基于自适应预测箱的风电场景分析方法 |
点此下载全文 |
引用本文:杨娴1,汪柳兵2,李德林1.基于自适应预测箱的风电场景分析方法[J].电网与清洁能源,2020,36(8):82~90 |
摘要点击次数: 202 |
全文下载次数: 95 |
|
|
中文摘要:针对风电出力的不确定性特点,提出一种场景分析方法,通过生成风电的时序场景集刻画其出力的不确定性信息。将场景分析方法分为场景生成部分和场景消减部分。在场景生成部分,构建自适应预测箱描述不同预测出力幅值下预测误差的概率分布,结合多元标准正态分布和逆变换技术得到满足自相关性的初始场景集;至于场景消减方面,改进传统K-means算法提高聚类算法的聚类水平和聚类稳定性。实例分析表明,该场景分析技术对风电概率信息的刻画具有较高的准确性,适用于涉及大规模风电并网的电力系统调度问题。 |
中文关键词:场景分析方法 自适应预测箱 多元标准正态分布 场景消减 |
|
A Wind Power Scenarios Analysis Method Based on AdaptivePrediction Box |
|
|
Abstract:In view of the uncertain characteristics of wind power output, this paper proposes a scenarios analysis method, which depicts the uncertainty information of wind power output by generating time series scene sets of wind power. The scenarios analysis method is divided into a scene generation part and a scene reduction part. In the scene generation part,an adaptive prediction box is constructed to describe the probability distribution of prediction errors under different predicted output amplitudes, combined with multivariate standard normal distribution and inverse transform technology to obtain an initial scene set that meets autocorrelation; as for the scene reduction, the traditional K-means algorithmis improved to enhance the clustering level and stability of the clustering algorithm. The example analysis shows that the scenarios analysis method has a high accuracy in describing the probability information of wind power, and it is suitable for the dispatching of the power system involving large-scale wind power grid connection. |
keywords:scenarios analysis method adaptive prediction box multivariate standard normal distribution scenarios reduction |
查看全文 查看/发表评论 下载PDF阅读器 |