Parameter Estimation and Adaptive Design of an Autonomous Racing Car

Abstract

This thesis proposes a parameter estimation method for a scale-down autonomous racing car. The dynamic single-track model of Euler integration method is used for parameter estimation based on the measurements generated by the multi-body model in state-space form. By implementing optimization methods, the vehicle mass, moment of inertia for entire mass about z axis, distance from centre of gravity to rear/front axle, centre of gravity height of total mass and cornering stiffness coefficients of front/rear tires have been successfully estimated. The accuracies of the estimated parameters were also demonstrated. Moreover, an idea of adaptive design for the estimated parameters have been proposed in this thesis. By adjusting the selected parameters, the vehicle can achieve the performance of a complex model under a simple model.