Keywords: Automotive control, Fault tolerant control, Predictive control
Abstract: In the paper, a complex estimation and control algorithm is presented for autonomous vehicles. The algorithm has an observer layer that aims to estimate the slip of the vehicle on both the front and rear axles. The observer design is based on the combination of the Linear Parameter Varying (LPV) approach and the ultra-local model technique. The goal of the ultra-local model is to update the cornering stiffness in order for the lateral dynamics of the vehicle model to be updated. A Model Predictive Control (MPC) is developed based on the updated model and the estimated slips, that can maintain the lateral stability of the vehicle even under extreme circumstances, e.g., moving on a surface with a low adhesion coefficient. This stabilization is achieved by limiting the slips of the vehicle. The proposed algorithm is tested through a complex test scenario using a high-fidelity simulation software, CarMaker.

Keywords: Automotive control, Fuzzy logic and fuzzy control, Nonlinear control
Abstract: This article presents an observer-based control framework designed for four-wheel steering (4WS) vehicles, aimed at improving lateral stability and handling. Relying on active front/rear steering and direct yaw moment control, the system guides the vehicle to closely follow the desired lateral sideslip angle and yaw rate of the reference model. Different from previous approaches, this work takes into account longitudinal speed variations within a given range, adopting the Takagi-Sugeno (T-S) fuzzy method to deal with the resulting non-linearities and build an appropriate T-S fuzzy model. The design of the control and observer is based on the common Lyapunov function method, with stability conditions expressed as linear matrix inequalities (LMIs). Simulation results validate the effectiveness of the proposed method in achieving the main objectives of this study.

Keywords: Complex systems, Computational intelligence, Automotive control
Abstract: Network control theory (NCT) offers a robust analytical framework for understanding the influence of network topology on dynamic behaviors, enabling researchers to decipher how certain patterns of external control measures can steer system dynamics towards desired states. Distinguished from other structure-function methodologies, NCT's predictive capabilities can be coupled with deploying Graph Neural Networks (GNNs), which have demonstrated exceptional utility in various network-based learning tasks. However, the performance of GNNs heavily relies on the expressiveness of node features, and the lack of node features can greatly degrade their performance. Furthermore, many real-world systems may lack node-level information, posing a challenge for GNNs. To tackle this challenge, we introduce a novel approach, NCT-based Enhanced Feature Augmentation (NCT-EFA), that assimilates average controllability, along with other centrality indices, into the feature augmentation pipeline to enhance GNNs performance. Our evaluation of NCT-EFA, on six benchmark GNN models across two experimental setting—solely employing average controllability and in combination with additional centrality metrics—showcases an improved performance reaching as high as 11%. Our results demonstrate that incorporating NCT into feature enrichment can substantively extend the applicability and heighten the performance of GNNs in scenarios where node-level information is unavailable.

Keywords: Mechatronic systems, Modelling and simulation, Automotive control
Abstract: This paper summarizes the methodology for developing a Formula Student (FS) vehicle based on Systems Engineering principles. It showcases a condensed version of an automotive Development Process focused on Vehicle Dynamics and Suspension Design. It highlights key vehicle Goals, System, and subsystem level Requirements in the form of vehicle attributes and finally the Verification and Validation method. The intent is to offer a compass for teams and individuals building mechatronic prototypes (not limited to FS vehicles), where the tools and methods are not production ready and are on tight resource constraints (financial and time) as we normally see in FS competitions.

Keywords: Robotics, Robust control, Automotive control
Abstract: This paper introduces a novel Nussbaum function-based PID control for robotic manipulators. The integration of the Nussbaum function into the PID framework provides a solution with a simple structure that effectively tackles the challenge of unknown control directions. Stability is achieved through a combination of neural network-based estimation and Lyapunov analysis, facilitating automatic gain adjustment without the need for system dynamics. Our approach offers a gain determination with minimum parameter requirements, significantly reducing the complexity and enhancing the efficiency of robotic manipulator control. The paper guarantees that all signals within the closed-loop system remain bounded. Lastly, numerical simulations validate the theoretical framework, confirming the effectiveness of the proposed control strategy in enhancing robotic manipulator control.

Keywords: Automotive control, Mechatronic systems, Robotics
Abstract: In this paper, we present the design and the development of an active control system for adjusting the front wheels’ camber angle of an electric prototype vehicle according to the instant yaw rate measurement. The vehicle is equipped with a simple independent suspension structure, where each wheel is attached to the vehicle body through an upper and a lower arm. With the proposed mechanism in place, the upper arms of the front wheels can be translated horizontally, adjusting the camber angle of the front wheels. Conforming to the theory, the proper camber angle adjustment can reduce the camber thrust, improving the yaw rate and accordingly the lateral acceleration of the vehicle, which are directly associated with its stability and traction. Based on this hypothesis, simple circular maneuvers with constant longitudinal velocity and steering angle were designed for the mathematical proof of con- cept. In accordance with the experimental results, the proposed system, on average, improves the yaw rate by approximately 10%, enhancing the vehicle capacity to navigate corners at higher speeds.

Keywords: Feedback stabilization, Time-delay systems, Biologically inspired systems
Abstract: The issue of human balancing in the sagittal plane using fractional order time delayed acceleration feedback is studied in this contribution. The problem of asymptotic stability of closed-loop fractional order neutral time delay system is solved by applying the D - decomposition approach. Stability regions in the control parameters space are determined using this method. Special attention was focused on the consideration of the influence of the time delay on the asymptotic stability of the system. Finally, simulation results are presented to show the effectiveness of the proposed method.

