Wenjie Xu

Doctoral student in Electrical Engineering
École polytechnique fédérale de Lausanne (EPFL)
[Google Scholar] [ResearchGate] [LinkedIn]

Biography

I received my PhD in Electrical Engineering from EPFL as part of NCCR automation, under the supervision of Prof. Colin Jones and Dr. Bratislav Svetozarevic from Empa. Before that, I received my MPhil from the Department of Information Engineering, The Chinese University of Hong Kong, where I worked with Prof. Minghua Chen. Prior to that, I received my BEng in Electronic Engineering and (dual) BSc in Math, both from Tsinghua University in 2018.

Research Interests

My current research interest lies in Data/AI-driven control and optimization (particularly LLM-powered agentic AI), with applications to sustainable autonomous cyber-physical-human systems (CPHS), e.g., buildings.

Collaborations and Consultancy

Academic collaboration: I am generally open to academic collaborations. See my publications for our work. Feel free to contact me if you are interested in some discussions.

Industrial consultancy: I can give consultancy and tutorials on data science and optimization for energy systems (building systems in particular) and logistics systems. Please email me if you are interested.

Services

Reviewer for: IEEE Transactions on Automatic Control, IEEE/ACM Transactions on Networking, IEEE Transactions on Neural Networks and Learning Systems, Optimal Control Applications and Methods.

Reviewer for: NeurIPS, ICML, ICLR, IEEE Conference on Decision and Control, American Control Conference.

Contact

Lab: EPFL STI IGM LA3 ME C2 399 (Bâtiment ME) Station 9 CH - 1015 Lausanne
Email: wenjie.cuhk AT gmail.com, wenjie.xu AT epfl.ch

News

04/2026. Paper "What price to pay? Auto-tuning a building MPC controller for optimal economic cost" accepted to Control Engineering Practice.
03/2026. Paper "Smart building HVAC control challenge: experience and solutions from the ADRENALIN project" accepted to the Journal of Building Performance Simulation. (Our solution won the challenge!)
03/2026. Paper "Human-in-the-loop: Real-time Preference Optimization" accepted to 2026 European Control Conference (ECC).