Comparative Metaheuristic Optimization of <scp>PI</scp> Controllers for Electric Vehicle Speed Control

Abstract

ABSTRACT This paper presents the design, modeling, and simulation validation of an electric vehicle system, using speed variation as a reference to ensure optimal operation in terms of acceleration and deceleration. First, a conventional PI controller controls the vehicle's speed. However, after a detailed analysis of the speed response, this controller is limited by a large initial overshoot, sensitivity to controller gains, and delayed response to disturbances. Therefore, adopting a more efficient controller, based on a robust design strategy using a metaheuristic algorithm, metaheuristic optimization, particle swarm optimization (PSO), and ant colony optimization (ACO), is essential to adjust the PI parameters. The ACO algorithm has several advantages, such as a simple structure, reduced control parameters, and easy implementation. It relies on indirect communication between ants to find the fastest path to their food source. The electric vehicle was simulated in the MATLAB/Sim Power System software environment. Simulation results show that the proposed controller achieves faster speed tracking, reduces overshoot, and improves robustness compared to conventional PI controllers and PSO‐tuned PI controllers. The results demonstrate that the PI‐ACO controller effectively enhances EV speed regulation and performance, such as statistical analysis, computational complexity, and computational cost.