: Recently, event-triggered control methods have attracted much attention due to their communication efficiency. The methods usually require a plant model although its derivation needs a lot of effort such as time and cost. As methods to reduce such effort, data-driven control methods are known to be effective. By combining these methods, we have proposed a data-driven event-triggered control method for reducing communication amount and design effort in a previous research. However, there are still some issues that we have not considered noise in a plant and an output feedback case yet. Therefore, in this paper, we propose a method that has comparable performance to the previous research's one even in a case that noise is affected in a plant and only the output is measurable. We verify the effectiveness of the proposed method through experiments with a DC motor.

Keywords: Adaptive and Optimal Control, Mechanical Systems Control, Mechatronics Systems
Abstract: This paper proposes a combined controller of simple adaptive and integral terminal sliding mode control (ITSMC) for precise motion of mechanical systems. This design aims to alleviate the problem of requiring comprehensive understanding of plant dynamics and parameters for ITSMC application and benefits from the inclusion of the fractional power and switching control terms. Simulation results demonstrate the adaptive ability to outperforms the conventional simple adaptive technique.

Keywords: Adaptive and Optimal Control, Modeling, System Identification and Estimation, Mechanical Systems Control
Abstract: A method is proposed to enhance the performance of the Linear Active Disturbance Rejection Controller (LADRC) by utilizing a Parameter Adaptive Law (AL). Additionally, a Particle Swarm Optimization algorithm based on dual-position chaotic mapping is introduced to address the controller tuning issue, with the goal of minimizing the Integral of Time-weighted Absolute Error (ITAE) to achieve the desired control objective. The effectiveness of the proposed parameter optimization method and the AL in improving the performance of LADRC is demonstrated through trajectory control simulation experiments conducted on a single-degree-of-freedom (DOF) robotic arm.

Keywords: Adaptive and Optimal Control, Multivariable Control, Computer Aided Design
Abstract: An infinite-horizon optimal control problem is considered with an objective function consisting of a quadratic form of an input and a state and the 1-norm of an input. This problem is shown to have an optimal solution if it has a feasible solution that makes the objective function value finite. Moreover, the optimal input is shown sparse in the sense that it is equal to zero after some time step. These properties are expected to be useful for explicit generation of the optimal input.

Keywords: Adaptive and Optimal Control, Robotic and Automation Systems
Abstract: This paper presents a coverage control algorithm specialized to image sampling for 3D model reconstruction of environment using a drone network. It is widely known that the sampling the environment from rich viewing angles enhances the 3D model quality. To this end, the authors presented so-called angle-aware coverage control based on the concept of constraint-based control, which enables drones to determine its translational motion and camera rotation so that a performance constraint is enforced. In this paper, we combine this control methodology with the battery charging constraint in order to ensure mission persistency. Now, dynamically assigning charging stations to drones is expected to enhance the coverage performance. We thus design a distributed assignment algorithm of the stations based on so-called alternating direction method of multipliers (ADMM). The algorithm is then integrated with the above constraint-based controller, and the integrated system is demonstrated through simulation in Robot Operating System 2.

Keywords: Computer Aided Design, Adaptive and Optimal Control, Intelligent Control
Abstract: Recent dynamical systems, such as mobile robots and drones, are controlled via wireless networks, where communication resources are limited. Sparse control has been recognized to be effective in such systems because it maximizes the duration over which the control signal is precisely zero. In this study, we introduce a novel sparse control design method utilizing the minimax concave penalty function. Through a numerical example, we compare our proposed method with the conventional L1-based method and demonstrate the superiority of the proposed method.

Keywords: Computer Aided Design, Signal and/or Image Processing
Abstract: This paper proposes two methods for computing fractional-order calculus in discrete time. While previous research has presented a calculation method for discrete-time differences using matrices, the applicable order of differentiation is constrained by the form of an integer reciprocal. In contrast, our study addresses this limitation by deriving calculation methods capable of handling all fractional orders of differentiation in discrete-time. We propose two distinct approaches: one leveraging matrix exponential and logarithmic functions, and the other employing binomial series expansion. Furthermore, we conduct simulations to investigate the correlation between sampling frequency and the accuracy of our proposed calculations comparing with the continuous-time fractional-order differentiation.

Keywords: Multivariable Control
Abstract: This paper proposes discrete-state predictive control for multi-input-time-delay system. Discrete-state-predictive control is already established for single-input-multi-output systems. Unlike the single-input-time-delay systems, the future state of the systems cannot be determined because of the gap between each time delay. This paper establishes the discrete-state-predictive control for Multi-input systems and validates it in simulation.

Keywords: Multivariable Control
Abstract: This paper clarifies the applicability of the forwarding approach to a stabilization problem of a class of linear cascade systems involving distributed state delay. The system consists of an upper subsystem and a lower subsystem. The lower subsystem is cascaded to the upper subsystem through delayed state. The control input is applied to both the subsystems. A stabilizing control law is successfully designed based on the forwarding approach.

