Tianzi Tian;Ning Wang;Jun Yang;Zhuqing Miao;Lei Li
There has been increasing attention to the maintenance optimization of balanced systems in recent years. However, existing studies mostly neglect state-dependent inspection intervals and group maintenance, which inadequately addresses the maintenance challenges of balanced systems. Thus, we propose an availability evaluation and maintenance optimization method for balanced systems considering state-dependent inspection intervals. First, multiple maintenance thresholds are introduced to characterize the maintenance strategy considering preventive maintenance and state-dependent inspection intervals, where the next inspection interval is determined based on the post-maintenance system state. Then, the system availability is evaluated by combining semi-regenerative theory and universal generating functions, where calculations are simplified by merging the same system states. Meanwhile, this study also explores the system availability under group maintenance to better reflect reality. Second, the average maintenance cost per unit of time is calculated using the renewal theory. The optimal maintenance thresholds are given by minimizing the maintenance cost under the constraint of minimum system availability. To improve the optimization efficiency, a tabu list-based two-stage iterative partial optimization algorithm is proposed. Finally, the effectiveness of the proposed method is demonstrated through a numerical example involving a lithium-ion battery pack.
{"title":"Availability Evaluation and Maintenance Optimization of Balanced Systems Considering State-Dependent Inspection Intervals","authors":"Tianzi Tian;Ning Wang;Jun Yang;Zhuqing Miao;Lei Li","doi":"10.1109/TR.2024.3394862","DOIUrl":"10.1109/TR.2024.3394862","url":null,"abstract":"There has been increasing attention to the maintenance optimization of balanced systems in recent years. However, existing studies mostly neglect state-dependent inspection intervals and group maintenance, which inadequately addresses the maintenance challenges of balanced systems. Thus, we propose an availability evaluation and maintenance optimization method for balanced systems considering state-dependent inspection intervals. First, multiple maintenance thresholds are introduced to characterize the maintenance strategy considering preventive maintenance and state-dependent inspection intervals, where the next inspection interval is determined based on the post-maintenance system state. Then, the system availability is evaluated by combining semi-regenerative theory and universal generating functions, where calculations are simplified by merging the same system states. Meanwhile, this study also explores the system availability under group maintenance to better reflect reality. Second, the average maintenance cost per unit of time is calculated using the renewal theory. The optimal maintenance thresholds are given by minimizing the maintenance cost under the constraint of minimum system availability. To improve the optimization efficiency, a tabu list-based two-stage iterative partial optimization algorithm is proposed. Finally, the effectiveness of the proposed method is demonstrated through a numerical example involving a lithium-ion battery pack.","PeriodicalId":56305,"journal":{"name":"IEEE Transactions on Reliability","volume":"74 1","pages":"2241-2254"},"PeriodicalIF":5.0,"publicationDate":"2024-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141942436","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}
Monitoring the health of complex industrial assets is crucial for safe and efficient operations. Health indicators that provide quantitative real-time insights into the health status of industrial assets over time serve as valuable tools for, e.g., fault detection or prognostics. This article proposes a novel, versatile, and unsupervised approach to learn health indicators using contrastive learning, where the operational time serves as a proxy for degradation. To highlight its versatility, the approach is evaluated on two tasks and case studies with different characteristics: wear assessment of milling machines and fault detection of railway wheels. Our results show that the proposed methodology effectively learns a health indicator that follows the wear of milling machines (0.97 correlation on average) and is suitable for fault detection in railway wheels ($mathbf {88.7%}$ balanced accuracy). The conducted experiments demonstrate the versatility of the approach for various systems and health conditions.
