基于公共变量的双变量逆高斯过程的湿式离合器RUL预测

doi:10.12382/bgxb.2024.1043

摘要/Abstract

摘要： 湿式离合器是车辆综合传动装置中重要的组成部分，其剩余使用寿命(Remaining Useful Life，RUL)的准确在线预测对设备的运行安全和视情维护策略的制定起着关键作用。基于逆高斯过程构建具有公共变量的双变量退化模型，通过公共变量表征变量间的相关性，利用贝叶斯方法进行模型参数的后验估计，并提出拟合优度检验指标。考虑离合器设备的运行维护过程，推导其动态失效率和可用度的求解方法，提出基于蒙特卡洛模拟的离合器RUL在线预测方法，通过仿真研究和湿式离合器试验退化数据分析，验证了模型在不同样本量数据下的有效性。结果表明：相较于传统伽马模型，所建立的双变量逆高斯过程模型能够准确估计湿式离合器设备的失效率和可用度，以及实现更准确的RUL在线预测。

关键词: 湿式离合器, 退化分析, 逆高斯过程, 剩余使用寿命预测

Abstract:

Wet multi-disc clutch is a critical component within the vehicle’s integrated transmission system，where the accurate online prediction of remaining useful life (RUL) plays a key role in ensuring the operational safety and formulating the condition-based maintenance strategies.A bivariate degradation model with common random variable based on inverse Gaussian processes is developed.The correlation between variables is characterized by a common random variable.The Bayesian method is used for the posterior estimation of model parameters，and the index for goodness-of-fit test is proposed.Considering the operation and maintenance process of wet clutch equipment，the solving method for its dynamic failure rate and availability is derived，and an online RUL prediction method based on Monte Carlo simulation is proposed.The effectiveness of the proposed model is verified with different sample sizes through simulation study and degradation analysis of wet clutch tests.Compared to traditional Gamma models，the proposed bivariate inverse Gaussian process model is enable to accurately estimate the failure rates and availability of wet clutch and complete more precise online RUL prediction.

Key words: wet clutch, degradation analysis, inverse Gaussian process, remaining useful life prediction

冯毓庆, 郑长松, 于亮, 张丁戈, 张金乐, 张玉东. 基于公共变量的双变量逆高斯过程的湿式离合器RUL预测[J]. 兵工学报, 2025, 46(S1): 241043-.

FENG Yuqing, ZHENG Changsong, YU Liang, ZHANG Dingge, ZHANG Jinle, ZHANG Yudong. RUL Prediction of Wet Clutches based on Bivariate Inverse Gaussian Degradation Process with Common Random Variable[J]. Acta Armamentarii, 2025, 46(S1): 241043-.

图/表 13

图1 湿式离合器结构和摩擦元件

Fig.1 Schematic diagram of wet clutch structure and friction components

图2 钢片使用前后表面形貌对比

Fig.2 Surface morphologies of steel sheet before and after use

表1 模型参数估计结果

Table 1 Estimation results of model parameters

(n，m)	指标	参数
(n，m)	指标	ω	κ	λ₁	λ₂	α₁	α₂
(1，30)	Mean	14.761	10.006	0.309	0.363	1.264	0.827
(1，30)	MSE	3.392	5.977	0.0069	0.0405	0.00275	0.00691
(5，50)	Mean	16.461	10.082	0.280	0.185	1.313	0.927
(5，50)	MSE	0.840	5.962	0.0016	0.0005	0.00028	0.00086
(15，30)	Mean	16.188	10.049	0.273	0.210	1.307	0.908
(15，30)	MSE	0.192	6.223	0.0011	0.0003	0.00015	0.00018
(25，50)	Mean	16.225	9.928	0.285	0.205	1.305	0.907
(25，50)	MSE	0.106	6.265	0.0003	0.0001	0.00004	0.00007

图3 模型参数后验样本的遍历平均图，(n，m)=(5，50)

Fig.3 Ergodic mean plots of posterior samples of the proposed model parameters，(n，m)=(5，50)

图4 2个离合器设备的性能参数的累计退化路径

Fig.4 Cumulative degradation paths of PCs of two wet clutches

图5 所提方法实施流程

Fig.5 Implementation flow chart of the proposed method

表2 离合器1退化数据的参数估计结果

Table 2 Parameter estimation results with clutch-1 data

参数	均值	95%置信区间
ω	6.080	(2.775，9.998)
κ	0.091	(0.008，0.185)
λ₁	9.571	(5.707，13.266)
λ₂	1.789	(1.073，2.568)
α₁	0.853	(0.819，0.888)
α₂	0.524	(0.483，0.564)

图6 离合器1退化路径的拟合结果

Fig.6 Fitting results of degradation paths of clutch-1

表3 不同模型的拟合性能对比

Table 3 Fitting performances of different models

设备	退化模型	WAIC	MAE
离合器1	文献[20]方法	455.77	0.02
离合器1	本文方法	444.23	0.0098
离合器2	文献[20]方法	472.74	0.0206
离合器2	本文方法	468.76	0.0049

