Detection and Risk Analysis with Lane-Changing Decision Algorithms for Autonomous Vehicles

Abstract

Despite the great technological advances in ADAS, autonomous driving still faces many challenges. Among them is improving decision-making algorithms so that vehicles can make the right decision inspired by human driving. Not only must these decisions ensure the safety of the car occupants and the other road users, but they have to be understandable by them. This article focuses on decision-making algorithms for autonomous vehicles, specifically for lane changing on highways and sub-urban roads. The challenge to overcome is to develop a decision-making algorithm that combines fidelity to human behavior and that is based on machine learning, with a global structure that allows understanding the behavior of the algorithm and that is not opaque such as black box algorithms. To this end, a three-step decision-making method was developed: trajectory prediction of the surrounding vehicles, risk and gain computation associated with the maneuver and based on the predicted trajectories, and finally decision making. For the decision making, three algorithms: decision tree, random forest, and artificial neural network are proposed and compared based on a naturalistic driving database and a driving simulator.