Reliability Engineering & System Safety最新文献_第4页

Universal domain adaptation in rotating machinery fault diagnosis: A self-supervised orthogonal clustering approach

IF 9.4 1区工程技术

Reliability Engineering & System Safety Pub Date : 2025-01-22 DOI: 10.1016/j.ress.2025.110828

Yang Liu, Aidong Deng, Geng Chen, Yaowei Shi, Qinyi Hu

{"title":"Universal domain adaptation in rotating machinery fault diagnosis: A self-supervised orthogonal clustering approach","authors":"Yang Liu, Aidong Deng, Geng Chen, Yaowei Shi, Qinyi Hu","doi":"10.1016/j.ress.2025.110828","DOIUrl":"10.1016/j.ress.2025.110828","url":null,"abstract":"<div><div>The development of fault diagnosis has been significantly advanced by progress in domain adaptation (DA). Universal Domain Adaptation (UniDA) has garnered considerable attention for its ability to eliminate the assumptions about the target labeling space, effectively handling various scenarios including closed set, partial set, open set, and open-partial set. However, existing UniDA methods often rely heavily on supervised learning within the source domain and fail to adequately explore the intrinsic data structure of the target domain. This limitation hinders the model’s ability to recognize unknown faults and reduces its domain adaptation performance. To address this issue, we propose a self-supervised orthogonal clustering network (SSOCN) for UniDA. The core idea of SSOCN fully leverages the structure of the target data to learn discriminative features and achieves adaptive clustering of target domain samples. By using source class centers as clustering points, SSOCN facilitates instance-level feature alignment, enabling the model to effectively address arbitrary category gaps. Furthermore, orthogonal regularization and a known–unknown separation strategy are incorporated to ensure feature orthogonality across different classes and to enhance the recognition of unknown samples, respectively. Extensive experiments across all sub-cases of UniDA demonstrate the effectiveness and superiority of the proposed method.</div></div>","PeriodicalId":54500,"journal":{"name":"Reliability Engineering & System Safety","volume":"257 ","pages":"Article 110828"},"PeriodicalIF":9.4,"publicationDate":"2025-01-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143286324","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Real-time adaptation for time-series signal prediction using label-aware neural processes

IF 9.4 1区工程技术

Reliability Engineering & System Safety Pub Date : 2025-01-22 DOI: 10.1016/j.ress.2025.110833

Seokhyun Chung , Raed Al Kontar

{"title":"Real-time adaptation for time-series signal prediction using label-aware neural processes","authors":"Seokhyun Chung , Raed Al Kontar","doi":"10.1016/j.ress.2025.110833","DOIUrl":"10.1016/j.ress.2025.110833","url":null,"abstract":"<div><div>Building a predictive model that rapidly adapts to real-time condition monitoring (CM) time-series data is critical for engineering systems/units. Unfortunately, many current methods suffer from a trade-off between representation power and agility in online settings. In this paper, we propose a neural process-based approach that addresses this trade-off. It encodes available observations within a CM signal into a representation space and then reconstructs the signal’s history and evolution for prediction. Once trained, the model can encode an arbitrary number of observations without requiring retraining, enabling on-the-spot real-time predictions along with quantified uncertainty and can be readily updated as more online data is gathered. Furthermore, our model is designed to incorporate partial information on qualitative factors (e.g., missing labels) from individual units. This integration not only enhances individualized predictions for each unit but also enables joint inference for both signals and their associated labels. Numerical studies on both synthetic and real-world data in degradation modeling highlight the advantageous features of our model in real-time adaptation, enhanced signal prediction with uncertainty quantification, and joint prediction for labels and signals.</div></div>","PeriodicalId":54500,"journal":{"name":"Reliability Engineering & System Safety","volume":"257 ","pages":"Article 110833"},"PeriodicalIF":9.4,"publicationDate":"2025-01-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143286633","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Knowledge distillation-based domain generalization enabling invariant feature distributions for damage detection of rotating machines and structures

IF 9.4 1区工程技术

Reliability Engineering & System Safety Pub Date : 2025-01-22 DOI: 10.1016/j.ress.2025.110842

