In order to improve the speed control performance of the PMSM speed control system, a fractional order integral neural network sliding mode speed controller is designed based on fractional order integral sliding mode. The robustness of the system is enhanced by the slow decay of fractional integration over time, and the chattering phenomenon of the sliding surface is weakened by the exponential approach law, analyzed the influence of fractional order on the dynamic performance of the system, and the RBF neural network is introduced to dynamically adjust the fractional integral order in the sliding surface function. The system comprehensive performance is improved. The simulation proves that compared with the ordinary fractional order integrated sliding mode control, the designed variable-order fractional-order integral sliding mode controller based on RBF neural network can acquire better dynamic tracking performance and stable accuracy.