Customer Oriented Vehicle Dynamics Assessment for Autonomous Driving in Highway 2019-01-1020
Nowadays, thanks to continuous improvement in IT, carmakers spend great efforts on studies regarding Autonomous driving. Autonomous driving is the most advanced and focused research area for what concerns the automotive field and Carmakers are mainly investing money in this field but in most cases the analysis is centered towards the efforts on control logic, legal issues or possible benefits of autonomous driving.
This paper’s objective is to evaluate the vehicle dynamic response of an autonomous driven vehicle, in particular, evaluation of vehicle/driver performance for a highway overtake maneuver. Obviously in this new scenario the vehicle will be driven autonomously, so performance evaluation means comfort evaluation. For this reason, the acceleration magnitude of peaks must be observed in a complete different logic eventually with new thresholds, as well as new magnitudes. Lateral acceleration variation (jerk) for example is going to be a key parameter for this evaluation. RMSs remain interesting indicators for comfort like the best practice for NVH evaluations, but in this kind of maneuvers horizontal accelerations should also be considered.
In this very wide and complex field, a lot of variables may interfere with the car’s behavior so at the beginning the most effective solution is to concentrate on a simple overtake maneuver in a more predictable environment like a straight motorway.
While the right parameters have been analyzed and evaluated, an optimum trajectory has been defined as a reference to evaluate and understand what could be the minimum physical (or mathematical) limit of the system. In this study, in particular, bunches of minimization algorithms have been developed. Different cost functions have been evaluated for the definition of the smoothest possible trajectory and in order to choose the best result performance indicators should be properly used.
Massimiliana Carello, Alessandro Ferraris, Lorenzo Bucciarelli, Silvio Data, Giovanni Gabiati
Politecnico di Torino, Fiat Chrysler Automobiles, Centro Ricerche Fiat