Keywords: Marine control, Computational methods, Modelling and simulation
Abstract: The use of autonomous underwater vehicles (AUVs) for the real-time estimation of a plume generated by an underwater stationary or moving source finds important applications. A mathematical estimation framework is presented that includes the process-state (plume) model described by the 3D advection-diffusion partial differential equation and the exosystem (moving source) with a 6 DOF dynamical model and a guidance scheme. The process-state model is discretized using a non-overlapping domain decomposition finite volume method with Runge-Kutta integration. The source is following a physics-based prescribed trajectory with a PID controller. Numerical results show grid sensitivity analysis for the process-state under realistic underwater tidal currents and diffusivities. Plant data are shown for a typical case of a moving source in a large simulation domain and realistic ambient conditions.

Keywords: Marine control, Modelling and simulation
Abstract: Simulation can be useful to the marine engine designer in identifying necessary improvements. During the initial design of a complex engine, even inaccurate physical modelling can give important information to quickly change configuration and parameters, if necessary, without waiting for experimental tests. This paper considers supercharged diesel engines that are widely used in marine applications. It reports the preliminary design of the speed control system for one turbine of a more complex engine based on a sequential turbine architecture. The authors show how to improve a speed PID controller by using a simplified model to represent an engine turbine in simulation. An accurate physical representation is not required to choose the control scheme and set the parameters. Namely, the specifications on the minimum turbine speed and the limitation of speed oscillations are met, and performance is improved compared to a first control scheme.

Keywords: Nonlinear control, Nonlinear systems, Algebraic and geometric methods
Abstract: We consider a kinematic model of a controlled car with two trailers by assuming that each trailer is attached at some distance from the preceding axle ("off-hooked trailers"). For this model, we derive the transformation towards privileged coordinates and present the corresponding nilpotent quasihomogeneous approximate system. The components of this nilpotent approximation are written explicitly in terms of mechanical parameters of the original system. The constructed system does not satisfy the Brockett necessary stabilizability condition, and the design of time-varying feedback controllers with oscillating components is proposed. It is proved that these controllers ensure the exponential convergence of solutions to the trivial equilibrium, and simulation results are presented to illustrate the behavior of the closed-loop system.

Keywords: Nonlinear systems, Modelling and simulation, Real-time control
Abstract: Besides an efficient operation of a gas boiler, ensuring a safe combustion across the operating range is essential. This becomes even more critical when burning hydrogen instead of natural gas. The distinct characteristics of hydrogen and their effects on flame behavior necessitate stricter safety requirements. This paper investigates a novel method to contribute to the safe operation of a gas boiler by using control barrier functions (CBFs). Thereby, it demonstrates the potential of CBFs in the heating industry as a novel field of application. The system is represented by a Hammerstein model. In order to ensure the operation remains within a safe range, two CBFs and a high order control barrier function (HOCBF) are set up. The resulting safety filter is integrated into the combustion control loop and verified in simulation.

Keywords: Predictive control, Computational intelligence, Optimisation
Abstract: Atmospheric pressure plasma jets (APPJs) hold significant promise in biomedical applications, where safe and efficient operation is critical. In this study, a new data-driven robust control paradigm is proposed for APPJs in the Linear Parameter-varying (LPV) framework. By leveraging Bayesian Neural Networks (BNNs), a state space LPV model is identified to capture the intricate nonlinear dynamics of APPJs while providing statistical insights into the system's behavior. This approach allows for the adaptation of the uncertainty region at each time step, enhancing closed-loop control adaptability, and alleviating the conservativeness of the control design compared to the conventional robust controllers. The proposed robust Model Predictive Control (MPC) design method operates through an online process, where an optimization problem, formulated using Linear Matrix Inequalities (LMIs), computes a time-varying feedback control law. Through extensive simulations, the LPV-based robust control's efficacy in handling modeling discrepancies and external disturbances is assessed. Furthermore, a comparison is made with an alternative control scheme employing MPC with a given LTI model, demonstrating the superior robustness and tracking capabilities of the proposed LPV-MPC-based approach. These findings underscore the potential of the proposed technique to enhance APPJ control across diverse practical scenarios.

Keywords: Education and training, Modelling and simulation, Mechatronic systems
Abstract: Teaching mechatronic systems engineering should include theories as well as real world experiments. A rotating disc driven by two solenoids is presented as a new experimental setup. Mathematical equations for modelling of the system are given. Quantization and simplification are considered in the model. A validation of the model is carried out and checks the correctness but also the limitation of operation for the model. Finally, the use of the setup in education is presented.

Keywords: Integrated control and diagnostics, Education and training, Decentralized control
Abstract: The proliferation of technological advancements, coupled with the introduction of electric vehicles (EVs), has led to a substantial surge in battery utilization. Battery systems represent intricate multi-domain systems with a significant potential for hazards. Effectively mitigating these hazards necessitates rigorous supervision. This supervision is orchestrated by battery management systems (BMS) to ensure the safe and optimal operation of the battery systems. Design engineers must be educated on hazards, diagnostic, and fault protection solutions. The hazardous conditions inherent in testing or researching new battery technologies create an environment that is conducive to remote testing. The purpose of this paper is to demonstrate and discuss the development of a graphics-based operating battery management system instrument. This system can manage several battery packs in a safe operation and can be controlled and supervised via a remote human interface. It is used as an educational tool to demonstrate the operation of a battery management system with various functions such as unsafe operation diagnostic, controller failure, safe switch, battery charge, discharge, battery balancing, etc. These characteristics also designate it as a remote testing tool for hazardous environments). This remote-control system is constructed of various “off the shelf” laboratory and industrial equipment that is specialized in one of the building block functions. Thus, the paper demonstrates how different manufactured modular components that consist of a variable energy source, variable electronic load, and BMS were assembled, and then interconnected via one main operations controller, running one GUI.

