Trajectory tracking of autonomous vehicles via model predictive control aided by sliding mode control

Abstract

In this work, we propose a composite control system for stable and robust trajectory tracking of autonomous ground vehicles (AGVs) in the presence of bounded disturbances and uncertainties. A nominal model predictive control (MPC) system is combined with a sliding mode controller (SMC) to formulate the proposed control system under the umbrella of tube-based MPC approach, with the aim of tackling the trajectory tracking challenge for AGVs in uncertain environments. The control scheme is further modified by replacing the classical first-order sliding mode control (FOSMC) with a second-order one, the super twisting sliding mode controller (STSMC), to obtain a smooth control signal by diminishing the chattering phenomena. The proposed system's stability is analyzed and guaranteed via Input-to-State Stability (ISS) in coordination with Lyapunov stability theory.

For the first time, this combined control structure is applied to the nonlinear kinematic model of AGVs, where STSMC plays the role of an auxiliary controller in the feedback loop to handle disturbances and uncertainties that cause deviation from the nominal model. In particular, the auxiliary STSMC approach is used to produce a control action that reduces the difference between the nominal predicted states and the actual ones, with the main performance metric being the error between the vehicle's position and orientation against a desired trajectory, in addition to fulfilling all of the optimization constraints. A comparative simulation study is presented and demonstrates the effectiveness and robustness of the proposed composite control system in the presence of disturbance effects.