大规模风电多尺度出力波动性的统计建模研究

Research on statistical modeling of large-scale wind farms output fluctuations in different spacial and temporal scales

掌握风电波动性规律是解决大规模风电并网运行难题的基础,从相关性和平滑性两个方面,研究了不同时空尺度下风电出力波动性的统计学规律。首先基于华北地区的实测数据对风电功率波动概率分布模型进行假设检验分析,证明了风电功率变化率采用混合高斯分布而非正态分布具有更好的拟合精度,然后利用数理统计方法建立了华北风电功率波动率概率密度分布的二分量一维混合高斯分布模型和置信区间模型,分析了模型参数受风电规模的影响规律。在此基础上,提出了能够有效反应风电群聚对出力波动性平滑效果的评价指标,并进一步建立该指标与相关系数之间的一次函数,对平滑效果与相关性之间的消长关系进行了描述。给出了评价规模化风电波动性、平滑性与相关性等主要特性的基本统计学方法,有效提升了风电波动性统计模型的精确性以及指标的完整性。

To model the fluctuation characteristics of wind power is the basis for solving wind power integration problems. Considering the correlation and smoothing effect of wind power, the paper studies statistical models of wind power fluctuations in different spatial and temporal scales. It first uses hypothesis testing to analyze the distribution model of wind power fluctuations based on measured data, verifying that the combined Gaussian distribution model has higher fitting precision than normal distribution. The paper then establishes the combined Gaussian distribution model and the confidence interval model for the probability density distribution of wind power variation pattern, and studies the law about the scale influence on parameters. Then the evaluation index of smoothing effect which can reflect the effect of wind-farm aggregation on smoothing wind power fluctuations is defined. And the relationship between such index and the correlation coefficient modeled as the linear function reflects the growth and decline relationship between smoothing effect and correlation. The paper proposes the statistical methods for evaluating large-scale wind power fluctuation and smoothing effect and correlation, which improves the accuracy of the fluctuation models and the integrality of indices.