{"title":"Learning Informative Health Indicators Through Unsupervised Contrastive Learning","authors":"Katharina Rombach;Gabriel Michau;Wilfried Bürzle;Stefan Koller;Olga Fink","doi":"10.1109/TR.2024.3397394","DOIUrl":"10.1109/TR.2024.3397394","url":null,"abstract":"Monitoring the health of complex industrial assets is crucial for safe and efficient operations. Health indicators that provide quantitative real-time insights into the health status of industrial assets over time serve as valuable tools for, e.g., fault detection or prognostics. This article proposes a novel, versatile, and unsupervised approach to learn health indicators using contrastive learning, where the <italic>operational time</i> serves as a proxy for degradation. To highlight its versatility, the approach is evaluated on two tasks and case studies with different characteristics: wear assessment of milling machines and fault detection of railway wheels. Our results show that the proposed methodology effectively learns a health indicator that follows the wear of milling machines (0.97 correlation on average) and is suitable for fault detection in railway wheels (<inline-formula><tex-math>$mathbf {88.7%}$</tex-math></inline-formula> balanced accuracy). The conducted experiments demonstrate the versatility of the approach for various systems and health conditions.","PeriodicalId":56305,"journal":{"name":"IEEE Transactions on Reliability","volume":"74 1","pages":"2408-2420"},"PeriodicalIF":5.0,"publicationDate":"2024-03-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141061802","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}
Dequan Zhang;Zhijie Hao;Yunfei Liang;Fang Wang;Weipeng Liu;Xu Han
With the ever-increasing complexity and scale of advanced modern engineering systems, multifailure modes coupling and input parameter correlations become important and inevitable challenges that hinder efficient reliability analysis of complex mechanical systems. To tackle this problem, in this article, a system reliability analysis method based on evidence theory considering parameter correlations is proposed. First, the optimal Copula function is selected by the Akaike information criterion using existing samples and the joint basic probability assignment considering parameter correlations is calculated. Second, engineering systems with multifailure modes are divided into series systems or parallel systems. The corresponding belief and plausibility measures of system reliability are derived, respectively. Moreover, support vector regression models are constructed by Latin hypercube sampling and genetic algorithm to replace the real performance functions. Therefore, the probability interval consisting of belief and plausibility measures is obtained through fewer performance function calls. Finally, two numerical examples and an engineering application of a 6-DoF industrial robot are exemplified to verify the effectiveness of the currently proposed method.
{"title":"An Efficient System Reliability Analysis Method Based on Evidence Theory With Parameter Correlations","authors":"Dequan Zhang;Zhijie Hao;Yunfei Liang;Fang Wang;Weipeng Liu;Xu Han","doi":"10.1109/TR.2024.3391252","DOIUrl":"10.1109/TR.2024.3391252","url":null,"abstract":"With the ever-increasing complexity and scale of advanced modern engineering systems, multifailure modes coupling and input parameter correlations become important and inevitable challenges that hinder efficient reliability analysis of complex mechanical systems. To tackle this problem, in this article, a system reliability analysis method based on evidence theory considering parameter correlations is proposed. First, the optimal Copula function is selected by the Akaike information criterion using existing samples and the joint basic probability assignment considering parameter correlations is calculated. Second, engineering systems with multifailure modes are divided into series systems or parallel systems. The corresponding belief and plausibility measures of system reliability are derived, respectively. Moreover, support vector regression models are constructed by Latin hypercube sampling and genetic algorithm to replace the real performance functions. Therefore, the probability interval consisting of belief and plausibility measures is obtained through fewer performance function calls. Finally, two numerical examples and an engineering application of a 6-DoF industrial robot are exemplified to verify the effectiveness of the currently proposed method.","PeriodicalId":56305,"journal":{"name":"IEEE Transactions on Reliability","volume":"74 1","pages":"2200-2213"},"PeriodicalIF":5.0,"publicationDate":"2024-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141061801","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 aids the reliability assurance team in identifying defect-prone code and assists the team in reasonably allocating limited testing resources. Recently, researchers assumed that the topological associations among code fragments could be harnessed to construct defect prediction models. Nevertheless, existing graph-based methods only concentrate on features of single-view association, which fail to fully capture the rich information hidden in the code. In addition, software defects may involve multiple code fragments simultaneously, but traditional binary graph structures are insufficient for representing these multivariate associations. To address these two challenges, this article proposes a multiview hypergraph representation learning approach (MVHR-DP) to amplify the potency of code features in defect prediction. MVHR-DP initiates by creating hypergraph structures for each code view, which are then amalgamated into a comprehensive fusion hypergraph. Following this, a hypergraph neural network is established to extract code features from multiple views and intricate associations, thereby enhancing the comprehensiveness of representation in the modeling data. Empirical study shows that the prediction model utilizing features generated by MVHR-DP exhibits superior area under the curve (AUC), F-measure, and matthews correlation coefficient (MCC) results compared to baseline approaches across within-project, cross-version, and cross-project prediction tasks.
