基于LSTM循环神经网络的城轨车辆制动系统故障预测  

Fault Prediction of Urban rail Vehicle Braking System Based on LSTM Cyclic Neural Network

作　　者：赵振乾 王鹏程 ZHAO Zhenqian;WANG Pengcheng(Shaanxi College of Communications Technology,Xi'an Shaanxi 710000,China)

机构地区：[1]陕西交通职业技术学院,陕西西安710000

出　　处：《信息与电脑》2023年第16期67-69,共3页Information & Computer

基　　金：2021年度校级科研项目“城市轨道交通车辆制动系统故障预测方法研究”(项目编号:YJ21003);陕西省教育厅2022年度一般专项科研计划项目(项目编号:22JK0286)。

摘　　要：文章基于长短期记忆网络(Long Short Term Memory,LSTM)循环神经网络,研究了城轨车辆制动系统故障预测问题。通过采集车辆制动系统的传感器数据,建立一个基于LSTM的故障预测模型,并利用该模型预测城轨车辆制动系统中的故障,最后进行实验验证。实验结果表明,所提出的模型可以有效地预测城轨车辆制动系统的故障,具有一定的实用价值。The article is based on Long Short Term Memory(LSTM)recurrent neural network to study the fault prediction problem of urban rail vehicle braking system.By collecting sensor data from the vehicle braking system,a fault prediction model based on LSTM was established,and the model was used to predict faults in the braking system of urban rail vehicles.Finally,experimental verification was conducted.The experimental results indicate that the proposed model can effectively predict the faults of the braking system of urban rail vehicles,and has certain practical value.

关 键 词：城轨交通 车辆制动系统 故障预测 长短期记忆网络(LSTM) 

分 类 号：U279.32[机械工程—车辆工程]