Keywords: Iterative learning control, System identification and modelling
Abstract: The future of manufacturing equipment and scientific instruments hinges on the ability to perform precise and fast motions. Examples of such mechatronic systems include wafer scanners, printing systems, pick-and-place machines, microscopes, and telescopes. These systems are subject to ever-increasing speed, accuracy, and flexibility requirements. Learning from data provides major opportunities to meet these requirements. In this presentation, I will outline new identification, learning, and control methodologies that can deal with the increasing requirements and large complexity in envisaged future mechatronic systems. I will also show their successful implementation on a selection of state-of-the-art mechatronic systems. The new results pave the way for new, revolutionary, data-intensive mechatronic designs with a massive number of actuators and sensors.

Keywords: Artificial intelligence methods for control, Renewable energy and smart grids, Systems engineering
Abstract: This paper presents a model-free control approach using deep reinforcement learning(DRL) which aims at optimizing the performance of wave energy converter(WEC). Several researchers commonly utilize a linear WEC model, neglecting the inherent non-linear static friction of the WEC system. This oversight may result in a decline in control performance. Consequently, static friction is deliberately included as a modeling error in the control framework to compare the resilience between model-based reactive control (RC) and DRL method against disturbances. To further address the challenge of real-time wave information being difficult to acquire in practical applications of the DRL method, a novel DRL controller is proposed with no wave information. The simulations are conducted to compare the proposed DRL controller with the RC approach based on a nonlinear WEC system in MATLAB/Simulink. It is shown that when considering modeling errors of static friction force, the DRL-based controller can outperform reactive control in both regular and irregular wave conditions - by 20.0% and 6.2%. Preliminary experimental validation of the DRL controller has been obtained through wave tank testing in regular wave conditions, showing satisfactory performance and consistency with simulation results. This is the first successful practice of running the RL algorithm on a practical controller and proves the feasibility of the proposed approach.

Keywords: Artificial intelligence methods for control, System identification and modelling, Renewable energy and smart grids
Abstract: Genetic algorithms use the ideas of Darwinian evolutionary theory to find the optimal solution to a design problem. Here they are utilised in two scenarios. Firstly, finding the optimal power take-off (PTO) force for maximising the electrical power output of a device by accounting for PTO efficiencies. The genetic algorithm finds a solution marginally faster than a brute forcing method with the added benefit of not being constrained to a discrete grid of test points, hypothetically leading to a more accurate result.

Secondly, these optimal power take-off forces are used with another genetic algorithm to fit a transfer function for use as part of a previously designed adapted optimal velocity tracking controller that accounts for PTO efficiencies. Along with the reduced requirement for control engineering expertise, the resultant transfer function is found to have a smaller average phase error, when compared to a manually fitted transfer function. Simulations are undertaken that find that using a genetic algorithm derived transfer function results in approximately the same, or better energy capture when compared to the manually fitted transfer function, depending on the sea state, with the largest improvement being an increase of 5.93%. These methods form the basis of a potential control co-design methodology.

Keywords: Energy and Power systems, Predictive control, Artificial intelligence methods for control
Abstract: To harness maximal wave energy, control and optimization for wave energy converters(WECs) have been investigated for decades.It has been long recognized that WEC control is essentially a non-causal control problem, in which future wave determines current control decisions. This paper introduces a double deep Q network into the foundation of the non-causal time variant PD control system, enabling real-time parameter adjustments for dynamic control responses. Additionally, this paper delves into a comparative assessment of the influence of different prediction horizons on the efficiency of energy harvesting. The primary objective of this study is to elevate the control performance of wave energy converters, facilitating more efficient capture and conversion of wave energy into usable electrical power. The integration of deep reinforcement learning empowers researchers to adapt swiftly to fluctuating waves and ocean conditions, fine-tuning control parameters to enhance overall system efficiency and stability. Taking the point absorber as an example, the effectiveness of the proposed method has been verified. This method can be straightforwardly applied to other types of WEC, such as Dielectric Elastomer Generators and Dielectric Fluid Generators.

Keywords: Predictive control, Renewable energy and smart grids, Energy and Power systems
Abstract: Combining multiple direct-drive wave energy converters (WECs) into WEC arrays can effectively improve power generation and energy utilization, which is a promising wave energy utilization scheme. However, unlike wind and photovoltaic systems, the output power of WEC arrays fluctuates frequently and varies greatly, which can easily threaten the reliability and security of power grids. Therefore, it is necessary to use energy storage systems (ESSs) with long cycle life and high power density to realize stable power output. To this end, this paper employs supercapacitors (SCs) to smooth output power fluctuations of WEC arrays and proposes a model predictive control (MPC) strategy aiming at minimum power output and optimal state of charge (SOC) of supercapacitors. To characterize the power output requirements of supercapacitors under different SOCs, this paper introduces a weight factor for the SOC increment in the MPC objective function. By dynamically adjusting the weight factor during rolling horizon optimization, the self-recovery of the supercapacitor SOC can be ensured while the output power of supercapacitors is as small as possible. Finally, a simulation model is built in MATLAB/Simulink. Compared with the low-pass filtering (LPF) strategy, the MPC strategy proposed in this paper can obtain significantly smoother grid-connected power with smaller supercapacitor throughput, and the supercapacitor SOC can be stabilized within the target range.

