Research on the Trading Strategy of VPP in the Power Elctricity Market with an Improved Proximal Policy Optimization Method

To enhance the flexible regulation capability of Virtual Power Plants(VPPs) and incentivize their participation in power electricity market operations, it is imperative to investigate how VPPs can rapidly respond to electricity demand in the market transactions. In this study, a bi-level trading model is constructed to represent market clearing and bidding games of VPPs. Moreover, a dynamic bidding strategy is proposed to maximize individual profit for the VPP during uncertain electricity market clearing process. Then, an improved proximal policy optimization (PPO) algorithm is employed for reinforcement learning training. During the training process, the VPP can dynamically adjust its bidding strategies in response to competitive market conditions. Finally, the simulation results demonstrate that the proposed strategy substantially improves VPP regulation revenues and facilitates economically efficient operation of the power electricity market.