摘要
以我国南方某地区为例,收集了该地区连续48个月的电力事故数据及其对应的15个气象要素资料。首先用因子分析法消除了15个气象要素的多重共线性,提取了温度因子、降水因子、湿度因子、风力因子4类主要因素,其次应用Logistic回归建立了气象因素对电力事故的影响模型。模型探究了气象条件与电力事故的内在联系,并用2010年的检验样本验证了模型拟合的准确性,为电力事故预警机制的建立进行了积极的探讨。
Taking a certain region in South China for example, the power system accident data in continuous 48 months of this region and corresponding data of 15 meteorological factors are collected. Firstly, the multicollinearity of the 15 meteorological factors are eliminated by factor analysis, and four kinds of main factors such as temperature factor, precipitation factor, humidity factor and wind power factor are extracted; secondly, an impact model of meteorological factors on power system accidents is established by Logistic regression. Using the established impact model the internal relations between meteorological conditions and power system accidents are explored, and the fitting accuracy of the proposed model is validated by test samples in 2010. The proposed impact model is available for reference to the establishment of early warning mechanism for power system accidents.
出处
《电网技术》
EI
CSCD
北大核心
2013年第6期1683-1687,共5页
Power System Technology
基金
教育部人文社会科学研究项目(10YJA630088)~~
关键词
电力事故
气象条件
影响模型
power accidents
meteorological conditions
impact model
