Compressor Stall Warning via Nonlinear Feature Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting imminent compressor stall in gas turbine engines are unreliable due to changes in tip clearance and eccentricity, leading to false alarms and inadequate warning systems, especially in aero-engines where tip clearance varies during flight cycles.
Innovation Solution
A novel method using nonlinear feature extraction algorithms, specifically phase space reconstruction and evaluation of approximate entropy from casing-mounted pressure transducer data, to identify flow disturbances prior to compressor stall, providing an effective stall warning signal.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If stall warning systems are based on blade passing signature irregularity using pressure transducers, then flow disturbances can be detected, but the systems produce false alarms and unreliable results due to changes in tip clearance and eccentricity during flight cycles
Solution Approach 1:
The patent segments the blade passing signature into individual blade signatures and analyzes each blade's contribution separately. By comparing each blade's signature against a reference and identifying deviations, the system can detect stall-related flow disturbances while being less sensitive to overall tip clearance changes that affect all blades uniformly. This segmentation approach isolates the anomaly detection from the confounding effects of clearance variations.
Solution Approach 2:
The patent applies local quality by focusing analysis on specific circumferential locations where blade passing signatures are measured, rather than attempting a global analysis of the entire compressor annulus. The system identifies local irregularities in blade passing signatures that indicate stall inception, allowing it to detect localized flow disturbances even when overall tip clearance conditions vary during flight operations.
2Measurement precision
If multiple pressure transducers are used at different locations to detect blade passing irregularities, then flow disturbances can be detected, but the system complexity increases and false alarms increase
Solution Approach 1:
The patent merges the information from multiple pressure transducer measurements by combining individual blade passing signatures into a composite analysis. Rather than treating each transducer location independently, the system integrates the signals to form a comprehensive blade passing signature that captures flow disturbance characteristics while reducing the impact of location-specific variations and minimizing false alarms through signal fusion.
Solution Approach 2:
The patent creates a universal blade passing signature analysis method that can be applied regardless of the specific number or location of pressure transducers. The system processes signals from multiple locations through a unified algorithmic framework that identifies blade-specific irregularities, making the measurement system adaptable and reducing complexity by providing a single analytical approach that works across different sensor configurations.
3Reliability
If stall warning systems rely on correlation measure or event rate parameters, then pre-stall flow irregularities can be detected, but the systems are sensitive to tip clearance changes and eccentricity variations
Solution Approach 1:
The patent performs preliminary action by establishing reference blade passing signatures during known stable operating conditions before attempting to detect stall. These reference signatures, which account for the specific compressor's tip clearance and eccentricity characteristics, are stored and used as a baseline for comparison during actual operation. This preliminary characterization allows the system to adapt to the specific compressor configuration and reduce sensitivity to normal operating variations.
Solution Approach 2:
The patent applies parameter changes by transforming the raw pressure signals into blade-specific signature parameters that are less sensitive to tip clearance and eccentricity variations. By converting the continuous pressure time-series into discrete blade passing signatures and analyzing their statistical properties, the system changes the parameter space to one where stall-related changes are more distinct from those caused by clearance variations, improving detection reliability across different operating conditions.
Data Source
AI summary
The present disclosure relates to a novel method to detect an imminent compressor stall by using nonlinear feature extraction algorithms. The present disclosure focuses on the small nonlinear disturbances prior to deep surge and introduces a novel approach to identify these disturbances using nonlinear feature extraction algorithms including phase-reconstruction of time-serial signals and evaluation of a parameter called approximate entropy. The technique is applied to stall data sets from a high-speed centrifugal compressor that unexpectedly entered rotating stall during a speed transient and a multi-stage axial compressor with both modal- and spike-type stall inception. In both cases, nonlinear disturbances appear, in terms of spikes in approximate entropy, prior to surge. The presence of these pre-surge spikes indicates imminent compressor stall.