Production System Control for Pareto-Based Design Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current multi-dimensional optimization methods for complex product design are computationally intensive and rely on surrogate models with low accuracy, making it challenging to efficiently optimize product characteristics across multiple target variables.
Innovation Solution
A method and system that utilize design evaluation modules, including machine learning and simulation modules, to predict target values and select optimal design datasets by reproducing rankings and accounting for statistical distributions and uncertainty, allowing for more accurate and robust design optimization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multidimensional optimization methods simulate a large number of design variants to optimize target variables, then optimization accuracy is improved, but computational complexity increases considerably
Solution Approach 1:
The patent segments the design evaluation process into multiple independent design evaluation modules, each responsible for predicting specific target variables. This allows parallel processing of different design variants across multiple modules, reducing the sequential computational burden while maintaining comprehensive evaluation coverage.
Solution Approach 2:
The system performs preliminary ranking of design variants using predicted target values from design evaluation modules before conducting detailed simulations. This preliminary filtering eliminates clearly inferior designs early in the process, reducing the number of computationally intensive simulations required while preserving optimization accuracy for promising candidates.
2Use of energy by stationary object
If surrogate models are used to predict simulation results without detailed simulation, then computational complexity is reduced, but prediction accuracy becomes low or variable
Solution Approach 1:
The patent introduces design evaluation modules as intermediary components between rough surrogate models and final design decisions. These modules aggregate predictions from multiple surrogate models and incorporate uncertainty quantification, providing more reliable intermediate assessments that guide subsequent detailed simulations without requiring exhaustive computational resources.
Solution Approach 2:
The system dynamically adjusts the level of detail in evaluation based on predicted performance and uncertainty. Designs with high predicted performance and low uncertainty undergo less detailed evaluation, while those with moderate performance or high uncertainty trigger more detailed simulations. This adaptive parameter adjustment optimizes the trade-off between computational effort and prediction reliability.
3Reliability
If multiple design evaluation modules are used to predict target values, then prediction reliability is improved, but device complexity increases
Solution Approach 1:
The patent merges multiple design evaluation modules into a unified system that processes design variants collectively. The modules share common infrastructure for data management, uncertainty quantification, and result aggregation. This merging approach maintains the reliability benefits of multiple independent predictions while reducing overall system complexity through shared components and coordinated operation.
Data Source
AI summary
A plurality of test data sets include: a first design data set specifying a design variant of a product; and first target values, which quantify target variables of the design variant which are to be optimized and ranked. Furthermore, a plurality of design evaluation modules for predicting target values on the basis of design data sets is provided. For each of the design evaluation modules, a second ranking of the first design data sets with respect to the predicted target values and a deviation of the second ranking from the first ranking are then determined. One design evaluation module is then selected in accordance with the determined deviations. Furthermore, a plurality of second design data sets is generated, and are predicted by the selected design evaluation module. A target-value-optimized design data set is then derived from the second design data sets and is output for the manufacturing of the product.