PLS Regression Aircraft Flight Parameter Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current aircraft flight control systems face challenges in accurately estimating flight parameters in real-time, particularly with anemometric, clinometric, and inertial data, which affects the availability and reliability of flight control, and existing solutions require redundant sensors and complex monitoring processes.
Innovation Solution
A method using Partial Least Squares (PLS) regression to automatically estimate flight parameter vectors by receiving observed values, determining a coefficient vector that establishes a linear relationship between explanatory variables and flight parameters, and transmitting the estimated values for consolidation and fault detection, without the need for new sensors or specific gauges.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If redundant sensors and complex monitoring processes are used to improve flight parameter availability, then reliability is improved, but device complexity increases
Solution Approach 1:
The patent introduces an intermediary estimation device that acts as a mediator between existing sensors and the flight control system. This device uses PLS regression to process data from available sensors and generate estimated flight parameters, thereby improving parameter availability without requiring redundant sensors or complex monitoring processes. The intermediary nature of this estimation device resolves the contradiction by providing reliable parameter estimates through a simplified intermediate processing layer.
Solution Approach 2:
The patent replaces the mechanical approach of adding redundant physical sensors with a computational approach using PLS regression algorithms. Instead of physically duplicating sensor systems to improve reliability, the invention substitutes a mathematical modeling approach that processes existing sensor data to generate reliable flight parameter estimates, thereby reducing device complexity while maintaining or improving reliability.
2Measurement precision
If PLS regression is used to estimate flight parameters, then measurement precision is improved, but computing power requirements increase
Solution Approach 1:
The patent applies preliminary action by pre-calculating and storing the PLS regression model coefficients during the design and calibration phase. These pre-computed coefficients are then used during flight operations to quickly estimate flight parameters through simple matrix operations, rather than performing complex real-time regression calculations. This approach maintains high measurement precision while significantly reducing the computing power and energy requirements during actual flight operations.
3Productivity
If real-time estimation is implemented, then productivity is improved, but device complexity increases
Solution Approach 1:
The patent extracts the complex PLS regression calculations from real-time operation and separates them into an offline model development phase. The estimation device only needs to implement the simpler application of pre-computed coefficients to current sensor data, which can be performed efficiently in real-time. This extraction of computational complexity to the modeling phase enables real-time productivity improvement without requiring complex real-time computation infrastructure.
Data Source
Figure 1~2
Figure 3
Figure 4~5
AI summary
The detection device (1) includes means (5) for estimating, over an observation window, a coefficient vector which allows a linear relationship to be determined between the flight parameter vector sought and explanatory values, by implementing a PLS regression, and means (5) for using this coefficient vector to calculate, using linear modeling, an estimated value of said flight parameter vector.