Due to current progresses in the field of driver assistance systems and the continuously growing electrification of vehicle drive trains, the evaluation of driver behavior has become an important part in the development process of modern cars. Findings from driver analyses are used for the creation of individual profiles, which can be permanently adapted due to ongoing data processing. A benefit of data-based dynamic control systems lies in the possibility to individually configure the vehicle behavior for a specific driver, which can contribute to increasing customer acceptance and satisfaction. In this way, an optimization of the control behavior between driver and vehicle and the resulting mutual system learning and -adjustment hold great potential for improvements in driving behavior, safety and energy consumption.
The submitted paper deals with the analysis of different methods and measurement systems for the identification and classification of driver profiles as well as with their potential to optimize both vehicle driving behavior and energy consumption on the example of a hybrid drive train.
A literature research results in a number of different approaches of evaluation, which are analyzed, linked and adapted in the publication. As a result, an evaluation of the connection between different methods of driver profile determination is given. Data collection and interviews have been performed during twenty test drives on a defined route profile with different measurement systems and methods. The acquired data form the basis for a comparison and an analysis of a comprehensive driving style classification. Subsequently, a framework for computer-aided investigations of the influences of driver behavior on the control of drive trains is established by use of an existed simulation model of a hybrid drive train. Finally, a driver model is implemented based on the learnings out of analyzing the measurements and surveys. The evaluation of the measurement campaigns delivers detailed information about vehicle longitudinal acceleration behavior in different driving scenarios. This information is used to classify the individual driving styles into the types calm, normal and aggressive. This driving style-related information can be integrated into the control strategy of a hybrid power train to support operation strategy optimization regarding both driver satisfaction and reduction of energy-, respectively fuel consumption.