Multi-agent system for managing a power distribution system with Plug-in Hybrid Electrical vehicles in smart grid

Demand response programs play a much important role in smart grid. Plug-in Hybrid Electrical Vehicles (PHEV) have potential to increase the ability of residential customers to participate in demand response programs. Therefore, a new infrastructure that enables vehicle to participate in demand response programs is needed. Smart charging and discharging plans will lower the operational cost of electric vehicles and reduces peak system load. In order to identify the best solution, a decentralized Multi-Agent System (MAS), and a hybrid algorithm combined with an Evolutionary Algorithm (EA) and a Linear Programming (LP) were developed to manage a power distribution system with PHEVs in this paper. Outcomes of simulation studies show that a central scheduling system can obtain an optimal way of charging and discharging of PHEVs, but it is unfeasible in practice. Centralized methods require complete information on when and how much PHEVs are needed to charge and discharge beforehand. This is not available in practical systems. The multi-agent system approach proves that it is a scalable decentralized methodology which can adapt to incomplete and unpredictable information while the numerical outcomes are not much worse than that from a central scheduler.