Domestic load scheduling using genetic algorithms

Abstract

An approach using a genetic algorithm to optimize the scheduling of domestic electric loads, according to technical and user-defined constraints and input signals, is presented and illustrative results are shown. The aim is minimizing the end-user's electricity bill according to his/her preferences, while accounting for the quality of the energy services provided. The constraints include the contracted power level, end-users' preferences concerning the admissible and/or preferable time periods for operation of each load, and the amount of available usable power in each period of time to account for variations in the (non-manageable) base load. The load scheduling is done for the next 36 hours assuming that a dynamic pricing structure is known in advance. The results obtained present a noticeable decrease of the electricity bill when compared to a reference case in which there is no automated scheduling. © Springer-Verlag Berlin Heidelberg 2013.