Keywords: Renewable energy and smart grids, Autonomous systems
Abstract: This paper introduces an innovative autooptimisation control framework for wave energy converters (WECs) where the concept of dual control for exploration and exploitation (DCEE) is employed to effectively address this challenge in the realm of WECs. The control problem for WECs is characterised by its dynamic and unpredictable nature, demanding strong adaptivity and robustness based on wave predictions. A sophisticated automatic control framework is proposed that transforms the inherently periodic WEC control problem into an optimal operational parameter search problem. A DCEE approach is developed to optimally search the best operational condition through trading off between exploitation and exploration. More specifically, the DCEE approach contributes to the reduction of belief uncertainty in the identification of wave parameters, which is achieved by actively exploring the operating environment. It also facilitates the tracking of optimal operational conditions for power takeoff force. Simulation results validates the effectiveness of this novel framework featuring the DCEE approach.

Keywords: Renewable energy and smart grids, Predictive control, Distributed and decentralised control
Abstract: Wave Energy Converters (WECs) deployed in an array can reduce investment and operational costs, such as infrastructures, operations, and maintenance. Control for WEC arrays becomes more complex due to the dynamic behaviour of individual WECs and the interactions between neighbouring devices, including their defective and radiating effects, which impact the overall system performance. This paper investigates nonlinear model predictive control (NMPC) approaches for controlling an array of wave energy converters (WECs) to extract the maximum amount of energy while respecting safety constraints. Three schemes are benchmarked, namely: (i) Centralised NMPC, where the array is considered as an augmented system; (ii) Independent NMPC, where each WEC is considered independently, neglecting all interactive effects; and (iii) Distributed NMPC (D-NMPC), where a WEC only considers the interactive effects of its nearest neighbour. The proposed approaches are demonstrated and benchmarked through comparative simulation studies based on an array of homogeneous point absorbers. Simulation results highlight the limitations of the independent and centralised approaches and reveal the potential of D-NMPC in achieving most of the performance of C-NMPC, with a computational burden similar in scale to D-NMPC.

Keywords: Active noise and vibration control, Embedded systems, Linear control design
Abstract: An alternative method for the determination of a transfer function of a single-input single-output mechanical system is given, illustrated with a model of a train suspension system taken from [4], and its application in H2 norm analysis following [6]. This example is demonstrated with a MATLAB workbook, which follows the example step-by-step, calculating the H2 norm from this transfer function numerically and symbolically. Building on this, this method is then embedded within an interior point optimisation scheme within MATLAB, demonstrating how a workflow using these methods may look, and demonstrating a speed up factor of around 20 times. Computation time is given and compared to existing methods built in to MATLAB. The associated code to reproduce all results in this paper can be found on Github [5].

Keywords: Linear control design, Cyber-physical systems, Linear control theory
Abstract: This paper presents conditions in the form of Linear Matrix Inequalities to design state-feedback controllers for discrete-time uncertain positive linear systems in the presence of denial of service (DoS) attacks. The system under attack is modeled after a switched model that allows us to derive the synthesis conditions. Two strategies for control are considered, a hold strategy and a packet-based approach. Numerical experiments illustrate the efficacy of the proposed method of keeping the positiveness and stability of the closed-loop system under the presence of DoS attacks.

Keywords: Linear control design, Linear control theory
Abstract: In this paper, we present an event-based control technique for discrete-time linear parameter-varying (LPV) systems by employing an emulation approach. A new condition to design an event-triggering mechanism is proposed, and an optimisation procedure is presented to minimise the number of events. The conditions are formulated in the form of parameter-dependent linear matrix inequalities. A polynomial parameter-dependent controller is employed to stabilize the system. The impact of increasing the polynomial degree of the controller matrices on the number of events is investigated. The effectiveness of the method is illustrated through a numerical example.

