Keywords: Robotics
Abstract: This paper proposes a novel method of Chinese character stroke extraction and a framework for human robot skill transfer through vision-based observation. By analyzing the structure of Chinese characters, a direction vector update rule and a pixel finding rule were devised to find the basic strokes. Then we designed a basic stroke connection algorithm to achieve stroke extraction. Afterward, to adapt to human interference in real-time, we adopt dynamical movement primitives (DMPs) to model writing skills. Finally, the adaptive capability of the method was verified by experiments in which the robot writes Chinese characters on a randomly moved writing board.

Keywords: Flexible Manufacturing Systems, Robotics, Man-machine Interactions
Abstract: (Re-)Programming industrial robots today still remains an intricate process requiring expert knowledge. Therefore enterprises often avoid modifications to existing robot cells, even if processing could be further improved. This contribution presents a projector-based Augmented Reality support system that supports shop-floor-level inspection and intuitive modification of milling robot operations. The system enables visualization of the programmed milling processes directly on top of the workpiece and allows even for non-experts to verify or change the robotic work program directly in the work cell if needed. Technical challenges and proposed solutions are discussed, and the developed demonstrator system is presented. The system was tested in an industrial shop-floor work environment processing large-scale metal workpieces with high production tolerances, making it an ideal candidate for evaluating the practicability of this AR system for quick and user-friendly robot work program modifications.

Keywords: Robotics, Flexible Manufacturing Systems, Motion Control
Abstract: This paper proposes a framework to make the robot programming faster, user-friendly, and flexible for kitting operations. This is achieved by innovating at three different levels; (i) at the motion stack level, it integrates the robot with a constraint-based robot programming paradigm in which a novel approach for obstacle avoidance is introduced, (ii) at the discrete planning level, it introduces adaptive planning to derive an optimal sequence of robot kitting operations, (iii) and at the knowledge representation level, the overall operation is aided by a world-model that formally represents and provides services from the knowledge of the task, the robot state, and the sensor data. The framework provides industrially relevant capabilities like, collision avoidance independent of the obstacle location, optimal discrete planning of the actions based on the costs and goals, and dealing with complex conditions such as object blocking and object graspability. The performance of the proposed framework is evaluated through a set of experiments.

Keywords: Estimation and Identification, Energy Efficiency, Robotics
Abstract: Flight time due to available energy plays a crucial role in unmanned aerial vehicle (UAV) applications in industry. It has not yet been established a general admitted UAV energy model and previous works highlight the importance of experimental tests to determine the parameters of an energy consumption model. This paper presents the experimental results of several energy consumption tests of an UAV when few flight data are accessible. We have followed the methodology proposed in other works in the literature studying the energy consumption in four types of flights: hovering, steady level flight, vertical upward movement, and vertical downward movement. The energy consumption model is obtained without the need to add new sensors, only with flight data. Our study provides an approximation as a first degree linear model and a guideline to determinate the hovering time threshold (HTT), that is, when to land an UAV is more convenient than keep it on hovering.

Keywords: Nonlinear Systems and Control, Motion Control, Control Applications
Abstract: In the training of patients assisted by lower limb exoskeleton rehabilitation robot (LLERR), precise control is desired under a given trajectory to achieve a certain training effect. In this paper, a nonsingular terminal sliding mode (NTSM) control strategy is put forward for a 2-degree-of-freedom (DOF) LLERR. Initially, a Lagrange equation is used to establish the dynamic model of the 2-DOF LLERR. Then, the NTSM controller is designed for the LLERR, where the stability of closed-loop control system is substantiated in the Lyapunov sense. For comparison, a conventional PID controller is also designed as the benchmark. Finally, MATLAB-OpenSim co-simulations are carried out to test the control performance with and without disturbance. The simulation results illustrate that the proposed NTSM control method possesses apparent advantages, which is reflected in higher tracking precision and tougher robustness compared with the benchmark controller.

Keywords: Sensor/Data Fusion, Estimation and Identification, Multi-agent Systems
Abstract: This paper addresses the localization problem for multiple disjoint targets using noisy time difference of arrival measurements in wireless sensor networks. In comparison with the classic two step weighted least squares method which firstly ignores the non-linear relation between each auxiliary variable and the corresponding target, the proposed method follows the weighted least squares strategy but takes these relationships as non-linear constraints in the formulation. An iterative approximate procedure is proposed to solve it efficiently. The Monte Carlo numerical simulation results demonstrate the good localization performance of the proposed method.

Keywords: Multi-agent Systems, Sensor Networks
Abstract: This paper studies the multisensor-based distributed localization of a multi-agent system given the orientations and positions of the leaders and inter-agent distance measurements. Each agent is equipped with multiple range sensors. Different from the existing distributed localization in static sensor networks, this paper focuses on distributed localization in dynamic sensor networks, i.e., each agent is allowed to move freely. Both position and orientation estimation protocols based on distance measurements are proposed, which can estimate the positions and orientations of the followers globally. Moreover, the proposed method can also be applied in the angle-based distributed localization.

