Advanced Load Management Techniques with the Inclusion of Distributed Energy Resources in a Smart Grid

Abstract

Smart grid has been under continuous development since 2008. It requires the re-construction of traditional power systems. As an important component of a smart grid, load management has been diffusely known as a bright solution to enhance the demand side energy efficiency and optimize energy consumption. In this project, load management techniques on the demand side are studied at two levels in a smart grid: the smart home level and the load aggregator level. At the smart home level, this project studies the development of home energy management systems (HEMSs), which optimally schedule home energy resources to minimize home electricity costs. The potential of plug-in electric vehicles (EVs) is harnessed by the developed HEMS to supply power to the home and absorb surplus residential renewable power output. At the load aggregator level, this project studies the feasibility of aggregating thermostatically controlled loads (TCLs) in multiple buildings to bid in the wholesale power market. An optimal scheduling model for TCLs is proposed in this project to minimize imbalance costs of the load aggregators in the power market. To address the uncertainties in imbalance penalty prices, this project introduces the rolling horizon optimization (RHO) technique to continuously update TCL ON/OFF plans with the realization of imbalance prices. This research also includes techniques for solving load management optimization models. A new heuristic optimization method, Natural Aggregation Algorithm (NAA), is used to solve the aforementioned HEMSs and TCLs scheduling models. The encoding scheme and constraint handling strategies are studied, and the efficiency of NAA in solving the models is also investigated.