无刷直流电机反电势的人工神经元网络预测

BACK ELECTROMOTIVE FORCE PREDICTION USING ARTIFICIAL NEURAL NETWORK FOR BRUSHLESS DC MOTOR

提出了一种采用人工神经元网络（ＡＮＮ）预测永磁无刷直流电机反电势的新方法，给出了用于反电势预测的人工神经元网络的结构，预测的样机反电势波形与实测反电势比较结果表明，该方法是一种快速、简便、准确的反电势预测方法，在无刷直流电机的计算机优化设计及性能分析中有着广阔的应用前景，并对其它类型电机同类问题的解决具有一定的参考价值。

A new back electromotive force prediction method using Artificial Neural Network for Permanent Magnetic Brushless DC Motor is presented in this paper. The structure, training data patterns and training method of the ANN used is discussed. The comparison between predicted EMF wave and experiment result has proved the effectiveness of the ANN predicting method. Its simplicity and exactness warrant its wide usage in the field of the computer aided design and performance analysis of Permanent Magnetic BLDCM.