Deterministic prediction method using recurrent neural network optimized by genetic algorithm

In this paper, we propose the deterministic prediction method using recurrent neural network optimized by genetic algorithm Several methods with neural network have been proposed for identification of deterministic rule so far. However, with neural networks, there are two serious problems, one of which is that it is possible to fall into a local optimal solution, the other one is that there is no specific method to determine the structure of a neural network. To overcome these problems, we determine the optimal structure of the recurrent neural network using genetic algorithm. After optimization of structure of the recurrent neural network, we apply deterministic prediction method to short-term load forecasting. The suitability of the proposed approach is illustrated through an application to actual load data of Okinawa Electric Power Company in Japan.