Research Projects

A summary of my research projects in reverse chronological order.

Risk-aware game framework for coordinating microgrids in the energy and reserve markets

Prof. Hobbs’ group, Whiting School of Engineering, Johns Hopkins University (JHU) April 2022 – October 2022

• Developed deep learning neural networks for the weather-informed probabilistic predictions of net loads.
• Evaluated different risk-aware formulations considering uncertain reserve service provisions from renewable microgrids.
• Modeled players’ strategic bidding and offerings in the power markets using the game-theoretical framework.
• The research has been submitted to journal and conference.

Distributionally robust joint chance-constrained optimization for networked microgrids

Energy and Power Group (Led by Prof. McCulloch) & Power Electronics Group (Led by Prof. Rogers), University of Oxford

• The project is supported by EPSRC project, Robust Extra low-cost solar nano-grid (RELCON), to develop microgrids for energy-poor communities.
• Modeled the power flow in the networked microgrids considering droop control and smart load shedding.
• Constructed a distributionally robust joint chance-constrained optimization framework for controlling networked microgrids considering solar generation uncertainty and utility contingencies.
• Developed an easy and fast solving method for the NP-hard optimal Bonferroni approximation.
• The research is published at IEEE Smart Grids and CDC conference.

Green Radio: Dynamic power saving configuration for mobile networks

The Alan Turing Institute, London Sep 2019

• The project collaborator is one of the world’s largest mobile telecommunications companies, Telenor.
• Forecasted day-ahead demands of 200 signal towers using five state-of-the-art ML algorithms.
• Designed the activation policies for recommending day-ahead power-saving schemes.
• Clustered sectors of signal towers for improving scalability and summarized implications of reinforcement learning (RL) in future works.
• The research is published as a co-authored report on the Turing website page.