Pub Date : 2022-10-13DOI: 10.1109/PHM-Yantai55411.2022.9942123
Wang Luo, Chao Lou, Yuan Xia, De-Quan Gao, Ji-Wei Li, Ziyan Zhao, Fenggang Lai, Chao Ma
The Fault Diagnosis Knowledge Graph (FDKG) of Cloud Data Center (CDC), in a broad sense, is the knowledge Digital Twin of the fault phenomenon, reasoning, and maintenance process of Cloud Data Center in the physical world. The key to the Digital Twin is to establish the information interface between the physical space and the virtual space, and the key to the construction of the fault diagnosis KG is also here. FDKG of The State Grid Cloud Data Center needs to integrate multi-source knowledge to establish an information interface with the CDCs of subsidiaries in each province. However, in the process of updating the FDKG, the entity name attribute represented by long sentences reduces the accuracy of entity alignment, and it is difficult to efficiently integrate knowledge into the FDKG without increasing knowledge redundancy. This paper proposes an entity alignment method based on the fusion of attribute and relationship similarity, which will use the clearly defined relationship information in the FDKG to effectively improve the accuracy of entity alignment. The knowledge update tool developed based on this, effectively improves the entity alignment accuracy of the FDKG, and improves the information interface connection efficiency of the FDKG of the CDC.
{"title":"Research on Entity Update Technology for Fault Diagnosis Knowledge Graph of Cloud Data Center","authors":"Wang Luo, Chao Lou, Yuan Xia, De-Quan Gao, Ji-Wei Li, Ziyan Zhao, Fenggang Lai, Chao Ma","doi":"10.1109/PHM-Yantai55411.2022.9942123","DOIUrl":"https://doi.org/10.1109/PHM-Yantai55411.2022.9942123","url":null,"abstract":"The Fault Diagnosis Knowledge Graph (FDKG) of Cloud Data Center (CDC), in a broad sense, is the knowledge Digital Twin of the fault phenomenon, reasoning, and maintenance process of Cloud Data Center in the physical world. The key to the Digital Twin is to establish the information interface between the physical space and the virtual space, and the key to the construction of the fault diagnosis KG is also here. FDKG of The State Grid Cloud Data Center needs to integrate multi-source knowledge to establish an information interface with the CDCs of subsidiaries in each province. However, in the process of updating the FDKG, the entity name attribute represented by long sentences reduces the accuracy of entity alignment, and it is difficult to efficiently integrate knowledge into the FDKG without increasing knowledge redundancy. This paper proposes an entity alignment method based on the fusion of attribute and relationship similarity, which will use the clearly defined relationship information in the FDKG to effectively improve the accuracy of entity alignment. The knowledge update tool developed based on this, effectively improves the entity alignment accuracy of the FDKG, and improves the information interface connection efficiency of the FDKG of the CDC.","PeriodicalId":315994,"journal":{"name":"2022 Global Reliability and Prognostics and Health Management (PHM-Yantai)","volume":"223 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121441412","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-13DOI: 10.1109/phm-yantai55411.2022.9941867
Yang Wang, Guanghan Bai, Yun-An Zhang, J. Tao, Li Zhang
In the long-lasting war game, a continuous and stable power supply is of significant importance for giving full play to its superiority. Once the electric power system is attacked, the cascading effect will arise which caused faults to spread within the power network and between operation networks, and eventually lead to large-scale power outages. As a result, the operation loops will rupture and the operational system effectiveness will decline dramatically. In this paper, according to the characteristics of power network and combat network, a cascading failure model of heterogeneous interdependent network considering load characteristics is proposed from the perspective of complex network dependency, and a vulnerability assessment index for electric power system-operation networks(E-O) coupling networks is proposed based on the degree of network survival and operational system effectiveness. Finally, the cascading failure process and vulnerability of E-O coupling networks under different attack strategies are simulated and verified by a modified IEEE 39-bus power system. The results show that the proposed dependent network model of the E-O coupling system can reflect the characteristics and reveal failure rules of the E-O coupling system.
