D-Optimal Strain Measurement Placement for Severity Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Composite work cycles often suffer from redundancy and lack of specified events, leading to inaccurate mapping to desired severity percentiles due to high variability in equipment application and severity understanding.
Innovation Solution
A method involving the construction of a damage rate basis matrix, followed by a D-optimal row selection calculation to optimize strain measurement device locations and solve for unknown coefficients, ensuring accurate severity estimates and detection of missing or redundant events.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional composite work cycles are used with defined test events, then the testing process is straightforward, but the mapping to desired severity percentiles is inaccurate due to high variability in equipment application
Solution Approach 1:
The patent transforms the testing approach by changing from fixed predefined events to a mathematical optimization framework using D-optimal design. This allows the test parameters to be dynamically selected based on maximizing information gain about severity percentiles, directly improving measurement precision while managing complexity through systematic methodology
Solution Approach 2:
The patent replaces traditional mechanical trial-and-error testing with a mathematical optimization system. By using D-optimal design algorithms to select test events, the system substitutes empirical testing with a rigorous mathematical framework that systematically identifies the most informative test conditions for estimating severity percentiles
2Reliability
If more strain measurement devices are placed on the machine, then more data is collected for better analysis, but hardware costs and system complexity increase
Solution Approach 1:
The patent extracts only the most critical measurement locations needed for accurate severity estimation using D-optimal design. Instead of instrumenting the entire machine, the method identifies and selects a minimal subset of strain measurement locations that provide maximum information about the desired percentiles, reducing hardware quantity while maintaining reliability
Solution Approach 2:
The patent applies partial action by collecting measurements only at strategically selected locations rather than comprehensive monitoring. The D-optimal design determines the minimum necessary measurements to achieve accurate percentile estimation, avoiding excessive instrumentation while ensuring sufficient data quality for reliable severity assessment
3Adaptability or versatility
If composite work cycles include all possible machine operations, then complete coverage is achieved, but redundancy increases and accuracy decreases
Solution Approach 1:
The patent extracts and removes redundant events from the composite work cycle by using D-optimal design to select only the most informative test events. This extraction process eliminates unnecessary redundancy while preserving adaptability, as the selected events are specifically chosen to maximize their contribution to estimating the desired severity percentiles
Solution Approach 2:
The patent applies partial action by including only the essential machine operations needed for accurate percentile estimation rather than all possible operations. The D-optimal design identifies the minimum necessary subset of events that provide maximum information, achieving adequate coverage without the degradation in precision caused by excessive redundancy
Data Source
AI summary
A method for providing improved composite work cycle damage estimates includes constructing a damage rate basis matrix, performing a D-optimal row selection calculation on the damage rate basis matrix, selecting, based on the D-optimal row selection calculation, a finite number of strain measurement device locations on the machine, extracting a target percentile damage rate for each of the one or more strain measurement devices, and using the extracted damage rates to solve for the unknown coefficients and verify the weightings assigned to the machine operations.