This paper investigates the design of an adaptive event-triggered controller to facilitate the efficient auto-scaling of the distributed fog computing platform (DFP), while meticulously considering resource constraints, disturbances, and uncertainties. A data-driven discrete polytopic linear parameter-varying (PLPV) model effectively addresses the service requirements of DFP with time-varying characteristics. The controller gains are formulated utilizing linear matrix inequalities (LMIs) to guarantee the ${\mathbb{D}_R}$ stability of the system. Finally, the results are experimentally validated on a DFP to offload the images from an autonomous mobile robot (AMR) specifically designed for vision applications in a warehouse environment. Furthermore, the validation results are systematically compared to a static event-triggered control approach for discrete PLPV systems.

This paper delineates the development of two sophisticated control strategies: Adaptive Event-Triggered Control (AETC) and Adaptive Self-Triggered Control (ASTC), specifically designed for a discrete-time polytopic Linear Parameter Varying (LPV) system subjected to input saturation. Initially, a novel AETC algorithm is introduced, which utilizes a parameter-varying Lyapunov function specifically tailored for the polytopic LPV system with input saturation constraint. Subsequently, the ASTC is proposed to mitigate the necessity for continuously monitoring measurement errors. The theoretical advancements presented herein are experimentally validated on a distributed fog computing platform (DFP) for efficient offloading of vision data acquired from a mobile robot in a warehouse environment.

This paper focuses on adaptive prescribed performance control for nonlinear systems with parametric uncertainties. The proposed control scheme incorporates a certainty equivalence controller, a batch least-squares identifier (BaLSI) and a performance triggered condition. The off-line BaLSI, which utilizes all the previously appeared excitation information for parameter updating, is activated as intervals by the performance triggered condition. The effects of the parametric uncertainties are eliminate in finite times of updating, and the closed-loop system can achieve the prescribed performance without suffering from stiff differential equation problem. The simulation results are provided to demonstrate the effectiveness of the proposed control scheme.

This paper proposes a control-theoretic approach for discrete polytopic linear parameter varying (pLPV) systems with time-varying characteristics. The approach employs event-triggered model predictive control (ET-MPC) to uphold the performance while minimising the frequency of controller updates. Initially, a model predictive control (MPC) is designed to predict the system’s dynamics in the future. An event-triggering mechanism is also incorporated in the design, which uses a parameter-varying Lyapunov function to lessen the burden of online optimization. The design of an ET-MPC is based on the global stability and feasibility of the pLPV system. Finally, the control-theoretic approach is experimentally validated on a distributed fog computing platform to enable the auto-scaling of compute nodes for mobile robot vision applications.

In this article, the decentralized adaptive secure control problem for cyber-physical systems (CPSs) against deception attacks is investigated. The CPSs are formed as a type of nonlinear interconnected strict-feedback systems with uncertain time-varying parameters. The attack affects the information transmission between sensor and actuator in a multiplicative manner. A novel decentralized adaptive backstepping secure control strategy is established by exploiting a particular kind of Nussbaum functions and a flat-zone Lyapunov function analysis approach. It is shown that all of closed-loop signals remain globally bounded, and each output signal eventually converges into a small neighborhood of the origin. Simulation results on an illustrative example are provided to display the effectiveness of the proposed control scheme.

Safe handling of heavy payloads in an off-shore environment requires careful crane maneuvering to avoid collision with obstacles and other equipment. The residual swing from a hanging payload can ultimately lead to danger and high cost failures if not properly dealt with. This paper investigates an open-loop control method for eliminating the payload swing for hanging loads on offshore knuckle-boom cranes. A trajectory tracking method is designed for payload swing suppression for open-loop control and based on the iterative learning algorithm. It is shown that the proposed anti-swing control method guarantees asymptotic convergence of the swing, the angular velocity, and the angular acceleration of the payload using Lyapunov techniques. The simulation results show the superior performance of the proposed anti-swing control method.

This paper investigates the attack detection and reconstruction problem for a class of nonlinear systems with unknown parameters and actuator/state attacks. An adaptive sliding mode observer with online parameter estimation is designed to estimate the states and the convergence of estimation errors is guaranteed. By analyzing the features of adaptive sliding mode observers, an attack detection and reconstruction scheme is proposed. It is shown that the reconstruction signal can approximate the attack with any accuracy under a persistent excitation condition. Finally, simulation results are given to verify the effectiveness of the proposed scheme.

