Reinforcement Learning for H∞ Optimal Control of Unknown Continuous-Time Linear Systems.

In this paper, a novel synchronous Q-learning method is proposed for solving discrete-time linear quadratic regulator (LQR) problems. To begin with, the Bellman equation corresponding to the optimal Q-function is reformulated into a consistency equation on the parameters of the optimal Q-function and the optimal controller. Then an actor-critic structure is introduced to learn the optimal Q-function and the optimal controller online in real time by using the state samples generated by the behavior policy. Particularly, the proposed synchronous Q-learning scheme simultaneously updates the Q-function approximation and the optimal controller approximation, rather than iterating between policy evaluation and policy improvement. The proposed control scheme is proved to be uniformly ultimately bounded (UUB) under appropriate learning rates, provided that certain persistence of excitation (PE) conditions are satisfied. Besides, the PE conditions can be easily met by injecting appropriate exploration noise into the behavior policy without causing any excitation noise bias. Finally, one simulation example is provided to verify the effectiveness of the proposed synchronous Q-learning method.

Dynamic event-triggered-based single-network ADP optimal tracking control for the unknown nonlinear system with constrained input

1 A Survey on Computational Optimal Control.- Issues in the Direct Transcription of Optimal Control Problems to Sparse Nonlinear Programs.- Optimization in Control of Robots.- Large-scale SQP Methods and their Application in Trajectory Optimization.- Solving Optimal Control and Pursuit-Evasion Game Problems of High Complexity.- 2 Theoretical Aspects of Optimal Control and Nonlinear Programming.- Continuation Methods In Boundary Value Problems.- Second Order Optimality Conditions for Singular Extremals.- Synthesis of Adaptive Optimal Controls for Linear Dynamic Systems.- Control Applications of Reduced SQP Methods.- Time Optimal Control of Mechanical Systems.- 3 Algorithms for Optimal Control Calculations.- Second Order Algorithm for Time Optimal Control of a Linear System.- An SQP-type Solution Method for Constrained Discrete-Time Optimal Control Problems.- Numerical Methods for Solving Differential Games, Prospective Applications to Technical Problems.- Construction of the Optimal Feedback Controller for Constrained Optimal Control Problems with Unknown Disturbances.- Repetitive Optimization for Predictive Control of Dynamic Systems under Uncertainty.- Optimal Control of Multistage Systems Described by High-Index Differential-Algebraic Equations.- A New Class of a High Order Interior Point Method for the Solution of Convex Semiinfinite Optimization Problems.- A Structured Interior Point SQP Method for Nonlinear Optimal Control Problems.- 4 Software for Optimal Control Calculations.- Automated Approach for Optimizing Dynamic Systems.- ANDECS: A Computation Environment for Control Applications of Optimization.- Application of Automatic Differentiation to Optimal Control Problems.- OCCAL: A mixed symbolic-numeric Optimal Control CALculator.- 5 Applications of Optimal Control.- A Robotic Satellite with Simplified Design.- Nonlinear Control under Constraints of a Biological System.- An Object-Oriented Approach to Optimally Describe and Specify a SCADA System Applied to a Power Network.- Near-Optimal Flight Trajectories Generated by Neural Networks.- Performance of a Feedback Method with Respect to Changes in the Air-Density during the Ascent of a Two-Stage-To-Orbit Vehicle.- Linear Optimal Control for Reentry Flight.- Steady-State Modelling of Turbine Engine with Controllers.- Shortest Paths for Satellite Mounted Robot Manipulators.- Optimal Control of the Industrial Robot Manutec r3.

Sliding-mode surface-based decentralized event-triggered control of partially unknown interconnected nonlinear systems via reinforcement learning

This paper is concerned with the optimal output synchronization control problem for a class of heterogeneous complex dynamical networks (CDNs) with partially unknown system dynamics. By defining a new quadratic performance index including the information of state couplings, an optimal controller consisting of a feedback control and a compensated feedforward control is then designed. A necessary and sufficient condition for the optimal control minimizing the quadratic performance index is established. Especially, an algebraic Riccati equation (ARE) associated with a discount factor is obtained, and a novel online iteration algorithm by iteratively solving the ARE is proposed to compute the optimal controller gain for the case that the CDNs are with partially unknown system dynamics. Moreover, it is proved that the output synchronization errors converge to zero asymptotically by applying the optimal control laws. Finally, two simulation examples are provided to verify the effectiveness of the proposed control method.

