Evaluation of Path Tracking Performance of a Self-driving Tracked Vehicle

Tracked vehicles have been utilized for numerous agricultural projects albeit in an un-structured manner. Recently, self-driving technology has been developed for such tracked vehicles. However, conducting experiments on these self-driving tracked vehicles is expensive and time consuming. Moreover, it is difficult to consider different soil characteristics appropriately, due to the low repeatability of experiments. Therefore, to overcome such difficulties, a co-simulation environment that reflects the characteristics of both the soil and the self-driving vehicles is required. In this study, a co-simulation environment has been developed for the RecurDyn multibody plant model and the Simulink control model to evaluate the self-driving performance of a tracked vehicle. Path-following simulations for straight run, turning, and working path were tracked under various soil conditions. Furthermore, using these simulation results, a few tracking performance indices for self-driving tracked vehicles have been proposed.