Electricity price forecasting has gained a reputation for its importance in the deregulated energy market. The forecast process can be complicated as it depends on many elements. This paper proposes a hybrid of a neural network with a genetic algorithm for the electricity price forecasting. The Ontario energy market is select as the tested market for this model. The features for the neural network input are the actual historical demand and actual Hourly Ontario Energy Price (HOEP). The genetic algorithms help to select the number of features and to optimize the parameters of the neural network. This hybrid model helps to improve the accuracy of the forecasted price when comparing with the accuracy of the individual neural network itself. The mean absolute percentage error has represented the accuracy of the hybrid model, and it is used as a benchmark of the proposed hybrid model with other models.
Weights Optimization Based on Genetic Algorithm for Variable Weight Combination Model of BP-LSSVM for Day-ahead Electricity Price Forecasting
