Optimization Fitness Evaluation Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Time-intensive optimization processes often require significant computational resources and can lead to suboptimal solutions due to the complexity of simulated problems, with existing methods either simplifying models or using costly parallel processing.
Innovation Solution
A method that conducts preliminary evaluations of candidate solutions using a fitness function to determine their potential for improving the current optimal solution, allowing for faster and more focused complete evaluations, and updates selection criteria based on new optimal solutions, thereby trading off search fidelity for execution speed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If numerous processors are used to perform parallel calculations, then the time required to explore the solution space is reduced, but the cost becomes prohibitive
Solution Approach 1:
The patent applies preliminary action by performing a preliminary evaluation of candidate solutions before conducting complete evaluations. The fitness function evaluates candidate solutions at reduced fidelity first, identifying only those with potential to improve the current optimal solution. This preliminary filtering step reduces the number of complete evaluations needed, thereby reducing optimization run time without requiring a proportional increase in computational resources or processors.
2Productivity
If the simulation model is simplified or evaluation detail is reduced, then the run time is improved, but suboptimal solutions are obtained
Solution Approach 1:
The patent applies local quality by using different evaluation fidelities for different candidate solutions. The fitness function performs preliminary evaluations at reduced fidelity for all candidate solutions, then performs complete high-fidelity evaluations only for those identified as potential improvements. This creates localized high-precision evaluation only where needed, maintaining solution quality for promising candidates while improving overall optimization speed by avoiding exhaustive high-fidelity evaluation of all candidates.
Solution Approach 2:
The patent applies parameter changes by dynamically adjusting the fidelity parameter of the fitness function based on the candidate solution being evaluated. The fitness function can operate at different fidelity levels - reduced fidelity for preliminary screening and full fidelity for complete evaluation. This parameter adjustment allows the system to balance between evaluation speed and solution quality, ensuring that high-fidelity evaluation is applied only when necessary to maintain solution accuracy.
Data Source
AI summary
The performance of optimization algorithms operating with compute-intensive fitness functions is enhanced by constraining time-intensive fitness evaluations for candidate solutions that show low likelihood of being fit at early stages of the fitness evaluation. By prematurely discarding alternatives that could be potentially optimal upon complete fitness evaluation but with low likelihood, the running time of the overall optimization process is advantageously reduced substantially, thereby trading off time complexity for search fidelity.