A Data-Driven Modeling Method for Stochastic Nonlinear Degradation Process With Application to RUL Estimation

Zhang, Yuhan; Yang, Ying; Li, He; Xiu, Xianchao; Liu, Wanquan

doi:10.1109/tsmc.2021.3073052

Cited by 16 publications

(5 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…, θ m ] T . The Gaussian kernel function is usually selected in nonlinear model prediction [44]. Then, a generalized nonlinear Wiener process degradation (GNWP) model is formulated by combining ( 12) and ( 14) as follows…”

Section: Rul Predictionmentioning

confidence: 99%

A Deep Learning Feature Fusion Based Health Index Construction Method for Prognostics Using Multiobjective Optimization

et al. 2023

View full text Add to dashboard Cite

Degradation modeling and prognostics serve as the basis for system health management. Recently, various sensors provide plentiful monitoring data that can reflect the system status. A multitude of feature fusion techniques based on multisensor data have been proposed to generate a composite health index (HI) for prognostics, which can represent the underlying degradation mechanism. Most existing methods have used linear fusion models and neglected the practical requirements for HI construction, which are insufficient to reveal the nonlinear relations among features and difficult to obtain accurate HIs for complicated systems. This study proposes a novel feature fusion-based HI construction method with deep learning and multiobjective optimization. Multiple degradation features are fused by a deep neutral network (DNN). Several desired properties that the HIs should have for prognostics are adopted to formulate the objective functions of DNN training. To balance the spatial complexity and performance of the fusion model, a multiobjective optimization model is generated for training the DNN. Then, a generalized nonlinear Wiener process model is used to predict the remaining useful life with the resulted HIs. Finally, two cases are analyzed to illustrate the effectiveness and robustness of the proposed method.

show abstract

Section: Rul Predictionmentioning

confidence: 99%

A Deep Learning Feature Fusion Based Health Index Construction Method for Prognostics Using Multiobjective Optimization

et al. 2023

View full text Add to dashboard Cite

show abstract

“…Note that Zhang et al (2021) used a Gaussian basis function as part of RVM to represent the degradation drift increment for RUL estimation, which is the same as the proposed KWP model. However, the main limitation of KWP is that the prediction errors can accumulate in the long-term task, where the degradation mechanism is not explicit and the future data is unavailable.…”

Section: The Proposed Collaborative Modeling With Transfer Learning F...mentioning

confidence: 99%

“…Wang et al (2019) developed an improved Wiener process model with adaptive drift and diffusion for online RUL prediction. Zhang et al (2021) combined a nonlinear Wiener process model with a relevance vector machine to depict the nonlinearity of degradation without priors. Zhang et al (2022) studied a nonlinear Wiener process model with a random time-varying covariate for degradation modeling and RUL prediction.…”

Section: Introductionmentioning

confidence: 99%

Collaborative Kernel-based Nonlinear Degradation Modeling with Transfer Learning for Remaining Useful Life Prediction

Chen,

Liu,

Zio

et al. 2022

Book of Extended Abstracts for the 32nd European Safety and Reliability Conference

View full text Add to dashboard Cite

A novel nonlinear collaborative modeling method for remaining useful life (RUL) prediction is proposed. This method uses a kernel-based Wiener process (KWP) model, which formulates a nonlinear drift function with the weighted combination of kernel functions. Compared with the existing Wiener process models, this kind of modeling allows characterizing the nonlinearity of degradation more accurately and flexibly. To address the problem of error accumulation and lack of data in long-term prediction, a transfer learning scheme, based on the KWP models, is developed by leveraging multiple historical degradation trends from different units to collaboratively describe the degradation process of interest, even with limited data. The positive transfer learning is realized by introducing cross-correlations into the drift functions for obtaining more robust and accurate results than those obtained by constructing multiple individual models, one for each unit. The unknown model parameters are estimated by a spare estimation algorithm. Then, based on the KWP model, a close-form expression of the RUL distribution is derived for uncertainty quantification. An online framework is also proposed to iteratively predict the RUL. Finally, the proposed method is verified on lithium-ion battery datasets and compared to existing methods. The outcomes demonstrate the effectiveness and superiority of the proposed method for RUL prediction.

show abstract

“…To characterize the nonlinear degradation process, Si et al [30] first proposed a general nonlinear degradation model and obtained the probability density function (PDF) of RUL based on a well-known time-space transformation. After this pioneering work, extensive research has been conducted on the RUL prediction of nonlinear degradation devices [31][32][33][34].…”

Section: Introductionmentioning

confidence: 99%

Remaining Useful Life Prediction for Two-Phase Hybrid Deteriorating Lithium-Ion Batteries Using Wiener Process

Cui,

Lu,

Han

2024

IEEE Access

View full text Add to dashboard Cite

Owing to operating condition switching and internal degradation mechanisms, the degradation processes of some lithium-ion batteries (LIBs) exhibit non-monotone and two-phase patterns, which are composed of a linear first phase and a nonlinear second phase. The existing Gamma process and Inverse Gaussian process methods are limited to modeling the monotone degradation data. Besides, traditional singlephase nonlinear models and two-phase linear models are insufficient to describe such a degradation process effectively. Therefore, degradation modeling and remaining useful life (RUL) prediction of the hybrid deteriorating LIBs is still a compelling practical issue. In this paper, a two-phase hybrid degradation model with a linear first phase and a nonlinear second phase is formulated based on the widely used Wiener processbased model. Taking into account the random effects caused by the unit heterogeneity and the uncertainty of the degradation state at the changing point, we obtain the analytical solutions of the lifetime estimation and RUL prediction under the concept of the first passage time (FPT). In addition, to conduct model parameter identification, the expectation maximization (EM) algorithm in conjunction with a profile log-likelihood function method are utilized for offline parameter estimation. Subsequently, the Bayesian rule is adopted to conduct the online parameter updating. Finally, the numerical and practical experiments are provided for verification and show that the proposed method could achieve high estimation accuracy for the RUL prediction of the two-phase hybrid deteriorating LIBs.INDEX TERMS Lithium-ion batteries, RUL prediction, two-phase degradation, unit-to-unit variability, Wiener process.

show abstract

A Data-Driven Modeling Method for Stochastic Nonlinear Degradation Process With Application to RUL Estimation

Cited by 16 publications

References 30 publications

A Deep Learning Feature Fusion Based Health Index Construction Method for Prognostics Using Multiobjective Optimization

A Deep Learning Feature Fusion Based Health Index Construction Method for Prognostics Using Multiobjective Optimization

Collaborative Kernel-based Nonlinear Degradation Modeling with Transfer Learning for Remaining Useful Life Prediction

Remaining Useful Life Prediction for Two-Phase Hybrid Deteriorating Lithium-Ion Batteries Using Wiener Process

Contact Info

Product

Resources

About