图7 离合器1的失效率与可用度估计

Fig.7 Failure rate and availability estimation of clutch-1

图8 离合器设备的RUL在线预测

Fig.8 RUL prediction results for two clutch devices

图9 不同模型下的MAEs

Fig.9 MAEs of different models

图10 离合器设备RUL预测的PDF

Fig.10 Clutch RUL’s PDFs under two models

参考文献 26

[1]	YU L, MA B, CHEN M, et al. Investigation on the failure mechanism and safety mechanical-thermal boundary of a multi-disc clutch[J]. Engineering Failure Analysis, 2019, 103:319-334.
[2]	YU L, MA B, CHEN M, et al. Experimental study on the friction stability of paper-based clutches concerning groove patterns[J]. Industrial Lubrication and Tribology, 2020, 72(4):541-548.
[3]	OST W, DE BAETS P, DEGRIECK J. The tribological behaviour of paper friction plates for wet clutch application investigated on SAE#II and pin-on-disk test rigs[J]. Wear, 2001, 249(5-6):361-371.
[4]	ZHENG H L, YANG J, KANG W D, et al. Accelerated degradation data analysis based on inverse Gaussian process with unit heterogeneity[J]. Applied Mathematical Modelling, 2024, 126:420-438.
[5]	WU X, HUANG T T, LIU J. Common stochastic effects induced multivariate degradation process with temporal dependency in degradation characteristic and unit dimensions[J]. Reliability Engineering & System Safety, 2023, 239:109505.
[6]	MA Z H, NIE S L, YIN F L, et al. Lifetime prediction of WC-6Ni/SiC friction pair under seawater lubrication using an Inverse Gaussian model[J]. Ceramics International, 2022, 48(1):463-471.
[7]	LI H, ZHANG Z X, LI T M, et al. A review on physics-informed data-driven remaining useful life prediction:challenges and opportunities[J]. Mechanical Systems and Signal Processing, 2024, 209:111120.
[8]	ZHANG Z X, SI X S, HU C H, et al. Degradation data analysis and remaining useful life estimation:a review on Wiener-process-based methods[J]. European Journal of Operational Research, 2018, 271(3):775-796.
[9]	DAI X L, QU S, SUI H, et al. Reliability modelling of wheel wear deterioration using conditional bivariate gamma processes and Bayesian hierarchical models[J]. Reliability Engineering & System Safety, 2022, 226:108710.
[10]	POONIA N, PRAJAPATI D, AZAD S. The bivariate exponentiated additive Weibull distribution and its multivariate extension with applications[J]. Computers & Industrial Engineering, 2024, 188:109886.
[11]	BARUI S, MITRA D, BALAKRISHNAN N. Flexible modelling of a bivariate degradation process with a shared frailty and an application to fatigue crack data[J]. Reliability Engineering & System Safety, 2024, 242:109722.
[12]	SARI J K, NEWBY M J, BROMBACHER A C, et al. Bivariate constant stress degradation model:LED lighting system reliability estimation with two-stage modelling[J]. Quality and Reliability Engineering International, 2009, 25(8):1067-1084.
[13]	PAN Z Q, BALAKRISHNAN N, SUN Q, et al. Bivariate degradation analysis of products based on Wiener processes and copulas[J]. Journal of Statistical Computation and Simulation, 2013, 83(7):1316-1329.
[14]	WANG X, BALAKRISHNAN N, GUO B, et al. Residual life estimation based on bivariate non-stationary gamma degradation process[J]. Journal of Statistical Computation and Simulation, 2015, 85(2):405-421.
[15]	WANG X, XU D H. An inverse Gaussian process model for degradation data[J]. Technometrics, 2010, 52(2):188-197.
[16]	PENG W W, LI Y F, YANG Y J, et al. Bivariate analysis of incomplete degradation observations based on inverse Gaussian processes and copulas[J]. IEEE Transactions on Reliability, 2016, 65(2):624-639.
[17]	DUAN F J, WANG G J, WANG H. Inverse Gaussian process models for bivariate degradation analysis:a Bayesian perspective[J]. Communications in Statistics-Simulation and Computation, 2018, 47(1):166-186.
[18]	YE Z S, CHEN N. The inverse Gaussian process as a degradation model[J]. Technometrics, 2014, 56(3):302-311.
[19]	YANG Z H, HUANG W N, DONG S, et al. Mixture bivariate distribution of wind speed and air density for wind energy assessment[J]. Energy Conversion and Management, 2023, 276:116540.
[20]	SONG K, CUI L R. A common random effect induced bivariate gamma degradation process with application to remaining useful life prediction[J]. Reliability Engineering & System Safety, 2022, 219:108200.
[21]	XU A C, SHEN L J, WANG B X, et al. On modeling bivariate Wiener degradation process[J]. IEEE Transactions on Reliability, 2018, 67(3):897-906.
[22]	SONG K, SHI J. A gamma process based in-play prediction model for national basketball association games[J]. European Journal of Operational Research, 2020, 283(2):706-713.
[23]	CUI J Z, XIE F W, WANG C T. Numerical investigation on thermal deformation of friction pair in hydro-viscous drive[J]. Applied Thermal Engineering, 2015, 90:460-470.
[24]	GONG T M, YAO P P, XIAO Y L, et al. Wear map for a copper-based friction clutch material under oil lubrication[J]. Wear, 2015, 328:270-276.
[25]	LIU X. Planning of accelerated life tests with dependent failure modes based on a Gamma frailty model[J]. Technometrics, 2012, 54(4):398-409.
[26]	陈旭, 张书锋, 李岳, 等. 正态型寿命分布设备的可用性分析[J]. 兵工学报, 2024, 45(3):986-996. doi: 10.12382/bgxb.2022.0606
	CHEN X, ZHANG S F, LI Y, et al. Availability analysis for equipment with normal life distribution[J]. Acta Armamentarii, 2024, 45(3):986-996 doi: 10.12382/bgxb.2022.0606