{"title":"Code Multiview Hypergraph Representation Learning for Software Defect Prediction","authors":"Shaojian Qiu;Mengyang Huang;Yun Liang;Chaoda Peng;Yuan Yuan","doi":"10.1109/TR.2024.3393415","DOIUrl":"10.1109/TR.2024.3393415","url":null,"abstract":"Software defect prediction technology aids the reliability assurance team in identifying defect-prone code and assists the team in reasonably allocating limited testing resources. Recently, researchers assumed that the topological associations among code fragments could be harnessed to construct defect prediction models. Nevertheless, existing graph-based methods only concentrate on features of single-view association, which fail to fully capture the rich information hidden in the code. In addition, software defects may involve multiple code fragments simultaneously, but traditional binary graph structures are insufficient for representing these multivariate associations. To address these two challenges, this article proposes a multiview hypergraph representation learning approach (MVHR-DP) to amplify the potency of code features in defect prediction. MVHR-DP initiates by creating hypergraph structures for each code view, which are then amalgamated into a comprehensive fusion hypergraph. Following this, a hypergraph neural network is established to extract code features from multiple views and intricate associations, thereby enhancing the comprehensiveness of representation in the modeling data. Empirical study shows that the prediction model utilizing features generated by MVHR-DP exhibits superior area under the curve (AUC), F-measure, and matthews correlation coefficient (MCC) results compared to baseline approaches across within-project, cross-version, and cross-project prediction tasks.","PeriodicalId":56305,"journal":{"name":"IEEE Transactions on Reliability","volume":"73 4","pages":"1863-1876"},"PeriodicalIF":5.0,"publicationDate":"2024-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141061804","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}
Code summarization is a process of creating a readable natural language from programming source codes. Code summarization has become a popular research topic for software maintenance, code generation, and code recovery. Existing code summarization methods follow the encoding/decoding approach and use various machine learning techniques to generate natural language from source codes. Although most of these methods are state of the art, it is difficult to understand the complex encoding and decoding process to map the tokens with natural language words. Therefore, these coding and decoding approaches are treated as opaque models (black box). This research proposes explainable AI methods that overcome the black box features for the token mapping in code summarization process. Here, we created an abstract syntax tree (AST) from the tokens of the source code. We then embedded the AST into natural language words using a bilingual statistical probability approach to generate possible statistical parse trees. We applied a page rank algorithm among the parse trees to rank the trees. From the best-ranked tree, we generate the comment for the corresponding code snippet. To explain our code generation method, we used Takagi–Sugeno fuzzy approach, layerwise relevance propagation and a hidden Markov model. These approaches make our method trustworthy and understandable to humans to understand the process of source code token mapping with natural language words.
