Sliding mode control of electromechanical actuators based on an improved RBF network

DOI:	10.11809/bqzbgcxb2023.04.031
Keywords:	electromechanical actuator; radial basis function neural network; sliding mode control; flat top; ballistic chattering
Abstract:	Aiming at the ballistic chattering caused by the small angle flat top of an electric actuator, this paper studies the flat top phenomenon by focusing on the small angle flat top issue, and proposes an improved sliding mode control method of the radial basis function (RBF) network. The sliding mode controller is designed to weaken the adverse effects of friction and backlash so as to improve the tracking precision of the system. Aiming at the uncertainties of the model, the RBF network is introduced to approximate them, which improves the control accuracy and, at the same time, eliminates the chattering of the sliding mode control algorithm. In order to improve the iterative efficiency of the RBF network, two single parameters are used to replace the weight estimation matrix, which greatly reduces the number of parameters that need iterative estimation. Finally, the experimental platform of the electromechanical actuator system is built, and the PID controller is compared with the improved RBF sliding mode controller. The experimental results show that, compared with the traditional PID controller, when tracking 0.1° 4 Hz sinusoidal signals, the improved RBF sliding mode controller algorithm reduces the flat top time from 64 ms to 9 ms, and the position tracking error from 0.123° to 0.04°. In the semi physical simulation, the frequency of ballistic chattering reduces from 14.7 Hz to 6.8 Hz, and the chattering amplitude reduces from 0.079 ° to 0.034 °. The improved RBF sliding mode controller can effectively improve the tracking performance of the system, weaken the flat top and restrain ballistic limit cycle oscillation.
Issue:	Vol. 44 No. 4 (2023)
Published:	2023-04-28
PDF HTML