基于小波神经网络的电机转子位置预测模型
novel wavelet neural network for rotor position detecting
无刷直流电机的控制中，转子位置的检测至关重要。基于小波函数与神经网络的优点提出了一种电机转子位置预测方法，可以用于无刷直流电机转子位置的检测，克服了传统检测方法检测范围狡窄的缺陷。对检测模型的原理以及算法进行了阐述，设计出了模型网络拓扑结构。应用遗传算法对模型进行训练，得出了模型的各个参数。应用MATLAB对进行了电机转角预测仿真实验，仿真结果与实际结果进行比对，证明此检测模型可以较好的完成电机转子位置的预测，具有更高的精度和检测范围。
he rotor position detecting is important in the control of brushless direct current motor. On basis of the advantage of neural network (NN) and wavelet function, a wavelet neural network (WNN) was designed for brushless direct current motor\'s rotor position detecting. It overcame the shortage of the traditional methods. The algorithm of WNN was given and the topology of the network was designed. Parameters were settled by training with genetic algorithm (GA). Some emulation experiments were done. The results prove that this WNN can detect the rotor position correctly and has higher precision and detection range.
胡沥丹、尹雯、袁云
电机
无刷直流电机遗传算法小波神经网络转子位置检测
brushless direct current motor (BLDCM)Genetic Algorithmwavelet neural networkrotor position detecting
胡沥丹,尹雯,袁云 .基于小波神经网络的电机转子位置预测模型[EB/OL].(2010-06-24)[2025-08-02].http://www.paper.edu.cn/releasepaper/content/201006-457.点此复制
评论