结合SVM和改进证据理论的多信息融合故障诊断

PDF(1384 KB)

振动与冲击 ›› 2015, Vol. 34 ›› Issue (13) : 71-77.

论文

向阳辉，张干清，庞佑霞

作者信息 +

Fault Diagnosis of Multi-information Fusion by SVM & Improved Evidence Theory

XIANG Yang-hui, ZHANG Gan-qing, PANG You-xia

Author information +

文章历史 +

摘要

为了综合合理利用设备多个方面特征信息来提高故障诊断的准确性，提出一种结合支持向量机(Support vector machine, SVM)和改进证据理论的多信息融合故障诊断方法。该方法通过混淆矩阵获取各SVM局部诊断证据对各故障模式的可靠度，赋予不同的权重系数，并对由各SVM局部诊断硬输出判决矩阵构造出的基本概率分配进行加权处理，从而实现SVM和改进证据理论在多信息融合故障诊断中的有效结合。实验结果表明，各SVM局部诊断证据的加权融合处理能够显著降低各局部诊断间的冲突，所提方法可以有效提高故障诊断的准确率。

Abstract

To comprehensively reasonably utilize much feature information of the equipment to improve the accuracy of the fault diagnosis, a method of multi-information fusion fault diagnosis is proposed which is based on support vector machine(SVM) and improved evidence theory. It gets the reliability of the local diagnosis evidence of each SVM to every failure mode by the confusion matrix, by which gives different weight coefficient. The basic probability assignment that is constructed by hard output decision matrix from the local diagnosis of each SVM, which effectively combines SVM and improved evidence theory in multi-information fusion fault diagnosis. The experimental results show that the weighted fusion treatment of the local diagnosis evidence from each SVM can significantly reduce the conflict between the local diagnoses, and that the proposed method can effectively improve the accuracy of fault diagnosis.

导出引用

向阳辉，张干清，庞佑霞. 结合SVM和改进证据理论的多信息融合故障诊断[J]. 振动与冲击, 2015, 34(13): 71-77

XIANG Yang-hui, ZHANG Gan-qing, PANG You-xia. Fault Diagnosis of Multi-information Fusion by SVM & Improved Evidence Theory[J]. Journal of Vibration and Shock, 2015, 34(13): 71-77

参考文献

[1] 胡金海，余治国，翟旭升，彭靖波，任立通. 基于改进D-S证据理论的航空发动机转子故障决策融合诊断研究[J]. 航空学报, 2014, 35(2): 436-441.
(HU Jinhai, YU Zhiguo, ZHAI Xusheng, PENG Jingbo, REN Litong. Research of Decision Fusion Diagnosis of Aero-engine Rotor Fault Based on Improved D-S Theory[J]. Acta Aeronautica Et Astronautica Sinica, 2014, 35(2): 436-441.)
[2] 张恒,赵荣珍.故障特征选择与特征信息融合的加权KPCA方法研究[J].振动与冲击,2014,33(9):89-93.
(ZHANG Heng, ZHAO Rong-zhen. Weighted KPCA based on fault feature selection and feature information fusion[J]. Journal of Vibration and Shock, 2014, 33(9): 89-93.)
[3] 蒋玲莉，刘义伦，李学军，陈安华. 基于SVM与多振动信息融合的齿轮故障诊断[J].中南大学学报(自然科学版), 2010,41(6): 2184-2188.
(JIANG Ling-li , LIU Yi-lun, LI Xue-jun, CHEN An-hua. Gear fault diagnosis based on SVM and multi-sensor information fusion[J]. Journal of Central South University (Science and Technology), 2010, 41(6): 2184-2188.)
[4] 王维刚，刘占生. 多目标粒子群优化的支持向量机及其在齿轮故障诊断中的应用[J]. 振动工程学报, 2013,26(5): 743-749.
(WANG Wei-gang, LIU Zhan-sheng. Support vector machine optimized by multi-objective particle searm and application in gear fault diagnosis[J]. Journal of Vibration Engineering, 2013, 26(5): 743-749.)
[5] 李巍华, 张盛刚. 基于改进证据理论及多神经网络融合的故障分类[J].机械工程学报, 2010, 46(9): 93−99.
(LI Weihua, ZHANG Shenggang. Fault Classification Based on Improved Evidence Theory and Multiple Neural Network Fusion[J]. Chinese Journal of Mechanical Engineering, 2010, 46(9): 93−99.)
[6] 韩德强, 杨艺, 韩崇昭. DS 证据理论研究进展及相关问题探讨[J]. 控制与决策, 2014, 29(1): 1-8.
(HAN De-qiang, YANG Yi, HAN Chong-zhao. Advances in DS evidence theory and related discussions [J]. Control and Decision, 2014, 29(1): 1-8.)
[7] 姜万录，吴胜强．基于SVM和证据理论的多数据融合故障诊断方法 [J]．仪器仪表学报，2010，31(8)：1738-1743．
(Jiag Wanlu, Wu Shengqiang. Multi-data fusion fault diagnosis method based on SVM and evidence theory[J]. Chinese Journal of Scientific Instrument, 2010, 31(8): 1738-1743.)
[8] 车红昆,吕福在,项占琴.多特征SVM-DS融合决策的缺陷识别[J].机械工程学报,2010,46(16): 101-105.
(CHE Hongkun , LÜ Fuzai , XIANG Zhanqin. Defects Identification by SVM-DS Fusion Decision-making with Multiple Features[J]. Journal of Mechanical Engineering, 2010, 46(16): 101-105.)
[9] Denoux T. Conjunctive and disjunctive combination of belief functions induced by non distinct bodies of evidence[J]. Artificial Intelligence, 2008, 172(2/3): 234-264.
[10] Chao F, Yang S L. The combination of dependencebased interval-valued evidential reasoning approach with balanced scorecard for performance assessment[J]. Expert Systems with Applications, 2012, 39(3): 3717-3730.
[11] 韩东，李洪儒，许葆华．基于广义证据一致量和信息量因子的证据组合规则[J]．仪器仪表学报，2010，31(12)：2724-2727．
(Han Dong, L iH ong ru, Xu Baohu. Combination rule of evidence based on general evidence consistency degree and information content factor[J]. Chinese Journal of Scientific Instrument, 2010, 31(12): 2724-2727.)
[12] Jousselme A L, Maupin P. Distances in evidence theory: Comprehensive survey and generalizations[J]. Int J of Approximate Reasoning, 2012, 53(2): 118-145.
[13] 张锟, 张昌芳, 李杰. 基于新冲突度量的属性信息相关算法 [J]. 控制与决策, 2011, 26(4): 601-605.
(ZHANG Kun, ZHANG Chang-fang, LI Jie. Attribute information correlation algorithm based on new conflict Measure[J]. Control and Decision, 2011,26(4): 601-605.)
[14] Platt J. Probabilistic outputs for Support Vector Machines and Comparison to Regularized Likelihood Method[C]. Advance in large margin classifier. Cambridge: MIT Press, 2000: 61-74.
[15] LIN H, LIN C, WENG R C. A note on platt’s probabilistic outputs for support vector machines[J]. Machine Learning , 2007, 68(3): 267-276.
[16] 雷蕾，王晓丹，邢雅琼，毕凯．结合SVM和DS证据理论的多极化HRRP分类研究[J]．控制与决策，2013，28(6)：861-866．
(LEI Lei, WANG Xiao-dan, XING Ya-qiong, BI Kai. Multi-polarized HRRP classification by SVM and DS evidence theory [J]. Control and Decision, 2013, 28(6): 861-866.)