{"title":"Cascading failure analysis and robustness assessment of the operational system and electric power system based on dependent network","authors":"Yang Wang, Guanghan Bai, Yun-An Zhang, J. Tao, Li Zhang","doi":"10.1109/phm-yantai55411.2022.9941867","DOIUrl":"https://doi.org/10.1109/phm-yantai55411.2022.9941867","url":null,"abstract":"In the long-lasting war game, a continuous and stable power supply is of significant importance for giving full play to its superiority. Once the electric power system is attacked, the cascading effect will arise which caused faults to spread within the power network and between operation networks, and eventually lead to large-scale power outages. As a result, the operation loops will rupture and the operational system effectiveness will decline dramatically. In this paper, according to the characteristics of power network and combat network, a cascading failure model of heterogeneous interdependent network considering load characteristics is proposed from the perspective of complex network dependency, and a vulnerability assessment index for electric power system-operation networks(E-O) coupling networks is proposed based on the degree of network survival and operational system effectiveness. Finally, the cascading failure process and vulnerability of E-O coupling networks under different attack strategies are simulated and verified by a modified IEEE 39-bus power system. The results show that the proposed dependent network model of the E-O coupling system can reflect the characteristics and reveal failure rules of the E-O coupling system.","PeriodicalId":315994,"journal":{"name":"2022 Global Reliability and Prognostics and Health Management (PHM-Yantai)","volume":"06 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129075505","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-13DOI: 10.1109/PHM-Yantai55411.2022.9942101
Jian Nie, Qing Shi, Yuanyuan Dai
Based on the current situation of rail operation and maintenance of urban rail transit, combined with the technical system of fault diagnosis and health management (PHM), the operation and maintenance requirements of urban rail transit vehicles are investigated in depth. Due to the feasibility and superiority of PHM technology in rail operation and maintenance management, an intelligent operation and maintenance system for urban rail transit vehicles based on PHM technology is designed. Through PHM technology framework, the key hardware of the system is designed. Based on the hardware system, the vehicle information management module, operation day plan management module, data preprocessing module, health status identification module, operation and maintenance decision module and construction operation management module are established. The results show that the effectiveness of the designed system is proved through functional and performance tests, which provides solutions for the intelligent operation and maintenance of urban rail transit vehicles, and improves the safety and economy of operation and maintenance.
{"title":"Research on Intelligent Operation and Maintenance System of Urban Rail Transit Vehicles Based on PHM Technology","authors":"Jian Nie, Qing Shi, Yuanyuan Dai","doi":"10.1109/PHM-Yantai55411.2022.9942101","DOIUrl":"https://doi.org/10.1109/PHM-Yantai55411.2022.9942101","url":null,"abstract":"Based on the current situation of rail operation and maintenance of urban rail transit, combined with the technical system of fault diagnosis and health management (PHM), the operation and maintenance requirements of urban rail transit vehicles are investigated in depth. Due to the feasibility and superiority of PHM technology in rail operation and maintenance management, an intelligent operation and maintenance system for urban rail transit vehicles based on PHM technology is designed. Through PHM technology framework, the key hardware of the system is designed. Based on the hardware system, the vehicle information management module, operation day plan management module, data preprocessing module, health status identification module, operation and maintenance decision module and construction operation management module are established. The results show that the effectiveness of the designed system is proved through functional and performance tests, which provides solutions for the intelligent operation and maintenance of urban rail transit vehicles, and improves the safety and economy of operation and maintenance.","PeriodicalId":315994,"journal":{"name":"2022 Global Reliability and Prognostics and Health Management (PHM-Yantai)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129281168","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-13DOI: 10.1109/PHM-Yantai55411.2022.9941882
Cai Zhongyi, Wang Zezhou, Xia Keqiang, Lin Shaoliang, Yang Bingye
Aiming at the problem that aviation products generally have multiple performance degradation indicators, which lead to difficulty in extracting degradation features and inaccurate prediction of remaining useful lifetime (RUL), a method for building health index and predicting RUL of aviation products based on multi-source degradation data fusion is proposed. This method is based on the accuracy of degradation modeling to construct the construction principle of health index and realizes the deep fusion of multi-source degradation data. On this basis, the distribution expression of the RUL is derived based on the distribution of the first hitting time of the Wiener process. The effectiveness of the method is verified through the example analysis.
