Robust digital position control algorithm of PMSM using simple neural network compensator

Abstract

A very simple control approach using a neural network for the robust position control of a permanent magnet synchronous motor (PMSM) is presented. The linear quadratic controller plus feedforward neural network compensator is employed to obtain the robust PMSM system approximately linearized using a field-orientation method for an AC servo. The neural network is trained in on-line phases and a feedforward recall and error back-propagation training compose this neural network. Since the total number of nodes is only eight, this system is realized easily by the general microprocessor. During normal operation, the input–output response is sampled and the weighting value is trained multi-times by an error back-propagation method at each sample period to accommodate the possible variations in the parameters or load torque. The state space analysis is also performed to obtain the state feedback gains systematically. In addition, the robustness is also obtained without affecting overall system response. A floating-point digital signal processor DS1102 board (TMS320C31) realizes this method. The basic DSP software is used to write C-program, which is compiled by using ANSI-C style function prototypes.