Data-driven sliding mode tracking control for unknown Markovian jump non-linear systems

This study aims to investigate the sliding mode control (SMC) problem for non-linear Markovian jump systems (MJSs) in which the transition probability could only be partly obtained and the system models and orders are unknown. Firstly, a mode-dependent data-driven sliding surface is established and by using the high-order SMC strategy, a data-based SMC law is proposed to guarantee the reachability of the SMC system. With the linear matrix inequalities technique, the tracking error of the closed-loop non-linear MJS is demonstrated to be asymptotically stochastically stable. Furthermore, a simulation experiment is carried out to prove the effectiveness of this method.