【摘要】：風(fēng)電功率的概率預(yù)測能提供風(fēng)電功率的預(yù)測區(qū)間或分布函數(shù),國內(nèi)相關(guān)的研究和應(yīng)用尚處于起步階段。文中對風(fēng)電功率概率預(yù)測的基本框架、主要模式、難點和熱點進(jìn)行了綜述。首先,明確了概率預(yù)測的概念及其適用問題。然后,對概率預(yù)測的建模方法提出了兩種不同的分類方式:按照是否進(jìn)行條件化假設(shè)或參數(shù)化假設(shè)進(jìn)行分類,并介紹了概率預(yù)測中涉及的新型算法和概率預(yù)測的評價指標(biāo)。最后,結(jié)合概率預(yù)測發(fā)展現(xiàn)狀,針對誤差分析不精細(xì)、概率預(yù)測與電力系統(tǒng)結(jié)合不充分等不足,總結(jié)了今后的發(fā)展方向和需要進(jìn)一步探索的研究內(nèi)容。
[Abstract]:The probability prediction of wind power can provide the prediction interval or distribution function of wind power, and the related research and application in China is still in the initial stage. In this paper, the basic framework, main modes, difficulties and hotspots of the probability prediction of wind power are reviewed. Firstly, the concept of probabilistic prediction and its application are clarified. Then, two different classification methods are proposed for the modeling method of probabilistic prediction: the classification is based on the conditional hypothesis or parametric hypothesis, and the new algorithm and the evaluation index of probabilistic prediction are introduced. Finally, according to the present situation of probabilistic prediction, aiming at the deficiency of error analysis and insufficient combination of probabilistic prediction and power system, this paper summarizes the future development direction and research contents that need to be further explored.
【作者單位】： 清華大學(xué)電機(jī)工程與應(yīng)用電子技術(shù)系;電力系統(tǒng)及發(fā)電設(shè)備控制和仿真國家重點實驗室清華大學(xué);甘肅省電力公司風(fēng)電技術(shù)中心;
【基金】：國家自然科學(xué)基金重大項目(51190101) 國家重點研發(fā)計劃資助項目(2016YFB0900101) 國家電網(wǎng)公司科技項目(522727160002)~~
【分類號】：TM614
，

本文編號：2426121