A comparison of free trajectory quasi-steady-state and transient vehicle models in minimum time manoeuvres

Abstract

Lap-time optimisation programs utilising Quasi-Steady-State (QSS) vehicle models are computationally efficient, though inherent simplifications produce inaccuracies. An analysis of the prediction capabilities of a new QSS lap-time optimisation method is presented. A transient vehicle model is formulated with seven degrees-of-freedom (DOF) dynamics. The QSS vehicle model utilises point-mass dynamics, where accelerations are constrained by a bi-quintic piecewise spline surface produced with an identical transient vehicle model. Notable changes to the development of the new QSS model include, evaluation of the acceleration envelope in natural coordinate frame, and the implementation of a normal jerk limit which acts to artificially mimic the yaw inertia resistance during change of direction manoeuvres. Results of the minimum time solution are shown for left-hand turn, double lane change and full circuit analysis based on Adria International Raceway. An accuracy improvement against previous QSS models is presented. Minimum time results for the new QSS model are within for left turn, for Adria full circuit, and for double lane change. Optimal accelerations, track centreline offset and velocities are presented and compared.