论文部分内容阅读
阐述了椭球单元（ＥｌｉｐｓｏｉｄａｌＵｎｉｔ）网络的原理及其结构，研究了网络权重初始化方法和网络的训练算法，借助这种高阶网络泛化的有界性，针对大型旋转机械多故障同时性诊断问题，构造了一种由多个子网络组成的分级诊断网络（ＨＤＡＮＮ）。测试结果表明：用基于椭球单元网络的ＨＤＡＮＮ网络分级诊断策略解决大规模故障诊断问题是合理有效的，且具有较高的诊断精度，可用于旋转机械工况实时监测和诊断场合。
The principle and structure of the Elipsoidal Unit network are described. The method of network weight initialization and the training algorithm of the network are studied. With the help of the boundedness of the generalization of the higher-order network, aiming at the problem of multi-fault simultaneous diagnosis of large rotating machinery , Constructed a hierarchical diagnostic network (HDANN) composed of multiple sub-networks. The test results show that it is reasonable and effective to solve the large-scale fault diagnosis problem based on the HDANN network hierarchical diagnosis strategy based on the ellipsoid unit network, and it has high diagnostic accuracy and can be used in real-time monitoring and diagnosis of rotating machinery conditions.