中国电机工程学报
中國電機工程學報
중국전궤공정학보
ZHONGGUO DIANJI GONGCHENG XUEBAO
2013年
34期
17-24
,共8页
茆美琴%龚文剑%张榴晨%曹雨%徐海波
茆美琴%龔文劍%張榴晨%曹雨%徐海波
묘미금%공문검%장류신%조우%서해파
集合经验模态分解%支持向量机%光伏电站%短期预测%组合预测模型
集閤經驗模態分解%支持嚮量機%光伏電站%短期預測%組閤預測模型
집합경험모태분해%지지향량궤%광복전참%단기예측%조합예측모형
ensemble empirical mode decomposition%support vector machines%photovoltaic power station%short-term prediction%combined forecasting model
针对光伏电站日前小时短期出力预测问题,提出一种基于集合经验模态分解(ensemble empirical mode decomposition,EEMD)和支持向量机(support vector machines,SVM)的 EEMD-SVM 组合模型预测方法。该方法将天气类型分为突变天气和非突变天气。首先采用EEMD分解法将历史光伏电站小时出力数据分解为一系列相对平稳的分量序列,对不同的天气类型考虑不同的气象因素,然后采用SVM法对所分解的各分量序列分别建立预测模型,选用不同的核函数和参数以取得单个分量序列的最佳预测精度。算例结果表明,分类建模思想和EEMD-SVM组合预测法能够使突变天气预测结果的平均绝对百分比误差减少5%,非突变天气的减少3%。
針對光伏電站日前小時短期齣力預測問題,提齣一種基于集閤經驗模態分解(ensemble empirical mode decomposition,EEMD)和支持嚮量機(support vector machines,SVM)的 EEMD-SVM 組閤模型預測方法。該方法將天氣類型分為突變天氣和非突變天氣。首先採用EEMD分解法將歷史光伏電站小時齣力數據分解為一繫列相對平穩的分量序列,對不同的天氣類型攷慮不同的氣象因素,然後採用SVM法對所分解的各分量序列分彆建立預測模型,選用不同的覈函數和參數以取得單箇分量序列的最佳預測精度。算例結果錶明,分類建模思想和EEMD-SVM組閤預測法能夠使突變天氣預測結果的平均絕對百分比誤差減少5%,非突變天氣的減少3%。
침대광복전참일전소시단기출력예측문제,제출일충기우집합경험모태분해(ensemble empirical mode decomposition,EEMD)화지지향량궤(support vector machines,SVM)적 EEMD-SVM 조합모형예측방법。해방법장천기류형분위돌변천기화비돌변천기。수선채용EEMD분해법장역사광복전참소시출력수거분해위일계렬상대평은적분량서렬,대불동적천기류형고필불동적기상인소,연후채용SVM법대소분해적각분량서렬분별건립예측모형,선용불동적핵함수화삼수이취득단개분량서렬적최가예측정도。산례결과표명,분류건모사상화EEMD-SVM조합예측법능구사돌변천기예측결과적평균절대백분비오차감소5%,비돌변천기적감소3%。
A combined prediction method based on ensemble empirical mode decomposition (EEMD) and support vector machine (SVM) was proposed to tackle the problem of the short-term forecast of hourly output photovoltaic system (PVS) a day ahead. Weather types were classified into abnormal day (weather changed suddenly) and normal day. Firstly, the history data for hourly output of PVS was decomposed into a series of components by using EEMD method. Considering different factors for different types of weather, the different models were built and different kernel functions and parameters were chosen to deal with each component of the data by using SVM. Simulation results show that the proposed classification modeling ideas and EEMD-SVM combination forecasting method enable the mean absolute percentage error results for the abnormal days decreased by 5%, and normal day decreased by 3%compared with the traditional SVM method.