Xiaoyou Wang , Jinyang Jiao , Xiaoqing Zhou , Yong Xia

{"title":"Knowledge distillation-based domain generalization enabling invariant feature distributions for damage detection of rotating machines and structures","authors":"Xiaoyou Wang , Jinyang Jiao , Xiaoqing Zhou , Yong Xia","doi":"10.1016/j.ress.2025.110842","DOIUrl":"10.1016/j.ress.2025.110842","url":null,"abstract":"<div><div>The poor generalization ability of machine learning (ML) models in the absence of sufficient labeled data remains a major challenge hindering their practical application. Domain generalization (DG) allows ML models trained on a set of source domains to generalize directly to related but unseen target domains. This capability makes DG particularly suitable for migrating ML model to unseen structures for online structural damage detection. A main branch of DG methods is domain-invariant feature learning, which involves aligning extracted embedding to make it invariant across domains. However, as the domain variety increases, extracting domain-invariant features becomes more challenging. This study rethinks DG from a novel three-stage knowledge distillation perspective. The first stage learns features with domain-invariant conditional distribution based on variational Bayesian inference. Multiple auxiliary domain-specific student models are designed to establish a bridge between unconditional and conditional variational inference. In the second stage, auxiliary student models learn from each other to additionally excavate features with domain-invariant marginal distribution. In the third stage, a student leader model distills knowledge from all auxiliary student models to enhance the model's robustness for final decision-making. The developed method is applied to civil and mechanical structures for damage detection. Results demonstrate that the developed method outperforms the state-of-the-art DG methods. Besides, the designed mechanism enables the student leader model to achieve superior performance compared to the teacher model.</div></div>","PeriodicalId":54500,"journal":{"name":"Reliability Engineering & System Safety","volume":"257 ","pages":"Article 110842"},"PeriodicalIF":9.4,"publicationDate":"2025-01-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143352253","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Mission optimal assignment of multi-mission systems under multiple phases with a shared component following an exponential lifetime distribution

IF 9.4 1区工程技术

Reliability Engineering & System Safety Pub Date : 2025-01-22 DOI: 10.1016/j.ress.2025.110824

Lirong Cui , Hang Ma , He Yi

引用次数: 0

Cascading dynamics in double-layer hypergraphs with higher-order inter-layer interdependencies

IF 9.4 1区工程技术

Reliability Engineering & System Safety Pub Date : 2025-01-22 DOI: 10.1016/j.ress.2025.110841

Chun-Xiao Jia, Run-Ran Liu

{"title":"Cascading dynamics in double-layer hypergraphs with higher-order inter-layer interdependencies","authors":"Chun-Xiao Jia, Run-Ran Liu","doi":"10.1016/j.ress.2025.110841","DOIUrl":"10.1016/j.ress.2025.110841","url":null,"abstract":"<div><div>Higher-order interactions are ubiquitous in complex systems where groups of elements interact collectively rather than through simple pairwise connections. This study investigates how higher-order inter-layer interdependencies affect the robustness and cascading dynamics in double-layer hypergraphs, in which groups of nodes in one layer are interdependent with groups of nodes in the other layer. We analyze how the average size of these interdependency groups across layers influences system robustness and the nature of phase transitions. First, as the average size of higher-order inter-layer interdependency groups increases, the system undergoes a change from a second-order to a first-order phase transition. Second, in scale-free hypergraphs, larger interdependency groups can cause two consecutive first-order transitions, indicating nontrivial dynamical behavior that may affect the predictability and controllability of the system. Finally, the impact of higher-order inter-layer interdependencies on system robustness is not straightforward, i.e., it does not monotonically increase or decrease across both scale-free and random hypergraphs. Specifically, intermediate-sized interdependency groups result in the most vulnerable system, as indicated by the highest critical point for system disintegration. These results give useful views into how multilayer network architectures with higher-order interdependencies affect the resilience and stability of interconnected systems, offering practical guidance for designing more robust infrastructure networks.</div></div>","PeriodicalId":54500,"journal":{"name":"Reliability Engineering & System Safety","volume":"257 ","pages":"Article 110841"},"PeriodicalIF":9.4,"publicationDate":"2025-01-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143286321","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Integrating physical knowledge for generative-based zero-shot learning models in process fault diagnosis

IF 9.4 1区工程技术

Reliability Engineering & System Safety Pub Date : 2025-01-22 DOI: 10.1016/j.ress.2025.110852