Keywords: Networked Control, Modeling and Control of Complex Systems, Discrete Event Systems
Abstract: A new concept called the Nash equilibrium set (NES) is proposed. Using it, the set stability and stabilization of networked evolutionary games (NEGs) are investigated. Necessary and sufficient conditions for stabilizability to an NES are obtained, and concrete control design techniques are provided. Some numerical examples are presented to illustrate the theoretical results.

Keywords: Multi-agent Systems, Nonlinear Systems and Control, Networked Control
Abstract: In this paper, we consider the safety verification problem for a large-scale nonlinear system. Existing tools for safety verification are not scaled well for an interconnected system with nonlinear dynamics and strong interconnection. To solve the problem, we propose a cyclic-small-gain condition to verify the safety of a large-scale nonlinear interconnected system. We show that the higher-dimensional safety set of the large-scale nonlinear system can be constructed from the lower dimensional safety set corresponding to each subsystem if the local input-to-state safety and the cyclic-small-gain conditions are satisfied.

Keywords: Multi-agent Systems, Networked Control, Nonlinear Systems and Control
Abstract: In this paper, we consider the cooperative robust output regulation problem for a class of nonlinear multiagent systems with relative degree two over jointly connected switching networks. First, by attaching each subsystem with an input driven filter, we obtain an extended system. Second, by integrating the distributed observer approach and the distributed internal model approach, we convert the cooperative output regulation problem of the original system into a distributed robust stabilization problem of an augmented error system which is also a multi-agent system. Finally, we solve the original problem by stabilizing the augmented error system via a distributed output feedback control law.

Keywords: Discrete Event Systems, Linear Systems, Nonlinear Systems and Control
Abstract: For a linear system with multiple linear controllers, the problem of stabilization is to seek proper control inputs and a control law that steer the system exponentially stable. Based on the system canonical decomposition and cyclic matrix theory, a common state-feedback controller is designed to stabilize the system. The proposed controller is always continuous along the switching law, which avoids discontinuity of the control input signal. A numerical example is presented to verify the effectiveness of the proposed scheme.

Keywords: Nonlinear Systems and Control, Linear Systems, Control Applications
Abstract: This paper investigates the output regulation problem for switched strict-feedback nonlinear systems. Besides the regulation of output error of the closed-loop systems, the state constraints of the system are addressed during the regulation control. By the constructed switched regulator equations and coordinate transformation, the output regulation problem is converted to a constrained stabilization problem. Moreover, based on a barrier Lyapunov function technique and a time dependent switching rule, a state feedback control scheme is constructed to achieve that the error output tracks a desired trajectory asymptotically and the system states keep in given bounds. The effectiveness of the result is demonstrated by a simulation example.

Keywords: Fault Detection and Diagnostics, Estimation and Identification, Learning Systems
Abstract: Time series anomaly detection has attracted great attention due to its widespread existence in real life. With the increasing development and advancement of deep learning, many unsupervised deep learning methods have been proposed for time series anomaly detection since labeling time series is prohibitively expensive. In this paper, we propose an unsupervised anomaly detection method: Gaussian Mixture Variational Autoencoder with Whitening Distance Anomaly Score (WGVAE). Concretely, we employ a Long Short-Term Memory (LSTM)-based variational autoencoder (VAE) to capture the long-term dependence of time series and learn the lowdimensional feature representation and distribution, in which the Gaussian mixture prior are used to characterize multimodal time series. Further, whitening distance anomaly scores are used to make the multidimensional time series independently and identically distributed among each dimension, which combines the distribution characteristics of the samples to measure the degree of outliers. When the anomaly score is below the threshold, the sample is detected as anomalous. Finally, comprehensive experiments are given to verify the effectiveness of our method.

Keywords: Fault Detection and Diagnostics, Learning Systems
Abstract: High-dimensional and high-complexity unsupervised abnormal data recognition with neural networks remains an open research question. We propose a multi-level Variational Auto Encoder (VAE) structure in this work. First, based on traditional VAE, we deconstruct the original data using convolution kernels of different specifications to establish multiple feature latent spaces, thereby reducing the excessive data aggregation in the latent space. Then we use an equal-window convolution kernel to fuse feature blocks of different sizes, thereby transforming local features into globally identifiable features. Moreover, the model structure shortens the distance between the sub-feature block and the output endpoint, reducing the training inefficiency caused by the vanishing gradient.

Keywords: Fault Detection and Diagnostics, Learning Systems
Abstract: As an indispensable part of carrier-based aircraft, the actuator system plays an important role in ensuring the flight safety. Fault detection and diagnosis of actuator are necessary for improving actuator system reliability. Motivated by solving the uncertainty problem in fault diagnosis of actuator system, which is caused by various reasons, such as bias and noise of sensors, this paper proposes a deep stacked autoencoder network-based (DSAEN) deep learning fault diagnosis method for flight control system. The fault diagnosis performance under different network structures is further studied and the accuracy of the diagnosis model is improved.

