摘要
为解决有轨电车用燃料电池电堆系统（fuelcellstack system,FCSS）故障分类问题,提出基于多分类相关向量机（multi-class relevance vector machine,m RVM）和模糊C均值聚类（fuzzy C means clustering,FCM）的有轨电车用FCSS故障诊断新方法。该方法利用FCM形成标准聚类中心,采用m RVM对测试样本实现多分类,能有效剔除奇异数据并提高模型分类正确率。实例分析表明,所提方法可快速识别氢气泄漏、去离子水加湿泵低压、空气压力过低和正常共4种健康状态,分类准确率可达96.67%,为有轨电车用FCSS在线故障诊断研究提供参考。
In order to solve the fault classification problem of fuel cell stack systems （FCSS） for tramways, a new fault diagnosis method for FCSS of tramways based on multi-class relevance vector machine （mRVM） and fuzzy C means clustering （FCM） was proposed. FCM was used to form the standard clustering center and mRVM was adopted to classify the test samples. This method can effectively eliminate the singular data and improve the classification accuracy of the model. The results show that the proposed method can quickly identify the leakage of hydrogen, the low pressure of deionized water humidification pumps, the low pressure of input air and the normal health state with the classification accuracy of up to 96.67%. The proposed method is an important foundation for on-line fault diagnosis of FCS S for tramways.
作者
刘嘉蔚
李奇
陈维荣
燕雨
LIU Jiawei;LI Qi;CHEN Weirong;YAN Yu(School of Electrical Engineering,Southwest Jiaotong University,Chengdu 611756,Sichuan Province,China)
出处
《中国电机工程学报》
EI
CSCD
北大核心
2018年第20期6045-6052,共8页
Proceedings of the CSEE
基金
国家自然科学基金项目(61473238,51407146)
四川省杰出青年基金项目(2015JQ0016)~~
关键词
多分类相关向量机
质子交换膜燃料电池
故障诊断
混合动力有轨电车
模糊C均值
multi-class relevance vector machine （mRVM）
proton exchange membrane fuel cell
fault diagnosis
hybridtramway
fuzzy C means clustering
