基于特定频带的Prony暂态振荡检测方法  

The Prony Detection Method of Transient Oscillation Based on the Specific Frequency Band

作　　者：陈晓芳[1] 赵成勇[1] 

机构地区：[1]华北电力大学电力系统保护与动态安全监控教育部重点实验室,河北保定071003

出　　处：《电测与仪表》2011年第5期1-5,共5页Electrical Measurement & Instrumentation

摘　　要：小波分析和Prony算法都是分析暂态信号的有效方法,但小波分析不能得到信号频点信息,Prony算法的实现需要事先判断信号发生时刻。该文通过比较二者的优缺点,结合小波分析时频局部化和Prony算法可以准确获取振荡模式特征的优点,提出基于特定频带的Prony暂态振荡检测方法,克服了基于特定频带的STFT检测方法由于振荡信号衰减过快、频带能量缺失导致的幅值参数检测精度低的问题。该方法利用小波多分辨率分析Mallat算法对暂态振荡信号进行分解重构,选取符合能量要求、包含暂态信号的特定细节空间作为待分析信号;在提取特定频带的基础上,利用Prony算法对信号进行检测。通过仿真对比基于特定频带的STFT和Prony暂态振荡检测新方法在提取信号特征参数上的误差,验证了基于特定频带的prony检测方法的有效性和准确性。Wavelet and Prony algorithm are both the effective methods for the analysis of transient signals.However,wavelet analysis can notgetthe signal frequency information,and Prony algorithmalso need to determine the signal oc-curred moment in advance.In this paper,by comparing the advantages and disadvantages of both,and combining thewavelet analysis with time-frequency localization and the Prony algorithmwith high accuracy to get the transient oscil-lation mode characteristics,a newProny detection method based on the specific frequency band is proposed.It over-comes problem of lowaccuracy of the STFTdetection method based on specific frequency band,when the oscillationsignal decays too fast to reserve energy.First,it utilizes the Mallat algorithm for the transient oscillation signal’s de-composition and reconstruction;and then,selects the appropriate frequency band,including specific details of thetransient signal meeting the energy requirements.Based on the extraction of the specific frequency band,Prony algo-rithm are used to detectthe signal.Simulation is done to compare with the STFTand Prony detection methods based onthe specific frequency band about accuracy of the signal characteristic parameters,and verified the validity and accura-cy of the Prony detection method based on the specific frequency band.

关 键 词：暂态振荡 MALLAT算法 短时傅里叶变换 PRONY算法 

分 类 号：TM935[电气工程—电力电子与电力传动]