Reliability Engineering & System Safety最新文献

{"title":"Enhancing battery health estimation using incomplete charging curves and knowledge-guided deep learning","authors":"Kui Zhang ,&nbsp;Safwat Khair Rayeem ,&nbsp;Weijie Mai ,&nbsp;Jinpeng Tian ,&nbsp;Liang Ma ,&nbsp;Tieling Zhang ,&nbsp;C.Y. Chung","doi":"10.1016/j.ress.2025.111211","DOIUrl":"10.1016/j.ress.2025.111211","url":null,"abstract":"<div><div>Accurately monitoring the health degradation of batteries is critical for ensuring the safety and reliability of electrochemical energy storage systems. While great progress has been made in this area by machine learning, a significant but overlooked challenge is the prohibitive demand for large high-quality degradation data for model training. Here, we reveal that although ubiquitous incomplete charging curves cannot derive state of health (SOH) labels, they provide valuable degradation knowledge that can effectively contribute to the development of SOH estimation models. We propose a knowledge-guided method that has high flexibility in using charging segments to estimate the SOH. More importantly, it takes advantage of incremental capacity indicators obtained from incomplete charging curves to guide the training of a deep neural network (DNN). The resulting pre-trained DNN can quickly adapt to SOH estimation with very few SOH labels. Validations show that using only 50 charging segments with SOH labels, our knowledge-guided DNN achieves an SOH estimation root mean square error of 21.08 mAh for 0.74 Ah batteries, which is 60 % less compared with conventional methods. Further validations on 3 datasets covering different battery chemistries and operating conditions confirm the superiority of the proposed method. Our work highlights the promise of employing domain knowledge to advance machine learning models for health monitoring purposes.</div></div>","PeriodicalId":54500,"journal":{"name":"Reliability Engineering & System Safety","volume":"262 ","pages":"Article 111211"},"PeriodicalIF":9.4,"publicationDate":"2025-05-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143929102","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}

{"title":"A non-probabilistic reliability-based design optimization method via dimensional decomposition-aided Chebyshev metamodel","authors":"Heng Zhao ,&nbsp;Chao Fu ,&nbsp;Yaqiong Zhang ,&nbsp;Zhiqiang Wan ,&nbsp;Kuan Lu","doi":"10.1016/j.ress.2025.111208","DOIUrl":"10.1016/j.ress.2025.111208","url":null,"abstract":"<div><div>In high-dimensional interval uncertainty optimization problems, traditional methods often face the challenge of the “curse of dimensionality”. To address this problem, this paper proposes an efficient non-probabilistic reliability-based design optimization method to improve the reliability and safety of the system. First, the coefficients of Chebyshev polynomials are efficiently computed by decomposing the high-dimensional problem into multiple low-dimensional subproblems. A dimensional decomposition-aided Chebyshev metamodel balances the accuracy and efficiency of interval analysis, replacing inner-layer optimization in the traditional two-layer nested optimization framework. Furthermore, the proposed method transforms the uncertainty optimization problem into a deterministic bi-objective optimization problem by using the interval order relation and non-probabilistic reliability theory. Then, the bi-objective optimization problem is reduced to an unconstrained single-objective optimization problem using the linear weighting method and the penalty function approach. To enhance the stability and global convergence of the optimization process, a new meta-heuristic optimization algorithm, the snake optimizer, is introduced in this paper. The effectiveness and accuracy of the proposed method in improving the safety and reliability of engineering systems are verified through numerical examples and an aero-engine dual-rotor system. The proposed method does not depend on the derivative information of the objective function or constraints, which is especially suitable for complex “black-box” engineering uncertainty optimization problems and has a wide range of engineering applications.</div></div>","PeriodicalId":54500,"journal":{"name":"Reliability Engineering & System Safety","volume":"262 ","pages":"Article 111208"},"PeriodicalIF":9.4,"publicationDate":"2025-05-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143924395","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}