{"title":"Interpretable Code Summarization","authors":"Md Sarwar Kamal;Sonia Farhana Nimmy;Nilanjan Dey","doi":"10.1109/TR.2024.3392876","DOIUrl":"10.1109/TR.2024.3392876","url":null,"abstract":"Code summarization is a process of creating a readable natural language from programming source codes. Code summarization has become a popular research topic for software maintenance, code generation, and code recovery. Existing code summarization methods follow the encoding/decoding approach and use various machine learning techniques to generate natural language from source codes. Although most of these methods are state of the art, it is difficult to understand the complex encoding and decoding process to map the tokens with natural language words. Therefore, these coding and decoding approaches are treated as opaque models (black box). This research proposes explainable AI methods that overcome the black box features for the token mapping in code summarization process. Here, we created an abstract syntax tree (AST) from the tokens of the source code. We then embedded the AST into natural language words using a bilingual statistical probability approach to generate possible statistical parse trees. We applied a page rank algorithm among the parse trees to rank the trees. From the best-ranked tree, we generate the comment for the corresponding code snippet. To explain our code generation method, we used Takagi–Sugeno fuzzy approach, layerwise relevance propagation and a hidden Markov model. These approaches make our method trustworthy and understandable to humans to understand the process of source code token mapping with natural language words.","PeriodicalId":56305,"journal":{"name":"IEEE Transactions on Reliability","volume":"74 1","pages":"2280-2289"},"PeriodicalIF":5.0,"publicationDate":"2024-03-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141064018","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 system reliability assessment, expert opinions are oftentimes elicited to cope with the issue of poor quantity of failure data. However, information from expert subjective judgments may exhibit imprecision and be elicited from multiple physical levels of a system. Moreover, the elicited information may be conflicting as experts own varying backgrounds of knowledge as well as differentiated cognitive levels, which, thereby, cannot reach a consistent reliability estimate. In this article, we aim at conducting the reliability assessment of multistate systems by fusing conflicting multisource imprecise information (MSII) from the consensus reaching perspective. We utilize an aggregation operator to fuse individual opinions into a collective opinion in the form of mass functions. Evidence distance is, then, adopted to quantify the dissimilarity between individual and collective opinions, and accordingly, a measurement of the average degree of consensus is proposed. In this way, the consensus reaching model is formulated by minimizing the total calibration of MSII with the constraint of a predetermined consensus threshold. The consensus reaching model is resolved by a feasibility-based particle swarm optimization algorithm. A numerical example, along with an application of control rod drive mechanism in nuclear reactors, is used for the demonstration of the effectiveness of the proposed method.
{"title":"Multisource Imprecise Information Calibration for Reliability Assessment of Multistate Systems: A Consensus Reaching Perspective","authors":"Ruijie Liu;Tangfan Xiahou;Yu Liu","doi":"10.1109/TR.2024.3393985","DOIUrl":"10.1109/TR.2024.3393985","url":null,"abstract":"In system reliability assessment, expert opinions are oftentimes elicited to cope with the issue of poor quantity of failure data. However, information from expert subjective judgments may exhibit imprecision and be elicited from multiple physical levels of a system. Moreover, the elicited information may be conflicting as experts own varying backgrounds of knowledge as well as differentiated cognitive levels, which, thereby, cannot reach a consistent reliability estimate. In this article, we aim at conducting the reliability assessment of multistate systems by fusing conflicting multisource imprecise information (MSII) from the consensus reaching perspective. We utilize an aggregation operator to fuse individual opinions into a collective opinion in the form of mass functions. Evidence distance is, then, adopted to quantify the dissimilarity between individual and collective opinions, and accordingly, a measurement of the average degree of consensus is proposed. In this way, the consensus reaching model is formulated by minimizing the total calibration of MSII with the constraint of a predetermined consensus threshold. The consensus reaching model is resolved by a feasibility-based particle swarm optimization algorithm. A numerical example, along with an application of control rod drive mechanism in nuclear reactors, is used for the demonstration of the effectiveness of the proposed method.","PeriodicalId":56305,"journal":{"name":"IEEE Transactions on Reliability","volume":"74 1","pages":"2226-2240"},"PeriodicalIF":5.0,"publicationDate":"2024-03-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141061840","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}
Due to the strong uncertainty in the retrieval order of containers and the complex coupling relationship between container yard production equipment, it is challenging for yards to formulate an appropriate slot allocation strategy to control the proportion of rehandling containers. Meanwhile, the unpredictable performance of the slot allocation strategy results in the yard lacking the means to adjust the allocation strategy. To address these issues, an efficient container rehandling probability prediction model based on deep learning has been proposed to assist yards in formulating and adjusting slot allocation strategy. Moreover, we design a container slot allocation strategy driven by the predictive container rehandling probability for reducing the proportion of rehandling container at yards. Extensive experiments on the container storage dataset demonstrate that: 1) the prediction model based on deep learning enables to efficiently and precisely predict the container rehandling, 2) taking Seq2Seq network as the prediction layer of model outperforms other deep sequence models on MSE, MAE, and accuracy, and 3) the slot allocation strategy based on the predictive container rehandling probability can effectively reduce the probability of the rehandling container.
