基于数据特征提取的风电功率预测误差估计方法
An Estimation Method for Wind Power Forecast Errors Based on Numerical Feature Extraction
-
摘要: 估计风电功率预测误差对电力系统的调度与控制、安全与防御等方面具有重要意义。从风电历史数据和日前预测数据特征提取的角度,研究了日前风电功率预测误差的估计方法。首先,提取并分析影响风电功率预测误差的主要因素,包括风电出力波动程度、风电功率幅值、预测方法等,并通过数据统计分析其相关性。然后,结合风电历史运行数据,采用多元线性回归方法建立风电功率预测误差的估计模型。最后,基于比利时电力运营商Elia公开的风电场实际运行数据,进行了仿真算例分析。所述方法也在中国西北部某省调度系统上应用于备用需求分析,并实现了试运行。Abstract: Estimating wind power forecast errors is very important for the dispatch and control as well as security and defense of wind power systems.Based on numerical feature extraction of historical data and day-ahead forecast data,a method of estimating wind power day-ahead forecast errors is proposed.Firstly,the main factors which influence the wind power forecast errors are analyzed.These include the fluctuation and amplitude of wind power,and the forecast method used.The correlations are calculated by data statistics.Then through analyzing the historical wind power data series,the model of wind power forecast error estimation is established using the multivariate linear regression method.Finally,the performance of the proposed method is verified using the wind power data from the Belgium grid’s system operator Elia.This method has also been tested in the dispatch system of a northwest province of China.All of the results demonstrate the feasibility and practicality of this method.
下载:
