Advancing EV fast charging: Addressing power mismatches through P2P optimization and grid-EV impact analysis using dragonfly algorithm and reinforcement learning

In the wake of escalating interest in sustainable transportation, electric vehicles (EVs) have emerged as a pivotal solution. With their rising prominence, the need for expeditious and effective charging solutions has intensified. However, current paradigms fail to deal with fast charging system integration issues into the existing power grid. EV fast charging, power mismatch amelioration, and pioneering optimization are studied in the paper. In the case when the local grid's power capacity is not enough to support the total power demand of many fast-charging stations, power mismatch is addressed. Whereas the complexities of battery longevity and how fast they are all studied. As a means to equitably distribute load among the stations, we study the problem of peer-to-peer (P2P) charging localization. Using dynamic power allocation, we apply the Dragonfly algorithm to improve P2P localization and cope with power mismatch. At the same time, a new Deep Q Network (DQN) model modifies charging strategies upon real-time conditions of the grid. DA is empirically studied on different sizes and commonly used shapes of charging stations with different power mismatch levels, and it is demonstrated that the DA can considerably eliminate the power mismatches, optimize the charging station allocation, and enhance the grid resiliency. Additionally, the DQN model adapts to the fluid grid dynamics and improves charging efficiency. As a result of these innovations, a charging optimization framework beyond conventional charging optimization and a charging optimization scheme that is consistent with the grid infrastructure are provided. Through this new paradigm, we supply our contribution to sustainable energy transportation based on efficient EV charging, lowered grid stress, and maximal power utilization.