Path planning and global trajectory tracking control assistance to autonomous vehicle

This paper presents two contributions for the path planning for motion, and convergent global trajectory tracking, which assistance to autonomous vehicle. The path planning for motion is processed by two stages: road network detection and the shortest path estimation for vehicle motion. A road network is detected using the road map images based on image-processing techniques such color filter, segmentation technique. The road map images are collected from online free charge maps services. The road network method estimates not only the shape of road network but also the directed road network, which could not be estimated by the use of only aerial/satellite images. Some lack road segments, which are not annotated by map service, are detected using satellite images based on some filter techniques. The shortest path for motion is estimated using the Dijkstra combining with heuristic based on greedy breadth-first search technique. The road network is converted to the global coordinate, which provides a convenience for online auto-navigation task. The stable and robust control method is used for global trajectory tracking to navigate vehicle motion. The results of simulation and experiment demonstrate the effectiveness of this method under a large scene of the outdoor environments.