Smith Predictor Compensation and Fuzzy Incremental Control for Delay of Space Docking Hardware-in-the-loop Simulation System
Simiao Yu,Junwei Han,Wenming Zhang,Dongmei Xu
DOI: https://doi.org/10.1177/09544100211053305
2021-01-01
Proceedings of the Institution of Mechanical Engineers Part G Journal of Aerospace Engineering
Abstract:Hardware-in-the-loop (HIL) simulation for space manipulator docking is an important means to simulate real space docking on the ground. The HIL simulation system in this paper utilizes the contact force measured by force sensor to calculate the dynamics of the mechanisms, and the docking process is simulated by the parallel robot. The measurement delay of force sensor and dynamic response delay of the parallel robot are inevitable, which not only affect the accuracy of simulation but also lead to the instability of the HIL simulation system. The traditional first-order phase compensation is the most commonly used force sensor compensator; but when the force changes with a high frequency, its compensation effect becomes bad, which will lead to the divergence of the HIL simulation system. Most control methods of the parallel robot are based on the model of the parallel robot, but the forces of the parallel robot are complex during the docking process, and the system parameters, motion frequency, and dynamic response characteristics are time-varying; thus, it is difficult to design the controller based on the model. In this paper, the Smith predictor compensation (SPC) method and fuzzy incremental control (FIC) method are utilized to decrease the delays of the force sensor and parallel robot, respectively. The effectiveness of the Smith predictor compensation and fuzzy incremental control method in reducing the delay of the HIL system and in improving the stability of the system is verified by simulation and experiment; compared with the traditional first-order phase compensation and proportional-integral-differential control methods, the advantages of the proposed methods are illustrated. The research in this paper provides an important technical means for accurately simulating the real docking process.