Compressor Stall Margin Detection From High-Frequency Sensor Signals
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional compressor active stability management (CASM) systems for gas turbine engines struggle to accurately and promptly capture degrading stall margin features from high-frequency sensor data, leading to slow and inaccurate stall margin generation, which hinders active control and operability.
Innovation Solution
A control system utilizing machine-learned models, specifically neural networks, to process high-frequency sensor data in real-time, determining stall margin remaining and adjusting engine systems to maintain optimal stability, thereby enhancing the accuracy and speed of stall margin detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional CASM systems process sensor data using traditional methods, then the system structure is simpler, but the processing time is too long and accuracy is insufficient to support active control
Solution Approach 1:
The patent replaces conventional signal processing methods with machine learning models (neural networks) to process high-frequency sensor data. The machine-learned model directly maps sensor inputs to stall margin predictions, eliminating the need for complex traditional signal processing algorithms while achieving both high accuracy and real-time performance.
Solution Approach 2:
The patent transforms the processing approach by changing from traditional algorithmic processing to data-driven machine learning processing. The system uses trained neural network models with optimized architectures (including convolutional and recurrent layers) to process sensor data, fundamentally changing the parameter space from algorithm complexity to model structure and training data characteristics.
2Measurement precision
If conventional CASM systems use traditional processing methods, then computational resources are easier to manage, but the system cannot capture degrading stall margin features accurately
Solution Approach 1:
The patent replaces complex traditional signal processing algorithms with machine learning models that automatically learn feature representations from raw sensor data. The neural network architecture (including convolutional layers for spatial features and recurrent layers for temporal dependencies) automatically captures degrading stall margin features without manual feature engineering, reducing algorithmic complexity while improving accuracy.
Solution Approach 2:
The patent uses trained machine learning models as virtual copies of expert stall margin assessment knowledge. The models are trained on extensive sensor data to create a digital representation of stall margin behavior, enabling accurate feature capture through pattern recognition rather than complex computational algorithms.
3Productivity
If machine learning models are used to process sensor data in real-time, then processing speed and accuracy improve, but computational resource usage increases
Solution Approach 1:
The patent implements a hybrid processing architecture where machine learning models handle only the critical stall margin prediction function, while other system functions use conventional processing. The model processes high-frequency sensor data at selected phases of the compressor rotation cycle, performing partial processing rather than continuous full-data processing, thereby reducing overall computational resource usage while maintaining real-time performance for safety-critical functions.
Data Source
AI summary
A control system for active stability management of a compressor element of a turbine engine is provided. In one example aspect, the control system includes one or more computing devices configured to receive data indicative of an operating characteristic associated with the compressor element. For instance, the data can be received from a high frequency sensor operable to sense pressure at the compressor element. The computing devices are also configured to determine, by a machine-learned model, a stall margin remaining of the compressor element based at least in part on the received data. The machine-learned model is trained to recognize certain characteristics of the received data and associate the characteristics with a stall margin remaining of the compressor element. The computing devices are also configured to cause adjustment of one or more engine systems based at least in part on the determined stall margin remaining.