电力系统自动化
電力繫統自動化
전력계통자동화
Automation of Electric Power Systems
2015年
20期
12-18
,共7页
风电功率预测%动态组合预测%Cook 距离%自适应遗忘因子
風電功率預測%動態組閤預測%Cook 距離%自適應遺忘因子
풍전공솔예측%동태조합예측%Cook 거리%자괄응유망인자
wind power prediction%dynamic combination forecast%Cook's distance%adaptive forgetting factors
风电功率预测对电力系统运行、调度计划具有重要意义。针对目前单体预测模型的局限性,文中给出了基于可变向量遗忘因子的自适应指数动态优选组合预测模型。模型首先采用数值天气预报作为3种单体预测模型的主要输入,所选模型结合物理和统计模型的优点,同时兼有线性与非线性特点。然后,在单体预测结果的基础上,分别采用递归最小二乘方法、协方差优选组合方法及多层感知器网络对单体模型的预测结果进行组合。最后,引入基于 Cook 距离的向量遗忘因子,利用 Cook 距离评估新观测值对参数估计的影响,采用基于可变向量遗忘因子的自适应指数组合模型动态分配模型权系数,对加权组合得到的3种组合预测结果进行再次组合,在自适应过程中实现模型优选,得到最终的预测结果。算例结果表明,所述优选组合预测模型能够在超短期组合预测的过程中实现模型优选,可有效提高风电功率预测精度。
風電功率預測對電力繫統運行、調度計劃具有重要意義。針對目前單體預測模型的跼限性,文中給齣瞭基于可變嚮量遺忘因子的自適應指數動態優選組閤預測模型。模型首先採用數值天氣預報作為3種單體預測模型的主要輸入,所選模型結閤物理和統計模型的優點,同時兼有線性與非線性特點。然後,在單體預測結果的基礎上,分彆採用遞歸最小二乘方法、協方差優選組閤方法及多層感知器網絡對單體模型的預測結果進行組閤。最後,引入基于 Cook 距離的嚮量遺忘因子,利用 Cook 距離評估新觀測值對參數估計的影響,採用基于可變嚮量遺忘因子的自適應指數組閤模型動態分配模型權繫數,對加權組閤得到的3種組閤預測結果進行再次組閤,在自適應過程中實現模型優選,得到最終的預測結果。算例結果錶明,所述優選組閤預測模型能夠在超短期組閤預測的過程中實現模型優選,可有效提高風電功率預測精度。
풍전공솔예측대전력계통운행、조도계화구유중요의의。침대목전단체예측모형적국한성,문중급출료기우가변향량유망인자적자괄응지수동태우선조합예측모형。모형수선채용수치천기예보작위3충단체예측모형적주요수입,소선모형결합물리화통계모형적우점,동시겸유선성여비선성특점。연후,재단체예측결과적기출상,분별채용체귀최소이승방법、협방차우선조합방법급다층감지기망락대단체모형적예측결과진행조합。최후,인입기우 Cook 거리적향량유망인자,이용 Cook 거리평고신관측치대삼수고계적영향,채용기우가변향량유망인자적자괄응지수조합모형동태분배모형권계수,대가권조합득도적3충조합예측결과진행재차조합,재자괄응과정중실현모형우선,득도최종적예측결과。산례결과표명,소술우선조합예측모형능구재초단기조합예측적과정중실현모형우선,가유효제고풍전공솔예측정도。
Wind power prediction plays a very important role in the dispatching and operation to maintain power system safety and stability.Due to the limitation of single prediction model with large errors at some individual points,a novel two-step dynamic optimal combination model of wind power prediction based on variable adaptive vector forgetting factor is proposed, which takes the numerical weather prediction (NWP) data as inputs.Three single prediction models are adopted considering advantages of physical and statistical models.Recursive least square (RLS),covariance optimization combination and multi-layer perceptron (MLP) network are proposed to calculate the weights of each single prediction model.Effect of new observations on parameter estimation is further evaluated based on Cook”s distance,and the weight coefficients of three combined prediction models are calculated through adaptive exponential dynamic combination model by use of variable vector forgetting factor.The proposed model is used to combine the results from the first step to obtain the optimal model in the adaptive process.Case studies show that the optimal forecasts can be obtained in ultra-short term wind power prediction using the proposed model,which improves the accuracy of prediction effectively.