Development of an Optimization Scheme for A Fixed-Wing UAV Long Endurance with PEMFC and Battery

Abstract

Unmanned aerial vehicles (UAVs) and fuel cell industries are seeking to enhance the capability and performance of UAVs powered by fuel cells as propulsion systems. Tasks such as surveillance and land surveying increased the need to improve UAVs flight characteristics especially its endurance. The design for long endurance UAVs along with its integrated propulsion system components should be carefully designed and optimized. In this thesis, an optimization approach is developed to obtain optimal flight endurance. The proposed approach includes a component-level, a subsystem-level as well as a system-level modeling. For the component level modeling, models of all propulsion system components are developed based on real physical models. In the subsystem level modeling, fuel cell subsystems are integrated in a single model that is verified through experiments to obtain its polarization curve. The hybrid subsystem is then integrated through a developed energy management scheme. For the system level modeling, multi-disciplinary design analysis (MDA) is developed for different case studies that include steady level mission and climb flights. This model is solved using a nonlinear solver function of the MATLAB tool box (fsolve). The optimization scheme for all UAV design variables uses a genetic algorithm on the outer loop of the optimization routine to search for the optimal solution using the developed MDA as a fitness function. The optimization selects wing airfoil, hydrogen tank, battery, motor, gear reduction and propeller from different possible combinations to maximize the endurance of a UAV. The combination of propellers with larger diameters and motors with lower voltage constants resulted in higher overall system performance. The fuel cell model that depends on a simple polarization curve test demonstrated good agreement with experimental results. To reduce the uncertainty in the propeller modeling, the propeller-fuselage interference is considered to produce more accurate results. The contribution of integrating the propeller interference reduces the uncertainty from 30% to less than 4%. The results of this study indicated that using a battery for climb rate mission and a fuel cell for steady level flight mission, along with the proposed energy management scheme, increased the overall UAV flight endurance by 19.4% compared to classical approaches in design methods.