To enhance the accuracy of microgrid load forecasting,a method is proposed that utilizes grey correlation analysis to optimize the input layer data of the BP neural network,employs a novel multidimen-sional universe algorithm to optimize the hidden layer weights,and utilizes an imperial competitive algorithm to optimize the output layer results in a layer-by-layer optimization model.The method is applied to the analysis of two sets of actual measurement data.The results indicate that the proposed GCA-MVO-ICA opti-mization method for the BP network can improve the prediction accuracy of microgrid load and exhibits good universality.
关键词
BP神经网络/灰色关联度/多元宇宙算法/帝国竞争算法/负荷预测
Key words
BP neural network/gray correlation/multiverse algorithm/empire competition algorithm/load forecasting
维普
万方数据