Ising Model Optimization Apparatus Temperature Setting Method
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing optimization apparatuses face challenges in setting an optimal highest temperature for solving multivariate optimization problems, as excessively high temperatures lead to frequent state transitions while excessively low temperatures hinder escape from local solutions, making it difficult to converge to an optimum solution.
Innovation Solution
The optimization apparatus employs a temperature setting method that calculates a maximum temperature value (Tmax) based on continuous energy increases, allowing for a higher probability of escaping local solutions by setting Tmax=-Dsum*log(A), where Dsum is the total energy increase and A is the transition acceptance probability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a high temperature is set to escape local solutions, then the probability of escaping local solutions increases, but the system undergoes frequent state transitions which reduces convergence efficiency
Solution Approach 1:
The patent dynamically adjusts the temperature parameter based on the energy landscape characteristics of the specific optimization problem. By calculating the maximum energy increase (Dsum) required to escape local minima and using it to determine an optimal maximum temperature (Tmax), the system adapts the temperature parameter to match the problem's energy barriers, thereby balancing exploration capability with convergence efficiency.
Solution Approach 2:
The patent performs preliminary analysis of the energy landscape by calculating energy increases for bit flips between different solutions before setting the temperature. This preliminary calculation of Dsum allows the system to pre-determine an appropriate Tmax that is tailored to the specific problem structure, avoiding both excessive temperature (which causes frequent transitions) and insufficient temperature (which fails to escape local minima).
2Productivity
If a low temperature is set to reduce state transitions, then convergence efficiency improves, but the system cannot escape local solutions
Solution Approach 1:
The patent dynamically adjusts the temperature parameter based on the energy landscape characteristics of the specific optimization problem. By calculating the maximum energy increase (Dsum) required to escape local minima and using it to determine an optimal maximum temperature (Tmax), the system adapts the temperature parameter to match the problem's energy barriers, thereby balancing exploration capability with convergence efficiency.
3Ease of operation
If a fixed temperature schedule is used, then the optimization process is simple to implement, but it cannot adapt to problems with different energy landscape characteristics
Solution Approach 1:
The patent automatically adapts the temperature parameter to match the specific energy landscape of each optimization problem by calculating Dsum (the maximum energy increase required to escape local minima) and deriving Tmax from it. This eliminates the need for manual temperature scheduling while providing problem-specific optimization, thereby achieving both ease of implementation and adaptability.
Solution Approach 2:
The system performs self-adjustment by automatically analyzing its own energy landscape and determining the appropriate temperature parameters without external intervention. The calculation of Dsum and subsequent determination of Tmax is performed autonomously by the optimization apparatus, allowing it to adapt to different problems without requiring complex external control or manual tuning.
Data Source
AI summary
An optimization apparatus includes a processor. The processor configured to execute search for a ground state of energy in an Ising model by using a plurality of temperature values, select a first solution and a second solution from a plurality of solutions, sequentially perform processing of calculating a change of the energy in the Ising model caused when one value of a plurality of bits of a bit sequence of the first solution different from bits of a bit sequence of the second solution is changed on each of the plurality of bits, computing, when an increase of the energy continuously occurs a plurality of number of times, a total value of the energies increased in the plurality of number of times, decide a maximum value of the plurality of temperature values based on the total value, and setting the decided maximum value.