{"title":"Container Rehandling Probability Prediction Model Based on Seq2Seq Network","authors":"Guojie Chen;Weidong Zhao;Xianhui Liu;Mingyue Wei","doi":"10.1109/TR.2024.3392919","DOIUrl":"10.1109/TR.2024.3392919","url":null,"abstract":"Due to the strong uncertainty in the retrieval order of containers and the complex coupling relationship between container yard production equipment, it is challenging for yards to formulate an appropriate slot allocation strategy to control the proportion of rehandling containers. Meanwhile, the unpredictable performance of the slot allocation strategy results in the yard lacking the means to adjust the allocation strategy. To address these issues, an efficient container rehandling probability prediction model based on deep learning has been proposed to assist yards in formulating and adjusting slot allocation strategy. Moreover, we design a container slot allocation strategy driven by the predictive container rehandling probability for reducing the proportion of rehandling container at yards. Extensive experiments on the container storage dataset demonstrate that: 1) the prediction model based on deep learning enables to efficiently and precisely predict the container rehandling, 2) taking Seq2Seq network as the prediction layer of model outperforms other deep sequence models on MSE, MAE, and accuracy, and 3) the slot allocation strategy based on the predictive container rehandling probability can effectively reduce the probability of the rehandling container.","PeriodicalId":56305,"journal":{"name":"IEEE Transactions on Reliability","volume":"73 3","pages":"1569-1580"},"PeriodicalIF":5.0,"publicationDate":"2024-03-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141061914","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}
As the core of cloud computing, the data center network (DCN) provides services and decision support for smart cities by providing powerful data storage and computing capabilities. As a server-centric DCN, the high scalability data center network architecture (HSDC) can cope with the rapid growth of data volume and provide an effective service foundation for smart cities. However, the rapid development of smart cities requires that DCNs can still operate reliably in the presence of faulty while expanding its scale. Therefore, it is very important to design reliable fault-tolerant communication algorithms in DCNs. This article investigates the fault-tolerant communication algorithm in HSDC. Initially, we propose an $O(n^{2})$ algorithm to establish $n$ disjoint paths between any pair of nodes in the $n$ -dimension HSDC. Simulation experiments show that the disjoint paths generated by our algorithm in HSDC have a maximum length only 3 longer than the diameter, which guarantees a small communication delay in the worst case. In addition, we propose an $O(n)$ algorithm to establish a fault-tolerant unicast path between any pair of fault-free nodes in the $n$ -dimension HSDC. Moreover, simulation experiments indicate that as the scale of HSDC increases, the algorithm performs well in both running efficiency and the length of constructed paths.