{"title":"Health index construction and remaining useful lifetime prediction of aviation products based on multi-source degradation data fusion","authors":"Cai Zhongyi, Wang Zezhou, Xia Keqiang, Lin Shaoliang, Yang Bingye","doi":"10.1109/PHM-Yantai55411.2022.9941882","DOIUrl":"https://doi.org/10.1109/PHM-Yantai55411.2022.9941882","url":null,"abstract":"Aiming at the problem that aviation products generally have multiple performance degradation indicators, which lead to difficulty in extracting degradation features and inaccurate prediction of remaining useful lifetime (RUL), a method for building health index and predicting RUL of aviation products based on multi-source degradation data fusion is proposed. This method is based on the accuracy of degradation modeling to construct the construction principle of health index and realizes the deep fusion of multi-source degradation data. On this basis, the distribution expression of the RUL is derived based on the distribution of the first hitting time of the Wiener process. The effectiveness of the method is verified through the example analysis.","PeriodicalId":315994,"journal":{"name":"2022 Global Reliability and Prognostics and Health Management (PHM-Yantai)","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128638843","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-13DOI: 10.1109/phm-yantai55411.2022.9941755
Jinyuan Li, Jixing Wang, Wenjun Ji, Yannan Yu, Chang-Bin Chen
Blockchain technology has unique advantages in quality management and traceability. It has had successful cases in food, medicine, e-commerce and other fields, but it is still less applied in the construction industry. According to the requirements of cscec-2020-z-12 funded by the science and technology R & D plan of CSCEC, this paper introduces blockchain technology into the construction industry and constructs a blockchain based construction project quality management system to help improve the quality of construction projects. Design the blockchain technical architecture, select the alliance chain to build the system, complete the setting of data layer, network layer, consensus layer, incentive layer and contract layer, and ensure the normal operation of the system. Then, four functional modules of the system are designed in the application layer, which are user management module, contract management module, quality management module and quality traceability module, and the business processes of quality management and quality traceability using the system are displayed. The experimental results show that the construction project quality and safety management system based on blockchain technology proposed in this paper has a shorter management time, a management range of more than 90%, and a strong management ability.
{"title":"Construction Project Quality and Safety Management System Based on Blockchain Technology","authors":"Jinyuan Li, Jixing Wang, Wenjun Ji, Yannan Yu, Chang-Bin Chen","doi":"10.1109/phm-yantai55411.2022.9941755","DOIUrl":"https://doi.org/10.1109/phm-yantai55411.2022.9941755","url":null,"abstract":"Blockchain technology has unique advantages in quality management and traceability. It has had successful cases in food, medicine, e-commerce and other fields, but it is still less applied in the construction industry. According to the requirements of cscec-2020-z-12 funded by the science and technology R & D plan of CSCEC, this paper introduces blockchain technology into the construction industry and constructs a blockchain based construction project quality management system to help improve the quality of construction projects. Design the blockchain technical architecture, select the alliance chain to build the system, complete the setting of data layer, network layer, consensus layer, incentive layer and contract layer, and ensure the normal operation of the system. Then, four functional modules of the system are designed in the application layer, which are user management module, contract management module, quality management module and quality traceability module, and the business processes of quality management and quality traceability using the system are displayed. The experimental results show that the construction project quality and safety management system based on blockchain technology proposed in this paper has a shorter management time, a management range of more than 90%, and a strong management ability.","PeriodicalId":315994,"journal":{"name":"2022 Global Reliability and Prognostics and Health Management (PHM-Yantai)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128588977","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-13DOI: 10.1109/PHM-Yantai55411.2022.9941774
Ran Fu, Lei Xiao, Baiteng Ma
Industrial robots have been widely used in various industrial manufacturing companies to improve production efficiency. With the service time gained of an industrial robot, the possibility of failure or fault of an industrial robot joint gains. Due to the motion propagation among the joints, some industrial robot arms show abnormal performance even though there is no fault in their joints. Although some vibration-based detection methods for industrial robot joint faults have been successfully established, it is still difficult to detect industrial robot joint faults by using only the current signal, especially, there is no sufficient label to classify fault or normal current signal. To deal with the above issues, this paper proposes an unsupervised fault detection method based on spectral clustering and the sensitive features of the current signal. To enlarge the samples, the collected current signal in a certain time is divided into several pieces according to the peak finding function. Then widely adopted time-domain features are selected according to the sensitivity. The selected features are fed into the spectral clustering to detect the fault location among the industrial robot joints. The proposed method is validated by a reliability-test industrial robot.