Keywords: Multivariable Control, Adaptive and Optimal Control
Abstract: An optimal control problem (OCP) in linear model predictive control (MPC) is split into small parallelizable subproblems. A formula for splitting with maximum parallelization degree is given. An optimal solution of the original OCP is iteratively computed by solving the parallelizable subproblems. The performance of the proposed method is numerically investigated.

Keywords: Multivariable Control, Identification and Estimation, Process Control
Abstract: Estimated response iterative tuning(ERIT), a method of data-driven control, predicts the control output when the feedforward controller of a two-degree-of-freedom (2-DOF)control system is changed. The main specification of ERIT is based on predicting a system output/input signals from a set of experimental data. In this study, we consider a 2-DOF integral servo system and predict not only output but also input and state signal prediction for feedforward controller changes. We proposed a feedforward controller design method. Finally, we verify the usefulness of the system with numerical examples.

Keywords: Modeling, System Identification and Estimation
Abstract: This paper studies a backward representation for a linear-time-varying (LTV) stochastic system. A backward representation for an LTV stochastic system is derived from the computation of covariance matrices by extending stochastic realization theory for linear-time-invariant systems.

Keywords: Modeling, System Identification and Estimation, Identification and Estimation
Abstract: During the operation of brushless dc motors, angular velocity is estimated based on the Hall sensor signals. To improve estimation performance, using a Kalman filter designed to mitigate the negative influence of Hall sensors' misalignment is effective. However, accurate estimation remains difficult during low-speed rotations due to the low resolution of Hall sensor signals. This paper proposes a method for solving this problem. In the proposed method, the estimated angular velocity as follows: the predicted state, calculated using the Kalman filter, is used when Hall sensor signals are unavailable during low-speed rotation. On the other hand, when the sensor signals are available, the angular velocity is estimated using the Kalman filter based on these observed signals. The effectiveness of this method was experimentally verified.

Keywords: Modeling, System Identification and Estimation, Identification and Estimation, Robotic and Automation Systems
Abstract: This paper discusses safety concerns and technological advancements in human-robot collaboration. It emphasizes the importance of safety standards like ISO/TS 15066 and explores collision detection techniques. It highlights challenges in obtaining accurate dynamic model parameters and proposes methods for their estimation. Excitation trajectory design and matrix condition number play a crucial role in accurate parameter identification. A parameter identification approach is presented by integrating an iterative Newton-Euler method, a least squares method, a procedure for eliminating linear dependencies, and an exciting trajectory design with a condition number for a robotic arm. The effectiveness of the proposed approach is exemplified by carrying out a numerical simulation with a 6-Dof industrial robotic arm, HIWIN RA605-710 robotic manipulator. MATLAB's symbolic math toolbox facilitates symbolic computations, incorporating dynamic parameters such as Coulomb friction, viscous friction, inertial tensor, and link masses.

Keywords: Modeling, System Identification and Estimation, Identification and Estimation, Signal and/or Image Processing
Abstract: We propose a new singular value decomposition-based recursive PI-MOESP identification algorithm with fixed input-output data size (PI-MOESP identification algorithm: the multiple-input multiple-output output-error state-space model (MOESP) identification algorithm using an instrumental variable consisting of past input (PI)). The proposed recursive PI-MOESP identification algorithm (RPI-MOESP) has the advantage that its accuracy is hardly affected by the choice of a data length. In addition, since the proposed RPI-MOESP updates an inverse matrix via two matrix inversion lemmas that compute addition and subtraction separately, it can reduce computational complexity. The effectiveness of the computation time and the accuracy of the proposed RPI-MOESP is statistically examined through numerical experiments.

Keywords: Modeling, System Identification and Estimation, Nonlinear Control
Abstract: System identification is widely applied in various frameworks to derive a system model from measured data. Whilst numerous system identification approaches were investigated, less studies concerning the data basis generation used in the identification were performed. Harnessing only little prior system knowledge, a novel approach is proposed to generate the data basis to be used in an arbitrary system identification in an advantageous way, utilizing random input generation and automated measurements. As almost no prior knowledge is necessary to apply the presented approach it is applicable for an arbitrary single input single output system. The efficacy of the proposed approach is illustrated by a simulation example and a comparison to the use of standard amplitude modulated pseudo-random binary sequences.

Keywords: Flow Measurement and Control, Signal and/or Image Processing, Modeling, System Identification and Estimation
Abstract: In the paper, we suggest a set of verification and validation methods for scaled sloshing and visual feedback. First, we check the scaled experimental setup via the planar approximation and Archimedes’ principle. Second, we compare the computation time by the proposed method (normalized Legendre basis) and that by the conventional method (Taylor basis) for obtaining the projected free surface in real time. At the same time, we also compare the spatial error between the measured free surface and the projected free surface. The effectiveness of the proposed method (normalized Legendre basis) is confirmed experimentally.

Keywords: Signal and/or Image Processing, Intelligent Systems, Social Systems
Abstract: Traffic accidents are a serious problem, and they often occur at intersections. The police use intersec-tion cameras to manually inspect an accident scene, which is, however, very costly. In this paper, we propose an effective method for computing vanishing points. In this method, white lines are detected using an image pro-cessing method, and they are interpolated to obtain straight lines. After processing, we can easily compute the vanishing points from the image.

