Electric Load Analysis and Forecasting Using Artificial Neural Networks

Power grids around the world are experiencing a major shift from non-renewable sources of energy to renewable fuels. However, due to the nature of renewable sources of energy like wind and solar power, improper use could either lead to wastage or shortage of energy. It is thus beneficial for power producers to have prior knowledge of electricity consumption. In this paper, historical load trends and effects of ambient temperature on load are analyzed. Knowledge of this historical data combined with local temperature information is used to build a prediction model for short term and medium term load forecasting using neural networks. The prediction model performed well with a mean square error less than 1.