Finite-time observer-based model-free time-varying sliding-mode control of disturbed surface vessels

In this study, the model-free robust tracking control problem of a surface vessel in the presence of unknown dynamics and disturbances is solved by creating an innovative finite-time observer-based time-varying sliding mode control approach. By developing a trend data model and a regulation index function, an efficient time-varying sliding-mode surface is devised such that the asymptotic convergence of the plant's outputs with small overshoots can be guaranteed. With the aid of data-driven strategy and finite-time control technique, a data-driven finite-time observer is further proposed to quickly estimate the total unknowns including unknown dynamics, external disturbances and couplings, and whereby a model-free time-varying sliding mode controller is thereafter synthesized such that these unknowns can be accommodated well since sliding mode can be exactly reached. Finally, the efficiency of the developed approach is evaluated via theoretical analysis and simulation studies on a benchmarked surface vessel.