Vibration Sensor Defect Detection via Spectral Whitening
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Vibration sensors in machines, such as aircraft turbine engines, often detach under severe operational conditions, leading to false alarms and reduced reliability in predictive maintenance due to changes in their transfer function and signal structure.
Innovation Solution
A method involving spectral whitening of the vibration signal and calculation of an asymmetry indicator based on the difference between positive and negative aberrant points, which helps distinguish defects in the sensor from other sources of asymmetry, using cepstral whitening and an aberration threshold to detect sensor defects accurately.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If vibration sensors are fixed using adhesive under severe operational conditions, then the sensor can be installed and function initially, but the sensor may detach over time causing false alarms and reduced reliability
Solution Approach 1:
The patent applies preliminary action by detecting sensor defects through asymmetry analysis before they cause false alarms. The system continuously monitors the vibration signal asymmetry and triggers a defect indication when asymmetry exceeds a threshold, allowing preventive maintenance before the sensor completely detaches and causes unreliable measurements
Solution Approach 2:
The patent implements feedback by using the vibration signal asymmetry as a diagnostic indicator that feeds back to the monitoring system. When asymmetry exceeds the threshold, the system generates a defect indication that alerts operators to potential sensor issues, creating a closed-loop feedback mechanism for maintaining sensor reliability
2Difficulty of detecting and measuring
If vibration sensors are monitored using conventional asymmetry detection, then defect detection is possible, but false alarms occur due to asymmetry from other sources
Solution Approach 1:
The patent applies local quality by analyzing specific characteristics of the vibration signal - specifically the asymmetry in the time domain and spectral content in the frequency domain. By focusing on localized features (asymmetry metrics) rather than overall signal analysis, the system can distinguish sensor defects from other vibration sources with different asymmetry patterns
Solution Approach 2:
The patent uses parameter changes by establishing an asymmetry threshold that changes based on operational conditions. The threshold is determined through analysis of healthy sensor signals under various conditions, allowing the detection system to adapt to changing operating environments while maintaining reliable defect detection
3Measurement precision
If spectral whitening and asymmetry indicator calculation are used for defect detection, then detection precision is improved, but computational complexity increases
Solution Approach 1:
The patent applies segmentation by breaking down the vibration signal into frequency components through spectral whitening, then analyzing asymmetry in specific frequency bands. This segmentation allows the system to focus computational resources on relevant frequency components where sensor defects manifest, improving precision while managing complexity through targeted analysis
Data Source
AI summary
A method detects a defect in a vibration sensor including the whitening of the vibratory signal delivered by the vibration sensor, and the calculation of an indicator of asymmetry of the whitened vibratory signal. The whitening may be a cepstral whitening implemented simply by dividing the Fourier transform of the signal by its modulus. The asymmetry indicator is a counter of aberrant points in the whitened vibratory signal.