Keywords: Signal and/or Image Processing, Remote Sensing, Identification and Estimation
Abstract: Currently, the management of power lines is essential, and there are needs to reduce costs and improve inspection quality by automating verification process. The usage of 3D-point cloud data (3D-PCD) is promising to them, and we had proposed automatic tracking methods for power lines. In this study, a novel method for extracting inherent information of power lines in order to automate power line tracking, and its validity in a real environment is confirmed.

Keywords: Signal and/or Image Processing, Robotic and Automation Systems, Entertainment Systems
Abstract: Table tennis spin affects trajectory and is useful information for improving play and controlling table tennis robots. The ball spin that can reach 150 rps is very fast, and accurate and high-speed spin measurement is difficult. In this paper, we propose a high-speed measurement method for table tennis ball spin using symmetrical dot markers and structured light of M-sequence flickering pattern. We have experimentally showed sufficient measurement accuracy of spin amount using table tennis ball spinner with a motor of 90 rps, and also low computational cost less than 3 ms.

Keywords: Signal and/or Image Processing, Robotic and Automation Systems, Intelligent Systems
Abstract: A set of Gabor filters with selectivity for various spatial frequencies, phases, and orientations is one of the most important spatial features, not only for implementing neuronal network models of the visual cortex, but also for developing purely engineering applications such as image recognition. In this study, we designed an efficient neuro-inspired multi-scale Gabor filtering circuit that provides Gabor filtered images with four orientations and two orthogonal phases for each scale. To reduce the computational cost, we employed the following techniques: neuro-inspired hierarchical filtering that starts with a simple low-pass filter and progressively achieves Gabor-like characteristics, separating one two-dimensional filter into two one-dimensional filters by using the separability of Gaussian and Gabor functions, and multi-scale filter kernels that differ in width but not in the weight values. We implemented the filter circuit into a field-programmable gate array (FPGA), and developed an image sensor system that consists of the FPGA and a CMOS image sensor. The system outputs a set of Gabor filtered images with 160 x 120 pixels at approximately 110 frames per second. The set of filtered images was also used to visualize responses of a model of the visual nervous system, and the response of a neuron model that respond selectively to particular spatial frequencies and orientations independent of phase was adequately reproduced.

Keywords: Signal and/or Image Processing, Sensors and Transducers, Mechatronics Systems
Abstract: Gas sampling mechanisms often draw inspiration from nature, but gas-source-seeking robots aren't confined to two sensors like nostrils in mammals or antennae in insects. More sensors can be placed at various heights to detect gas sources in three dimensions. This study introduces a High-Degree-of-Freedom gas sensor device trained on data collected in obstructed environments. The mapping process, after location inference, demonstrates high model accuracy. Robots can utilize this model to extract valuable environmental information using only a set of gas sensors.

Keywords: Signal and/or Image Processing, Sensors and Transducers, System Engineering
Abstract: Society 5.0 is characterized by the sophisticated integration of cyberspace with physical space, that is, cyber-physical systems (CPS), which are integrations of computation and physical processes. Embedded computers and networks monitor and control the physical processes, usually with feedback loops. We have been applying category theory to describe an SoS architecture that is crucial to develop and operating an object SoS and to achieve sustainability. Presenting and previous papers have been developed a universal formulation of measurement and sensing technologies based on the framework of category theory.

Keywords: Mechanical Systems Control, Robotic and Automation Systems, Mechatronics Systems
Abstract: In this research, we study on posture control of a planar one-link torque unit manipulator which has a revolute joint with viscous friction. The torque unit is an easy maintenance module that consists of a motor and a reaction wheel, and it can be set to an arbitrary position of the link as an actuator to control the angle of the rotating link with free joint. However, the system of a planar one-link torque unit manipulator has an uncontrollable mode. There arises a problem such that the angular momentum of the reaction wheel remains non-zero after controlling to a target angle of the link. Furthermore, it is essentially inevitable that the link of the manipulator goes back to the initial angle after stopping of the control. In this paper, we investigate the eigenvalues of the control system in order to understand these phenomena. Then, we consider using a reaction wheel with air resistance and show that the link of the manipulator can be controlled to an arbitrary target angle even though the system has an uncontrollable mode. We conclude that the air resistance of the reaction wheel functions as a kind of implicit control.

Keywords: Mechanical Systems Control, Virtual Reality Systems, Nonlinear Control
Abstract: Given that teaching emphasizes the process of building knowledge frameworks, virtual reality-based education resources and materials can effectively preserve details of teaching processes, consequently contributing to sustainable education. Digitized courses in virtual reality, particularly those involving engineering practices and experiments, offer the advantages of being accessible and replayable after class sessions. This position paper describes a design of a virtual laboratory for control engineering education and collaboration with industries. MATLAB® Live Editor is used for creating interactive scripts for different control engineering topics. In addition, Simulink 3D animation tool is applied for visualization of dynamic systems in a photorealistic 3D environment.