{"title":"Fault-Tolerant Communication in HSDC: Ensuring Reliable Data Transmission in Smart Cities","authors":"Hui Dong;Mengjie Lv;Weibei Fan","doi":"10.1109/TR.2024.3371953","DOIUrl":"10.1109/TR.2024.3371953","url":null,"abstract":"As the core of cloud computing, the data center network (DCN) provides services and decision support for smart cities by providing powerful data storage and computing capabilities. As a server-centric DCN, the high scalability data center network architecture (HSDC) can cope with the rapid growth of data volume and provide an effective service foundation for smart cities. However, the rapid development of smart cities requires that DCNs can still operate reliably in the presence of faulty while expanding its scale. Therefore, it is very important to design reliable fault-tolerant communication algorithms in DCNs. This article investigates the fault-tolerant communication algorithm in HSDC. Initially, we propose an \u0000<inline-formula><tex-math>$O(n^{2})$</tex-math></inline-formula>\u0000 algorithm to establish \u0000<inline-formula><tex-math>$n$</tex-math></inline-formula>\u0000 disjoint paths between any pair of nodes in the \u0000<inline-formula><tex-math>$n$</tex-math></inline-formula>\u0000-dimension HSDC. Simulation experiments show that the disjoint paths generated by our algorithm in HSDC have a maximum length only 3 longer than the diameter, which guarantees a small communication delay in the worst case. In addition, we propose an \u0000<inline-formula><tex-math>$O(n)$</tex-math></inline-formula>\u0000 algorithm to establish a fault-tolerant unicast path between any pair of fault-free nodes in the \u0000<inline-formula><tex-math>$n$</tex-math></inline-formula>\u0000-dimension HSDC. Moreover, simulation experiments indicate that as the scale of HSDC increases, the algorithm performs well in both running efficiency and the length of constructed paths.","PeriodicalId":56305,"journal":{"name":"IEEE Transactions on Reliability","volume":"73 4","pages":"1933-1945"},"PeriodicalIF":5.0,"publicationDate":"2024-03-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140155424","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}
This article introduces a data-driven approach to assessing failure mechanisms and reliability degradation in outdoor photovoltaic (PV) string inverters. The manufacturer's stated PV inverter lifetime can vary due to the impact of operating site conditions. To address limitations in degradation estimation through accelerated testing, condition monitoring, or degradation modeling, we propose a machine learning (ML) oriented approach. Utilizing data from a 1.4 MW PV power plant operational since 2016, with 46 string PV inverters tied to the grid, we employ the unsupervised one-class support vector machine ML technique to analyze inverter and sensor data, capable of classifying humidity cycling and temperature fluctuations as dominant failure mechanisms. Utilizing the anomaly alert relationship and alert details specific to the inverter, the level of PV inverter output is considered as its availability or available reliability. Subsequently, a continuous Markov model is applied to six-month alert data, revealing an average stated reliability of 20% after 20 years of continuous operation. These results support recommendations for time-bound preventive measures to enhance PV inverter reliability under diverse outdoor conditions. The approach provides a nondestructive, top–down, and generalized method for analyzing any commercial PV inverter exposed to outdoor conditions, contingent on the availability of relevant data.
本文介绍了一种数据驱动方法，用于评估户外光伏（PV）组串逆变器的故障机制和可靠性退化。制造商标明的光伏逆变器使用寿命会因运行地点条件的影响而变化。为了解决加速测试、状态监测或退化建模在退化估计方面的局限性，我们提出了一种以机器学习（ML）为导向的方法。利用自 2016 年起运营的 1.4 兆瓦光伏电站的数据，我们采用无监督单类支持向量机 ML 技术来分析逆变器和传感器数据，能够将湿度循环和温度波动划分为主要故障机制。利用逆变器特有的异常警报关系和警报细节，光伏逆变器的输出水平被视为其可用性或可用可靠性。随后，一个连续的马尔可夫模型被应用到六个月的警报数据中，结果显示，在连续运行 20 年后，所述平均可靠性为 20%。这些结果支持对有时限的预防措施提出建议，以提高光伏逆变器在各种户外条件下的可靠性。该方法提供了一种无损、自上而下和通用的方法，可用于分析任何暴露在室外条件下的商用光伏逆变器，但这取决于相关数据的可用性。
