The Key Technology Development For Autonomous Localization Of Moving Agents

Moving agent can perform tasks without human intervention or less intervention with high flexibility and autonomy,which can help people complete the dangerous or repetitive work.Because of the diversity and complexity of the environment,the ability to determine their own position by sensing the surrounding envirment of moving agent is essential,this is the premise and key to realize complex functions and perform multiple tasks.Therefore,it is very important to make clear the relative position of moving agent in the environment.In order to solve this problem,this paper mainly focuses on the key technologies of autonomous localization of moving agents.Firstly,the autonomous localisation platform of moving agents is built.Then the SLAM method based on laser radar and SLAM method based on visual sensor are analyzed and developed.The experimental result shows that the method can basically satisfy the need of autonomous llocalization of moving agents.The specific research content of this paper is divided into the following parts:1.Firstly,the background and significance of autonomous localisation of moving agents are investigated and analyzed,and the research status of agents localisation is analyzed.It is clear that the SLAM method is an important key technology to solve the problem of autonomous localisation.2.The autonomous localisation platform of moving agent was built.In order to verify the key technologies of autonomous localisation,according to the different functions of the platform is divided into motion behavior layer,environment perception layer,core algorithm layer and communication layer of four layers,and build the autonomous localization platform.3.The key technologies of autonomous localisation based on laser radar are researched and developed.Firstly,the related theory algorithms of laser SLAM are studied and analyzed.On this basis,the initial map is established by bicubic interpolation method,and the Levenberg-Marquardt method is used to solve the least square problem to complete the scan matching.Finally,the positioning effect of laser SLAM is verified by experiments,and the pose estimation of mobile robot and the establishment of environment map are realized.4.The key technology of autonomous localization based on vision sensor is researched and implemented.Firstly,SLAM method based on RGB-D sensor,realizes the real-time appearance of dense mapping and pose estimation based on working principle;then with monocular vision equipment and semi direct visual odometry is studied,finally realizes the real-time estimation of camera pose transform,verify the reliability and validity of the algorithm.