Aiming at the difficulty of extracting the composite fault features of the gearbox, the lack of automatic identification in the diagnosis, and the failure information is often not fully represented by a single channel, a multi-channel gearbox composite fault diagnosis method based on the fusion of random forest and evidence theory is proposed. Firstly, the composite faults signal of each channel is decomposed by wavelet packet transform (WPT), and the eigenvectors of the fault signal are obtained; Next, a new feature set combination framework is introduced to construct feature datasets for different faults, and a single classification model is divided by random forest algorithm; Then comprehensively consider each classification model, synthesize the ensemble classifier under each channel, and propose a new iterative self-update strategy to continuously improve the performance of the classifier; Finally, an improved D-S evidence theory algorithm based on Lance distance is designed, the algorithm uses the Lance distance to measure the evidence distance between each spatial evidence, and constructs the Lance matrix, from this, a similarity matrix is obtained to measure the similarity and support between the various evidence bodies, the final diagnostic fusion result is obtained by calculating the sensitivity weight coefficients of each channel for BPA correction. The algorithm is verified by the gearbox experimental platform, and the results show that the method can effectively identify each type of fault contained in the composite faults, and can fully integrate the fault redundancy information of different channels to realize the accurate diagnosis of the gearbox composite faults.
Key words
gearbox /
composite faults /
random forest algorithm /
multi-channel /
evidence theory
{{custom_keyword}} /
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
References
[1] 林琳. 旋转机械运行趋势预测与故障诊断通用方法研究[D].大连理工大学,2021.
LIN Lin. Research on general method of Running Trend prediction and Fault Diagnosis for Rotating Machinery[D]. Dalian University of Technology,2021.
[2] 周东华,纪洪泉,何潇.高速列车信息控制系统的故障诊断技术[J].自动化学报,2018,44(07):1153-1164.
ZHOU Donghua, JI Hongquan, HE Xiao. Fault diagnosis technology of high-speed train information control system[J]. Acta Automatica Sinica,2018,44(07):1153-1164.
[3] He S L, Chen J L, Zhou Z T, et al. Multifractal entropy based adaptive multiwavelet construction and its application for mechanical compound-fault diagnosis[J].Mechanical Systems and Signal Processing,2016,76(08):742-758.
[4] Teng W, Ding X, Cheng H, et al. Compound faults diagnosis and analysis for a wind turbine gearbox via a novel vibration model and empirical wavelet transform[J].Renewable Energy, 2019, 136(06): 393-402.
[5] 付强. 风电齿轮箱早期复合故障信息提取技术研究[D].哈尔滨工业大学,2015.
FU Qiang. Research on early complex fault information extraction technology of Wind Turbine Gearbox[D]. Harbin Institute of Technology,2015.
[6] LIN H, WU F, HE G. Rolling bearing fault diagnosis using impulse feature enhancement and nonconvex regularization[J]. Mechanical Systems and Signal Processing, 2020, 142:106790.
[7] 赵晓平, 王逸飞, 张永宏, 吴家新, 王丽华. 一种基于Triplet loss的齿轮箱复合故障识别方法[J]. 振动与冲击, 2021, 40(05):46-54.
ZHAO Xiaoping, WANG Yifei, ZHANG Yonghong, WU Jiaxin, WANG Lihua. A compound fault identification method for gearbox based on Triplet loss[J]. Journal of Vibration and Shock，2019，33( 6) : 2573-2586．
[8] XU Y B，CAI Z Y，CAI X Y，et al．An enhanced multipion optimal minimum entropy deconvolution approach for bearing fault detection of spur gearbox[J]. Journal of Mechanical Science and Technology，2019，33( 6) : 2573-2586．
[9] HONG L H，LIU X B，ZUO H Y．Compound fault diagnosis of rotating machinery based on adaptive maximum correlated kurtosis deconvolution and customized multiwavelet transform[J]. Measurement Science and Technology, 2018, 29( 11) :115007．
[10] GEOFF L, MCDONALD, QING ZHAO, MING J. ZUO. Maximum correlated Kurtosis deconvolution and application on gear tooth chip fault detection[J]. Mechanical Systems and Signal Processing, 2012, 33:237-255.
[11] MCDONALD G L, ZHAO Q. Multipoint Optimal Minimum Entropy Deconvolution and Convolution Fix: Application to vibration fault detection[J]. Mechanical Systems & Signal Processing, 2017, 82:461-477.
[12] 李兵,韩睿,何怡刚,等.改进随机森林算法在电机轴承故障诊断中的应用[J].中国电机工程学报,2020,40(04):1310-1319+1422.
LI Bing, HAN Rui, HE Yigang, et al. Application of improved random forest algorithm in fault diagnosis of motor bearing[J]. Proceedings of the CSEE, 2020,40(04):1310-1319+1422.
[13] Samanta B, Nataraj C. Use of particle swarm optimization for machinery fault detection[J]. Engineering Applications of Artificial Intelligence, 2009, 22(2):308-316.
[14] 周兴康,余建波.基于深度一维残差卷积自编码网络的齿轮箱故障诊断[J].机械工程学报,2020,56(07):96-108.
ZHOU Xingkang, YU Jianbo. Gearbox fault diagnosis based on deep one-dimensional Residual convolutional autocoding network[J]. Journal of Mechanical Engineering, 2020,56(07):96-108.
[15] 李凤英. 基于多标签学习和极限学习机的复合故障诊断方法研究[D].北京化工大学,2021.
LI Fengying. Research on Compound fault diagnosis method based on multi-label learning and extreme learning Machine[D]. Beijing University of Chemical Technology,2021
{{custom_fnGroup.title_en}}
Footnotes
{{custom_fn.content}}
PDF(3257 KB)