Keywords: Learning Systems, Fuzzy and Neural Systems, Intelligent and AI Based Control
Abstract: The International Classification of Disease (ICD) coding scheme authorized by the World Health Organization, is a standard used to classify medical diagnosis in health insurance claims. One of the key challenges in adopting machine learning techniques for health insurance claims processing lies in the sparse and complex nature of ICD codes. There are over 69,000 unique diagnosis codes in the 10th version of the ICD coding scheme where multiple codes are used to represent complex medical diagnosis. In this work, we address the challenge of embedding both single and multiple ICD codes into a vector space. This representation addresses the sparse representation of ICD codes by providing a more compact representation that captures their diagnosis relationship. Our proposed inductive multi-code representation introduces a novel application of inductive node embedding by representing multiple ICD codes as a multi-code node. This enables both single and multiple ICD codes to be represented in the same vector space. For multi-code relevance, our proposed approached surpassed the benchmark node2vec implementation by 34%, highlighting its effectiveness

Keywords: Signal Processing
Abstract: X-ray fluorescence (XRF) is an effective technique for element analysis which has gained increasing attention in a variety of applications. However, the continuous background is inevitably introduced to X-ray spectra collected from samples because of device-induced reflections and experimental environment. The existence of continuous background will significantly worsen the analysis precision of the spectra. Thus, in this paper, an improved background estimation algorithm based on the variational mode decomposition (VMD) is proposed. First, the spectral signal can be decomposed into a specified number of modes, which are characterized by different frequencies ranging from the high to the low. These modes are all have actual physical meaning and low frequency mode signal is utilized to estimate the continuous background information. Then, the original spectral signal serves for correcting estimated background after each iteration until the fit converges close to the real background. Finally, the clean spectral signal can be achieved by deducting the estimated background. The simulated data demonstrate the effectiveness, stability and convenience of the VMD. Furthermore, the results of the experimental data show that VMD has a remarkable performance even in removing the complex background of the spectra of the ore samples.

Keywords: Learning Systems, Real-time Systems, Smart Buildings
Abstract: Respiration, a vital basis for life, is a key indicator of health status for human beings. Recently, with contact-based devices, some breathing signal detection methods have been proposed, which can achieve high accuracy and signal-to-noise ratio performance, empowering many healthcare applications. However, these methods require users to be contacted with the devices, leading to a series of problems, such as hindering the movement of users. Therefore, there is an urgent need to call for a contactless solution for respiration detection. With the popularity of the indoor WiFi devices, respiration detection with WiFi sensors has drawn a lot of attention. Nevertheless, the multipath effects, which commonly exist in indoor environments, have serious impacts on the propagation of wireless signals, leading to signal attenuation and poor signal quality. Moreover, although the channel state information (CSI) can be readily collected from commercial off-the-shelf (COTS) WiFi devices, the received CSI is distorted due to various offsets introduced during the propagation of the wireless signals and hardware imperfections. In this paper, we try to resolve the challenges mentioned above, and propose a device-free respiration detection system, ResFi, utilizing the CSI data from COTS WiFi devices. The final evaluation shows an accuracy of 96.05% for human respiration detection, which is up to 15% higher than that of the traditional machine-learning methods.

Keywords: Nonlinear Systems and Control, Motion Control, Control Applications
Abstract: This paper presents quaternion based twisting and backstepping controller for quadrotor. The quaternion model is free from singularity problem unlike the Newton Eulers model. There are two types of controller is designed, one for inner loop control and second is the outer loop control. In the proposed work inner loop controller is backstepping and outer loop control is twisting control. Inner loop control is also termed as attitude control. In attitude control reference velocity is generated using reference quaternion model. Both inner loop and outer loop controller are robust enough to handel disturbances. Stability is guaranteed using Lyapunov method. Simulation results are presented for different generated paths to show the advantage of the proposed method.

Keywords: Automated Guided Vehicles, Motion Control, Robotics
Abstract: This paper presents simulation-based realisation of an autopilot algorithm for UAV to approach the mounting position for weather station on the transmission tower. The specified corner leg serves to install and maintain the sensor network in the electric grid, aims to build a weather-dependent capacity monitoring and forecasting system. The algorithm is used to detect the transmission tower while a UAV is approaching, divided into the following steps: with consideration of the mathematical attribute of mast corner legs, the line detection based on gradient and length filter is used to extract and describe its features; the relative initial pose of UAV with respect to specified mounting leg is determined to pursue the rough adjustment; the thresholds are set to trigger the safety algorithm, to prevent the deviation of tower outside the field of view of camera, while the UAV is approaching toward the target; the fine adjustment is considered based on the pose and depth estimation of the diverse cross-section profile of mast leg. In addition, the algorithm for collision avoidance and preservation of the target under the observation is integrated. As a result, the algorithm performs good in the gazebo simulator and achieves its initial purpose, in which the transmission towers, overhead lines, light source and other objects with real scale are added.

