Asynchronous Singular MJLSs Modeling and Attitude Stabilization of a Class of Hybrid Terrestrial–Aerial Quadrotors

Abstract

This study is concerned with the modeling and stabilization of the hybrid terrestrial–aerial quadrotors (HyTAQs), a class of robots with terrestrial/aerial movement capability. Affected by the obstacles or rough terrain randomly distributed in the environment, the locomotion mode of HyTAQs is generally subject to stochastic switching, modeled by a continuous-time Markov process in this work. To precisely describe the practical mode switching, a dwelling stage is introduced into the Markov process such that the sojourn time can be lower bounded. Also, allowing for the practical uncertain lag in controller switching, the asynchronous switching phenomenon is modeled under the singular Markov jump linear system framework, such that the controller is designed with robustness to the asynchrony of terrestrial/aerial mode, which is accompanied by state-dimension mismatching. Compared with existing studies, this article presents a more practical switched system model for HyTAQs and guarantees stochastic stability under the effect of both dwelling stage and asynchronous switching. Experimental results are provided to demonstrate the effectiveness and merits of the proposed approach.