基于证据融合算法的地铁车辆轴承故障检测方法研究
|
摘要
转向架轴承作为地铁车辆的重要组成部分,其故障会对车辆运行安全构成严重威胁。现有的轴承故障诊断方法准确性较低,存在漏诊、误诊的可能。针对这一问题,分析地铁车辆转向架轴承的故障模型,对经典证据融合算法进行改进,提出一种新的均值加权融合法,该方法提取轴承振动信号的特征频率谱峰比值与频率均值,设置对应的概率分配函数来分配上述两种特征参数对不同元素的支持度,并进行证据融合。为了验证所提出方法的有效性,设计搭建了轴承实验平台,对已有的均值K系数法、吸收法及所提出的均值加权融合算法进行了对比实验。实验结果表明,本文提出的均值加权融合算法的漏诊率及误诊率明显小于已有的均值K系数法和吸收法,提高了地铁车辆轴承诊断的准确率,更加适用于地铁车辆转向架轴承的故障诊断。
The failure of bogie bearing, an important part of metro vehicles, will pose a serious threat to the safety of vehicle operation. Due to the low accuracy of the existing fault diagnosis methods, there may be misdiagnosis or missed diagnosis. To address this deficiency, this paper first analyzed the fault model of the metro vehicle bogie bearing. Based on the improvement of the classical evidence fusion algorithms, a new mean weighted fusion algorithm was proposed, which extracted the characteristic frequency spectrum peak ratio and the mean value of frequency of bearing vibration signal. The corresponding probability distribution functions were set to distribute the support degree of these two characteristic parameters to different elements and the evidence was fused. To verify the effectiveness of the proposed method, the bogie bearing experimental platform was designed and constructed. The mean K coefficient algorithm, the Absorption algorithm and the proposed mean weighted fusion algorithm were compared on that platform. The results show that missed diagnosis rate and misdiagnosis rate of the mean weighted fusion algorithm proposed in this paper are significantly lower than those of the other two existing improved algorithms, which improves the accuracy of the metro vehicle bearing diagnosis and is more applicable to the fault detection of metro vehicle bogie bearings.
The failure of bogie bearing, an important part of metro vehicles, will pose a serious threat to the safety of vehicle operation. Due to the low accuracy of the existing fault diagnosis methods, there may be misdiagnosis or missed diagnosis. To address this deficiency, this paper first analyzed the fault model of the metro vehicle bogie bearing. Based on the improvement of the classical evidence fusion algorithms, a new mean weighted fusion algorithm was proposed, which extracted the characteristic frequency spectrum peak ratio and the mean value of frequency of bearing vibration signal. The corresponding probability distribution functions were set to distribute the support degree of these two characteristic parameters to different elements and the evidence was fused. To verify the effectiveness of the proposed method, the bogie bearing experimental platform was designed and constructed. The mean K coefficient algorithm, the Absorption algorithm and the proposed mean weighted fusion algorithm were compared on that platform. The results show that missed diagnosis rate and misdiagnosis rate of the mean weighted fusion algorithm proposed in this paper are significantly lower than those of the other two existing improved algorithms, which improves the accuracy of the metro vehicle bearing diagnosis and is more applicable to the fault detection of metro vehicle bogie bearings.
引文
[1] 李培玉,郑俊.基于多通道振动信号的港机车轮轴承故障诊断[J].轴承,2007(9):35-38.LI Peiyu,ZHENG Jun.Fault Diagnosis Based on Multi-channel Vibration Signal for Port Machine Trolley Wheel Bearings[J].Bearing,2007(9):35-38.
[2] 邢宗义,俞秀莲,程晓卿,等.基于证据理论的城轨车辆走行系融合故障诊断[J].机电工程,2014,31(12):31-32.XING Zongyi,YU Xiulian,CHENG Xiaoqing,et al.Approach for Fault Diagnosis of the Metro Vehicle Running Gears Based on Evidential Theory[J].Journal of Mechanical & Electrical Engineering,2014,31(12):31-32.
