V2G Scheduling Scheme Based on Integral Reinforcement Learning Considering Battery Aging and Owner Benefits

As the penetration of renewable energy sources continues to increase, their intermittency poses new challenges to grid stability. In this context, vehicle-to-grid (V2G) technology is regarded as an effective means to cope with grid fluctuations. In this paper, we propose a V2G scheduling model that integrates grid frequency regulation (FR), battery aging and vehicle owner's revenue to achieve optimisation through a unified cost function. An integral reinforcement learning (IRL) algorithm with actor-critic neural network (NN) is employed to learn the optimal scheduling scheme under model-free conditions. Simulation results show that the scheme can significantly reduce grid frequency fluctuations, enhance owner participation through economic incentives, and effectively inhibit battery aging. In this paper, a practical solution is proposed to support the application of V2G technology in complex grid environments.