Guoqing Mu , Ching-Lien Liu , Junghui Chen

{"title":"Integrating physical knowledge for generative-based zero-shot learning models in process fault diagnosis","authors":"Guoqing Mu , Ching-Lien Liu , Junghui Chen","doi":"10.1016/j.ress.2025.110852","DOIUrl":"10.1016/j.ress.2025.110852","url":null,"abstract":"<div><div>Despite longstanding operational processes, persistent undiagnosed issues and inefficiencies continue to exist. Specifically, the collected data with specified faults is often either non-existent or sparse. The challenge lies in the absence of specified faults for reliable training of fault diagnosis models in such processes. This study proposes an innovative physical knowledge-guided zero-sample fault diagnosis method, which decomposes process variables into states and defines attributes based on domain knowledge. This transformation of variables into the attribute space replaces the traditional label space. The method involves three key steps: (1) Constructing the Seen Fault Latent Space: Utilizing seen fault data through the Conditional Variational Auto-Encoder model and Linear Discriminant Analysis in the latent space to classify seen faults. (2) Extending the Model Space with Unseen Fault Attributes: Using attributes to extend the unknown fault space and introducing a discriminator to ensure accurate separation of seen and unseen faults. (3) Retraining the Model: Using the data generated in the second step to retrain the model, enabling the diagnosis of both seen and unseen faults by the encoder. Experiments on numerical, continuous stirred tank reactor, and three-level tank examples demonstrate a significant 11 % improvement in classification accuracy for unseen fault samples compared to traditional methods.</div></div>","PeriodicalId":54500,"journal":{"name":"Reliability Engineering & System Safety","volume":"258 ","pages":"Article 110852"},"PeriodicalIF":9.4,"publicationDate":"2025-01-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143379304","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

SDCGAN: A CycleGAN-based single-domain generalization method for mechanical fault diagnosis

IF 9.4 1区工程技术

Reliability Engineering & System Safety Pub Date : 2025-01-22 DOI: 10.1016/j.ress.2025.110854

Yu Guo , Xiangyu Li , Jundong Zhang , Ziyi Cheng

{"title":"SDCGAN: A CycleGAN-based single-domain generalization method for mechanical fault diagnosis","authors":"Yu Guo , Xiangyu Li , Jundong Zhang , Ziyi Cheng","doi":"10.1016/j.ress.2025.110854","DOIUrl":"10.1016/j.ress.2025.110854","url":null,"abstract":"<div><div>In recent years, fault diagnosis based on domain generalization has attracted increasing attention as an effective approach to address the challenge of domain shift. most existing approaches depend on learning domain-invariant representations from multiple source domains, limiting their practical application in fault diagnosis. To address this issue, this paper introduces a single-domain generalization method for mechanical fault diagnosis, the Single-Domain Cycle Generative Adversarial Network (SDCGAN). A CycleGAN-based domain generation module is introduced to produce extended domains that exhibit substantial divergence from the source domain, enhancing the model's generalization capability. The diagnostic task module subsequently extracts domain-invariant features from both the source and extended domains. Furthermore, an adversarial contrastive training strategy is employed to learn generalized features robust to unknown domain shifts. Comprehensive experiments on two mechanical datasets verify the effectiveness of the proposed method, while ablation studies validate the contributions of its components, highlighting its potential for real-world applications.</div></div>","PeriodicalId":54500,"journal":{"name":"Reliability Engineering & System Safety","volume":"258 ","pages":"Article 110854"},"PeriodicalIF":9.4,"publicationDate":"2025-01-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143387582","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Nataf-based probabilistic buffeting prediction of overhead transmission lines under multi-dimensional turbulent wind excitation

IF 9.4 1区工程技术

Reliability Engineering & System Safety Pub Date : 2025-01-21 DOI: 10.1016/j.ress.2025.110846

Wen-Long Du , Xing Fu , Deng-Jie Zhu , Gang Li , Hong-Nan Li , Zeng-Hao Huang , Ying-Zhou Liu

{"title":"Nataf-based probabilistic buffeting prediction of overhead transmission lines under multi-dimensional turbulent wind excitation","authors":"Wen-Long Du , Xing Fu , Deng-Jie Zhu , Gang Li , Hong-Nan Li , Zeng-Hao Huang , Ying-Zhou Liu","doi":"10.1016/j.ress.2025.110846","DOIUrl":"10.1016/j.ress.2025.110846","url":null,"abstract":"<div><div>The overhead transmission line (OTL) is a complex nonlinear system susceptible to turbulence-induced buffeting. Classical one-dimensional buffeting analysis overlooks potential vertical turbulence. Additionally, the turbulence parameters associated with stochastic wind fields are simply assumed to be deterministic. Such simplified schemes might cause severe misestimation of extreme responses. Therefore, this paper presents a Nataf-based probabilistic buffeting prediction framework of OTLs under multi-dimensional turbulent wind. Initially, a two-dimensional influence line method is introduced to replace the time-consuming nonlinear finite element analysis (NFEA). Subsequently, Nataf transformation is employed to generate turbulence parameter samples, effectively preserving the marginal probability distribution and correlation structure. Then, the three-dimensional wind speeds are simultaneously synthesized, and the impact of turbulence parameter uncertainty on wind field characteristics is investigated. Next, the time-frequency buffeting responses are analytically estimated, and a quantile-based approach is proposed to quantify the uncertainty of extreme buffeting responses. Finally, the stochastic sensitivity analyses are performed to determine the relative sensitivity of various responses to each turbulence parameter. The results demonstrate that the two-dimensional influence line method exhibits significant efficiency advantages; All deterministic extreme responses are less than the 0.5 quantile of uncertain results; The top three turbulence parameters in sensitivity ranking are identical under different conditions.</div></div>","PeriodicalId":54500,"journal":{"name":"Reliability Engineering & System Safety","volume":"257 ","pages":"Article 110846"},"PeriodicalIF":9.4,"publicationDate":"2025-01-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143219616","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Designing random collaborative warranty and customizing maintenance strategies for systems subject to mission cycles

