基于证据理论的概率盒融合改进方法
|
摘要
针对证据理论对不同特征融合缺乏统一的融合框架及在矛盾数据处理上的不合理性,通过分析证据理论与概率盒核心DSS之间的关系,论证了利用证据理论融合概率盒的可行性.通过对概率盒融合前的预处理得到了统一的融合框架,在对比了证据理论和加权平均法融合概率盒方法的基础上,给出了利用加权平均法改进证据理论的概率盒融合算法.提出利用加权平均矛盾因子矩阵迫使矛盾焦元融合后的区间变窄,以此降低矛盾数据的不合理影响力.通过仿真实验证明了改进方法的有效性,并以滚动轴承故障信号为实例进行了图表对比与分析,结果表明改进方法能在不失证据合成规则推理性的前提下有效地处理矛盾焦元.
Using evidence theory to fuse different features lacks unified fusion framework,and the evidence theory is not reasonable in the process of contradictory data. The relationship between evidence theory and probability box's core DSS was analyzed. The feasibility of using evidence theory to fuse probability box was demonstrated.By pretreatments before probability box fusion,the unified fusion framework was obtained. The probability box fusion methods based on Evidence theory and weighted average method were compared. A probability box fusion algorithm using the weighted average method to improve evidence theory was proposed. Weighted average contradiction factor matrix was proposed to force fused contradiction focus' interval narrower. The measure was used to reduce the irrational influence of conflicting data. The effectiveness of improved method was proved by simulation experiments. Experimental results based on rolling bearing fault signals were compared and analyzed. The results show that the improved method can effectively deal with focal elements' contradiction without loss of evidence theory reasoning rule.
Using evidence theory to fuse different features lacks unified fusion framework,and the evidence theory is not reasonable in the process of contradictory data. The relationship between evidence theory and probability box's core DSS was analyzed. The feasibility of using evidence theory to fuse probability box was demonstrated.By pretreatments before probability box fusion,the unified fusion framework was obtained. The probability box fusion methods based on Evidence theory and weighted average method were compared. A probability box fusion algorithm using the weighted average method to improve evidence theory was proposed. Weighted average contradiction factor matrix was proposed to force fused contradiction focus' interval narrower. The measure was used to reduce the irrational influence of conflicting data. The effectiveness of improved method was proved by simulation experiments. Experimental results based on rolling bearing fault signals were compared and analyzed. The results show that the improved method can effectively deal with focal elements' contradiction without loss of evidence theory reasoning rule.
引文
[1]王健,何卫平,王伟,等.基于DS证据理论的产品多信息融合标识失效补救方法[J].计算机集成制造系统,2015,21(4):1143-1148.
[2]丛林虎,徐廷学,董琪,等.基于改进证据理论的导弹状态评估方法[J].系统工程与电子技术,2016,38(1):70-76.
[3]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):71-76.
[4]梅检民,赵慧敏,肖云魁,等.基于多分类支持向量机和DS证据理论的轴承故障诊断[J].汽车工程,2015,37(1):114-119.
[5]龚本刚,张孝琪,郭丹丹.基于证据理论-DEA交叉效率的混合型多属性决策方法[J].控制与决策,2016,31(5):943-948.
[6] Savic. Neural generation of uncertainty reliability functions bounded by belief and plausibility frontiers[C]∥European Conference on Safety and Reliability,2005,1757-1762.
[7] Nguyen,WMlker. A First Course in Fuzzy Logic[M]. Boca Raton,Florida:CRC Press,2006.
[8] Berleant,Bounding. The Times to Failure of 2-Components Systems[J]. IEEE Transaction on Reliability,2004,53(4):542-550.
[9] Limbourg,Savic. Fault tree analysis in an early design stage using the Dempster-Shafer theory of evidence[J]. Risk,Reliability and Societal Safety,2007(3):99-105.
[10] Michael Oberguggenberger,Julian King,Bernhard Schmelzer. Imprecise probability methods for sensitivity analysis in engineering[C]∥In 5th International Symposium on Imprecise Probability:Theories and Applications,Prague,Czech Republic,2007,6:1130-1138.
[11] Luis G Crespo,Sean P Kenny,Daniel P Giesy. Reliability analysis of polynomial systems subject to p-box uncertainties[J]. Mechanical Systems and Signal Processing,2012(9):111-124.
