基于模糊神经网络的高速列车悬挂系统故障诊断

Fuzzy Neural Network Based High-speed Train Suspension System Fault Diagnosis

以高速列车垂向悬挂系统为研究对象,提出了一种基于模糊神经网络的故障诊断方法。首先通过多体动力学软件SIMPACK搭建仿真平台,在加上轨道不平顺情况(包括横向、垂向、侧滚)下获得实验数据,再将实验数据导入模糊神经网络中,对模糊神经网络进行训练。通过对比预测输出和实际输出,从而验证了模糊神经网络在高速列车悬挂系统的故障诊断中的可行性,并具有一定准确度,为高速列车悬挂系统故障诊断提供了良好的方法。

This paper based on the high-speed train vertical suspension system as the research object and concerns a diagnosis method based on fuzzy neural and verified it.The proposed method is divided into four steps：Set up a simulation platform by using simulation tool SAMPACK.Combined with track irregularity（Lateral,Vertical,Roll）to acquire the experimental data.The effectiveness of the proposed approach is demonstrated by compare the predicted output with the actual output,and the algorithm is also accuracy.