基于概率盒理论的滚动轴承故障信号建模方法

PDF(1748 KB)

振动与冲击 ›› 2016, Vol. 35 ›› Issue (19) : 31-37.

论文

杜奕 1 ，丁家满 2，刘力强 1

作者信息 +

A Rolling Bearing Fault Signal Modeling Method Based on Probability Box Theory

DU Yi 1 DING Jiaman 2 LIU liqiang 1

Author information +

文章历史 +

摘要

为了解决机械故障诊断存在特征提取带来的信息丢失问题和多段平均丢弃数据不确定性的问题，提出了一种基于概率盒理论的机械故障信号建模方法。以滚动轴承故障信号为研究对象，分析原始信号的概率分布类型，获得概率分布类型参数的不确定性区间，提出基于确定概率分布类型的概率盒建模方法。针对故障信号概率分布类型难确定问题，提取原始信号的特征，利用特征信号的有序性，提出基于特征提取的概率盒建模方法，以歪度和峭度特征为例，对比两种特征概率盒的异同点。基于概率盒定义，将原始数据的不确定性直接映射到概率盒的上下界，提出无需验证数据概率分布类型的原始数据概率盒直接建模方法。通过滚动轴承实测数据，对比三种方法的有效性及适用性，与传统特征提取方法对比，证明了方法的有效性。

Abstract

Feature extractions lead to information loss，and multi-segments-averages lead to data uncertainties discarding. In order to solve the above problems in mechanical fault diagnosis, a new modeling method for mechanical fault signals based on probability box (p-box) theory was proposed. The fault signals of the rolling bearing were the research objects. The raw data’s probability types were analyzed. The uncertainty interval of probability distribution’s parameter was gotten. The p-box modeling method based on normal distribution was proposed. In order to solve the identification difficulty of fault signal data’s probability distribution, the raw data’s characteristics were extracted, and the orderliness of characteristics was used, and a p-box modeling method based on feature extraction was proposed. The similarities and differences of skewness p-box and kurtosis p-box were contrasted. Based on the p-box’s definition, the raw data’s uncertainties were projected into the p-box’s bounds, and another more effective p-box modeling method which was directly based on raw data was proposed, which does not need data’s probability distribution identification. The effectiveness and applicability of the three methods were compared by the rolling bearing’s measure data. The method’s validity is verified by compared with conventional feature extraction method.

导出引用

杜奕 1,丁家满 2，刘力强 1. 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

DU Yi 1 DING Jiaman 2 LIU liqiang 1 . A Rolling Bearing Fault Signal Modeling Method Based on Probability Box Theory[J]. Journal of Vibration and Shock, 2016, 35(19): 31-37

参考文献

[1] 赵志宏,杨绍普,刘永强. 多小波系数特征提取方法在故障诊断中的应用[J]. 振动、测试与诊断，2015,35(2):276-280.
Zhao Zhihong, Yang Shaopu, Liu Yongqiang．Multi-wavelet coefficient feature extraction method in fault diagnosis．Journal of Vibration, Measurement & Diagnosis, 2015, 35(2):276-280.
[2] 向丹, 葛爽. 基于样本熵和流形学习的故障特征提取方法[J]. 航空动力学报，2014, 29(7):1535-1542
Xiang Dan, Ge Shuang. Method of fault feature extraction based on EMD sample entropy and manifold learning [J]. Journal of Aerospace Power, 2014, 29(7): 1535-1542.
[3] 任立通，胡金海，谢寿生等. 基于随机共振预处理的振动故障特征提取研究[J]. 振动与冲击，2014，33(2):141-146
Ren Litong, Hu Jinhai, Xie ShouSheng, etc. Vibration fault feature extraction based on stochastic resonance pretreatment. Journal of Vibration and Shock, 2014，33(2):141-146.
[4] 唐贵基，王晓龙．基于局部均值分解和切片双谱的滚动轴承故障诊断研究[J]. 振动与冲击, 2013, 32(24): 83-88.
TANG GuiJi, WANG XiaoLong. Fault diagnosis of roller bearings based on local mean decomposition and slice bispectrun[J]. Journal of Vibration and Shock, 2013, 32(24): 83-88.
[5] 隋文涛, 张丹. 平稳小波变换在轴承振动信号去噪中的应用[J]. 轴承, 2012, 1:38-40
SUI Wen tao, ZHANG Dan. Application of stationary wavelet transform in de-nosing of bearing vibration signal[J]. Bearing, 2012, 1:38-40.
[6] Kaufmann. Introduction to Fuzzy Arithmetic: Theory and Applications[M]. New York : Van Nostrand Reinhold, 1985.
[7] Shafer. The Combination of Evidence[J]. International Journal of Intelligent Systems, 1986, 1: 155–179.
[8] Hailperin. Boole’s Logic and Probability[M]. North-Holland, Amsterdam, 1986.
[9] Frank, Nelsen and Schweizer. Best-possible bounds for the distribution of a sum - a problem of Kolmogorov[J]. Probability Theory and Related Fields, 1987, 74: 199-211.
[10] Savic. Neural generation of uncertainty reliability functions bounded by belief and plausibility frontiers[C]. European Conference on Safety and Reliability, 2005, 1757-1762.
[11] Nguyen, Walker. A First Course in Fuzzy Logic[M]. Boca Raton, Florida: CRC Press, 2006.
[12] Scott Ferson, Vladik Kreinovich, Lev Ginzburg, etal. Constructing Probability Boxes and Dempster-Shafer Structures[R]. California：Sandia National Laboratories, 2003.
[13] Berleant, Bounding. The Times to Failure of 2-Components Systems [J]. IEEE Transaction on Reliability, 2004, 53(4):542-550.
[14] Fulvio Tonon. Using random set theory to propagate epistemic uncertainty through a mechanical system [J]. Reliability Engineering and System Safety, 2004, 85:169-181.
[15] Limbourg, Savic. Fault tree analysis in an early design stage using the Dempster-Shafer theory of evidence [J]. Risk, Reliability and Societal Safety. 2007, 1(3): 99-105.
[16] 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.
[17] Destercke, Dubois, Chojnacki. Unifying practical uncertainty representations – I: Generalized p-boxes [J]. International Journal of Approximate Reasoning, 2008, 49(3): 649-663.
[18] 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.
[19] Matthias C.M. Troffaes, Enrique Miranda, Sebastien Destercke. On the connection between probability boxes and possibility measures Original Research Article [J]. Information Sciences, 2013, 224 (3): 88-108.
[20] N. Ben Abdallah, N. Mouhous-Voyneau, T. Denoeux. Combining statistical and expert evidence using belief functions: Application to centennial sea level estimation taking into account climate change [J]. International Journal of Approximate Reasoning, 2014, 55(1): 341-354.