This paper established a co-simulation environment for the development of autonomous driving technology of a multi-purpose crawler platform and evaluated the simulation results. The upland or field work environment is an atypical terrain in which there are various cases depending on the soil condition and topography. Accordingly, tracked vehicles are being used for the field work, and autonomous driving technology of field work machines is being developed to automate agricultural work. However, since experiments using actual tracked vehicles are expensive and have low reproducibility, it is difficult to consider different soil characteristics for each field when developing autonomous driving technology. Therefore, in order to reflect the characteristics of the soil, a co-simulation considering multi-physical factors is required. In the co-simulation of this study, the vehicle''s multi-body dynamics model, hydraulic drive model, path-following controller model, and ground model were implemented. These models were integrated into the RecurDyn-Simulink environment. Through the simulation, it was possible to compare the track-ground interaction and work performance of the tracked vehicle according to the type of soil and the work path.
In this study, the ground was set as a simple contact model and three soil models and simulations were performed. Moreover, the dynamic characteristics of tracked vehicles were analyzed in straight run, concentric turn, work path tracking, and field driving simulations. In the straight run and concentric turn unit tests, the performance of the path-following controller and the effect of the ground on the track were confirmed. Simulation of working path driving, a combination of straight run and turn, demonstrates the required state when a tracked vehicle enters the workspace. Finally, this paper validates co-simulation in the development of autonomous driving techniques through path tracking simulations on fields. Based on the analysis of simulation results, this paper proposes an index that can validate the straight run performance, turning performance, and work performance of tracked vehicles. As a result, the co-simulation method of tracked vehicles analysis is expected to contribute to the development of agricultural automation and autonomous driving technology.