Terminal Constraint-Free Model Predictive Longitudinal Control for Unmanned Ground Vehicles With Driving Force Table

In this paper, a hierarchical architecture is presented on the longitudinal control of unmanned ground vehicles (UGVs). A terminal constraint-free model predictive control (MPC) algorithm is developed as the upper controller to calculate the desired acceleration. For the lower controller, through modeling experiments, we establish a driving force table (DFT) to represent the relationship among throttle opening, speed and driving force of UGVs. Moreover, in order to improve the disturbance rejection capability, a DFT-based feedforward controller is proposed to convert the desired acceleration into the expected throttle opening. The modeling errors and external disturbances are integrated and an extended state observer (ESO) is designed for estimation. Besides, the recursive feasibility and stability of the MPC without terminal constraint are guaranteed through rigorous proof. Several experiments are conducted to show the effectiveness of the proposed strategy.