Simulation Model Surrogate Creation via Sensitivity Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Simulation models often require extensive computational resources and are inefficient for evaluation, particularly when dealing with complex technical systems, as they involve a large number of parameters, making it challenging to assess their sensitivity and quality effectively.
Innovation Solution
A method is introduced to identify a subset of parameters with high sensitivity, create a substitute model using chaos polynomials, Gaussian processes, or artificial neural networks, and iteratively refine it based on quality assessments, allowing for a more efficient evaluation compared to the original simulation model.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a simulation model with a large number of parameters is used to simulate complex technical systems, then the model can capture system behavior more accurately, but the computational resources required and evaluation time increase significantly
Solution Approach 1:
The patent extracts only the most sensitive parameters from the full parameter set by performing sensitivity analysis. This identifies a subset of critical parameters that have the greatest impact on simulation output, allowing the model to be simplified by removing or fixing less sensitive parameters, thus reducing computational complexity while maintaining accuracy
Solution Approach 2:
The patent creates a substitute model (surrogate model) that replicates the behavior of the complex simulation model but with reduced computational cost. This substitute model is trained on data from the original simulation model and can provide similar predictions much faster, effectively copying the essential functionality without the computational burden
2Measurement precision
If all parameters of a simulation model are considered in the evaluation, then the quality assessment is more comprehensive, but the complexity of the evaluation process increases
Solution Approach 1:
The patent extracts and focuses evaluation only on the most sensitive parameters identified through sensitivity analysis. By prioritizing parameters that have the greatest impact on simulation output, the evaluation process becomes less complex while maintaining comprehensive assessment of the critical aspects of the model
Solution Approach 2:
The patent segments the parameter space into sensitive and non-sensitive parameters. This segmentation allows the evaluation process to be divided into manageable parts, focusing computational effort on the most important parameters while simplifying the overall evaluation structure
Data Source
AI summary
A computer-implemented method for processing data associated with a simulation model, the simulation model being designed to simulate at least one aspect of a technical system. The method includes: ascertaining a first parameter group of parameters of the simulation model, whose sensitivity exceeds a predefinable sensitivity limiting value, ascertaining a substitute model for the simulation model based on the first parameter group, assessing a quality of the substitute model, a quality measure characterizing the quality of the substitute model being obtained.