摘要
针对设备出现重载和长期超负荷运行容易导致故障甚至设备烧损的问题,提出了基于主变负载状况的变压器负荷预测和风险评估模型。分析了影响主变负载的因素,为研究不同因素对主变负载的影响程度,计算各因素间的相关系数矩阵和梯度提升树,以便相关人员对主变负载预测和风险评估的结果进行分析,找出过载原因,及时处理。采用奇异谱分析和BP神经网络对变压器负载进行预测,利用预测的负荷建立主变负荷风险评估模型,综合考虑负载相关因子和后果严重度。最后,算例分析证明了所提模型的可行性。
Aiming at the problem that the heavy load and long-term overload operation of equipment can easily lead to failures and even equipment burnout,this paper proposes a transformer load forecasting and risk assessment model based on the load status of the main transformer.Firstly,the factors affecting the load of the main transformer are analyzed.In order to study the degree of influence of different factors on the load of the main transformer,the correlation coefficient matrix and the gradient promotion tree between the factors are calculated,so that the relevant personnel can carry out the analysis of the main transformer load prediction and risk assessment results,find out the cause of overload,and deal with it in time.Secondly,the singular spectrum analysis and BP neural network are used to predict the load of the transformer,and a risk assessment model for the main transformer load is established based on the predicted load.The load-related factors and the severity of the consequences are considered comprehensively.Finally,the analysis of a calculation example proves the feasibility of the proposed model.
作者
范强
文屹
吕黔苏
张迅
肖宁
刘君
FAN Qiang;WEN Yi;LV Qiansu;ZHANG Xun;XIAO Ning;LIU Jun(Electric Power Research Institute of Guizhou Power Grid Co.,Ltd.,Guiyang 550002,China;Guizhou Power Grid Co.,Ltd.,Guiyang 550002,China)
出处
《电测与仪表》
北大核心
2024年第11期99-106,共8页
Electrical Measurement & Instrumentation
基金
中国南方电网有限责任公司科技项目(GZKJXM2 0200537)。
关键词
奇异谱分析
神经网络
风险评估
负荷预测
singular spectrum analysis
neural network
risk assessment
load forecasting