摘要
在智能电网的发展背景下,电磁式断路器的健康管理对提高电力系统的供电可靠性、安全性和稳定性具有重要意义,因此受到了研究者们的广泛关注。状态监测和智能评估是电磁式断路器健康管理的主要内容。首先,通过对近些年相关文献的梳理,从传感监测、特征提取与降维、故障诊断、健康评估和剩余寿命预测方面总结当前的研究现状,并着重分析了现有技术中存在的局限。然后,简要介绍电磁式断路器健康管理系统的硬件实现方案,阐明了在实际应用上的不足。最后,讨论了状态监测与智能评估所面临的研究挑战以及未来的发展趋势,希望能给该领域的研究人员提供一定参考。
As the key switchgear in a power system,an electromagnetic circuit breaker plays a vital role in controlling energy flow,protecting the system circuit,and isolating fault current.It is of great significance to realize the health management of electromagnetic circuit breakers to improve the reliability,security,and stability of power supply systems.With the development of advanced sensor technology,industrial big data technology,and artificial intelligence technology,the“data-driven approach”has gradually replaced the previous“mechanism modeling approach”to become the mainstream research direction of electromagnetic circuit breaker health management technology.Condition monitoring and intelligent assessment are the main contents in this field.Through a detailed review of the main references or published technology patents in the past decade,the current research status of“sensing monitoring”,“feature extraction and dimension reduction”,“fault diagnosis”,“health assessment”,and“residual life prediction”were analyzed.For“sensing monitoring”,almost all sensing signals and their monitoring methods were covered from electrical,mechanical,and temperature characteristics.The advantages and disadvantages of each sensing signal were expounded,respectively.It is pointed out that the sensitivity of various signals to the deterioration of different parts of the circuit breaker is different.It is difficult to reflect the overall deterioration state of the circuit breaker by relying only on a single type of signal,so it may be necessary to monitor multi-source sensing signals for fusion analysis.For“feature extraction and dimension reduction”,typical feature extraction methods,such as the short-time energy method,wavelet transform,and empirical mode decomposition,were mainly analyzed from the time domain and time-frequency domain mixture two aspects.Furthermore,two types of feature dimension reduction methods were summarized;For“fault diagnosis”,building expert systems and applying machine learning are the mainstream research in this field.The existing typical algorithms were summarized,and it is pointed out that the above two kinds of methods need rich expert experience and comprehensive data sources,respectively,to achieve better diagnosis results.For“health assessment”,the existing methods were summarized into two categories:level assessment and quantitative assessment.It is pointed out that there are few related studies in this field,and further research is needed.For“residual life prediction”,the existing methods of“direct mapping based mapping model”and“indirect prediction based deterioration model”were introduced and summarized respectively,and their shortcomings were discussed.In addition,the hardware implementation schemes of condition monitoring and intelligent assessment techniques in existing research were briefly summarized.Finally,considering the existing problems,the following research methods and development trends of scientific and technological problems are discussed.(1)Combined with the working conditions of circuit breakers and the Internet of things technology,developing a special sensor monitoring method and constructing a distributed multi-source sensor network with high reliability and low cost will become the mainstream way.(2)Digital twin technology,multi-physical field coupling simulation technology,and virtual simulation technology combined will become a major trend to explore the deterioration law under the comprehensive influence in multiple ways.(3)Multi-disciplinary integration can enhance the completeness of intelligent assessment methods.(4)The application of cloud computing,5G,and other new technologies should be accelerated.The embedded integration,real-time control,distributed control,and system security should be comprehensively considered as far as possible to jointly promote the practical process of electromagnetic circuit breaker health management system platform.
作者
韩翔宇
纽春萍
何海龙
吴佳宝
陈子薇
Han Xiangyu;Niu Chunping;He Hailong;Wu Jiabao;Chen Ziwei(State Key Laboratory of Electrical Insulation and Power Equipment Xi’an Jiaotong University,Xi’an 710000 China)
出处
《电工技术学报》
EI
CSCD
北大核心
2023年第8期2191-2210,共20页
Transactions of China Electrotechnical Society
关键词
断路器
状态监测
智能评估
数据驱动
人工智能
Circuit breaker
condition monitoring
intelligent assessment
data driving
artificial intelligence
