With the increasing penetration of renewable energy, modern power systems face significant challenges in terms of uncertainty, stability, and operational efficiency. This project aims to explore advanced deep learning approaches—such as deep neural networks and deep reinforcement learning—to enhance the analysis and operation of power systems with high levels of renewable energy integration.

Key research topics include probabilistic power flow analysis, robust optimal power flow (OPF), and network topology optimization. By leveraging state-of-the-art computational tools, this project seeks to improve both the accuracy and efficiency of traditional power system analysis methods.

Participants will engage in hands-on numerical experiments using widely adopted platforms such as Optuna, MATPOWER, and MATLAB. Through these experiments, they will evaluate the performance of deep learning models and develop potential improvements for real-world applications.

This project is designed to:

．provide participants with a solid understanding of power system operation and analysis under renewable energy integration,

．develop practical skills in applying deep learning techniques to engineering problems, and

．enhance participants’ ability to interpret and synthesize experimental results.

Duration and Participation:

The project will host 1 participant and will run for 4 months, from September 2026 to December 2026.

Implementation Method:

The project will be conducted through laboratory-based research and numerical simulations, offering a hands-on and research-oriented learning experience.

Financial Support:

The participant will receive a monthly stipend of at least NTD 15,000 ( USD 500), with a maximum total funding of NTD 60,000 ( USD 2000) for the entire program.