{"title":"Unsupervised Fault Detection of Industrial Robot Joints Using Current Signal","authors":"Ran Fu, Lei Xiao, Baiteng Ma","doi":"10.1109/PHM-Yantai55411.2022.9941774","DOIUrl":"https://doi.org/10.1109/PHM-Yantai55411.2022.9941774","url":null,"abstract":"Industrial robots have been widely used in various industrial manufacturing companies to improve production efficiency. With the service time gained of an industrial robot, the possibility of failure or fault of an industrial robot joint gains. Due to the motion propagation among the joints, some industrial robot arms show abnormal performance even though there is no fault in their joints. Although some vibration-based detection methods for industrial robot joint faults have been successfully established, it is still difficult to detect industrial robot joint faults by using only the current signal, especially, there is no sufficient label to classify fault or normal current signal. To deal with the above issues, this paper proposes an unsupervised fault detection method based on spectral clustering and the sensitive features of the current signal. To enlarge the samples, the collected current signal in a certain time is divided into several pieces according to the peak finding function. Then widely adopted time-domain features are selected according to the sensitivity. The selected features are fed into the spectral clustering to detect the fault location among the industrial robot joints. The proposed method is validated by a reliability-test industrial robot.","PeriodicalId":315994,"journal":{"name":"2022 Global Reliability and Prognostics and Health Management (PHM-Yantai)","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115778332","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-13DOI: 10.1109/PHM-Yantai55411.2022.9941857
Hengwei Zhao, Jiakai Ding, Dongming Xiao
Acoustic emission(AE) signals are generated during the SLM process, which contains much information about the spatter phenomenon. In this paper, an experimental platform from of SLM process is built. It is used acquisition the AE signals of the spatter phenomenon in the SLM process to realize the feature extraction of the spatter phenomenon. A method combining the improved Genetic Algorithm(GA) with the Variational Mode Decomposition(VMD) algorithm is proposed. First, The AE signals are analyzed in the time domain, frequency- domain, and time-frequency domain. Obtain the time-frequency feature of the AE signals of the spatter phenomenon. Then, the VMD algorithm is optimized by the improved GA, and the optimal parameter combination of the VMD algorithm is obtained. Finally, the feature extraction of AE signals of spatter phenomenon by optimized VMD algorithm. The results show that the feature frequency of the AE signals of the spatter phenomenon mainly ranges from 169.448KHz.
{"title":"Feature Extraction Of Acoustic Emission Signal Of Spatter Phenomenon In The SLM Process Based On Improved GA-VMD Algorithm","authors":"Hengwei Zhao, Jiakai Ding, Dongming Xiao","doi":"10.1109/PHM-Yantai55411.2022.9941857","DOIUrl":"https://doi.org/10.1109/PHM-Yantai55411.2022.9941857","url":null,"abstract":"Acoustic emission(AE) signals are generated during the SLM process, which contains much information about the spatter phenomenon. In this paper, an experimental platform from of SLM process is built. It is used acquisition the AE signals of the spatter phenomenon in the SLM process to realize the feature extraction of the spatter phenomenon. A method combining the improved Genetic Algorithm(GA) with the Variational Mode Decomposition(VMD) algorithm is proposed. First, The AE signals are analyzed in the time domain, frequency- domain, and time-frequency domain. Obtain the time-frequency feature of the AE signals of the spatter phenomenon. Then, the VMD algorithm is optimized by the improved GA, and the optimal parameter combination of the VMD algorithm is obtained. Finally, the feature extraction of AE signals of spatter phenomenon by optimized VMD algorithm. The results show that the feature frequency of the AE signals of the spatter phenomenon mainly ranges from 169.448KHz.","PeriodicalId":315994,"journal":{"name":"2022 Global Reliability and Prognostics and Health Management (PHM-Yantai)","volume":"273 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115287936","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-13DOI: 10.1109/phm-yantai55411.2022.9942158
Meng Li
According to the requirements of the environment adaptability of the equipment installation platform and from the point of reliability,a modular design method is adopted to design a stacked equipment structure. The finite element software is used to simulate and verify the design scheme, and according to the simulation results, the design scheme is iteratively optimized. The optimized scheme is effective and feasible, reducing the risk of equipment failure and improving the reliability of the equipment.
{"title":"Thermal Design and Simulation Optimization of an Airborne Equipment Structure","authors":"Meng Li","doi":"10.1109/phm-yantai55411.2022.9942158","DOIUrl":"https://doi.org/10.1109/phm-yantai55411.2022.9942158","url":null,"abstract":"According to the requirements of the environment adaptability of the equipment installation platform and from the point of reliability,a modular design method is adopted to design a stacked equipment structure. The finite element software is used to simulate and verify the design scheme, and according to the simulation results, the design scheme is iteratively optimized. The optimized scheme is effective and feasible, reducing the risk of equipment failure and improving the reliability of the equipment.","PeriodicalId":315994,"journal":{"name":"2022 Global Reliability and Prognostics and Health Management (PHM-Yantai)","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114905669","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-13DOI: 10.1109/PHM-Yantai55411.2022.9941966
Tianyu Gao, Ye Li, Xue Bai, Jingli Yang
With the rapid development of electronic technology, accurately identifying the incipient faults of analog circuits has become an important measure to improve the reliability and safety of electronic equipment. In recent years, deep learning is extensively applied to fault diagnosis because of its powerful feature mining ability. Therefore, a method based on spatio-temporal center network (STCN) is proposed to identify incipient faults for analog circuits, which includes a feature extraction module and a classification module. In the former, a spatio-temporal backbone network is designed to comprehensively mine the effective feature representation, including multi-scale spatial information and temporal information in the response signals of analog circuits. In the classification module, the spatio-temporal feature representation is imported into the Softmax layer for fault identification. Finally, in addition to the commonly used cross entropy loss, the central loss is also constructed for the STCN model. By reducing the intra class distance among similar feature representations, the discrimination of feature representation is further improved. In order to assess the effectiveness of the proposed method, the Sallen-key bandpass filter circuit is selected for experimental verification. Experimental results indicate that STCN is superior to some typical fault diagnosis approaches in incipient fault diagnosis of analog circuits.
