摘要
为了有效识别电网运行的关键风险要素,基于典型电网企业2014—2019年代表性事故案例,采用贝叶斯网络的机器学习方法分别构建了电网事故、设备事故及人身事故致因的贝叶斯网络模型,分析各风险因素对事故的影响程度并反向诊断事故发生的关键诱因。结果表明:(1)3种贝叶斯网络模型预测精度分别达到87.85%、89.24%、96.88%;(2)不同类型事故的关键风险因素存在差异,但人因仍是主要致因。电网事故的关键影响因素为巡检不到位、检修质量不良和验收不合格;设备事故的关键风险因素为处理不当、巡检不到位和施工质量不良;人身事故的关键风险因素为安全意识缺乏、施工质量不良、监护不到位和验收不合格。最后,对电网系统安全运行提出了针对性建议。
The present paper is aimed to identify the key risk factors of the grid operation effectively based on the investigation of the 288 representative accidents of the typical power grid enterprises from 2014 to 2019 machine learning method of Bayesian Network.Its aim is to build up the Bayesian Network Models of the grid accident causes,the equipment accident causes and the accident its own cause due to the inefficient control done respectively.As the result of our investigation,the risk factors for the grid accidents can be analyzed and identified from the following 4 aspects:the human factors;the equipment factors;the environmental factors,as well as those of the management.And,then,we have managed to analyze the impact of each risk factor on the accident with the key influential factors of the accident diagnosed in a reversed way.The results of our analysis can be shown as follows:(1)The 3 Bayesian Network Models can help to forecast and predict the likely phenomena to be occurring in the accidents,with the forecast and predicted accuracy being up to87.85%,89.24%,96.88%,respectively;And,(2)there may exist differences or gaps in the prediction of the principal risk factors for the different types of the accidents,though the human factors should still serve as the chief ones.At the same time,the key influential factors of the grid accidents remain far from being enough made out in the profound root-cause searching and examination,from the point of view of maintenance qualified and/or unqualified being accepted.As to the key risk factors of the equipment and facility accidents,the chief causes can be attributed to the improper handling,inadequate inspections,in addition to the poor self-construction quality.The principal risk factors on account of the accidents for the inadequate training or short of the safety sense of responsibility,or of the poor construction quality of the machine or equipment of their own,the inadequate monitoring and unqualified acceptance.(3)The following nodes are also to be ranked in order
作者
王东
王佳琪
陈红
苏峰
杨文清
吴梅芬
彭启伟
龙如银
WANG Dong;WANG Jia-qi;CHEN Hong;SU Feng;YANG Wen-qing;WU Mei-fen;PENG Qi-wei;LONG Ru-yin(State Grid Corporation of China,Beijing 100031,China;School of Economics and Management,China University of Mining and Technology,Xuzhou 221116,Jiangsu,China;School of Business,Jiangnan University,Wuxi 214122,Jiangsu,China;Nanrui Group Co.,Ltd./State Grid Electric Power Research Institute,Nanjing 211111,China;The Institute for Jiangnan Culture,Jiangnan University,Wuxi 214122,Jiangsu,China)
出处
《安全与环境学报》
CAS
CSCD
北大核心
2021年第5期1947-1956,共10页
Journal of Safety and Environment
基金
国家电网公司总部科技项目(SGJSWX00FZJS1901915)
中国矿业大学文化传承创新双一流建设专项(2018WHCC03)。
关键词
安全工程
风险评价
电网
事故致因
贝叶斯网络
关联模型
safety engineering
risk assessment
power grid
cause of accidents
Bayesian network
correlation model
