7 results on '"fault correlation"'
Search Results
2. A hazard analysis of heavy CNC machine tools based on fusion fault correlation.
- Author
-
Chen, Hongxia, Zhang, Junfeng, Guo, Chuncheng, Li, Hongyue, and Li, Chenguang
- Subjects
- *
NUMERICAL control of machine tools , *MACHINE tools , *HAZARDS , *MACHINE parts , *DECISION making , *WEIGHING instruments , *BEVERAGES - Abstract
This paper presents an analysis of the fault correlation between CNC machine tool subsystems using decision laboratory analysis and the House of Reliability (HoR) methodology. Initially, the maintenance information of the machine tool and other hazard indicators is comprehensively analyzed based on the cognitive best worst method (CBWM) to assign weights of hazard indicators. The HoR method is then employed to perform fault correlation fusion, which results in comprehensive hazard analysis results of the machine tool subsystems considering the fault correlation. Finally, the results obtained from the fusion fault correlation and the traditional hazard analysis are compared. The results indicate that the proposed method can highlight the relatively important subsystems and the most vulnerable parts of the machine tools. Overall, the hazard analysis results of this paper contribute to the safe and reliable operational and maintenance optimization of CNC machine tools. [ABSTRACT FROM AUTHOR]
- Published
- 2023
- Full Text
- View/download PDF
3. Research on data mining model of fault operation and maintenance based on electric vehicle charging behavior
- Author
-
Bin Zhu, Xiaorui Hu, Min He, Lei Chi, and Tingting Xu
- Subjects
association rules ,charging equipment ,renewable energy ,data mining ,electric car ,fault correlation ,General Works - Abstract
In recent years, with the development of new energy technology and the country’s strong support for electric vehicles, there is a lack of effective electric vehicle charging fault analysis and diagnosis methods at this stage. A comprehensive analysis of the working principle of the charging process of electric vehicles, based on the clarification of the failure mechanism of the power battery and charging equipment, analyzes the fault-related factors affecting the power battery and charging equipment from multiple angles, and summarizes the relationship between the power battery and charging equipment. The feature parameters related to equipment failure are discretized by the k-means clustering algorithm. Using the optimized FP-Growth algorithm based on weights, the association rules between the power battery and charging equipment failures and the characteristic parameters of the failure factors are mined, and the correlation of the failures is analyzed based on the association rules, and the correlation between the failure factors and the failures is obtained relevant level.
- Published
- 2023
- Full Text
- View/download PDF
4. Forecasting Method of Consumption Spare Parts of Mutual Support System Based on Stochastic Process.
- Author
-
Shenyang, Liu, Zhijie, Huang, Qian, Zhu, and Chen, Zhu
- Subjects
AEROSPACE industries ,SPARE parts ,DEBUGGING ,COMMUNICATION ,SUPPORT groups ,STOCHASTIC analysis - Abstract
Considering the characteristics of mutual support system in which unit fault leads to the failure rate mutation, through the analysis of practical problems, using the theory of stochastic process, the paper gives the procedure and method of consumption forecast model of spare parts to establish the relay board mutual support system of a communication device. The applicability of the model is illustrated by examples. [ABSTRACT FROM AUTHOR]
- Published
- 2017
- Full Text
- View/download PDF
5. Machine Learning applied to fault correlation
- Author
-
Lima, Pedro Jorge Rito, Mendes, Rui, Araújo, Carlos Guilherme, and Universidade do Minho
- Subjects
Machine Learning ,Aprendizagem máquina ,Fault correlation ,Gestão de alarmes ,Automatização de regras ,Artificial Intelligence ,Alarm manager ,Root cause analysis ,Rules automation ,Inteligência Artificial ,Topologias de rede ,Correlação entre falhas ,Network topologies - Abstract
Dissertação de mestrado em Engenharia Informática, Over the last years, one of the areas that have most evolved and extended its application to a multitude of possi bilities is Artificial Intelligence (AI). With the increasing complexity of the problems to be solved, human resolution becomes impossible, as the amount of information and patterns that can be detected is limited, while AI thrives on the dimension of the problem under analysis. Furthermore, as nowadays more and more traditional devices are computerized, an increasing number of elements are producing data that has many potential applications. Consequently, we find ourselves at the height of Big Data, where huge volumes of data are generated, being entirely unfeasible to process and analyze them manually. Additionally, with