ANFIS controller based on RBF identification for piezoelectric actuator in a positioning system

Abstract

Piezoelectric actuator is widely used in precision positioning mechanism for the advantages of ultra-high resolution, easy miniaturization, and rapid dynamic performance. But piezoelectricity material has nonlinear hysteresis effect. Due to this characteristic, accurate tracking is difficult to be achieved. In this paper, a model for piezoelectric actuator with hysteretic nonlinear model is developed. Also to improve the performance of piezoactuator, two control algorithms, adaptive PID control based on RBF neural network and ANFIS controller based on RBF neural network identification, are proposed. In both controllers, an identification algorithm is used to adjust timely the weights and coefficients of the controllers. Then simulation results of these algorithms are compared with together. The results show that an average error of ANFIS controller and a PID-RBF is close together; however, ANFIS controller is capable to reduce the effect of hysteresis better than the PID-RBF controller.