Xu Zhao;He Jiang;Xiaochen Li;Shikai Guo;Zhilei Ren;Peiyu Zou;Huijiang Liu
In electronic design automation (EDA), printed circuit board (PCB) design plays a crucial role. Ensuring the reliability of the PCB design tool chain is essential, as bugs in the tool chain can cause significant issues and losses during design and production. To improve reliability, a key process is to generate numerous PCB schematics and execute them in the tool chain, to test the correctness of each tool chain functionality. However, it is a challenge to automatically generate valid schematics to simulate the actual use of the PCB design tool chain. To this end, we propose PCBSmith, an effective schematic generator for PCB design tool chain. PCBSmith mimics the steps of a PCB designer for schematic design. PCBSmith first selects the appropriate electronic components from a comprehensive library and connects them according to the constraints of different components. PCBSmith then sets electrical parameters and simulation models for each component, eventually generating simulatable schematics. Experiments show that PCBSmith demonstrates high efficiency in schematic generation, averaging only one schematic per second. PCBSmith maintains a success rate over 61.44% for generating schematics, which outperforms the baseline method by 30.68%. The generated schematics have successfully identified unknown bugs in PCB design tools.
{"title":"PCBSmith: An Effective Schematic Generator for Testing PCB Design Tool Chain","authors":"Xu Zhao;He Jiang;Xiaochen Li;Shikai Guo;Zhilei Ren;Peiyu Zou;Huijiang Liu","doi":"10.1109/TR.2025.3529303","DOIUrl":"https://doi.org/10.1109/TR.2025.3529303","url":null,"abstract":"In electronic design automation (EDA), printed circuit board (PCB) design plays a crucial role. Ensuring the reliability of the PCB design tool chain is essential, as bugs in the tool chain can cause significant issues and losses during design and production. To improve reliability, a key process is to generate numerous PCB schematics and execute them in the tool chain, to test the correctness of each tool chain functionality. However, it is a challenge to automatically generate valid schematics to simulate the actual use of the PCB design tool chain. To this end, we propose PCBSmith, an effective schematic generator for PCB design tool chain. PCBSmith mimics the steps of a PCB designer for schematic design. PCBSmith first selects the appropriate electronic components from a comprehensive library and connects them according to the constraints of different components. PCBSmith then sets electrical parameters and simulation models for each component, eventually generating simulatable schematics. Experiments show that PCBSmith demonstrates high efficiency in schematic generation, averaging only one schematic per second. PCBSmith maintains a success rate over 61.44% for generating schematics, which outperforms the baseline method by 30.68%. The generated schematics have successfully identified unknown bugs in PCB design tools.","PeriodicalId":56305,"journal":{"name":"IEEE Transactions on Reliability","volume":"74 3","pages":"3281-3295"},"PeriodicalIF":5.7,"publicationDate":"2025-02-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144998034","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Qian Yang;Shailesh N. Joshi;Raymond Viviano;Hiroshi Ukegawa;Krishna R. Pattipati
Power electronic (PE) reliability is critical to electric vehicle performance and safety. Thus, it is vital to predict the remaining useful life (RUL) of components that are subject to predictable degradation. Here, we propose a RUL estimation framework for PE components. The framework has two consecutive phases: Generation of distance-based health indicators through an unsupervised learning procedure, such as self-organizing map (SOM) or K-means clustering, and subsequent deployment of interacting multiple model (IMM) that integrate linear and extended Kalman filters with varied degradation profiles to forecast future values of the indicator and RUL. Specifically, a nominal SOM or K-means model is learned, using the on-state median signal data from the PE component. The indicator is then calculated by measuring the distance between the test vector and the cluster center. To adaptively track the health indicator and its rate of change, accounting for the noise intrinsic to degradation processes, various degradation profiles, and the measurement system, the IMMs are applied. The RUL is evaluated as the difference between a predefined threshold and the health indicator estimate, divided by the present degradation rate. Validation of the framework involved accelerated aging experimental datasets, encompassing both low-frequency and high-frequency switching scenarios. The results reveal the framework's versatility and potential for implementation across diverse applications.
