Temperature Scheduling for Energy-Based Optimization Search

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Determining appropriate temperature values for methods like simulated annealing and replica exchange is challenging in solving optimization problems.

Innovation Solution

An information processing apparatus and method that acquires and adjusts temperature values by monitoring energy function updates during the search process, determining suitable temperature values based on whether the energy value is smaller than the previous smallest value, and outputting new temperature ranges for more efficient searches.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If temperature values are set arbitrarily for simulated annealing or replica exchange methods, then the optimization search can proceed, but the efficiency of finding the ground state deteriorates due to inappropriate temperature selection

Engineering Contradiction:
Improveoptimization search efficiencyVSAvoidtime for temperature value determination
Core Design Contradiction:
ProductivityVSLoss of time

Solution Approach 1:

The system automatically determines appropriate temperature values by having the search unit execute preliminary searches at candidate temperature values and having the determination unit select optimal temperatures based on the energy function values obtained, eliminating the need for manual temperature value determination while optimizing search efficiency

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The system performs preliminary searches at multiple candidate temperature values before the main optimization process to evaluate which temperature values are most suitable, allowing the determination of optimal temperature parameters in advance to guide the subsequent efficient search

Inventive Principle:
Principle #10Preliminary action

2Measurement precision

If a wide range of temperature values is tested to find suitable values, then the accuracy of temperature selection improves, but the computational overhead and time consumption increase

Engineering Contradiction:
Improvetemperature value selection accuracyVSAvoidcomplexity of temperature determination process
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system changes the temperature parameter systematically by testing a plurality of candidate temperature values and selecting those that yield appropriate energy function values, allowing precise temperature selection through controlled parameter variation rather than exhaustive searching

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The determination unit uses feedback from the search unit's results at different temperature values to identify which temperature values are most suitable, creating a closed-loop system where temperature selection is refined based on actual search performance rather than arbitrary selection

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP3905073B1Information processing apparatus, information processing method, and program
Publication Date: 2026.04.22 FUJITSU LTD
  • EP3905073B1 patent drawingFigure 1
  • EP3905073B1 patent drawingFigure 2
  • EP3905073B1 patent drawingFigure 3

AI summary

A method includes: acquiring, from a search node configured to perform a search for a ground state represented by plural state variables included in an energy function by using plural temperature values and hold a value of the energy function for the plural state variables, a value of the energy function obtained for the plural state variables at a first temperature value among the plural temperature values; determining whether the value acquired is smaller than a smallest value of the energy function obtained for the plural state variables before reaching the first temperature value; recording update information indicating that the smallest value has been updated at the first temperature value in a case where the value is smaller than the smallest value; and outputting a second temperature value based on the first temperature value at which the update information has been recorded among the plural temperature values.