Swarm-Based Design of Dynamic Sliding Mode Control for Wireless Charging of Hybrid Energy Storage Systems

Abstract

The increasing demand for sustainable and intelligent energy solutions in electric vehicles (EVs) has led to a significant interest in the development of advanced hybrid energy storage systems (HESS) and efficient wireless charging architectures. In this work, a dynamic sliding mode control (DSMC) technique is optimized through a swarming heuristics framework for a battery-ultracapacitor HESS integrated with a wireless power transfer (WPT) system. Leveraging an LCC-S topology, the WPT system enables high-efficiency, contactless energy transfer to the storage modules under varying load and alignment conditions. To address the nonlinearities and parameter uncertainties inherent in such systems, a robust DSMC approach is formulated to ensure smooth system tracking and disturbance rejection. The control design is further refined using a bio-inspired moth–flame optimization algorithm hybridized with gravitational search and fractional-order PSO (MFOGSAPSO)—enhanced with adaptive entropy regulation and fractal-based memory—to dynamically tune the sliding-surface coefficients and switching gains. The proposed methodology is validated through comprehensive simulations in MATLAB/Simulink and a controller hardware-in-the-loop (C-HIL) setup on TI F28379D LaunchPads. Among the three MFO variants, MFOGSAPSO-A achieves the fastest objective function convergence, stabilizing near 685 within 10 iterations and substantially outperforming the optimized PID (715) and Optimized SMC (708). The proposed DSMC attains an overall RMSE of 0.1081, reducing the tracking error by 60.69% relative to PID and 14.84% relative to SMC, while shortening the settling time to 0.102 ms against PID (84.84%) and SMC (23.88%) improvements. The C-HIL results closely match the offline simulation waveforms without retuning, confirming superior energy management, improved power sharing between the battery and ultracapacitor, and enhanced overall efficiency of the wireless charging process under realistic embedded execution.