基于分段滤波的Prony算法在谐波检测中的应用

现代电力系统中的谐波与间谐波频率范围广,直接利用Prony算法辨识得到的结果往往不是很精确;随着信号中谐波含量的增多,模型阶数的变大,计算时间变得过于冗长。针对这些问题,文中提出一种基于分段滤波的Prony算法,首先根据谐波的频率范围进行频段划分,确定各频段的采样参数;然后由低频段到高频段依次进行Prony分析,在进行下一频段检测之前,将已辨识出的所有谐波含量从原始信号中滤除;最后综合各段分析结果,得出各频段谐波的模态参数。仿真试验和应用实例表明,对含量较多且复杂的谐波信号,改进后的方法能有效地提高计算速度和辨识精度。

Detection Method of Harmonic Using Prony Algorithm Based on Segmentation Filtering

The range of harmonic and inter-harmonic frequencies in modern power system is so wide that the direct Prony analysis are often inaccurate;besides,with the increase of harmonic content,the order of model becomes too large and the time of calculation becomes too long.In order to solve these problems,this paper proposes a Prony algorithm based on segmentation filtering.Firstly,the proposed method divides frequency bands based on the range of harmonic frequency,and determines the sampling parameters of various frequency bands.Then use the Prony to analyze the frequency bands from low to high.Before the detection of next band,we have to filter the harmonic which have been detected from the original signal.At last,analyze each bands comprehensively to get the modal parameters of all harmonics.For complex harmonic signal,as the simulation and application examples showed,the proposed method can effectively improve the computing speed and recognition accuracy.