Digital Adaptive Filter for Fluid Flow Sensor Signal Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing measuring systems for fluid flow parameters like flow velocity, volume flow rate, pressure, and density face challenges in processing sensor signals efficiently and precisely, particularly due to the limitations of discrete Fourier transformation, which hampers high-frequency resolution and requires significant computing power, and are prone to disturbances that complicate the extraction of accurate measurement variables.
Innovation Solution
A measuring system employing a tube arrangement with a flow obstruction and sensor arrangements to generate frequency spectra, coupled with digital adaptive filtering using a microprocessor to filter and process sensor signals, allowing for rapid and precise extraction of measurement variables by determining filter coefficients through algorithms like LMS or RMS, enabling recursive adaptation to disturbances.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If discrete Fourier transformation is used to process sensor signals, then frequency analysis capability is provided, but computing power requirements increase and processing speed decreases
Solution Approach 1:
The patent changes the processing parameter from batch Fourier transformation to recursive filtering with adaptively updated coefficients. The filter coefficients are updated recursively based on incoming signal data, transforming the computational approach from high-computation batch processing to efficient incremental processing, thereby improving both speed and real-time capability
Solution Approach 2:
The patent performs preliminary action by pre-adapting filter coefficients to match the spectral characteristics of the signal before full measurement begins. This preliminary adaptation phase prepares the filter for efficient processing, allowing the system to achieve high frequency resolution without requiring continuous high-computation Fourier transforms during operation
2Measurement precision
If discrete Fourier transformation is used for signal processing, then frequency spectrum analysis is achieved, but computational burden increases significantly
Solution Approach 1:
The patent substitutes the mechanical/computational Fourier transformation process with an electronic/algorithmic recursive filtering system. Instead of performing heavy mathematical transformations, the system uses adaptive filters that recursively compute spectral characteristics with minimal computational energy, replacing energy-intensive processing with efficient incremental updates
Solution Approach 2:
The patent introduces dynamics by making filter coefficients adaptive and time-varying rather than static. The coefficients dynamically adjust to track changing spectral characteristics, allowing the system to maintain accurate spectral analysis with reduced computational burden compared to repeated Fourier transforms
3Measurement precision
If sensor signals are processed without adaptive filtering, then processing simplicity is maintained, but measurement accuracy decreases due to disturbances
Solution Approach 1:
The patent implements feedback by using the first sensor signal to continuously adapt the filter coefficients that process the second sensor signal. The system monitors spectral characteristics in real-time and adjusts filter parameters accordingly, creating a closed-loop system that automatically compensates for disturbances and maintains high measurement accuracy
Solution Approach 2:
The patent performs preliminary adaptation of filter coefficients based on initial spectral analysis before full measurement operation begins. This preliminary setup phase configures the filter to match expected signal characteristics, simplifying subsequent processing while ensuring high accuracy from the start
4Measurement precision
If high frequency resolution is achieved through traditional methods, then spectral component separation is improved, but processing time increases
Solution Approach 1:
The patent uses periodic action through recursive updates at optimized intervals rather than continuous heavy computation. The filter coefficients are updated periodically based on accumulated data, achieving high spectral resolution through incremental improvements rather than continuous full-spectrum analysis, thereby reducing processing time
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enables faster and more precise processing of sensor signals, allowing for real-time measurement of fluid flow parameters with improved accuracy and reduced computational burden, effectively filtering out disturbances and enhancing the measurement system's dynamic range.
Implementation Method 1
the flow obstruction is adapted to induce in the fluid flowing past vortices having a shedding rate (1/fVtx) dependent on the flow velocity of the fluid, in such a manner that a Kármán vortex street is formed in the fluid flowing downstream of the flow obstruction
Implementation Method 2
a measuring transducer serving for registering pressure fluctuations in the flowing fluid, for example, for registering pressure fluctuations in a Kármán vortex street formed in the flowing fluid
Data Source
AI summary
A measuring system includes: a lumen forming a flow path and a flow obstruction arranged in the flow path for effecting a disturbance in a flowing fluid; a sensor arrangement adapted to produce a first sensor signal and a second sensor signal; and transmitter electronics. The transmitter electronics are adapted to receive both the first and second sensor signals and to convert such into first and second sensor signal sampling sequences approximating the first and second sensor signals, respectively, the transmitter electronics further adapted using a digital adaptive filter to ascertain from the first sampling sequence a filter coefficients set and therewith to form a z-transfer function for filtering the second sampling sequence such that the z-transfer function is determined by the filter coefficients set, the signal filter and the second sampling sequence to produce a wanted signal sequence, to produce therefrom digital measured values representing a measurement variable.