期刊文献+

基于支持向量机的地铁故障类型预测 被引量:1

Subway fault type prediction based on support vector machine
下载PDF
导出
摘要 针对城市轨道交通运营压力增大带来的故障隐患和诸多问题,基于城市轨道交通故障隐患数据,构建支持向量机实验预测模型,将筛选处理后的数据与模型相结合,探究未来一定时间段内城市轨道交通可能发生故障的概率和类型,在不同维度、数据量的情况下,对模型运算的效果进行分析,找出其对模型准确率的影响,确定最佳数据维度和使用数据量,为运营公司应对突发故障提供决策辅助方法.以现有数据通过支持向量机模型进行概率预测估计,同时加入噪声数据测试对模型准确率的影响,最后与BP神经网络模型和极限学习机方法进行对比.研究结果表明:本文构建的模型进行地铁故障概率预测的准确率能够保持在60%左右,说明具有一定的可行性,同时能够在含有噪声数据的情况下进行故障发生概率预测. In view of the potential failures and many problems caused by the increasing operation pressure of urban rail transit,this paper builds an experimental prediction model of support vector machine based on the potential failure data of urban rail transit.The filtered and processed data is combined with the model to explore the future probability and types of possible failures in urban rail transit during a period of time.The effect of model operation is analyzed under different dimensions and data volumes to identify its impact on the accuracy of the model,determine the best data dimension and the amount of data used,thus providing decision-making methods for operating companies to deal with sudden failures.The existing data are used to estimate the probability prediction by the support vector machine model.At the same time,the noise data are added to test the impact on the accuracy of the model.Finally,it is compared with the BP Neural Network model and the Extreme Learning Machine method.The research results show that The accuracy of the model built in this paper for subway fault probability prediction can be maintained at about 60%,which shows that it is feasible,and can predict the failure probability with noise data.
作者 张仪鹏 周玮腾 韩宝明 ZHANG Yipeng;ZHOU Weiteng;HAN Baoming(School of Traffic and Transportation,Beijing Jiaotong University,Beijing 100044,China)
出处 《北京交通大学学报》 CAS CSCD 北大核心 2023年第1期90-97,共8页 JOURNAL OF BEIJING JIAOTONG UNIVERSITY
基金 中央高校基本科研业务费专项资金(2022JBMC057) 北京市自然科学基金(L201013) 交控科技设计创新和学科发展基金(9907006511)。
关键词 城市轨道交通 故障预测 机器学习 支持向量机 urban rail transit fault prediction machine learning support vector machine
  • 相关文献

参考文献8

二级参考文献45

共引文献79

同被引文献5

引证文献1

二级引证文献1

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部