Universal Path Planning for Autonomous Vehicles Considering Real-Time and Obstacle-Avoidance Performance

Effectively handling various scenarios with autonomous vehicles involves the problem of planning paths. Starting from the practical application, we propose a universal path planning method, which can respond to the dynamic traffic environment in a timely manner. The method achieves long-term driving behavioral objectives through the construction of a linear optimization framework that combines a reactive obstacle avoidance strategy and spatial discretization techniques. As for the obstacle avoidance strategy, we introduce the exogenous parameter “time” to realize the consideration of bistatic (static and dynamic) obstacles, while the spatial discretization of path solving is realized by determining the optimal end position through constructing a dynamic objective function. The feasibility and validity of the proposed planning method are verified by simulation experiments.