Unknown configuration autopilot platform design using machine learning /
Adham Esmat Abdelaziz Mousa,
Unknown configuration autopilot platform design using machine learning / تصميم منظومة طيار آلى للطائرات غير محددة التكوين باستخدام تعلم الآلة / by Adham Esmat Abdelaziz Mousa ; Supervision of Prof. Dr. Gamal Mahmoud El-Bayomi, Prof. Dr. Osama Saaid Mohamady. - 172 pages : illustrations ; 30 cm. + CD.
Thesis (M.Sc)-Cairo University, 2024.
Bibliography: pages 80
One of the ways to realize unknown configuration autopilot is the usage of system identification to estimate the aircraft parameters based on the variation of A/C configuration and flight conditions. The system identification needs an initial condition rather than using random values to find the parameters, to avoid the probability of converging to the wrong set of parameters.
Our approach is the usage of machine learning to identify the initial conditions based on aircraft data. Also, the approach discussed in this thesis acts as a basis for real-time system identification where the aircraft model becomes more accurate by feeding real-time data during the flight. This thesis implements a machine learning life cycle on a dataset which is built from scratch as part of the thesis contribution
Text in English and abstract in Arabic & English.
Machine Learning
Unknown Configuration Autopilot Machine Learning Flight Dynamics Aircraft Dataset System Identification
005.31
Unknown configuration autopilot platform design using machine learning / تصميم منظومة طيار آلى للطائرات غير محددة التكوين باستخدام تعلم الآلة / by Adham Esmat Abdelaziz Mousa ; Supervision of Prof. Dr. Gamal Mahmoud El-Bayomi, Prof. Dr. Osama Saaid Mohamady. - 172 pages : illustrations ; 30 cm. + CD.
Thesis (M.Sc)-Cairo University, 2024.
Bibliography: pages 80
One of the ways to realize unknown configuration autopilot is the usage of system identification to estimate the aircraft parameters based on the variation of A/C configuration and flight conditions. The system identification needs an initial condition rather than using random values to find the parameters, to avoid the probability of converging to the wrong set of parameters.
Our approach is the usage of machine learning to identify the initial conditions based on aircraft data. Also, the approach discussed in this thesis acts as a basis for real-time system identification where the aircraft model becomes more accurate by feeding real-time data during the flight. This thesis implements a machine learning life cycle on a dataset which is built from scratch as part of the thesis contribution
Text in English and abstract in Arabic & English.
Machine Learning
Unknown Configuration Autopilot Machine Learning Flight Dynamics Aircraft Dataset System Identification
005.31