[3] 文振华,陈果.基于D-S证据理论的航空发动机磨损故障智能融合诊断方法[J].机械科学与技术,2005,24(9):1018-1020.WEN Zhenhua,CHEN Guo.An Intelligent Fusion Technique for Diagnosis of Engine Wear Fault Based on D-S Evidence Theory[J].Mechanical Science and Technology,2005,24(9):1018-1020.
[4] 李伟彤.一种冲突情况下的自适应证据理论合成公式[J].弹箭与制导学报,2006,26(3):242-243.LI Weitong.A Kind of Adapting Evidence Theory Combination Rule in Conflict[J].Journal of Projectiles,Rockets,Missiles and Guidance,2006,26(3):242-243.
[5] 屈志宏,杨传道,李方.基于D-S证据理论信息融合的故障诊断方法[J].火炮发射与控制学报,2008,29(4):105-109.QU Zhihong,YANG Chuandao,LI Fang.Fault Diagnosis Method Based on D-S Evidence Theory Informationg Fusion[J].Journal of Gun Launch & Control,2008,29(4):105-109.
[6] 雷萍.小波神经网络技术在齿轮箱轴承故障诊断中的应用[D].兰州:兰州理工大学,2009:12-15.
[7] 徐敏.设备故障诊断手册[M].西安:西安交通大学出版社,1998.
[8] 雷剑宇,廖明夫.基于旋转坐标系转轴振动信号的滚动轴承故障诊断方法[J].航空动力学报,2007,8(8):1340-1344.LEI Jianyu,LIAO Mingfu.A Diagnostic Approach to Rolling Element Bearing Based on Vibration Signal of Shaft in a Rotating Reference Frame[J].Journal of Aerospace Power,2007,8(8):1340-1344.
[9] JIN X H.Motor Bearing Fault Diagnosis Using Trace Ratio Linear Discriminant Analysis[J].IEEE Transactions on Industrial Electronics,2014,64(5):2441-2445.
[10] 刘建强,赵治博,章国平,等.地铁车辆转向架轴承故障诊断方法研究[J].铁道学报,2015,37(1):32-38.LIU Jianqiang,ZHAO Zhibo,ZHANG Guoping,et al.Research on the Fault Diagnosis Method for Bogie Bearings of the Metro Vehicle[J].Journal of the China Railway Society,2015,37(1):32-38.
[11] 安学利,蒋东翔,李少华.基于决策融合的直驱风力发电机组轴承故障诊断[J].电网技术,2011,35(7):36-39.AN Xueli,JIANG Dongxiang,LI Shaohua.Fault Diagnosis of Spherical Roller Bearing of Direct-drive Wind Turbine Based on Decision Fusion[J].Power System Technology,2011,35(7):36-39.
[12] 曾承志,姚兴佳,唐德尧,等.多物理量融合的轴承齿轮故障诊断方法研究[J].机车电传动,2015(2):86-90.ZENG Chengzhi,YAO Xingjia,TANG Deyao,et al.Research on Bearing-gear Fault Diagnosis Method Based on Multiple Physical Quantities Fusion[J].Electric Drive for Locomotives,2015(2):86-90.
[13] 刘娟,胡泽,葛亮.一种改进型冲突证据合成算法研究[J].制造业自动化,2013,35(4):89-91.LIU Juan,HU Ze,GE Liang.Research of a Modified Combination Rule of Conflict Evidence[J].Manufacturing Automation,2013,35(4):89-91.
[14] 王俊松,李建林.D-S证据理论改进方案综述[J].信息化研究,2011,37(6):4-6.WANG Junsong,LI Jianlin.Review on the Improved Scheme of D-S Evidence Theory[J].Informatization Rearch,2011,37(6):4-6.
