improving short-term wind power prediction with neural network and ica algorithm and input feature selection

according to this fact that wind is now a part of global energy portfolio and due to unreliable and discontinuous production of wind energy; prediction of wind power value is proposed as a main necessity. in recent years, various methods have been proposed for wind power prediction. in this paper the prediction structure involves feature selection and use of artificial neural network (ann). in this paper, feature selection tool is applied in filtering of inappropriate and irrelevant inputs of neural network and is performed on the biases of mutual information. after determining appropriate inputs, the wind power value for the next 24-hours is predicted using neural network in which bp algorithm and pso and ica evolutionary algorithms are used as training algorithm. with investigation and compare numerical results, better performance of pso and ica evolutionary algorithm is deduced with respect to bp algorithm. more accurate survey will result in more proper efficiency of imperialist competitive algorithm (ica) in comparison to swarm particle algorithm. thus, in this paper; accuracy of the wind power prediction for the next 24-hours is improved considerably using mutual information and providing an irrelevancy filter for reducing the input dimension by eliminating the irrelevant candidates and more effectively using imperialist competitive evolutionary algorithm for training the neural network.