Keywords: Control Applications, Robotics, Nonlinear Systems and Control
Abstract: This paper proposes a robust backstepping control for a quadrotor in the presence of uncertainties and packet dropouts. The proposed work uses the concept of super twisting and twisting reaching law for designing the novel backstepping control. Using backstepping control, virtual control input can be developed, which rejects the unmatched uncertainty. The real control input counteracts the matched uncertainty via the input channel. The quadrotor is prone to both matched and unmatched uncertainty for real-time applications; hence the proposed novel controller offers more robustness towards uncertainty. For reducing the computation effort of the controller and communication burden, event-triggering conditions are derived. Furthermore, the allowable number of packet dropouts between two triggering instants are derived. Simulations are conducted to validate the proposed technique.

Keywords: Fault Detection and Diagnostics, Intelligent and AI Based Control, Signal Processing
Abstract: Quadrotor has been developed and applied for various civil aviation and military services. Since its wide applications, this raises a safety concern. For example, rotor or sensor may have faults which may cause the quadrotor to crash. In this paper, we present a fault-tolerant control (FTC) scheme that can be used to handle GPS faults. The proposed control uses neural networks to learn dynamical behaviors and estimate the GPS value. Simulation studies have been given to illustrate the effectiveness of the proposed FTC method

Keywords: Nonlinear Systems and Control, Robotics, Control Applications
Abstract: This work presents the development of a back-stepping nonlinear integral sliding mode controller (BSNISMC) for a 3 DoF vision augmented quadrotor, in the presence of system uncertainties and bounded external disturbances. Vision augmented models play a significant role in visual servoing by enabling the use of torque based commands to directly alter the position of the quadrotor with respect to the desired point of interest (POI), in the camera frame. Since the model, at hand, is a vision augmented model, the overall control heavily relies upon the availability of quadrotor and camera parameters, which, in most practical cases, cannot be determined accurately. In addition, camera noises as well as external disturbances hinder the performance of the system significantly. To overcome these issues, this work develops a back-stepping nonlinear integral sliding mode controller which is known to be robust against bounded external disturbances. The inclusion of the nonlinear integral sliding manifold ensures better steady state response of the system. To cater to the system uncertainties, adaptation law's for these parameters are obtained using Lyapunov's theory, which also aids in establishing the stability of the system. Numerical simulations are presented to demonstrate the efficacy of the controller and the adaptation laws.

Keywords: Estimation and Identification, Control Applications, Control of Smart Power Delivery Systems
Abstract: The on-line inertia identification based anti-disturbance control method is concerned for the attitude loop of quadrotor system in this paper. The inertia matrix is important in the attitude-loop stabilization, state estimation and failure diagnosis for quadrotor or spacecraft generally, however, it is challenge to measure directly or calculate from the model due to the geometry complexity. An online identification method is proposed for the inertia matrix without specific command signals, which can converge to real one with small storage and incremental computation. Compared with the existing algorithm, the proposed approach works well in scenario where the inertia is time-variant, such as variable payload. Besides, an extended state observer is designed with identified inertia to compensate the external disturbance existing inherently in attitude loop of quadrotor system. The results of simulation and experiment are presented to show the validity of proposed approach, including inertia identification and disturbance observation.

Keywords: Control Applications, Energy Efficiency, Process Control & Instrumentation
Abstract: The present work describes a Model Predictive Control application aimed at satisfying the production plan of a hydroelectric plant. The considered process is characterized by two reservoirs (connected through a regulation gate) and a set of turbines for energy production. The predictive control approach has been motivated by the availability of forecasts on water requests by the hydroelectric plant and by the goodness of fit performances of the obtained linear process model. This model is characterized by first principles equations together with tailored relationships between the sensor level information and reservoir water volume (obtained from data). Unknown water flows and model uncertainties have been considered through ad hoc model mismatch compensation strategies. The precise fulfillment of water requests aimed at production plan satisfaction represented the main goal of the controller, together with the minimization of the wasted water. Virtual environment simulations based on significant scenarios have shown the soundness of the proposed approach. The developed proprietary framework has been tested on the real plant, obtaining a significant service factor and remarkable results.

Keywords: Control Applications, Signal Processing, Learning Systems
Abstract: A Harmonic Acoustic Pneumatic Source (HAPS) is an acoustic device developed to perform active harmonic sound control in turbofans. In order to develop a demonstrator, a ring of multiple HAPS (RHAPSODI) is assembled around a cylindrical duct and a dedicated MIMO harmonic controller was developed to perform active sound power attenuation. The control objective is to develop a MIMO feedback controller adapted to multiple HAPS and using in-duct microphones close to them. Three problems are addressed : the control of HAPS as a mechanical modulator, the in-duct error microphones located in the near field of HAPS, and the tuning of the controller. The presented method is based on a dedicated MIMO feedback controller of complex envelops with compensation of near field. The compensation matrix of the near field is tuned during a learning phase based on the measurement of a set of optimal commands experimentally obtained with a dedicated controller using the out-duct microphones. When the matrix of compensation is estimated, the MIMO controller using in-duct microphones can be used. Preliminary experimental results show that a RHAPSODI of 5 HAPS can attenuate the first harmonic of a radiated tonal sound power by more than 20 dB.

