Abstract-High speed unmanned ground vehicles evolving on natural terrains can exhibit a significant slip and skid. An estimation of both friction and traction forces can allow to achieve a better control. In order to implement a control architecture based on the vehicle dynamic model and the wheel-soil interaction model, the knowledge of the wheels slip rate is required. The wheel angular velocities can be precisely measured. But the true measurement of the ground speed of the vehicle is much more challenging. A low-cost Doppler radar is used, in conjunction with an accelerometer, to obtain the ground speed. Thus, the knowledge of the slip rate allows us to set up an in-situ procedure for the estimation of soil parameters that is based on the measurement of the motors torques. A wheel slippage controler has also been implemented, which is a first step toward high-level dynamic control.

Abstract-A model-based control for fast autonomous fourwheel mobile robots on soft soils is developed. This control strategy takes into account slip and skid effects to extend the mobility over planar granular soils. Each wheel is independently actuated by an electric motor. The overall objective is to follow a path roughly at relatively high speed. Some results obtained in dynamic simulation are presented.