Decentralized Tracking Optimization Control for Partially Unknown Fuzzy Interconnected Systems via Reinforcement Learning Method

In this paper, a novel neural network-based error-track iterative learning control scheme is proposed to tackle trajectory tracking problem for tank gun control systems. Firstly, the system modeling for tank gun control systems is introduced as a preparation of controller design. Then, the reference error trajectory is constructed to deal with the nonzero initial error of iterative learning control. The adaptive iterative learning controller for tank gun control systems is designed by using Lyapunov approach. Adaptive learning neural network is adopted to approximate nonlinear uncertainties, with robust control technique being used compensate the approximation error and external disturbances. As the iteration number increases, the system error can follow the desired error trajectory over the whole time interval, which makes the system state accurately track the reference error trajectory during the predetermined part time interval. Numerical simulations demonstrate the effectiveness of the proposed iterative learning control scheme. INDEX TERMS Tank gun control systems, iterative learning control, neural network, Lyapunov approach.

show abstract

“…respectively. In (10) and (11), θ θ θ * (t) is the optimal weight of neural network, (X X X k ) is the approximation error,…”

Section: Control System Designmentioning

confidence: 99%

Neural Network-Based Error-Tracking Iterative Learning Control for Tank Gun Control Systems With Arbitrary Initial States

et al. 2020

View full text Add to dashboard Cite

show abstract

“…In Algorithm 1, repeated iteration between (17) and (18) will be performed until convergence. In contrast to offline algorithm [35], the policy improvement step is conducted by the learned kernel matrixH j+1 .…”

Section: End Proceduresmentioning

confidence: 99%

“…Generally speaking, reinforcement learning provides adaptive optimal control design philosophy, which brings new insight into the Control System Community [13], [14]. It has given development to an alternative optimal control strategy known as adaptive dynamic programming (ADP), which is a (partially) model-free method achieving the optimal performance index [14]- [18]. Solutions to optimal control based on the idea of ADP have been extensively investigated for both linear quadratic regulator (LQR) and linear quadratic tracking (LQT) problems, see [19]- [22] and the references therein.…”

Section: Introductionmentioning

confidence: 99%

Reinforcement Q-Learning Incorporated With Internal Model Method for Output Feedback Tracking Control of Unknown Linear Systems

et al. 2020

View full text Add to dashboard Cite

This paper investigates the output feedback (OPFB) tracking control problem for discretetime linear (DTL) systems with unknown dynamics. With the approach of augmented system, the tracking control problem is first turned into a regulation problem with a discounted performance function, the solution of which relies on the Q-function based Bellman equation. Then, a novel value iteration (VI) scheme based on reinforcement Q-learning mechanism is proposed for solving the Q-function Bellman equation without knowing the system dynamics. Moreover, the convergence of the VI based Q-learning is proved by indicating that it converges to the Q-function Bellman equation and it brings out no bias of solution even under the probing noise satisfying the persistent excitation (PE) condition. As a result, the OPFB tracking controller can be learned online by using the past input, output, and reference trajectory data of the augmented system. The proposed scheme removes the requirement of initial admissible policy in the policy iteration (PI) method. Finally, effectiveness of the proposed scheme is demonstrated through a simulation example. INDEX TERMS Adaptive dynamic programming (ADP); optimal conrol; Bellman equation; on-policy; internal model

show abstract

“…PI Algorithm is used to seek the optimal value of υ i . Following [27], [28], the details of PI Algorithm are as follows:…”

Section: A Pi Algorithmmentioning

confidence: 99%

“…However, few works investigate optimal synchronization for CNs with unknown dynamics, which makes nodes not only achieve consensus but also minimizes the cost. As is known to all, optimal synchronization depends on the solution of HJB equation [27], [28]. Since the dynamics of each node in CN VOLUME 4, 2016 are always unknown and nonlinear, it is difficult to obtain the solution of HJB equation analytically.…”

Section: Introductionmentioning

confidence: 99%

Data-Driven Optimal Synchronization for Complex Networks With Unknown Dynamics

Gao

Dong

2020

IEEE Access

View full text Add to dashboard Cite

Decentralized Tracking Optimization Control for Partially Unknown Fuzzy Interconnected Systems via Reinforcement Learning Method

Cited by 61 publications

References 38 publications

Neural Network-Based Error-Tracking Iterative Learning Control for Tank Gun Control Systems With Arbitrary Initial States

Neural Network-Based Error-Tracking Iterative Learning Control for Tank Gun Control Systems With Arbitrary Initial States

Reinforcement Q-Learning Incorporated With Internal Model Method for Output Feedback Tracking Control of Unknown Linear Systems

Data-Driven Optimal Synchronization for Complex Networks With Unknown Dynamics

Contact Info

Product

Resources

About