Driving behavior classification for Heavy-Duty vehicles using LSTM Networks

Official URL: http://risc01.sabanciuniv.edu/record=b2315127_ (Table of contents)

Despite growing autonomous driving trend, human is still a major factor in the current vehicle technology. Drivers have a great impact on both fuel economy and accident prevention. Therefore, identi cation and evaluation of driving behaviors are crucial to improve the performance, safety and energy management of vehicle technologies, particularly for heavy-duty vehicles. In this thesis, several driving behaviors with di erent acceleration and car following characteristics are generated on a realistic truck model in IPG's TruckMaker simulation environment. A Long Short Term Memory (LSTM) classi er is then utilized to recognize driving behaviors. First, six drivers are de ned based on their longitudinal and lateral acceleration limits. The classi er is trained using driving signals acquired from the simulated truck which follows an arti cial training road with di erent trailer loads. The training road is designed to cover possible road curves that can be seen in highways. The model is tested with driving signals that are collected from a realistic road using the same method. Then, three drivers (calm, normal and aggressive) are de ned based on their longitudinal acceleration pro les in car following and the classi er is trained and tested using driving signals of these drivers in di erent tra c scenarios. Results show that the proposed LSTM classi er is capable of successfully capturing the dynamic relations encoded in driving signals and recognizing di erent driving behaviors in small time samples