Superiority of quantum-generated starting solutions in HQTS

Determine whether initializing the Hybrid Quantum Tabu Search (HQTS) algorithm for the Capacitated Vehicle Routing Problem with a quantum computing–derived starting solution is uniformly superior to initializing with a classical heuristic starting solution, such as the Clarke–Wright savings heuristic, across problem instances.

Background

The paper introduces a Hybrid Quantum Tabu Search (HQTS) algorithm for the Capacitated Vehicle Routing Problem that integrates quantum annealing–based routing within a classical tabu search framework. The authors systematically study the impact of different starting solution strategies, including classical heuristics (e.g., Clarke–Wright) and quantum-generated initializations based on prior hybrid methods.

Across experiments on the CMT benchmark set, the authors observe that while quantum components generally improve performance when used more frequently, the relative advantage of quantum-generated starting solutions over classical heuristic starting solutions is not established. This leaves unresolved whether a quantum-derived initialization consistently provides a measurable advantage for HQTS.