基于最小二乘法的电机转速测量

Motor Speed Measurement Using the Least-squares Fit

伺服控制系统中,转速测量的准确度和实时性非常重要,但传统M法测速时,测量精度和响应速度往往是一对不可调和的矛盾。为此将最小二乘曲线拟合应用于电机转速测量,即通过编码器获取转子位置,然后利用最小二乘法拟合得到位置函数,进而求导得到电机转速。比较了两种不同阶次最小二乘法的测速特点,最终提出了一种综合最小二乘法来提高测速性能。仿真和实验对比了该法和M法的测速性能,结果表明这种方法在提高测量精度的同时,也保证了测量的实时性。

The accuracy and response time of measuring speeds are very significant in the servo control system.But in traditional fix-time method（M method）, the accuracy and response time are incompatible. To mitigate theproblem, a method to estimate speed using the least-squares fit was adopted. This method was to use optical encoder toget a series of rotor position data,and then obtained an estimation function of positions using the least-squares fit, andthe speed could be further obtained by differentiation. Two least-squares fits of different degrees were compared andthen a new version of the method was proposed by comprehensively using different least-squares fits to improve themeasurement performance. Simulation and experiment were carried out to compare the performance of the new methodwith that of the M method. Results show that the new method has an improved accuracy while maintaining the sameresponse time.