基于组合模型预测短期风电功率

Prediction of short-term wind power based on combination model

为提高风电功率短期预测精度,提出一种基于时间序列与天牛须搜索算法改进BP神经网络的组合预测模型。首先,分别利用时间序列法与天牛须搜索算法改进的BP神经网络得到两个风电功率预测模型;其次,根据误差平方和最小的原则利用粒子群算法寻找两个模型的权重系数,建立时间序列与改进BP神经网络的组合预测模型;最后,以某风电场的实测数据为基础,进行数据分析、预测与对比。仿真结果表明,文中所提出的组合预测模型比单一预测模型的预测精度更高。

In order to improve the short-term prediction accuracy of wind power,a combined prediction model is proposed based on time series and BP neural network improved by beetle antennae search algorithm.Firstly,two wind power forecasting models are obtained by using time series method and BP neural network improved by beetle antennae search algorithm.Secondly,according to the principle of minimum square error,the weight coefficients of the two models are gained by particle swarm optimization algorithm.The combined forecasting model of time series and improved BP neural network are established.Finally,Data analysis,prediction and comparison are carried out,based on the data measured from a wind farm.The simulation result shows that the combined prediction model proposed in this paper has higher prediction accuracy than the single prediction models.