Adaptive Velocity PSO-Based Parameter Optimization for a Permanent Magnet DC Motor Drive in Light Electric Vehicles

This study proposes an adaptive velocity particle swarm optimization (AVPSO) technique for tuning the proportional-integral (PI) controllers of a permanent magnet DC (PMDC) motor drive system in light electric vehicles (LEVs). The AVPSO algorithm adaptively adjusts the inertia weight and acceleration coefficients to improve convergence and balance between exploration and exploitation during the optimization process. A MATLAB/Simulink model is used for offline evaluation of the cost function, which penalizes steady-state and transient errors in angular velocity and armature current. Simulation results show that the AVPSO-based tuning achieves faster settling times reduced by 15.6% in forward and 2.1% in reverse operation while eliminating steady-state errors observed in classical PSO. Performance comparisons under four-quadrant operation and torque disturbance scenarios confirm the method’s effectiveness in improving control accuracy, reducing power losses, and maintaining system stability. These results demonstrate the potential of AVPSO as a robust and efficient tool for optimizing motor drive performance in LEV applications. Overall, the AVPSO-based optimization provides a robust and efficient solution for optimizing motor drive systems in LEVs.