Speed Control System Based on Fuzzy Neural Network of BLDCM

Abstract

The controller uses the BP network as the basic structure. Fuzzy rules, fuzzy variables and membership functions are integrated by Mamdani fuzzy control method in the network structure and parameters; BP algorithm is applied to adjust the weights of the network, which is equal with membership, enabling the features of fuzzy control with fuzzy reasoning, fuzzy decision-making, self-learning neural networks, adaptive and parallel data processing. The simulation result shows fuzzy neural network control method improves the static and dynamic performance compared to the other two control methods (traditional PI and fuzzy control). The speed increases faster, the overshoot reduces significantly, the motor speed reaches stability quickly at rated speed, added anti-interference capability. The speed control system overall shows high reliability, stability and robustness. In conclusion, the fuzzy neural network control, when used in permanent magnet brushless DC motor speed control system, greatly improves the dynamic and static performance and the anti-interference capability of the system. It is with high practical value.