Cabin Air Compressor Spectrogram AI for Incipient Surge Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Aircraft cabin air conditioning and temperature control systems (CACTCS) face damage and reliability issues due to compressor surge, which occurs when air flow reverses, causing vibrations and potential mechanical failure, and existing systems can only detect severe surges after damage has begun, leading to costly repairs and reduced aircraft usability.
Innovation Solution
Implementing a machine learning model trained on vibration and speed data to detect incipient compressor surge events in real-time using edge-based processing, applying a fast Fourier transform to generate spectrograms, and deploying a reduced model on a microcontroller to take corrective actions before surge develops.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional surge detection systems are used, then severe compressor surge can be detected, but damage has already begun and costly repairs are needed
Solution Approach 1:
The system performs preliminary detection of incipient surge conditions before full surge occurs. The machine learning model analyzes vibration and operational parameters to identify early warning signs, enabling corrective action to be taken before damage begins, thus preventing the worsening of compressor reliability
Solution Approach 2:
The patent replaces traditional mechanical surge detection methods with an artificial intelligence-based detection system. The ML model processes vibration data and operational parameters to detect incipient surge, providing superior measurement precision compared to conventional mechanical or threshold-based systems
2Measurement precision
If a full machine learning model is deployed, then accurate incipient surge detection is achieved, but the microcontroller lacks sufficient processing power and memory
Solution Approach 1:
The system extracts only the essential features and parameters needed for incipient surge detection from the complete machine learning model. By identifying and retaining only the critical detection logic and removing redundant components, the model achieves accurate detection while fitting within the microcontroller's limited resources
Solution Approach 2:
The patent applies local quality by optimizing the machine learning model specifically for the microcontroller's processing capabilities and memory constraints. The model is tailored with appropriate complexity, data structures, and computation methods that match the local characteristics of the embedded system, enabling accurate detection without exceeding device limits
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Prevents compressor surge by detecting incipient events early, reducing damage and improving aircraft reliability and usability by implementing corrective actions before full surge occurs, thereby extending the lifespan of cabin air compressors.
Implementation Method 1
generating, at the microcontroller and using the operating data, an operating spectrogram
Data Source
AI summary
Examples described herein provide a computer-implemented method that includes receiving training data indicative of incipient compressor surge for cabin air compressors. The method further includes generating, using the training data, a training spectrogram. The method further includes training, by a processing system, a machine learning model to detect incipient compressor surge events for the cabin air compressors using the spectrogram. The method further includes receiving, at a microcontroller associated with a cabin air compressor, operating data associated with the cabin air compressor. The method further includes generating, at the microcontroller and using the operating data, an operating spectrogram. The method further includes detecting, by the microcontroller associated with the cabin air compressor, an incipient compressor surge event by applying the machine learning model to the operating spectrogram. The method further includes implementing a corrective action to correct the incipient compressor surge event.