Parameter Identification of Vehicle Dynamics Models using CAN Communication on Real-life Driving Data

The topic of this paper is the simple acquisition of vehicle data, which is necessary for efficient generation of sufficiently precise vehicle models for the assessment of vehicle dynamics. One method is the identification of vehicle parameters using parameter identification from vehicle motion data. Although modern and especially highly equipped vehicles have a large number of sensors and control units that collect a large amount of data, these data are still very difficult to access. In this paper a method for accessing this data and for modelling vehicle dynamics based on real driving data is presented. The data is recorded exclusively via the vehicle sensors and are picked up via the DLC interface and then processed. In a first step, it is investigated to what extent a minimal data set based on driver inputs and vehicle accelerations and speeds is sufficient to model the vehicle dynamics.