Enhanced Sensorless Current Control in Boost Converters for EV Fast Charging: Equilibrium Optimization-Tuned Cascaded ANN-Fuzzy Logic Control

Efficient current control in boost converters is critical for electric vehicle (EV) battery charging applications, where dynamic operating conditions challenge conventional control strategies. This paper presents a novel cascaded artificial neural network (ANN) and fuzzy logic (ANFIS) controller designed to enhance current regulation in boost converters. The performance of the proposed approach is validated through comparison with single PI, cascaded PI, and single ANFIS control schemes. To improve system performance, a sensorless current estimation technique is implemented, reducing hardware dependency and enhancing robustness, while lowering system cost. Experimental evaluations demonstrate that the proposed cascaded ANFIS controller maintains consistent current levels under rapid load changes, providing improved transient response and steady-state accuracy. Specifically, the proposed method achieves a settling time of 23 ms, compared to 185 ms for single PI and 66 ms for single ANFIS control, corresponding to an 87.6% improvement over PI. In addition, overshoot is reduced from 37.25% (PI) and approximately 20% (ANFIS) to negligible levels, while a fast rise time of 3.62 ms is obtained. These improvements reduce current deviation and transient stress, contributing to more stable and reliable charging operation. Although charging time, efficiency, and battery lifetime are not directly measured, the results indicate strong potential for improved practical charging performance. The proposed approach provides valuable insights into the integration of intelligent control methods in power electronics and supports the development of high-performance EV charging systems.