Keywords: Control Applications, Motion Control
Abstract: Laser-Assisted Charge Exchange (LACE) requires the position and angle of a high-power laser beam to be stable at the sub-millimeter and sub-milliradian level at a point 65 meters from the laser. Due to the high laser power and improvised nature of the laser transport line, the laser beam suffers from pointing instabilities in the form of drift and pulse-to-pulse jitter. A closed-loop Laser Pointing Control System (LPCS) has been developed in the lab and deployed to the field to control and stabilize both the position and angle of the laser beam. The LPCS relies on feedback between CMOS cameras and beam relay mirrors on kinematic mounts controlled by a PC running custom LabVIEW software. The system has been demonstrated to reduce the standard deviation of beam position to 200 micrometers and of pointing angle to 300 microradians at the interaction point, corresponding to a reduction of ~2x.

Keywords: Control Applications, Process Control & Instrumentation, Process Automation
Abstract: In textile processes, accurate temperature control is crucial in order to achieve right first time dyeing. The widely used PI (Proportional-Integral) controllers in the industry are difficult to tune by using standard methods as the dyeing process is highly complex and batch-varying. In this paper, a novel PI auto-tuning method suitable for temperature control in a variety of textile dyeing processes is developed and implemented. Simulation and field application results are provided in order to confirm the success of the proposed technique.

Keywords: Sensor/Data Fusion, Signal Processing
Abstract: A FMCW(Frequency-Modulated Continuous Wave) LiDAR recently attracts a lot of interest because it is much robust to harsh environmental conditions and sun light because it takes the advantage of the wave properties of the laser. It measures the distance by calculating the optical interference frequency generated by interfering a reference beam that is modulated with a received beam that is reflected from an object. A tunable laser can be used for modulating the frequency of the laser beam. However, it is difficult to have a linear frequency modulation by the applied current to the tunable laser. For example, when the current is increased proportionally, the interference frequency varies with respect to time. Hence, a linear frequency modulating control of the tunable laser is proposed to improve the dista

Keywords: Control Applications, Motion Control
Abstract: A closed-loop compensation control structure was designed for permanent-magnet synchronous motor (PMSM) servo drives, to improve the performance of PMSMs in performing constant-speed motion and low-speed reversal motion. PMSMs have been used extensively in various types of industrial mechanical equipment. To mitigate the adverse effects of friction and external disturbances caused by loading conditions, PMSM servo drives developed in recent years have been provided with an additional compensation control structure that can improve the motion performance of PMSMs. However, PMSMs are still significantly affected by nonlinear friction during low-speed reversal motions. Therefore, in this study, compensation control structures, i.e., a disturbance observer and compensator, a coulomb friction compensator, and a low-speed friction compensator, were designed with consideration of the friction and external disturbance characteristics of the loading conditions. Constant-speed motion and reversal-motion tests were performed on a PMSM test bench. The experimental results indicated that the rate of improvement for the root-mean-square speed error of the compensation control structure designed in this study was 46.44%, confirming the feasibility and compensation effects of the proposed method.

Keywords: Robotics, Real-time Systems, Learning Systems
Abstract: The LiDAR and inertial sensors based localization and mapping are of great significance for Unmanned Ground Vehicle related applications. In this work, we have developed an improved LiDAR-inertial localization and mapping system for unmanned ground vehicles, which is appropriate for versatile search and rescue applications. Compared with existing LiDAR-based localization and mapping system such as LOAM, we have two major contributions: the first is the improvement of the robustness of particle swarm filter-based LiDAR SLAM, while the second is the loop closure methods developed for global optimization to improve the localization accuracy of the whole system. We demonstrate by experiments that the accuracy and robustness of the LiDAR SLAM system are both improved. Finally, we have done systematic experimental tests at the Hong Kong science park as well as other indoor or outdoor real complicated testing circumstances, which demonstrates the effectiveness and efficiency of our approach. It is demonstrated that our system has high accuracy, robustness, as well as efficiency. Our system is of great importance to the localization and mapping of the unmanned ground vehicle in an unknown environment.

