Aircraft Engine Anomaly Detection Corrected by Wear Factor
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current health monitoring systems for aircraft engines require manual expert analysis and data downloading after raising alarms, leading to significant processing time and cost, especially for false alarms, which reduces efficiency and increases costs.
Innovation Solution
A health monitoring system that includes an anomaly detection unit for analyzing engine parameters and raising alarms based on threshold exceedance, combined with an engine operating condition monitoring unit to determine wear rates and weight the probability of degradation, thereby reducing false alarms through automated decision-making.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual expert analysis and data downloading are performed after raising alarms, then reliability of anomaly detection is maintained through expert verification, but processing time and operational costs increase significantly
Solution Approach 1:
The system performs preliminary automated analysis of engine data against multiple threshold rules before raising alarms. This preliminary filtering action reduces the volume of data requiring manual expert review, thereby maintaining detection reliability while reducing processing time for true anomalies.
Solution Approach 2:
An automated intermediate analysis layer is introduced between the anomaly detection unit and expert reviewers. This intermediary system applies wear rate models and multiple threshold rules to pre-process and prioritize alarms, serving as a mediator that filters out false alarms before they reach human experts.
2Reliability
If manual expert analysis is performed for all alarms, then false alarms can be identified and filtered, but operational costs and resource requirements increase
Solution Approach 1:
The system implements self-service through automated wear rate calculation and threshold rule evaluation that performs initial false alarm filtering without human intervention. The automated system serves itself by processing and prioritizing alarms based on engine-specific wear characteristics, reducing the need for expensive manual expert review of every alarm.
Solution Approach 2:
The system changes parameters by applying engine-specific wear rates and multiple threshold rules to dynamically adjust alarm evaluation criteria. This parameter-based automation allows the system to adapt to different engine conditions and wear patterns, maintaining false alarm reduction capability while reducing operational costs.
3Ease of operation
If simple thresholding algorithms are used for anomaly detection, then system complexity is reduced and ease of operation is improved, but the rate of false alarms increases
Solution Approach 1:
The anomaly detection system is segmented into multiple independent threshold rules (Type I, II, III) with different sensitivity levels. Each rule segment handles specific detection scenarios, allowing the system to maintain simplicity while reducing false alarms through diversified detection approaches rather than a single complex algorithm.
Solution Approach 2:
The system incorporates feedback mechanisms where alarm outcomes and engine response data are fed back into the analysis. This feedback loop allows the system to learn from previous detections and adjust threshold applications, maintaining operational simplicity while improving reliability by reducing false alarms through experience-based adjustments.
4Measurement precision
If multiple threshold exceedance confirmation rules are applied, then detection precision is improved and false alarms are reduced, but device complexity increases
Solution Approach 1:
The system applies dynamic threshold selection based on engine wear rates and operating conditions. Instead of using fixed complex algorithms, the system dynamically adjusts which threshold rule (Type I, II, or III) applies based on real-time engine state, achieving high detection precision while maintaining algorithmic simplicity through conditional logic rather than complex mathematical models.
Data Source
AI summary
A system for monitoring the state of health of a monitored aircraft engine comprises an anomaly detection unit which analyses engine operating parameters and raises an alarm if a result of the analysis of one of the engine operating parameters crosses a threshold, the alarm being associated with a probability of a given type of engine damage occurring. The system further comprises an engine operating conditions monitoring unit which determines a state of wear of the engine, and an alarm corroboration unit which weights the said probability of occurrence by the determined state of wear. The invention is applicable to the preventive maintenance of turbomachines.