Driving behavior has an important impact on vehicle fuel consumption and traffic safety. Firstly, driving behaviors were classified using the derivative data of speed, including acceleration, jerk and snap. Because of the discreteness and low resolution (1 km/h) of speed data from On-board Diagnostics, cubic spline interpolation was used. The distributions of acceleration, jerk and snap were visualized, based on which, a quantitative description index of drive cycle was given. Then the relationship between the quantitative indicators of driving behaviors in drive cycle and the corresponding fuel consumption was analyzed. Finally, application based on. NET platform was developed to make it easy for the driver to know his driving behavior and adjust accordingly in real time while driving. The application can read and parse the data from On-board Diagnostics in real time, such as speed, engine speed, fuel consumption in 100km, instantaneous fuel consumption, and mileage. Meanwhile, the application performed cubic spline interpolation and data visualization and provided driving suggestions based on optimal energy consumption. The results of this study will also assist autonomous vehicle engineers in designing more energy-efficient driving behaviors for the controller of the autonomous vehicles.

CICTP