基于改进的VPSO-ELman神经网络的短期负荷预测

Short Term Load Forecasting Based on Improved VPSO-ELman Neural Network

提出一种基于改进的VPSO-Elman神经网络的短期负荷预测方法,在经典的Elman动态网络中引入VPSO算法对网络的训练过程做进一步的优化,提高Elman网络的泛化能力和适应能力,使预测模型的精度提高、鲁棒性变强。利用辽宁某地区的历史负荷数据作为训练样本,通过与经典的Elman预测模型进行对比,对预测结果与实际数据进行比较,得出该方法辨识能力优于经典网络,泛化误差明显小于经典网络。

A short-term load forecasting method based on improved VPSO-Elman neural network is proposed.The VPSO algorithm is introduced into the classical Elman dynamic network to further optimize the training process of the network and to improve the generalization ability and adaptability of the Elman network.The accuracy of the prediction model is improved and the robustness becomes stronger.Using the historical load data of a certain region in Liaoning as a training sample,the comparison between the predicted results and the actual data is carried out by comparing with the classical Elman prediction model.It is concluded that the method is superior to the classical network and the generalization error is obviously smaller than the conclusion of the classical network.