Noise-Cancellation Sensor Screening for Corrupted Signal Exclusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Noise-cancellation systems face performance degradation due to corrupted sensor signals, which can occur from loose wiring connections or other causes, leading to 'crackling' sounds and requiring effective detection and exclusion to maintain system efficacy.
Innovation Solution
A noise-cancellation system and method that utilize sensors to determine the power spectral density of sensor signals at various frequencies, calculate a measure of association using correlation coefficients, and exclude signals exceeding a predetermined threshold to generate a noise-cancellation signal, while notifying users of corrupted sensors and disengaging them to prevent performance degradation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If all sensor signals are used to compute noise-cancellation signal, then system productivity is improved, but reliability deteriorates due to corrupted sensor signals causing crackling sounds and performance degradation
Solution Approach 1:
The system performs preliminary detection of corrupted sensor signals by analyzing power spectral density and calculating measure of association (correlation coefficient) before the noise-cancellation computation. This preliminary action identifies and flags corrupted signals in advance, allowing the system to exclude them from subsequent processing and prevent performance degradation
Solution Approach 2:
The system extracts and removes corrupted sensor signals from the set of signals used for noise-cancellation computation. By identifying signals with measure of association exceeding the threshold and excluding them, the system separates harmful corrupted signals from useful clean signals, maintaining high reliability while preserving productivity
2Reliability
If sensor signals are excluded based on measure of association threshold, then reliability is improved by preventing crackling sounds, but device complexity increases due to additional detection and analysis steps
Solution Approach 1:
The system transforms sensor signals into the frequency domain by calculating power spectral density, and further transforms the analysis by computing correlation coefficients between frequency and power values. These parameter changes convert the detection problem into a mathematical comparison against a threshold, enabling automated reliability assessment without requiring complex physical modifications
Solution Approach 2:
The system replaces manual sensor validation and corruption detection with automated computational methods. Instead of physical inspection or manual testing, the system uses algorithmic analysis of power spectral density and correlation coefficients to automatically identify and exclude corrupted signals, reducing operational complexity
Data Source
AI summary
A noise-cancellation system, including: a plurality of sensors, each sensor outputting a sensor signal; a controller configured to receive each sensor signal, and, for each sensor signal, to: determine a power of the sensor signal at a plurality of frequencies; determine a measure of association between the power of the sensor signal at the plurality of frequencies and frequency; and determine whether the measure of association exceeds a predetermined threshold, wherein the processor is further configured to compute a noise-cancellation signal using the sensor signals, wherein the noise-cancellation signal is computed excluding sensor signals that were determined to exceed the predetermined threshold; and at least one actuator receiving the noise-cancellation signal and producing a noise-cancellation audio signal.