{"title":"An active learning method based on Monte Carlo dropout neural network for high-dimensional reliability analysis","authors":"Huabin Sun,&nbsp;Yuequan Bao","doi":"10.1016/j.ress.2025.111169","DOIUrl":"10.1016/j.ress.2025.111169","url":null,"abstract":"<div><div>In structural reliability analysis, the AK-MCS method, which combines Kriging and Monte Carlo Simulation, is well-acknowledged for its effectiveness but struggles with accuracy and efficiency in high-dimensional and nonlinear scenarios. To leverage the advantages and circumvent the limitations of AK-MCS, an active learning method based on the Monte Carlo dropout (MC-dropout) neural network is proposed. The MC-dropout neural network-based surrogate model provides both predictive mean and standard deviation in complex scenarios with a limited number of samples. By identifying candidate samples and utilizing a learning function that considers predictive mean and standard deviation, the method selects new samples close to the limit state surface with significant uncertainties to update the surrogate model. An ensemble of MC-dropout neural networks is then used to obtain a reliable failure probability. Two convergence criteria are introduced to determine the termination of the active learning process. Two numerical examples, a cantilever beam and an actual cable-stayed bridge are used to demonstrate the efficacy of the proposed method. The results show that the MC-dropout neural network-based surrogate model exhibits adaptivity and flexibility in handling high-dimensional and nonlinear scenarios and the proposed method achieves a relatively accurate failure probability with a limited number of samples.</div></div>","PeriodicalId":54500,"journal":{"name":"Reliability Engineering & System Safety","volume":"262 ","pages":"Article 111169"},"PeriodicalIF":9.4,"publicationDate":"2025-05-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143908106","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}

{"title":"Evolutionary safety evaluation of a truss bridge using dynamic Bayesian networks","authors":"Jia-Li Tan ,&nbsp;Sheng-En Fang","doi":"10.1016/j.ress.2025.111201","DOIUrl":"10.1016/j.ress.2025.111201","url":null,"abstract":"<div><div>Timely safety evaluation of a real-world complex truss structure is difficult because of complex failure modes, uncertainty influence and difficulty in theoretical deduction. Therefore, a dynamic Bayesian network (DBN) has been designed to demonstrate the safety evolution process of a truss bridge under a time-varying load. The DBN comprises a prior network and a transition network, forming different time slices. Its network nodes represent the truss members and system, and the discrete nodal variables indicate the probabilities for safety and failure states. An effective network topology definition method is proposed by incorporating a hybrid topology learning strategy with a virtual substructure division strategy. The two strategies provide a rational topology with the reduced dimensions of conditional probability tables for complex truss structures. Numerical observation data are generated for learning the conditional probabilities between connected nodes in both the prior and transition networks. Subsequently, state probability inference between different time slices can be achieved using measured observation data from a limited number of members at a given time as the evidence. Afterwards, the failure state probability evolution curve of the truss bridge system can be described. The validation on an experimental truss bridge model has successfully demonstrated its state evolution under the different loading periods. The failure time of the truss system was predicted, which well accorded with the experimental observations.</div></div>","PeriodicalId":54500,"journal":{"name":"Reliability Engineering & System Safety","volume":"262 ","pages":"Article 111201"},"PeriodicalIF":9.4,"publicationDate":"2025-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143912203","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}

{"title":"A quantitative framework for performance-based reliability prediction for a multi-component system subject to dynamic self-reconfiguration","authors":"Zhiyi Huang,&nbsp;Yian Wei,&nbsp;Yao Cheng","doi":"10.1016/j.ress.2025.111188","DOIUrl":"10.1016/j.ress.2025.111188","url":null,"abstract":"<div><div>In a multi-component system, the performance of all components might be restricted by the most degraded component. This dependency results in an undesirable performance loss of the system. To date, engineers have developed a performance-maximization-oriented technique that enables dynamic isolation and retrieval of the components from and back to the system to mitigate the dependency-induced negative impact. Despite its engineering application, the technique’s effectiveness in system performance enhancement still lacks systematic explorations. In this paper, we fill the gap by developing a quantitative framework for the system’s performance-based reliability metrics prediction, considering the technique (defined as dynamic self-reconfiguration mechanism in this paper) may function perfectly or imperfectly, and the real-time system information may be unavailable or partially available with biases. First, we analytically characterize the mechanism by modeling the probability distribution of the system configuration, building on which we proactively predict the system’s performance-based reliability metrics. Afterward, we develop a particle filtering algorithm to utilize the noisy multi-dimensional-multi-type real-time information for progressive system state estimation and reliability prediction. Based on the prediction models, we quantify the effectiveness of the dynamic self-reconfiguration mechanism, which assists operators in system reliability enhancement. A case study of a photovoltaic system is provided.</div></div>","PeriodicalId":54500,"journal":{"name":"Reliability Engineering & System Safety","volume":"262 ","pages":"Article 111188"},"PeriodicalIF":9.4,"publicationDate":"2025-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143924269","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}

