Compact Aero-Thermal State Estimation for Gas Turbine Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing fluid-based engineering control systems face inefficiencies and inaccuracies due to computational limitations and the need for real-time estimation of system parameters, particularly in gas turbine engines, where precise control is crucial for safe and efficient operation.
Innovation Solution
A control system incorporating an engine parameter on-board synthesis (EPOS) module using a compact aero-thermal model (CAM) that processes input and output vectors to estimate engine parameters, employing fault detection and accommodation, and gain scheduling to improve accuracy and robustness.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If piecewise linear state variable representations are used, then device complexity is reduced, but measurement precision and reliability of parameter estimation deteriorate
Solution Approach 1:
The patent transforms the control approach by changing from piecewise linear representations to physics-based mathematical models with explicit analytical solutions. This parameter change enables accurate real-time estimation of system states while maintaining computational efficiency suitable for embedded control systems.
Solution Approach 2:
The patent replaces traditional mechanical sensor-based measurement systems with a virtual sensing approach using physics-based mathematical models. This substitution eliminates the need for additional physical sensors while providing accurate parameter estimation through model-based calculations.
2Measurement precision
If semi-empirical relationships with additional control sensors are used, then measurement precision improves, but device complexity and cost increase
Solution Approach 1:
The patent creates virtual copies of physical sensor measurements through mathematical modeling. Instead of installing additional physical sensors, the system uses physics-based models to calculate and estimate unmeasured parameters, providing the same information as physical sensors would without the hardware overhead.
Solution Approach 2:
The patent introduces physics-based mathematical models as intermediaries between available sensor measurements and desired system parameters. These models act as mediators that transform limited sensor data into comprehensive system state information without requiring direct physical measurement of all parameters.
3Measurement precision
If stationary simulations are deployed, then measurement precision improves, but productivity and reliability deteriorate due to computational load
Solution Approach 1:
The patent segments the complex stationary simulation model into component-level physics-based models with explicit analytical solutions. By dividing the system into manageable subsystems (compressor, turbine, heat exchangers, etc.), each with its own analytical model, the system achieves real-time computational speed while maintaining accuracy.
Solution Approach 2:
The patent transforms static stationary simulation models into dynamic real-time models by developing explicit analytical solutions that adapt to changing operating conditions. This enables the system to respond dynamically to real-time inputs while maintaining the accuracy benefits of physics-based modeling.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Systems and methods for controlling a fluid based engineering system are disclosed. The systems and methods may include a model processor for generating a model output, the model processor including a set state module for setting dynamic states of the model processor, the dynamic states input to an open loop model based on the model operating mode. The model processor may further include an estimate state module for determining an estimated state of the model based on a prior state model output and the current state model of the open loop model, the estimate state module determining estimator gain associated with the current state model and applying the estimator gain to determine the estimated state of the model.