Research on Global Path Planning based on Improved A* Algorithm

This paper looks at demand issues in path planning for mobile robots. The discussion of the A* algorithm in path planning algorithms has been important, although the algorithm is relatively computationally mature and simple, it suffers from a number of shortcomings, such as long search time, large turning angle, occupy a large amount of memory and insufficiently smooth paths, among other prominent problems. One limitation of previous research is that it usually focuses on algorithm improvement and neglects environmental map modeling. In this paper, the improved A* algorithm is combined with a positive hexagonal raster map. Environmental map modeling is the basis for path planning. This paper investigates the construction and movement directions of two different grid types, and analyzes the effects of different grid types on the final planning under the same environment and the same path planning algorithm. For the A* algorithm, the forward and reverse bidirectional cross-search strategy and dynamic weighting strategy are introduced to obtain the dynamic weighted bidirectional A* algorithm. The experimental results show that the dynamic weighted bidirectional A* algorithm based on honeycomb raster can reduce the number of search rasters in path planning, shorten the total length of the path and effectively improve the planning efficiency.