ETD

• Electric Propulsion System
• Fault-tolerant control
• health-monitoring
• Hybrid Propulsion
• Li-Po Battery Pack
• lightweight UAVs
• PMSM

Within the broader context of aviation’s objective to achieve net zero carbon dioxide emissions, the partial or full electrification of Unmanned Aerial Vehicle (UAV) propulsion systems has gained considerable interest in both industrial and academic sectors. Electrification offers several advantages, including power on demand, improved efficiency, reduced noise and emissions, and lower operational and maintenance costs. However, the transition to electric propulsion also introduces significant technical challenges. These include the limited energy density of batteries and the need to meet stringent reliability requirements for airworthiness certification. Electric propulsion systems, despite their benefits in torque density, compactness, and control precision, often fall short of meeting the reliability standards required for safe UAV operation. In particular, standards such as those issued by NATO (AEP 83) mandate relative low probabilities of catastrophic failure. However, current fully electric propulsion architectures often exceed these limits, primarily due to failure modes in the electric drive subsystem including inverter switch faults, motor stator winding issues, and control electronics malfunctions. These types of faults are especially critical, as they can result in immediate loss of thrust and vehicle control. On the other hand concerning the endurance bottleneck of LiPo barriers two complementary long-term and a short-term strategies can be pursued. In the first case advanced health monitoring techniques of LiPo batteries can improve efficiency, safety, and longevity. In the second case, propulsion system hybridization can improves both performance and reliability while minimizing architectural redesign, since it leverages existing propulsion hardware.
To address these challenges, the first major contribution of this research focuses on developing fault tolerant control strategies for electric drives based on high speed permanent magnet synchronous motors. This includes current signature based fault detection and isolation techniques, elliptical fitting methods for inverter fault diagnosis in the Clarke plane, and post fault current control strategies using neutral point modulation in four leg inverter topologies. Additionally, a method is introduced for estimating permanent magnet demagnetization or torque imbalance in dual motor configurations using only current and speed feedback. The fault tolerant control and health monitoring algorithms are validated via simulation on high fidelity dynamic models of reference electric drives, which is experimentally validated in nominal conditions.
The second major contribution targets the challenges in energy storage reliability. Multi cell lithium polymer battery packs are widely used but are limited by their relatively low energy density and sensitivity to operational conditions. This research proposes an observer based estimation framework for state of charge and state of temperature, built on a lightweight electrothermal model. Several estimation techniques are compared, including extended Kalman filters, sliding mode observers, and Luenberger observers. Hardware in the loop testing validates the proposed approach, achieving estimation errors within five percent for state of charge and two degrees Celsius for state of temperature.
As a complementary approach, the third contribution addresses the previous two limitations by exploring hybrid electric propulsion as a practical solution. The aim is not to design a novel hybrid system but to retrofit conventional internal combustion engine platforms. In this strategy, existing onboard electric generators are repurposed to operate as booster motors during high power flight phases such as take-off or climb. This configuration combines the efficiency of electric drives with the endurance of fuel-based propulsion, while improving reliability through power source redundancy. A simulation framework and a preliminary supervisory hybrid control computer unit are developed to manage operating modes and coordinate power flow, enabling optimum power split between the electric and combustion subsystems.
Together, these contributions form an integrated reliability and performance oriented framework for electric and hybrid UAV propulsion. The proposed solutions enhance fault resilience, energy storage safety, and hybrid performance, supporting the development of certifiable and high performance UAV systems aligned with future aviation standards. The validation platforms are primarily small fixed wing and fixed wing vertical take-off and landing UAVs.