基于ARMA模型的受损限幅器模型建立方法

Auto-regressive moving average modeling algorithm of impaired limiter

针对限幅器遭受功能或物理损伤而发生的非线性效应,给出了一种基于信号分析的受损限幅器非线性效应研究新视角,并从该视角出发,提出了模型建立方法。该方法首先测得受损限幅器幅频响应序列,将其转化为零均值实平稳序列;然后计算该序列的自相关函数和偏相关函数,估计自回归滑动平均(ARMA)模型的阶数;最后采用最小二乘法估计出模型的表达式,并采用最终预测误差(FPE)准则对该表达式进行校验。试验结果和理论分析验证了该方法的良好效果。

A new perspective of the impaired limiter based on signal analysis is presented for the nonlinear effect of the limiter subjected to functional or physical damage.For this perspective,the modeling algorithm is proposed.Firstly,an amplitude-frequency response sequence of the impaired limiter is measured,and converted into a zero-mean real stationary sequence.Then,the autocorrelation function and the partial autocorrelation function of the sequence are calculated along with the estimated order of the auto-regressive moving average(ARMA)model which is estimated by least squares method,and verified by the Final Prediction Error(FPE)criterion.Experiment results agree well with theoretical analysis.