Convolutional Neural Network for Battery System Monitoring and SOC Estimation for EV Applications

Global power consumption is rising, necessitating the search for an alternative energy source to address the looming energy crisis, which is likely accomplished by grid-synchronized electric vehicles (EV). One of the intelligent grid's technological developments that enables energy transfer between EVs and the grid is called vehicle to grid (V2G) technologies. The EVs and grid plays a major role in power system leading to imbalances in frequencies, voltages, and supply-demand. This research introduces workable optimisation strategies to reduce these problems. Lithium-ion batteries are first charged via smart grid, and their State of Charge (SOC) is determined by Convolutional Neural Networks (CNN). In this study, the voltage is changed in both directions by a bidirectional DC-DC converter. With the help of PI, the calculated SOC is used to construct the gating sequence for the converter. When the SOC is below 80%, the battery gets power from the grid using the Buck method, and when the SOC is over 80%, the battery is discharged using the Boost method. Proportional plus Integral controllers are used to control both the active and reactive power. Through MATLAB simulation, the suggested tactic is verified.