Disturbance Observer Based Lateral and Longitudinal NMPC for Agricultural Harvester Trajectory Tracking Control

In this paper, a disturbance observer based nonlinear model predictive controller is proposed for autonomous agricultural harvester with complex environments. Firtstly, a disturbance observer based radial basis function neural network technical is designed to handle external disturbances, such as friction of tires. Then, a nonlinear model predictive controller with pseudo-Huber function is used to handle the trajectory tracking control issues of the agricultural harvesters. The simulation results show that the proposed controller has good tracking accuracy and robustness.