Challenge 3 - Quantum Mads

Quantum-Enhanced Optimization for Renewable Energy Systems

PROBLEM DESCRIPTION 

The growing integration of renewable energies, particularly wind power combined with ultracapacitor-based energy storage, presents significant control and optimization challenges to ensure both stability and efficiency. Although the Linear Quadratic Regulator(LQR) approach is widely used for optimal control, solving it at large scale can lead to high computational complexity. To explore new paradigms, we propose mapping the LQR problem to a Quadratic Unconstrained Binary Optimization (QUBO) formulation for solution on digital annealers (e.g., Fujitsu Digital Annealer). This challenge leverages direct methods of optimal control, inspired by the reference article [1]. The goal isto demonstrate how quantum computing
and annealing-basedmethods can enhance the speed and robustness of control decisionsin hybrid energy systems.

OBJECTIVES / EXPECTED OUTCOMES 

The expected outcomes of this challenge are: 

1. Quantum LQR Formulation

• Define the LQR problem (using indirect methods) in a manner suitable for translation into a QUBO model.
• Incorporate stability and feasibility constraints particular to a wind generation system with ultracapacitors.

2. Implementation and Validation on Digital Annealers

• Encode the QUBO representation in the Fujitsu Digital Annealer (or another annealer) and compare the results with open-license classical solvers.
• Assess computational time, solution accuracy, and sensitivity to variations in system dynamics (wind, load, etc.).

3. Demonstrations and Numerical Results

• Present implementation examples in different scenarios, highlighting benefits (or limitations) of the quantum-oriented approach.
• • Offer recommendations for scalability and possible extensions to other hybrid energy systems.

REFERENCES 

1. Majid A. Abdullah. et al. Linear quadratic regulator controllers for regulation of the dc-bus voltage in a hybrid energy system: Modeling, design and experimental validation. Sustainable Energy Technologies and Assessments, Volume 50, March 2022, 101880.

2. Pelofske, E., Hahn, G. & Djidjev, H. Solving Larger Maximum Clique Problems Using Parallel Quantum Annealing. Quantum Information Processing, Volume 22, article number 219, (2023).