Experiment Design Optimization for Uncontrolled Factors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In experimental design, there is a challenge in efficiently selecting a subset of candidate design cases for a set of factors, where uncontrolled factors have specified options and controlled factors have assignable options, while ensuring that the design criterion is optimized, especially when resource constraints limit the number of design cases that can be used.
Innovation Solution
A computer-program product and method that access design information to generate an initial subset of design cases, determine a criterion measure, and index data elements corresponding to uncontrolled and controlled factors, assessing whether substituting or changing options improves the design criterion, thereby updating the criterion measure to optimize the experiment design.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If a sample of design cases is selected due to resource constraints, then the quantity of design cases is reduced, but the design criterion optimization is compromised
Solution Approach 1:
The system changes the parameters of uncontrolled factors by substituting specified options with alternative options from the candidate design cases. This allows optimization of the design criterion (e.g., D-optimality, A-optimality) while working within a reduced set of design cases constrained by resource limitations.
Solution Approach 2:
The system performs preliminary indexing of data elements corresponding to uncontrolled and controlled factors before the actual design selection. This preliminary organization enables efficient evaluation and substitution of options to improve the design criterion without requiring exhaustive search through all candidate cases.
2Manufacturing precision
If substitution of options for uncontrolled factors is performed to improve design criterion, then the design optimization is improved, but the complexity of the design process increases
Solution Approach 1:
The system segments the design factors into controlled factors and uncontrolled factors, with distinct handling procedures for each. Controlled factors maintain their assigned options while uncontrolled factors are candidates for substitution. This segmentation simplifies the overall process by focusing optimization efforts only on the uncontrolled factors.
Solution Approach 2:
The system uses design criterion evaluation as feedback to guide the substitution process. After substituting options for uncontrolled factors, the design criterion is re-evaluated to determine if the substitution improved the design. This feedback loop ensures that complexity is justified by actual improvement in design quality.
3Measurement precision
If indexing of data elements is performed for all factors, then the design criterion evaluation is improved, but the computational complexity increases
Solution Approach 1:
The system applies different levels of indexing and processing to different types of factors. Uncontrolled factors, which are candidates for substitution, receive full indexing attention. Controlled factors, which have fixed assigned options, receive minimal processing. This local differentiation reduces unnecessary computational complexity while maintaining accurate criterion evaluation.
Data Source
AI summary
A computing system generates a subset of design cases of candidate design cases. The system indexes, in the subset, data elements. The system generates a design of an experiment by, for each respective data element, determining a status indicating whether the respective data element corresponds to an uncontrolled factor or a controlled factor. When the status indicates the uncontrolled factor, the system determines if substituting a respective set of specified options of a respective candidate design case comprising the respective data element with a different set of specified options of the candidate design cases improves a criterion measure according to a design criterion. When the status indicates the controlled factor, the system determines if changing an assigned option of the respective data element improves the criterion measure. The system updates the criterion measure with an updated criterion measure according to a change of the subset based on generating the design.