摘要
提高电力系统可靠性评估的准确度是改善电网安全运行的关键之一,元件可靠性计算又是系统可靠性评估的基石。传统的元件可靠性采用统计模型,无法表征元件实时状态;而考虑潮流、电压、天气等运行条件影响的元件分析模型,又受到数据来源的限制。综合前2类模型的特点,基于支持向量机和灰色预测技术,根据元件运行时间、输电线路所在区域的污秽等级和落雷密度,提出了2种实用的输电线路运行可靠性预测模型。采用2种预测模型对华东电网500 kV线路可靠性预测的结果表明,文中提出的预测模型较为真实地反映了元件的运行可靠性水平。
The accuracy of power system reliability evaluation based on the evaluation of component reliability is one of the key factors for improving secure operation of power systems.The calculation of component reliability,in turn,is the foundation for system reliability evaluation.The classical statistical model of component reliability cannot characterize the component＇s real-time state,while some new models that take power flow,voltage,weather and other operating conditions into consideration may experience difficulty in obtaining the actual data.This paper presents two practical models for predicting operating reliability of transmissions,the grey prediction and supporting vector machine regression models,both taking the operating time,lightning density and polluting parameters into account.Test results based on the historical operating data on the 500 kV transmission system of East China Grid show that the models proposed can fairly accurately reflect the components＇ operating state in reliability evaluation.
出处
《电力系统自动化》
EI
CSCD
北大核心
2010年第24期18-22,共5页
Automation of Electric Power Systems
基金
国家高技术研究发展计划(863计划)资助项目(2007AA05Z458)~~
关键词
元件可靠性
统计模型
预测模型
灰色预测
支持向量机回归
reliability of components
statistical model
prediction model
gray prediction
supporting vector machine regression