Keywords: Linear control theory, Process control systems, Linear control design
Abstract: Slot coating is a widely used technique in various manufacturing processes. The efficiency of the process mostly depends on the precise control of diverse inputs such as pump rate and gap. However, this method can be sensitive to minor disturbances in process conditions, which can disrupt the uniformity of the film thickness. To overcome the challenge, through the implementation of the generalized disturbance observer and compensator, the impact of disturbances can be attenuated in the output channel. In addition, by combining with an output tracking controller, the proposed composite architecture effectively compensates the disturbances while guaranteeing the tracking of the specified film thickness across a wide range of situations. The simulation results were illustrated to showcase the effectiveness of the disturbance observer-based optimal tracking control system.

Keywords: System identification and modelling, Renewable energy and smart grids, Nonlinear control
Abstract: In the real world, model gaps always exist because models cannot perfectly match the objective physical plants. Model gaps reflect the integrity of models, which is vital for model validation and calibration. This letter proposes a novel approach for model gap quantification and evaluation. First, a comprehensive metric is developed to quantify dynamic model gaps. Next, the model gaps are evaluated from a panoramic view of the probability distribution of the comprehensive metric for multiple scenarios. Finally, we demonstrate the proposed quantification and evaluation approach through synchronous machine models.

Keywords: Estimation and filtering, Linear control theory, Classical and optimal control methods
Abstract: This paper considers the problem of structured observer synthesis for linear-time-invariant (LTI) systems. The importance of the problem is demonstrated through two motivational examples: utilizing high-fidelity black-box models as part of observers, and anti-windup gain design. The problem is connected with its control counterpart: static output feedback, and the process is illustrated through the design of an anti-windup gain for a multi-variable PID controller. A numerical example shows how the resulting closed-loop system converges faster than a recent method in the literature, with the added benefit of solving the problem with a non-iterative method, which decreases the computational effort.

Keywords: Multi-agent methods and distributed control, Robotics and autonomous systems, Human-machine systems
Abstract: A previously-proposed semi-autonomous search and rescue system is investigated experimentally. A stationary vision system identifies and tracks a victim using an object detection neural network. The vision system then guides an autonomous unmanned ground vehicle to the victim’s location using a ROS 2 publisher-subscriber communication protocol. A series of experiments are conducted with off-the-shelf components to investigate the impact of several design parameters. The resulting experimental data demonstrates that a fast and accurate response can be achieved. Furthermore, guidelines are established for parameter selection in future experiments.

Keywords: Autonomous systems, Artificial intelligence methods for control, Transportation and vehicle systems
Abstract: Typical autonomous driving systems are a combination of machine learning algorithms (often involving neural networks) and classical feedback controllers. Whilst significant progress has been made in recent years on the neural network side of these systems, only limited progress has been made on the feedback controller side. Often, the feedback control gains are simply passed from paper to paper with little re-tuning taking place, even though the changes to the neural networks can alter the vehicle's closed loop dynamics. The aim of this paper is to highlight the limitations of this approach; it is shown that re-tuning the feedback controller can be a simple way to improve autonomous driving performance. To demonstrate this, the PID gains of the longitudinal controller in the TCP autonomous vehicle algorithm are tuned. This causes the driving score in CARLA to increase from 73.21 to 77.38, with the results averaged over 16 driving scenarios. Moreover, it was observed that the performance benefits were most apparent during challenging driving scenarios, such as during rain or night time, as the tuned controller led to a more assertive driving style. These results demonstrate the value of developing both the neural network and feedback control policies of autonomous driving systems simultaneously, as this can be a simple and methodical way to improve autonomous driving system performance and robustness.

Keywords: Robotics and autonomous systems
Abstract: This paper presents a novel control approach for ensuring safety in a mobile manipulator system consists of a fixed-base manipulator mounted on a mobile platform. Initially, the trajectory tracking problem of a non-holonomic mobile manipulator (NH-MM) by employing decoupling dynamic control is addressed. This enables the independent control of the end-effector position and the mobile platform position. Subsequently, an obstacle avoidance method based on the control barrier function (CBF) is introduced. The objective is to utilize the system’s redundancy for obstacle avoidance by the mobile platform of the NH-MM. By adopting this approach, the mobile platform can successfully navigate around obstacles while the end-effector continues to perform its intended task autonomously. The effectiveness of the proposed method is demonstrated through simulation results.

Keywords: Robotics and autonomous systems, Autonomous systems, Artificial intelligence methods for control
Abstract: Motion planning plays an important role in autonomous manipulators, especially when operating in confined spaces with various constraints. The problem is particularly challenging in unstructured and cluttered environments where there is no a priori information on the surroundings available. In this paper, a method for constrained motion planning of a hydraulically actuated manipulator is proposed within the context of laser cutting applications in nuclear decommissioning. This method enables the robot to move the end-effector to follow the surface of an inspected workpiece, based on the assumption that cutting points are generated from a 3D vision system. The algorithm works in two phases: in the first step, the ConstrainedRRT* motion planning algorithm computes a collision-free path for the manipulator to reach the initial cutting point above the surface of the target object whilst respecting robot kinematic constraints. In the second step, a laser cutting path is generated such that it satisfies the constraints imposed by the user and avoids singularities. The results indicate that the proposed method is robust against a degraded camera with 40 dB SNR under a 10 Gy/h radiation dose. The algorithm's performance is also compared against some well-known approaches in terms of planning time and computational complexity.

