Research On GGRRT-Based Robot Adaptive Grid Map Creation And Path Planning

With the boom of artificial intelligence sweeping the world,the development of robots has attracted wide attention.Indoor mobile robots,as an important branch of robotics research,have attracted many scholars' attention.Map building and path planning are hot and difficult issues in the research of indoor mobile robots.At present,the technology of robot navigation in known environment is relatively mature,but there are still some key problems to be solved in unknown environment.Firstly,this paper studies the localization problem of robots in unknown environments,using beacon-based UWB localization technology to provide accurate position and pose estimation for robots.Secondly,it studies the target search problem in unknown environments,and proposes a sub-target search algorithm based on GGRRT.Thirdly,it studies the problem of map construction of robots,and proposes a new method to save the storage space of raster maps.An adaptive grid map creation algorithm based on multiway tree is proposed.Finally,the path planning problem under the known environment map is studied,and the improved MMAS global path planning algorithm is adopted.The specific research work and results are as follows:(1)Subtarget search algorithm based on GGRRT in unknown environmentThe position and pose estimation of mobile robots in indoor environment is studied.UWB positioning technology with high positioning accuracy and good comprehensive performance is adopted to obtain the accurate positioning of mobile robots.In order to avoid the blindness of searching in the unknown environment for the first time,a subtarget searching algorithm based on GGRRT is designed.This algorithm introduces the goal-guide function on the basis of the traditional RRT algorithm to reduce redundant searching and improve the planning efficiency.At the same time,three sub-target search strategies in different environments are formulated,aiming at the following three situations: no obstacles can be scanned,boundary points can be scanned and only the edges of obstacles can be scanned.The corresponding rules are set to determine the location of sub-targets respectively,and gradually guide the robot to move towards the final target point.The sub-target search strategy overcomes the local minima problem that often occurs in the process of environmental exploration.In this way,the robot can reach the target point smoothly without the environment map,and complete the process of exploring the environment.(2)Adaptive grid map creation algorithmAiming at the problem that the traditional uniform scale grids occupy a large amount of storage space in the process of map creation,an adaptive grid map creation algorithm for mobile robot based on multi-fork tree is proposed.The algorithm can segment and represent the environment adaptively according to the complexity of obstacles.Initialize the unknown environment as a partially occupied grid,and then divide the grid into nine-palace structure.Each grid is represented by one of the three states of full occupancy,partial occupancy and vacancy.The partially occupied grid would be further subdivided.Repeat the above segmentation process until the whole map has only two states: vacancy and full occupancy.Then the map has been created.Compared with the traditional grid,the proposed algorithm saves an order of magnitude of storage space.Compared with the quadtree-based grid,it has advantages in running time and storage space in dealing with a large area of office environments.(3)Global path planning based on improved MMAS algorithmBased on the self-adaptive grid map,this paper carries out global path planning.MMAS ant colony algorithm has better robustness and global optimization function.In order to further improve the convergence speed and optimization ability of MMAS algorithm,the pheromone volatilization factor r in MMAS are ameliorated.Parameterr is designed as a dynamic parameter in accordance with the variation of Gaussian distribution.Comparing the MMAS algorithm with the improved MMAS algorithm,the improved MMAS algorithm takes less time,has shorter path length and faster convergence speed.It achieves the purpose of improving the algorithm.Comparing the improved MMAS algorithm with the A* algorithm,A* takes less time and can find the optimal path,but it can only be used for the path planning of uniform grid map.The improved MMAS can be applied to adaptive raster map,and it also shows faster convergence and better global optimization ability.