Kalman Filter Initialization for Gas Turbine Engine Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional gas turbine engine modeling requires initial thermal stability, which can take 30-60 minutes to achieve, leading to residual errors if dynamic states are not accurately initialized, necessitating adaptive control systems that can update in real-time and automate the initialization process.
Innovation Solution
The use of a Kalman or similar type filter to initialize and configure dynamic model states of a gas turbine engine, allowing for real-time adaptation and automation of engine control based on measured performance parameters such as temperature, pressure, and heat flux.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional models wait for thermal stability to initialize dynamic states, then model accuracy is improved, but initialization time increases significantly (30-60 minutes)
Solution Approach 1:
The Kalman filter performs preliminary estimation of dynamic states using available measurements before thermal stability is achieved. By predicting thermal states ahead of time based on measured parameters and system models, the filter eliminates the need to wait for thermal equilibrium, reducing initialization time from 30-60 minutes to near-real-time operation.
Solution Approach 2:
The Kalman filter continuously updates dynamic state estimates by comparing predicted states with actual measurements and adjusting accordingly. This feedback mechanism allows the system to converge to accurate thermal state estimates rapidly without requiring the slow thermal stability process, maintaining model accuracy while dramatically reducing initialization time.
2Loss of time
If dynamic states are initialized without accurate thermal stability, then initialization time is reduced, but residual errors in performance parameters increase
Solution Approach 1:
The Kalman filter continuously refines dynamic state estimates by comparing model predictions with actual sensor measurements and adjusting the estimates accordingly. This feedback loop rapidly converges to accurate thermal states without requiring waiting for thermal stability, eliminating residual errors that would otherwise persist for extended periods.
Solution Approach 2:
The patent replaces the passive thermal stability waiting process (mechanical/thermal system) with an active estimation algorithm (Kalman filter) that uses mathematical models and measurements to compute thermal states. This substitution transforms a slow physical process into a rapid computational process, maintaining accuracy without the time penalty.
3Measurement precision
If manual initialization methods are used to achieve accurate dynamic states, then model precision is improved, but system complexity and operator burden increase
Solution Approach 1:
The Kalman filter implementation performs self-initialization by automatically computing dynamic states using available sensor measurements and system models. The algorithm autonomously handles the initialization process without requiring operator intervention to select thermal stability points or manually configure parameters, reducing system complexity while maintaining high accuracy through automated estimation.
Data Source
AI summary
Embodiments of the invention can provide systems and methods for initializing dynamic model states using a Kalman or similar type filter. In one embodiment, a system for controlling a gas turbine engine is provided. The system can include at least one sensor adapted to obtain dynamic-type information about a current state of the engine. The system can also include an engine model adapted to receive information from the sensor, and further adapted to reflect the current state of the engine. Furthermore, the system can include a model filter adapted to initialize the engine model with at least a portion of the dynamic-type information, wherein at least one value based at least in part on the filtered dynamic-type information is input to the engine model to determine an engine control action.