Benchmarking quantum computing for the wind farm layout optimization problem

We address the Wind Farm Layout Optimization (WFLO) problem using a Quadratic Unconstrained Binary Optimization (QUBO) formulation based on the Jensen wake model to account for wake-induced power losses. The goal is to maximize energy production through turbine placement. We compare three approaches: the classical Gurobi solver, D-Wave’s quantum annealer, and the Quantum Approximate Optimization Algorithm (QAOA) on a gate-based quantum computer, evaluating solution quality and computational time.

Gurobi delivers high-quality solutions with optimality guarantees but at a higher computational cost. D-Wave provides near-optimal solutions in seconds, making it ideal for rapid iterations, though it lacks optimality guarantees. QAOA, while flexible with tunable accuracy, remains constrained by current quantum hardware limitations but shows potential as technology advances.

Our results show that Gurobi excels on larger instances but is slower, while D-Wave’s speed is suited for quick feedback. Additionally, D-Wave’s approximate solutions can be used as a warm start for Gurobi, potentially speeding up convergence. QAOA holds promise for the future. This study highlights trade-offs between solution quality and time, showcasing quantum computing’s potential in large-scale energy system optimization.