A Multiscale Similarity Ensemble Methodology for Remaining Useful Life Prediction in Multiple Fault Modes

doi:10.16183/j.cnki.jsjtu.2021.024

Abstract

Abstract:

Traditional similarity-based methods generally ignore the diversity of equipment fault modes, the difference in degradation rates, and the inconsistency among monitoring data lengths. Thus, a similarity-based multi-scale ensemble method in multiple fault modes (MFM-MSEN) is proposed to improve remaining useful life (RUL) prediction accuracy and characterize prediction uncertainty. By training the fault mode classification model, designing the time-series weighted prediction strategy, and recognizing the fault mode of equipment, the test equipment is matched with the training equipment with the same fault mode to reduce matching complexity, based on which, a multi-scale ensemble strategy is proposed to overcome the data utilization limitation caused by single-scale matching methods and enhance the generalization ability of the proposed MFM-MSEN method. This strategy matches the similarities between test equipment and training equipment at multiple time scales, and then multiscale prediction results are integrated to fit accurate RUL probability distribution by employing kernel density estimation. Experimental results demonstrate the superiority of the proposed MFM-MSEN method in dealing with the differences in equipment degradation.

Key words: remaining useful life (RUL), fault mode identification, multiscale ensemble strategy, kernel density estimation, similarity-based method

CLC Number:

TH17

SHU Junqing, XU Yuhui, XIA Tangbin, PAN Ershun, XI Lifeng. A Multiscale Similarity Ensemble Methodology for Remaining Useful Life Prediction in Multiple Fault Modes[J]. Journal of Shanghai Jiao Tong University, 2022, 56(5): 564-575.

Figures/Tables 16

Tab.1

Fig.1

Fig.2

Fig.3

Fig.4

Tab.2

Fig.5

Fig.6

Fig.7

Tab.3

Fig.8

Fig.9

Tab.4

Tab.5

Fig.10

Fig.11

References 20

[1]	XIA T B, DONG Y F, XIAO L, et al. Recent advances in prognostics and health management for advanced manufacturing paradigms[J]. Reliability Engineering & System Safety, 2018, 178: 255-268. doi: 10.1016/j.ress.2018.06.021 URL
[2]	ZHUO P C, ZHU Y, WU W X, et al. Real-time fault diagnosis for gas turbine blade based on output-hidden feedback Elman neural network[J]. Journal of Shanghai Jiao Tong University (Science), 2018, 23(Sup. 1): 95-102.
[3]	YU W N, KIM I Y, MECHEFSKE C. Remaining useful life estimation using a bidirectional recurrent neural network based autoencoder scheme[J]. Mechanical Systems and Signal Processing, 2019, 129: 764-780. doi: 10.1016/j.ymssp.2019.05.005 URL
[4]	KHELIF R, CHEBEL-MORELLO B, MALINOWSKI S, et al. Direct remaining useful life estimation based on support vector regression[J]. IEEE Transactions on Industrial Electronics, 2017, 64(3): 2276-2285. doi: 10.1109/TIE.2016.2623260 URL
[5]	LIU Q M, DONG M, LV W, et al. A novel method using adaptive hidden semi-Markov model for multi-sensor monitoring equipment health prognosis[J]. Mechanical Systems and Signal Processing, 2015, 64/65: 217-232. doi: 10.1016/j.ymssp.2015.03.029 URL
[6]	LI X, DING Q, SUN J Q. Remaining useful life estimation in prognostics using deep convolution neural networks[J]. Reliability Engineering & System Safety, 2018, 172: 1-11. doi: 10.1016/j.ress.2017.11.021 URL
[7]	ZHENG S, RISTOVSKI K, FARAHAT A, et al. Long short-term memory network for remaining useful life estimation[C]// 2017 IEEE International Conference on Prognostics and Health Management. Dallas, TX, USA: IEEE, 2017: 88-95.
[8]	ZHANG C, LIM P, QIN A K, et al. Multiobjective deep belief networks ensemble for remaining useful life estimation in prognostics[J]. IEEE Transactions on Neural Networks and Learning Systems, 2017, 28(10): 2306-2318. doi: 10.1109/TNNLS.2016.2582798 URL
[9]	SONG Y, SHI G, CHEN L Y, et al. Remaining useful life prediction of turbofan engine using hybrid model based on autoencoder and bidirectional long short-term memory[J]. Journal of Shanghai Jiao Tong University (Science), 2018, 23(Sup.1): 85-94.
[10]	宋亚, 夏唐斌, 郑宇, 等. 基于Autoencoder-BLSTM的涡扇发动机剩余寿命预测[J]. 计算机集成制造系统, 2019, 25(7): 1611-1619.
	SONG Y, XIA T B, ZHENG Y, et al. Remaining useful life prediction of turbofan engine based on Autoencoder-BLSTM[J]. Computer Integrated Manufacturing Systems, 2019, 25(7): 1611-1619.
[11]	WANG T Y, YU J B, SIEGEL D, et al. A similarity-based prognostics approach for remaining useful life estimation of engineered systems[C]// 2008 International Conference on Prognostics and Health Management. Denver, CO, USA: IEEE, 2008: 1-6.
[12]	CANNARILE F, BARALDI P, ZIO E. An evidential similarity-based regression method for the prediction of equipment remaining useful life in presence of incomplete degradation trajectories[J]. Fuzzy Sets and Systems, 2019, 367: 36-50. doi: 10.1016/j.fss.2018.10.008 URL
[13]	CAI H S, JIA X D, FENG J S, et al. A similarity based methodology for machine prognostics by using kernel two sample test[J]. ISA Transactions, 2020, 103: 112-121. doi: 10.1016/j.isatra.2020.03.007 URL
[14]	LIU Y C, HU X F, ZHANG W J. Remaining useful life prediction based on health index similarity[J]. Reliability Engineering & System Safety, 2019, 185: 502-510. doi: 10.1016/j.ress.2019.02.002 URL
[15]	YU W N, KIM I Y, MECHEFSKE C. An improved similarity-based prognostic algorithm for RUL estimation using an RNN autoencoder scheme[J]. Reliability Engineering & System Safety, 2020, 199: 106926. doi: 10.1016/j.ress.2020.106926 URL
[16]	HOU M R, PI D C, LI B R. Similarity-based deep learning approach for remaining useful life prediction[J]. Measurement, 2020, 159: 107788. doi: 10.1016/j.measurement.2020.107788 URL
[17]	XIA T B, SONG Y, ZHENG Y, et al. An ensemble framework based on convolutional bi-directional LSTM with multiple time windows for remaining useful life estimation[J]. Computers in Industry, 2020, 115: 103182. doi: 10.1016/j.compind.2019.103182 URL
[18]	PENG W W, YE Z S, CHEN N. Bayesian deep-learning-based health prognostics toward prognostics uncertainty[J]. IEEE Transactions on Industrial Electronics, 2020, 67(3): 2283-2293. doi: 10.1109/TIE.2019.2907440 URL
[19]	SAXENA A, GOEBEL K, SIMON D, et al. Damage propagation modeling for aircraft engine Run-to-failure simulation[C]// 2008 International Conference on Prognostics and Health Management. Denver, CO, USA: IEEE, 2008: 1-9.
[20]	RAMASSO E. Investigating computational geometry for failure prognostics in presence of imprecise health indicator: Results and comparisons on C-MAPSS datasets[C]// Second European Conference of the Prognostics and Health Management Society 2014. Nantes, France: CCSD, 2014: 1-13.

