Annealing Optimization Device for Stepwise Solution Ranking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods struggle to obtain stepwise optimal solutions in combinatorial optimization problems using annealing, as they primarily focus on finding the optimal value of the objective function rather than ranked solutions.
Innovation Solution
The optimization device and method employ an annealing process that executes solution calculations multiple times, excluding spin sets within a predetermined distance from previously obtained optimal solutions. Additionally, a learning mechanism is used to create an objective function model through machine learning, which is then incorporated into the annealing process to refine solution calculations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional annealing methods are used to solve combinatorial optimization problems, then the optimal value of the objective function can be obtained, but stepwise optimal solutions (ranked solutions) cannot be obtained
Solution Approach 1:
The patent segments the search space by excluding spin sets within a predetermined distance from previously obtained optimal solutions. This segmentation allows the annealing process to systematically explore different regions of the solution space and retrieve stepwise optimal solutions in ranked order, thereby achieving both solution ranking precision and retrieval efficiency.
Solution Approach 2:
The patent performs preliminary actions by storing previously obtained optimal solutions and using them to define exclusion zones in subsequent annealing iterations. This preliminary action enables the system to systematically retrieve ranked solutions without repeating the search for already-found optima, improving both measurement precision and productivity.
2Measurement precision
If multiple annealing calculations are performed to obtain stepwise optimal solutions, then ranked solutions can be retrieved, but the computational time and complexity increase
Solution Approach 1:
The patent changes the search parameters dynamically by adjusting the exclusion distance threshold and modifying the Hamiltonian based on previously found solutions. This allows the annealing process to efficiently navigate the solution space and retrieve ranked solutions with reduced computational time compared to brute-force methods.
Solution Approach 2:
The patent implements feedback mechanisms where the results of each annealing calculation are fed back into the system to define exclusion zones for subsequent calculations. This feedback loop enables the system to learn from previous results and avoid redundant computations, reducing overall computational time while maintaining solution ranking accuracy.
3Measurement precision
If the search range is restricted to exclude nearby spin sets, then stepwise optimal solutions can be obtained, but the exploration of the solution space is limited
Solution Approach 1:
The patent employs dynamic adjustment of the exclusion distance parameter based on the annealing progression and solution characteristics. This dynamic approach allows the search range to adaptively expand or contract, ensuring that stepwise optimal solutions are obtained while maintaining adequate exploration of the solution space through flexible parameter modification.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach allows for the systematic retrieval of stepwise optimal solutions, enhancing the capability to find ranked solutions in optimization problems, particularly in applications like material development where multiple optimal values are relevant.
Implementation Method 1
an annealing means which executes solution calculation by annealing multiple times
Implementation Method 2
a learning means which learns an objective function model expressed by a spin polynomial by machine learning using training data which includes pairs of a spin set as an explanatory variable and a value obtained by applying the spin set to a black box function
Data Source
AI summary
The optimization device includes an annealing means. The annealing means executes solution calculation so as to exclude from search range spin sets whose distances from the spin sets indicating optimal solutions obtained by the n−1st solution calculation are within a predetermined range in the nth solution calculation.