摘要
由于已有方法不能完成非线性驱动信息的转换,导致电机驱动系统故障检出率偏低,提出在PLC技术下进行新能源汽车电机驱动系统故障检测的方法。利用小波包分析获取驱动电机频带分量,通过分析各个频带分量的变化提取出故障特征参数。根据故障特征,在PLC技术下,融合核方法和主成分分析方法,将电机驱动故障特征参数非线性信息转换为线性化信息。最后,采用模糊核聚类方法对线性化信息进行聚类处理,完成新能源汽车电机驱动系统故障检测。实验结果表明,所提方法具有较低的漏检率与误检率,同时能够提升故障检出率,实际应用性能较强。
Because the existing methods can not complete the conversion of nonlinear drive information,resulting in the low fault detection rate of motor drive system,a method of fault detection of motor drive system of new energy vehicles under PLC technology is proposed. The frequency band components of the driving motor are obtained by wavelet packet analysis,and the fault characteristic parameters are extracted by analyzing the changes of each frequency band component. According to the fault characteristics,under PLC technology,the nonlinear information of motor drive fault characteristic parameters is transformed into linearized information by integrating kernel method and principal component analysis method. Finally,the fuzzy kernel clustering method is used to cluster the linearized information to complete the fault detection of motor drive system of new energy vehicles.The experimental results show that the proposed method has low missed detection rate and false detection rate,can improve the fault detection rate,and has strong practical application performance.
作者
王良成
汪源
张永辉
WANG Liang-cheng;WANG Yuan;ZHANG Yong-hui(Science and Technology Department,Sanya University,Hainan Sanya 572022,China;School of Information and Communication Engineering,Hainan University,Hainan Haikou 570228,China)
出处
《机械设计与制造》
北大核心
2022年第6期199-202,207,共5页
Machinery Design & Manufacture
基金
海南省教育厅教学改革项目资助—对标“双一流”的校企协同新能源汽车综合实践教学平台建设与共享机制研究(Hnjg2021-100)
海南省教育厅教学改革项目资助—金课建设背景下基于信息技术的理工系列课程建设的研究与实践(Hnjg2021ZD-42)。
