基于BP-AHP风机状态评估的超短期风电功率动态预测研究

Dynamic Prediction of Ultra-short-term Wind Power based on BP-AHP Wind Turbine Condition Assessment

针对传统风电功率预测仅考虑气象因素,且无法计及风电机组真实出力状态导致预测精度较差问题,本文提出一种计及风机状态的超短期风电功率动态预测方法。首先,为能够精确评估风机状态,将BP(error back propagation, BP)算法引入层次分析法(analytic hierarchy process, AHP)的评估结构中,构建BP-AHP风机状态评估模型,实现单台风机状态评估;然后,综合考虑地形及机组排布等因素,将风电场所有风机的状态取均值作为风电场状态,利用皮尔逊相关系数衡量所评估状态与功率之间的相关性以验证评估模型合理性,并采用XGBoost构建计及风机状态的动态预测模型;最后,以陕西地区某风电场实测数据进行算例分析,验证了所提方法的可行性及有效性。

To address the problem that the traditional wind power prediction only considers meteorological factors and cannot take into account the real output condition of the wind turbine,which leads to poor prediction accuracy,this paper proposes an ultra-short-term wind power dynamic prediction method considering the turbine condition.First,in order to accurately evaluate the turbine condition,the BP algorithm is introduced into the evaluation structure of analytic hierarchy process(AHP)to build a BP-AHP turbine condition evaluation model to realize the condition evaluation of a single turbine.Then,consider the terrain and turbine arrangement and other influencing factors,take the average value of the condition of all wind turbines in the wind farm as the wind farm condition,and the correlation between the assessed condition and power is measured using Pearson correlation coefficient to verify the reasonableness of the assessment model,and XGBoost is used to construct a dynamic prediction model of the turbine condition.Finally,an arithmetic analysis is carried out with actual measured data from a wind farm in Shaanxi to verify the feasibility and effectiveness of the proposed method.