This repository is dedicated to the dissemination of the source code for a pioneering research project on the application of data-driven deep reinforcement learning (DRL) for the control of DC-DC buck converters feeding Constant Power Loads (CPLs).
This work aims to demonstrate the feasibility and efficiency of applying advanced reinforcement learning techniques directly to power electronics systems.
Our focus is on developing a controller that can adaptively manage the dynamic environment of DC-DC converters to ensure voltage stability under various operating conditions.
The methodology and experimental results presented in this repository are based on comprehensive studies detailed in the following papers:
Paper 1: Voltage Regulation of DC-DC Buck Converters Feeding CPLs via Deep Reinforcement Learning
Paper 2: Implementation of Transferring Reinforcement Learning for DC–DC Buck Converter Control via Duty Ratio Mapping
Paper 3: Robustness enhancement of DRL controller for DC–DC buck convertersfusing ESO
These papers provide the theoretical foundation and empirical evidence supporting the effectiveness of deep reinforcement learning in controlling DC-DC converters with constant power loads.
Matlab 2023b
- Intelligent and Autonomous Systems Laboratory (IAS Lab) at Shanghai University of Electric Power.