Keywords: Safety, Environment and Eco-Systems, Mechatronics Systems, Robotic and Automation Systems
Abstract: The passive dynamic control (PDC) is a control method for a mechanical system, and it executes the moving operation which is the purpose control, after executing the balancing operation which takes balance with external force. It ensures high safety by achieving stopping, position hold and speed control using a brake mechanism. However, it has been difficult to apply the PDC to a case where the balancing state is broken due to movement. This paper discusses the application of the PDC to a pneumatic artificial muscle manipulator set in vertically. By using a balancing operation device and a moving operation device separately, the balancing operation can treat large force with the controlled object fixed by the brake, and the moving operation can be executed by small force. On this principle, control with high safety can be realized. The effectiveness of the proposed motion control method is verified through experiments in positioning control.

Keywords: Mechanical Systems Control
Abstract: A design method for a driver-assist control system that aims to maintain occupant comfort is presented. The comfort index is the magnitude of the vehicle acceleration passing through a frequency filter. The assist torque is provided only when the index exceeds a threshold value. The method is formulated as a model predictive control (MPC) problem with quadratic constraints. Numerical simulations demonstrate the effectiveness of the proposed method.

Keywords: Modeling, System Identification and Estimation
Abstract: Cornering stiffness and sideslip angle are the crucial parameters for designing vehicle dynamical control systems. This paper discusses the problem of estimating these quantities based on data obtained from in-vehicle IMU sensors. A vehicle is described as a linear parameter varying (LPV) system with cornering stiffness as a variable parameter. A moving horizon estimation (MHE) method is applied to the LPV system, and the result of the application to a small electric vehicle is presented.

Keywords: Nonlinear Control, Robust Control, Modeling, System Identification and Estimation
Abstract: This paper presents some results to develop dissipative systems theory in terms of asymmetrically scaled sectorial supply rates. The supply rates, which describe strict passivity on asymmetric spaces, were proposed recently by the author to simultaneously verify stability and positivity of dynamical systems whose equilibria are interior points of positive state spaces. This paper demonstrates that systems defined by an asymmetrically scaled sectorial supply rate have finite asymmetric gain if their positivity is guaranteed by the supply rate. For cyclic networks whose component systems are defined by asymmetrically scaled sectorial supply rates, this paper shows that the networks are globally asymptotically stable on the positive space whenever positivity is guaranteed by the individual supply rates. The developments deal with general nonlinear systems. It is illustrated that positive systems design benefits a lot from asymmetrically scaled sectorial supply rates and their implications even for affine compartmental systems for which symmetric methods are ineffective.

Keywords: Nonlinear Control, Modeling, System Identification and Estimation, Mechatronics Systems
Abstract: While drone technology is developing rapidly today, the landing of drones is of critical importance in sensitive missions. Soft landing allows drones to land safely and precisely on the ground. Soft landing of drones is an essential component of safety and mission success. To this end, the use of appropriate design and control strategies is a fundamental requirement for a wider range of applications of drone technology. This article examines the control strategies that can be used to ensure a fast and soft landing of the designed drone solution. More specifically, we designed a controller based on MATLAB/Simulink and applied it to the Raspberry pi. The Raspberry Pi is connected to a Time of Flight (ToF) sensor, which is used for calculating altitude. Landing tests were carried out with the designed control system, and flight test data shows that the quadcopter landed smoothly with this control technology in a minimum of time. Soft landing is important both for the drone itself and the payloads it carries in terms of minimizing damage, performing sensitive tasks and ensuring minimal impact on the environment. This article has been prepared in order to emphasize the importance of drones in making soft landings and to provide a basis for future developments.

Keywords: Nonlinear Control, Transportation Systems, Mechanical Systems Control
Abstract: This study shows the effectiveness of global path following control using a locally semi-concave practical control Lyapunov function (LS-PCLF). Within this framework, an LS-PCLF can be designed by allocating virtual inputs to one of equilibrium points in a virtual space known as étale space. However, there can be more than one equilibrium point in the étale space corresponding to asymptotic stabilization. Therefore, based on the minimum projection method, we arbitrarily choose one of the corresponding equilibrium points and give a locally semi-concave practical control Lyapunov function. The effectiveness of the proposed method is confirmed by experimental results using an actual electric wheelchair.

Keywords: Nonlinear Control, Human-Machine Systems, Transportation Systems
Abstract: There is no end to vehicle accidents caused by human errors. To enhance the control system's safety, Nakamura et al. proposed a reciprocal control barrier function (RCBF)-based on human-assist control. A strict zeroing CBF (S-ZCBF) was proposed by Tezuka and Nakamura, which may deal with a control system crossing the boundary. Moreover, we need to consider the control system with input constraints. In this study, we propose a relaxed ZCBF-based human-assist control using the notion of viability kernels, which satisfies the input constraint and the state constraint simultaneously. Moreover, we validate the proposed law and design separate functions outside of the viability kernel through computer simulation.

