Aircraft Engine Valve Malfunction Detection via Pressure Regression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Aircraft engine NAI-type valves lack position detection means, making it difficult to supervise their working condition and detect malfunctions, which can lead to overheating or icing issues, necessitating a reliable and simple method to detect early signs of malfunction.
Innovation Solution
A system that acquires output pressure measurements and contextual/command data to define indicators of malfunction, uses a regression model to determine estimators, computes distances between indicators and estimators, and compares these distances to thresholds to detect malfunctions, primarily relying on output pressure measurements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If position detection means are installed on NAI valves to enable supervision, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent uses pressure measurements as an intermediary variable to indirectly infer valve position and detect malfunctions. Instead of directly measuring valve position with sensors, the system measures output pressure and uses this intermediate data, combined with regression models, to detect valve status and malfunctions, thereby avoiding the need for additional position detection means
Solution Approach 2:
The patent replaces mechanical position detection means (sensors, switches) with a computational approach using pressure measurements and regression models. The mechanical system for direct position measurement is substituted by a computational model that infers valve status from pressure data, reducing hardware complexity while maintaining detection capability
2Reliability
If complex supervision systems are implemented to detect valve malfunctions, then reliability is improved, but device complexity increases
Solution Approach 1:
The system uses existing pressure measurement data and contextual data that are already available from the aircraft's normal operations. The regression model is trained offline using historical data, and the actual detection process requires minimal additional computational resources, allowing the system to essentially supervise itself using data it already collects
Solution Approach 2:
The patent changes the approach from direct position monitoring to parameter-based inference using pressure measurements. By monitoring pressure parameters and comparing them against modeled expectations, the system achieves reliable malfunction detection through parameter analysis rather than direct physical measurement
3Ease of operation
If direct position monitoring is used for valve supervision, then ease of operation is improved, but measurement precision deteriorates due to lack of position sensors
Solution Approach 1:
The system establishes a feedback loop where pressure measurements are continuously monitored, compared against expected values from the regression model, and used to detect deviations indicating valve malfunction. This feedback mechanism enables continuous supervision without requiring direct position sensors, maintaining both simplicity and precision
Data Source
AI summary
A method and a system for detecting the first signs of malfunction of an aircraft engine valve, including an acquisition device to acquire the output pressure measurements of the valve, and contextual and command data of the valve, a processor to define a set of indicators of the first signs of malfunction, according to the output pressure measurements and the contextual and command data. A set of estimators corresponding to the set of indicators of the first signs of a malfunction, is determined using a previously produced regression model. At least one distance between the set of indicators and the set of estimators is computed. The distance is compared to a threshold of the first signs of a malfunction, in order to detect the first signs of malfunction of the said valve.