Keywords: Robotics, Intelligent and AI Based Control, Learning Systems
Abstract: The vision of unmanned aerial vehicles is very significant for UAV-related applications such as search and rescue, landing on a moving platform, etc. In this work, we have developed an integrated system for the UAV landing on the moving platform, and the UAV object detection with tracking in the complicated environment. Firstly, we have proposed a robust LoG-based deep neural network for object detection and tracking, which has great advantages in robustness to object scale and illuminations compared with typical deep network-based approaches. Then, we have also improved based on the original Kalman filter and designed an iterative multi-model-based filter to tackle the problem of unknown dynamics in real circumstances of motion estimations. Next, we have implemented the whole system and do ROS Gazebo-based testing in two complicated circumstances to verify the effectiveness of our design. Finally, we have deployed the proposed detection, tracking, and motion estimation strategies into real applications to do UAV tracking of a pillar and do obstacle avoidance. It is demonstrated that our system shows great accuracy and robustness in real applications.

Keywords: Automated Guided Vehicles, Robotics
Abstract: Automated Guided Vehicles (AGVs) have played an import part in many areas, such as transporting and delivery, domestic cleaning, search and rescue. In many industrial scenarios, 2D LiDAR is a popular localization and navigation device due to its relatively low cost and high accuracy. However, AGVs equipped with 2D LiDAR are unable to detect obstacles in 3D space which limits their applications in complex environment. Vision-based approaches are able to solve this problem, however, they are not efficient in map building and navigation capability. To enhance the intelligence and capabilities of traditional AGVs equipped with 2D LiDAR sensors and make it more robust in various environments, we propose a vision-aided localization and navigation system which integrates the advantages of camera and 2D LiDAR. We propose an efficient obstacle detection method in 3D space and reflect them in a 2D map produced by a 2D LiDAR. A model predictive control (MPC) based path planning is then utilized for autonomous navigation with collision free capability. Experiments with industrial AGVs in warehouse demonstrate that the AGVs are able to localize accurately and navigate with collision avoidance capability with our proposed method.

Keywords: Robotics, Automated Guided Vehicles, Real-time Systems
Abstract: In recent years, the rapid development of Visual Simultaneous Localization and Mapping (vSLAM) technology has provided improvements for robot localization. However, for autonomous underwater vehicles, feature extraction methods in underwater environments have not been studied much due to the specificity of their application scenarios. For this purpose, this paper evaluates and analyzes the performance of feature detecting and matching in underwater environments by using well-known interest points detectors and descriptors in vSLAM. The performance of different well-known interest points detectors and descriptors is compared and analyzed by simulation experiments using four sets of underwater image data and combining histogram equalization-based and deep learning-based image enhancement algorithms. The experimental results have verified the effectiveness of the image enhancement algorithm and feature extraction methods in underwater environments, and the quantitative comparison concludes that ORB algorithm has an advantage in underwater environments.

Keywords: Integrated Manufacturing, Flexible Manufacturing Systems, Modeling and Control of Complex Systems
Abstract: In this paper, a three-phase hybrid heuristic algorithm based on product supply relationship (TP-HHA) is proposed to deal with the problem of textit{order and rack assignment for picking stations} (ORAPS). The ORAPS occurs in the order picking process of robotic mobile fulfillment system (RMFS). It is an NP-hard problem so that the optimal algorithm is difficult to effectively solve it with larger scale instances, which cannot meet the requirements of real applications. To overcome this difficulty, we propose the TP-HHA. It firstly uses a problem decomposition frame to divide the ORAPS into three subproblems. Then, in each stage, a meta-heuristic or constructive heuristic is designed to solve the corresponding subproblem. In addition, a post-optimization strategy is employed to find a better solution. The sufficient neighborhood search operators with problem-specific knowledge are designed for both subproblem optimization and post-optimization. The computational results are given to verify the effectiveness of our algorithm.

Keywords: Automated Guided Vehicles, Control Applications, Networked Control
Abstract: This paper investigates the co-design problem of the integrated event triggering (IET) scheme and integral sliding mode control (ISMC) for the nonlinear connected vehicles (CV) with communication constraints. The main goal is to improve the tracking performance and increase the bandwidth utilization of vehicular ad-hoc networks (VANET) with various external disturbances. Firstly, a nonlinear CV dynamic model is established by considering the disturbance from the preceding vehicle, internal and external resistance, and uncertainty of the engine time constant. Secondly, an IET transmission mechanism based on both absolute and relative restriction conditions of sampled-data error is proposed to reduce unnecessary data transmissions among vehicles. Thirdly, co-design IET mechanisms and ISMC controller based on estimated of H∞ observer is designed, which can achieve string stability. Finally, extensive simulations are conducted to verify the effectiveness of the proposed method.

Keywords: Multi-agent Systems, Control Applications, Nonlinear Systems and Control
Abstract: This paper gives an experimental verification for circle formation indoors. A circle formation platform is constructed by quadrotor UAVs, which consists of the UWB positioning system, the control system of quadrotor UAVs and the ground station. The Kalman filter is used to achieve the stable and accurate results of ultra-wideband (UWB) positioning. A leader-following control law of circle formation is realized by using the yaw angle velocity of each follower. Experimental results show that the positioning errors are less than 20 cm and at the same time the random error is less than 3 cm. Two cases, that is, the stationary leader and the moving leader, are under consideration, which indicates that all followers achieve the circle formation motion successfully.