{"title":"A time-efficient solution approach for multi/many-task reliability redundancy allocation problems using the online transfer parameter estimation based multifactorial evolutionary algorithm","authors":"Md. Abdul Malek Chowdury ,&nbsp;Rahul Nath ,&nbsp;Amit Rauniyar ,&nbsp;Amit K. Shukla ,&nbsp;Pranab K. Muhuri","doi":"10.1016/j.ress.2025.111175","DOIUrl":"10.1016/j.ress.2025.111175","url":null,"abstract":"<div><div>This paper introduces a time efficient solution approach for multi/many-task RRAP under the framework of the novel online transfer parameter estimation based multi-factorial evolutionary algorithm (MFEA-II). To represent similarity between tasks, the basic MFEA utilizes a single value for transfer parameter leading to negative knowledge transfer during the evolution process as different pair of tasks often have different level of similarity. Proposed MFEA-II based solution approach avoids above problem while solving RRAPs simultaneously by employing online transfer parameter estimation based MFEA-II. To demonstrate the efficiency of the proposed approach, two set of problems (or test sets) are considered with more than two RRAPs. The test set-1 (TS-1) portray the scenario of multi-tasking by considering three problems while test set-2 (TS-2) considers the many-tasking scenario with four problems. The TS-1 includes three RRAP problems: a series system, a complex bridge system, and a series-parallel system. The TS-2 includes these three problems plus a new RRAP problem: the over-speed protection system of a gas turbine. We address each test set using the MFEA-II framework by incorporating the solution structures of all problems into a single solution. For comparison, basic MFEA is utilized to solve each test sets similar to MFEA-II. Subsequently, each problem is also solved independently using genetic algorithms (GA) and particle swarm optimization (PSO). The simulation results are evaluated based on the average of the best reliability, total computation time, performance ranking, and statistical significance tests. The outcome shows that even if the number of tasks increases in a multi-tasking environment, our proposed approach can generate better results compared to basic MFEA as well as single-task optimizer. Moreover, in terms of computation time, the proposed approach provides 6.96 % deteriorated and 2.46 % improved values compared to basic MFEA in TS-1 &amp; TS-2, respectively. In comparison to single task optimizer, proposed MFEA-II provides 40.60 % and 53.43 % faster than GA and 52.25 % and 62.70 % faster than PSO for TS-1 and TS-2, respectively. Further, to rank the algorithm in terms of quality of reliability values and computation time, the multi-criteria decision-making method named TOPSIS method is utilized, where the proposed approach secured the top rank.</div></div>","PeriodicalId":54500,"journal":{"name":"Reliability Engineering & System Safety","volume":"264 ","pages":"Article 111175"},"PeriodicalIF":9.4,"publicationDate":"2025-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144205420","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}

{"title":"Data-driven decision support system for the safety management of railway bridge networks","authors":"Giorgio Monti ,&nbsp;Raihan Rahmat Rabi ,&nbsp;Luca Marella ,&nbsp;Sergio Tremi Proietti","doi":"10.1016/j.ress.2025.111202","DOIUrl":"10.1016/j.ress.2025.111202","url":null,"abstract":"<div><div>Ensuring the safety of railway bridges across extensive networks requires automated, reliable systems that combine precise monitoring with actionable decision-making. This study introduces a fully automated, model-free, data-driven Decision Support System (DSS) designed to assist railway network operators in managing bridge safety efficiently and proactively. The DSS uses data from diverse sensors to monitor critical damage mechanisms, such as differential settlement, pier tilt, abnormal deck torsion, and flexibility retention ratio in bridge decks. These damage indicators are assessed against mechanics-based and code-based thresholds, ensuring accuracy and robustness in detection. The DSS adopts a probabilistic framework that defines multi-level alarms—low, medium, and high—triggered by pre-established misclassification probabilities. This approach minimizes both false and missed alarms, enhancing the system reliability. A global damage index aggregates local damage indicators, enabling the ranking of monitored bridges by urgency for intervention. This index guides operators in prioritizing inspections, maintenance, or safety measures, optimizing resource allocation across large-scale networks. The methodology was validated through case studies on steel and reinforced concrete bridges, demonstrating its scalability and effectiveness in identifying structural damage, reducing false alarms, and supporting timely decision-making. The proposed DSS represents a significant step toward smart, self-diagnosing railway networks, offering a scalable solution to enhance infrastructure safety and operational efficiency. Future advancements could include integrating predictive algorithms and expanding applicability to diverse environmental conditions and bridge typologies.</div></div>","PeriodicalId":54500,"journal":{"name":"Reliability Engineering & System Safety","volume":"262 ","pages":"Article 111202"},"PeriodicalIF":9.4,"publicationDate":"2025-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143935831","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}

