Prediction of the Wind Speed Series Based on Genetic Hysteretic Elman Network

A hysteretic Elman neural network is proposed to improve the prediction performance of the wind speed series. A hysteretic function which is constructed by shifting the Sigmoid function is used as the activation function of the neuron in the context layer. The historical information can be used in the process of choosing the activation response branch. Genetic algorithm is used to train the parameters in the hysteretic Elman neural network. The known data are used to construct the training samples. Multi-step prediction of the wind speed series are performed by the hysteretic Elman neural network. Simulation results show that the better prediction results can be got by the hysteretic Elman neural network and prediction errors can be reduced.