[12] Ljung L. System Identification-theory for the User[M]. NJ:Prentice Hall,Englewood Cliffs,1987.
[13] Sun Lei,Jin Xiaoming. Model-predictive-control Based on Subspace Identification and Its Application[J]. Control Theory and Application,2009,26(3):313-315.
[14]张永富.一种子空间辨识对权重矩阵加权平均新方法[J].北华大学学报(自然科学版),2014,15(5):588-592.
[15]陈亮,傅闯,李兴源.基于加权平均组合评价法的电网直流融冰计划决策[J].电网技术,2014,38(10):2913-2918.
[16]王斌,张洪波,辛琛,兰甜.基于贝叶斯模型加权平均法的径流序列高频分量预测研究[J].水力发电学报,2016,35(5):75-83.
[17]焦荣,许录平,张华,等. X射线脉冲星光子序列频域加权测相方法[J].西安交通大学学报,2016,50(6):152-158.
[18]顾雪平,王大江,梁海平,等.电力系统扩展黑启动方案动态综合评估[J].电工技术学报,2015,30(3):44-52.
[19]杜奕,丁家满,刘力强,等.基于概率盒理论的滚动轴承故障信号建模方法[J].振动与冲击,2016,35(19):31-37.
[2]丛林虎,徐廷学,董琪,等.基于改进证据理论的导弹状态评估方法[J].系统工程与电子技术,2016,38(1):70-76.
[3]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):71-76.
[4]梅检民,赵慧敏,肖云魁,等.基于多分类支持向量机和DS证据理论的轴承故障诊断[J].汽车工程,2015,37(1):114-119.
[5]龚本刚,张孝琪,郭丹丹.基于证据理论-DEA交叉效率的混合型多属性决策方法[J].控制与决策,2016,31(5):943-948.
[6] Savic. Neural generation of uncertainty reliability functions bounded by belief and plausibility frontiers[C]∥European Conference on Safety and Reliability,2005,1757-1762.
[7] Nguyen,WMlker. A First Course in Fuzzy Logic[M]. Boca Raton,Florida:CRC Press,2006.
[8] Berleant,Bounding. The Times to Failure of 2-Components Systems[J]. IEEE Transaction on Reliability,2004,53(4):542-550.
[9] Limbourg,Savic. Fault tree analysis in an early design stage using the Dempster-Shafer theory of evidence[J]. Risk,Reliability and Societal Safety,2007(3):99-105.
[10] Michael Oberguggenberger,Julian King,Bernhard Schmelzer. Imprecise probability methods for sensitivity analysis in engineering[C]∥In 5th International Symposium on Imprecise Probability:Theories and Applications,Prague,Czech Republic,2007,6:1130-1138.
[11] Luis G Crespo,Sean P Kenny,Daniel P Giesy. Reliability analysis of polynomial systems subject to p-box uncertainties[J]. Mechanical Systems and Signal Processing,2012(9):111-124.
[12] Ljung L. System Identification-theory for the User[M]. NJ:Prentice Hall,Englewood Cliffs,1987.
[13] Sun Lei,Jin Xiaoming. Model-predictive-control Based on Subspace Identification and Its Application[J]. Control Theory and Application,2009,26(3):313-315.
[14]张永富.一种子空间辨识对权重矩阵加权平均新方法[J].北华大学学报(自然科学版),2014,15(5):588-592.
[15]陈亮,傅闯,李兴源.基于加权平均组合评价法的电网直流融冰计划决策[J].电网技术,2014,38(10):2913-2918.
[16]王斌,张洪波,辛琛,兰甜.基于贝叶斯模型加权平均法的径流序列高频分量预测研究[J].水力发电学报,2016,35(5):75-83.
[17]焦荣,许录平,张华,等. X射线脉冲星光子序列频域加权测相方法[J].西安交通大学学报,2016,50(6):152-158.
[18]顾雪平,王大江,梁海平,等.电力系统扩展黑启动方案动态综合评估[J].电工技术学报,2015,30(3):44-52.
[19]杜奕,丁家满,刘力强,等.基于概率盒理论的滚动轴承故障信号建模方法[J].振动与冲击,2016,35(19):31-37.