Skip to Main Content
This work presents a novel adaptive fuzzy cerebellar model articulation controller (AFCMAC) to regulate the speed of a switched reluctance motor (SRM). The proposed controller comprises two parts - a fuzzy cerebellar model articulation controller (CMAC) and a compensating controller. The fuzzy CMAC learns and approximates system dynamics; the compensating controller compensates the approximation error of the fuzzy CMAC. The parameters of the AFCMAC are adjusted online according to adaptive rules, which are derived from Lyapunov stability theory, so that both the stability of the control system and error convergence can be guaranteed. The effectiveness and robustness of the proposed AFCMAC are investigated by numerical simulation and experimental studies. Three control strategies, AFCMAC, ACMAC and proportional-integral (PI) control, are experimentally investigated and the performance index, root mean square error (RMSE) of each scheme is evaluated. The experimental results indicate that AFCMAC provides a much better system performance than the other compared schemes. The proposed AFCMAC performs well in tracking ability, parameter variation capacity and load disturbance rejection capability. The effectiveness and practicability of the proposed control scheme in a practical SRM drive are experimentally verified.