基于灰色神经网络的微电网短期负荷预测

Short-term Load Prediction Based on Gray Neural Network in Microgrid Environment

综合考虑日期类型以及温度和湿度等天气因素对微电网负荷的影响,利用灰色理论处理随机性数据的优势以及神经网络的高度非线性,提出了一种基于灰色神经网络的微电网短期负荷预测方法。采用灰色模型、反向传播(BP)神经网络模型和灰色神经网络模型对2个微电网进行短期负荷预测。结果表明:灰色神经网络模型的预测结果精度更高,该方法可为微电网的经济可靠运行提供参考。

With a consideration of the effects of date type and temperature and humidity on microgrid load,using the advantages of gray theory in random data and the high nonlinearity of neural network,a short-term prediction model of microgrid load based on gray neural network was proposed.Then,the loads of two microgrids were predicted with gray model,back propagation（BP） neural network model and gray neural network model.The simulation results showthat the proposed method is accurate and provides a practical reference for the economical operation of microgrid.