Research On Automatic Navigation System Of Orchard Management Robot Platform

For raising automation level of orchard precision management, decreasing the management cost, and promoting the application of robot products, an orchard tracked mobile robot is designed in this paper. The orchard robot is composed of robot body and robot guiding mechanism. A method based on the principle of slider-crank mechanism is used in this paper. A rubber pipeline whose diameter is 4 centimeters is laid down in the orchard as the path of the robot, robot body, rubber pipeline and robot guiding mechanism will constitute a slider-crank mechanism after the robot guiding wheel contacts with the rubber pipeline. In the slider-crank mechanism, the robot body is the crank, the shaft is the linkage and the guide wheel is the slide. This robot will be used in orchard management, such as variable spraying and fruit picking. The main result of this paper is as the following:(1) A method based on the principle of slider-crank mechanism is presented in this paper, and kinematics model of the robot is built, analyses the dynamical characteristics of the robots and presents the dynamical differential equations of the system. The driving drag and driving power are calculated for the robot.(2) Putting up the robot mobile platform, including the installation of attitude angle sensor, position angle sensor, DC servo driven system and MCU control system. The software controller is realized with VC++ 6.0 and Keil uVision2.(3) The detailed hardware circuit of control system based on AT89S52 is designed, including power circuit part and the microcontroller circuit part.(4) A series of experimental and analytical research on the riding performance of the robot are conducted. The robot velocity calibrated, the reasons of the robot slip forward during the straight driving is analyzed.(5)Based on the analysis of the path characteristics, the article performs the path planning and designs the Fuzzy-PID controller of the robot. Lines and curve tracking were carried out, the experiments showed that the maximum tracking error was less than 0.02 when the robot moved straight at the speed of 0.15m/s, and the maximum tracking error was less than 0.05m when the turning radius was 2m.