风电场发电功率短期预测组合模型研究

Study on the combination model of short-term wind power generation prediction

随着风电机组装机容量的持续高速增加以及大规模风电场的建设,各个国家(地区)的电网对风电的重视程度也在增加,风电场发电功率的短期预测对于风电场并网以及电网的调度起着至关重要的作用。提出基于相空间重构理论RBF神经网络功率预测模型,通过判断功率时间序列的混沌属性,还原其规律性,以达到提高预测准确度的要求;结合时间序列模型,建立了组合预测模型。通过对结果进行对比分析,显示组合模型可以得到较高的短期发电功率预测准确度,更好地满足实际现场需要。

With the rapid increase of installed capacity of wind generation and the construction of large scale wind farms, the power grid of every country （area） pay more attention to wind power. The short-term prediction of wind power generation for wind integration and grid dispatching plays a vital role. A RBF neural network power prediction model is proposed in this paper based on the theory of phase space reconstruction. Firstly, by juding chaos attribute of power time series, restoring the regularity, the prediction accuracy is improved. Secondly, the combination forecast model is established by combining time series model. Finally, through comparative analysis of the results, combination model can get a better short-term power generation power prediction accuracy, and better meets the actual needs.