Airplane Engine Behavior Model Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for constructing behavior models of airplane engines either fail to account for individual engine differences, leading to inaccurate and non-robust models when using generic databases, or lack the breadth of validity when using specific databases, resulting in poor performance and limited context applicability.
Innovation Solution
A method that combines training a generic regression function from a multi-engine database with a resetting step using specific engine data to create a re-set behavior model that accounts for specific features, reducing the number of parameters and required training data, thereby enhancing accuracy and robustness.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a generic behavior model is constructed from a database concerning a plurality of engines, then the model provides a broad range of validity and robustness when faced with rare contexts, but the model does not take account of differences between engines such as manufacturing disparities, resulting in relatively poor accuracy
Solution Approach 1:
The patent segments the model construction process into two distinct phases: first constructing a generic behavior model from fleet-wide data to capture general patterns and ensure robustness, then creating an adjusted behavior model specific to individual engines by subtracting the generic model from the specific engine's actual behavior. This segmentation allows each model to serve its purpose - the generic model provides broad validity while the adjusted model captures individual engine characteristics for accurate degradation tracking.
Solution Approach 2:
The patent merges the generic behavior model with engine-specific data to create an adjusted behavior model. By combining the strengths of both approaches - the broad coverage of the generic model and the specificity of individual engine data - the solution achieves both robustness across rare contexts and accuracy for individual engine degradation detection.
2Measurement precision
If a specific behavior model is constructed from a database specific to one engine, then the model is more accurate for that engine, but the model does not benefit from data about other engines, resulting in a narrower range of validity and poor robustness when faced with rare contexts
Solution Approach 1:
The patent segments the model construction into a generic component (derived from fleet data) and a specific adjustment component (derived from individual engine data). This allows the specific engine model to benefit from the robustness of fleet-wide data while maintaining accuracy through the engine-specific adjustment, resolving the contradiction between specificity and robustness.
Solution Approach 2:
The adjusted behavior model serves multiple functions: it maintains the broad validity and robustness of the generic model for rare contexts, while simultaneously providing engine-specific accuracy for degradation tracking. The model is universally applicable across different engines yet customized for individual characteristics.
3Adaptability or versatility
If a specific behavior model is constructed from a database specific to one engine, then the model is tailored to that engine's features, but the database is limited by the duration over which the engine has been tracked, resulting in insufficient data coverage
Solution Approach 1:
The patent merges individual engine data with fleet-wide generic data to compensate for the limited duration of individual engine tracking. By combining these data sources, the system achieves both engine-specific adaptability and sufficient data quantity for robust model training, overcoming the limitation of short tracking periods.
Solution Approach 2:
The generic behavior model is constructed in advance from extensive fleet data before being applied to individual engines. This preliminary action provides a pre-trained baseline that can be adjusted for specific engines, eliminating the need to accumulate extensive data for each individual engine before deployment.
Data Source
AI summary
A method of constructing a behavior model (32) of an airplane engine, in particular in order to track the operation of the engine, the method comprising a training step (12) of training at least one statistical regression function on the basis of a generic database (14) containing data from a plurality of airplane engines in order to establish a generic behavior model (10) of the airplane engines, and an additional step of resetting the generic behavior model from data in a database (24) specific to the above-mentioned airplane engine, in order to establish a behavior model (32) that takes account of features specific to that engine.