Uncertainty propagation in flight performance of multirotor with parametric and model uncertainties

Abstract

This paper focuses on the uncertainty propagation in the flight performance of multirotor-type unmanned aerial vehicles from the systematic perspective in conceptual design phase. The multirotor performance is estimated by a conceptual design and analysis framework which is capable of predicting the performance for given mission profiles and multirotor specifications. In this study, not only are parametric uncertainties considered in multirotor components such as rotor, motor and battery, but also model uncertainties in aerodynamics and electric propulsion system analysis models of the framework are taken into account. The model uncertainties in the analysis methods are considered by adopting model uncertainty parameters calculated from experimental data, which is particularly important in conceptual design phase where mathematical methods have relatively lower accuracy. The performance variance under various flight conditions with different multirotor configurations is quantified using Monte Carlo Simulation with 10,000 samples determined by Latin Hypercube Sampling. It is found that the uncertainties in the electric components and circuit analysis are far more dominant than those related with the rotor aerodynamics. Also, the uncertainty effect is nearly independent of the multirotor configuration, yielding similar results in the quadrotor and hexarotor configuration.