Keywords: Safety, Environment and Eco-Systems, System Engineering, Modeling, System Identification and Estimation
Abstract: This paper proposes STPA for Complex System (STPA_CS), an extension of System-Theoretic Process Analysis (STPA) so that it can be applied to the causal analysis of hazards caused by systems with complex functions and configurations.　STPA_CS has the following features: (1) Class diagrams in the Unified Modeling Language (UML) can be used to describe the elements that are the parts of the system, enabling appropriate description of class behavior, the control signals, and data sending/receiving from the classes. (2) The configuration of the system is described accurately and hierarchically by using class diagrams. (3) By regarding the exchange of control signals (methods) between classes as a directed graph, the process from the occurrence of the control signal that causes the hazard to the occurrence of the causes of the hazard can be traced. (4) By analyzing the causes of hazards while developing class diagrams, the causes of hazards can be analyzed seamlessly throughout the system development process. As a result, the proposed method enables consistent analysis of the causes of hazards throughout the development process, and the safety of the system is improved.

Keywords: Social Systems, Safety, Environment and Eco-Systems, Intelligent Systems
Abstract: Agent-based simulation (ABS) has emerged as a method to analyze human-driven technology selections. Currently, studies on energy and environmental systems represent the major fields adopting ABS. A well-known limitation of ABS is that machine agents comprising ABS find it difficult to deal with the psychological aspects of decision-makings by humans, such as norm, trust, and risk perception. Therefore, this study aims to develop a method called clustering and learning (CAL) to design machine agents that learn decision-making strategies of humans from their gameplay data. The gameplay data were obtained by a large-scale online experiment where human participants play the Lumberjack game. The designed machine agents played the game with two bots as the original experiment. The results indicate that the machine agents had a probability of 76% in performing the same actions as those performed by human participants. The time-series behavior of humans was reproduced well by the machine agents. Further, the CAL method is revealed to be useful in extracting typical strategies of humans' gameplay thorough its classifying processes.

Keywords: Transportation Systems, Autonomous Decentralized Systems, Network System Integration
Abstract: In recent years, natural disasters such as earthquakes and typhoons have occurred frequently in Japan. The 2024 Noto Peninsula Earthquake occurred on January 1st, 2024, and caused extensive damage. In the case of a large-scale disaster, prompt and efficient rescue activities are required to protect the lives of the victims. In response to such natural disasters, the local governments in Japan formulate disaster prevention plans while considering the characteristics of each region to prepare for the occurrence of disasters. When the number of transportation vehicle increases, waiting queues form at hospitals and it limits the transportation performance. This paper focuses on a bottleneck analysis to estimate the formation of waiting queue, as well as introduces a transport simulator and analytical solution to estimate transportation performance considered in the disaster prevention plan. Using these tools, the analyses are carried out for a city, assuming a Tokyo Inland Earthquake. Transportation capability for the injured people is quantitatively clarified under multiple scenarios based on the regional disaster prevention plan and disaster medical relief plan. Users of these analytical tools can consider more effective measures of their rescue operation.

Keywords: Transportation Systems, Human-Machine Systems, Autonomous Decentralized Systems
Abstract: 近年、自動運転の開発 削減を目指して技術が追求されてきました 交通事故と渋滞。ただし、必要です 安全性への影響を徹底的に考慮し、 装備車両時の交通の流れの円滑さ 自動運転技術と従来の技術が混在 車。本稿では、環境の構築を目指します 混合自動運転車と実験の実施 ドライビングシミュレーターを使用して、運転スキルを習得します。 混合環境でのドライバーと安全性の評価と 円満。

Keywords: Transportation Systems, Intelligent Systems
Abstract: In recent years, research on autonomous ships has gained significant traction, resulting in the development of various automatic avoidance systems. However, these systems typically offer only one avoidance route, which may not match the navigator's preferences. This discrepancy arises from differences between the evaluation function utilized by the automated system and the factors considered by the navigator when selecting a route. Additionally, the navigator's considerations vary depending on the ship's maneuvering phase and prevailing conditions. To overcome these limitations, this study proposes an automatic avoidance system that generates multiple avoidance routes tailored to the navigator's preferences. By integrating multiple evaluation functions, the system presents a range of options, allowing the navigator to choose the most appropriate route based on their assessment.

Keywords: Social Systems
Abstract: This study explores the potential and effect of utilizing proactive mechanisms newly introduced in the Gift & Circulation Model of relationality assets. Specifically, we investigate the model's ability to create opportunities for interaction among strangers and stimulate casual communication in a community. We present the results of a multi-agent simulation study that tests the effectiveness of a new mechanism leveraging the sunk cost effect and discuss the adjustment of the redistribution rate to prevent public account divergence.