Multi-agent deep reinforcement learning (MADRL), where a group of agents inside multi-agent systems cooperate to achieve a common goal, has been shown useful in many applications such as collaborative robots, autonomous driving or video games involving teams. In this paper, we propose two multi-agent deep reinforcement learning (MADRL) frameworks for value function factorization built by using Graph Convolutional Neural Networks (GCNN) based on neighborhood graph filters (NGFs). These MADRL frameworks are based on the paradigm of centralized training with decentralized execution (CTDE). In this work, we show that the superior stability of the NGFs as compared to standard graph filters leads also to superior performance for the MADRL algorithms. In the first MADRL framework, the NGF-based GCNN is used to predict the local q-value at each of the agents, while in the other one, the NGF-based GCNN is used to mix the local q-values to generate a global Q-value. We have compared the performance of NGF-based GCNNs over state-of-the-art graph neural networks for value function factorization in the MADRL framework for the StarCraft II and Coalition Structure Generation problems. The results show that the proposed MARL frameworks outperform the existing state-of-art architectures.

In this paper, we consider a 2-degrees-of-freedom (DOF) helicopter system subject to input delays and uncertain system parameters. To address this challenge in control design, we develop an adaptive predictor-feedback control law. The control law is designed to compensate for a known delay while considering the system uncertainty. Stability of the closed-loop system is established, where tracking is achieved. We demonstrate the effectiveness of our proposed control approach through simulations of the helicopter system, where the input delays are compensated in the control-loop.

While Deep Learning based methods can solve complex problems by employing Neural Networks to act as powerful function approximators, they often suffer from inflexibility in terms of deployment beyond the training scenario and also include irrelevant data priors in the form of an ordered array of input values. This problem is quite evident in the field of Multi-Agent Reinforcement Learning (MARL), where most research covers methods that are trained on a fixed number of agents, restricted by the fixed size of the input vector. In this paper, we argue that this is not a reasonable assumption, both in terms of the inflexible amount of environmental information and the restrictive nature of the structure of the information. We explore DeepSets and Set Transformers as two powerful set function approximators to address the problem of cardinality invariance and permutation invariance in the observation space of a reinforcement learning agent. We explain Set-Input Reinforcement Learning (SIRL) in detail and evaluate the performance of the DeepSets and Set Transformer methods through simulated experiments on a challenging multi-agent environment, that otherwise yields sub-optimal policies through traditional function approximation approaches. We demonstrate that both DeepSets and Set Transformer based encoders scale well to increasing the number of agents from training to evaluation.

This paper focuses on motor joint control using learning based Adaptive Dynamic Programming (ADP). The dynamics of a 2-link robot manipulator is presented. The system is regarded as a linear system and the optimal control is developed using linear quadratic regulator (LQR) and ADP methods. An off-line, off-policy controller iteration approach is used to find two optimal controllers, one for each motor joint of the robot. The optimal controller is developed to control the joint position without knowing the dynamic model. Simulation results show that two independent optimal linear control policies are learned simultaneously under a common exploration strategy, and the efficiency of ADP is verified.

In this paper the attitude tracking control problem of a 2 degrees-of-freedom helicopter system with network induced constraints is studied. A predictor feedback control law is developed to compensate a known delay in the communication, where the inputs are quantized before transmitted over the network. Stability of the closed-loop system is established, where tracking is achieved with bounded tracking errors due to the network issues. The developed predictor-based controller is experimentally tested on the helicopter system, where we demonstrate that tracking is achieved in presence of both input delay and quantization.

Multi-Agent Reinforcement Learning (MARL) algorithms based on the Centralised Training Decentralised Execution (CTDE) approach have seen a great deal of interest in recent years. Most of the recent works focus on a specific class of environments built on the StartCraft Multi-Agent Challenge (SMAC) environment suite. However, experiments with the PettingZoo multi-particle environments show poor performance in an interesting subset of tasks. In this paper, the nature of these environments and the reward structures are analyzed. It shows that poor performance in these tasks is not due to a lack of representation power in the individual Q function or mixing functions, but rather a result of convergence to a suboptimal equilibrium of dual channel rewards and the issue of agent-reward decoupling that can be a common problem for many MARL environments. We present reward curriculum-based versions of QMIX and VDN as a solution to these problems and compare their results with the standard algorithms. The results show a clear performance gain in terms of a common cumulative reward metric.