Keywords: Artificial intelligence methods for control, Medical and biological systems, Human-machine systems

Keywords: Transportation and vehicle systems, Robotics and autonomous systems, Autonomous systems
Abstract: Numerous methodologies based on velocity obstacles have been developed for marine collision avoidance over the past decade. They are typically limited to one or two ship encounters, sequential ship encounters and/or lack safety guarantees in scenarios involving multiple obstacles. This paper proposes an adaptive collision avoidance strategy based on the velocity obstacle method to safely avoid multiple dynamic obstacles (one, two or more) simultaneously while navigating towards the waypoint in partial compliance with the Convention on International Regulations (COLREGs) rules. Our avoidance strategy is based on the construction of the adaptive velocity cone according to the motion of the dynamic obstacles that ensures safety with the obstacles at all times. We implement the proposed avoidance strategy on a set of standardised scenarios, namely, Imazu problems for multi-vessel encounter situations. We compare our proposed strategy with the standard velocity obstacle method for multiple dynamic obstacles and found that our approach is better than standard in terms of minimum safety distance between the ship and the obstacles.

Keywords: Iterative learning control, Robotics and autonomous systems
Abstract: A refined control scheme is presented to optimise the trajectory-tracking performance of robotic manipulators. This strategy integrates a linear feedback controller and a feedforward learning controller. The former mitigates unknown disturbances and desensitises the system to unidentified parameters, while the latter enhances tracking efficiency by incorporating past tracking errors. Traditionally, a fixed learning gain is used in the learning law to update the control input. However, we modify the learning law in this study by applying an adaptive learning gain. Our simulations on a robotic manipulator demonstrate that the adaptive ILC algorithm surpasses the classical ILC algorithm concerning convergence speed. These findings highlight the advantages of our approach, showcasing its extensive applicability in trajectory tracking.

Keywords: Iterative learning control, Distributed and decentralised control, Linear control design
Abstract: High performance collaborative tracking problem, requiring a group of independent subsystems to generate a global output that can precisely track the desired reference in a repetitive manner, has found lots of applications in practice. However, for such an important control task, existing iterative learning control (ILC) methods have not considered the constraint on the entire system's output, which leads to potential risk during the control process. This paper proposes a novel optimisation based ILC method to address the high performance collaborative tracking problem with output constraints. The proposed ILC framework can guarantee that not only the entire system's output constraints are always satisfied during the control process, but also the monotonic convergence of the collaborative tracking error norm to a possibly minimum value. To avoid huge computational complexity for large scale systems, we further apply the idea of the alternative direction method of multipliers (ADMM) to implement the proposed ILC framework in a decentralised manner, which allows the resulting decentralised methods to be applied to large-scale and changing systems. Moreover, the decentralised ILC method proposed in this paper is suitable for non-minimum phase, heterogeneous and/or homogeneous systems, which is appealing in practice. Convergence properties of the proposed ILC algorithms are analysed rigorously, and numerical examples are given to demonstrate the algorithms' effectiveness.

Keywords: Artificial intelligence methods for control, Process control systems, Autonomous systems
Abstract: We propose a self-learning real-time optimization framework with safety constraints for power target allocation of parallel hydrogen fuel cell co-generation stacks, where the performance characteristics of the fuel cells for electrical and thermal power output are initially unknown. For this purpose, Gaussian Processes (GPs) are utilized to model the unknown fuel cell characteristics while safety constraints for total thermal power output and total hydrogen consumption are enforced using upper confidence intervals of the learned GPs. Additionally, the estimation uncertainty of GPs is incorporated into the objective function as a contributing term to enable exploration, where the weight of this term is varied according to the process operating conditions. The expected tracking errors between the desired and achieved electrical and thermal power outputs are incorporated into the objective function as well and are penalized with predetermined weights. Simulation results show that the proposed framework can reduce the overall hydrogen consumption and meet both electrical power and thermal power targets – when they are feasible – while always satisfying the safety constraint. Comparison with an oracle algorithm with access to the true values of the fuel cell performance curves, reveal that the proposed approach attains comparable performance following a small number of external reference changes and independent exploration moves by the algorithm.

