Research Of Optimal Energy Management Strategies For Microgrid Based On Multi-agent System

With the continuous consumption of fossil fuels, industrial progress of mankind hasbeen questioned in acquiring economic benefits. Energy shortage and environmentalpollution become the hottest researching fields asking for clean fuels and the applicationof distributed generation (DG) which is famous for the high-efficiency in powergeneration and minimum losses on the line transfer. A DG device can be located veryclose to the demand so as to make certain of the power reliability and system stability.However, DGs in large scale paralleling into the real power grid may have an impact onthe safety of power system. To solve this, scientists propose a new method to viewgeneration and associated loads as a subsystem which has become the focus of extensiveresearches what we called a Microgrid (MG). The concept of MG involves thecombination of DERs, storage devices and corresponding loads in a reduced space by anadvanced control system. It is designed either in online mode to switch power to the maingrid, or in island mode as an independent power supply when there is system error indisturbances. Although the MG occupies an important position in taking full advantage ofDG units and saves the loss of line, the increasing problem from emissions caused bypower stations also need to be focused on.In this work, the internal structure and operating mode of MAS based on marketingoperation are first introduced. Inside communications between Agents are established forthe bidding of the market players. The second task is to construct the bidding model andoptimize the energy management of a MG. The optimization is implemented by animproved particle swarm optimization (PSO) algorithm when fully considering thecharacteristics of distributed resources. The effectiveness of the proposed optimizationmethod is demonstrated by means of simulation results.