{"title":"Functionality assessment and improvement of medical buildings during the emergency: Incorporating resource restructuring and patient priority scheduling","authors":"Zhuoru Song ,&nbsp;Changhai Zhai ,&nbsp;Jin Liu ,&nbsp;Shunshun Pei ,&nbsp;Cheng Zhang","doi":"10.1016/j.ress.2025.111205","DOIUrl":"10.1016/j.ress.2025.111205","url":null,"abstract":"<div><div>This study proposes a comprehensive framework for post-earthquake functionality assessment of medical buildings. This framework incorporates the cascading damages of structural and non-structural components, medical equipment, and the functional interdependencies of subsystems by combining Bayesian networks and fault tree methods. Considering the emergency characteristics of medical buildings, the maximum treatable nominal number of earthquake casualties is proposed as the functionality indicator. The indicator is quantified by applying a Genetic Algorithm to a discrete event simulation model based on the Markov process. In this study, resource restructuring, medical room redistribution, and dynamic priority scheduling of casualties are introduced into the framework for the first time, which addresses the gaps in the dynamic changes and self-adaptation adjustments of emergency treatment scenarios. The operability and effectiveness of this framework are verified through numerical research on a hospital. This study analyzes the effects of cascading damages of components on hospital functionalities, revealing that hospital functionality is overestimated when the interdependence of components is ignored. Additionally, the study also validated the effectiveness of organizational management methods such as resource restructuring and redistribution, as well as dynamic priority scheduling of casualties, in improving hospital functionalities through parameter analysis methods.</div></div>","PeriodicalId":54500,"journal":{"name":"Reliability Engineering & System Safety","volume":"262 ","pages":"Article 111205"},"PeriodicalIF":9.4,"publicationDate":"2025-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143922446","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}

{"title":"Fractional wavelet synchrosqueezed transform for linear chirp signal: theory and damage detection by the electromechanical impedance based nonlinear wave modulation","authors":"Naserodin Sepehry ,&nbsp;Mohammad Ehsani ,&nbsp;AmirMasoud Chavoshi ,&nbsp;Hamidreza Amindavar","doi":"10.1016/j.ress.2025.111203","DOIUrl":"10.1016/j.ress.2025.111203","url":null,"abstract":"<div><div>The electromechanical impedance-based nonlinear wave modulation (EMI-NWM) presented in this paper is a novel damage detection technique based on the modulation of chirp signals. The approach deals with the difficult and time-consuming process of selecting the optimal frequency for the pump and carrier waves in the NWM-based structural health monitoring, making it more appropriate for real-time deployment. However, compared to when monoharmonic signals are utilized as excitations, the processing of the EMI-NWM recorded signals is more challenging. Time-frequency signal processing can assist in this regard, but some of the existing methods do not offer sufficient resolution to analyze EMI-NWM signals effectively. The fractional wavelet synchrosqueezed transform (FrWSST), an innovative time-frequency analysis approach that combines the advantages of fractional wavelet and synchrosqueezed transforms, is developed to address the issue. FrWSST parameters are tuned for linear chirp signals such as those used in EMI-NWM. EMI-NWM along with FrWSST is used to detect bolt loosening in sandwich beams. The proposed method's resistance to external noise and its effectiveness in damage identification is being examined. The results show that FrWSST is at least twofold more robust to noise than other time-frequency methods, making it a promising technique for real-world applications.</div></div>","PeriodicalId":54500,"journal":{"name":"Reliability Engineering & System Safety","volume":"262 ","pages":"Article 111203"},"PeriodicalIF":9.4,"publicationDate":"2025-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143924397","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}

{"title":"Response flow graph neural network for capacitated network reliability analysis","authors":"Yan Shi ,&nbsp;Cheng Liu ,&nbsp;Michael Beer ,&nbsp;Hong-Zhong Huang ,&nbsp;Yu Liu","doi":"10.1016/j.ress.2025.111198","DOIUrl":"10.1016/j.ress.2025.111198","url":null,"abstract":"<div><div>Capacitated network reliability (CNR) analysis is essential for computing the reliability of diverse networks. The NP-hard nature of CNR problems makes exact solutions through exhaustive permutations impractical for many real-world engineering networks. In this research, a new graph-based neural network termed the response flow graph neural network (RFGNN) is developed to address CNR problems. The innovation of the proposed method comprises three key components. Firstly, an iteration equation is proposed to update network link weights by identifying nodes where flow is obstructed during propagation. Secondly, a novel expression is developed to amalgamate local neighborhood information for each node by incorporating the updated link weights, culminating in the creation of the RFGNN. Thirdly, an adaptive framework is developed to improve the prediction accuracy of the RFGNN in solving CNR problems. Several CNR problems are presented to assess the efficacy of the developed method. The results unequivocally demonstrate the effectiveness of the developed method. Furthermore, the RFGNN exhibits remarkable computational accuracy when estimating CNRs across various sub-networks once it is appropriately constructed from the original network. This represents a capability that conventional non-machine learning methods typically struggle to attain.</div></div>","PeriodicalId":54500,"journal":{"name":"Reliability Engineering & System Safety","volume":"262 ","pages":"Article 111198"},"PeriodicalIF":9.4,"publicationDate":"2025-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143908107","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}