Overview

This presentation explored how hybrid quantum–classical computing can be applied to large-scale energy system optimization, with a focus on power systems and data center energy management. The talk highlighted the growing computational challenges posed by mixed-integer linear programming (MILP) models used in modern energy systems and proposed leveraging quantum optimization techniques—integrated with classical solvers—to improve solution efficiency, scalability, and robustness.

 
Expert Insights & Key Takeaways

Optimization bottlenecks in modern energy systems
Energy system operations—such as unit commitment, data center energy management, and power–hydrogen coordination—are commonly modeled using MILP formulations. While continuous subproblems are efficiently solvable, the integer master problem remains NP-hard and computationally expensive at scale.

Hybrid quantum–classical framework
The work integrates quantum optimization into a classical Benders decomposition framework:

• The quantum computer addresses the integer master problem by solving a QUBO (Quadratic Unconstrained Binary Optimization) formulation.
• The classical CPU solves the continuous subproblem and generates feasibility and optimality cuts.
• This division exploits the strengths of both platforms.

Performance robustness
Classical MILP solvers exhibited wide variance in solve time depending on instance characteristics, while the quantum-assisted approach showed tighter runtime distributions and greater consistency.
 

Future Outlook

Hybrid quantum–classical optimization presents a promising pathway for next-generation energy system management, particularly as systems grow in size, complexity, and coupling (e.g., power–hydrogen–data center integration). While near-term quantum hardware remains limited, strategic use of quantum solvers for combinatorial bottlenecks—paired with mature classical optimization—can already deliver measurable gains. As quantum hardware scales, these hybrid approaches could become a cornerstone for real-time, large-scale, and uncertainty-aware energy system optimization.

Guest Speakers

Lei Fan

Assistant Professor

Department of Engineering Technology

Read Bio