{"title":"An Incipient Fault Diagnosis Method Based on Spatio-Temporal Center Network for Analog Circuits","authors":"Tianyu Gao, Ye Li, Xue Bai, Jingli Yang","doi":"10.1109/PHM-Yantai55411.2022.9941966","DOIUrl":"https://doi.org/10.1109/PHM-Yantai55411.2022.9941966","url":null,"abstract":"With the rapid development of electronic technology, accurately identifying the incipient faults of analog circuits has become an important measure to improve the reliability and safety of electronic equipment. In recent years, deep learning is extensively applied to fault diagnosis because of its powerful feature mining ability. Therefore, a method based on spatio-temporal center network (STCN) is proposed to identify incipient faults for analog circuits, which includes a feature extraction module and a classification module. In the former, a spatio-temporal backbone network is designed to comprehensively mine the effective feature representation, including multi-scale spatial information and temporal information in the response signals of analog circuits. In the classification module, the spatio-temporal feature representation is imported into the Softmax layer for fault identification. Finally, in addition to the commonly used cross entropy loss, the central loss is also constructed for the STCN model. By reducing the intra class distance among similar feature representations, the discrimination of feature representation is further improved. In order to assess the effectiveness of the proposed method, the Sallen-key bandpass filter circuit is selected for experimental verification. Experimental results indicate that STCN is superior to some typical fault diagnosis approaches in incipient fault diagnosis of analog circuits.","PeriodicalId":315994,"journal":{"name":"2022 Global Reliability and Prognostics and Health Management (PHM-Yantai)","volume":"43 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114704898","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-13DOI: 10.1109/PHM-Yantai55411.2022.9941970
Yifei Ding, M. Jia
Transfer learning (TL), especially domain adaptation (DA), has greatly enhanced the cross-domain fault diagnosis of rotating machines. However, the existing methods based on feature alignment at a single scale are still inadequate for complex cross-domain generalization, and thus have much room for improvement. Therefore, this work proposed a multi-scale domain adaptation network (MSDAN) to achieve representation alignment with multiple scales. By minimizing the uniquely designed combined mean maximum discrepancy (CoMMD) metrics, MSDAN is able to learn more domain-invariant representations on multi-scale branches. The case study that learns multi-scale domain adaptation (MSDN) with vibration signals of cross-domain bearings fully validates the feasibility of this method. Comparison with state-of-the-art methods also shows the necessity and advantages of simultaneous domain adaptation on multi-scale representations.
{"title":"Cross-Domain Fault Diagnosis for Rotating Machines with Multi-Scale Domain Adaptation","authors":"Yifei Ding, M. Jia","doi":"10.1109/PHM-Yantai55411.2022.9941970","DOIUrl":"https://doi.org/10.1109/PHM-Yantai55411.2022.9941970","url":null,"abstract":"Transfer learning (TL), especially domain adaptation (DA), has greatly enhanced the cross-domain fault diagnosis of rotating machines. However, the existing methods based on feature alignment at a single scale are still inadequate for complex cross-domain generalization, and thus have much room for improvement. Therefore, this work proposed a multi-scale domain adaptation network (MSDAN) to achieve representation alignment with multiple scales. By minimizing the uniquely designed combined mean maximum discrepancy (CoMMD) metrics, MSDAN is able to learn more domain-invariant representations on multi-scale branches. The case study that learns multi-scale domain adaptation (MSDN) with vibration signals of cross-domain bearings fully validates the feasibility of this method. Comparison with state-of-the-art methods also shows the necessity and advantages of simultaneous domain adaptation on multi-scale representations.","PeriodicalId":315994,"journal":{"name":"2022 Global Reliability and Prognostics and Health Management (PHM-Yantai)","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127240755","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","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：