Collaborative Optimization Scheduling of Distribution Network and Microgrids Based on Multi Agent Deep Reinforcement Learning
In recent years,microgrids have played an increasingly important role in the construction of new type power systems.However,with the massive access of distributed flexible resources,the communication and computing tasks faced by the dispatch center are becoming increasingly heavy.Centralized scheduling strategies are difficult to effectively balance the interests and demands of multiple parties and ensure computational efficiency.The collaborative optimization scheduling strategy for microgrid is proposed based on multi-agent deep reinforcement learning algorithm.Firstly,aiming at the collaborative scheduling problem with multiple types of resource clusters,the model is set up with the objective function of minimizing the operating cost of microgrids;Secondly,a multi-agent deep reinforcement learning framework with centralized training and decentralized execution is designed to transform the collaborative scheduling problem into Markov game,the algorithm with entropy constraints is improved.During the strategy training phase,global state information is used as input,each agent makes optimization decisions through local observations after training;Finally,a case study is conducted based on the modified distribution network system to verify the effectiveness of the proposed method.
distribution networkmicrogridscollaborative optimization schedulingMADRL
万方数据
维普
