Electricity Demand Forecasting Using Neural Networks

摘要

This paper introduces a methodology to forecast electricity consumption. A Multi-Version System (MVS) methodology combines different data mining methods to compensate weaknesses of each individual method and to improve the overall performance of the system. The current benchmark forecasting system in use is a Regression system, which is in need of improvement after the structural as well as operational changes in the electricity supply markets. Experiments are Modelled on the Regression and the prototype Network system. The results indicate that, in some cases, the Neural Network Model has performed better than the Regression System.