Abstract
Keywords:Race line optimisation, setup variation, motor sports, vehicle dynamics simulation, track model
The amount of computer simulation in the field of vehicle development is increasing since years. It plays an essential role in the area of passenger car development as well as in motor sports and racing car development. In both application areas a shift of development effort can be seen from classic prototyping to virtual testing via numerical simulations. Besides economic questions which are the main moving force behind this trend, there is one more reason especially in motor sports. Motor sport associations publish new and more restrictive regulations every year. Among other things the number of test runs has been limited several times. Because of this and an unmanageable variety of setup possibilities of a modern racing car, the teams need tools to virtually adjust the setup of their cars to the respective race track before they do the first "real" test run. To satisfy this needs the AVL-List GmbH developed AVL-VSM, a vehicle simulation system that provides high modelling accuracy, as well as the possibility to automatically vary the simulation vehicles properties in order to calculate the desired optimum setup.
Such simulation systems do not only need an accurate description of vehicle and track, but also the coordinates of the racing line, which is different for each vehicle class (touring car, formula car, ...). Methods to calculate the curvature of the racing line only from telemetric data lead to insufficient results.
For this reason the Virtual Vehicle Competence Center and the AVL-List GmbH developed the RLO - Race Line Optimiser, a tool that calculates the racing line dependent on the vehicle setup and the geometry of the race track.
This paper gives a more detailed view on the problems mentioned above. The focus is laid on the functionality of the RLO, which uses the Optimal Control Theory for its path planning task. An Optimal Control application needs a model of the system, which's behaviour should be optimised. For that reason the RLO needs a simplified model of the simulation system's vehicle and race track. The structure and functionality of the RLO represents the main part of this paper. Finally an example of the whole process will show the way from 3D photogrammetric track data to the RLO racing line.