Machine Learning-Driven Optimization of Power Electronics for Electric Vehicle Applications

EasyChair Preprint no. 12720

Abstract

Electric vehicles (EVs) are becoming increasingly prevalent as a sustainable transportation solution, necessitating efficient power electronics to manage energy conversion and distribution within these vehicles. This paper proposes a novel approach for optimizing power electronics in EVs using machine learning techniques. By harnessing the capabilities of machine learning algorithms, such as neural networks and genetic algorithms, we aim to enhance the performance, efficiency, and reliability of power electronics systems in EV applications. The methodology involves training neural networks to predict optimal configurations for power electronic components based on various input parameters, such as vehicle speed, battery state of charge, and environmental conditions. Additionally, genetic algorithms are employed to evolve and refine these configurations over time, adapting to changing operational requirements and environmental factors.

Keyphrases: electric vehicles, machine learning, neural networks, Optimization, Power Electronics

@Booklet{EasyChair:12720,
  author = {M Khan},
  title = {Machine Learning-Driven Optimization of Power Electronics for Electric Vehicle Applications},
  howpublished = {EasyChair Preprint no. 12720},

  year = {EasyChair, 2024}}