Open AccessDesigning Active Disturbance Rejection Control of Permanent Magnetic Spherical Actuator Based on Nonlinear Extended State Observer
Author(s) -
Xiwen Guo,
Xianda Liao,
Qunjing Wang,
Yan Wen,
Tao Wu
Publication year - 2022
Publication title -
elektronika ir elektrotechnika
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.224
H-Index - 26
eISSN - 2029-5731
pISSN - 1392-1215
DOI - 10.5755/j02.eie.30102
Subject(s) - control theory (sociology) , nonlinear system , actuator , differentiator , active disturbance rejection control , decoupling (probability) , robustness (evolution) , state observer , trajectory , tracking error , engineering , computer science , mathematics , control engineering , control (management) , filter (signal processing) , physics , biochemistry , chemistry , quantum mechanics , artificial intelligence , astronomy , electrical engineering , gene
Aiming at the problems of low control precision and poor anti-disturbances ability of permanent magnet spherical actuator (PMSA), an active disturbance rejection trajectory tracking control method based on nonlinear extended state observer (NESO) is proposed in this paper. Firstly, the multivariable and strong coupling mathematical model of the PMSA is obtained by rigid body rotation coordinate transformation and the Lagrange dynamic equation. Then, the tracking differentiator is used to reduce the influence caused by the expected trajectory jump. In addition, NESO is designed to observe and calculate coupling of the system, external disturbance, and model error online. Finally, the nonlinear control law is used to compensate for the observed disturbance, and the stability is proved based on the Lyapunov equation. Simulation and experimental results show that the proposed control scheme can achieve decoupling and tracking control of the complex system in the presence of model error, random disturbance, and other uncertain factors, has good control accuracy and response speed, and has strong robustness to uncertain disturbances.