摘要
为分析短时间尺度下光伏出力爬坡事件,建立光伏出力爬坡率的广义高斯混合模型,并通过在不同时间尺度和天气类型下与不同分布模型对比检验,得出3阶广义高斯混合模型最适合用于描述短时间尺度光伏出力爬坡率的概率分布。其次,基于建立的爬坡率模型,对短时间尺度下光伏电站发生爬坡事件的概率进行总体评估。最后,采用旋转门算法,并提出一种改进的得分函数和定义的标签向量的方法,对不同天气类型下光伏爬坡事件进行识别筛选,发现长时间尺度下光伏出力爬坡持续时间、爬坡速率和爬坡幅值均呈幂律分布,且晴天与非晴天下3种指标的分布规律有差异。
Photovoltaic(PV)power fluctuations and PV rate events have significant influence on the operation of the grid. In order to analyze PV ramp events in short time scales,the generalized Gaussian mixture model(GGMM)of PV power ramp rate is set up. By comparison with different models under different time scales and weather types,the results show that 3 order GGMM performs best to describe the probability distribution of PV ramp rate under short time scales.Then,based on the proposed model of PV ramp rate,the probability of PV ramp events is assessed in short time scales.Finally,applying the swinging door algorithm,modified score function and defined label vector and so on,PV ramp events are identified under different weather types and it is concluded that ramp duration,ramp speed and ramp magnitude in a long time scales are all in power-law distribution. The distribution of three kinds of indicators under the sunny days and other days is different.
作者
李芬
李春阳
闫全全
王丽娟
杨兴武
段善旭
Li Fen;Li Chunyang;Yan Quanquan;Wang Lijuan;Yang Xingwu;Duan Shanxu(School of Electrical Engineering,Shanghai University of Electric Power,Shanghai 200090,China;Maintenance Company of SMEPC,Shanghai 200063,China;Meteorological Service Center of Hubei Province,Wuhan 430074,China;State Key Laboratory of Advanced Electromagnetic Engineering and Technology,Huazhong University of Science and Technology,Wuhan 430074,China)
出处
《太阳能学报》
EI
CAS
CSCD
北大核心
2019年第11期3289-3298,共10页
Acta Energiae Solaris Sinica
基金
国家自然科学基金青年项目（51307105）
上海绿色能源并网工程技术研究中心（13DZ2251900）
上海市经济和信息委员会专项资金（沪CXY-2016-012)
关键词
光伏出力波动
爬坡率
爬坡事件
广义高斯混合模型
旋转门算法
PV power fluctuation
ramp rate
ramp event
generalized Gaussian mixture model
swinging door algorithm
作者简介
通信作者:李芬(1984-),女,博士、副教授,主要从事可再生能源发电评估及功率预报技术方面的研究。beckyhust@163.com。
