Research On Fusion Algorithm Of Dynamic Path Planning For Mobile Robots

With the rapid rise of emerging technologies such as unmanned driving and artificial intelligence,many experts and scholars have shown unprecedented enthusiasm for the research of robotics.As an indispensable key technology in the field of robotics research,path planning has been widely used in intelligent logistics,smart home,maritime search and rescue,cruise and other fields in recent years.A robot with excellent path planning ability can not only quickly plan a safe and smooth path in an environment full of obstacles,but also ensure that the path is optimal on a certain evaluation index.This paper takes mobile robot as the research object,and conducts in-depth research and analysis on the global path planning problem of A*algorithm in static obstacle environment and the local path planning problem of dynamic window approach in dynamic obstacle environment.The two algorithms are correspondingly improved and integrated to make the advantages of these two algorithms are complemented,then the problems of poor real-time obstacle avoidance ability and poor global search ability that cannot be solved by a single algorithm have been solved.The main contents of this study are as follows:(1)The basic theory and implementation steps of the standard A* algorithm is analyzed in detail.Aiming at the problem that the standard A* algorithm frequently calculates a large number of meaningless nodes in the path-finding process,which reduces the efficiency of its path planning,this paper proposes to optimize it by introducing a search strategy of jump points search.The optimized A* algorithm no longer performs invalid calculations on a large number of meaningless nodes in the path search process,but only performs a jump search for a small number of selected jump points,thereby effectively improving the search speed of the global path.And through the comparison and analysis of multi-dimensional simulation experiments on the path planning process and performance indicators of the A* algorithm before and after optimization,the effectiveness and generalization of the optimized A* algorithm are proved.(2)Aiming at the problems that the dynamic window approach is easy to fall into the local area and cannot complete the navigation in the path planning process,the dynamic window approach is improved,and a new azimuth angle evaluation index is designed.It enables the mobile robot to fully evaluate the global path nodes and target points at the same time during the movement process,which makes up for the shortcomings of the dynamic window approach in global path planning.At the same time,this paper also proposes a fusion algorithm based on the optimized A* algorithm and the improved dynamic window method.The global path jump points obtained by the optimized A* algorithm are integrated into the improved dynamic window approach,which effectively improves the path planning performance of mobile robots.(3)A software and hardware experimental platform is built based on the ROS system,and three different practical experimental environments,such as simple and complex static obstacle scenes and dynamic obstacle scenes,are designed.Using the established ROS-based experimental platform,the SLAM environment map was constructed for the above experimental environment.On these three actual environment maps,the static and dynamic physical verification experiments of the path planning fusion algorithm based on the A* algorithm and the dynamic window method are carried out respectively,and the effectiveness and feasibility of the fusion algorithm in the static environment and the dynamic environment are verified.