Probabilistic Uncertainty Estimator for Radiation Sensor Calibration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for quantifying radioactivity face challenges in accurately accounting for uncertainties in sample measurement conditions that deviate from calibration processes, particularly due to variations in sample density, composition, non-uniformity, and other factors, which are often unpredictable and require subjective, time-consuming, and costly approaches.
Innovation Solution
A method using mathematical modeling and numerical calculations to simulate various sample conditions by defining a model with variable parameters, randomly selecting values within defined ranges, and computing calibration factors for multiple configurations to determine mean and uncertainty values, mimicking the results of constructing numerous calibration sources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional methods are used to evaluate uncertainty by considering one variable at a time, then the evaluation process is straightforward and manageable, but the results are subjective and do not account for combined effects of multiple variables
Solution Approach 1:
The patent creates virtual copies of calibration sources through mathematical modeling, generating synthetic calibration sources that replicate the characteristics of physical sources. This allows uncertainty evaluation to account for multiple variables simultaneously by simulating numerous hypothetical scenarios, thereby improving accuracy without requiring complex experimental setups for each variable combination.
Solution Approach 2:
The patent replaces the mechanical/experimental approach of creating physical calibration sources with a computational model. Instead of physically constructing multiple calibration sources to test variable combinations, the system uses mathematical algorithms to simulate their effects, substituting computational processing for physical experimentation and reducing overall system complexity.
2Measurement precision
If a large number of radioactive calibration sources are constructed to account for variable variations, then the uncertainty evaluation becomes more accurate, but the process becomes technically challenging, very time consuming, and expensive
Solution Approach 1:
The patent creates virtual representations of calibration sources through mathematical models, generating synthetic calibration data that mimics the behavior of physical radioactive sources. This approach allows the system to evaluate uncertainties across numerous variable combinations instantaneously through computation, eliminating the time-consuming process of physically constructing and measuring multiple calibration sources while maintaining evaluation accuracy.
Solution Approach 2:
The patent systematically varies model parameters representing different source and sample characteristics (dimensions, densities, compositions, geometries) to simulate the effects of multiple calibration sources. By changing these parameters computationally across many iterations, the system achieves comprehensive uncertainty evaluation without the time and resource expenditure required to physically create corresponding calibration sources for each parameter combination.
3Productivity
If mathematical modeling is used to compute calibration factors, then the calibration process becomes more efficient, but uncertainties related to sample variations from calibration conditions are not adequately accounted for
Solution Approach 1:
The patent performs preliminary uncertainty analysis by integrating variance propagation calculations into the calibration model itself. Before final calibration factor computation, the system pre-calculates how variations in source and sample parameters affect the calibration factors, allowing uncertainties to be accounted for systematically throughout the computational process rather than as a separate post-processing step.
Solution Approach 2:
The patent creates virtual calibration sources through mathematical modeling that incorporate parameter variations representing real-world uncertainties. These synthetic calibration sources replicate the statistical and physical characteristics of physical sources, allowing the computational model to account for sample variations and calibration condition differences while maintaining the efficiency of mathematical computation throughout the calibration process.
Data Source
AI summary
A system for determining the calibration factor uncertainty of a radiation sensor. A computing device accepts a mathematical model of the sensor, sample, and other items affecting the calibration factor, wherein each known mathematical model parameter is assigned its normal dimensions or values and each not-well-known parameter is assigned a variable with an upper and lower limit, and a shape parameter that describes the parameter may vary within or about the limits. Random values consistent with the upper and lower limits and shape parameters for each of the variable parameters in the model are then selected, to create a mathematical model of one possible variation of source-detector measurement configuration. The selection and calibration factor computation are then repeated a large number of times and statistical parameters describing the calibration factor and uncertainty are then computed. Another embodiment performs the steps at different energies for spectroscopic detectors.