基于无监督神经网络的故障模式识别

Fault Pattern Recognition Based on Unsupervised Neural Network

无监督学习的竞争式神经网络是一种数据聚类方法,能保持输入空间的拓扑关系不变,借助于一维或二维输出平面的一组有序的向量,实现高维数据的聚类和可视化;探讨了一种无监督神经网络———SOFM网络原理、思想和算法步骤,研究了无监督网络在模式识别中的应用,提出了基于SOFM网络的故障模式识别和状态监测方法;通过实例研究了SOFM网络在机械设备故障模式识别和状态监测中的应用。

The unsupervised learning network is a data cluster method, it can transform an incoming signal pattern of arbitrary dimension into a one-or two-dimensional discrete map, and perform this transformation adaptively in a topologically ordered fashion. The prin- ciple and algorithm of a unsupervised neural network--SOFM is be discussed, the unsupervised network is applied to pattern recognition of the bearing fault, the new pattern recognition technique has been introduced, the high--dimensional input vectors are projected into a two-dimensional space. Numeric experimentation results show reasonable agreements to that the proposed method could recognize the fault pattern clearly and distinctly than common visualization method.