Keywords: Social Systems, System Engineering, Relationality-Oriented Systems Design
Abstract: This research introduces an innovative incentive system to enhance intrinsic motivation and active participation in community activities among older adults. Recognizing challenges posed by internal and external factors, the study employs an individuality-oriented motivational design framework. This framework incorporates personalized rewards, recognition, social support, and individual goal-setting to address the psychological nuances of older individuals. Given the social losses and disruptions to support networks experienced by most older adults, the study emphasizes the importance of social connections for mental well-being. The framework, emphasizing outdoor social spaces, stands out as a pioneering approach tailored to the unique psychological landscape of older adults. Central to the study is recognizing that a one-size-fits-all strategy is insufficient for sustained engagement. The framework creates a dynamic motivational environment by combining personalized elements, such as rewards, recognition mechanisms, and social support. The system's profound influence on mental health and social well-being is underscored, fostering a sense of belonging and purpose among older community members. The incentive system's versatility notably accommodates diverse preferences, promoting achievement and social connectivity. The study employs a multifaceted methodology, including multi-agent simulations, field experiments, and comprehensive questionnaires, revealing a significant impro

Keywords: Human-Machine Systems
Abstract: In this paper, we propose a method to detect automation surprises caused by autonomous driving of electric wheelchairs using image processing. Automation surprises refer to the general confusion caused by discrepancies between the driver’s and the system’s perception of the driving situation, or by the driver’s inability to fully grasp which mode the system is operating in. Therefore, we propose a detection method of automation surprises based on facial expression recognition and the operational input values. By identifying sections where the driver’s expression changes and excessive operational inputs to correct the vehicles unintended movement, we confirm that automation surprises can be detected from the racing game developed in Unity

Keywords: Human-Machine Systems, Electrophysiology and Kinesiology, Signal and/or Image Processing
Abstract: Responses to multiple-choice questionnaires for personality tests may be influenced by social desirability bias, impacting the reliability of personality assessment results. In this study, we propose a method to determine the presence or absence of deliberate responses influenced by social desirability bias in multiple-choice questionnaires using the beta wave content rate of EEG in the occipital lobe during responses. We conducted a verification experiment with 23 participants aged between 19 and 23. The results indicated an accuracy of 0.653, a positive predictive value of 0.996, a recall of 0.655, and an F score of 0.786, respectively.

Keywords: Human-Machine Systems, Intelligent Systems, Autonomous Decentralized Systems
Abstract: The ability of autonomous vehicles to recognize different objects in their natural environments, such as cars, trucks, pedestrians, bicyclists, and traffic lights, is made possible in large part by object detection. Enhanced object detection in self-driving traffic with the modified YOLOv8 architecture model is proposed. The attention mechanism is one of improve method detection object performance. The Efficient Attention Channel (EAC) integrates the attention mechanism into the original YOLOv8 architecture. The improved YOLOv8+EAC algorithm is tested through the self-driving traffic detection dataset to identify 10 class objects such as biker, car, traffic light-green, traffic light-green left, traffic-light red, traffic-light red left, traffic-light yellow and truck. The results show a precision confidence of 95.6% and mAP value of 72.4%. In addition, the comparative performance is superior to other architecture models. The proposed method indeed improves the accuracy of recognizing traffic signs in complex scenarios requirements of intelligent vehicles for traffic sign recognition tasks.

Keywords: Human-Machine Systems, Intelligent Systems, Robotic and Automation Systems
Abstract: Humans combine complex motions of the teeth, tongue, and cheeks during mastication to crush food, mix it with saliva, and form a swallowable food bolus. The fields of food science and industry desire to reproduce such a human food bolus formation with machines to evaluate the texture, taste, and aroma of food objectively and quantitatively. In this study, we propose a robotic mastication simulator dedicated to the faithful reproduction of food bolus formation. First, three fundamental operations involved in food bolus formation are defined as primitives: crushing, mixing, and gathering. Subsequently, the structure and motions of the robot are designed to execute these primitives, and the robotic mastication simulator is developed. Sequentially performing the primitives, the robot attempts to reproduce the human food bolus formation. Finally, human and the proposed robot masticate test foods, and the resulting food bolus images are analyzed using a convolutional neural network. It is shown that the robot using the appropriate sequence of primitives has a potential to reproduce human food bolus formation.

Keywords: Human-Machine Systems, Robotic and Automation Systems, Autonomous Decentralized Systems
Abstract: This work utilizes a NVIDIA Jetson NX board and a depth camera integrating with the Siamese Region Proposal Network (SiamRPN) algorithm for a human-following robot. To this purpose, a PID controller is employed to implement real-time control of a mecanum wheeled robot. The Jetson NX processes real-time imagery and depth information, while the SiamRPN learning model accurately locates and tracks the target. Finally, comparisons using AlexNet and VGG network algorithms in the proposed SiamRPN configuration are conducted. The result shows that the performance of AlexNet is better than that of VGG in terms of success and precision rates as well as frame rate. Moreover, the system's efficiency in achieving precise and reliable human-following is validated under different operation conditions through the conducted experiments.

Keywords: Human-Machine Systems, System Engineering
Abstract: Steady-State Visual Evoked Potential (SSVEP) is an　electroencephalography (EEG) signal synchronized with the flickering frequency of the visual stimulus. This feature is caused by neurons in the visual cortex response synchronized with the stimulus, and different responses are shown depending on the frequency and contrast of the stimulus. Conventional SSVEP studies have used pattern-reversal stimuli depending on the refresh rate of common displays or flash stimuli from a single point using light-emitting diodes (LEDs). However, these flashing stimuli implicitly contain spurious spectral components that may affect the biological response. In this study, we built an experimental system using a high-speed projector that enables precise stimuli with regard to spatio-temporal structure. Experiments using two types of visual stimuli –square waves and sinusoidal waves– showed that some harmonic components in the EEG were suppressed by sinusoidal wave stimuli compared with square wave stimuli.

