The availability of accurate aircraft models is one of the key elements in ensuring aircraft improvements. These models are used to improve flight controls and design new aerodynamic systems for the design of deformable aircraft wings.
This project consists of designing a method for identifying certain parameters of the engine model of the US Cessna Citation X business aircraft for the cruise phase from the flight tests. These tests were performed on the designed flight simulator manufactured by CAE Inc. which has flight dynamics D level. Level D is the highest level of accuracy given by the FAA Civil Aviation Authority in the United States.
A methodology based on optimized neural networks using an algorithm called the "Extended Great Deluge" is used in the design of this identification model. Several flight tests for different altitudes and Mach numbers were performed to serve as databases for learning neural networks. Model Validation was carried out using the simulator data. Despite the non-linearity and complexity of the system, engine parameters were predicted very well for a particular flight envelope. This estimated model could be used for engine performance analyzes and could provide aircraft control during this cruise phase.
Engine model identification could also be carried out for the other phases of climb and descent in order to obtain its complete model for the whole Cessna Citation X aircraft flight envelope (climb, cruise, descent). This method used in this work could also be efficient to realize a model to identify aerodynamic coefficients of the same airplane always from flight tests.
| Date | 17 Mar 2017 |
|---|
| Original language | French |
|---|
| Awarding Institution | - École de technologie supérieure
|
|---|
| Supervisor | Ruxandra Botez (Supervisor) |
|---|
Zaag, M. (Author),
Botez (Supervisor),
17 Mar 2017Student thesis: Master's thesis › Master in Engineering: Automated Manufacturing Engineering