Acoustic frequency-based method for high-speed aircraft combustion analysis and hybrid artificial intelligence diagnostics
[ 1 ] Instytut Napędów i Lotnictwa, Wydział Inżynierii Lądowej i Transportu, Politechnika Poznańska | [ 2 ] Instytut Automatyki i Robotyki, Wydział Automatyki, Robotyki i Elektrotechniki, Politechnika Poznańska | [ P ] pracownik
[2.2] Automatyka, elektronika, elektrotechnika i technologie kosmiczne[2.7] Inżynieria lądowa, geodezja i transport
2024
artykuł naukowy
angielski
- On-board aircraft diagnosis
- Acoustic empirical characteristics
- Highly turbulent combustion
- Early symptoms identification
- Artificial intelligence
EN In the paper, a method of high-speed continuous combustion monitoring for jet aircrafts, based on acoustic data processing, is proposed. As part of the research, acoustic signals corresponding to the specific phases of the combustible mixture formation and its combustion were separated. They were assigned to the generated process energy value. In the result, parametric functions were obtained in multidimensional spaces of states and processes. Parameters and acoustic characteristics in the amplitude, frequency and JTFA domains constituted information about particular symptoms in the jet engine, thanks to which reference waveforms of signals and forms for specific malfunctions were mapped. The above graphic characteristics have been parameterized to determine their diagnostic reliability factors. The method is verified by empirical signals obtained from turbojet propulsion system in time and power dependent conditions. Appropriate exothermic combustion phases were assigned for the determined acoustic spectra, taking into account the obtained power. Then, representative diagnostic parameters transformed in the process values domain were calculated. Finally, monitoring functions and algorithms of combustion efficiency with determination of malfunctions sources (at dynamic conditions) are proposed. As a result, it is possible to identify first outbreaks of design and process failures during fuel injection and combustion in aircrafts, taking into account artificial intelligence methods. The obtained results indicate it is possible to adapt the acoustic characteristics of the spectra and their discrete representatives, for appropriate states and failures, to detect first dangerous symptoms in the aircraft during flight conditions (supporting OBD with the acoustic adaptive diagnostic procedures).
115304-1 - 115304-28
CC BY (uznanie autorstwa)
czasopismo hybrydowe
ostateczna wersja opublikowana
w momencie opublikowania
200
5,2 [Lista 2023]