Behavioral Classification of Drivers for Driving Efficiency Related ADAS Using Artificial Neural Network

The driver states, driving styles and aggressiveness strongly influences vehicle control, and energy efficiency. If the driving patterns can be collected and effectively analyzed the resulting classification can greatly improve the effectiveness and design of active safety system, advanced driving assistance system (ADAS) or energy efficient control. For an efficiency oriented analysis, artificial neural network (ANN) is used to classify drivers into aggressive, normal, and calm states through three different driving inputs: vehicle acceleration, speed and throttle pedal angle. The resultant models have fairly accurate classification according to different driving scenarios, with overall accuracy of 90%. The classification can be a reminder for the drivers of their current behavior, in-order for the drivers to take necessary actions to improve the driving condition.