4. Machine Learning-Based Grid-Interactive Chargers for Optimizing Power Quality Under Non-grid Conditions

This chapter delves into the application of machine learning techniques to optimize power quality in grid-interactive electric vehicle chargers. It begins with an overview of the bi-directional power flow between the grid and electric vehicles, highlighting the benefits of Vehicle-to-Grid (V2G) technology. The article then explores the system configuration of an On-Board Bi-Directional Charger (OBBC), detailing its components and their roles. A significant focus is placed on the design of controllers, particularly the use of machine learning-based algorithms for DC-link voltage control. The chapter discusses various control strategies, including the Cascaded Non-Identical Second-Order Generalized Integrator (CNISOGI) filter and decision tree algorithms, and their effectiveness in managing power flow and maintaining grid stability. The results and discussion section presents simulation outcomes, demonstrating the performance of the proposed control strategies under different grid conditions. The conclusion emphasizes the superior performance of decision tree regressors in capturing complex data patterns and improving grid stability. Overall, the chapter provides a comprehensive guide to enhancing the efficiency and reliability of electric vehicle charging systems through advanced machine learning techniques.