基于爬坡特征识别的短期风电功率集成预测方法

Research on Integrated Short Term Wind Power Prediction Methods Based on Climbing Feature Recognition

大规模风电并网给电力系统稳定运行带来了极大挑战,准确预测风电功率可以促进消纳、保证系统可靠运行;而风电不确定性及功率突变的爬坡现象会导致预测精度不足的问题。提出一种考虑功率爬坡特征识别的短期风电功率集成预测方法。基于风电功率历史数据,利用爬坡定义将功率序列划分为上爬坡、下爬坡和非爬坡;考虑不同爬坡时序特点,选择适应其特征的预测方法,对上爬坡、下爬坡、非爬坡阶段分别使用SSA-BiLSTM模型、CNN-BiLSTM模型、LSTM模型进行预测;集成各段预测结果,得到最终的短期风电功率预测结果。算例验证结果表明:相比于传统的预测模型,所提集成预测方法能有效提升爬坡事件下风电功率预测的准确性,大大提高了短期风电功率预测的整体精度。

Large-scale wind power grid integration will bring great challenges to the normal operation of the power system.Accurate wind power prediction can improve the level of wind power consumption and ensure the normal operation of the power system.To address an insufficient prediction accuracy in sudden changes in wind power,a short-term wind power integrated prediction method based on power ramp feature identification is proposed in this paper.First,based on the historical data of wind power,the power sequence is divided into uphill climb,downhill climb and non-climb by using the climbing definition.Second,considering the characteristics of different climbing sections,the prediction method adapted to its characteristics is selected:SSA-BiLSTM model,CNN-BiLSTM model and LSTM model are used to predict the upper,lower and non-climbing stages,respectively.Finally,the prediction results of each segment are linearly superimposed to obtain the prediction results of wind power in the entire period.Taking the power sequence provided by a wind farm in Belgium as an example for verification,the results show that,compared with the traditional prediction model,the prediction method proposed can improve the accuracy and precision of wind power prediction effectively.