Keywords: Iterative learning control
Abstract: Iterative learning control (ILC) improves the tracking performance of a system working in a repetitive mode by learning from previous trials. The existing ILC algorithms can achieve high performance but often with the use of a system model or careful parameter tuning. To address this limitation, we propose an alternative approach: stochastic zeroth-order (ZO) optimisation-based ILC. The proposed algorithm can achieve good convergence performance without using a system model or deliberate parameter tuning. A convergence analysis is provided, and the effectiveness of the proposed algorithm is verified by a simulation example.

Keywords: Energy and Power systems, Renewable energy and smart grids, Predictive control
Abstract: Abstract: Model predictive control (MPC) is a widely used control strategy in the field of power system, which has good performance in power system frequency control, voltage control and power system stability enhancement. In this paper, a fast control strategy of virtual power plant based on model predictive control is proposed. The MPC controller is used to optimize the operation of the virtual power plant, and the optimal operation curve is predicted by the optimization algorithm. The MPC controller can adjust the operation of each distributed power plant in the virtual power plant according to the predicted optimal operation curve, so as to improve the overall operation efficiency and stability of the virtual power plant. In order to improve the response speed and control accuracy of the MPC controller, a fast control strategy based VPP is proposed. The controller improves the dynamic performance and robustness of the VPP. The simulation results show that the fast control strategy of virtual power plant based on model predictive control can effectively improve the overall operation efficiency and stability of the virtual power plant.

Keywords: virtual power plant, model predictive control, operation optimization, dynamic performance.

Keywords: Iterative learning control, Multi-agent methods and distributed control, Predictive control
Abstract: This paper addresses high-performance consensus tracking of repetitively operating networked dynamical systems using an iterative learning control (ILC) algorithm. It circumvents the need for precise model information in traditional methods and guarantees the high-performance by the predictive framework with a novel performance index that takes into account both current and future performance. The proposed algorithm ensures geometric convergence of the tracking error norm to zero and can be applied to both heterogeneous and non-minimum-phase systems. A distributed implementation of the algorithm is developed using the Alternating Direction Method of Multipliers, with detailed convergence analysis and numerical examples confirming its effectiveness.

Keywords: Classical and optimal control methods, Nonlinear control

Keywords: Nonlinear control, Autonomous systems, Robotics and autonomous systems
Abstract: Guaranteeing safety and stability is important in controller design for dynamical systems. Stability can be certified by Control Lyapunov Functions (CLFs), while safety can be certified by Control Barrier Functions (CBFs). These functions constrain the system vector field to certify the corresponding property. However, these constraints may be in conflict with each other at some points. In this paper, we propose a relaxed compatibility condition for a CBF-CLF pair, which can then be used to design a safe and locally stable controller. The proposed concept is demonstrated on a linear system.

Keywords: Nonlinear control, Artificial intelligence methods for control, Embedded systems
Abstract: Many systems involving neural networks (NNs) can be framed as Lurie systems: feedback systems consisting of a linear time-invariant (LTI) part and a static nonlinearity. Examples of these include the interconnection of LTI systems with L-layer feedforward NNs and continuous time recurrent neural networks (RNN). Stability analysis of a Lurie system lends itself to a range of criteria from absolute stability; however, in NN analysis, the size of m is typically large. As a result, existing absolute stability criteria suffer from greater conservatism and/or computational complexity. This paper addresses this problem by strengthening the low complexity classical Circle and Popov Criteria for the specialised case of the repeated ReLU nonlinearity (a popular NN activation function). The results are cast as a set of linear matrix inequalities (LMIs) with less restrictive conditions on the matrix variables than their classical counterparts.

Keywords: Nonlinear control, Classical and optimal control methods
Abstract: There exist third-order systems with input-output delay that do not satisfy the Kalman Conjecture. We provide a counterexample by simulation and show that there is no corresponding O'Shea-Zames-Falb multiplier.

Keywords: Nonlinear control, Control education
Abstract: Interconnection and damping assignment (IDA) is a popular passivity-based control (PBC) technique. This nonlinear control approach has proven suitable for stabilizing a wide range of systems. However, like many other nonlinear control techniques, implementing the IDA-PBC approach is often hampered by the necessity of solving partial differential equations (PDEs). This paper studies systems for which the solutions to these equations are always guaranteed, and closed-form solutions are known. To this end, we provide straightforwardly verifiable conditions over the open-loop system such that IDA-PBC can be applied without solving PDEs.