the increasing complexity of network topologies, it is necessary to ensure the correct func tioning of all equipment, avoiding cascade failures among devices, which can lead to catastrophic consequences depending on their use. Thus, Root Cause Analysis (RCA) tools become fundamental since these are developed to automatically, through rules established by its users, realize the underlying causes when some equipment mal functions. However, with the growing network complexity, the definition of rules becomes exponentially more complicated as the possible points of failure scale drastically. In this context, framed by the Altice Labs RCA and network environment use case, the main objective of this research project is defined. The aim is to use Machine Learning (ML) techniques to extrapolate the relationship between different types of equipment alarms, gathered by the Alarm Manager tool, to have a better understanding of the impact of a failure on the entire system, thus easing and helping the process of manual implementation of RCA rules. As this tool manages millions of daily alarms, it becomes unfeasible to process them manually, making the application of ML essential. Furthermore, ML algorithms have tremendous capabilities to detect patterns that humans could not, ideally exposing which specific failure causes a series of malfunctions, thus allowing system administrators to only focus their attention on the source problem instead of the multiple consequences. The ML approach proposed in this project is based on the causality among alarms, instead of their features, and uses the cartesian product of a specific problem, the involved technology, and the manufacturer, to extrap olate the correlations among faults. The results achieved reveal the tremendous potential of this approach and open the road to automatizing the definition of RCA rules, which represents a new vision on how to manage network failures efficiently., Ao longo dos últimos anos, uma das áreas que mais tem evoluído e estendido a sua utilização para uma infinidade de possibilidades é a Inteligência Artificial (IA). Com a crescente complexidade dos problemas, a resolução humana torna-se impossível, uma vez que a quantidade de informação e padrões que podem ser detectados é limitada, enquanto a IA prospera na dimensão do problema em análise. Além disso, como hoje em dia cada vez mais dispositivos tradicionais são informatizados, um número crescente de elementos está a pro duzir dados com muitas potenciais aplicações. Consequentemente, encontramo-nos no auge do Big Data, onde enormes volumes de dados são gerados, sendo totalmente inviável processá-los e analisá-los manualmente. Esta é uma das razões que tem levado à prosperidade da IA. Além disso, com a crescente complexidade das topologias de rede, é necessário assegurar o correcto fun cionamento de todos os equipamentos, evitando falhas em cascata entre dispositivos, o que pode levar a con sequências catastróficas dependendo da sua utilização. Assim, as ferramentas de Root Cause Analysis (RCA) tornam-se fundamentais, uma vez que são desenvolvidas para, através de regras estabelecidas pelos seus utilizadores, se aperceberem automaticamente das causas subjacentes quando algum equipamento apresenta anomalias. No entanto, com a crescente complexidade da rede, a definição de regras torna-se exponencial mente mais complicada, uma vez que os pontos possíveis de falha escalam tremendamente. Neste contexto, enquadrado pelo ambiente de rede e cenários de RCA da Altice Labs, foi definido o principal objectivo deste projecto de investigação. Este objectivo consiste na aplicação de técnicas de Machine Learning (ML) para extrapolar a relação entre os diferentes tipos de alarmes dos equipamentos, geridos pela ferramenta Alarm Manager, para ter uma melhor compreensão do impacto de uma falha em todo o sistema, facilitando e ajudando assim o processo de implementação manual das regras RCA. Como esta ferramenta gere milhões de alarmes diários, torna-se inviável processá-los manualmente, tornando essencial a aplicação do ML. Além disso, os algoritmos ML têm uma enorme capacidade para detectar padrões que os humanos não conseguem detectar, idealmente expondo quais as falhas específicas que causam uma série de falhas, permitindo assim que os administradores do sistema apenas concentrem a sua atenção no problema de raiz em vez das suas múltiplas consequências. A abordagem ML proposta neste projecto baseia-se na causalidade entre os alarmes, em vez das suas car acterísticas, e utiliza o produto cartesiano de um problema específico, da tecnologia envolvida, e do fabricante, para extrapolar as correlações entre falhas. Os resultados alcançados revelam o enorme potencial desta abor dagem e abrem o caminho para automatizar a definição de regras RCA, o que representa uma nova visão sobre como gerir eficazmente as falhas da rede.
