低频振动模态参数辨识方法与预警指标研究

Study on Identification Method and Early Warning of Low Frequency Oscillation Modal Parameter

为实现环境激励下多通道随机响应的低频振荡模态辨识及预警,将航空航天领域用于分析航天器结构振动特性的特征系统实现算法(ERA)与随机减量技术(RDT)相结合。利用RDT从随机响应系统中提取自由振荡信号,进而采用ERA算法对得到的信号进行辨识,可获取低频振荡频率,阻尼比等参数,并将其应用于预警系统。通过仿真实验与ESPRIT和Prony方法的对比分析,表明该方法对于多通道信号能一次性进行低频振荡模式识别,并具有更高的计算速度、辨识精度和抗噪能力。该方法给多通道低频振荡模态参数辨识提供了一种更加快速准确有效的手段,能够满足电力系统低频振荡在线辨识与预警的要求,具备很好的应用前景。

Combining ERA(eigensystem realization algorithm) and RDT(random decrement technique) which are previously used to analyze the vibration characteristics of spacecraft structures to realize low frequency oscillation mode identification and early warning based on multi-channel random response under environmental excitation.We use RDT to extract free oscillation signals from the random response system,and then use the ERA algorithm to identify low frequency oscillation in the data obtained.Damping ratio and other parameters can be obtained and later used in early warning system.Comparing the proposed method with ESPRIT and Prony method via simulation,results show that the proposed method can identify low-frequency oscillation modes for multi-channel signals at one time and with higher calculation speed,identification accuracy and noise rejection.This method provides a more rapid,accurate and effective method for multi-channel low-frequency oscillation modal parameter identification.It complies the requirements for on-line identification and early warning of low-frequency oscillation in power system and has a good application prospect.