| 引用本文: | 杨茂,张罗宾,崔杨,杨琼琼,黄宾阳.基于混合半云模型的风速-功率曲线建模方法[J].电力自动化设备,2020,40(5): |
| YANG Mao,ZHANG Luobin,CUI Yang,YANG Qiongqiong,HUANG Binyang.Wind speed-power curve modeling method based on hybrid half-cloud model[J].Electric Power Automation Equipment,2020,40(5): |
|
| 摘要: |
| 风速-功率曲线的准确建模是风电机组出力态势评估和风电功率预测的关键基础之一。计及风电映射关系的不确定性及功率曲线的分布形态，提出一种基于混合半云模型的建模策略来实现对风功率数据固有和随机分布特征的挖掘和建模。引入最优组内云熵算法快速有效地剔除异常数据；采用逆向云发生器求取期望、熵与超熵数字特征来定量刻画风速-功率对应关系的不确定性，构建腰部数据的半云模型；通过X条件云发生器和正向云发生器分别求取腰部和上部数据的功率云滴，实现定性数字特征向定量数据的转换。以中国东北某大型风电场的实测数据为例，从数据质量、频率分布和风功率预测等维度分析混合半云模型，验证了所提方法的可行性。 |
| 关键词: 混合半云模型 风速-功率曲线 X条件云发生器 最优组内云熵算法 不确定性预测 |
| DOI：10.16081/j.epae.202005004 |
| 分类号:TM614 |
| 基金项目:国家重点研发计划项目(促进可再生能源消纳的风电/光伏发电功率预测技术及应用)(2018YFB0904200) |
|
| Wind speed-power curve modeling method based on hybrid half-cloud model |
|
YANG Mao1, ZHANG Luobin1, CUI Yang1, YANG Qiongqiong2, HUANG Binyang3
|
|
1.Key Laboratory of Modern Power System Simulation and Control & Renewable Energy Technology, Ministry of Education, Northeast Electric Power University, Jilin 132012, China;2.Baoji Power Supply Company of State Grid Shaanxi Electric Power Company, Baoji 721004, China;3.Maintenance Branch of State Grid Chongqing Electric Power Supply Company, Chongqing 400039, China
|
| Abstract: |
| Accurate wind speed-power curve modeling is one of the key basics for situation assessment of wind turbine output and wind power forecasting. Considering the uncertainty of wind power mapping relationship and the distribution pattern of power curve, a modeling strategy based on HHC(Hybrid Half-Cloud model) is proposed for mining and modeling of the inherent and random distribution characteristics of wind power data. The OSCE(Optimal Segmentation Cloud Entropy) algorithm is introduced to eliminate the abnormal data quickly and effectively. The backward cloud generator is adopted to obtain the digital characteristics of expectation, entropy and hyper-entropy for quantify the uncertainty of wind speed-power relationship, and a semi-cloud model of waist data is constructed. The power cloud droplets of waist and upper data are calculated by X condition cloud generator and forward cloud generator respectively, and the trans-formation from the qualitative features to quantitative data is realized. Taking the measured data of a large scale wind farm in Northeast China as an example, HHC is analyzed from the perspectives of data quality, frequency distribution and wind power forecasting, which verifies the feasibility of the proposed method. |
| Key words: hybrid half-cloud model wind speed-power curve X condition cloud generator optimal segmentation cloud entropy algorithm uncertainty forecasting |
