Deep Reinforcement Learning-Based Control Strategy with Direct Gate Control for Buck Converters

TL;DR

This paper introduces a deep reinforcement learning-based control method that directly generates gate signals for buck converters, achieving faster response, improved stability, and robustness compared to traditional control schemes.

Contribution

It presents a novel DRL-based direct gate control approach for buck converters, enhancing control speed, stability, and robustness over conventional methods.

Findings

Faster transient response than PWM control

Stable output voltage regulation under parameter variations

Robustness against sensor noise

Abstract

This paper proposes a deep reinforcement learning (DRL)-based approach for directly controlling the gate signals of switching devices to achieve voltage regulation in a buck converter. Unlike conventional control methods, the proposed method directly generates gate signals using a neural network trained through DRL, with the objective of achieving high control speed and flexibility while maintaining stability. Simulation results demonstrate that the proposed direct gate control (DGC) method achieves a faster transient response and stable output voltage regulation, outperforming traditional PWM-based control schemes. The DGC method also exhibits strong robustness against parameter variations and sensor noise, indicating its suitability for practical power electronics applications. The effectiveness of the proposed approach is validated via simulation.