Research on diesel engine fault diagnosis based on feature correlation analysis

doi:10.16731/j.cnki.1671-3133.2018.12.025

Abstract

Abstract: Aiming at the problem of failure diagnosis of diesel engine working status based on steady process and linear system statistical method,a fault diagnosis method of diesel engine based on feature extraction and correlation analysis is proposed.Firstly,the time domain,frequency domain and time-frequency characteristics of vibration signals of diesel engine cylinder head under different working conditions are extracted,and a standard characteristic matrix and an unknown characteristic matrix are established.Relevant analysis methods are used to analyze the correlation between different working condition signals.Finally,a model six diesel engine test bench was used to simulate the four conditions of normal,collision cylinder,misfire and small head tile wear respectively.The experimental results show that this method can accurately diagnose four operating conditions of normal,collision cylinder,misfire and small head tile wear,and prove the effectiveness of this method in diesel engine fault diagnosis.At the same time,sensitivity coefficients are used to evaluate and rank the fault features.The sensitive characteristics of the fault of the impact cylinder and the small head tile wear are mainly concentrated in the time domain,while the normal and misfire are also concentrated in the time and frequency domains.

Key words: diesel engine, feature extraction, correlation analysis, sensitivity coefficients, fault diagnosis

CLC Number:

TK421

Yang Dong, Song Hanjiang, Huo Baiqi, Zhao Haipeng, Ma Weihao. Research on diesel engine fault diagnosis based on feature correlation analysis[J]. Modern Manufacturing Engineering, 2018, 459(12): 147-152.

References

[1] 严莉.基于多源信息融合的发动机故障诊断研究[D].太原:中北大学,2010.
[2] 刘世元,杜润生,杨叔子.柴油机缸盖振动信号的小波包分解与诊断方法研究[J].振动工程学报,2000(4):85-92.
[3] SHAO Y,DING Y,MECHEFSKE C K.Engine fault detection using angle domain signal envelope algorithm[J].Proceedings of the Institution of Mechanical Engineers Part I Journal of Systems & Control Engineering,2013,227(6):541-551.
[4] 王奉涛.非平稳信号故障特征提取与智能诊断方法的研究及应用[D].大连:大连理工大学,2003.
[5] 张宇飞,么子云,唐松林,等.一种基于主成分分析和支持向量机的发动机故障诊断方法[J].中国机械工程,2016,27(24):3307-3311.
[6] 别锋锋,刘杨,裴峻峰,等.EEMD近似熵和SVM在柴油机传动系统中的故障诊断研究[J].机械设计与制造,2015(3):24-27,31.
[7] PORTEIRO J,COLLAZO J,PATINO D,et al.Diesel engine condition monitoring using a multi-net neural network system with nonintrusive sensors[J].Applied Thermal Engineering,2011,31(17):4097-4105.
[8] FU Y,YAN S,HUAMG T S.Correlation metric for generalized feature extraction[J].IEEE Transactions on Pattern Analysis & Machine Intelligence,2008,30(12):2229-2235.
[9] 熊武勇,丁光明.机械振动诊断中信号的相关分析与故障定位[J].江西电力职业技术学院学报,2003,16(2):21-22,28.
[10] 侯新国,吴正国,夏立,等.基于相关分析的感应电机定子故障诊断方法研究[J].中国电机工程学报,2005,25(4):83-86.
[11] 袁佳胜,冯志华.基于相关分析与小波变换的齿轮箱故障诊断[J].农业机械学报,2007,38(8):120-123.
[12] 林瑞霖,孙云岭.基于瞬时转速相关分析的柴油机工作均匀性检测[J].内燃机,2006(4):18-20.
[13] 耿志强,胡海霞,韩永明.基于余弦相似度的复杂网络故障检测方法及应用[J].北京化工大学学报(自然科学版),2017,44(2):81-86.
[14] LEE Y K,KIM W Y,CAI Y D,et al.CoMine:Efficient Mining of Correlated Patterns.IEEE International Conference on Data Mining[C].Melbourne:IEEE Computer Society,2003:581-584.
[15] XIONG H,TAN P N,KUMAR V.Hyperclique pattern discovery[J].Data Mining and Knowledge Discovery,2006,13(2):219-242.
[16] LIU H,YU L.Toward integrating feature selection algorithms for classification and clustering[J].IEEE Transactions on Knowledge & Data Engineering,2005,17(4):491-502.
[17] SUN Y.Iterative relief for feature weighting:algorithms,theories,and applications[J].IEEE Transactions on Pattern Analysis & Machine Intelligence,2007,29(6):1035-51.