A Transfer Learning-Based Intelligent Voltage Control Algorithm

With the rapid advancement of digital technologies, the development of intelligent, adaptive, and autonomous algorithms has become crucial for decision-making and performance optimization under dynamic conditions. Such algorithms leverage data-driven strategies to learn from past experiences, enhance efficiency, and reduce reliance on precise system models. Transfer learning (TL), as an advanced paradigm in machine learning, enables models to reuse knowledge from previous tasks, leading to faster convergence, improved generalization, and reduced training requirements. In this study, we design and implement an enhanced TL-based algorithm integrated with metaheuristic optimization. The proposed method operates without requiring an exact system model and achieves faster, more efficient responses with limited data. To validate its effectiveness, the algorithm is applied to voltage control in BUCK converters—a long-standing challenge in control and power systems. Simulation results demonstrate that the TL-based framework combined with metaheuristic optimization offers a robust and scalable solution, outperforming conventional control methods in both accuracy and computational efficiency. These findings highlight the potential of the proposed approach for future applications in intelligent and high-precision control systems.

کلیدواژه ها:

Transfer Learning ، Intelligent Control ، DC-DC Converter ، Adaptive Model Tuning