In this article, we investigate the output consensus tracking problem for a class of high-order nonlinear systems with unknown parameters, uncertain external disturbances, and intermittent actuator faults. Under the directed topology conditions, a novel distributed adaptive controller is proposed. The common time-varying trajectory is allowed to be totally unknown by part of subsystems. Therefore, the assumption on the linearly parameterized trajectory signal in most literature is no longer needed. To achieve the relaxation, extra distributed parameter estimators are introduced in all subsystems. Besides, to handle the actuator faults occurring at possibly infinite times, a new adaptive compensation technique is adopted. It is shown that with the proposed scheme, all closed-loop signals are globally uniformly bounded and asymptotically output consensus tracking can be achieved.

This paper proposes a new adaptive controller for a 2-Degree of Freedom (DOF) helicopter system in the presence of input quantization. The inputs are quantized by uniform quantizers. A nonlinear mathematical model is derived for the 2-DOF helicopter system based on Euler-Lagrange equations, where the system parameters and the control coefficients are uncertain. A new adaptive control algorithm is developed by using backstepping technique to track the pitch and yaw position references independently. Only quantized input signals are used in the system which reduces communication rate and cost. It is shown that not only the ultimate stability is guaranteed by the proposed controller, but also the designers can tune the design parameters in an explicit way to obtain the required closed loop behavior. Experiments are carried out on the Quanser helicopter system to validate the effectiveness, robustness and control capability of the proposed scheme.

In this paper, we investigate the output tracking control problem for a class of uncertain nonlinear systems in parametric strict feedback form with quantized input. A novel backstepping based adaptive quantized control scheme is proposed. Different from the existing results, the true quantization parameters are allowed to be unknown in the design of adaptive controller. It is shown that with the proposed control scheme, the system output can track the desired trajectory asymptotically and all the closed-loop signals are globally uniformly bounded.

In this paper, we address adaptive predictor feedback design for a simplified drilling system in the presence of disturbance and time-delay. The main objective is to stabilize the bottomhole pressure at a critical depth at a desired set-point directly. The stabilization of the dynamic system and the asymptotic tracking are demonstrated by the proposed adaptive control, where the adaptation employs Lyapunov update law design with normalization. The proposed method is evaluated using a high fidelity drilling simulator and cases from a North Sea drilling operation are simulated. The results show that the proposed predictor controller is effective to stabilize the bottom hole pressure within the desired margins and compensate the effects of the delay and disturbance.

In this paper, we consider a class of uncertain magnetic suspension system preceded by Bouc-Wen type of hysteresis nonlinearity. A new perfect inverse function of the hysteresis is constructed and used to cancel the hysteresis effects in controller design with backstepping technique. For the design and implementation of the controller, no knowledge is assumed on system parameters. It is shown that the proposed controller not only guarantees asymptotic stability, but also transient performance.

In this paper, the results of stabilizing a large scale nonlinear systems with uncertain dynamic interactions and unmodelled dynamics depending on both subsystem inputs and outputs in [9] and [10] are extended to highly nonlinear systems. Certain modifications on standard adaptive backstepping controllers are proposed to compensate for the effects of interactions from other subsystems in designing local controllers. .

In this paper, we present a new scheme to design adaptive controller for uncertain nonlinear systems in the presence of input saturation. The control design is achieved by using backstepping technique and neural network. Unlike some existing control schemes for systems with input saturation, the developed controller does not require uncertain parameters within a known compact set. Besides showing stability, transient performance is also established and can be adjusted by tuning certain design parameters

In this paper, we present an adaptive backstepping control algorithm for a class of uncertain nonlinear systems under state-feedback form including hysteretic nonlinearity. The result is applied to a system found in base isolation schemes for seismic active protection of building structures. This system exhibits a hysteretic nonlinear behavior which is described by the so-called Bouc-Wen model. Unlike other control schemes, the developed backstepping control does not require the model parameters within known intervals. It is shown that not only global stability is guaranteed by the proposed controller, but also both transient and asymptotic performances are quantified as explicit functions of the design parameters so that designers can tune the design parameters in an explicit way to obtain the required closed loop behavior.