Demand-supply balancing using multi-agent system for bus-oriented microgrids

To balance demand and supply in a microgrid can be considered as a demand-supply balancing problem, and it can be solved in either a static way or a dynamic way. In static algorithms knowledge about the system is known ahead of time. The connection of load and generator is made at the designing phase. A demand-supply balancing algorithm using the knowledge of system state at runtime has potentials of performing better than static algorithms. Demand-supply balancing problem is a large optimization problem with many variables for consumption or generation power flow at connected buses. We developed a market-based multi-agent system solving the problem in a decentralized way. Our approach is iterative with a fixed number of inter-agent communication steps. Consumer agents exchange simple messages with producers in the power network and then solve their own optimization problem by maximizing their own benefits. It has been shown that this approach converges to a solution when the agents´ objectives are achieved without a global coordinator. We have simulated different scaled bus-oriented microgrids and the simulation results show that the performance of the approach is promising.