[2] 邢宗义,俞秀莲,程晓卿,等.基于证据理论的城轨车辆走行系融合故障诊断[J].机电工程,2014,31(12):31-32.XING Zongyi,YU Xiulian,CHENG Xiaoqing,et al.Approach for Fault Diagnosis of the Metro Vehicle Running Gears Based on Evidential Theory[J].Journal of Mechanical & Electrical Engineering,2014,31(12):31-32.
[3] 文振华,陈果.基于D-S证据理论的航空发动机磨损故障智能融合诊断方法[J].机械科学与技术,2005,24(9):1018-1020.WEN Zhenhua,CHEN Guo.An Intelligent Fusion Technique for Diagnosis of Engine Wear Fault Based on D-S Evidence Theory[J].Mechanical Science and Technology,2005,24(9):1018-1020.
[4] 李伟彤.一种冲突情况下的自适应证据理论合成公式[J].弹箭与制导学报,2006,26(3):242-243.LI Weitong.A Kind of Adapting Evidence Theory Combination Rule in Conflict[J].Journal of Projectiles,Rockets,Missiles and Guidance,2006,26(3):242-243.
[5] 屈志宏,杨传道,李方.基于D-S证据理论信息融合的故障诊断方法[J].火炮发射与控制学报,2008,29(4):105-109.QU Zhihong,YANG Chuandao,LI Fang.Fault Diagnosis Method Based on D-S Evidence Theory Informationg Fusion[J].Journal of Gun Launch & Control,2008,29(4):105-109.
[6] 雷萍.小波神经网络技术在齿轮箱轴承故障诊断中的应用[D].兰州:兰州理工大学,2009:12-15.
[7] 徐敏.设备故障诊断手册[M].西安:西安交通大学出版社,1998.
[8] 雷剑宇,廖明夫.基于旋转坐标系转轴振动信号的滚动轴承故障诊断方法[J].航空动力学报,2007,8(8):1340-1344.LEI Jianyu,LIAO Mingfu.A Diagnostic Approach to Rolling Element Bearing Based on Vibration Signal of Shaft in a Rotating Reference Frame[J].Journal of Aerospace Power,2007,8(8):1340-1344.
[9] JIN X H.Motor Bearing Fault Diagnosis Using Trace Ratio Linear Discriminant Analysis[J].IEEE Transactions on Industrial Electronics,2014,64(5):2441-2445.
[10] 刘建强,赵治博,章国平,等.地铁车辆转向架轴承故障诊断方法研究[J].铁道学报,2015,37(1):32-38.LIU Jianqiang,ZHAO Zhibo,ZHANG Guoping,et al.Research on the Fault Diagnosis Method for Bogie Bearings of the Metro Vehicle[J].Journal of the China Railway Society,2015,37(1):32-38.
[11] 安学利,蒋东翔,李少华.基于决策融合的直驱风力发电机组轴承故障诊断[J].电网技术,2011,35(7):36-39.AN Xueli,JIANG Dongxiang,LI Shaohua.Fault Diagnosis of Spherical Roller Bearing of Direct-drive Wind Turbine Based on Decision Fusion[J].Power System Technology,2011,35(7):36-39.
[12] 曾承志,姚兴佳,唐德尧,等.多物理量融合的轴承齿轮故障诊断方法研究[J].机车电传动,2015(2):86-90.ZENG Chengzhi,YAO Xingjia,TANG Deyao,et al.Research on Bearing-gear Fault Diagnosis Method Based on Multiple Physical Quantities Fusion[J].Electric Drive for Locomotives,2015(2):86-90.
[13] 刘娟,胡泽,葛亮.一种改进型冲突证据合成算法研究[J].制造业自动化,2013,35(4):89-91.LIU Juan,HU Ze,GE Liang.Research of a Modified Combination Rule of Conflict Evidence[J].Manufacturing Automation,2013,35(4):89-91.
[14] 王俊松,李建林.D-S证据理论改进方案综述[J].信息化研究,2011,37(6):4-6.WANG Junsong,LI Jianlin.Review on the Improved Scheme of D-S Evidence Theory[J].Informatization Rearch,2011,37(6):4-6.