超短期风电功率爬坡事件对风电功率实时预测误差的影响研究  

ULTRA-SHORT-TERM WIND POWER CLIMBING EVENT RESEARCH ON THE EFFECT OF WIND POWER REAL-TIME PREDICTION ERROR

作　　者：杨茂[1] 刘红柳[1] 

机构地区：[1]东北电力大学电气工程学院,吉林132012

出　　处：《太阳能学报》2017年第3期571-577,共7页Acta Energiae Solaris Sinica

基　　金：国家重点基础研究发展(973)计划(2013CB228201);国家自然科学基金(51307017);吉林省科技发展计划(20140520129JH);吉林省产业技术研究与开发专项(2014Y124)

摘　　要：阐述了超短期风电功率爬坡事件的研究背景及定义,建立超短期风电功率爬坡事件的检测和统计方法,利用持续法、支持向量机(SVM)和组合预测法3种风电功率实时预测方法,分析发生超短期风电功率爬坡事件时风电功率预测误差指标的变化,定量给出超短期风电功率爬坡事件对风电功率预测误差的影响。实例表明,风电功率爬坡事件具有小概率高风险特性,风电功率预测精度随着超短期风电功率爬坡事件频率的增大而降低。The background and definition of the ultra-short-term wind power climbing events were described, and the detection and statistical methods were established, and the detection and statistical method of ultra-short-term wind power climbing events was established. Using continuous method, the BP neural network and SVM （support vector machine） three wind power real-time prediction method to analyze the change of the wind power prediction error when the ultra-short-term wind power climbing event occurs, and give the influence of ultra-short-term wind power climbing event on the prediction error of wind power quantitatively. An example of a wind farm showed that the accuracy of wind power prediction can be greatly reduced when the ultra-short-term wind power climbing event occurs. At the same time, it is difficult to guarantee the high accuracy of the wind power prediction. Examples showed that wind power prediction accuracy decreases with increasing frequency of occurrence wind power climbing events in the wind farm.

关 键 词：爬坡事件 风电功率 实时预测 预测误差 

分 类 号：TM71[电气工程—电力系统及自动化]