Recurrent fuzzy neural network controller design using sliding-mode control for linear synchronous motor drive

A sliding-mode recurrent fuzzy neural network control (SMRFNNC) is proposed to control the mover of a permanent-magnet linear synchronous motor (PMLSM) servo drive so as to track a periodic sinusoidal reference trajectory. First, the PMLSM drive system is identified by a recurrent fuzzy neural network identifier (RFNNI) to provide sensitivity information of the drive system to a recurrent fuzzy neural network controller (RFNNC). Next, a sliding-mode adjuster (SMA) is determined according to the sliding mode condition. Then, the SMA is backpropagated through the RFNNI to train the parameters of the RFNNC online. Simulated and experimental results show that the control effort and chattering of the SMRFNNC are smaller than those of sliding-mode control. Moreover, a robust control performance is achieved when uncertainties occur including a nonlinear friction force.