Overtaking assistant algorithm based on model predictive control

It is generally accepted that an anticipatory driving style can yield a substantial contribution to improve efficiency and safety of transport. However, such an anticipatory driving style requires high attention of the driver to assess the traffic situation accurately and presume the correct reactions. Considering country road traffic a high risk is present keeping in mind the high velocities as well as the oncoming traffic participants. This work focuses on overtaking maneuvers on country roads, and a stochastic model predictive control (MPC) algorithm is implemented to manage this task. The behavior of the surrounding vehicles is predicted stochastically by using Bayesian networks. Simulation results show that the control algorithm performs safe maneuvers for all the observed scenarios. Finally, the performance of the controller is compared to the human driver, with a study in a stochastic traffic environment. The results show that the control algorithm provides the safest trip in an acceptable travel time.