Keywords: Nonlinear control, Linear control theory
Abstract: Nonlinear negative imaginary systems find application in a range of engineering fields, including the control of flexible structures and air vehicles. Nevertheless, unlike their linear counterparts, the theory for nonlinear negative imaginary systems is not as well-established. In this paper, we propose a generalized k-th order dissipativity framework with respect to a supply rate which is a function of the k-th time-derivative of the system output. It is shown that positive realness and negative imaginaryness can be defined in this general framework in a unified manner. Then, necessary and sufficient conditions for first order dissipativity of nonlinear systems are obtained. These capture and are more general than the negative imaginary property. Moreover, the concept of exponentially negative systems for both linear and nonlinear systems is developed and the required conditions are obtained.

Keywords: System identification and modelling, Estimation and filtering, Intelligent manufacturing and production
Abstract: The estimation of bulk density and mass flow in fibreboard production is important for prediction of fiber quality and energy consumption. Therefore, a new lumped-parameter model of fibreboard production is developed and exploited by a Moving Horizon Estimator to estimate the bulk densities and mass flows. The estimation results are verified in simulated data, showing convergence to the reference values. In the next steps, the estimations need to be experimentally validated.

Keywords: Health monitoring, Fault detection and diagnosis, Robotics and autonomous systems
Abstract: A non-invasive method of characterising bowel lesions using machine learning (ML) and a robotic capsule has been proposed and investigated. As the capsule travels and encounters lesions in the lumen, its exhibited dynamics are envisaged to vary significantly in response to any change in the biomechanical properties of encountered lesions. Resulting capsule dynamics in the form of displacement signals have been processed for features that may be indicative of variation in biomechanical property such as stiffness (E). In the first stage ML, multi-layer perceptron (MLP) and support vector regression (SVR) were trained to predict E-values from the processed features. In the second stage, unsupervised K-means clustering is used to group the lesions into clusters of high intra-cluster similarities but low inter-cluster similarities using the MLP and SVR predicted E-values. The method achieved greater than 97 % clustering accuracies for both simulation and experimental data.

Keywords: System identification and modelling
Abstract: This work aims to develop a new method for estimating the stiffness of linear constraint of a single-degree-of-freedom impacting system. It was done by studying its free vibration response through extracting backbone curves by using the zero-crossing method. Efficacy of the proposed approach was demonstrated by comparing the backbone curves obtained from simulations and experiments. Both results showed different backbone curves under a variety of stiffnesses, where larger impacting stiffness may lead to a measurable lowering of the backbone curve, and the effective stiffness of the entire system can be estimated quantitatively.

Keywords: Aerospace and space systems, Autonomous systems
Abstract: Inverse Synthetic Aperture Radar (ISAR) images are a popular and effective tool used in the modern age to identify moving targets, particularly in the airborne and space arenas. Much research has been undertaken on the automatic recognition of targets in this area, applying computer vision algorithms to the two dimensional image maps produced when measuring targets via this method. In this document we discuss an on-going programme of work to fully automate space target recognition, and specifically here outline a methodology proposed for automating the identification of specific features of space targets, in order to aid the confidence of an operator making the final decisions. Large scale results are still currently being collected for the project.

Keywords: System identification and modelling, Linear control design, Active noise and vibration control
Abstract: Control systems capable of damping resonant modes while ensuring stability have been a major field of development as the requirements for high scanning speeds and precision become more widespread in modern engineering applications. Integral Resonant Control (IRC) is a promising approach that offers simplicity, ease of design and robustness. The ideal IRC controller displays the Butterworth pattern. Unfortunately, due to the inherent internal time delays of the systems, if not accounted for during the design stage, real world experimental trials show deviation from the Butterworth pattern and the emergence of highly resonant behaviour. This paper showcases what effect not accounting for time delays during the design procedure has on a system as well as offers a novel design methodology to reach as close as possible to the closed-loop Butterworth filter pattern within the IRC scheme in the presence of known time delays using the 1st order Pade approximation. The mathematical explanation for why the Butterworth pattern cannot be achieved is also explained. The proposed approach is compared to existing methodologies, demonstrating its effectiveness and superior performance in nanopositioning applications. Through simulations and experimental trials on a nanopositioning platform, it is shown that accounting for time delays during design significantly enhances the system’s performance and reduces tracking error. The presented methodology offers a practical solution for enhancing the precision of positioning systems with inherent time delays.

Keywords: System identification and modelling, Robotics and autonomous systems, Fault detection and diagnosis
Abstract: Accurate characterization of radiation hotspots is a critical requirement for monitoring and decommissioning operations in the nuclear industry, particularly where the arrangement of contamination is complex, and the availability of ground-truth data is limited. This article develops a novel stochastic modelling approach that alleviates challenges often present in such operations. Initially, the experimentally derived angular responses of a collimated single detector apparatus at different energy regions (counts over particular radiation footprints) are expressed by two functions: the Fourier transform of a rectangular pulse (approximated by a sinc function) and a Moffat function. Subsequently, these are both framed within a Dynamic Linear Regression (DLR) model. The resulting Moffat/sinc-DLR models enhance the quality of the fit to experimental data, and improve the accuracy and resolution of radiation localization, thus showcasing the value of such methods for radiation characterization tasks.

