Energy Management System of a Zero Energy Building using Multi-Agent Systems

Abstract

Rising trend towards urbanization has resulted in the emergence of urban buildings as major energy consumers. This has increased the burden on the grid which may eventually result into severely interrupted power supply and outages. In order to assist the over-burdened grid, the concept of zero energy buildings (ZEBs), which consist of individual generation systems, has been developed. Traditionally, ZEBs have a negligible net energy consumption when accounted on an annual basis. However, this concept of annual energy accounting does not reduce the building’s dependency on the grid. Thus, in order to minimize their grid-dependency, this work makes ZEBs self-sufficient by incorporating instantaneous power management techniques. Accordingly, a power management system has been designed which controls the instantaneous flow of generated power in the ZEB depending upon the load variations. This involves decentralized control of dispersed parameters which has been achieved through the technique of multi-agent system (MAS). For the purpose of simulations, agent models have been developed using JAVA Agent Development Framework (JADE). Extensive simulations have been performed and results have been obtained for optimal power management and load scheduling using Genetic algorithm (GA).