基于支持向量机的汽车制动防抱死系统故障检测方法  

Fault Detection Method for Automotive anti Lock Braking System Based on Support Vector Machine

作　　者：冯俊侃 罗永星 胡剑锋 秦园 Feng Jun-kan;Luo Yong-xing;Hu Jian-feng;Qin Yuan(Yiwu Zhengjie Motor Vehicle Performance Testing Co.,Ltd,Yiwu 322000,China;Taizhou Hongxing Vehicle Testing Co.,Ltd,Taizhou 317700,China;Hangzhou Hongbang Automotive Testing Co.,Ltd,Hangzhou 310000,China;Haining Junda Motor Vehicle Testing Service Center Co.,Ltd,Jiaxing 314050,China)

机构地区：[1]义乌市正捷机动车性能检测有限公司,浙江义乌322000 [2]台州红兴车辆检测有限公司,台州317700 [3]杭州鸿邦汽车检测有限公司,杭州310000 [4]海宁骏达机动车检测服务中心有限公司,嘉兴314050

出　　处：《内燃机与配件》2024年第18期78-80,共3页Internal Combustion Engine & Parts

摘　　要：为实现对防抱死系统(ABS)故障的精准检测,降低汽车行驶安全隐患,引进支持向量机(SVM)展开分析。在仅考虑汽车纵向运动的条件下,建立汽车行驶过程的动力学方程,提取ABS系统运行的力学特征参数;将SVM作为二分类方法,设定原始数据集合,应用SVM进行超平面信息检索,实现基于最小化经验的参数学习、参数的学习与决策函数的构建;建立系统正常运行因变量参考序列,根据参数的时间序列分布,进行系统序列的无量纲比较,以此实现系统故障的诊断与检测结果输出。对比实验结果表明:设计的故障检测方法实际应用效果良好。In order to achieve precise detection of anti lock braking system(ABS)faults and reduce driving safety hazards,support vector machine(SVM)is introduced for analysis.Under the condition of only considering the longitudinal motion of the car,establish the dynamic equation of the car driving process and extract the mechanical characteristic parameters of the ABS system operation;Using SVM as a binary classification method,setting the original data set,applying SVM for hyperplane information retrieval,achieving parameter learning based on minimizing experience,parameter learning,and decision function construction;Establish a reference sequence of dependent variables for the normal operation of the system,and perform dimensionless comparison of the system sequence based on the time series distribution of parameters,in order to achieve the diagnosis and detection results output of system faults.The comparative experimental results show that the designed fault detection method has good practical application effects.

关 键 词：支持向量机 决策函数 检测方法 故障 制动防抱死系统 汽车 

分 类 号：U463.526[机械工程—车辆工程]