Multi-Objective Particle Swarm Optimization-Tuned Fuzzy Logic Control for Active Power Filter Integration in Electric Vehicle Charging Systems

Abstract:With the rapid expansion of electric vehicles (EVs) and their integration with power distributionnetwork, the power quality, harmonic distortion, and the control of reactive power are major signalling issues.A novel control approach, which integrates Fuzzy Logic Control (FLC) with Multi-Objective Particle SwarmOptimization (MOPSO) in maximizing the performance of Active Power Filters (APFs) for EV chargingsystems is presented in this paper. The framework covers the most critical power quality problems due to thenon-linearity of high power EV chargers considering grid stability, harmonics mitigation and the optimalcompensation of the reactive power. The fuzzy logic controller is arranged to dynamically regulate theswitching signals of the APF in response to the instantaneous state of the grid, including the load currentharmonics, the voltage fluctuations and changes in network impedance. MOPSO is utilized to further improvethe FRB and MFs by considering multiple competing objectives like THD, power factor correction and energyloss minimization. The optimization problem is designed to manage these trade-offs and still maintain solidcontroller performance under different load and grid conditions. For evaluating the proposed system in variousEV charging scenarios such as single and multiple fast charging cases, a full simulation platform is constructedin MATLAB/Simulink. Comparison results reveal that the FLC, optimized by the MOPSO, is superior to theconventional PI and heuristic controllers, with up to 45% THD reduction, 30% power factor improvement, andbetter dynamic response to load changes. Moreover, the system has strong resistance against grid disturbancesand it satisfies IEEE-519 requirements for harmonic distortion. This study demonstrates the effectiveness ofintelligent-optimization-driven fuzzy control to improve grid-vehicle interaction and promote smart and robustEV charging infrastructures. The proposed method can be easily expanded to renewable-integrated systems andvehicle-to-grid (V2G) configurations, and hence is a scalable solution for the smart grid of the future.