Multi-Agent Optimal Allocation of Energy Storage Systems in Distribution Systems

Abstract

A variety of optimal methods for the allocation of a battery energy storage system (BESS) have been proposed for a distribution company (DISCO) to mitigate the transaction risk in a power market. All the distributed devices are assumed to be owned by the DISCO. However, in future power systems, more parties in a distribution system will have incentives to integrate BESS to reduce operational cost. In this paper, an enhanced BESS optimal allocation method is proposed for multiple agents in a distribution system. First, the electricity market mechanism is extended to a distribution system, and the corresponding energy transaction process is modeled for different agents, such as wind farms, solar power stations, demand aggregators, and the DISCO. The uncertainties of renewable energy and demand are addressed using stochastic methods. In the proposed transaction model, the integration of BESS can help an agent to reduce the operational cost, also defined as the payoff function. Next, game theory is introduced in this paper to investigate the interactions among the agents and to determine the BESS integration plans. The agents are built as players who are willing to minimize their payoff functions in the proposed non-cooperative game. The Nash equilibrium, which is the best strategy for the players, is proved to exist. Such equilibrium can be solved using an iterative algorithm. The proposed BESS allocation method for the multi-agent system is verified for two cases, and the payoff reductions are quantified based on the proposed distribution energy transaction mechanism.