{"title":"A Distance-Based Health Indicator and Its Use in an Interacting Multiple Model for Failure Prognosis in Power Electronic Devices","authors":"Qian Yang;Shailesh N. Joshi;Raymond Viviano;Hiroshi Ukegawa;Krishna R. Pattipati","doi":"10.1109/TR.2025.3526594","DOIUrl":"https://doi.org/10.1109/TR.2025.3526594","url":null,"abstract":"Power electronic (PE) reliability is critical to electric vehicle performance and safety. Thus, it is vital to predict the remaining useful life (RUL) of components that are subject to predictable degradation. Here, we propose a RUL estimation framework for PE components. The framework has two consecutive phases: Generation of distance-based health indicators through an unsupervised learning procedure, such as self-organizing map (SOM) or K-means clustering, and subsequent deployment of interacting multiple model (IMM) that integrate linear and extended Kalman filters with varied degradation profiles to forecast future values of the indicator and RUL. Specifically, a nominal SOM or K-means model is learned, using the <sc>on</small>-state median signal data from the PE component. The indicator is then calculated by measuring the distance between the test vector and the cluster center. To adaptively track the health indicator and its rate of change, accounting for the noise intrinsic to degradation processes, various degradation profiles, and the measurement system, the IMMs are applied. The RUL is evaluated as the difference between a predefined threshold and the health indicator estimate, divided by the present degradation rate. Validation of the framework involved accelerated aging experimental datasets, encompassing both low-frequency and high-frequency switching scenarios. The results reveal the framework's versatility and potential for implementation across diverse applications.","PeriodicalId":56305,"journal":{"name":"IEEE Transactions on Reliability","volume":"74 3","pages":"4032-4046"},"PeriodicalIF":5.7,"publicationDate":"2025-02-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144998177","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The existing methods for distribution system reliability assessment mainly adopt the multiple iteration calculation modes of traversal search of failure scenarios one by one to get the failure effect area and degree. However, these methods are time-consuming and low-efficiency, due to the repeated searches and the complexity of failure effect analysis. Meanwhile, the reliability assessment results only representing the system's average level always cannot provide weak link information for the operators. In this article, an analytical calculation method for distribution system reliability index based on the improved failure effect incidence matrix (FEIM) is proposed. First, an incidence matrix-based modeling of the complete topological association of source-network-load of the distribution system is conducted, including node-branch relationships, segment switch's locations, and fuse locations, which are closely related to reliability analysis. And the rules for calculating the source-load power supply association matrix are also provided. Next, an improved FEIM model is established to analytically express the correlation between failure components and affected loads. Finally, the distribution system reliability index analytical calculation method based on the improved FEIM is presented. The proposed method is validated using the IEEE RBTS bus-6 case and a modified 96-node case. The results demonstrate that the method can significantly improve the computation efficiency while ensuring the accuracy of the results. Additionally, it can conveniently provide more efficient information on system bottlenecks and weak points for the reliability improvement of distribution systems.
