Design of an Adaptive Power Balancing Model with Energy Recovery & Powertrain Control via Fuzzy Bio-inspired Optimizations

Abstract

This paper presents a comprehensive approach for optimizing power balancing, energy recovery, and powertrain control in hybrid vehicles using adaptive algorithms and fuzzy logic-based optimization techniques. The approach combines three internal models to address limitations in deep learning techniques. The first is an Adaptive Power Splitting Algorithm for Battery Degradation Mitigation with Elephant Herding Optimization, which considers temperature, charge level, and historical usage patterns. The second is an Energy Recovery Algorithm for Regenerative Braking Optimization based on Genetic Algorithms, which optimizes energy recovery during braking. The third is a powertrain control algorithm based on fuzzy logic, considering driver preferences, traffic conditions, and vehicle speed. The results show significant improvements over current deep learning techniques.We performed extensive experiments to assess the performance of our suggested model using data from the National Renewable Energy Laboratory (NREL) Vehicle Testing and Integration Database (VTID), the University of California's Hybrid Vehicle Dataset, and the U.S. Environmental Protection Agency's (EPA) Fuel Economy Dataset. Our findings show notable advancements over current deep learning techniques, including an 8.5% increase in fuel efficiency, a 10.4% increase in energy recovery efficiency, a 4.5% decrease in emissions, and a 3.5% increase in cost efficiency levels.