Download files
Abstract
Investigation of Unmanned Agricultural Ground Vehicles (UAGVs) has been on the in- crease in recent decades as UAGVs have great potential in agricultural applications, and thus UAGVs are expected to rule farms in the future. The use of UAGVs benefits from replacing human operators to do tiresome as well as hazardous work, thus reducing the risk. It can also significantly improve the efficiency to solve food shortages due to the dramatic growth of world. Path Tracking has been an important topic in the development of UAGVs, however the automatic guidance of farm vehicle becomes more difficult and challenging than that of standard mobile robots as farm vehicles are subjected to significant disturbances due to rough terrain and ground engaging operations. The controller of autonomous farm vehicles is required to be sufficiently robust to both guarantee high path tracking accuracy and stability.
This thesis mainly researches path tracking control methods for three kinds of UAGVs in the presence of significant wheel slip. In path tracking, UAGVs are guided to follow a de- sired path from an initial position while the controller keeps minimizing offsets with respect to the reference path. To achieve the accuracy required in agriculture, this work utilizes offset models derived based on kinematic models, which take both lateral and longitudinal slips into account. A model predictive controller with receding min-max optimization is then proposed to address the problem of wheel slip in UAGVs. This adaptive min-max model predictive controller provides both robustness and adaptation for path tracking.
The superior performance of the proposed controller is verified by kinematic simulation, dynamic simulation as well as field testing, compared to that of the classical model predictive control. Then, the proposed controller is also compared with a well-known robust sliding mode controller and a well-performing backstepping controller, which is carried out by implementing controllers on a UAGV at Elizabeth Macarthur Agricultural Institute. Results from simulations and experiments validate that the proposed adaptive min-max model predictive controller ensures the required accuracy and robustness in the presence of wheel slip without any slip measurement or estimation.
Page vii