Keywords: Education and training, Cyber-physical systems, Robotics
Abstract: In this paper we explore the use of a humanoid robot named LINA as an educational tutor in primary schools. The study focuses on understanding student acceptance and engagement with LINA within a formal classroom setting, examining their reactions to both the robot itself and the educational activities it facilitated. Utilizing a combination of questionnaires, we investigated students' perceptions of LINA's appearance, voice, and instructional methods. The results demonstrate a positive overall picture, with students expressing appreciation for LINA's physical design, clear communication, and interactive learning activities. Despite finding some tasks moderately challenging, they reported high levels of engagement and enjoyment. The findings highlight the potential of social robots in education but emphasizes the importance of considering design and social capabilities for effective implementation.

Keywords: Education and training
Abstract: STEM education is of great importance and al- though idely adopted in the last decades it still lacks affordable tools with extended capabilities. In this work we introduce an enhanced and significantly modified version of the system proposed by Kakkaras et al. in [1]. The new systems presents a low cost solution, where custom hardware is enclosed into functional shells designed and developed using a 3D printing approach. These shells are connected using a custom connection mechanism with enhanced capabilities, which allow the design and development of various working prototypes. The system is accompanied by a custom software adapted to the age, educa- tional level and skills of the end users, which in our case are students of primary and secondary education. An autonomous mobile robot was developed using the aforementioned approach to highlight the robustness and ease of use of our approach.

Keywords: Education and training, Embedded control systems, Robotics
Abstract: The increasing adoption of ROS-based control systems across diverse industries can be attributed to its middleware functionalities and open-source accessibility, prompting educational institutions to integrate comprehensive robot programming courses into their curricula. The students lack the necessary programming skills but there are plenty of tools such as MOOL (Massive Open Online Laboratories) and MOOC (Massive Open Online Courses) available which can support this. They still lack traditional hands-on learning experience which has proven to improve the pedagogical approach to teaching and learning. Programming with embedded systems has always piqued the interest of students to develop their hands-on skills and the learning experience. On the other hand, the embedded systems are cost-effective and highly energy efficient making it an affordable option to control most electrical and electronic devices. Since embedded systems are seeing rapid integration in industries, and ROS-based support to control the embedded systems, microROS (micro Robot Operating System) was introduced. This paper focuses on controlling a 4 DOF (Degree Of Freedom) self-made robotic arm using the microROS framework of ROS2, visualization performed through RViz and motion planning performed using the MoveIt framework. Overall this paper discusses various educational advantages and the control engineering aspects devised as an education module.

Keywords: Image processing, Neural networks, Computational intelligence
Abstract: Localisation and disposal of unexploded ordnance (UXO) is a crucial task that can save the lives of both military personnel and civilians. In comparison to immediate post-war intervention situations, EOD (Explosive Ordnance Disposal) teams can now leverage emerging technologies based on computer vision architectures, mitigating the perceived risks associated with hands-on inspection of ammunition. The paper analyzes the use of new convolutional neural network architectures, in detection and identification of unexploded ordnance by combining specialised domain knowledge with computer vision models and methods. Additionally, it presents image preprocessing methods, research techniques, results, and conclusions. Moreover, we propose a complementary approach to previous research, often based on the interpretation of external sensor signals, representing the missing link in a com- prehensive and extensive identification. Standardized metrics such as mean average precision, precision, recall, and F1-score are reported to evaluate the outcomes. Results, using the YOLOv8 architecture, achieve up to 80.8% mAP for the binary classification task (detection problem) and up to 81.4% mAP performance for the subsequent identification task. The work aims to offer a baseline study and new perspective on addressing a highly significant issue: mitigating by computer vision the risks associated with unexploded ordnance.

Keywords: Image processing, Signal processing
Abstract: In this work, an attempt is made for the first time to use the measurement pattern generated by morphological transformation quantified by Hausdorff fractal dimension (HFD) and classified with ensemble learning based on bagging. The proposed work uses three morphological transformations for image preprocessing: hit-and-miss transform (HMT), white (WHT), and black top-hat (BHT). The pattern texture of US breast images is described by extracting the HFD from the regions of interest (ROI) after the ultrasound (US) images have been preprocessed. The main objective of this study was achieved by comparatively analyzing the classification performance of features using the Random Forest (RF), Extra Trees (ET) classifier, and bagging ensemble method based on XGBoot classifier. In the presented study, the XGBoost classifier and BHT image processing method give an accuracy of 89.8% in a binary classification, benign versus malignant breast cancer.

Keywords: Neural networks, Image processing, Industrial automation, manufacturing
Abstract: The increasing challenges associated with waste management necessitate advanced technologies to enhance the efficiency of recycling processes. In this work, we present a comprehensive study utilizing the WaRP (Waste Recycling Plant) dataset for waste detection and classification within an industrial waste sorting plant. The dataset is available on Kaggle, and it encompasses 28 recyclable waste categories, including plastic and glass bottles, cardboard, detergents, canisters, and cans. Notably, objects in the dataset exhibit complexities such as overlap, deformation, and varying lighting conditions. For the classification task, we leverage the WaRP-C dataset, consisting of cut-out image areas from the WaRP-D set with class labels. Employing diverse deep learning architectures, including CNN, Le-Net5, AlexNet, VGG16, Mobile-Net-v2, Inception, and DenseNet, we explore the efficacy of each of them in accurately classifying the waste. For the task of waste detection, we focus on the WaRP-D dataset, comprising images with full HD resolution. The YOLO v8 architecture is employed for its efficiency in detecting waste items in cluttered and complex scenes. The dataset includes 2452 training images and 522 test images, each annotated with bounding boxes for precise object localization. The integration of both classification and detection tasks contributes to a holistic approach in automating waste sorting processes, fostering advancements in sustainable waste management practices.

