基于相似日聚类和IHGWO-WNN-AdaBoost模型的短期光伏功率预测

Short-term Photovoltaic Power Forecasting Based on Similar Weather Clustering and IHGWO-WNN-AdaBoost Model

为进一步提升光伏输出功率短期预测的准确性和稳定性,提出一种基于相似日聚类的小波神经网络(wavelet neural network,WNN)和AdaBoost的混合预测模型。首先利用模糊C均值聚类(fuzzy C-means algorithm,FCM)算法将初始数据集按照不同的季节和天气类型进行划分;其次选用WNN作为改进AdaBoost算法的基学习器,构建WNN-AdaBoost模型,并使用改进混合灰狼优化(improved hybridizing grey wolf optimization,IHGWO)算法优化WNN的小波因子和权值;最后选用我国中部地区某光伏电站实采的输出功率数据进行算例分析,通过与其他模型的对比,验证了所提模型的预测效果。实验结果表明:在不同季节和天气类型下,所提模型均能得到较好的预测结果,在有效提升光伏短期输出功率预测精度的同时,兼备了较强的适应性和鲁棒性。

To further improve the accuracy and stability of short-term forecasting of photovoltaic output power, an adaptive hybrid prediction model based on similar weather clustering, wavelet neural network(WNN) and AdaBoost is proposed. Firstly, the fuzzy C-means(FCM) algorithm is used to divide the initial data set according to different seasons and weather types. Secondly, WNN is selected as the base learner of the improved AdaBoost to build the WNN-AdaBoost model, and the improved hybridizing grey wolf optimization(IHGWO) algorithm is used to optimize the wavelet factor and weight of WNN. Finally, the actual measured output power data of a photovoltaic power station in the central region of China is selected to analyze the calculation example, and the prediction effect is verified by comparison with other models. The results show that the proposed model can obtain better prediction results under different weather types, and it has strong adaptability and robustness while effectively improving the accuracy of photovoltaic short-term output power prediction.