文献类型：期刊文献

中文题名：基于有效性评价机制的小波包特征提取技术

英文题名：Feature Extraction on the Basis of Effective Evaluation

作者：张颖[1];姚晓东[1]

机构：[1]华东理工大学信息工程学院,上海200237

年份：2010

卷号：26

期号：4

起止页码：195

中文期刊名：微计算机信息

外文期刊名：Control & Automation

语种：中文

中文关键词：小波包变换;特征提取;信号处理;压缩机

外文关键词：lifting wavelet packet transformation; feature extraction; signal processing; compressor

摘要：本文介绍了如何应用提升小波包变换对信号进行特征提取,并在此基础上提出了四条定量的评价标准,能够全面地对此类特征提取方法的有效性进行评价。通过这四个标准,就能更科学地选取合适的特征小波包,从而进一步提高原方法的效率,减少不必要的计算复杂度,使之更加适用于压缩机信号的实时监测。
Feature extraction on the basis of lifting wavelet packet transformation is introduced. Four assessment criterions are proposed, through which the effectiveness of feature extraction can be assessed completely and the chosen of feature wavelet packet will be more scientific and practical. Feature extraction and assessment can upgrade the effectiveness of original wavelet packet algorithm with the decrease of calculation. Through that, wavelet packet analysis can be applied to compressor monitoring more effective.