Keywords: Neural networks, Image processing, Prognostics and diagnostics
Abstract: Face emotion recognition is an important subject in fields such as psychology and cognitive sciences, as well as in applications that use machine learning methods to create intelligent models capable of understanding human behavior. This paper addresses the challenge of negative facial emotion recognition by employing a comprehensive framework. Our approach leverages 68 facial landmarks and integrates three pre-trained convolutional neural network models (VGG19, DenseNet201, Inception V3) into a unified weighted voting system. The efficacy of our system is evaluated using the Extended Cohn-Kanade (CK+) and Karolinska Directed Emotional Faces (KDEF) databases. The results showcase the effectiveness of our approach in accurately identifying negative facial emotions. This research contributes to the advancement of emotion recognition systems, offering a promising methodology with implications for fields such as psychology, cognitive sciences, and the broader spectrum of intelligent systems reliant on facial expression analysis.

Keywords: Robotics, Energy efficient systems, Algebraic and geometric methods
Abstract: The problem of efficient motion planning is of significant interest in precision agriculture. We propose a mechanism to estimate it via Bezier function parametrization of the position, velocity and acceleration profiles for a quadcopter system. This is done by providing closed-form descriptions of the cost components, which are further incorporated into the overall motion planning problem to analyze the effect of various tuning parameters (such as total flight time, maximum velocity and acceleration). The ideas are tested over a proof-of-concept orchard navigation setup (challenging due to the close grouping of the way-points) with the final goal being thee comparison between horizontal-first and vertical-first path construction.

Keywords: Aerospace control
Abstract: In this paper, we introduce a trajectory planning method based on a modification of the well-known Dubins path. The goal is to ensure the interception of a stationary target at both a desired impact or approach angle and a desired impact or terminal time. We satisfy these constraints of impact time and angle simultaneously by appropriately selecting the radius of a relevant circle. This work outlines the range of values over which this radius (design parameter)may vary, highlighting its dependence on the initial engagement geometry. The simplicity of the proposed strategy lies in the requirement for constant levels of lateral acceleration to be applied over three distinct time intervals during the engagement. Additionally, we present a closed-loop implementable version of the control law, and simulation results demonstrate the superiority of our proposed strategy over some existing methods.

Keywords: Aerospace control, Nonlinear control, Robust control
Abstract: This paper considers robust Sliding Mode Control (SMC) methods for spacecraft attitude maneuvers. The quaternions representation is used to describe the spacecraft attitude dynamics, while the actuation system consists of a cluster of four reaction wheels in pyramidal configuration. The saturation values of the cluster are used to design the control law, which is based on the Super-Twisting (STW) SMC algorithm. A novel strategy is proposed to design the parameters of the STW control law so that it does not overload the actuator limits while maneuvering the spacecraft. The proposed control strategy is validated through extensive Monte Carlo simulations, aimed at testing the robustness of the controller under a wide range of initial conditions, external disturbances and values of system parameters.

Keywords: Aerospace control, System identification, Neural networks
Abstract: An on-line aerodynamic parameters identification method is proposed based on improved Physics-Informed Neural Network (PINN) to address aerodynamic parameters error problem during flight control. An integration-based loss function is utilized to ensure that the neural network can learn the correct physical equation information, and adopts a parallel neural network architecture to reduce network complexity. To ensure the feasibility of the network, the input and output data are measurable by the Integrated Navigation System. The improved PINNs is used to identify the aerodynamic parameters of the Reentry Gliding Vehicle in numerical simulation. Simulation results demonstrate that the network can effectively identify aerodynamic parameters during the flight process and the proposed method is insensitive to measurement noise. The proposed method can provide information for the design of multi constraints guidance laws for flight vehicle.

Keywords: Distributed systems, Neural networks, Aerospace control
Abstract: This paper studies the problem of three-dimensional (3-D) cooperative guidance for speed-varying interceptors. A 3-D cooperative guidance strategy is designed by adding a cooperative term to proportional navigation guidance (PNG). First, a deep neural network (DNN) is constructed for predicting the time-to-go of the speed-varying interceptor under PNG. Then, a cooperative term is derived on the basis of predicted time-to-go information. Moreover, the time-to-go consensus error system is proven to be input-to-state stable (ISS) under the designed guidance law. Finally, numerical simulations are conducted to illustrate the validity of the proposed method.

Keywords: Guidance, Aerospace control, Nonlinear control
Abstract: Aiming to cope with the problem of impact-time control (ITC) for engaging maneuvering targets, a nonlinear optimal guidance method is proposed in this paper. Utilizing the classical differential geometric curve theory and the relative virtual frame, the ITC is realized by the combination of the relative trajectory-length control and the relative speed prediction. The characteristics of the ideal proportional navigation (IPN) guidance law is fully leveraged in this paper: 1) with the series solution of the relative trajectory length-to-go under IPN, the guidance law is derived using the optimal error dynamic method; 2) to provide the real-time and stable estimation of the future relative speed profile, the implementation framework of ITC is developed based on IPN. Besides, the linear solutions of the relative lead angle and the relative guidance curvature are derived, which are compared with the exact numerical solutions to explore the properties of the proposed law. Finally, numerical simulations are conducted to substantiate the effectiveness of guidance method and the validity of relevant theoretical findings.

