Finite-Time Model-Free Adaptive Control for MIMO Nonlinear Systems With Adjustable Fast Convergence Domain: Theory and Experiment

In this article, the substantial finite-time tracking control problem for multiple input multiple output (MIMO) nonaffine nonlinear systems is innovatively solved by developing a novel full-form dynamic linearization-based finite-time model-free adaptive control (FFDL-FMFAC) method. Unlike previous data-driven finite-time control approaches, the fast convergence effect is considered within a larger and adjustable convergence domain. Inspired by the finite-time control technique, a new scaled tracking error vector with time-varying gain matrix is designed to address complex couplings and achieve adjustability in the fast convergence domain. Within this scheme, by devising a FFDL-based finite-time scaled cost function, an efficient FFDL-FMFAC control law with an adjustable fast convergence domain is subsequently developed. Furthermore, by constructing an equivalent linear time-varying system from the original closed-loop FFDL-FMFAC system using the equivalent system transformation strategy, it is shown that the bounded input and bounded output with finite-time convergence property can be rigorously guaranteed, and the convergence speed can be significantly improved compared to conventional model-free finite-time control methods. Finally, a practical experiment on a quadrotor unmanned aerial vehicle is conducted to validate the efficiency and superiority of the proposed FFDL-FMFAC approach.