Detecting soil parameters from a small tracked vehicle

Abstract

Tracked robots have been studied for years, to high-level applications, such as autonomous navigation in areas of difficult access and maneuvering strategies that allow better performance compared to other types of vehicles. The full operation of a robotic system depends on ensuring the functioning of all its parts, which can be divided hierarchically or functionality. There is a consensus that the kinematic model can be simplified for most analyzes. Likewise, the dynamic model is greatly simplified because of its complexity and uncertainty, but not allowing use of models that approximate the actual behavior. The dynamic approach proposed in this work includes a lateral drag force, causing the tracks moves in curvilinear, considering their non-linearity and strong coupling between the tracks. The traction performed by the vehicle is from electric DC motors. Also it is presented the implementation of a procedure for detecting parameters of the interaction between the tracks and the soil, by estimating the coefficients of kinetic and dynamic friction, from the measurement of current supplied to the engines while performing a maneuver.