IMPLEMENTING A WEIGHTED LEAST-SQUARES PROCEDURE IN-TRAINING A NEURAL-NETWORK TO SOLVE THE SHORT-TERM LOAD FORECASTING PROBLEM

The use of a weighted least squares procedure when training a neural n etwork to solve the short-term load forecasting (STLF) problem is inve stigated. Our results indicate that a neural network that implements t he weighted least squares procedure outperforms a neural network that implements the least squares procedure during the on-peak period for t he two performance criteria specified; MAE% and COST, and during the e ntire period for the COST criterion. It is, therefore, recommended tha t the weighted least squares procedure be further studied by electric utilities which use neural networks to forecast their short-term load, and experience large variabilities in their hourly marginal energy co sts during a 24-hour period.