| With the increasing size of cities in China,cable tunnels have become the main force of power transmission in large and medium-sized cities,and undertake the important task of power transmission in cities.In order to ensure the safe and stable operation of the equipment in the cable tunnel,the conventional way is to manually inspect the cable tunnel regularly,but with the increasing scale of the cable tunnel,the manual inspection way is gradually unable to meet the requirements.With the development of smart grid,more intelligent inspection methods have come into being,such as intelligent inspection robot for substation and unmanned aerial vehicle(UAV)for overhead network.Because the environment of cable tunnel is very special,the intelligent patrol mode for cable tunnel is relatively few.The current inspection robot mostly runs on the track,the track needs to be erected in advance,which increases the cost,and the track erection is difficult and the robot's agility is poor.As the running environment of the newly built cable tunnel becomes better,it is very necessary to study the patrol robot for the cable tunnel without track.In this thesis,the wheeled mobile robot is used as the carrier of the cable tunnel inspection robot.In view of the long and narrow operation environment of cable tunnel,in this thesis,the positioning and navigation method of inspection robot is deeply studied,especially the localization of robots in a closed underground environment.The main contents of this thesis are as follows:(1)First of all,this thesis studies the positioning and navigation method of the current cable tunnel inspection robot.In view of the fact that most of the cable tunnel inspection robots currently use track to locate and navigate,this thesis takes the inspection robot based on wheeled mobile robot platform as the research object.In this thesis,a positioning and navigation method suitable for cable tunnel environment is proposed.(2)In this thesis,the kinematic characteristics of the mobile robot are analyzed and the kinematics model of the robot is established.The principle and application environment of various sensors and positioning methods are discussed,which provides the basis for the positioning and navigation of the patrol robot in the cable tunnel.(3)The mileage model and inertial sensor model are established,and the sensor information is pre-processed.The Calman filter algorithm is used to complete the estimation of the robot pose.And the IMU is used to complete the actual test of robot pose estimation.(4)In this thesis,the problem of robot positioning is analyzed from the perspective of probability.According to the error characteristics of each sensor model,the sensor information is selectively used,and the multi-sensor information fusion is realized in a tightly coupled way.A localization algorithm of patrol robot based on extended Calman filter is constructed.(5)In this thesis,the system of the cable tunnel inspection robot is realized.The positioning and navigation system of inspection robot is realized with ROS system as the framework.The location effect of multi-sensor fusion of inspection robot is measured and analyzed by visualization technology,which fully proves that multi-sensor fusion can provide better location estimation for robot. |
PDF Full Text Request