基于FFT的伺服系统谐振频率特性的在线辨识

Online Identification of Resonance Characteristic Based on FFT in Servo System

交流伺服系统中弹性传动装置引起的机械振荡严重影响其稳定性和加工精度，目前商业用伺服主要采用陷波滤波器来抑制机械谐振。为了正确配置陷波滤波器的参数以达到理想的抑制效果，准确获取伺服系统机械谐振频率特性至关重要。本文分别用离线快速扫频法和基于FFT的在线辨识法实现了伺服系统的谐振频率特性辨识。在线辨识算法在DSP芯片TMS320F28335上实现，实验结果表明，此算法易于实现，运行快速准确，适合实际工程应用。

The mechanical vibrations caused by the elastic transmission device in AC servo system seriously affects the system stability and precision. The current commercial servo mainly uses the notch filter to suppress mechanical resonance. In order to configure the parameters of notch filter properly to get the desired effects, it’s of critical importance to obtain the resonance frequency characteristic of the servo system. This paper achieves the identification of resonant frequency by using methods ofoff-line fast frequency sweep and online FFT. The online identification algorithm is realized on DSP TMS320F28335, theexperimental results show that this algorithm is fast, accurate and easy to implement , so it’s suitable for practical engineering applications.