Keywords: Human-Machine Systems, Human Interfaces, Biological and Physiological Engineering
Abstract: Typical symptoms of stroke include motor dysfunction. This can be interpreted as a state in which there is a difference between ''movement intention'' and ''actual body''. Therefore, to improve symptoms, it is important to consider the patient's ''awareness regarding movement intentions'' and ''awareness regarding the actual body''. In terms of these two types of consciousness, what has been focused on so far has been the movement that contemplates the initiation of action: conscious movement. However, investigating unconscious movements, which are movements that do not intend to initiate an action, will provide important clues to understanding these consciousnesses more deeply.

Keywords: Agricultural and Bio-Systems, Identification and Estimation, Signal and/or Image Processing
Abstract: This study proposes a method to measure crop row errors for crop row following control, considering the statistical properties of those errors. This method uses maximum likelihood estimation, assuming a normal distribution of errors and a straight tractor’s trajectory during sowing. A numerical example showed that this method can measure crop row errors. Additionally, when the trajectory was sinusoidal, shorter measurement intervals yielded results closer to those under straight-line conditions.

Keywords: Agricultural and Bio-Systems, Integrative Biophysical Engineering and Informatics, Networked Sensor System
Abstract: In recent years, the number of farmers in Japan has been decreasing and aging, and abnormal weather conditions such as typhoons and global warming have become problems. In addition, a stable supply and high quality of agricultural products are required. Plant factories are attracting attention as a solution to these problems and as a means of meeting customer needs. In our laboratory, we have constructed and demonstrated the temperature and VPD control system that can control the spray pressure of fine mist cooling. However, to adapt to different regions, changing weather conditions, cultivars and plans, and quality objectives, the relationship between plants and water needs to be clarified. Therefore, in this study, we developed an IoT system that can measure VPD and the amount of weight change due to transpiration in a plant canopy. Leaf area was measured using 3D model construction, and transpiration rate and vapor conductance were derived. As a result, the relationship between VPD and stomata was partially clarified.

Keywords: Agricultural and Bio-Systems, Intelligent Control
Abstract: This paper addresses a control problem of a sunlight-type plant factory with a special input saturation. The authors have studied its environmental control, especially, by focusing on vapor pressure deficit so far. However, it has not been explicitly considered that the input port has a saturation element including a discrete-valued section. This paper investigates what effect such an input saturation brings into the control system. Also, to improve the control performance, inserting a static quantizer between a controller and the saturation is proposed. The effectiveness of the proposed method is demonstrated via simulation.

Keywords: Agricultural and Bio-Systems, Signal and/or Image Processing, Flow Measurement and Control
Abstract: This study investigates the optimization of drone-based crop spraying for small-scale farms using computed tomography (CT) imaging to analyze droplet distribution from market-available nozzles. Small-scale farms often face challenges related to limited land area, tight budgets, and the risk of chemical drift from traditional spraying methods. Drone technology offers a potential solution for precision agriculture, but existing systems are primarily designed for large-scale operations. Our previous work established optimal nozzle placement for drone spraying, demonstrating effective spray confinement even under wind disturbances. However, we have noticed that the distribution of droplets may not always conform to the usual assumption of being uniform. To address this, we used CT imaging to capture precise droplet distributions. Images were acquired at various angles using a high-speed camera and back-light illumination. We employed both Filtered Back Projection (FBP) and Simultaneous Iterative Reconstruction Technique (SIRT) algorithms for CT reconstruction. This position paper presents the methodology of image capture and CT reconstruction techniques and illustrates preliminary reconstructed images. The detailed spatial droplet distribution data from the CT images will be integrated into our simulation models.

Keywords: Agricultural and Bio-Systems, Signal and/or Image Processing, Human Interfaces
Abstract: In this study, we developed a low-cost smartphone application that is easy to use and can be implemented in the field to assist strawberry farmers in harvesting their crop. This application can determine the ripeness and predict future ripeness of strawberries from their images, enabling inexperienced strawberry pickers to make accurate harvest decisions. The ability to quickly visualize ripeness leads to farmer's unification of harvesting standards. It is expected to lead to the production of high-quality, homogeneous, and high-brand strawberries. The developed system evaluates the ripeness of strawberries by calculating the ratio of ripe to unripe areas based on HSV color information from the acquired images. In order to achieve future ripeness prediction, the system additionally looks into the connection between ripeness and cultivation, and then builds a prediction model of ripeness change. Through basic verification experiments, we confirmed that the developed system can accurately determine ripeness, predict future ripeness, and provide real-time user feedback.

