Multi-Objective Optimization of Aerial Vehicle Powertrain Architectures based on Energy, Efficiency and Emissions.

The burgeoning demand for efficient and sustainable Aerial Vehicles (AVs) requires the development of powertrain systems tailored to specific mission requirements. Current AVs are incorporating hybrid-electric powertrains, to handle the trade-offs between power density, endurance, and fuel efficiency. However, the challenge remains in selecting which powertrain architecture is most suitable for specific AV missions, given the complex design objectives. Existing literature delves into generalizing architecture suggestions by comparative analyses of limited powertrain variants or analytical sizing. However, it remains difficult to draw definitive conclusions, as current methods often fail to simultaneously account for the multi-objective nature of AV powertrain sizing and selection, a gap that this work aims to address. The proposed framework establishes a novel methodology for the optimal sizing and selection of AV powertrain components, leveraging a simulation framework that unifies aerodynamics, flight dynamics, control systems, and powertrain models. The versatile tool allows us to optimize powertrain systems for a diverse range of applications, enhancing the performance of AVs. The tool is designed to evaluate various powertrain architectures—focusing particularly on a configurable Series-Parallel Hybrid Electric (SPHE) architecture—by sizing components based on mission specific requirements and user-defined objectives. The single SPHE architecture can switch to mimic operation of Internation Combustion Engine (ICE) only, Battery Electric (BEV) only, Series-Hybrid-Electric (SHE), Parallel-Hybrid-Electric (PHE), or combined flexible Series- Parallel-Hybrid-Elected (SPHE) modes. This helps dynamically select the optimal powertrain architecture based on real-time mission data, constraints, and objectives. Results from the framework demonstrate the tool’s ability to identify viable and effective powertrain configurations for different mission profiles with optimized component sizing.