Multiagent study of smart grid customers with neighborhood electricity trading

Abstract The smart grid of the future may equip customers with distributed generation and storage systems that can change their overall demand behavior. Indeed, the smart grid's infrastructure provides new opportunities for the grid and its customers to exchange information regarding real-time electricity rates and demand profiles. Here we report on innovative agent-based modeling and simulation of a smart grid where active customers are modeled as self-interested, autonomous agents with their own specific load profiles and generation/storage capacities. They may choose to use locally generated power, charge/discharge their batteries, and manipulate their loads. A unique scenario for the customers analyzed for this paper is one in which customers are allowed to trade electricity within their neighborhood in order to minimize their electricity costs. Meanwhile, the grid prefers an overall uniform demand from all customers. To achieve this, we propose an effective demand flattening management scheme for the customers. A model of the active customers within the smart grid environment is used to determine the impact of the neighborhood power transactions, demand diversity, and load shifting on the customers and the utility. A number of case studies and sensitivity analyses have determined how and to what extent these parameters affect customer electricity costs and power system metrics.