Keywords: Navigation, Linear systems, Aerospace control
Abstract: Derivations of the Kalman filter that ensure optimality with respect to minimum variance typically assume that the process and measurement noise are uncorrelated. However, several works have presented generalized formulations of the Kalman filter where correlated measurement and process noise is permitted. This paper revisits the problem of correlated noise in the context of initialization and leveling for inertial navigation systems. We show how the discretized process and measurement noise covariance matrices can be calculated and applied in a discrete-time Kalman filter. These results are used in an error-state extended Kalman filter implementation which is experimentally tested using a data set from a fixed-wing unmanned aerial vehicle flight together with a simulation study. The simulation and experimental results both indicate that formally addressing the process and the measurement noise correlation during inertial navigation system initialization has little to no impact in our use case.

Keywords: Automotive control, Navigation, Robotics
Abstract: This paper investigates the guidance issue of the two-dimensional target following of a general robot with only discrete waypoints information of the target. In particular, a velocity guidance strategy is proposed by introducing an artificial vector field. First, thin plate spline interpolation is employed to fit the waypoints into a continuous curve in real time. Due to the spatial relationship between the robot and the target, a guiding vector field and an additional vector field are proposed separately to ensure the convergence of the robot onto the target curve. Finally, the velocity signal is obtained by combine the two vector fields to an artificial vector field. Simulation results show that the proposed strategy achieves target following with ultimate bound.

Keywords: Automotive control, Time-delay systems, Navigation
Abstract: The remote control of automated vehicles introduces various challenges, with communication latency being a prominent concern. The mitigation of this latency is crucial for ensuring safe and effective mobility. Within this context, this paper aims to address the impact of communication latency by introducing the Vehicle Dynamics Estimator (VDE). This approach is designed to enhance the accuracy of vehicle trajectory estimation and contribute to the overall improvement of remote control systems for automated vehicles. To validate the effectiveness of the proposed approach, comprehensive simulations have been conducted. The results of these simulations demonstrate the successful enhancement of vehicle trajectory estimation through the application of the VDE.

Keywords: Robotics, Navigation, Nonlinear control
Abstract: Navigation functions combine geometric path planning and control of mobile robots in the presence of obstacles. This paper extends the theory of mobile robot navigation functions by proposing a navigation function feedback control law that ensures safe navigation of a point mass robot under bounded control inputs. Simulations describe the robot state variables navigating under bounded control in planar environments populated by disc obstacles.

Keywords: Autonomous systems, Modelling and simulation, Navigation
Abstract: As the world of transportation goes towards fundamental transformations, the urban mobility experiences the dawn of autonomous electric vehicles. Under this scope we aim to develop a framework which will facilitate research around digital twins, autonomous agents and efficient electrified transportation. This effort is build around a two seat compact electric vehicle, which is currently getting modified to act as an autonomous platform. The current work has a twofold purpose. Initially to describe the simulated approach towards building a realistic model of the proposed vehicle, which will allow realistic experimentation. Subsequently, we validate our approach by implementing a two step procedure towards an autonomous agent, which will be able to follow a predefined path in an urban environment. Initially we use a standard path planner, namely the A∗ algorithm, and once we have the desired trajectory we train an autonomous agent to perform the navigation task. The whole process is integrated in CARLA simulator by using a tailored GYM environment.

Keywords: Robotics, Autonomous systems, Navigation
Abstract: This paper presents a robust control strategy for point-to-point navigation of mobile robotic systems under prescribed time specifications and with state and input constraints. Based on the concept of Nonlinear Model Predictive Control (NMPC), we propose a stable dynamically decreasing horizon robust optimal control strategy that achieves the aforementioned task while considering state and input constraints as well as system uncertainties and disturbances. We guarantee that, under the proposed control scheme, the mobile robotic system reaches the desired set in the specific given time, provided an upper bound of the disturbances. This can have great applications in collaborative tasks including a team of mobile robots, where in addition to mobility and flexibility, the mobile robots are required to be enhanced individually with abilities to satisfy time-constrained navigation tasks, in order to fulfill high-level planning specifications. Finally, simulation results verify the performance of the proposed framework.

Keywords: Intelligent transportation systems, Robotics, Navigation
Abstract: The advent of autonomous driving technology highlights the crucial need for reliable and real-time localization, which is essential for ensuring vehicle safety and efficient driving. Techniques relying solely on traditional sensors like GNSS and IMU often present significant challenges in localization accuracy. Moreover, LiDAR-based point cloud localization methods demand extensive storage and computational resources. To address these issues, this paper introduces a map-based localization system for autonomous vehicles that integrates LiDAR, GNSS, and IMU data within an Extended Kalman Filter framework together with sparse map-based features, offering a more efficient and robust solution for vehicle positioning and navigation. Real-world experiments and comparison results demonstrate the efficiency of our proposed method in urban and suburban environments.

