Research On Mapping And Path Planning Of Camellia Fruit Picking Robot

In order to improve the intelligence of agricultural machinery in Camellia oleifera forest,reduce labor consumption,and ensure that agricultural machinery can determine its own and target position in complex and unknown environments,provide highquality maps,and plan the best path in the established map.This paper studies the mapping and path planning of Camellia oleifera planting environment,and builds a lidar crawler experimental robot based on the robot operating system.Through the comparative analysis of three indoor mapping algorithms,the mapping method is determined,and then the global and local planning are carried out,and finally the autonomous navigation of the picking machine is completed,and the intelligentization of agricultural machinery is realized.The main work is divided into the following parts:(1)According to the target analysis of the Camellia oleifera woodland robot,the sensor model used by the robot(RPLIDAR A1 is selected for the lidar)is selected,and the chassis hardware structure and inertial module design of the experimental robot machine are completed according to the walking requirements.And based on ROS to build the relevant software platform on the Linux system,design and configure the software modules of the picking machine system,configure the data of the lidar,the code of the control chassis,the code related to the mapping,the global and the best planning and other related nodes,and carry out a control strategy to control the walking speed of the picker.(2)According to the advantages and disadvantages of the mapping algorithms on the market,three commonly used mapping algorithms are determined,and the implementation principles and functions of these three SLAM mapping algorithms(specifically:cartographer,karto,hector)are explained,and the respective advantages and disadvantages.According to the experimental site,design a walking route,and make the experimental car conduct a lidar mapping experiment according to the walking route through remote operation,and record the formation of the map when the experimental car is walking.Basic build.The mapping algorithm suitable for the Camellia oleifera environment is screened by measuring the mapping speed and calculation accuracy.In terms of accuracy,three indicators are selected for calculation,namely root mean square error(RMSE),peak signal-to-noise ratio(PSNR),structure Similarity(MSSIM).The comprehensive comparison found that compared with the other two algorithms,the map construction based on the cartographer algorithm is the most suitable method for map construction in the Camellia oleifera planting environment.(3)Compare and analyze various planning algorithms,and finally determine the A^*algorithm as the research method for the optimal path of Camellia oleifera forest.First,configure the code related to the A^*algorithm,and conduct simulation experiments through MATLAB,and conduct real-time obstacle avoidance experiments for the straight obstacles and turning obstacles existing in the path.The simulation results show that the A^*algorithm can avoid obstacles well.,suitable for Camellia oleifera planting environment.Then,autonomous navigation experiments are carried out in the experimental site,and global and local field experiments are carried out.It can be seen from the results that the built crawler-type experimental trolley can be planned in the above-mentioned map,meet the requirements of obstacle avoidance in autonomous navigation,and can accurately reach the area around the Camellia oleifera tree to realize autonomous navigation.