Optimization of a refinery scheduling process with column generation and a quantum annealer

Abstract

This study focuses on the optimization of a refinery scheduling process with the help of an adiabatic quantum computer, and more concretely one of the quantum annealers developed by D-Wave Systems. We present an algorithm for finding a global optimal solution of a MILP that leans on a solver for QUBO problems, and apply it to various possible cases of refinery scheduling optimization. We analyze the inconveniences found during the whole process, whether due to the heuristic nature of D-Wave or the implications of reducing a MILP to QUBO, and present some experimental results.