光伏发电功率预测方法研究综述  

Review of PV power generation prediction

作　　者：钱振[1] 蔡世波[1] 顾宇庆[2] 童建军[1] 鲍官军[1] 

机构地区：[1]浙江工业大学,特种装备制造与先进加工技术教育部/浙江省重点实验室,浙江杭州310032 [2]上海汽车集团股份有限公司技术中心,上海200041

出　　处：《机电工程》2015年第5期651-659,共9页Journal of Mechanical & Electrical Engineering

基　　金：国家高技术研究发展计划“863”计划资助项目(SS2013AA050405);教育部博士点基金资助项目(20123317110004);国际科技合作项目(S2013ZR0609);浙江省重点科技创新团队资助项目(2011R50011-16);特种装备制造与先进加工技术教育部/浙江省重点实验室(浙江工业大学)开放基金资助项目(2011EM001)

摘　　要：针对太阳能光伏发电输出功率不稳定的问题,分析了国内外光伏发电功率预测方法的研究现状,总结并提出了今后研究的建议方向,介绍了太阳辐射、晴空指数、日照时数、云、温度、风速和面板积尘等影响光伏发电的各种因素,着重综述了2008年以来光伏发电功率预测尤其是短期预测的研究概况,对光伏功率预测的时间尺度和精度这两个重要指标进行了讨论,分析了数据基础、影响因素、精度指标、时间响应和评价标准等光伏发电预测中需要重点关注的技术难点。研究结果表明,近年来的研究工作主要集中在短期预测,到目前为止对于各种预测方法尚没有被公认的评价标准,难以对各种预测算法进行评估比较。最后结合综述分析,提出了今后光伏发电功率预测研究工作的建议方向。Aiming at the problems of unstable solar PV power generation, the related prediction researches were analyzed and the suggestions of researching aspects in the future were proposed. The factors that influencing the PV power generating, such as solar radiation, clearness index, sunshine duration, cloud, temperature, wind speed and dust accumulation, were introduced. Recent progress of PV power prediction from 2008, especially short-term prediction research, was emphasized and overviewed. The two important index of PV power prediction, Temporal scale and accuracy were analyzed. The technical challenges that were needed to focus on in the PV power generation prediction, like data support, factors identification, accuracy metrics, time response and evaluation standard, were explored. The results indicate that most of recent research work are dedicated to short-term prediction and there is no approved standard for all kind of prediction methodologies till now, which makes it impossible to evaluate the algorithms. Finally, the suggestions for PV power generation prediction research in the future are proposed by the summary of the review analysis.

关 键 词：功率预测 光伏系统 评价标准 

分 类 号：TM615[电气工程—电力系统及自动化]