韩耀辉,刘波澜,王文泰,等.基于支持向量机的柴电混合动力故障诊断研究[J].内燃机工程,2022,43(1):101-108.
基于支持向量机的柴电混合动力故障诊断研究
Research on Fault Diagnosis of Diesel Electric Hybrid Based on Support Vector Machine
DOI：10.13949/j.cnki.nrjgc.2022.01.012
关键词:柴电混合动力系统  支持向量机  故障诊断  实时仿真
Key Words:diesel electric hybrid system  support vector machine  fault diagnosis  real-time simulation
基金项目:
作者单位E-mail
韩耀辉* 北京理工大学 机械与车辆学院，北京 100081 hanyaohui2019@163.com 
刘波澜* 北京理工大学 机械与车辆学院，北京 100081 liubolan@bit.edu.cn 
王文泰 北京理工大学 机械与车辆学院，北京 100081  
刘凡硕 北京理工大学 机械与车辆学院，北京 100081  
张俊伟 北京理工大学 机械与车辆学院，北京 100081  
摘要点击次数: 1340
全文下载次数: 555
摘要:对柴电混合动力系统级故障诊断进行了研究，利用仿真软件搭建了实时整车模型，并构建了基于支持向量机的柴电混合动力系统的诊断框架。采用一对一方法构建多分类器，故障识别准确率达到 98%。构建了柴电混合动力系统故障诊断实时仿真平台，进行了基于支持向量机的柴电混合动力系统故障诊断实时仿真，验证了实时环境下基于支持向量机诊断算法能有效实现对混合动力系统多故障并发模式诊断。
Abstract:The system level fault diagnosis of diesel electric hybrid power was studied. A real-time vehicle model was built by using the GT-Suite software to meet the accuracy requirements， and the diagnosis framework of diesel electric hybrid system based on support vector machine（SVM） was constructed. The one-verse-one（OVO） method was used to construct multiple classifiers， and the accuracy of fault recognition was 98%. The real-time simulation platform of diesel electric hybrid system fault diagnosis was constructed， and the fault diagnosis real-time simulation of diesel electric hybrid system based on SVM was carried out. Results show that the diagnosis algorithm based on support vector machine can effectively realize the multi fault concurrent mode diagnosis of hybrid system in real-time environment.
查看全文  HTML   查看/发表评论