基于小波包的故障信号特征提取的研究  

Study on the Extraction of Fault Signal Symptom Based on Wavelet Packets

作　　者：王新[1] 黄建[1] 

机构地区：[1]河南理工大学电气工程与自动化学院,河南454003

出　　处：《电子器件》2007年第3期999-1002,共4页Chinese Journal of Electron Devices

基　　金：国家自然科学基金资助(60474043);河南省自然科学基金资助(004060100);河南理工大学博士基金资助(B050801)

摘　　要：讨论了交——交变频调速系统故障诊断的重要性,在当前的检测方法与故障诊断手段研究的基础上,提出了一种改进的小波包算法,并运用该算法对变频调速系统输出电流进行分析.该算法的计算量和占用的存储空间较标准的小波包算法均减少80%.同时,运用基于改进的小波包的频带能量法提取变频调速系统的故障特征,分析了小波包分解层次对特征量提取的影响.经过仿真证明,该方法适用于变频调速系统故障信号的特征量提取.The importance of fault diagnosis for AC-AC frequency conversion drivers is discussed. Based on the present study state at home and abroad, an improved algorithm of the wavelet packets is proposed. The output current in the frequency conversion drivers is analyzed with the improved algorithm of the wavelet packets. The calculating time and the using memory of the improved algorithm is less than that of the standard algorithm of the wavelet packets by 80%. The frequency band energy method based on the wavelet packets analysis is used for the feature extraction of signal. In the wavelet packet decomposition, the impact of the feature extraction of signal is analyzed. It is proved that the improved algorithm of the wavelet packet is feasible by the simulation results.

关 键 词：小波包 交-交变频 小波能量 故障诊断 

分 类 号：TM464[电气工程—电器]