Genetic Algorithm Constraint Adaptation for Optimization Speed
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing optimization calculation methods face challenges in efficiently updating constraint conditions during genetic algorithm-based optimization processes, leading to either non-conforming solutions or excessive calculation time, as the pace of constraint updates is difficult to predict and may be too fast or too slow.
Innovation Solution
The method dynamically changes constraint conditions for each generation by setting the constraint condition value based on the previous generation's value and the provisional value achieved by half or more of the individuals used for generating child individuals, using the equation Li = (1-k) × Li-1 + K × Li', where k is a coefficient controlling the evolution pace.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the constraint condition is updated using a predetermined number of generations rule, then the optimization process can proceed systematically, but the update pace is rigid and may be too fast or too slow leading to non-conforming solutions or excessive calculation time
Solution Approach 1:
The patent implements feedback by monitoring the satisfaction rate of constraint conditions across the population and using this information to dynamically adjust the constraint condition for the next generation. The constraint condition is updated when a predetermined proportion of individuals satisfy it, creating a closed-loop control system that adapts the optimization pace based on actual population performance rather than following a rigid generation-based schedule
Solution Approach 2:
The patent makes the constraint condition dynamic by changing it adaptively during the optimization process based on the satisfaction rate. Instead of using a fixed update schedule, the constraint condition value is adjusted in real-time according to how well the current population meets the constraints, allowing the system to speed up or slow down the convergence pace as needed
2Reliability
If the constraint condition is updated when all individuals fall below a predetermined value, then the optimization can ensure constraint satisfaction, but the calculation time increases significantly
Solution Approach 1:
The patent applies partial action by updating the constraint condition when a predetermined proportion (not necessarily all) of individuals satisfy the constraint. This partial threshold approach allows the optimization to proceed without waiting for complete satisfaction across the entire population, reducing calculation time while still maintaining adequate constraint satisfaction through the proportional criterion
3Adaptability or versatility
If the constraint condition value is set too high initially, then the optimization can start with loose constraints allowing population evolution, but the convergence to the target constraint value becomes slower
Solution Approach 1:
The patent makes the constraint condition dynamic and adaptive, allowing it to evolve from initial loose constraints to tighter constraints as the optimization progresses. The constraint condition value is automatically adjusted based on the satisfaction rate, creating a dynamic trajectory that balances initial population flexibility with eventual convergence speed without requiring manual intervention to change the constraint level
Data Source
AI summary
An optimization calculation method includes, generating individuals of a present generation with an individual selected in a previous generation as a parent individual using an algorithm for obtaining optimum solutions while evolving a plurality of individuals for each generation, evaluating each individual of the present generation, setting a constraint condition value of the present generation based on a constraint condition value of one generation before with respect to the present generation, and a constraint condition provisional value achieved by a half or more of individuals used for generating child individuals of a next generation, determining whether an evaluation result of each individual satisfies a constraint condition value of the present generation, changing for lowering an evaluation result of individuals which do not satisfy the constraint condition value of the present generation, and selecting individuals to be solution candidates from among individuals on which the determination or change process is performed.