Keywords: Artificial intelligence methods for control, Cyber-physical systems
Abstract: Deep Reinforcement Learning (DRL) techniques have received significant attention in control and decision-making algorithms. Most applications involve complex decision-making systems, justified by the algorithms' computational power and cost. While model-based versions are emerging, model-free DRL approaches are intriguing for their independence from models, yet they remain relatively less explored in terms of performance, particularly in applied control. This study conducts a thorough performance analysis comparing the data-driven DRL paradigm with a classical state feedback controller, both designed based on the same cost (reward) function of the linear quadratic regulator (LQR) problem. Twelve additional performance criteria are introduced to assess the controllers' performance, independent of the LQR problem for which they are designed. Two Deep Deterministic Policy Gradient (DDPG)-based controllers are developed, leveraging DDPG's widespread reputation. These controllers are aimed at addressing a challenging setpoint tracking problem in a Non-Minimum Phase (NMP) system. The performance and robustness of the controllers are assessed in the presence of operational challenges, including disturbance, noise, initial conditions, and model uncertainties. The findings suggest that the DDPG controller demonstrates promising behavior under rigorous test conditions. Nevertheless, further improvements are necessary for the DDPG controller to outperform classical methods in all criteria. While DRL algorithms may excel in complex environments owing to the flexibility in the reward function definition, this paper offers practical insights and a comparison framework specifically designed to evaluate these algorithms within the context of control engineering.

Keywords: Artificial intelligence methods for control, Nonlinear control, Classical and optimal control methods
Abstract: In this paper, the optimal control problem of uncertain nonlinear systems is considered. A nonlinear disturbance observer is proposed to measure the lumped uncertainties present in the system. Disturbances that do not enter the same channel as the control signal, so-called mismatched disturbances, are difficult to reject directly within the control channel. To overcome the challenge, a generalized disturbance observer-based compensator is implemented to address the uncertainty compensation problem by attenuating its influence on the output channel. In real time, by augmenting the system states with the output tracking error, we develop a composite actor-critic reinforcement learning scheme for approximating the optimal control policy as well as the ideal value function pertaining to the compensated system by solving the Hamilton–Jacobi–Bellman equation. Concurrent learning is applied in this article by using the recorded data of the known model of the system, in order to enhance the robustness of the system by cancelling the influence of the probing signal. Simulation results demonstrate the effectiveness of the proposed scheme, offering an optimal solution for the output tracking problem in a second-order model with mismatched disturbances.

Keywords: Iterative learning control, Classical and optimal control methods, Robotics and autonomous systems
Abstract: Iterative learning control is a feedforward control scheme designed for systems operating in a repetitive setting to achieve high performance tracking for a single fixed reference, with fast learning of a control signal often only achieved when an accurate model of the system is known. On the other hand, biological control systems achieve fast learning without accurate a priori modelling, by learning dynamics and control signals simultaneously. Sensorimotor control studies the motion of humans and animals, and a key observation from this field is that a modular structure facilitates the generalisation of learnt skill, which inspires a new modular approach to iterative learning control design that accurately tracks trial-varying references.

Keywords: Iterative learning control, Nonlinear control
Abstract: In this work, a novel adaptive iterative learning control (AILC) scheme is proposed for a class of uncertain multi-input multi-output (MIMO) systems, where the control gain matrices are both unknown and time-varying. In order to develop the AILC scheme without requiring the exact knowledge of the control gain matrix, a directional parameter is firstly introduced to indicate the control direction, which thus paves the way to the utilization of the Nussbaum gain technique. Furthermore, the parametric uncertainty and the unknown control gain matrix are transformed into a norm-based function, based on which both the feedback control law and the parametric updating law are established to ensure the perfect tracking performance of the system states along the iteration axis. The convergence of the tracking error is rigorously analyzed under the framework of the composite energy function (CEF). Finally, a numerical example is illustrated to demonstrate the effectiveness of the proposed AILC scheme.

Keywords: Nonlinear control, Classical and optimal control methods
Abstract: Faster execution times achievable with explicit model predictive control (EMPC) promise to further extend the applicability of MPC. This work presents a novel approach for developing safe EMPC by training a neural network through a constrained optimization problem. Unique to this approach is the incorporation of closed-loop safety constraints directly into the neural network training step. Tractable training times are achieved since the number of points at which constraints are evaluated can be scaled with low computational cost, and the training problem is solved using a first order primal-dual method. The approach generalizes to nonlinear dynamics and constraints. Simulation experiments on three different systems demonstrate that the approach is capable of achieving approximately optimal, safe closed-loop control while requiring three to four orders of magnitude reduced time for online evaluation.

