DRIVING STYLE ANALYSIS AND DRIVER CLASSIFICATION USING OBD DATA OF A HYBRID ELECTRIC VEHICLE
Andrzej Puchalski, Iwona Komorska
Abstract
Ensuring the effectiveness of adaptive algorithms for advanced driver assistance systems(ADAS)requires online recognition of driving styles. The article discusses studies carried out during real driving cyclesbased onthe GPS parameters and OBD system dataof a hybrid vehicle. The work focuses on the search for measures of the speed and acceleration signals of the car and the measures determined on their basis thatbest describe the driving style responsible for thevehicle traffic safety and ecological safety.Relations betweenthe type of driver, driving dynamics,and fuel consumptionwere studied. The driver's categorization was based on astatistical analysis of input signals and mean tractiveforce (MTF) by clustering.