Agent-based multi-time-scale plug load energy management in commercial building

Abstract

Building energy management (BEM) system is a critical component to manage energy demand. The future vision of BEM is to promote the integration of renewable energy and demand response by exploiting smart grid technologies. However, it will introduce the uncertainty and randomness significantly both in energy supply and demand, which poses great challenge in scheduling the electricity procurement and load control efficiently. To remedy the critical issues, we present an agent-based BEM system framework, in which the objective of electricity procurement is decided and implemented by central control agent (CCA), and plug load control is separated to zone control agent (ZCA). This framework is designed to jointly optimize the electricity procurement and plug load control based on the collaboration of different agents. More specifically, we propose a low-complexity and effective distributed algorithm based on dual decomposition for electricity procurement. Zone level energy balance and plug load control are given by quadratic programming, 0-1 programming and heuristic algorithm, which can handle with sufficient speed and computational efficiency to make them practical for control and reduce information interaction. The simulation results show that our proposed agent-based multi-timescale BEM system can effectively address the randomness and uncertainties of renewable energy resources and plug load, and has the potential to reduce the electricity consumption in the building.