{"title":"Analytical Calculation Method for Power Supply Reliability of Distribution Systems With Multiple Tie Lines","authors":"Fengzhang Luo;Nan Ge;Jing Xu","doi":"10.1109/TR.2025.3530983","DOIUrl":"https://doi.org/10.1109/TR.2025.3530983","url":null,"abstract":"The existing methods for distribution system reliability assessment mainly adopt the multiple iteration calculation modes of traversal search of failure scenarios one by one to get the failure effect area and degree. However, these methods are time-consuming and low-efficiency, due to the repeated searches and the complexity of failure effect analysis. Meanwhile, the reliability assessment results only representing the system's average level always cannot provide weak link information for the operators. In this article, an analytical calculation method for distribution system reliability index based on the improved failure effect incidence matrix (FEIM) is proposed. First, an incidence matrix-based modeling of the complete topological association of source-network-load of the distribution system is conducted, including node-branch relationships, segment switch's locations, and fuse locations, which are closely related to reliability analysis. And the rules for calculating the source-load power supply association matrix are also provided. Next, an improved FEIM model is established to analytically express the correlation between failure components and affected loads. Finally, the distribution system reliability index analytical calculation method based on the improved FEIM is presented. The proposed method is validated using the IEEE RBTS bus-6 case and a modified 96-node case. The results demonstrate that the method can significantly improve the computation efficiency while ensuring the accuracy of the results. Additionally, it can conveniently provide more efficient information on system bottlenecks and weak points for the reliability improvement of distribution systems.","PeriodicalId":56305,"journal":{"name":"IEEE Transactions on Reliability","volume":"74 3","pages":"3180-3191"},"PeriodicalIF":5.7,"publicationDate":"2025-02-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144998262","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The rapid growth of interconnected smart devices and advanced computing technologies in the industrial Internet of Things (IIoT) has significantly enhanced operational resilience and performance but also increased cybersecurity risks. While deep learning shows promise in IIoT security, it faces challenges due to the lack of labeled data and reliance on human expertise for unsupervised anomaly detection. To address these challenges, a novel automated adversarial deep learning-based unsupervised anomaly detection method called EvoAAE is proposed to optimize the hyperparameters and neural architectures of adversarial variational autoencoder (VAE) for securing IIoT. Specifically, a generative adversarial network-based VAE is employed to adversarially generate multivariate time series. Then, particle swarm optimization with an efficient binary encoding strategy is designed to evolve hyperparameters and neural architectures in adversarial VAE including batch size, learning rate, the type of optimizer, the number of convolutional layer, the number of kernels of convolutional layer, kernel size, the type of normalization layer, and the type of active function. The experimental results indicate that EvoAAE achieves notable performance across four IIoT datasets in industrial control domain, i.e., secure water treatment, water distribution, Mars Science Laboratory, and power system domain, i.e., power system attack with precision of 0.949, 0.8356, 0.972, and 0.981, recall of 0.971, 0.9214, 0.964, and 0.979, and $F_{1}$-score of 0.960, 0.8764, 0.968, and 0.980, respectively.
{"title":"Evolutionary Adversarial Autoencoder for Unsupervised Anomaly Detection of Industrial Internet of Things","authors":"Guo-Qiang Zeng;Yao-Wei Yang;Kang-Di Lu;Guang-Gang Geng;Jian Weng","doi":"10.1109/TR.2025.3528256","DOIUrl":"https://doi.org/10.1109/TR.2025.3528256","url":null,"abstract":"The rapid growth of interconnected smart devices and advanced computing technologies in the industrial Internet of Things (IIoT) has significantly enhanced operational resilience and performance but also increased cybersecurity risks. While deep learning shows promise in IIoT security, it faces challenges due to the lack of labeled data and reliance on human expertise for unsupervised anomaly detection. To address these challenges, a novel automated adversarial deep learning-based unsupervised anomaly detection method called EvoAAE is proposed to optimize the hyperparameters and neural architectures of adversarial variational autoencoder (VAE) for securing IIoT. Specifically, a generative adversarial network-based VAE is employed to adversarially generate multivariate time series. Then, particle swarm optimization with an efficient binary encoding strategy is designed to evolve hyperparameters and neural architectures in adversarial VAE including batch size, learning rate, the type of optimizer, the number of convolutional layer, the number of kernels of convolutional layer, kernel size, the type of normalization layer, and the type of active function. The experimental results indicate that EvoAAE achieves notable performance across four IIoT datasets in industrial control domain, i.e., secure water treatment, water distribution, Mars Science Laboratory, and power system domain, i.e., power system attack with precision of 0.949, 0.8356, 0.972, and 0.981, recall of 0.971, 0.9214, 0.964, and 0.979, and <inline-formula><tex-math>$F_{1}$</tex-math></inline-formula>-score of 0.960, 0.8764, 0.968, and 0.980, respectively.","PeriodicalId":56305,"journal":{"name":"IEEE Transactions on Reliability","volume":"74 3","pages":"3454-3468"},"PeriodicalIF":5.7,"publicationDate":"2025-01-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144998176","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In real-world applications, the diagnostic efficiency of rolling bearings is commonly affected by operating conditions like fluctuating rotating speed and varying loads, especially, environmental disturbances like transient noises. These disturbances tend to mask the indicators of damage, presenting substantial obstacles for accurately pinpointing failures. Traditional diagnostic methods struggle with the complexity and the noise sensitivity of such scenarios, often failing to accurately identify failure signs amidst multivariate random transient noise. To address these challenges, the current study proposes a method known as short-term Markov transition frequency peak rate. This method focuses on precisely tracking temporal state changes and identifying abnormal signals. It is aimed at mitigating transient noise interference at its source and enhancing insensitivity to external transient noise, which facilitates a more accurate and reliable selection of demodulation bands. Furthermore, an amplitude interference-limiting mechanism is designed within this method to discern and mitigate the impact of transient noise that may adversely affect the demodulation band selection process. The experimental results validate the effectiveness of this approach, demonstrating that it can reliably diagnose bearing faults even in the presence of transient disturbances.