{"title":"Photovoltaic Inverter Failure Mechanism Estimation Using Unsupervised Machine Learning and Reliability Assessment","authors":"Sukanta Roy;Shahid Tufail;Mohd Tariq;Arif Sarwat","doi":"10.1109/TR.2024.3359540","DOIUrl":"10.1109/TR.2024.3359540","url":null,"abstract":"This article introduces a data-driven approach to assessing failure mechanisms and reliability degradation in outdoor photovoltaic (PV) string inverters. The manufacturer's stated PV inverter lifetime can vary due to the impact of operating site conditions. To address limitations in degradation estimation through accelerated testing, condition monitoring, or degradation modeling, we propose a machine learning (ML) oriented approach. Utilizing data from a 1.4 MW PV power plant operational since 2016, with 46 string PV inverters tied to the grid, we employ the unsupervised one-class support vector machine ML technique to analyze inverter and sensor data, capable of classifying humidity cycling and temperature fluctuations as dominant failure mechanisms. Utilizing the anomaly alert relationship and alert details specific to the inverter, the level of PV inverter output is considered as its availability or available reliability. Subsequently, a continuous Markov model is applied to six-month alert data, revealing an average stated reliability of 20% after 20 years of continuous operation. These results support recommendations for time-bound preventive measures to enhance PV inverter reliability under diverse outdoor conditions. The approach provides a nondestructive, top–down, and generalized method for analyzing any commercial PV inverter exposed to outdoor conditions, contingent on the availability of relevant data.","PeriodicalId":56305,"journal":{"name":"IEEE Transactions on Reliability","volume":"73 3","pages":"1418-1432"},"PeriodicalIF":5.0,"publicationDate":"2024-03-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140155423","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}
Mustafa Tosun;Umut Can Cabuk;Elif Haytaoglu;Orhan Dagdeviren;Yusuf Ozturk
Maintenance of connectivity in mobile ad hoc networks (MANETs) and especially in flying ad hoc networks, consisting of unmanned aerial vehicles (UAV), has crucial importance. Missions planned within these types of networks can be interrupted due to node failures, link errors, etc. This case becomes more critical when the application is heavily communication-dependent. To alleviate such problems, in this article, we propose a distributed proactive $k$ -connectivity recovery algorithm (DP $k$ CR) for UAV-based MANETs. To this end, the algorithm has been developed and tested by providing realistic scenarios. The time and message complexity analysis of the algorithm is presented. Moreover, to analyze the performance of the proposed algorithm, we compared it with other $k$ -connectivity restoration algorithms in the literature. Simulation results revealed that DP $k$ CR outperforms the alternatives in terms of convergence time for the recovery phase and, subsequently, in terms of energy consumption. Furthermore, DP $k$ CR provides improvements to the bandwidth requirements for the restoration.
{"title":"DPkCR: Distributed Proactive k-Connectivity Recovery Algorithm for UAV-Based MANETs","authors":"Mustafa Tosun;Umut Can Cabuk;Elif Haytaoglu;Orhan Dagdeviren;Yusuf Ozturk","doi":"10.1109/TR.2024.3370743","DOIUrl":"10.1109/TR.2024.3370743","url":null,"abstract":"Maintenance of connectivity in mobile ad hoc networks (MANETs) and especially in flying ad hoc networks, consisting of unmanned aerial vehicles (UAV), has crucial importance. Missions planned within these types of networks can be interrupted due to node failures, link errors, etc. This case becomes more critical when the application is heavily communication-dependent. To alleviate such problems, in this article, we propose a distributed proactive \u0000<inline-formula><tex-math>$k$</tex-math></inline-formula>\u0000-connectivity recovery algorithm (DP\u0000<inline-formula><tex-math>$k$</tex-math></inline-formula>\u0000CR) for UAV-based MANETs. To this end, the algorithm has been developed and tested by providing realistic scenarios. The time and message complexity analysis of the algorithm is presented. Moreover, to analyze the performance of the proposed algorithm, we compared it with other \u0000<inline-formula><tex-math>$k$</tex-math></inline-formula>\u0000-connectivity restoration algorithms in the literature. Simulation results revealed that DP\u0000<inline-formula><tex-math>$k$</tex-math></inline-formula>\u0000CR outperforms the alternatives in terms of convergence time for the recovery phase and, subsequently, in terms of energy consumption. Furthermore, DP\u0000<inline-formula><tex-math>$k$</tex-math></inline-formula>\u0000CR provides improvements to the bandwidth requirements for the restoration.","PeriodicalId":56305,"journal":{"name":"IEEE Transactions on Reliability","volume":"73 4","pages":"1918-1932"},"PeriodicalIF":5.0,"publicationDate":"2024-03-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140128736","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：