This article investigates the optimal tracking control problem for data-based stochastic discrete-time linear systems. An average off-policy Q-learning algorithm is proposed to solve the optimal control problem with random disturbances. Compared with the existing off-policy reinforcement learning (RL) algorithm, the proposed average off-policy Q-learning algorithm avoids the assumption of an initial stability control. First, a pole placement strategy is used to design an initial stable control for systems with unknown dynamics. Second, the initial stable control is used to design a data-based average off-policy Q-learning algorithm. Then, this algorithm is used to solve the stochastic linear quadratic tracking (LQT) problem, and a convergence proof of the algorithm is provided. Finally, numerical examples show that this algorithm outperforms other algorithms in a simulation.

The article proposes an optimized leader-follower formation control using a simplified reinforcement learning (RL) of identifier-critic-actor architecture for a class of nonlinear multiagent systems. In general, optimal control is expected to be obtained by solving Hamilton-Jacobi-Bellman (HJB) equation, but the equation associated with a nonlinear system is difficult to solve by analytical method. Although the difficulty can be effectively overcome by the RL strategy, the existing RL algorithms are very complex because their updating laws are obtained by carrying out gradient descent algorithm to square of the approximated HJB equation (Bellman residual error). For a multiagent system, due to the state coupling problem, these methods will become difficult implementation. In the proposed optimized scheme, the RL updating laws are derived from negative gradient of a simple positive function, which is the equivalence to HJB equation; therefore, the control algorithm is significantly simple. Furthermore, in order to solve the problem of unknown system dynamics, an adaptive identifier is integrated into the control. Finally, the theory and simulation demonstrate that the optimized formation scheme can guarantee the desired control performance.

In this paper, a learning-based predictive control (LPC) scheme is proposed for adaptive optimal control of discrete-time nonlinear systems under stochastic disturbances. The proposed LPC scheme is different from conventional model predictive control (MPC), which uses open-loop optimization or simplified closed-loop optimal control techniques in each horizon. In LPC, the control task in each horizon is formulated as a closed-loop nonlinear optimal control problem and a finite-horizon iterative reinforcement learning (RL) algorithm is developed to obtain the closed-loop optimal/suboptimal solutions. Therefore, in LPC, RL and adaptive dynamic programming (ADP) are used as a new class of closed-loop learning-based optimization techniques for nonlinear predictive control with stochastic disturbances. Moreover, LPC also decomposes the infinite-horizon optimal control problem in previous RL and ADP methods into a series of finite horizon problems, so that the computational costs are reduced and the learning efficiency can be improved. Convergence of the finite-horizon iterative RL algorithm in each prediction horizon and the Lyapunov stability of the closed-loop control system are proved. Moreover, by using successive policy updates between adjoint time horizons, LPC also has lower computational costs than conventional MPC which has independent optimization procedures between two different prediction horizons. Simulation results illustrate that compared with conventional nonlinear MPC as well as ADP, the proposed LPC scheme can obtain a better performance both in terms of policy optimality and computational efficiency.

An ADDHP-based Q-learning algorithm for optimal tracking control of linear discrete-time systems with unknown dynamics

This article addresses the online reinforcement Q-learning algorithms to design H <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">∞</sub> tracking controller for unknown discrete-time linear systems. An augmented system composed of the original system and the command generator is constructed, and a discounted performance function is introduced to establish a discounted game algebraic Riccati equation (GARE). The existence conditions of a solution to the GARE are proposed and a lower bound is found for the discount factor to assure the stability of the H <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">∞</sub> tracking control solution. The Q-function Bellman equation is then derived, based on which the reinforcement Q-learning algorithm is developed to learn the solution to H <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">∞</sub> tracking control problem without knowing the system dynamics. Both state-data-driven and output-data-driven reinforcement Q-learning algorithms toward finding the control policies are proposed. Unlike the value function approximation (VFA)-based approach, it is proved that the Q-learning scheme brings out no bias of solution to the Q-function Bellman equation under the probing noise satisfying the persistent excitation (PE) condition, and therefore, converges to the nominal discounted GARE solution. Moreover, the proposed output-data-driven method is more powerful than the state-data-driven method as it may not be available to completely measure the full system states in practical applications. A simulation example with a single-phase voltage-source UPS inverter is used to verify the effectiveness of the proposed Q-learning algorithms.

