Sensor Uncertainty Quantification via Close Match Span Calibration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for calibrating sensors, such as one-point through-origin (OPTO) calibration, are inadequate as they assume linearity through origin, which is not valid in many cases, and lack a clear method for deriving uncertainty estimation models using one-point close-match (OPCM) calibration, making it difficult to assess and prioritize uncertainty sources.
Innovation Solution
A generalized equation is developed to quantify measurement uncertainty using OPCM calibration, specifically for measuring sensors like gas sensors, which calculates combined standard uncertainty and relative contributions of uncertainty sources through equations I and II, allowing for improved uncertainty management.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If one-point through-origin (OPTO) calibration is used, then the calibration process is simple and widely applicable, but the measurement precision deteriorates because it assumes linearity through origin which is not valid in many cases
Solution Approach 1:
The patent changes the calibration model parameters from fixed through-origin linearity to a flexible close-match span model that allows the calibration line to not pass through the origin, thereby improving measurement precision while maintaining operational simplicity
Solution Approach 2:
The patent introduces a dynamic calibration approach where the calibration line is determined by two points (close match span points) rather than being fixed through the origin, allowing the model to adapt to actual sensor behavior and reduce measurement uncertainty
2Measurement precision
If one-point close-match (OPCM) calibration is used, then the measurement precision improves by not requiring linearity through origin, but the device complexity increases due to lack of standardized uncertainty estimation methods
Solution Approach 1:
The patent segments the uncertainty estimation into distinct components by identifying separate uncertainty sources (calibration standard uncertainty, repeatability uncertainty, etc.) and assigning specific equations to each, thereby managing complexity through systematic decomposition
Solution Approach 2:
The patent changes the approach from qualitative uncertainty assessment to quantitative parameter-based estimation by introducing specific mathematical models (Equations 1-4) that calculate uncertainty components, thereby providing a standardized method that manages complexity
3Reliability
If quantitative uncertainty estimation is implemented, then the reliability of measurement management improves, but the difficulty of detecting and measuring increases due to complex calculation requirements
Solution Approach 1:
The patent segments the complex uncertainty calculation into manageable components (Equations 1-4) that can be calculated separately and systematically combined, thereby reducing the difficulty of implementation while maintaining reliability
Solution Approach 2:
The patent establishes a feedback mechanism where measured values are compared with standard values, and the differences are used to calculate uncertainty components, creating a self-correcting system that maintains reliability while providing clear calculation pathways
Data Source
AI summary
Disclosed herein is a quantifying method of uncertainty of measured value by close match span calibration of measuring sensor comprising a step of:a quantifying using below equation I.uc(xbag)≃[1fcylu(Rbag)]2+[1fcylu(Rcyl)]2+u2(xcyl)(I)wherein, uc is a combined standard uncertainty, Xbag is a measured value of a test, xcyl is standard value, fcyl is standard response factor (sensitivity coefficient), Rbag is a signal value of a test, represents xbag·fbag, Rcyl represents xcyl·fcyl, and u is standard uncertainty.