【作者】 毛怿弘； 邹俊忠； 姚晓东； 王行愚；

【Author】 Mao Yihong Yao Xiaodong Zou Junzhong Wang Xingyu Department of Information, Shanghai 200237

【机构】 华东理工大学信息学院；

【摘要】 无刷直流电机(BLDCM)一般是通过位置传感器来产生正确的换相信息来进行控制。由于电动汽车空调压缩机制冷液对霍尔传感器具有腐蚀性,霍尔传感器不宜安装,而且在高温强腐蚀和频繁震动工作环境下,霍尔传感器位置变化和其它干扰会引起霍尔信号畸变,这都将降低控制性能和控制可靠性。因此有必要开发无位置无刷直流电机的控制系统。本文中基于TMS320LF2407DSP芯片,用硬件电路获得有效的转子位置信息,通用BP神经网络的方法得到正确的补偿相位,软硬件相结合从而进行无刷直流电机控制。系统结构简单、可靠,实验结果证明这种方法是可行的,使用TMS320LF2407芯片提高了控制器的实时控制性能,控制效果令人满意。更多还原

【Abstract】 Usually, the accurate commutation information is obtained through the sensors to control the brushless DC motor(BLDCM). However, in the air-conditioner compressor of the electric vehicle, due to the causticity of the refrigeration liquid and the frequent shaking of the vehicle motor, the fixing of position sensors is becoming difficult and the disturbs will also bring about the aberrance of the signals, all of which will destroy the control performance and its reliability. So it is necessary to develop the sensorless BLDCM control system. In this paper, a new hybrid solution is presented to achieve reliable sensorless BLDCM control with relative concise and low cost circuits as well as little software expense based on the TMS320LF2407. In order to get the right commutation instant. The neural network is used to compute the compensate time. The experimental results show that the scheme is effective and satisfactory.更多还原