Research on the Optimal Route Recommendation Algorithm System for Automatic Driving by Computer Artificial Intelligence Technology

In order to meet the requirements of fast driving in a dynamic environment, the existing unmanned vehicles generally use the method of adding local planning between the two-layer structure (path planning {path tracking) of the traditional planning system. Obstacle avoidance to reduce time-consuming global path re-planning. Based on this research background, this paper studies the computational performance of the Multi-layer Grid Neighborhood A* algorithm (MSA*) and the neighborhood grid layer by performing path planning on a random map. impact on it. In the numerical experiment, the path planning is carried out from A* to the 8th floor MSA*, and the path planning results are compared and analyzed. The numerical test results show that the multi-layer A* algorithm can improve the automatic driving planning path of the A* algorithm, but it also increases the calculation time. Considering the calculation time and the planning path comprehensively, it is suggested that the 4-MSA * algorithm can be used in the actual vehicle autonomous driving path planning problem.