{"title":"Interference Suppression of Nonstationary Signals for Bearing Diagnosis Under Transient Noise Measurements","authors":"Peng Chen;Yuhao Wu;Chaojun Xu;Cheng-Geng Huang;Mian Zhang;Junlin Yuan","doi":"10.1109/TR.2025.3527739","DOIUrl":"https://doi.org/10.1109/TR.2025.3527739","url":null,"abstract":"In real-world applications, the diagnostic efficiency of rolling bearings is commonly affected by operating conditions like fluctuating rotating speed and varying loads, especially, environmental disturbances like transient noises. These disturbances tend to mask the indicators of damage, presenting substantial obstacles for accurately pinpointing failures. Traditional diagnostic methods struggle with the complexity and the noise sensitivity of such scenarios, often failing to accurately identify failure signs amidst multivariate random transient noise. To address these challenges, the current study proposes a method known as short-term Markov transition frequency peak rate. This method focuses on precisely tracking temporal state changes and identifying abnormal signals. It is aimed at mitigating transient noise interference at its source and enhancing insensitivity to external transient noise, which facilitates a more accurate and reliable selection of demodulation bands. Furthermore, an amplitude interference-limiting mechanism is designed within this method to discern and mitigate the impact of transient noise that may adversely affect the demodulation band selection process. The experimental results validate the effectiveness of this approach, demonstrating that it can reliably diagnose bearing faults even in the presence of transient disturbances.","PeriodicalId":56305,"journal":{"name":"IEEE Transactions on Reliability","volume":"74 3","pages":"4047-4061"},"PeriodicalIF":5.7,"publicationDate":"2025-01-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144998139","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Traditional degradation-based reliability evaluation methods are typically based on rich data from a population of similar products, providing an average description of product performance. To capture individual characteristics for personalized maintenance, a dynamic reliability evaluation framework is proposed based on the individual monitoring data, which integrates a two-stage scheme and incorporates the physical model. The state-space model is first constructed based on Paris' Law to accurately describe bearing degradation, combining both physical mechanisms and secondary random factors. Then, an online stage division strategy based on an expanding time window is proposed, which implements change point detection and performs parameter estimation to serve as a priori information. Next, degradation state distributions and model parameters are adaptively estimated in the second stage using the extended Kalman filter, and the reliability is evaluated in real time based on the interval failure rate. Finally, to demonstrate the efficacy of the proposed framework, a comparative practical case study on bearing vibration data is presented.
