一种在线分离突发性脉冲干扰的小波分析法  

A Wavelet Analysis Method for SeparatingInterference of Sudden Pulse in On-Line Monitoring

作　　者：彭月祥[1] 杨少辰[1] 朱维新[2] 李铁香[2] 

机构地区：[1]北方交通大学物理系,北京100044 [2]北京石油化工学院自动化系,北京102617

出　　处：《数据采集与处理》2003年第1期36-39,共4页Journal of Data Acquisition and Processing

摘　　要：选取合适的奇异性小波基 ,抽取信号与小波基卷积序列中的某一项作为奇异性判断指标。从理论上找到了对有效信号变化最敏感的条件 ,并按照此条件推导出信号与噪声的小波变换模极大能有效地被区分。仿真结果表明 ,在合理选取判断阈值和进行小波变换的信号长度时 ,该方法能快速、准确地判断出真正的信号跳变。本文对提高测控仪表的可靠性有一定的借鉴作用。A new on-line wavelet analysis method for separating sudden pulse interference is presented. Strong pulse interference usually makes a mistaken operation of measuring-control instrument. However, this interference possesses less pulse width than real signal jump. So the singularities of these two singular signals are different too. This paper selects a proper wavelet base and a specific item as the judgment target of singularity from the sequence of convolution of signal with wavelet base. The condition which is most sensitive to the singularity of real signal jump against interference is found in theory. Under this condition, the wavelet transform modulus maximum of real signal jump can be effectively distinguished from that of pulse interference. Simulational result indicates that the real signal jump can be sped and accurately detected by this method if the judgement threshold and the length of signal used in wavelet-transform are properly selected, and the method has a certain reference effect on enhancing the reliability of measuring-control instrument.

关 键 词：小波分析 在线监测 信号处理 

分 类 号：TN911.4[电子电信—通信与信息系统]