Autonomous Navigation Of Multi-Sensor Based Crawler Robot In Outdoor Environment

The special mechanical construction of the crawler robot ensures its high adaptability in the field environment.Autonomous navigation can complete the capture and decision-making of environmental information,and provide the basis for the intelligence application of crawler robot.This research aims at studying the autonomous navigation of crawler robot in unknown environment and exploring its application.The main work is divided into three parts:Firstly,in order to realize autonomous navigation,the path planning of crawler robot is studied.The artificial potential field method is selected to carry out the path planning by considering the difficulties faced by the special situation of the crawler robot and the real-time requirement of engineering application.For the problem of complex environmental shock,fuzzy decision is added for the selection of the complex environmental parameter.By means of expert experience,fuzzy decision is made to adjust the size and direction of the mobile robot at all times,which can solve the problem of uncertainty of repulsive constant,gravitational direction deviation and robot speed.Secondly,in order to achieve the accurate positioning of the crawler robot in the open area,the information fusion of GPS and odometer is adopted.But the jump or loss of GPS signal under the tree-lined road will lead to the low positioning accuracy,which pushes us to correct positioning of the crawler robot.By means of the adjacent time environment information extracted by visual system,the SIFT algorithm is used to match the same feature point and measure the landmark depth information.Then the target position information is corrected by the EKF algorithm combined with the odometer status value and the visual system measurement value.In addition,in view of the obvious changes of covariance matrix in outdoor environment,fuzzy decision is used to adjust the covariance parameter.Finally,the autonomous navigation system hardware of crawler robot is realized.The crawler robot control system is analysis in detail,including mechanical structure and component specifications,industrial control computer and DC servo drive and other processing equipment.The kinematic and dynamic mathematical model of crawler robot is established,and sensors for system environment-aware are selected to achieve the autonomous navigation in outdoor environment.