摘要
为了解决核电站故障识别难度高、工作量大的问题,从核电站对故障预警和诊断的功能需求出发,通过分析多种智能算法和核电站应用场景的适配性,提出了基于卷积神经网络(CNN)、堆叠自编码(SAE)网络、故障树分析(FTA)等先进技术的核电站智能故障预警和诊断功能模块设计方案。该方案将提升核电站智能化水平,实现智能化、自动化的故障监视和诊断,为运行人员提供决策支持,减轻运行人员工作负担,提高故障处理的安全性和时效性。
To solve the problem of high difficulty and workload of fault identification in nuclear power plants, from the functional requirements of nuclear power plants for fault warning and diagnosis, the design scheme of intelligent fault warning and diagnosis function module of nuclear power plants based on advanced technologies such as convolutional neural network(CNN), stacked auto-encoder(SAE) network and fault tree analysis(FTA) and so on is proposed by analyzing the suitability of various intelligent algorithms and nuclear power plants application scenarios. The solution will enhance the intelligence level of nuclear power plants, realize intelligent and automated fault monitoring and diagnosis, provide decision support for operation personnel, reduce operation personnel workload, and improve the safety and timeliness of fault handling.
作者
王梦月
李鸣谦
万欣
WANG Mengyue;LI Mingqian;WAN Xin(China Nuclear Power Engineering Co.,Ltd.,Beijing 100080,China;School of Control and Computer Engineering,North China Electric Power University,Beijing 102200,China)
出处
《自动化仪表》
CAS
2023年第2期65-68,共4页
Process Automation Instrumentation
关键词
核电站
故障预警
故障诊断
卷积神经网络
堆叠自编码网络
故障树分析
Nuclear power plants
Fault warning
Fault diagnosis
Convolutional neural network(CNN)
Stacked auto-encoder(SAE)network
Fault tree analysis(FTA)
