A health evaluation method of multicopters modeled by Stochastic Hybrid System

For multicopters, failures may abort missions, crash multicopters, and moreover, injure or even kill people. In order to guarantee flight safety, a system of prognostics and health management should be designed to prevent or mitigate unsafe consequences of multicopter failures, where health evaluation is an indispensable module. This paper proposes a health evaluation method of multicopters based on Stochastic Hybrid System (SHS). In the SHS model, different working conditions (health statuses) of multicopters are modeled as discrete states, and system behaviors of different working conditions are modeled as continuous dynamics under discrete states. Then, the health of multicopters is quantitatively measured by a definition of health degree, which is a probability measure describing an extent of system degradation from an expected normal condition. On this basis, the problem of multicopter's health evaluation is transformed to a hybrid state estimation problem. In this case, a modified interacting-multiple-model algorithm is proposed to estimate the real-time distribution of hybrid state, and evaluate multicopter's health. Finally, a case study of multicopter with sensor anomalies is presented to validate the effectiveness of the proposed method.