Hybrid Classical-Quantum Combinatorial Optimization System

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current dedicated machines for solving combinatorial optimization problems using discrete binary variables face challenges in fixing all variables efficiently due to limited resources, particularly in quantum computers with a small number of qubits and couplings, leading to increased calculation time for large-scale problems.

Innovation Solution

A classical-quantum hybrid system is introduced, where a classical computer performs continuous optimization and extracts ambivalent variables, which are then solved by a quantum computer using binary optimization, optimizing resource usage and reducing computational time for large-scale combinatorial optimization problems.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Speed

If quantum computers are used to solve combinatorial optimization problems, then solution speed is improved, but resource limitations (number of qubits and couplings) prevent fixing all variables

Engineering Contradiction:
Improvesolution speedVSAvoidnumber of qubits
Core Design Contradiction:
SpeedVSQuantity of substance

Solution Approach 1:

The patent divides the set of variables into two segments: discrete variables (fixed to binary values) and continuous variables (not fixed). This segmentation allows the quantum computer to focus its limited qubit resources on the continuous variables while classical computers handle the discrete variables, resolving the contradiction between solution speed and qubit quantity limitations.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Instead of attempting to fix all variables (excessive action), the patent applies partial action by fixing only the discrete variables and leaving the continuous variables unfixed. This partial approach allows the quantum computer to operate within its resource constraints while still achieving accelerated solution performance for the critical continuous variables.

Inventive Principle:
Principle #16Partial or excessive action

2Measurement precision

If all variables are fixed to discrete binary values, then calculation accuracy is improved, but calculation time increases

Engineering Contradiction:
Improvecalculation accuracyVSAvoidcalculation time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent applies local quality by assigning different properties to different variables: discrete variables receive the property of being fixed to binary values (high precision requirement), while continuous variables are left unfixed (lower precision requirement). This localized differentiation maintains calculation accuracy for critical variables while reducing overall calculation time.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent changes the parameter state of variables from a uniform discrete binary state to a mixed state where some variables remain continuous. This parameter change allows the system to balance accuracy and time by maintaining discrete precision only where necessary while allowing continuous variables to be processed more rapidly.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP3859613A1Information processing system, combinatorial optimization method, and combinatorial optimization program
Publication Date: 2021.08.04 DENSO CORP
  • EP3859613A1 patent drawingFigure 1
  • EP3859613A1 patent drawingFigure 2A~2
  • EP3859613A1 patent drawingFigure 3

AI summary

An information processing system is used for solving combinatorial optimization problems for an objective function of a plurality of variables. The information processing system includes: two optimization systems that are a first optimization system and a second optimization system; and an extraction system. The first optimization system performs a first optimization process that allows, as continuous variables, the variables to continuously change in-between discrete values, and operates optimization and outputs evaluation which satisfies some restrictive conditions, using the continuous variables. The extraction system performs an extraction process that extracts variables, based on the continuous values of the first optimization system, and extracts, as ambivalent variables, the variables which cannot be decided to which discrete values should be taken. The second optimization system performs a second optimization process that solves the combinatorial optimization problem, based on the variables that are the ambivalent variables extracted in the extraction process.