Global path planning of unmanned vehicle based on improved A* algorithm
Hao Liang, Xiaofang Du · 2024
Aiming at the defects of the traditional A* algorithm in the global path planning of unmanned vehicles, for instance, low search efficiency, poor safety and non-conformity with vehicle kinematics principle, an improved A* algorithm that integrates JPS, virtual obstacles and cubic uniform B-spline curve is proposed in this paper. The JPS skip selection idea is integrated with the traditional A* algorithm to pare down the number of nodes needed to calculate and store in the path planning and enhance the search efficiency of the algorithm. The security of the algorithm is improved by preprocessing to increase virtual obstacles before pathfinding. Cubic uniform B-spline curve smooths the path. In this paper, MATLAB software is used to simulate on the raster map of 10*10, 20*20, 50*50 and 100*100. The simulation time of the improved algorithm is pared down by 72.91%, 50%, 32.41% and 36.5% respectively, the number of collisions is 0, and the curve and path are smoother. It is attested that the improved A* algorithm in this paper can enhance the efficiency of path search, ensure the safety of the path, and also make the path smoother, which can be better applied in the global path planning of unmanned vehicles.