Optimal semistable control for continuous-time linear systems

This article addresses the optimal tracking control problem with prescribed performance for uncertain nonlinear systems subject to input constraint and unknown disturbances. First, a fixed‐time monotonic convergence function is introduced to restrain tracking error, and a nonlinear mapping technique is employed to transform the constrained error into an unconstrained variable, then the fixed‐time output tracking issue is boiled down to the boundedness problem of the transformed variable. With the aid of a nonquadratic cost function, the input constraint is encoded into the optimization problem. To solve the unknown disturbances, an auxiliary system and an auxiliary disturbance policy are constructed, and the optimal control problem is formulated as a two‐player zero‐sum game. Moreover, a Hamilton–Jacobi–Isaacs (HJI) equation associated with this nonquadratic zero‐sum game is established to give the optimal control and the worst‐case disturbance policy solution. Subsequently, to avoid using knowledge of the system dynamics, three neural network approximators, namely, actor, critic, and disturbance, which are tuned online and simultaneously for approximating the solution of HJI, are constructed based on the integral reinforcement learning algorithm. Theoretical analysis shows that the reconstructed error system states and the weight estimation errors are semi‐globally uniformly ultimately bounded. Finally, the simulation study further tests the availability of the proposed control strategy.

This paper studies neural learning control with predefined tracking error bound for a marine surface vessel whose accurate dynamics could not be obtained a priori . With the introduction of an error transformation function, the constrained tracking control of the original vessel is transformed into the stabilization of an unconstrained system. A filtered tracking error is introduced based on the error transformation, and radial basis function (RBF) neural networks (NNs) are employed to approximate unknown vessel dynamics. Subsequently, stable adaptive NN control is proposed to ensure ultimate boundedness of all the signals in the closed-loop system and to guarantee prescribed tracking performances. Under persistent excitation (PE) condition, the proposed adaptive NN control is shown to be capable of acquiring knowledge on the vessel dynamics and of storing the learned knowledge in memory. The stored knowledge is reused to develop neural learning control such that the improved control performance with faster tracking convergence rate and less computational burden could be achieved, while prescribed transient and steady-state tracking control performances are guaranteed. Simulation studies are performed to demonstrate the effectiveness of the proposed design techniques.

In this paper, an iterative adaptive dynamic programming (ADP) algorithm is developed to solve the optimal cooperative control problems for residential multi-battery systems. To avoid solving high-dimensional optimal control problems, we first constrain all the batteries at their worst performance, which transforms the multi-input optimal control problem into a single-input one. Based on the worst-performance optimal control law, the optimal cooperative control law for the residential multi-battery systems is obtained, where in each iteration, only a single-input optimization problem is implemented. Finally, numerical results are given to illustrate the performance of the developed algorithm.

This paper presents a functional model predictive control (MPC) approach based on an adaptive dynamic programming (ADP) algorithm with the abilities of handling control constraints and disturbances for the optimal control of nonlinear discrete-time systems. In the proposed ADP-based nonlinear MPC (NMPC) structure, a neural-network-based identification is established first to reconstruct the unknown system dynamics. Then, the actor-critic scheme is adopted with a critic network to estimate the index performance function and an action network to approximate the optimal control input. Meanwhile, as the MPC strategy can effectively determine the current control by solving a finite horizon open-loop optimal control problem, in the proposed algorithm, the infinite horizon is decomposed into a series of finite horizons to obtain the optimal control. In each finite horizon, the finite ADP algorithm solves the optimal control problem subject to the terminal constraint, the control constraint, and the disturbance. The uniform ultimate boundedness of the closed-loop system is verified by the Lyapunov approach. Finally, the ADP-based NMPC is conducted on two different cases and the simulation results demonstrate the quick response and strong robustness of the proposed method.