滑动平均滤波在磁链函数转子位置检测法中的应用与设计

Application and design of moving average filters in rotor position detection based on flux function

基于磁链函数的无刷直流电机无位置传感器控制系统在中低速运行时,电压电流测量噪声会引起误换相,导致系统运行失败。为了准确提取转子位置,利用滑动平均滤波器抑制噪声,保留磁链函数的陡峭变化。为了减小滑动平均滤波器带来的相位延迟,提出在不同的转速范围内采用不同宽度的滑动平均滤波器,并进行了低速实验验证,证明了方法的有效性。

In the of sensorless control system of BLDC based on flux function, the random noise mainly due to that measurement of voltage and current induces fault rotor position and results in system operation failure in low speed. To get the rotor position accurately, moving average filters were used to attenuate random noise and keep the sharp change of the flux function. In order to reduce phase delay caused by moving average filters, different widths were adopted according to motor speed. The validity has been verified experimentally at low speeds.