Research On Archive Access Robot Based On Integrated Navigation And Vision Guided Grabbing

With the development of modern information technology,electronic archives have been used in more and more places.However,traditional paper archives still have incomparable advantages of electronic archives in terms of confidentiality,reliability,and authenticity.The existence of paper archives is still very necessary,but how to scientifically and effectively carry out a large number of paper archives Management remains a major challenge.Based on the background of file access in a large-scale paper archival storage scenario,this topic studies a file access robot based on integrated navigation and visual guidance grabbing.It focuses on the autonomous navigation of the robot in the archives and the function of visual positioning of archives in the cabinet,the main content of the paper is as follows:Firstly,Considering of the requirements of a large number of sites in archive storage scenarios and the need for flexible changes in the driving route,a navigation method combining vision and inertial sensors based on a two-dimensional code grid map was proposed.This method estimates the pose of the robot as it travels between the QR codes by using inertial sensor information,and performs steering control of the robot's motion based on the inertial guidance information.The accleration increment method and gravity offset filtering method are used to solve the problems of accelerometer zero drift and gravity acceleration offset respectively,and the fuzzy PID controller is designed for steering control.When the robot travels above the two-dimensional code,it obtains more accurate pose information of the robot by identifying the two-dimensional code features and decoding the stored coordinate information.Combined with the robot's motion model,the rough model estimated by the inertial sensor is used as the predicted value through Kalman filtering,and the pose information obtained by the two-dimensional code is used as the updated value,which combines the two types of navigation information.Experimental results show that the navigation method can meet the scenario requirements of this subject.Then,based on the difficulty of identifying the file box in the filing cabinet,a file identification and location method based on two-dimensional code was proposed.The camera's mathematical model and calibration method are analyzed,and the monocular camera positioning and binocular camera positioning are studied respectively.Experimental results show that the positioning accuracy of the monocular and binocular cameras on the files can meet the performance requirements of the application scenario of this subject.Finally,the architecture of the robot system based on multi-core ARM is studied,and the composition of the robot system is introduced.The hardware platform of the robot is analyzed,and then the software design of the robot system is described from the perspectives of operating system,video capture,multi-threading,etc.