{"title":"A Two-Stage Model-Based Dynamic Reliability Evaluation Method in Individual Monitoring: A Case Study on Bearing Vibration Data","authors":"Junling Wang;Xiaobing Ma;Yongbo Zhang","doi":"10.1109/TR.2025.3527128","DOIUrl":"https://doi.org/10.1109/TR.2025.3527128","url":null,"abstract":"Traditional degradation-based reliability evaluation methods are typically based on rich data from a population of similar products, providing an average description of product performance. To capture individual characteristics for personalized maintenance, a dynamic reliability evaluation framework is proposed based on the individual monitoring data, which integrates a two-stage scheme and incorporates the physical model. The state-space model is first constructed based on Paris' Law to accurately describe bearing degradation, combining both physical mechanisms and secondary random factors. Then, an online stage division strategy based on an expanding time window is proposed, which implements change point detection and performs parameter estimation to serve as a priori information. Next, degradation state distributions and model parameters are adaptively estimated in the second stage using the extended Kalman filter, and the reliability is evaluated in real time based on the interval failure rate. Finally, to demonstrate the efficacy of the proposed framework, a comparative practical case study on bearing vibration data is presented.","PeriodicalId":56305,"journal":{"name":"IEEE Transactions on Reliability","volume":"74 3","pages":"3799-3808"},"PeriodicalIF":5.7,"publicationDate":"2025-01-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144998072","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In recent years, many extended identity-based broadcast proxy re-encryption (IBPRE) schemes have been put forward. These schemes are flexible enough and feasible to various application scenarios, including conditional IBPRE, revocable IBPRE, and anonymous IBPRE. However, the existing extended IBPRE schemes are not able to simultaneously achieve fine-grained data sharing, identity privacy protection for authorized data users (DUs) and access privilege revocation. To this end, we put forward conditional identity-based broadcast proxy re-encryption with anonymity and revocation (CIBPRE-AR) and construct a concrete CIBPRE-AR scheme. The scheme implements fine-grained data sharing by associating conditions with the re-encryption key. The identity privacy protection for authorized DUs is provided by using Lagrange interpolation. Further, the access privileges of the violated DU are revoked by updating the re-encryption key. The indistinguishability of ciphertexts against chosen-plaintext attack and anonymity of the DUs are proved rigidly. Compared with existing similar schemes, only the CIBPRE-AR scheme simultaneously achieves fine-grained data access control, anonymity as well as revocation. The proposed scheme also has advantage with respect to computation cost.
{"title":"Conditional Identity-Based Broadcast Proxy Re-Encryption With Anonymity and Revocation","authors":"Liqing Chen;Meng Zhang;Jiguo Li","doi":"10.1109/TR.2024.3521507","DOIUrl":"https://doi.org/10.1109/TR.2024.3521507","url":null,"abstract":"In recent years, many extended identity-based broadcast proxy re-encryption (IBPRE) schemes have been put forward. These schemes are flexible enough and feasible to various application scenarios, including conditional IBPRE, revocable IBPRE, and anonymous IBPRE. However, the existing extended IBPRE schemes are not able to simultaneously achieve fine-grained data sharing, identity privacy protection for authorized data users (DUs) and access privilege revocation. To this end, we put forward conditional identity-based broadcast proxy re-encryption with anonymity and revocation (CIBPRE-AR) and construct a concrete CIBPRE-AR scheme. The scheme implements fine-grained data sharing by associating conditions with the re-encryption key. The identity privacy protection for authorized DUs is provided by using Lagrange interpolation. Further, the access privileges of the violated DU are revoked by updating the re-encryption key. The indistinguishability of ciphertexts against chosen-plaintext attack and anonymity of the DUs are proved rigidly. Compared with existing similar schemes, only the CIBPRE-AR scheme simultaneously achieves fine-grained data access control, anonymity as well as revocation. The proposed scheme also has advantage with respect to computation cost.","PeriodicalId":56305,"journal":{"name":"IEEE Transactions on Reliability","volume":"74 3","pages":"3573-3584"},"PeriodicalIF":5.7,"publicationDate":"2025-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144998358","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Engineered systems are increasingly integrating sensor techniques to trace their specific degradation behaviors, so as to facilitate their dynamic reliability assessment. Due to the hierarchical structure of these systems, sensing data can be collected at multiple physical levels, including the entire system, subsystems, and components. The quality of collected multilevel sensing data, however, decreases inevitably with the degradation of sensors mounted within each system, leading to a declining trustworthiness of dynamic reliability assessment for each specific individual system. This article develops a new dynamic reliability assessment framework of hierarchical multistate systems suffering from sensors’ degradation. The proposed framework mainly contains three steps: 1) utilizing discrete-state and continuous-state stochastic processes to, respectively, model the degradation behaviors of two types of sensors; 2) integrating these two types of sensors’ degradation models to update the joint state probability distribution of both the monitored objects and sensors by fusing multilevel sensing data; 3) deriving the marginal state probability distribution of the entire system to dynamically assess system reliability. A three-component system and an electromechanical actuator system in landing gear systems are exemplified to illustrate the performance of the proposed method.
