高速列车轴承复杂声学环境下道旁故障诊断关键理论研究

项目名称： 高速列车轴承复杂声学环境下道旁故障诊断关键理论研究

项目编号： No.51475441

项目类型： 面上项目

立项/批准年度： 2015

项目学科： 机械、仪表工业

项目作者： 何清波

作者单位： 上海交通大学

项目金额： 80万元

中文摘要： 高速列车速度不断提高以及无缝钢轨膨胀伸长导致的横向变形使得列车轴承承受着更大的动载和横向力，轴承发生故障的风险也随之上升。因此，列车轴承状态监测和故障诊断对于保障列车安全运行，减少人员财产损失具有重要的意义。目前，国内外对列车轴承道旁声学故障诊断的理论研究基础薄弱，发表文献很少。本项目针对此问题开展研究，首先从麦克风阵列采集到的复杂声学信号中分割和提取与每个轴承相关的信号；其次对分割导致的碎片化信号进行拼接和融合以得到能够反映轴承状态的完整信号；最后对具有小样本特征的轴承信号建立自学习专家系统，以了解和跟踪每个轴承的状态变化发展过程并对其健康进行监测和诊断。本研究旨在通过发展相应的理论方法探讨复杂声学环境下有效的列车轴承道旁故障诊断关键理论问题，为最终防范高速列车灾难性事故做出贡献。

中文关键词： 列车轴承；复杂声学环境；道旁故障诊断；信号碎片化；小样本

英文摘要： The high-speed train's speed becomes larger, and the transverse deformation occurs in the seamless rail because the steel expands with heat. These two factors lead to that the train bearing endures heavier dynamic load and transversal force, and finally lead to the rise of bearing fault risk. Thus, train bearing condition monitoring and fault diagnosis are significant in guarantee of safe train operation and reducing life property loss. At present, the theoretical basis of train bearing wayside acoustic fault diagnosis is still weak, and the relevant literatures are lack at home and aboard. This project is committed to address this issue, firstly, the bearing signals are segmented and extracted from the microphone array signals that were acquired from the complex acoustic environment; subsequently, the fragmental signals are spliced and merged to obtain the complete signal which reflects the bearing condition; finally, a self-learning expert system is constructed for the bearing signals with small sample characteristics, and then each bearing's time-varying conditions can be monitored and bearing fault diagnosis can be achieved. This project aims to develop the key theory methodologies for train bearing wayside fault diagnosis under the complex acoustic environment, and finally prevent the high-speed train catastrophes.

英文关键词： train bearing;complex acoustic environment;wayside fault diagnosis;signal fragmentary;small sample