Robust Fixed-Time Sliding Mode Attitude Control for a 2-DOF Helicopter Subject to Input Saturation and Prescribed Performance
Hao Shen, Xin Yu, Huaicheng Yan, Ju H. Park, Jing Wang
Abstract
This paper explores the issue of robust attitude control for a 2-DOF helicopter system under the fixed-time control rule. Thanks to the reinforcement-learning strategy, the optimization results for the attitude control objective have been achieved. Under the basic framework of the Actor-Critic Neural Networks, this paper not only solves a better solution of the cost-to-go function but also successfully estimates the external disturbance torque existed in the 2-DOF helicopter system. Furthermore, in conjunction with a sliding mode switching mechanism and a novel reaching law, this study introduces a new approach for effectively accomplishing the objective of attitude control while adhering to the constraints of input saturation and prescribed performance. Compared with other types of controllers, a fact can be validated that it has a better action performance of attitude control. In particular, under the action of the controller, each state variable has a stable bound over a specific fixed time. Finally, simulation and comparison examples offer evidence to demonstrate that the proposed control technique is stable.