Air Cycle Machine Deceleration Analysis for Failure Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Rotating machinery in aerial vehicles, such as Air Cycle Machines, experience reliability issues due to moisture-induced air bearing failure, leading to noncompliance conditions that are difficult to predict and prevent.
Innovation Solution
A computer-implemented method using historical records and sensor data to generate a predictive model that identifies deceleration time patterns and labels, allowing for the prediction of noncompliance conditions in rotating machines, specifically Air Cycle Machines, by analyzing time-to-zero values and sending alerts for maintenance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional monitoring methods are used for rotating machinery, then device complexity is reduced, but reliability decreases due to inability to predict noncompliance conditions
Solution Approach 1:
The system performs preliminary analysis by collecting historical deceleration time data and training predictive models before actual failure occurs. The model is trained offline using historical noncompliance condition data, enabling early prediction of potential failures before they manifest, thus improving reliability without requiring complex real-time intervention systems.
Solution Approach 2:
The patent replaces traditional mechanical monitoring systems with data-driven predictive analytics. Instead of using complex mechanical sensors and physical analysis methods, the system uses machine learning models that analyze deceleration time patterns from existing sensor data, substituting mechanical complexity with computational intelligence to predict noncompliance conditions.
2Measurement precision
If detailed analysis of deceleration time patterns is performed, then measurement precision improves, but loss of time increases due to data processing requirements
Solution Approach 1:
The system performs preliminary processing of deceleration time data by training predictive models offline using historical data. Feature extraction and model training are completed in advance, so that during operation, the system only needs to input new deceleration time values into the pre-trained model for rapid prediction, achieving high measurement precision without real-time processing delays.
Solution Approach 2:
The patent extracts only the most relevant features from deceleration time data for model training and prediction. By identifying and extracting key temporal patterns and characteristics from the raw deceleration data, the system reduces the dimensionality of the problem, enabling precise measurements while minimizing the time required for data processing and analysis.
Data Source
AI summary
Aspects of the present disclosure provide a method and apparatus for predicting noncompliance conditions of rotating machinery in aerial vehicles. Embodiments include receiving records related to historical noncompliance conditions of a plurality of rotating machines. The records include deceleration time values related to the plurality of rotating machines. Embodiments include generating, based at least on the deceleration time values and the historical noncompliance conditions, a predictive model. Embodiments include receiving sensor data related to a rotating machine of an aerial vehicle. Embodiments include determining, based on the sensor data, a series of deceleration time values for the rotating machine. Embodiments include using the predictive model to determine whether a noncompliance condition is predicted to occur for the rotating machine based on the series of deceleration time values.