论文部分内容阅读
本文介绍了一种模糊神经网络分类器，针对其在旋转机械工况辨识中的应用，论述了网络的输入和输出模糊化问题。文中就旋转机械中几种典型的故障模式采用模糊神经网络进行了识别，并与传统的ＢＰ网络进行了比较。研究结果表明：模糊神经网络方法应用于旋转机械工况识别是有效的，它比传统ＢＰ网络更适合于处理分类边界模糊的数据。
This paper introduces a fuzzy neural network classifier, in view of its application in the identification of rotating machinery working conditions, discusses the network input and output fuzzification problems. In this paper, several typical failure modes of rotating machinery are identified by fuzzy neural network, and compared with the traditional BP network. The results show that the fuzzy neural network method is effective in rotating mechanical condition recognition. It is more suitable than the traditional BP neural network in dealing with fuzzy boundary data classification.