Wind speed prediction based on chaotic operator network

A novel prediction approach based on chaotic operator network is proposed to predict wind speed series. Each chaotic operator has complex dynamic characteristic. When lots of chaotic operators are connected into a prediction network, the dynamic behavior of the network will depend on any tiny variations of each chaotic operator parameter. Thereafter, training samples are constructed by the known time series. Genetic algorithm is adopted to train the parameters of the prediction network by the training samples. In this way, the dynamic behavior of the prediction network can be changed to follow that of the real wind speed system, and good prediction results can be got. Simulation results prove the method valid.