The Research On Indoor And Outdoor Path Planning Based On Point Cloud Fusion And The Service Scheduling Technology Of Robot

General intelligent mobile robots should have the ability to work in indoor and outdoor environments,but path planning methods using map fusion technology are very different.The map information of the outdoor environment is known,and the path nodes required for path planning have been collected in advance,and high-precision navigation can be achieved by using only intelligent path planning algorithms.However,the indoor environment is unknown and the indoor vehicle positioning signals are poor,requiring the robot creates indoor environment maps and complete indoor path planning.At the same time,it is very difficult to solve the service scheduling problem among multiple robots by using cloud platform technology.Therefore,this thesis studies robot indoor and outdoor path planning and service scheduling technologies based on point cloud fusion.In this thesis,the 3D path planning method based on improved A* algorithm and the proposed Circular Area Search(CAS)algorithm of multi-robot services are innovative.It implements indoor and outdoor path planning and multi-robot service scheduling of intelligent mobile robots,which lays the foundation for the key technologies of intelligent mobile robot navigation and service scheduling.The main research contents are as follows:First,select a method based on feature extraction for point cloud map fusion.The fusion process includes the calculation of the point cloud normal vector,the establishment of FPFH feature histograms,and registration using a registration algorithm.In the point cloud registration phase,the Sample Consensus Initial Aligment(SAC-IA)coarse registration algorithm is used to obtain the initial transformation matrix,which solves the problem that the Iterative Closest Point(ICP)algorithm is trapped in the local optimal solution.For the problem of long registration time of SAC-IA and ICP,the KDTree data structure is used to accelerate respectively.Then,at the stage of indoor and outdoor map path planning,the fused indoor point cloud map is converted into a three-dimensional grid map in the form of an octree to reduce the time of indoor map path planning for the problem of large amount of indoor point cloud map data.Using the A* algorithm as an indoor map path planning algorithm,and aiming at the problem that the traditional A * algorithm treats the robot as an ideal point,this article extends the A* algorithm to consider the size of the robot to ensure the safety of the robot and its surrounding environment during the movement.In the outdoor path planning stage,the Gaode Map API is used as a development platform for robot outdoor path planning.Finally,the robot’s path planning is applied to multi-robot service scheduling,and a new CAS algorithm for multi-robot service scheduling is proposed.It uses theCloud platform as an intermediate processing platform and the SOA as a service architecture model.The CAS algorithm encapsulates different intelligent mobile robots into various granular services,the service applicant is the center of the service search,and the service type and latitude and longitude are the service search indicators,and the shortest distance and the scoring parameters are the optimal service screening indicators to obtain the optimal service by gradually expanding the search scope.At the same time,the simulation verification and comparative analysis of CAS algorithm are carried out.