基于改进二叉树支持向量机的内燃机故障识别

详细信息    查看全文 | 推荐本文 |

英文篇名：Internal Combustion Engine Fault Identification Based on Improved Binary Tree Support Vector Machine 作者：蔡艳平 ; 张恒 ; 石林锁 ; 张世雄 英文作者：Cai Yanping;Zhang Heng;Shi Linsuo;Zhang Shixiong;School of mechanical Engineering,Xi'an Jiaotong University;Rocket Force University of Engineering,Combat Support College; 关键词：内燃机 ; 改进二叉树 ; 支持向量机 ; 分类识别 英文关键词：internal combustion engine;;improved binary tree;;support vector machine;;classification recognition 中文刊名：NRJX 英文刊名：Transactions of CSICE 机构：西安交通大学机械工程学院;火箭军工程大学作战保障学院; 出版日期：2019-07-25 出版单位：内燃机学报 年：2019 期：v.37;No.184 基金：国家自然科学基金资助项目(51405498);; 中国博士后基金资助项目(2015M582642) 语种：中文; 页：NRJX201904011 页数：7 CN：04 ISSN：12-1086/TK 分类号：83-89

摘要

为提高内燃机在强耦合、弱信号条件下的故障诊断精度,提出一种基于改进二叉树支持向量机(SVM)的内燃机故障诊断方法.首先对样本的可分性测度进行重新定义,以此训练出的支持向量机模型更大程度上减少了样本错分的可能性,通过对仿真数据的分类识别,验证了有效性.以BF4L1011F型内燃机为诊断对象,分别提取振动信号的数据域及图像域特征,对比不同多分类算法识别结果,所提方法表现出更高的识别准确率.
        In order to improve the fault diagnosis accuracy of the internal combustion(IC)engine under strong coupling and weak signal conditions,a method based on improved binary tree support vector machine(SVM) of IC engine fault diagnosis was presented. This method redefines the divisibility measure of the samples between different classes,and the support vector machine model reduces the possibility of sample misclassification to a greater extent.The validity of the method was verified by the classification and identification of the simulation data. The BF4 L1011 F IC engine was used to extract the vibration signal domain and image domain characteristics and to compare the results of different multi-classification algorithms. The results show that the method presented gives a higher recognition accuracy.

引文

[1]Vapnik V N.An overview of statistical learning theory[J].IEEE Transactions on Neural Networks,1999,10(5):988-999.
    [2]余辉,赵晖.支持向量机多分类分类算法新研究[J].计算机工程与应用,2008,44(7):185-189.
    [3]张金泽,单甘霖.改进的SVM算法及其在故障诊断中的应用研究[J].电光与控制,2006,13(6):97-100.
    [4]谢娟英,张兵权,汪万紫.基于双支持向量机的偏二叉树多分类算法[J].南京大学学报(自然科学版),2011,47(4):354-363.
    [5]Vapnik V N.The Nature of Statistical Learning Theory[M].New York:Springer,1998:244-255.
    [6]Kressel U.Pairwise classification and support vector machines[J].Advances in Kernel Methods:Supports Vector Learning,1998:255-268.
    [7]王宪亮,吴志刚,杨金超,等.基于SVM一对一分类的语种识别方法[J].清华大学学报(自然科学版),2013,53(6):808-812.
    [8]Xia S,Li J,Xia L,et al.Tree-structured support vector machines for multi-class classification[J].2007,4493:392-398.
    [9]袁胜发,褚幅磊.次序二叉树支持向量机多类故障诊断算法研究[J].振动与冲击,2009,28(3):51-54.
    [10]Yang X,Yu Q,He L,et al.The one-against-all partition based binary tree support vector machine algorithms for multi-class classification[J].Neurocomputing,2013,113:1-7.
    [11]蔡艳平,李艾华,何艳萍,等.基于振动谱时频图像特征及SVM参数同步优化识别的内燃机故障诊断[J].内燃机学报,2012,30(4):377-383.
    [12]唐发明,王仲东,陈锦云.一种新的二叉树多类支持向量机算法[J].计算机工程与应用,2005,41(7):24-26.
    [13]赵海洋,许敏强,王金东.改进二叉树支持向量机及其故障诊断方法研究[J].振动工程学报,2013,26(5):764-770.
    [14]Asuncion A,Newman D J.UCI machine learning repository[EB/OL].http://www.ics.uci.edu/~mlearn/-MLRepository.html,2007.
    [15]蔡艳平,李艾华,王涛,等.基于小波包和LMS自适应降噪的柴油机振动诊断[J].噪声与振动控制,2010,1:104-110.
    [16]朱明,李志农,何旭平.基于广义S变换的滚动轴承故障诊断方法研究[J].机床与液压,2015,43(1):181-184.
    [17]李巍华,林龙,单外平.基于广义S变换与双向2DPCA的轴承故障诊断[J].振动、测试与诊断,2015,35(3):499-505.
    [18]蔡艳平,李艾华,王涛,等.基于时频谱图与图像分割的柴油机故障诊断[J].内燃机学报,2011,29(2):181-186.