Keywords: Multi-agent Systems, Learning-based Control, Networked Control
Abstract: In this paper, an optimal consensus control protocol based on Q-learning algorithm is proposed for a class of discrete-time multiagent systems with unknown system matrices and state delays. It is well known that coupled Hamilton Jacobi Bellman(HJB) equation is difficult to be solved especially for multiagent systems(MASs) with state delays. On the basis of the coordinate transformation, MASs with state delays can be converted to a corresponding delay-free system under certain conditions. Then, Q-learning algorithm is adopted by using the delay-free system data rather than the accurate system model. We formulate the Q-function Bellman equation and adopt policy iteration(PI) to calculate the optimal control iteratively. Finally, a simulation example is applied to demonstrate the validity of the optimal consensus control protocol.

Keywords: Modeling and Control of Complex Systems, Networked Control, Multi-agent Systems
Abstract: In this paper, a complex network of nodes with multidimensional dynamics is studied in terms of the minimum number of attacked nodes' state variables for the purpose of attacking the whole network. A multidimensional node is considered as a subnetwork and the multidimensional states of the node are considered for partial attack. Maximum matching principle is performed for multidimensional complex networks to determine the minimum number of attacked nodes. Based on the exact controllability theory, an algorithm is proposed to search the optimal set of driver nodes. For networks with root strongly connected components which are perfect matching (pm-rSCC), attacking the minimum set of driver nodes does not necessarily satisfy the requirements, and there are additional strategies for the selection of attacked nodes. Finally, a general flowchart for identifying the minimum set of attacked nodes for multidimensional complex networks is given by graphical approach and exact controllability theory. The effectiveness of the scheme is demonstrated by simulation examples.

Keywords: Multi-agent Systems, Nonlinear Systems and Control, Robotics
Abstract: This work is concerned with the fixed-time bipartite consensus tracking (BCT) problem for multiple Euler-Lagrange systems (MELSs) with asymmetric input constraint. Firstly, a fixed-time distributed observer (FTDO) is constructed to capture the leader's states. Then, the adaptive local control protocol (LCP) is designed to ensure that the followers' trajectories converge to the leader's trajectory within fixed time. Besides, a hyperbolic tangent function is used to deal with the issue of asymmetric saturation. With the Lyapunov stability theory, it is proved that all errors are semi-global fixed-time stable. Eventually, a numerical example is presented to show the validity of the proposed method.

Keywords: Multi-agent Systems, Control of Distributed Generation Systems, Fuzzy and Neural Systems
Abstract: This paper addresses the output consensus tracking control problem of leader-following multi-agent systems (MASs) under directed network topologies. It is assumed that the dynamics of the leader has an unknown nonlinear function and each follower is an unknown nonaffine nonlinear pure-feedback system. By translating each nonaffine nonlinear pure-feedback system into the lower triangle system by our previous transformation, a novel distributed observer for each follower is designed to estimate the unknown output consensus tracking errors of the lower triangle system meanwhile a Nussbaum-type function is used to stabilize the unknown sign of input of the lower triangle system. Then a novel adaptive fuzzy control law is designed without using any global information of topology. The asymptotic properties of the resulting adaptive fuzzy system are studied in detail. Simulations show the feasibility and effectiveness of the proposed scheme.

Keywords: Adaptive Control, Nonlinear Systems and Control, Control Applications
Abstract: In this study, a modified adaptive-gain super-twisting sliding mode guidance law with finite time convergence is proposed to intercept targets with impact angle constraint. The proposed guidance law introduces a linear term and an inner negative feedback so as to restrain the overshoot and enhance the performance of faster convergence. Furthermore, the adaptive-gain schedule can attenuate the chattering in the presence of the perturbations with the unknown boundaries. The finite time stability of the guidance law is analyzed. Simulation results show that the proposed guidance law is able to achieve more satisfactory interception. Index Terms—Super-twisting, impact angle constraint, adaptive-gain, inner feedback, linear term

Keywords: Automated Guided Vehicles, Control Applications
Abstract: Real-time motion planning is critical to real-world applications of autonomous surface vehicles (ASVs) moving in a complex sea environment. This paper presents an efficient motion planning method to generate collision-free trajectories for a fleet of ASVs subject to static obstacles. The safe motion planning algorithm is proposed based on high-order control barrier functions (HOCBFs), which are able to address the obstacle and inter-vehicle collision avoidance. The optimal collision-free trajectories are obtained by solving a quadratic program (QP) problem. The proposed safe motion planning algorithm is able to achieve the following two objectives: generating feasible yaw rates to guide each ASV toward its goal, and handling static obstacles and model constraints, simultaneously. Simulation results are given to substantiate the efficacy of the proposed safety-critical motion planning method for multiple ASVs subject to kinematic model constraints and collision avoidance constraints.

