Metoda pomiaru osiągów naziemnych samolotu z wykorzystaniem algorytmów atencji oraz ekscytacji w modelu q-kształtnej sztucznej sieci neuronowej
Monografia
ISBN: 978-83-7825-167-5
MNiSW
120
Poziom I
| Status: | |
| Warianty tytułu: |
Method of measuring aircraft ground performance using attention and excitation algorithms in a q-shaped artificial neural network model
|
| Autorzy: | Tomiło Paweł |
| Dyscypliny: | |
| Aby zobaczyć szczegóły należy się zalogować. | |
| Rok wydania: | 2024 |
| Serie: |
Monografie - Politechnika Lubelska
|
| Wersja dokumentu: | Drukowana | Elektroniczna |
| Język: | polski |
| Liczba stron: | 97 + Zał. 1 |
| Miejsce wydania: | Lublin |
| Wydawnictwo: | Wydawnictwo Politechniki Lubelskiej |
| Efekt badań statutowych | NIE |
| Materiał konferencyjny: | NIE |
| Publikacja OA: | TAK |
| Licencja: | |
| Sposób udostępnienia: | Otwarte repozytorium |
| Wersja tekstu: | Ostateczna wersja opublikowana |
| Czas opublikowania: | W momencie opublikowania |
| Data opublikowania w OA: | 17 stycznia 2025 |
| Abstrakty: | angielski |
| Measuring the take-off and landing distance of an aircraft is an important aspect, especially in the process of certification of new types of aircraft, as well as during production tests. The actual take-off and landing distance is important information for pilots in conditions that deviate from the typical ones. The aim of this monograph is to develop a method for measuring the ground performance of aircraft using an on-board measurement unit and software using artificial intelligence methods. The monograph presents the process of developing the discussed method, verification tests and the actual application of the method. An on-board measuring device equipped with appropriate sensors and a q-shaped artificial neural network model using attention and excitation algorithms were developed. The method uses an inertial measurement unit to acquire data in the form of acceleration, angular velocity and aircraft orientation. As part of the monograph, the process of developing the structure of the neural network and selecting appropriate algorithms was carried out. An on-board measuring device has been developed. Tests were carried out with the device installed inside the aircraft (PZL 104 Wilga, PZL 110 Koliber, PZL An-2, MS 880, Cessna 150, Cessna 172). These initial tests focused on collecting data for learning and improving the effectiveness of the neural network. Verification measurements were also carried out in relation to the reference methods, as well as actual measurements for selected aircraft and various types of runway surfaces. |
