Abstract
Vehicle dynamics simulations are widely used by automotive industry and race teams around the world to solve a magnitude of problems. One key aspect of these simulations is the driver’s vehicle control which has significant influence on the behavior of the entire system consisting of driver, vehicle and race track. This study’s aim is to provide a set of criteria to describe the individual driving style from the driver’s control inputs to the vehicle. Vehicle data recorded during race weekends is first processed to identify specific events and patterns during cornering of a race car. Following various metrics for the description of driving style are calculated. A driver detection algorithm is elaborated using supervised learning to prove relevance of the aforementioned objective metrics. The developed criteria can be used to successfully identify a driver by his interaction with the race car thus partially describe his driving style. The results of this study apply solely to a race car being driven at the limits of the vehicle’s dynamic capabilities which is a prerequisite for analysis of the control inputs of the driver by the discussed method.