Data-driven Fast Terminal Sliding Mode Formation Control of Unknown Multi-UAVs

Abstract

This paper studies the leader-follower formation control problem of unknown multiple quadrotors unmanned aerial vehicles (multi-QUAVs) with external disturbances and unmodeled dynamics in three-dimensional space, whereby a data-driven fast terminal sliding mode formation control (DFTSMFC) scheme is studied. Firstly, for multi-QUAV systems with unavailable dynamics, the equivalent full-format dynamic linearization based data models, via nonparametric dynamic linearization technology (DLT), are established. Then, with the aid of fast terminal sliding mode control (FTSMC) strategy and data-driven control (DDC) technique, a DFTSMFC scheme is further designed, which only relies on the input/output (I/O) data of multi-QUAVs, and also improves the robustness of the QUAV systems. At last, the effectiveness of the proposed method is verified by simulation studies.