摘要：AbstractThe Battery Energy Storage System (BESS) plays an important role in the smart grid and the ancillary market offers high revenues. It is reasonable for the owner of the BESS to maximise their profits by deciding how to bid with their rivals and balance between the different market offers. Therefore, this paper proposes an optimal bidding model of the BESS to maximise the total profit from the Automation Generation Control (AGC) market and the energy market, while taking the charging/discharging losses and the life of the BESS into consideration. Taking advantages of function approximation approaches, a reinforcement learning algorithm is introduced to the designed model, which can cope with the continuous and massive states of the proposed model and avoid the dimension curse. The resultant novel bidding model would help the BESS owners to decide their biddings and operational schedules profitably. Several case studies illustrate the effectiveness and validity of the proposed model.
关键词：KeywordsBattery Energy Storage System (BESS)optimal biddingreinforcement learning