{"title":"Dynamic Reliability Assessment of Hierarchical Multistate Systems With Sensors’ Degradation","authors":"Boyuan Zhang;Yu Liu;Yi-Xuan Zheng","doi":"10.1109/TR.2024.3524098","DOIUrl":"https://doi.org/10.1109/TR.2024.3524098","url":null,"abstract":"Engineered systems are increasingly integrating sensor techniques to trace their specific degradation behaviors, so as to facilitate their dynamic reliability assessment. Due to the hierarchical structure of these systems, sensing data can be collected at multiple physical levels, including the entire system, subsystems, and components. The quality of collected multilevel sensing data, however, decreases inevitably with the degradation of sensors mounted within each system, leading to a declining trustworthiness of dynamic reliability assessment for each specific individual system. This article develops a new dynamic reliability assessment framework of hierarchical multistate systems suffering from sensors’ degradation. The proposed framework mainly contains three steps: 1) utilizing discrete-state and continuous-state stochastic processes to, respectively, model the degradation behaviors of two types of sensors; 2) integrating these two types of sensors’ degradation models to update the joint state probability distribution of both the monitored objects and sensors by fusing multilevel sensing data; 3) deriving the marginal state probability distribution of the entire system to dynamically assess system reliability. A three-component system and an electromechanical actuator system in landing gear systems are exemplified to illustrate the performance of the proposed method.","PeriodicalId":56305,"journal":{"name":"IEEE Transactions on Reliability","volume":"74 3","pages":"3784-3798"},"PeriodicalIF":5.7,"publicationDate":"2025-01-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144996098","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Software defect prediction technology can discover potential errors or hidden defects by establishing prediction models before the use of products in the field of software engineering, so as to reduce subsequent problems and improve software quality and security. However, building predictive models requires enough software defect dataset support, especially defect samples. Due to the involvement of confidential information from various organizations or enterprises, software defect data cannot be shared and effectively utilized. Therefore, to achieve collaborative training of multiparty shared software defect prediction models while keeping the data local to various organizations, we made the federated learning framework for the issue of software defect prediction. Meanwhile, the nondefect and defect instances in software defect datasets are usually imbalanced, which can seriously affect the software defect prediction performance of the model. Therefore, this study designs a novel federated oversampling learning framework Fed-OLF. First, the TabDiT method based on deep generative model is proposed in Fed-OLF to expand and rebalance the local imbalanced software defect dataset of each client with a certain degree of privacy protection. Second, a parameter aggregation strategy based on local information entropy is proposed in Fed-OLF to further optimize the parameter aggregation effect of the global shared model, thereby achieving better model performance. We conduct extensive experiments on the PROMISE dataset and the NASA Promise repository, and experimental results on the PROMISE dataset and the NASA Promise repository show that, the proposed Fed-OLF exhibits better predictive performance under the F1-score, G-mean, and AUC metrics when compared with the advanced baseline methods. In addition, we verify that both the TabDiT method and the parameter aggregation strategy based on local information entropy in Fed-OLF are useful, and the combination of them can more effectively improve model performance.
