Hybrid stepping motor position servo system with on-line trained fuzzy neural network controller

Abstract

A digital signal processor (DSP) based two-phase hybrid stepping motor (HBSM) position servo system with integral-proportional (IP) position controller, vector control and on-line trained fuzzy neural network (FNN) controller is introduced in this paper. First, an optimized torque control mechanism and IP position controller are applied to the HBSM servo drive, but the performance of the control system is still influenced by parameter variations and external disturbances of the plant. To overcome this penalty, then, a fuzzy neural network controller is presented to generate an adaptive signal to preserve favourable module-following characteristics. Simulations and experimental results are provided to demonstrate the working of the proposed controller.