Estimation of Probability of Exceeding SC-VTOL Performance Requirements During Automatic Landing Using Subset Simulation

Abstract

With the rising interest in urban air mobility, new regulations are released to enable the certification of electric Vertical Take-off and Landing (eVTOL) vehicles. These novel performance requirements pose both technical challenges and opportunities for new design approaches. In this paper, we apply a Markov chain Monte Carlo method called subset simulation to the flight dynamic model of a prototype eVTOL vehicle to estimate the probability of violating performance and operational criteria. These requirements are derived from the means of compliance for the EASA 'Special Condition for small-category VTOL aircraft'; and 'Prototype Technical Design Specifications for Vertiports,' with additional aspects from the EASA 'Certification Specifications for All Weather Operations' transferred to eVTOL aircraft. The simulations are conducted with a closed-loop experimental flight guidance and control system for automatic landing in combination with varying parameters in the environmental and aircraft models. The computation time to estimate the failure probabilities is reduced compared to pure Monte Carlo methods by utilizing a subset simulation framework. Initial results show the sensitivity of exceeding limit thresholds to the parameters varied in the flight dynamic model.