Decision Tree Approaches for State of Charge Evaluation in Hybrid Electric Vehicles: A Comprehensive Study

This paper explores the application of decision tree frameworks for State of Charge (SoC) estimation in Hybrid Electric Vehicles (HEVs). The objective is to assess the accuracy and adaptability of decision trees in capturing SoC dynamics under dynamic driving conditions. The methodology involves a comprehensive review of existing studies, emphasizing feature selection, data preprocessing, and model training strategies. Key findings indicate that decision tree models offer superior accuracy but face challenges in limited dataset variability and model interpretability. Contributions of the paper include synthesizing knowledge on decision tree models in SoC estimation and proposing future directions for improvement. The potential impact of decision tree-based SoC estimation extends to increased efficiency, extended battery life, enhanced safety, and accelerated HEV adoption. The paper serves as a foundation for further research, encouraging innovative methodologies in the pursuit of accurate and adaptable SoC estimation for the next generation of HEV.