Keywords: Computational intelligence, Optimisation, System identification
Abstract: We introduce a constructive function approximation approach as a general tool, particularly useful in adaptive and data-driven methods for perception and control. The key idea is to estimate of a collection of simple local models as opposed to a single and complex regression model trained in the entire input space. We use principles from the Online Deterministic Annealing (ODA) optimization framework to construct an adaptive partition of the input space, which enables the introduction of local function approximation models within each subset of the partition. We show that both the partitioning and the local model training algorithms are stochastic approximation algorithms that operate online, and with the same observations, as part of a two-timescale stochastic approximation scheme. This process constitutes a heuristic method to gradually increase the complexity of the function approximation framework in a task-agnostic manner, giving emphasis to regions of the input space where the regression error is high. As a result this framework has inherent explainability properties, and is suitable for continuous learning applications where regression improvement without re-training from scratch is crucial. Simulation results illustrate the properties of the proposed approach.

Keywords: Computational methods, Optimisation
Abstract: The problem of global optimization of an objective function represented by a Random Forest (RF) is considered. A method to obtain an approximate solution at low computational complexity is proposed, resorting to the inherent structure of an RF, which is a non-parametric model that partitions the feature space in convex polytopes according to the training data. The approach selects the optimal solution inside the polytopes corresponding to the best data points. It is shown that the proposed approximate method is significantly more efficient, thus applicable at large scale, than extensive global search algorithms, such as gridding and Mixed Integer Linear Programming (MILP), which in turn provide exact solutions. The efficiency and sub-optimality of the approach are evaluated on RFs trained on a dataset generated by sampling a bivariate, discontinuous and non-convex benchmark function from the literature.

Keywords: Game theory, Linear systems, Optimisation
Abstract: An algorithm that locally converges to all the feedback Nash equilibria (F-NE) of a two-player, linear-quadratic (LQ) differential game is proposed. This objective is pursued by modifying classic Lyapunov iterations, which naturally arise by a straightforward interpretation of the coupled Algebraic Riccati Equations and which converge only to a subset of all the F-NE, so to enforce local asymptotic stability of all the F-NE of the underlying differential game.

Keywords: Linear systems, Optimisation
Abstract: In control engineering, effective implementations require to take into account the constraints for both the inputs and the outputs of the controlled system. A feedforward regulation problem is then set out to achieve a minimum-time rest-to-rest output transition for square MIMO (multi-input multi-output) linear time-invariant systems. The found time-optimal solution extends the well-known bang-bang control to a generalized bang-bang control in which some amplitude constraints on the outputs are active. A straightforward sufficient condition to have a solution is provided. The structure of the time-optimal input-output pair is also investigated. A feasibility linear programming approach to compute an approximate solution is proposed. An example highlights the found results.

Keywords: Nonlinear systems, Nonlinear control, Optimisation
Abstract: This paper presents a novel methodology for the synthesis of state observers for a class of systems subject to rational nonlinearities. The key idea is to represent both the system and the observer in the so-called differential-algebraic representation in order to obtain the error dynamics. Once achieved, design conditions can be formulated through optimization problems subject to constraints in the form of linear matrix inequalities. This approach enables the development of a systematic observer design framework with formal theoretical guarantees, including the asymptotic stabilization of the error within a specified exponential decay rate along with constraints on the observer gains. A post-synthesis analysis is presented to ensure the stability of the plant considering the feedback of the estimated states. An illustrative example is provided to showcase the proposed method.

Keywords: Predictive control, Optimisation, Robotics
Abstract: We introduce an alternative approach towards optimal proportional integral derivative (PID) control, consisting of model predictive control (MPC) based reference generation. To this end, we have integrated the reference as part of optimization variables of the resulting problem, where a deliberate sequence of errors is induced to obtain an optimal PID control action. In addition, the desired behavior of the PID controller is achieved without the need for internal modification of the PID gains. To better highlight the ability of coping with poor PID tuning, several test cases consisting of progressively degraded PID gains are presented. Validation of the proposed strategy is displayed by comprehensive simulations using two different plants.

Keywords: Event based systems, Process control
Abstract: This paper proposes a solution for controlling a gas flow rate using a diaphragm gas meter as a measuring device that emits a pulse after a certain volume of gas has passed through it, instead of relying on continuous flow rate measurement. The proposed method combines a continuous-time predictor with an event-triggered corrector, taking into account that the interval at which measurements are received is significantly larger than the time constant of the process. The presented algorithm demonstrates the ability to achieve offset-free reference tracking at steady state, as supported by both simulation and experimental results. Additionally, a stability proof is provided.

Keywords: Fault diagnosis, Prognostics and diagnostics
Abstract: This paper presents a fault prognosis approach using data-driven structurally generated residuals. The proposed approach assumes that a set of residuals generated using structural analysis (SA) and identified using data-driven approach are available. Residuals are used for fault detection purposes activating fault signals when residual values reach anomalous values. In addition, it is possible to predict future faults by means the detection of anomalous residual deviations. Once an anomalous change in the residual trend has been detected, it is proceed to estimate when this residual deviation will result in a fault detection and therefore what will be the Remaining Useful Life (RUL) time of the system. For this purpose, the future residual evolution is estimated by means of a regressor function. Nominal and interval parameters of regressor function are estimated with available residual data providing nominal an inteval values of the RUL of the system. A brushless direct current (BLDC) motor is used as the application case study to illustrate the performance of proposed method.

