Cloud Based Simultaneous Location And Mapping For Service Robot

Simultaneous localization and mapping are the hotspots in the field of robotics research in recent years.They are the basis of autonomous navigation and path planning for robots.The study of SLAM is of great significance to the autonomy and intellectualization of robots.Especially in the process of service robots being widely used,SLAM is very important for service robots to complete their tasks independently for service robots.Therefore,the localization and mapping of service robots are also widely concerned by academia and industry.Using the computational performance and mass storage capability of cloud robots,the pressure of SLAM on the resources of service robots can be alleviated,and a scheme for the positioning and composition of service robots can be provided.The main purpose of this paper is to alleviate the pressure of local robot SLAM task on resources,and to use the resources of cloud-based robots to solve the shortage of local robot resources.The main work of this paper is as follows:(1)Introduce the tasks of SLAM,analyze the requirements of computing,storage and real-time performance of tasks in ORB SLAM system,and analyze the feasibility of applying cloud robots to SLAM field.(2)Based on the analysis of ORB SLAM system,the SLAM scheme of cloud + robot is designed.Cloud SLAM system is built to localization and mapping in the mode of cloud and local robots.Cloud is responsible for close-loop detection,mapping and relocation of candidate frame information acquisition.Local robots focus on tracking,relocation and local map maintenance.Considering the non-real-time characteristics of ROS framework,a data transmission scheme between cloud server and local robot is designed by using Baidu RPC framework brpc.(3)The advantages and feasibility of the proposed algorithm are analyzed by using existing hardware to verify the advantages of cloud SLAM in resource utilization and network requirements,and to prove that real-time localization and mapping can be carried out based on this model.