Feasible path planning algorithm for city road network

Abstract

Because of the huge interest from the researchers and the industry, autonomous vehicles have become one of the most advancing research fields. Respectively, the future of autonomous vehicles, i.e., the realization of the envisaged benefits, such as increased driving safety or improved transportation efficiency and public acceptance, will depend on their ability to safely move in real road networks and various traffic scenarios. Due to this reason, this paper presents an effective path planning algorithm to generate a feasible and safe path in a city road network. The proposed algorithm is based on the Dubins path approach, a network map of a selected road and a set of waypoints, which represent intersections and their geometry. In order to ensure path feasibility and safety in a city road network, an objective function with inequality constraints is proposed. The performance of the proposed algorithm was demonstrated while applying the algorithm to different road networks of a selected city. The feasibility of the planned path was proven by executing experimental drives with an autonomous test vehicle in an enclosed lot, while imitating the moving in a city road network.