Keywords: Fault tolerant control, Fuzzy logic and fuzzy control, Nonlinear control
Abstract: This study focuses on the design of H∞ finite frequency (FF) fault tolerant (FF) static output feedback (SOF) problem of Diesel engine air-path system with consideration of actuator faults. The state space model of air-path system, firstly, is linearized around average an operating point of the Diesel engine. The aim is to regulate intake and exhaust manifold pressures to the desired reference pressures by controlling the Geometry Turbine (VGT) and Exhaust Gas Recirculation (EGR) valves. Then, the robust integrator-based control strategy is developed to track the desired reference signals despite the presence of disturbances and actuator faults. By constructing the Generalized Kalman Yakubovich Popov GKYP lemma, projection lemma and Lyapunov functions, sufficient conditions are established to ensure both the reference signal tracking and H∞ performance with FF domain. To corroborate the theoretical findings, a simulation results is employed, demonstrating the applicability of the proposed theorems.

Keywords: Multi-agent systems, Distributed systems, Event based systems
Abstract: This paper addresses the problem of measurement noise in event-triggered distributed control of multiagent systems. In the considered setup, each agent receives the state information from neighboring agents through a fixed, connected, and undirected graph, and the state of each agent is corrupted by measurement noise. To address this challenge, new and novel distributed control architectures predicated on local state observers are proposed in a leader-follower setting for both the infinite and finite horizon cases. The goal of these architectures is to effectively mitigate the effect of measurement noise while reducing agent-to-agent information exchange through event-triggering. The system-theoretical findings for both the infinite and finite horizon cases are presented, which are also numerically validated by two illustrative examples.

Keywords: Power systems and smart grid, Energy efficient systems, Renewable energy and sustainability
Abstract: Power system operators incentivize large consumers of electricity to reduce their demand during peak periods, where consumers can promise a fixed demand reduction in the form of a contract known as demand response (DR). Commercial building owners, particularly those who have electric vehicle (EV) charging infrastructure, can utilize vehicle to building (V2B) technology to participate in DR, while also aggregating other on-site distributed energy resources (DERs) such as solar and batteries. However, the provision of V2B-enabled DR creates trust issues between EV owners, building owners, and power system operators in ensuring that the contracted DR capacity is delivered, and that operational preferences, such as the minimum state of charge of EVs participating, is respected. Thus, this paper proposes a blockchain-based system to create a shared ledger that stores contract details, including contract bids and EV state of charge, and executes automated smart contracts to record, monitor, and dispatch the DERs to ensure that the contracted capacity is realized without violating constraints. Through real-world scalability testing and DR event demonstration, the proposed system shows that it can support up to 10,000 DERs and deliver 20 kW of contracted capacity for a 4 hour DR event.

Keywords: Prognostics and diagnostics, Computational intelligence, Industrial automation, manufacturing
Abstract: Anomaly Detection is a task in engineering aiming at identifying deviations from expected patterns in data. Data-driven approaches have emerged in past recent years due to the fact that a model of complex system may be hard or impossible to be derived in many scenarios. Moreover, unsupervised approaches have been particularly appealing for practitioners and scientists given the typical unavailability of tagged data.

Such approaches are often integrated in frameworks, like Decision Support Systems, that assist domain experts and operators in the monitoring task. Human presence, by providing a limited amount of feedback, can be leveraged as a valuable source of information to iteratively enhance detection performance. In this work we introduce Extended B-ALIF, a framework designed to incrementally select and integrate expert feedback into the Extended Isolation Forest anomaly detection model. This study extends B-ALIF, which originally proposed the same theoretical principles for another anomaly detection model, the Isolation Forest.

Keywords: Intelligent control systems, Neural networks, Image processing
Abstract: With the rise in frequency of catastrophic events, enviromental protection and risk management have become critical challenges for assuring both the safety of human populations and the sustainability of ecosystems. In this direction, the ARIES project has deveoped a suite of functionalities and software solution to support emergency operators in all phases of emergency management, from risk estimation to enviromental monitoring, early detection and response. The present paper provides an overview of the main functionalities that consistute the buidling blocks of the enviromental monitoring system developed by the project through a combination of dynamical systems, digital twins and convolutional neural networks, also reporting on its early validation activities.

Keywords: Neural networks, Computational intelligence, Image processing
Abstract: Preserving fish welfare is of major priority for fish farming companies and essential for the sustainability and future growth of the aquaculture industry. Wounds represent a serious welfare issue for the fish, both as a symptom of existing problems and a precursor to potential future vulnerabilities. Existing static or moving vision-based sensor systems paired with computer vision methods offer promise for monitoring different welfare indicators but have not yet been well adapted for wound detection. Targeting this challenge, this paper introduces a computer vision framework that leverages object detection and tracking algorithms to automate wound detection and tracking in video recordings. Tested on unseen video images from operational salmon fish farms, the model achieves 90% accuracy and showcases enhanced robustness in highly complex and dynamic scenarios due to the added tracking capabilities. Currently an assistive tool for detecting wounds during manual camera inspections, the framework has the potential for future development into a fully autonomous, data-driven approach to fish welfare monitoring.

Keywords: Neural networks, Image processing
Abstract: In this study, we address the challenge of text detection and recognition in low-resolution images. Addressing the dual constraints of limited training data and the necessity for real-time processing, we adopt a twofold strategy. Firstly, we introduce a pipeline for the generation of synthetic datasets that requires minimal manual annotation and is specifically designed for the context of low-resolution real-life text images. Secondly, we employ a streamlined neural network architecture based on the U-Net model, which concurrently executes text detection and recognition across multiple contextual layers. Our method demonstrates superior performance, achieving real-time processing speeds exceeding 120 fps, and is accurate even in challenging conditions where text characters are as small as five pixels in width. Our findings suggest that both the synthetic dataset generation pipeline and the neural network model are highly adaptable and can be easily modified for a broad range of applications.

