Multi-objective Bi-directional V2G Behavior Optimization and Strategy Deployment

Abstract

The mitigation of peak-valley difference and power fluctuations are of great significance to the economy and stability of the power grid. The concept of the vehicle to grid (V2G) technology makes it possible to integrate electric vehicles (EVs) into the grid as distributed energy resources and provide power balancing service to the grid. This chapter introduces a V2G scheduling approach that can provide power balancing services to the grid while mitigating the battery aging phenomenon. Firstly, an intelligent V2G behaviour management framework is presented, which enables the comprehensive utilization of prediction information in V2G scheduling. Then, a commonly used multi-objective V2G behavior optimization model is introduced, in which minimal battery degradation and grid load fluctuation are the optimization objectives. Meanwhile, a multi-population collaborative mechanism, which is particularly designed for the V2G scheduling problem and has been proved effective in previous literature, is also introduced to improve the performance of the heuristic optimization-based V2G scheduling model. Two commonly used real-time strategy deployment methods: fuzzy logic and neural network, are further introduced for online V2G scheduling. With the presented methods, grid-connected electric vehicle (GEV) energy storage capacity can be scheduled to provide power balancing services to the grid while significantly mitigating battery aging.