Artificial intelligence based smart sensing for electric aircraft propulsion

Lately, the preference of alternative fuels for aircrafts by diverging from traditionality has caused an increased percentage of electrified propulsion systems, encouraging scientific studies to target this research topic in particular. The purpose of this study is to bridge the gap between Unmanned Aerial Vehicles and electric aircrafts, taking into consideration the applicability and adaptability of UAV-based methodologies to more complicated systems. Studies done in this context provided an artificial intelligence-based approach for applicable advanced diagnostic techniques in electric aircraft. Concordantly, a comprehensive open-sourced UAV architecture was developed by using Model-Based Systems Engineering (MBSE) and System Modelling Language (SysML). A propulsion system that is found in the sub-systems of reference architecture allowed an acoustics diagnostics study for detecting propeller damage through machine learning. Following that; damaged propellers were identified via vibration data and a vibration-based diagnostic study was conducted by implementing an artificial neural network. The final part involved increasing the model accuracy by utilizing an ensemble learning methodology that is evaluating the acoustic and vibration data holistically and developing and smart sensing method. One of the aims of this study is to contribute to the manufacturing more optimized, more secure and safe electric aircraft through the integration of advanced diagnostic methods into propulsion systems.