永磁同步直线电动机伺服系统具有复杂耦合性、部件参数测量难度大、不确定等特点,使理论与真实模型存在差异.频域特性直观反映系统的响应和运动性能,快速自动频域模型辨识具有重要意义.在dSPACE/MAT-LAB实验平台上产生正弦扫频信号,简便灵活.对比对数扫频实验与稳态步进逐点扫描实验曲线,结果表明对数扫频可以快速精确地获得频域特性.采用最小二乘综合零极点自动优化配置的方法拟合模型,实现快速准确建模.
Some characteristics of permanent magnet synchronous linear motor (PMSLM) like complex coupling, difficult measurement of parameter and uncertain parameters have resulted in the difference between the theoretical model andthe actual one. Automatic and rapid model identification is of great significance as its frequency domain characteristics directly reflect the response and performance of the system. Different types of sine-sweep signals based on dSPACE/ MAT-LAB platform were generated and tested. The frequency response curves generated by logarithmic sweep and stepped sweepwere compared. The experimental results show that logarithmic sweep is a better option for acquiring frequency responsequickly and accurately. Curve fitting with least squares method in low frequency range and automatic zero-pole optimizedconfiguration in middle and high frequency range were implemented to fit the frequency response curve and tune the model.The results validate that the proposed modeling approach is a fast and accurate way.