Keywords: Process control, Image processing, Intelligent control systems
Abstract: Monitoring of the complex cold atmospheric plasma spraying process currently relies on technologically advanced systems, that often measure only a single parameter. This research introduces a more integrated approach, that uses a comprehensive set of parameters, consolidated in the characteristics of the plasma flame, is used to accurately classify particle temperature and coating materials. A simple optical camera system and image analysis based on a transfer learning model predicts with an 85% success rate. This approach simplifies and complements the monitoring process. Accuracy could be increased by using additional machine data already recorded, which could allow a complete replacement of additional the plasma flame monitoring.

Keywords: Robotics, Image processing, Autonomous systems
Abstract: Abstract— This study presents a novel approach to achieve sub-millimetric localization in intralogistic robotic applications by combining the simplicity of ArUco code detection with the enhanced accuracy of pose estimation using a structured light 3D camera. By evaluating the point cloud generated by the depth camera, the cartesian pose of the ArUco code can be directly extracted, eliminating the need for prior knowledge about camera intrinsics or calibration. This method enhances the accuracy of relative robot localization within warehouse environments up to 0.1 mm at a 1 m distance. To the best of our knowledge, this accuracy has not been reported in any previous work in the literature.

Keywords: Algebraic and geometric methods, Fault tolerant control, Linear systems
Abstract: We consider the problem of control allocation for weakly redundant systems subject to actuator faults. In particular, the design of a suitable allocator will be devised with the aim of compensating for the fault effects while, at the same time, keeping the control burden as low as possible. To this goal, two approaches can be followed for the synthesis of the allocation servomechanism: direct allocation using orthogonal projection and optimization-based allocation. Some numerical examples illustrate and highlight advantages, disadvantages and limitations of both strategies.

Keywords: Cyber-physical systems, Distributed systems, Linear systems
Abstract: Distributed control architectures are attractive for large-scale interconnected systems as they provide good trade-offs between control complexity and closed-loop performance. In such a context, it becomes crucial to ensure robustness against variations in the communication topology arising from connectivity failures or cyber-attacks. Based on Linear Matrix Inequalities, this paper introduces a novel design approach for distributed controllers that exhibit robustness to changes in the communication topology. The primary objective is to achieve resilience against potential disconnections of subsystems that may occur within the communication network. The control gains are structured, and reflect the nominal communication topology. Exponential stability with a prescribed decay rate is guaranteed within the sub-configurations of the nominal communication topology. The proposed approach does not make any assumptions regarding the connectivity of the communication graph. Moreover, leveraging on the projection lemma, the design of control gains is decoupled from the design of the Lyapunov matrix, thus minimizing the conservativeness of the solution. An illustrative example on a multimachine power system is presented to demonstrate the effectiveness of the proposed approach.

Keywords: Mechatronic systems, Feedback stabilization, Linear systems
Abstract: Observer design typically demands a comprehensive understanding of the system’s model, necessitating meticulous parameter tuning for each state. In this paper, we introduce an observer design tailored to a class of linear systems, mitigating the need for extensive tuning and intricate model knowledge. Our proposed observer integrates steady-state behavior estimation with a dynamic state-based estimation, effectively streamlining the design process. To illustrate its effectiveness, we apply this observer design to the control of running gears with independently rotating and driven wheels within the research project ’Next Generation Train’. This approach demonstrates a significant reduction in the requisite tuning parameters and model intricacy, offering a more practical and efficient solution for control systems.

Keywords: Modelling and simulation, Predictive control, Linear systems
Abstract: In this paper, we consider the problem of data-driven modelling and model predictive control (MPC) of a single machine infinite bus system (SMIB). When a linear state-space model of the plant is not available, a pair of Loewner matrices is constructed from a set of measured input-output data and a set of prescribed poles. We prove that the Loewner matrices are the solutions of two Sylvester equations. We compute the unique linear model of an order equal to the dimension of the data sets, that interpolates the input-output data and that has all the poles prescribed at the selected locations. As an application we perform MPC on both the original system and its approximant. The simulation show that the linear data-driven model is a good approximation of the SMIB and achieves good closed-loop MPC performance.

Keywords: Time-delay systems, Linear systems, Feedback stabilization
Abstract: This paper aims to explore the advantages and disadvantages associated with implementing derivative action in a PID controller using a delay-difference approximation scheme in continuous time. Despite the simplicity of the approximation, the characteristic function of the closed-loop system is a quasipolynomial including delay-dependent coefficients, which results in a challenging task. The proposed study primarily focuses on the way the controller gains and the delay affect the key characteristics of the dynamics such as asymptotic stability, noise rejection, and closed-loop performance for linear-time-invariant (LTI) minimum-phase second-order systems. Additionally, the paper provides some results concerning the delay margin of the delay-difference approximation of the closed-loop system. Finally, the proposed results are illustrated via numerical examples.

Keywords: Power systems and smart grid
Abstract: This article addresses the problem of robust stabilisation of uncertain systems with linear nominal part suffering from perturbations. The proposed composite controller, consisting of a linear state feedback (LQR) and a nonlinear Integral Sliding Mode Control (iSMC), guarantees that despite the presence of unmatched perturbations the closed loop system enters the sliding mode in finite time. The matched perturbations are rejected while the unmatched ones are attenuated in a quantifiable way. The suggested controller is applied on the Load Frequency Control (LFC) of power systems with two and three fully interconnected control areas. Problem specific tuning rules are offered enabling the stabilizing controller to handle also partial stability objectives. The numerical examples corroborate the applicability of the proposed nonlinear methodology.