基于赋权Apriori算法的新型列车运行控制系统故障定位方法  

Fault localization of novel train control systems based on weighted apriori algorithm

作　　者：张雁鹏[1] 左兴 ZHANG Yanpeng;ZUO Xing(School of Automation and Electrical Engineering,Lanzhou Jiaotong University,Lanzhou 730070,China)

机构地区：[1]兰州交通大学自动化与电气工程学院,甘肃兰州730070

出　　处：《铁道科学与工程学报》2024年第10期4226-4239,共14页Journal of Railway Science and Engineering

基　　金：国家自然科学基金资助项目(52062028);甘肃省高等学校青年博士基金资助项目(2022QB-068)。

摘　　要：新型列车运行控制系统作为保障我国西部高原铁路运营安全的核心技术装备,具有通信网络复杂、故障模式多、参数输入域广、参数维度高等特点。针对西部高原铁路新型列车运行控制系统多输入参数组合测试问题,选取临时限速场景,提出一种赋权关联规则的故障定位方法。首先,分析新型列车运行控制系统结构及临时限速功能的实现过程,按照层次化理论建立临时限速场景参数输入表,生成测试用例;然后,根据测试用例执行结果,以欧氏距离模型分别计算输入参数组合之间和测试用例之间的相似度,并为各个输入参数组合分配故障致因权重;接着,在输入参数组合与执行结果之间挖掘带权重的定向关联规则,得到不同维度下达到关联规则阈值的故障诱因可疑度,可疑度最高的n-way组合即为故障定位结果;最后,依托京张智能高速铁路列车运行控制系统仿真测试平台,验证所提方法的有效性。实验结果表明:临时限速场景下,故障组合与非故障组合之间的可疑度差值大于0.2,非故障组合之间的可疑度差值小于0.06,故障组合与非故障组合之间的可疑度差异明显;所提方法与FIC-BS、IterAIFL和Delta-debugging方法相比较,查全率、查准率和综合性评价指标的均值分别达到85%、70%和83%,均高于对比的方法,具有更好的故障定位效果。该方法可为新型列车运行控制系统功能测试提供一定的参考。Novel train control systems are of crucial importance for western plateau railway in China to make sure safe train operations,which have the intrinsic characteristics of complex communication networks,multiple fault modes,and input parameters with broad range and high dimensions.Aiming at the problem of multiple input parameters in combinatorial test for novel train control systems,temporary speed restriction(TSR)was taken as a typical scenario and a method of fault localization based on weighted association rules in data mining was proposed.First,the architecture of novel train control systems and the implementation process of TSR was thoroughly analyzed.The input parameter table of TSR was established in accordance with hierarchical theory to generate test cases.Subsequently,from the executing results of test cases,the similarity between input parameter combinations or between test cases was calculated by Euclidean distance formula,and fault-causing weights were assigned to each input parameter combination.Furthermore,directed association rules with weights were mined between the input parameter combinations and the executing results to achieve the fault-causing suspicion degree beyond the threshold of association rules in different dimensions.The n-way combination with the highest suspicion degree was also the result of fault localization.Finally,supported by the simulation testing platform of Chinese train control systems(CTCS)for Beijing-Zhangjiakou intelligent high-speed rail,the effectiveness of the proposed method was verified.The experimental results show that under the TSR scenario,the difference in suspicion degree is greater than 0.2 between fault combinations and non-fault combinations,and less than 0.06 between non-fault combinations,demonstrating a significant difference in suspicion degree between fault combinations and non-fault combinations.Compare with FIC-BS,IterAIFL and Delta-debugging,the average of evaluation indices involving recall,precision and f-measure in the proposed method reach 85%,

关 键 词：中国列车控制系统 故障定位 临时限速 组合测试 关联规则 

分 类 号：U284.48[交通运输工程—交通信息工程及控制]