IF 9.4 1区工程技术

Reliability Engineering & System Safety Pub Date : 2025-01-20 DOI: 10.1016/j.ress.2025.110843

Lijun Shang , Baoliang Liu , Rui Peng

{"title":"Designing random collaborative warranty and customizing maintenance strategies for systems subject to mission cycles","authors":"Lijun Shang , Baoliang Liu , Rui Peng","doi":"10.1016/j.ress.2025.110843","DOIUrl":"10.1016/j.ress.2025.110843","url":null,"abstract":"<div><div>Existing warranties are made by manufacturers subjectively and unilaterally with the aims of cutting costs, increasing sales volumes, and so on. However, they fail to meet users' individual needs such as effectively managing reliability or ensuring higher and stable production capacities. In response to this situation, the current paper puts forward random collaborative warranty models subject to mission cycles. In these models, manufacturers and users jointly set terms to fit users' actual needs. These warranties not only fulfill personalized needs but also function as value-added services, thus creating profit-making opportunities for manufacturers. Moreover, aiming to effectively manage the post-warranty reliability of systems, the paper devises post-warranty maintenance strategies such as trivariate random replacement with preventive maintenance and bivariate random combination replacement. The strategies' terms are based on different usage rate scenarios of the post-warranty system, which are determined by whether limited missions are completed before a specific time. The proposed solutions are modeled quantitatively, and numerical analyses are conducted on some typical aspects to uncover the underlying mechanisms and derive valuable insights. One notable finding is that users can achieve lower repair costs if they actively apply the two-scale approach following the 'whichever expires last' principle.</div></div>","PeriodicalId":54500,"journal":{"name":"Reliability Engineering & System Safety","volume":"257 ","pages":"Article 110843"},"PeriodicalIF":9.4,"publicationDate":"2025-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143219614","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A variational Bayesian deep reinforcement learning approach for resilient post-disruption recovery of distribution grids

IF 9.4 1区工程技术

Reliability Engineering & System Safety Pub Date : 2025-01-17 DOI: 10.1016/j.ress.2025.110840

Zhaoyuan Yin , Chao Fang , Yiping Fang , Min Xie

{"title":"A variational Bayesian deep reinforcement learning approach for resilient post-disruption recovery of distribution grids","authors":"Zhaoyuan Yin , Chao Fang , Yiping Fang , Min Xie","doi":"10.1016/j.ress.2025.110840","DOIUrl":"10.1016/j.ress.2025.110840","url":null,"abstract":"<div><div>The increasing frequency of natural hazards and the resulting disruptions in distribution systems emphasize the urgent need for resilient post-disruption recovery. Due to high computational complexities in solving large-scale models, traditional optimization methods face significant challenges in emergency responses for grid recovery planning. Deep reinforcement learning (DRL) is emerging as a promising alternative for tackling combinatorial problems but struggles with uncertainty quantification and the curse of dimensionality. To address these issues and promote resilient post-disruption restoration, this study proposes a variational Bayesian DRL approach for developing optimal repair strategies after system disruptions. Recovery planning is framed as a Markov Decision Process, integrating data-driven environmental interactions and optimization-based reward evaluation. The Bayesian deep Q-network, using stochastic variational inference, quantifies the inherent uncertainties of component repair times due to varying repair resources and environmental stochasticity. A novel variational Bayesian dueling double deep Q-network is designed to mitigate the challenges in estimating Q-values within large state and action spaces. Case studies of real-world emergencies in the Hong Kong region show that the proposed methodology is effective and robust compared to several alternative approaches. The analysis on numerical results provides valuable insights for emergency restoration.</div></div>","PeriodicalId":54500,"journal":{"name":"Reliability Engineering & System Safety","volume":"257 ","pages":"Article 110840"},"PeriodicalIF":9.4,"publicationDate":"2025-01-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143219610","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0