Dynamic Filter Switching for Analyte Sensing Noise Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing sensing devices face increased processing calculations and power consumption issues when attempting to continuously or intermittently quantify analyte concentrations, particularly in noise removal, which hinders device miniaturization and efficiency.
Innovation Solution
A sensing device and method that utilize a filtering unit in the frequency domain, switching between different filters based on the amount of temporal change in the measurement signal to effectively remove noise components while maintaining the capability to track concentration changes, using an identity transformation filter and filters with varying phase delay properties.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a filtering algorithm (particularly Kalman filter) is used to remove noise components, then the accuracy of quantifying analyte concentration is improved, but the amount of processing calculations increases
Solution Approach 1:
The patent divides the filtering process into multiple stages: first applying a moving average filter to remove high-frequency noise, then selectively applying a Kalman filter only when temporal change in signal is small. This segmentation reduces overall computational burden while maintaining measurement precision.
Solution Approach 2:
The patent dynamically adjusts the filtering strategy based on the temporal change of the measurement signal. When temporal change exceeds a threshold, no filtering is applied to preserve response speed; when temporal change is small, filtering is applied to improve accuracy. This dynamic approach optimizes the balance between calculation amount and measurement precision.
2Measurement precision
If filtering processing is applied to remove noise components, then the accuracy of quantifying analyte concentration is improved, but the response speed to temporal change decreases
Solution Approach 1:
The patent implements dynamic filter selection based on temporal change detection. When the temporal change of the measurement signal exceeds a predetermined threshold, filtering is suspended to maintain fast response speed. When temporal change is within the threshold, filtering is activated to improve accuracy. This resolves the contradiction by making the system adaptive to real-time signal characteristics.
Solution Approach 2:
Different filtering strategies are applied to different segments of the signal based on local characteristics. High temporal change regions receive no filtering to preserve response speed, while low temporal change regions receive filtering to improve accuracy. This local differentiation resolves the global contradiction between speed and precision.
3Measurement precision
If multiple types of filters are used to effectively remove noise components, then the accuracy of quantifying analyte concentration is improved, but the device complexity increases
Solution Approach 1:
The patent segments the filtering functionality into two distinct filter types: a moving average filter for high-frequency noise removal and a Kalman filter for comprehensive noise reduction. By separating these functions and selectively applying them based on signal characteristics, the system achieves high measurement precision without requiring all filters to operate simultaneously, thus managing device complexity.
Solution Approach 2:
The patent dynamically selects which filter type to apply based on the temporal change of the measurement signal. This dynamic selection mechanism allows the system to use simpler filtering when appropriate and more complex filtering only when needed, reducing the average computational complexity while maintaining high precision when required.
Data Source
AI summary
The present invention relates to a sensing device and a sensing method for continuously or intermittently quantifying a concentration of analyte. A measurement signal correlated with a concentration of analyte is sequentially acquired by use of a sensor (12). A filter processing is performed on a time sequence of the measurement signal acquired by the sensor (12) in a frequency domain via one type of filter among a plurality of types of filters (48, 49, 84). One type of filter used in the filter processing is switched depending on the amount of temporal change of the measurement signal.