Saiwei Wang

Saiwei Wang

Auto-generated profile, based on publicly available information

Saiwei Wang

Key Laboratory of Smart Manufacturing in Energy Chemical Process, Ministry of Education, East China University of Science and Technology, Shanghai, China

Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
9 November 2020 in IEEE Transactions on Network Science and Engineering
Verified
Journal Article
Closed access
SWSaiwei Wang
Saiwei Wang
XJXin Jin
Xin Jin
SMShuai Mao
Shuai Mao
AVAthanasios V. Vasilakos
Athanasios V. Vasilakos
YTYang Tang
Yang Tang
This paper develops a model-free approach to solve the event-triggered optimal consensus of multiple Euler-Lagrange systems (MELSs) via reinforcement learning (RL). Firstly, an augmented system is constructed by defining a pre-compensator to circumvent the dependence on system dynamics. Secondly, the Hamilton-Jacobi-Bellman (HJB) equations are applied to the deduction of the model-free event-triggered optimal controller. Thirdly, we present a policy iteration (PI) algorithm derived from RL, which converges to the optimal policy. Then, the value function of each agent is represented through a neural network to realize the PI algorithm. Moreover, the gradient descent method is used to update the neural network only at a series of discrete event-triggered instants. The specific form of the event-triggered condition is then proposed, and it is guaranteed that the closed-loop augmented system under the event-triggered mechanism is uniformly ultimately bounded (UUB). Meanwhile, the Zeno behavior is also eliminated. Finally, the validity of this approach is verified by a simulation example.

6 June 2022 in IEEE Transactions on Network Science and Engineering
Verified
Journal Article
Closed access
XJXin Jin
Xin Jin
SMShuai Mao
Shuai Mao
Ljupco KocarevLjupco Kocarev
Ljupco Kocarev
CLChen Liang
Chen Liang
SWSaiwei Wang
Saiwei Wang
YTYang Tang
Yang Tang
In this paper, an event-triggered Reinforcement Learning (RL) method is proposed for the optimal attitude consensus of multiple rigid body networks with unknown dynamics. Firstly, the consensus error is constructed through the attitude dynamics. According to the Bellman optimality principle, the implicit form of the optimal controller and the corresponding Hamilton-Jacobi-Bellman (HJB) equations are obtained. Because of the augmented system, the optimal controller can be obtained directly without relying on the system dynamics. Secondly, the self-triggered mechanism is applied to reduce the computing and communication burden when updating the controller. In order to address the problem that the HJB equations are difficult to solve analytically, an RL method which only requires measurement data at the event-triggered instants is proposed. For each agent, only one neural network is designed to approximate the optimal value function. Each neural network is updated only at the event-triggered instants. Meanwhile, the Uniformly Ultimately Bounded (UUB) of the closed-loop system is obtained, and Zeno behavior is also avoided. Finally, the simulation results on a multiple rigid body network demonstrate the validity of the proposed method.

1 January 2021 in Discrete and Continuous Dynamical Systems - S
Verified
Journal Article
Closed access
PDF file
YXYuan Xu
Yuan Xu
XJXin Jin
Xin Jin
SWSaiwei Wang
Saiwei Wang
YTYang Tang
Yang Tang
In this paper, an event-triggered reinforcement learning-based met-hod is developed for model-based optimal synchronization control of multiple Euler-Lagrange systems (MELSs) under a directed graph. The strategy of event-triggered optimal control is deduced through the establishment of Hamilton-Jacobi-Bellman (HJB) equation and the triggering condition is then proposed. Event-triggered policy iteration (PI) algorithm is then borrowed from reinforcement learning algorithms to find the optimal solution. One neural network is used to represent the value function to find the analytical solution of the event-triggered HJB equation, weights of which are updated aperiodically. It is proved that both the synchronization error and the weight estimation error are uniformly ultimately bounded (UUB). The Zeno behavior is also excluded in this research. Finally, an example of multiple 2-DOF prototype manipulators is shown to validate the effectiveness of our method.