Comparative study of velocity and computed torque control schemes for a differentially driven automated vehicle

This paper presents the development and performance of a computed torque control scheme for velocity control of automated vehicles. The required torque is pre-computed using a dynamic model of the vehicular system taking into account the effect of various factors that contribute to poor tracking performance of automated vehicles. Some of the factors that need to be considered are changes in the normal forces due to payload variations and changes in load distribution, forces due to lane change manoeuvres, forces and moments as a result of road curvature and gravitational effects while travelling uphill/downhill. Forward and inverse dynamic models are presented to compute the motor current required to drive a differentially driven vehicle at the desired linear and angular acceleration. Simulation and experimental results illustrate that the dynamic model based computed torque control scheme performs well compared to the conventional velocity control schemes based on vehicle kinematics