Experiment Design System Using Substitute Model Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for conducting experiments with technical systems or models lack efficiency in selecting optimal input data points, leading to suboptimal accuracy and increased resource utilization in simulations and real experiments.
Innovation Solution
A computer-implemented method that predefines a first set of input data points and determines a second set of input data points based on the first, using a substitute model to predict experiment results, where the margin between prediction statistics for the second set is minimized, allowing for more accurate test case generation and reduced resource requirements, employing Gaussian processes to enhance prediction accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If a first set of input data points is predefined for experiments, then the coverage of available test cases is improved, but the accuracy of predictions for experiment results deteriorates due to larger margins between prediction statistics
Solution Approach 1:
The first set of input data points is segmented into multiple second sets of input data points, where each second set is optimized for specific prediction tasks. This segmentation allows the system to maintain comprehensive coverage while achieving higher prediction accuracy for each subset by reducing the margin between different prediction statistics.
Solution Approach 2:
The system dynamically selects which second set of input data points to use based on the specific experiment or prediction task. This dynamic selection ensures that the most appropriate data subset is used for each prediction, optimizing accuracy while maintaining overall coverage across all possible experiments.
2Measurement precision
If more input data points are used to improve prediction accuracy, then the accuracy of experiment result predictions is improved, but the resource utilization and costs increase
Solution Approach 1:
Instead of using all available input data points for every prediction, the system applies partial action by selecting only the relevant second set of input data points needed for each specific prediction task. This approach achieves sufficient accuracy without the excessive resource consumption of using the complete first set for all predictions.
Solution Approach 2:
The system changes the parameter of data point selection by transitioning from a static use of all data points to a dynamic selection of optimized subsets. This parameter change enables the system to adapt the amount of data used to the specific requirements of each prediction task, optimizing the balance between accuracy and resource efficiency.
3Reliability
If all available test cases are carried out to ensure comprehensive testing, then the coverage and reliability are improved, but the time and resources required increase significantly
Solution Approach 1:
The system performs preliminary action by pre-defining and organizing input data points into optimized second sets before the actual experimentation begins. This preliminary organization enables efficient selection and execution of only the necessary test cases, maintaining comprehensive coverage while significantly reducing the time required to conduct all experiments.
Solution Approach 2:
The system creates optimized copies (second sets) of the original data set that are tailored for specific prediction tasks. These copied and optimized subsets can be used repeatedly for similar prediction tasks, eliminating the need to re-execute redundant test cases and thereby reducing overall experimentation time while maintaining reliability.
Data Source
AI summary
A device and computer-implemented method for carrying out an experiment using a technical system or using a model of a technical system. A first set of input data points for the experiment is predefined. A second set of input data points for the experiment is determined as a function of the first set of input data points. A substitute model for the technical system is configured to determine, as a function of the second set of input data points, predictions for a result of the experiment for a first prediction statistic, which is to be expected for the second set of input data points when carrying out the experiment using the technical system or using the model for the technical system.