Constraint Coefficient Control in Combinatorial Search
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing combinatorial optimization systems face challenges in determining appropriate constraint coefficients, leading to either narrowed search ranges that hinder better solutions or failure to satisfy constraints, thus affecting solving performance.
Innovation Solution
A data processing apparatus and method that dynamically adjusts constraint coefficients by comparing solutions using different evaluation functions, replacing constraint coefficients, and determining whether constraints are satisfied, to improve the search process and enhance the likelihood of finding better solutions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the constraint coefficient is increased to ensure constraint satisfaction, then the constraint satisfaction is improved, but the search range is narrowed and better solutions become difficult to reach
Solution Approach 1:
The patent applies dynamics by making the constraint coefficient adjustable rather than fixed. The system dynamically changes the constraint coefficient based on the search progress and solution quality, allowing it to adapt between ensuring constraint satisfaction and maintaining broad search range at different stages of the optimization process.
Solution Approach 2:
The patent changes the parameter of constraint coefficient to resolve the contradiction. By varying this parameter during the search process, the system can shift between prioritizing constraint satisfaction and prioritizing search exploration, thereby resolving the trade-off between reliability and productivity.
2Productivity
If the constraint coefficient is decreased to expand search range, then the search efficiency is improved, but constraint satisfaction becomes difficult to achieve
Solution Approach 1:
The system dynamically adjusts the constraint coefficient based on search progress, allowing it to be low during early exploration phases for broad search range, and high during later refinement phases for ensuring constraint satisfaction. This dynamic adjustment resolves the contradiction between search efficiency and constraint satisfaction.
Solution Approach 2:
The patent changes the constraint coefficient parameter adaptively during the search process. By modifying this parameter based on the current state of the search and solution quality, the system can achieve both broad search range and constraint satisfaction at appropriate times.
3Device complexity
If a fixed constraint coefficient is used, then the system simplicity is maintained, but the ability to adapt to different search stages is reduced
Solution Approach 1:
The patent introduces dynamics to the constraint coefficient, transforming it from a static fixed value to a dynamic adjustable parameter. This allows the system to adapt to different search stages without requiring complete system redesign, maintaining relative simplicity while gaining adaptability.
Solution Approach 2:
The system performs preliminary determination of appropriate constraint coefficients based on the current search state and solution characteristics. This preliminary action allows the system to pre-adjust the coefficient before it becomes a bottleneck, enabling smooth transition between search phases.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A storage unit stores first solutions each being a set of values of state variables. A processing unit obtains, from a search unit that searches for a solution based on a first evaluation function, a second solution obtained through search using an initial solution generated from the first solutions. The first evaluation function includes a constraint term indicating a violation degree of constraints on the state variables and a first constraint coefficient indicating a weight for the constraint term. The processing unit replaces any stored first solution with the second solution, depending on comparison between evaluation values of each first solution and the second solution, indicated by a second evaluation function obtained by replacing the first constraint coefficient with a second constraint coefficient, and determines the first constraint coefficient for next search, based on whether the constraints are satisfied by the second solution and/or other solutions obtained through searches.