Research On Map Construction And Location Technology Of Patrol Robot In Outdoor Scenes

Traditional inspection robots often use GPS technology or LIDAR for localization.GPS signals can be lost due to obstruction from tall buildings or trees in outdoor environments.2D LIDAR has limited range,typically measuring distances of only 10-30 meters.Therefore,a mapping and localization method for outdoor inspection robots based on the combination of 2D LIDAR and DETA navigation module is proposed.A point cloud map of complex outdoor environments is constructed using a combination of three sensors:LIDAR,inertial measurement unit,and odometer.The DETA navigation module and 2D odometer are combined to solve the localization problem in open outdoor environments,while the combination of LIDAR and odometer is used to solve the localization problem in complex outdoor environments.The main research work is as follows:An improved Hector-SLAM algorithm based on extended Kalman filter algorithm data fusion is proposed for complex outdoor environments.The robot’s real-time state is taken as the state variable,and the laser sensor’s scanning information,odometer,and IMU pose information are taken as the observation variables.These three sensor information are fused using the extended Kalman filter algorithm to correct the laser scan point cloud with odometer and IMU information and provide an accurate initial value for the system.The Hector-SLAM algorithm’s scanning matching method is used to represent the point cloud information as a grid map.For the localization problem in large outdoor scenes,a combination of DETA navigation module and odometer is used.An error model is established based on the statistical and dynamic characteristics of the DETA navigation module and odometer navigation errors.The Kalman filter is used to separate and estimate the various error sources of the system,and the optimal state estimation value is obtained based on the estimated error values to correct the output of the odometer navigation system,thereby achieving high-precision localization.For the localization problem in complex outdoor environments,a LIDAR and odometer combination localization method based on adaptive Monte Carlo is proposed.The two sensors are combined using the ROS operating system to achieve real-time localization of the inspection robot.A control system based on Free RTOS and STM32RCT6 is built to collect position information from the odometer and gyroscope,achieve parallel execution of multiple tasks,and a control system based on ROS and Raspberry Pi is built to transmit data between the 2D LIDAR,DETA navigation module,and inspection robot,and complete information exchange between the upper and lower computers.Experiments on the inspection robot were conducted in the campus flower bed and playground.The results show that an accurate grid map of complex outdoor environments can be constructed,and high-precision positioning information can be obtained in outdoor environments.