摘要
电动汽车中由于空调压缩机制冷液具有腐蚀性,所以采用无位置无刷直流电机。一般的换相控制通过检测反电动势过零点而间接获得换相时刻,但是由于电机的非线性动态特性、温度、转速等因素的影响,难以得到准确的信息。文中利用神经网络的非线性任意逼近特性,提出一种基于神经元网络的电机相位补偿控制。首先由硬件电路获得有效的反电动势信息,再利用BP神经网络进行正确相位补偿,实现无刷直流电机的无位置传感器控制,获得了较好的效果。
It is necessary to decide the commutation angle properly in the sensorless DC motor control system, however, due to the dynamic nonlinearity, temperature and the rotor speed in the air-conditioner compressor of the electric vehicle, it is difficult to get the right time to commutation. Inthis paper, BP neural networks is used to compensate the commutation angle after the circuits detection of the basic information of the back EMF. This method can obtain the proper time for current to change phase.
出处
《微电机》
2004年第3期43-45,共3页
Micromotors
