The Application of Heuristic Neural Network in Speed Estimation of Asynchronous Motor

This paper extends the BP neural network predictive control of the speed estimation in the three-phase asynchronous motor vector control system. After analyzing its mathematical model, we improve the input parameters of BP neural network while estimating the motor speed, and replace the stator voltage and current parameters as ANN training input. A new heuristic input method is proposed to preprocess the input parameters of the network due to the mismatch between the input and output frequencies of the traditional ANN. Simulation results demonstrate that the improved scheme has a good tracking performance and static/dynamic performance with a smaller overshoot.