{"title":"Fed-OLF: Federated Oversampling Learning Framework for Imbalanced Software Defect Prediction Under Privacy Protection","authors":"Xiaowen Hu;Ming Zheng;Rui Zhu;Xuan Zhang;Zhi Jin","doi":"10.1109/TR.2024.3524064","DOIUrl":"https://doi.org/10.1109/TR.2024.3524064","url":null,"abstract":"Software defect prediction technology can discover potential errors or hidden defects by establishing prediction models before the use of products in the field of software engineering, so as to reduce subsequent problems and improve software quality and security. However, building predictive models requires enough software defect dataset support, especially defect samples. Due to the involvement of confidential information from various organizations or enterprises, software defect data cannot be shared and effectively utilized. Therefore, to achieve collaborative training of multiparty shared software defect prediction models while keeping the data local to various organizations, we made the federated learning framework for the issue of software defect prediction. Meanwhile, the nondefect and defect instances in software defect datasets are usually imbalanced, which can seriously affect the software defect prediction performance of the model. Therefore, this study designs a novel federated oversampling learning framework Fed-OLF. First, the TabDiT method based on deep generative model is proposed in Fed-OLF to expand and rebalance the local imbalanced software defect dataset of each client with a certain degree of privacy protection. Second, a parameter aggregation strategy based on local information entropy is proposed in Fed-OLF to further optimize the parameter aggregation effect of the global shared model, thereby achieving better model performance. We conduct extensive experiments on the PROMISE dataset and the NASA Promise repository, and experimental results on the PROMISE dataset and the NASA Promise repository show that, the proposed Fed-OLF exhibits better predictive performance under the F1-score, G-mean, and AUC metrics when compared with the advanced baseline methods. In addition, we verify that both the TabDiT method and the parameter aggregation strategy based on local information entropy in Fed-OLF are useful, and the combination of them can more effectively improve model performance.","PeriodicalId":56305,"journal":{"name":"IEEE Transactions on Reliability","volume":"74 3","pages":"3266-3280"},"PeriodicalIF":5.7,"publicationDate":"2025-01-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144997978","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
During battery reliability tests, quality characteristic (QC) values like capacitance, voltage, or current are repeatedly observed during the cyclic charge-discharge processes. The battery's lifetime is determined by the first cycle where QC values drop below a specific threshold. Despite the recurrent nature of this cyclic data, performance declines with each charge-discharge cycle. The trend renewal process (TRP) transforms this periodic data through a trend function to ensure independent and stationary increments in the transformed data. However, combining the trend function with the renewal distribution complicates the resulting likelihood function. In typical battery reliability tests, sample sizes are small, and batteries exhibit heterogeneous differences. This article examines the inverse Gaussian accelerated trend-renewal process (ATRP) model for analyzing discharge-capacity battery data under various discharge currents, with model parameters being log-linear in discharge current. A hierarchical Bayesian approach is employed for three ATRP random-effects models, introducing latent variables to capture unit-to-unit variation among batteries. By selecting the most appropriate model based on the largest log marginal likelihood, predictive lifetime inference under normal discharging current is derived using the Markov chain Monte Carlo procedure. Monte-Carlo simulations validate the numerical calculations, and the proposed method is successfully applied to lithium-ion battery accelerated degradation test data.
{"title":"Bayesian Analysis of Accelerated Trend Renewal Processes With Application to Lithium-Ion Battery Data","authors":"Tsai-Hung Fan;Yi-Fu Wang;Chun-Kai Wu","doi":"10.1109/TR.2024.3523180","DOIUrl":"https://doi.org/10.1109/TR.2024.3523180","url":null,"abstract":"During battery reliability tests, quality characteristic (QC) values like capacitance, voltage, or current are repeatedly observed during the cyclic charge-discharge processes. The battery's lifetime is determined by the first cycle where QC values drop below a specific threshold. Despite the recurrent nature of this cyclic data, performance declines with each charge-discharge cycle. The trend renewal process (TRP) transforms this periodic data through a trend function to ensure independent and stationary increments in the transformed data. However, combining the trend function with the renewal distribution complicates the resulting likelihood function. In typical battery reliability tests, sample sizes are small, and batteries exhibit heterogeneous differences. This article examines the inverse Gaussian accelerated trend-renewal process (ATRP) model for analyzing discharge-capacity battery data under various discharge currents, with model parameters being log-linear in discharge current. A hierarchical Bayesian approach is employed for three ATRP random-effects models, introducing latent variables to capture unit-to-unit variation among batteries. By selecting the most appropriate model based on the largest log marginal likelihood, predictive lifetime inference under normal discharging current is derived using the Markov chain Monte Carlo procedure. Monte-Carlo simulations validate the numerical calculations, and the proposed method is successfully applied to lithium-ion battery accelerated degradation test data.","PeriodicalId":56305,"journal":{"name":"IEEE Transactions on Reliability","volume":"74 3","pages":"3083-3097"},"PeriodicalIF":5.7,"publicationDate":"2025-01-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144997993","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：