高级检索

基于生态博弈的含云储能微电网多智能体协调优化调度

Multi-agent Coordination and Optimal Dispatch of Microgrid with CES Based on Ecological Game

  • 摘要: 分布式储能可以缓解分布式电源大量接入微电网所带来的随机性问题,但高昂的初装成本和运维困难限制了其大规模推广应用。在微电网中引入“云储能”为用户提供高效的“虚拟分布式储能”服务,基于自然界生态系统思想,提出了含云储能微电网多智能体生态博弈协调优化调度模型。根据利益诉求关系,构建了微电网系统多智能体结构,得到微电网运营商、常规负荷代理、云储能运营商以及云储能用户四大智能体,建立了其优化模型;构建了微电网电力生态系统,建立了各智能体之间以及电力生态系统之间的博弈优化模型;采用基于纳什均衡的强化学习算法对多智能体生态博弈模型进行求解。算例结果表明,云储能服务优化了负荷曲线、降低了用电成本、云储能运营商也获得了收益,达到多方共赢效果。

     

    Abstract: Distributed energy storage can alleviate the randomness problem caused by a large number of distributed power sources connected to micro-grid, but high initial installation cost and operation and maintenance difficulties limit its large-scale promotion and application. In this paper, “cloud energy storage” system is introduced to micro-grid to provide users with efficient “virtual distributed energy storage” services. Based on the idea of natural ecosystems, a multi-agent ecological game coordination optimization dispatching model for microgrids with CES is proposed. According to interest appealing relationship, the multi-agent structure of the microgrid system is constructed with four intelligent agents, including micro-grid operator, general load aggregators, cloud energy storage and cloud storage users, and their optimization models were developed respectively. The micro-grid power ecosystem was constructed, and the game optimization model among agents and among power ecosystems was established. The reinforcement learning algorithm based on the Nash equilibrium was used to solve the multi-agent ecological game model. The simulation results show that the cloud energy storage service optimizes the load curve, reduces the electricity cost, and cloud energy storage operators also gain benefits, achieving a multi-party win-win effect.

     

/

返回文章
返回