2009 Volume 71 Issue 2 Pages 2_66-2_73
The main objective of this research work is to develop an autonomous crawler-type agricultural vehicle through the application of fuzzy control principles. A virtual steering angle was introduced in order to simplify the parameters of the control system from a two-input system into a single-input system. Heading angle and calculated values of lateral deviation were employed within if-then rule conditions as input variables to the fuzzy controller, while the virtual steering angle was set to output. A genetic algorithm was then introduced to optimize the parameters of the fuzzy controller. Experiments on lane change maneuver with the vehicle running on concrete surface were carried out to evaluate the viability of the control strategy. Initial results were fundamentally successful, although a slight delay in obtaining a specific control value for both heading angle and steering angle occurred in the process.