Computational Forecasting of Power Prices Using Artificial Neural Networks

Abstract

 In the restructured power markets, the primary responsibility is setting the price of electricity. Thus, it has become increasingly important to accurately and precisely forecast power prices. An Artificial Neural Network (ANN) that was developed specifically for temporary price prediction in restructured electricity markets is presented in this paper. An input level, two hidden layers, and output layer comprise the four levels of the suggested ANN model, which is a perceptron neural network. Instead of using traditional back propagation for ANN training, using Levenberg-Marquardt retrogression (LMBP) methodology is used to accelerate convergence. The performance and efficacy of the suggested ANN model may be shown by training it on the Ontario power market. MATLAB is used to train the model.