Keywords: Modeling and Control of Complex Systems, Control Applications, Nonlinear Systems and Control
Abstract: This paper focuses on designing and implementing the robust finite-time attitude tracking controller for small- scaled fixed-wing UAVs subject to nonlinearities, parameter perturbation, modeling error, and gust wind disturbances. Time-domain system identification approaches are introduced to obtain the longitudinal and lateral dynamic models of fixed- wing UAVs. By the feedback linearization technique, a linear equation with equivalent disturbances is derived to describe the attitude dynamics, for which a new robust finite-time controller is constructed. Based on the backstepping methodology and signal compensation theory, the controller incorporates two parts: a non-smooth nominal part using the sign function to achieve the desired finite-time convergence property and a robust compensator to restrain the influences of equivalent disturbances. Finally, the identification experiments are conducted and the effectiveness of the proposed controller is demonstrated by the Hardware-in-loop simulation studies.

Keywords: Multi-agent Systems, Learning-based Control, Nonlinear Systems and Control
Abstract: This paper investigates a distributed time-varying output formation control framework for heterogeneous multiagent systems (MASs) based on reinforcement learning (RL) method with multiple leaders and various disturbances. The outputs of followers are designed to accomplish robust tracking along with the movement of leaders’ convex combination while achieving a predefined configuration of time-varying formation. A three-layer framework, composed of a distributed adaptive finite-time observer, an off-policy RL-based optimal tracking controller and a robust formation controller, is proposed in a model-free manner while neither the global information of graph nor followers’ dynamics is utilized. The stability of this integrated observer-based controller is analyzed using Lyapunov theory which indicates that the formation tracking error asymptotically converges to zero under both external and internal unknown disturbances. Simulation results provide a detailed validation of the proposed control architecture.

Keywords: Estimation and Identification
Abstract: This paper focuses on the identification of graphical autoregressive moving-average (ARMA) models. Existing methods address the identification problem by estimating the graph topology, moving-average (MA) and autoregressive (AR) parameters in a separate way. To improve the identification efficiency, we design a two-stage identification algorithm, in which the AR and MA parameters are coupled together and can be estimated together with the graphical structure. Since a low-order ARMA model can be approximated by an AR model of appropriate high order, the identification object can be converted to the approximate graphical AR model, whose graph topology is identical to that of the primal graphical ARMA model. Based on l1-type nonsmooth regularized conditional maximum likelihood estimation and information theoretic model selection criterion, the simultaneous identification of the graphical structure and parameters of the approximate graphical AR model can be achieved. Then, the AR and MA parts of the primal graphical ARMA model are decoupled from the estimated parameters. Simulation results illustrate the effectiveness of the proposed algorithm.

Keywords: Control Applications, Motion Control, Multi-agent Systems
Abstract: In this paper, we present a novel framework and control laws for the formation of multiple unmanned rotorcraft systems, be it single-rotor helicopters or multi-copters, with physical constraints and with inter-agent collision avoidance, in cluttered environments. The proposed technique consists of a closed-form analytical consensus control solution in the free space and an optimization-based motion planning protocol for inter-agent and obstacle collision avoidance. More specifically, a consensus control law is designed to tackle a series of state constraints, that is, the physical limitations of velocity, acceleration and jerk, and an optimization-based motion planning technique is utilized to generate solutions when the consensus control fails to provide a collision-free trajectory. The stability of the switching process between the consensus control and motion planning is guaranteed by a sufficiency condition. The effectiveness of the proposed technique is successfully demonstrated through both simulation and real flight experiments.

Keywords: Sensor Networks, Sensor/Data Fusion, Signal Processing
Abstract: Indoor localization and tracking plays a vital role in the IoT field for its various types of applications, e.g. healthcare system, pedestrian navigation, emergency rescue. Among all the available sensors, the combination of IMU and WiFi-RSS is the most widely used scheme for its low cost and pedestrian compatibility. However, conventional indoor localization systems cannot achieve satisfactory accuracy. To cope with this issue, in this paper, we propose a novel indoor localization system that integrates the IMU sensing and the RSS fingerprinting via a proposed PSO (Particle Swarm Optimization) based algorithm. We develop a new WiFi-enabled platform which can obtain RSS from captured wireless messages in the WiFi traffic. Based on this platform, the RSS fingerprints and the displacement estimated by IMU serve as the guidance and soft constraints for the particle swarm. A fitness metric is also proposed to evaluate the likelihood of each particle finding the real position by incorporating the GPR (Gaussian Process Regression) method. To improve the localization accuracy, we leverage the layout of the map, which provides information on accessible and inaccessible areas. Besides, image processing techniques including line detection and contour extraction are utilized to constrain the particles that may fall into the inaccessible areas. Experiments are conducted in a multiple function laboratory and the results verify that the proposed approach outperforms the typical approaches, and the achieved mean localization error is 0.705m.