论文部分内容阅读
基于小波包变换能够把平稳和非平稳信号根据它们的时频特性分解到不同层次上不同频道内，与模糊数学能够处理工程中的不确定性问题，以及ＭＡＸＮＥＴ聚类结构的网络具有自组织聚类分析的优点，提出了小波包模糊聚类网络．该网络能够处理平稳和非平稳信号的不确定性问题，并且具有自适应、自组织聚类分析功能．最后举例说明了该网络在机械诊断实践中是一种行之有效的智能分类器．
Based on the wavelet packet transform, the stationary and non-stationary signals can be decomposed into different channels at different levels according to their time-frequency characteristics, and the fuzzy mathematics can deal with the uncertainties in engineering. The network of MAXNET clustering structure has the advantages of self-organizing Class analysis of the advantages of wavelet packet fuzzy clustering network. The network can deal with the uncertainty of stationary and non-stationary signals and has adaptive and self-organizing clustering analysis. Finally, an example is given to illustrate that this network is an effective intelligent classifier in the practice of mechanical diagnosis.