Speed Planning for Autonomous Driving in Dynamic Urban Driving Scenarios

Abstract

Trajectory planning is essential for autonomous vehicles when operating in dynamic traffic environments. A layered approach usually separates out into path planning and speed planning. In the work reported in this paper, speed profile planning over a given path, which is defined by a trajectory planner, is proposed. The relevant information is provided by vehicle-to-vehicle (V2V) communication. First, a speed planning optimization algorithm which considers safety, time efficiency, smoothness and comfort constraints is presented. This strategy can provide a safe, comfortable and feasible speed profile for autonomous driving via a S-T graph under a complex traffic environment. Secondly, a conventional non-convex optimization problem is translated into a quadratic programming problem. This has the advantage of a low computation requirement because it uses a CFS (convex feasible set) algorithm. The effectiveness of the proposed scheme is verified through simulation studies in various urban driving scenarios. This holistic approach provides a more effective approach to speed and trajectory planning.