Paper Title
Prediction of Power Supply Demand: Consumption Pattern Modeling

Abstract
This work is devoted to the problem of the local electricity supply management in the developing towns and areas. We propose a methodology that enables to predict the power demand by households. The method bases on the combination of market data, statistical information about local citizens’ behavior and agent based modelling. As an outcome, we obtain an hourly electricity demand that enables us to evaluate the necessary level of the power supply during a day and that enables us to predict about the further dynamics of the electricity demand and to adapt the power supply system to further changes beforehand. In the given work such simulation details as the size of the residential area, agents’ behavior and residence specification are taken from a real data comprised by the household survey from a developing town Shinchi, Japan. Index Terms - Consumption Pattering, Demand Prediction, Power Supply.