Abstract
Vehicle to grid (V2G) is an important way to enhance stability and reliability of the power grid and to promote renewable energy consumption. Optimal coordination of charging/discharging considering peak shaving and battery aging simultaneously is crucial to achieve whole-system social benefits. However, precise battery degradation, as a highly non-linear model, can hardly be embedded into mixed-integer linear programming. In this paper, a mixed integer linear programming (MILP) model based on hierarchical optimization is developed. The non-linear battery degradation model is linearized through ReLU-activated neural network and coupled with the V2G coordination MILP problem. First, the modeling of the behavioral boundary of the electric vehicle is obtained based on the Monte Carlo method. After that, the linearized replacement of the battery degradation model is realized by using segmented linear neural networks. Finally, a multi-objective optimization model of coordinated charging is developed for the load shifting and battery degradation problems. The simulation results show that compared with the optimization scenario with merely load regulation, the multi-objective algorithm considering battery degradation can effectively reduce the battery capacity decay by significantly decreasing intense EV charging/discharging profiles, while eliminating load fluctuations to a considerable extent.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call