Optimal tracking control for motion constrained robot systems via event-sampled critic learning

作者：Peng, Zhinan; Cheng, Hong; Shi, Kaibo; Zou, Chaobin^*; Huang, Rui; Li, Xiaoqing; Ghosh, Bijoy Kumar

来源：EXPERT SYSTEMS WITH APPLICATIONS, 2023, 234: 121085.

DOI：10.1016/j.eswa.2023.121085

• Optimal tracking control
• Robot system
• Motion constraints
• Dynamic event-sampled mechanism
• Critic learning

This article proposes a novel adaptive event-sampled learning method for addressing the optimal tracking control problem for robotic systems with motion constraints. In many practical systems, the joint movement of robots is restricted by mechanical structures or operational limitations, thus prescribed constraints are imposed on the joint states of the robot. To deal with the issue, a state-dependent transformation method is used to ensure motion restrictions and construct an unconstrained tracking error system. Then, for the transformed system, the constrained control problem is turned into a general optimal tracking control problem. To obtain the optimal solution, an only-critic learning-based framework is developed. Two novel event-sampled mechanisms are incorporated into the controller designs, reducing the state sampling times and computing costs. It is proven that the stability of the closed-loop system is ensured and the Zeno behavior is successfully eliminated via event-sampled learning control approaches. Finally, simulation and comparisons are conducted to validate the effectiveness of the theoretical results and proposed learning-based control approach.

• 单位
  电子科技大学