基于多源数据融合的电液车辆减速器监测系统研究  

Research on Electro-hydraulic Car Retarder Monitoring System Based on Multi-source Data Fusion

作　　者：胡淼 徐永新 宋少鹏 刘树栋 邱战国 安家丙 HU Miao;XU Yongxin;SONG Shaopeng;LIU Shudong;QIU Zhanguo;AN Jiabing

机构地区：[1]中国铁道科学研究院集团有限公司通信信号研究所,北京100081 [2]中国铁路北京局集团有限公司电务部,北京100045 [3]中国铁路北京局集团有限公司北京西电务段,北京100070 [4]北京华铁信息技术有限公司,北京100081

出　　处：《铁道通信信号》2025年第2期6-13,共8页Railway Signalling & Communication

基　　金：中国铁路北京局集团有限公司科技研究开发计划(2024CD07);中国铁道科学研究院集团有限公司基金课题(2022YJ217)。

摘　　要：为推动车辆减速器维修方式向更加高效、精准的状态修转变,设计了基于多源数据融合的电液车辆减速器监测系统。该系统集成数据采集分析和健康评估功能,通过整合压力、液位、温度、换向阀电信号、位置表示信息等多源数据,提升设备故障识别精度。首先通过数据清洗去除异常、跳变数据,归一化处理不同量纲数据,划分数据集并建立特征库;其次采用基于神经网络构建的Softmax回归模型进行数据融合分析,配合使用交叉熵损失函数,通过反向传播方法进行训练;最后结合双隐层结构和Relu激活函数,确保模型的收敛性和非线性处理能力,提高模型预测的准确性。试验结果表明,系统识别准确率达98.1%,为车辆减速器维修模式的转型提供了有力的技术支撑,也为电务基础设备智能化监测提供借鉴。In order to promote the transformation of the repair mode of car retarder to more efficient and precise condition-based repair,a electro-hydraulic car retarder monitoring system based on multi-source data fusion is designed.The system integrates data acquisition and analysis with health assessment functions,and improves the recognition accuracy of equipment faults by integrating multi-source data such as pressure,liquid level,temperature,electrical signal of reversing valve and position indication.Firstly,the abnormal and fluctuating data are removed by data cleaning,the different dimensional data are normalized,the data sets are divided and a feature library is established.Secondly,the neural-network-based Softmax regression model in conjunction with the cross-entropy loss function is used for data fusion analysis and trained using the back propagation method.Finally,by combining with the double hidden layer structure and the Relu activation function,the convergence and nonlinear processing ability of the model are ensured,and the prediction accuracy of the model is improved.The test results show that the recognition accuracy of the system is up to 98.1%,which provides a strong technical support for the transformation of the repair mode of the electro-hydraulic car retarder and offers a reference for intelligent monitoring of basic signal equipment.

关 键 词：电液车辆减速器 状态修 状态监测 多源数据融合 Softmax回归模型 健康评估 故障检测 

分 类 号：U284.63[交通运输工程—交通信息工程及控制]