Multi-objective energy management optimization in electric vehicles using fuzzy logic and particle swarm optimization

Abstract

This paper proposes a hybrid energy management system (EMS) for electric vehicles by integrating fuzzy logic control (FLC) with particle swarm optimization (PSO) to improve power-split decision-making under dynamic driving conditions. The FLC is designed using state of charge (SoC) and vehicle speed as input variables and power split as the output. A set of fuzzy rules defines the EMS behavior, while PSO is employed to fine-tune decisions by maximizing an efficiency objective function defined as the closeness of the power split to an ideal reference. The simulation is implemented in Python using Colab-compatible packages such as scikit-fuzzy, DEAP, and matplotlib, ensuring accessibility and reproducibility. A test grid covering 10 SoC levels (10-100%) and 10 speed levels (10-120 km/h) is used to evaluate the system. Visualization tools, including heatmaps, 3D surface plots, and contour plots, are employed to represent the EMS behavior. The PSO-enhanced system achieved a maximum efficiency of 98.2% at an optimized SoC of 61.7% and a speed of 53.6 km/h, outperforming standalone fuzzy logic control. Tabulated results and statistical summaries validate the effectiveness of the proposed system.