| With the rapid development of fields such as artificial intelligence and deep learning,robotics has gradually become an important part of economic development,people’s lives and social governance.In recent years,with the continuous increase in the scale of data centers and the number of data power computer rooms,the power computer room has become an important part of the stable work of the data center.However,due to indoor environmental factors,battery over-discharge,battery failure and other reasons,the stable operation of the power room is affected and electrical accidents are caused.Therefore,regular inspections of the power room are very important.However,there are a series of problems in the traditional inspection method,such as heavy workload,slow update of information in the computer room,paper-based work records,etc.With the rapid development of robot technology,robots will gradually replace manpower,and unmanned inspection has become the development trend of power system inspection.This thesis takes the power room inspection as the research background,researches and develops the indoor inspection robot system,which includes the inspection robot hardware,related algorithms and equipment monitoring website,and the inspection robot uses ROS(Robot Operating System)framework high-precision laser radar It is used to establish indoor maps and positioning and navigation,use multi-sensor fusion SLAM(Simultaneous Localization and Mapping)algorithm to assist lidar to obtain rich indoor environment information,and correct the robot pose by recognizing Apriltag two-dimensional code information,using the improved A~*(A-Star)algorithm is used as the robot’s global path planning algorithm to complete indoor navigation.In order to ensure the robot’s long-term work,infrared rays are used to realize the robot’s search for charging piles and complete autonomous charging.The cloud data platform equipment monitoring website is built using the popular Vue and Spring Boot frameworks,and uses visual components such as charts and pie charts to dynamically analyze the data collected by the inspection robot and display it on the page in real time.The hardware part of the robot includes infrared thermal imaging,high-definition camera,ultrasonic module,RGBD camera,charging pile,charging navigation module,etc.,and realizes the autonomous inspection of the robot indoors through multi-sensor information fusion.The main research content of this paper is as follows:Analyze and study the disadvantages of the traditional A~* algorithm,respectively improve the traditional A~* algorithm by selecting the heuristic solution method,optimizing the evaluation function,and dynamically weighting the A~* algorithm.The navigation path obtained by the improved A~* algorithm is passed through the Besser The Curve smoothes the planned path.Compared with the traditional A~* algorithm,the improved and optimized A~* algorithm has the advantages of small amount of calculation,fast calculation speed,smoother navigation path and more conducive to robot movement.An algorithm for correcting the robot’s pose by recognizing the Apriltag two-dimensional code is proposed.The angle information of the robot relative to the position of the charging pile is obtained by recognizing the angle of the two-dimensional code,and the distance information between the robot and the charging pile is obtained through the ultrasonic module at the tail of the robot.The data is processed and converted to convert the pose and distance information into world map information,and the robot position information is republished to correct the robot pose.A multi-sensing information fusion method is proposed,which obtains rich indoor environmental information by fusing the data information of lidar,ultrasonic,and depth camera sensors to establish reliable resolution grid map information,and detects special indoor obstacles through multi-sensors,effectively Avoid obstacles to prevent robots from colliding. |
PDF Full Text Request