数据集	子集	发动机数量	发动机最大运行周期	发动机最小运行周期	运行工况种类	故障模式
FD003	训练集	100	525	145	1	2
	测试集	100	475	38
FD004	训练集	249	543	128	6	2
	测试集	248	486	19

编号	结构名称	网络层	神经元数量	激励函数
1	编码器	输入层	14	ReLU
2		隐含层 1	7	ReLU
3		隐含层 2	1	Linear
4	解码器	隐含层 3	7	ReLU
5		输出层	14	tanh

数据集	时间尺度集合 {τ₁,τ₂,τ₃,τ₄}	组别 1({τ₁}) 发动机数量	组别 2({τ₁,τ₂}) 发动机数量	组别 3({τ₁,τ₂,τ₃}) 发动机数量	组别 4({τ₁,τ₂,τ₃,τ₄}) 发动机数量
FD003	{38, 58, 78, 98}	6	6	9	79
FD004	{19, 39, 59, 79}	15	10	17	206

方法种类	方法名称	RMSE			Score
方法种类	方法名称	FD003	FD004	总计	FD003	FD004	总计
基于相似性的方法	MFM-MSEN 方法	15.07	22.22	20.42	609.97	2641.05	3251.02
	传统单尺度方法	16.89	24.62	22.67	1280.43	5009.31	6289.74
	EN^[3]	19.16	22.15	21.33	1727	2901	4628
	BiLSTM-ED^[15]	17.48	23.49	21.93	574	3202	3776
	RULCLIPPER^[20]	16.00	24.33	22.26	317	3132	3449
RUL直接映射法	DCNN^[6]	12.64	23.31	20.81	284.1	12466	12750.1
	D-LSTM^[7]	16.18	28.17	25.31	852	5550	6402
	MODBNE^[8]	12.51	28.66	25.11	421.91	6557.62	6979.53

指标	FD003			FD004
指标	RMSE	Score	匹配时间/s	RMSE	Score	匹配时间/s
E_MSEN	15.77	1247.49	106	23.68	3808.86	718
E_MFM-MSEN	15.07	609.97	49	22.22	2641.05	374
Λ/%	4.43	51.1	53.8	6.17	30.7	47.9