- Published
- 2021
6. A Fault Correlation Approach to Detect Performance Anomalies in Virtual Network Function Chains
- Author
-
Domenico Cotroneo, Roberto Natella, Stefano Rosiello, Cotroneo, Domenico, Natella, Roberto, and Rosiello, Stefano
- Subjects
Service (systems architecture) ,Overload ,business.industry ,Computer science ,Quality of service ,Distributed computing ,020206 networking & telecommunications ,Cloud computing ,02 engineering and technology ,Anomaly detection ,Complex network ,Virtualization ,computer.software_genre ,NFV ,Software ,Fault correlation ,020204 information systems ,0202 electrical engineering, electronic engineering, information engineering ,business ,Safety, Risk, Reliability and Quality ,computer ,Virtual network - Abstract
Network Function Virtualization is an emerging paradigm to allow the creation, at software level, of complex network services by composing simpler ones. However, this paradigm shift exposes network services to faults and bottlenecks in the complex software virtualization infrastructure they rely on. Thus, NFV services require effective anomaly detection systems to detect the occurrence of network problems. The paper proposes a novel approach to ease the adoption of anomaly detection in production NFV services, by avoiding the need to train a model or to calibrate a threshold. The approach infers the service health status by collecting metrics from multiple elements in the NFV service chain, and by analyzing their (lack of) correlation over the time. We validate this approach on an NFV-oriented Interactive Multimedia System, to detect problems affecting the quality of service, such as the overload, component crashes, avalanche restarts and physical resource contention.
- Published
- 2017
7. A Statistical Approach for Estimating the Correlation between Lightning and Faults in Power Distribution Systems
- Author
-
Alberto Borghetti, Mario Paolone, M. Bernardi, Carlo Alberto Nucci, ROY BILLINTON, FRED TURNER, SAMY KRISHNASAMY, JAMES MCCALLEY, LINA BERTLING, LINA BERTLING, A. Borghetti, C.A. Nucci, M. Paolone, and M. Bernardi
- Subjects
Engineering ,Statistical methods ,Transient voltage suppressor ,Fault detection and isolation ,POWER QUALITY ,Electric power transmission networks ,Power distribution system (PDS) ,LIGHTNING INDUCED OVERVOLTAGES ,Electric appliances ,Lightning protection ,fault correlation ,Voltage dip (voltage sag) ,Electric power distribution ,Statistics ,Applied (CO) ,Local area networks ,Probabilistic methods ,simulation tools ,Distribution systems ,Correlation methods ,Electric load distribution ,MONTE CARLO METHOD ,Statistical approaches ,Uncertainty analysis ,Probability distribution ,Peak currents ,LIOV-EMTP code ,Fault (power engineering) ,Lightning ,Electric power systems ,power systems ,Electric power system ,Power system simulation ,Control theory ,ELECTROMAGNETIC TRANSIENTS ,international conferences ,Electronic engineering ,Accurate simulation ,Lightning flashes ,Probability ,Complex power ,transient voltages ,business.industry ,LIGHTNING LOCATION SYSTEMS ,Probability distributions ,Power transmission ,business ,Coordination reactions ,Estimation - Abstract
The paper deals with the subject of the source-identification of transient voltage disturbances in distribution system buses. In particular, a statistical procedure is proposed for the evaluation of the probability that a lightning flash detected by a Lightning Location System (LLS) could cause a fault and, therefore, relay interventions, generally associated with voltage dips. The proposed procedure is based on the coordinated use of the information provided by the LLS and the availability of an advanced simulation tool for the accurate simulation of lightninginduced voltages on complex power systems, namely the LIOV-EMTP code. The uncertainty levels of the stroke location and of the peak current estimations provided by the LLS are discussed and their influence on the lightning-fault correlation is analyzed. © Copyright KTH 2006.
- Published
- 2006
Catalog
Discovery Service for Jio Institute Digital Library
For full access to our library's resources, please sign in.