Keywords: Control education, Teaching laboratories, Linear control design
Abstract: There has a been a growing interest in virtual laboratories as a supplement to hardware laboratories in supporting student learning and experience. This paper focuses specifically on virtual laboratories built using the MATLAB environment and highlights some recent developments. Specifically, these virtual laboratories aim to give users an overview of the core content of an entire 1st course in control in a single virtual laboratory interface. This paper highlights two such laboratories and shows how they can be used to supplement other learning resources and activities.

Keywords: Teaching laboratories, Control education, Autonomous systems
Abstract: The teaching of autonomous vehicle control algorithm design involves the drawing together of multiple topics from both theory and practice. To enable an in-depth knowledge of how the control algorithms of an AV operate, a laboratory-scale approach using ‘lab-in-a-box’ has been developed at Aston University for a project-based module on the Future Vehicle Technologies MSc Course. The aim of the approach is such that students can learn the control engineering fundamentals before moving onto a larger vehicle platform. The adopted approach borrows ideas inspired from model-based design for embedded control. Continuous-time and discrete-time simulation tools are used to design what is in effect a real-time control system. Students are also taught ‘model shop’ skills and computer aided design (CAD), thus allowing the students to build a physical control system to operate in real-time. The focus is on the control of a DC motor, including speed control, temperature control and obstacle avoidance. Initial results presented here are promising, leading to increased levels of student satisfaction and engagement.

Keywords: Control education, Embedded systems, Teaching laboratories

Keywords: Aerospace and space systems, Linear control design, Transportation and vehicle systems
Abstract: This is a preliminary investigation into controlling the mass flow rate of gaseous hydrogen delivered to a turbine aeroengine, this being the key factor dictating the speed and thrust of the engine.

Keywords: Classical and optimal control methods, Estimation and filtering, Linear control design
Abstract: This work investigates a Floating Offshore Wind Turbine FOWT collective blade pitch controller aimed to reduce platform pitching and enhance generator power production. The work incorporates both, hydrodynamic state estimates and structural states in a Linear Parameter Varying LPV compensator. The controller design is based on linearised state space models of the wind turbine aided by OpenFAST, a high-fidelity wind turbine simulator. Method testing was on a full-order nonlinear model of the 15MW FOWT via OpenFAST using turbulent wind speed and an irregular wave disturbance. Compared to previous published work, the generator power and platform pitching RMSEs were reduced by 13% and 37%, respectively.

Keywords: Renewable energy and smart grids, Multi-objective methods and optimisation, Predictive control
Abstract: This study delves into the intricacies of integrating renewable energy sources (RESs) into microgrids. An energy management controller based on Model Predictive Control (MPC) is proposed, with the goal of optimising the balance between electricity demand and supply, hence improving adapt- ability and cost reduction. The study introduces a multi-objective optimisation framework that takes into account running cost, as well as an equivalent cost for battery degradation. Simulations on a 5 MW test microgrid using YALMIP and the MOSEK solver indicate the controller’s potential to improve adaptability and cost-effectiveness with energy storage integration.

Keywords: Renewable energy and smart grids, Predictive control
Abstract: Controlling batteries State of Charge (SoC) within operational constraints, while minimising the power exchange among microgrids and with the grid, is an important problem to maximise microgrids performance and extend batteries lives. To address this problem, this paper adopts an Unscented Predictive Control (UPC) to optimise the SoC control under uncertain conditions. Based on the model of the NMG and the principle of model predictive control, the design of the SoC control strategy is formulated as an Optimisation Problem (OP) with probability operation conditions. To deal with the latter, the unscented transformation is integrated with predictive control to derive the mean value and variance of system states. A tractable OP for NMGs is then obtained and the effectiveness of the proposed UPC-based SoC control strategy is verified by simulations with different NMG frameworks.

Keywords: Transportation and vehicle systems, Energy and Power systems, Predictive control
Abstract: The decarbonization of the commercial transport sector is a crucial part on the pathway to a fully green economy and the use of zero-emission Heavy Duty Vehicles (HDVs) is a major aim. Hydrogen Fuel Cells will probably represent the main technology to provide an alternative future fuel source to replace fossil fuels. This will involve combining fuel cells with batteries in HDVs, exploiting the full potential of these technologies in an economically effective way. To guarantee expected performance from fuel cell-based powertrains, these should be controlled by an appropriate Energy Management System to optimize the performance of the vehicle. This paper illustrates the performance achieved by applying optimal predictive control to the zero-emission Heavy-Duty Vehicles' power management problem.