Measurement Data Analysis Device Noise Variation Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for estimating peak shapes in measurement data from analyzers, such as chromatographs, assume uniform noise, which can lead to degraded reliability and accuracy due to noise suppression caused by device characteristics, affecting the predictive distribution of quantitative indices.
Innovation Solution
A measurement data analysis device and method that estimates noise variation or relative noise intensity, using a noise variation estimator and noise intensity estimator respectively, to correct and analyze measurement data, thereby improving estimation accuracy even in the presence of suppressed noise.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a frequency filter is applied in the analyzer to suppress noise, then the signal quality is improved, but the noise intensity is underestimated leading to degraded reliability of predictive distribution
Solution Approach 1:
The patent introduces a noise variation coefficient as an intermediary parameter that mediates between the filtered measurement data and the noise intensity estimation. This coefficient acts as a correction factor that accounts for the noise suppression effect of the frequency filter, allowing the system to recover the true noise intensity from the filtered data without being directly affected by the filter's noise reduction
Solution Approach 2:
The patent changes the parameter representation by estimating the noise variation coefficient that characterizes the filter's effect on noise, rather than directly measuring noise intensity. This parameter transformation allows the system to work with filtered data while still accurately representing the original noise characteristics for reliable predictive distribution
2Device complexity
If Bayesian inference is used for peak shape estimation assuming uniform noise, then the estimation process is simplified, but accuracy is degraded when noise is non-uniform due to device characteristics
Solution Approach 1:
The patent applies local quality by allowing the noise intensity parameter to vary locally through the noise variation coefficient, which captures the non-uniform noise characteristics at different frequencies. This enables the Bayesian inference to account for local noise variations caused by the frequency filter while maintaining the overall framework's simplicity
Solution Approach 2:
The patent performs preliminary action by pre-estimating the noise variation coefficient from the filtered measurement data before conducting the main peak shape estimation. This preliminary estimation of noise characteristics allows the subsequent Bayesian inference to proceed with accurate noise modeling without increasing the complexity of the main estimation process
Data Source
AI summary
A measurement data analysis device analyzes measurement data of a sample obtained in an analyzer, and includes a noise variation estimator that estimates a noise variation coefficient, the noise variation coefficient being applied to a noise included in the measurement data by a frequency filter included in the analyzer, an acquirer that acquires the measurement data to which the frequency filter has been applied in the analyzer, and a calculator that estimates, with use of the noise variation coefficient, a noise intensity included in the measurement data obtained before the frequency filter is applied, and analyzes, based on the estimated noise intensity, the measurement data.