Spectrometric Calibration Using Universal and Site-Specific Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Spectrometers of the same type exhibit variability in measurement properties due to differences in technical properties, measurement setups, conditions, and medium composition across different sites, leading to measurement errors in determining measurands, which existing methods struggle to efficiently and accurately address.
Innovation Solution
A spectrometric measurement method that involves performing reference measurements, using universal and supplementary training data to develop a dedicated model for each measurement situation, accounting for specific influencing factors through a universal model adapted or adjusted to reflect the characteristics of each site, thereby reducing the need for extensive training data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If a universal model is used across multiple spectrometers and measurement sites, then model development time is reduced, but measurement precision deteriorates due to spectrometer variability and site-specific conditions
Solution Approach 1:
The patent segments the measurement process into two distinct phases: a universal model development phase using reference spectra from multiple spectrometers, and a site-specific calibration phase using supplementary spectra from each measurement site. This segmentation allows the universal model to capture general spectral-measurand relationships while site-specific adjustments account for local variations, thereby resolving the contradiction between rapid model deployment and measurement precision.
Solution Approach 2:
The patent performs preliminary action by collecting and processing reference spectra from multiple spectrometers during the model development phase to establish a universal model. This preliminary work captures the core relationships between spectral features and measurands, enabling rapid deployment. The site-specific supplementary measurements then serve as fine-tuning adjustments, maintaining precision without requiring complete model redevelopment.
2Measurement precision
If extensive training data covering all possible variations is collected, then measurement precision improves, but the complexity and time required for model determination increases significantly
Solution Approach 1:
The patent applies partial action by collecting only the essential reference spectra needed to establish the universal model relationships, rather than attempting to collect exhaustive training data for all possible variations. The site-specific supplementary spectra provide the necessary adjustments for local conditions. This approach achieves sufficient measurement precision without the prohibitive complexity of comprehensive data collection and model determination.
3Measurement precision
If site-specific models are developed for each spectrometer and measurement location, then measurement precision improves, but the time and effort required for model determination increases
Solution Approach 1:
The patent creates a universal model that serves multiple spectrometers and measurement sites simultaneously, capturing the fundamental relationships between spectral features and measurands. This universal model can be deployed across numerous locations without requiring separate complete model developments. Site-specific supplementary spectra provide targeted adjustments, maintaining precision while avoiding the time-consuming process of developing entirely separate models for each location.
Data Source
AI summary
A spectrometric measurement method of, in measurement situation(s) within a predetermined application, determining measurement results of measurand(s) of a medium of an application specific type, wherein each measurement situation is predetermined by specifying at least one influencing factor affecting measured spectra in the respective measurement situation, includes: based on universal training data, determining a universal model for determining predictions of each measurand; for each measurement situation, based on the universal model and supplementary training data including at least one supplementary spectrum determined in the respective measurement situation and corresponding supplementary value(s) of each measurand, determining a dedicated model for determining measurement results of each measurand in the respective measurement situation; and based on measured spectra determined in the respective measurement situation and the dedicated model, determining measurement results of each measurand of the medium.