Integrating Quantum and Traditional Computing in Optimization

Abstract

This proposal will integrate quantum annealing with traditional computing technology to develop new optimization techniques and algorithms. Unfulfilled promises of quantum computers quickly solving complicated optimization problems have left much of the optimization community skeptical of the potential for quantum computing. There have been many times when researchers use esoteric problems to display quantum dominance, only to have the results reversed when optimization experts examine those problems. Solving challenging instances of well studied problems will help alleviate this skepticism. A unique part of the proposed approach is that we do not expect the quantum annealer to directly prove optimality, something that quantum annealers are not built to do. Instead, we take advantage of quantum annealer’s ability to quickly generate high quality solutions. This ability gives optimizers the potential to, almost instantly, arrive at an inner approximation of a given model’s feasible region. Such knowledge can be very useful. This proposal will demonstrate how these inner approximations can be used to verify the impact different formulations. In particular, we will use a quantum annealer to approximate the feasible regions of various formulations. Then, we can choose the formulation that is most ideal for solving the instance using traditional computing, yielding significant computational speedups.

Document Details

Document Type
DoD Grant Award
Publication Date
Jan 14, 2022
Source ID
FA95501910147

Entities

People

  • James Ostrowski

Organizations

  • Air Force Office of Scientific Research
  • United States Air Force
  • University of Tennessee

Tags

Readers

  • Distributed Systems and Data Platform Development
  • Educational Psychology
  • Operations Research

Technology Areas

  • Quantum Computing