Keywords: Biological and Physiological Engineering
Abstract: Sitting down on and standing up from a chair are essential daily activities. Existing prosthetic knees lack functions to perform these motions, especially in standing up. Therefore, unilateral transfemoral prosthetic users mostly rely more on their sound limbs for the motions than their prosthetic limbs, which potentially leads to injury or joint disease due to overload on the joints of the intact side. The present study aimed to develop a prosthetic knee mechanism that assists in sitting down on and standing up from a chair. We newly proposed a passive mechanism that was integrated into our previous prosthetic knee mechanism. The proposed mechanism stores mechanical energy in the spring with prosthetic knee flexion and provides flexion resistance. Then, the mechanism releases the stored energy by the user’s manipulation and assists prosthetic knee extension. The subject who was attached to a simulated thigh socket and the proposed prosthetic knee performed sitting down on and standing up from a chair. The proposed mechanism showed a higher symmetry index of the ground reaction force compared to the existing prosthetic knees in the literature. The proposed mechanism can be applied to other prosthetic knees. Transfemoral amputees may be able to gain the function for standing up without replacement of their prosthetic knee.

Keywords: Biological and Physiological Engineering, Autonomous Decentralized Systems, Embodied-Brain Systems Science
Abstract: In order for advanced driver assistance systems (ADAS) to be widely accepted, not only safety but also comfortability is important. Therefore, in the ADAS development process, a method to quantify the driver’s comfort level naturally is required. For this purpose, we investigated driver’s electroencephalogram (EEG) to identify a physiological marker that reflects discomfort. Participants wearing EEG electrodes sat in a driving simulator and were instructed to brake actively at a target location under automated driving situation. At a low frequency of the trials, the brake system was experimentally set to fail in order to induce the sense of discomfort. To analyze the sense of discomfort on EEG signals, we compared event-related spectral perturbations (ERSP) between normal and brake failure conditions. The results suggest that a relative power decrease in the low-beta frequency band at the channels on the longitudinal fissure of the brain, and a relative power increase in the alpha and beta frequency bands at the frontal region are related to the sense of discomfort.

Keywords: Biological and Physiological Engineering, Integrative Biophysical Engineering and Informatics, Human Interfaces
Abstract: Functional Magnetic Resonance Imaging (fMRI) has been one of the most popular approaches for investigating the sophisticated haptic perception of humans. Little attention has been paid to the source of　the haptic stimulation.The limitation is that the conventional actuators are usually not usable inside the scanner due to the strong magnetic field (1.5T, 3T, 7T). Therefore, the developed devices are usually complicated and designed for a specific task. In this study, we proposed and evaluated a high resolution MRI-compatible tactile array. We demonstrated that such a tactile array may help improve the passive tactile performance of participants during MRI scanning.

Keywords: Biological and Physiological Engineering, Robust Control
Abstract: This paper develops a speed-control treadmill for a rehabilitation trainer that employs the neuro-developmental treatment (NDT) principles. NDT is an effective rehabilitation technique for stroke patients, but it also demands the therapists' time and effort. Consequently, patients might not receive sufficient training in the golden rehabilitation period due to the lack of therapists' assistance. Previous studies proposed a stationary trainer that could automatically repeat the NDT intervention to relieve therapists' work burden and increase patients' practice time. However, the subjects often needed to adjust their walking speeds when receiving the training because the trainer employed a treadmill with constant velocities, resulting in walking with unnatural gait patterns. Therefore, this paper proposed a speed-control treadmill, which can modify velocities according to the subject's walking speed. The treadmill equips a sensor to detect the subject's position and a motor control system to keep the position. We derived the treadmill model and designed robust PID control to adjust the treadmill velocity according to the user's walking speed during NDT training. Finally, we invited ten healthy subjects and three stroke patients to conduct experiments. Based on the results, the stationary NDT trainer employing the velocity-control treadmill could improve subjects' longitudinal symmetry, lateral asymmetry, pelvic rotation, walking speed, and step length.

Keywords: Ubiquitous Healthcare, Identification and Estimation, Human-Machine Systems
Abstract: Accurate evaluation of a patient’s walking ability is crucial for successful rehabilitation treatment. Existing gait analysis methods, such as optical and inertial motion capture systems, can be employed to directly obtain gait parameters, including joint angles. However, these methods require complex and expensive equipment, which may limit their widespread use. In this research, a simple low-cost insole pressure sensor was developed for the estimation of lower limb sagittal plane joint angles, namely the hip, knee, and ankle, during walking. A 7 minutes walking insole pressure data and ground truth joint angle data, obtained from an optical motion capture, were collected from one subject. Two deep learning models, a 1-Dimensional Convolutional Neural Network (1D-CNN) and Long Short-Term Memory (LSTM) architectures, were adopted for the estimation and the results were compared. Both the 1D-CNN and LSTM models achieved average Mean Absolute Errors (MAEs) of 1.13° and 1.20°, respectively, for the six joint angles. Furthermore, the trained model was integrated with an iPad application for real-time estimation and visualization through a 2D human model animation. The proposed method demonstrates the potential for accurate gait analysis using a low cost and easily accessible equipment, suggesting its possibility of wider adoption for gait analysis in clinical practice.