Electric vehicles (EVs) are finding a hopeful home in contemporary grids thanks to Vehicle-to-Grid (V2G) technology. This work explains the scheme, functionalities, and possible advantages of V2G while also illustrating how it might be used on a microgrid. Explain how electric cars can act as a turning replacement by holding additional grid power stored in their batteries and reintroducing it to the system when needed. This aids in controlling the frequency of the grid within its nominal specifications. EVs are essential for achieving amazing features; however, integrating a large number of EVs also causes grid congestion. Uncontrolled charging could therefore result in under voltage and intricate grid congestion. To lessen the impacts, the causes of EVs charging uncontrollably have recently been examined. Because of various limitations on the part of the customer, it is very difficult to establish controlled charging; as a result, it is preferable to utilize power prediction systems for EV charging and discharging. This study’s main goal is to use a deep learning network to create an efficient and effective system for EV charging and discharging. Within this framework, a neural network-based model with a feed-forward algorithm is suggested to examine the effects of bi-directional charging within the microgrid setting.