在机械故障诊断学科的发展过程中，人们发现最重要、最关键而且也是最困难的问题之一是故障特征信息提取（简称特征提取）。在某种意义上，特征提取可以说是当前机械故障诊断研究中的瓶颈问题，它直接关系到故障诊断的准确性和故障早期预报的可靠性。为了从根本上解决特征提取 In the course of the development of mechanical fault diagnosis, it is found that one of the most important, the most critical and the most difficult one is the extraction of fault feature information (referred to as feature extraction). In a sense, feature extraction can be said to be a bottleneck in the current mechanical fault diagnosis research, which is directly related to the accuracy of fault diagnosis and fault early warning of reliability. In order to fundamentally solve the feature extraction