A Robust Artificial Intelligence-Empowered Adaptive Proportional-Integral Control for Wireless Power Transfer Systems

Coupling variations caused by misalignment between the transmitters and receivers threaten the stability, reliability, and efficiency of wireless power transfer (WPT) systems. By far, various hardware and software strategies have been proposed to address the misalignment issues. Compared to the hardware-based coil design schemes, software-based adaptive control schemes are more cost-effective since only algorithms in digital controllers are changed. However, conventional adaptive control, including the adaptive proportional-integral (PI) control, may suffer from overfitting to specific operating conditions. To this end, an Artificial Intelligence (AI)-empowered design (AI-D) approach, leveraging machine learning and meta-heuristic algorithms, is proposed in this paper to enhance the robustness of adaptive PI control against coupling variations of WPT systems. Optimal design cases tailoring for diverse dynamic performances are provided, which are validated with comprehensive scenario analysis.