Navigation Of An Agricultural Autonomous Vehicle Based On Laser Radar

With the development of science and technology and the improvement of the people’s living standards and the trend of population aging, the demand of agricultural mechanized production and refinement agricultural development is more and more urgent. Agricultural machinery intelligence operation is the future trend of production of agricultural mechanization. However, due to the complexity and particularity of the agricultural environment, intelligent agricultural machinery and equipment has not really universal access effectively. One reason is that many of the key technologies is not breakthrough, on the other hand is the development and maintenance costs high. In order to achieve autonomous operation of agricultural machinery, the autonomous navigation of agricultural vehicles is a key part. In order to achieve autonomous agricultural vehicles walk in the crop row, the navigation performance was studied for an agricultural robot based on laser radar between the metasequoia rows with intra-row spacing and the evergreen trees without intra-row spacing, especially in the absence of one-side crop row. The metasequoia was used to simulate the orchard and the evergreen trees were used to simulate the crop. According to the crop information from the laser radar, the different navigation methods were developed to calculate navigation path and thus to determine the lateral error and heading angle error for the different type of the crop rows. Then the fuzzy logic control was used to control the robot to travel along the calculated navigation path. Tests were conducted in different three test scene such the corridor, the metasequoia trees and the evergreen trees. In order to achieve this object, the mainly jobs in this paper are as following:1. The calibration and debugging of laser radarBy the calibration of the navigation system, the installation error is examined and the measurement error of laser radar was verified. In order to prepare for the following data acquisition and processing, debugging and parameter configuration for this laser radar were executed.2. Data acquisition and processingThe laser radar sensor is connected to the computer through the serial port. In order to obtain the data which meet the requirement, the matlab software was used to program data acquisition and processing, and then data acquisition and data receiving commands were sent to the laser radar and the collection data were processed and analyzed by the PC.3. The study of autonomous vehicle navigation controlThe different positioning methods were studied based on specific crops. By information extraction of the crops, the crops positioning was finished. The principle of least squares fitting method was used to fit the navigation path and the fuzzy control method was used to achieve the autonomous vehicle navigation control.4. The communication between PC and microcomputerThe PC and microcomputer was connected via serial to finish data transmission. The corresponding program was programmed in matlab software in order to send the navigation input command from the PC to the microcomputer and then the navigation control output parameter which was calculated based on the navigation input parameter was sent to the executive body belongs to the agricultural robot.5. Test and result analysisThe navigation test was conducted in order to verify the navigation performance of the agricultural robot. In the test, two parts were concluded such as the corridor scene and the outside scene. There were two test scene in outside scene such as the metasequoia rows with intra-row spacing and the evergreen trees without intra-row spacing, especially in the absence of one-side crop row. In order to verify its navigation performance, in different test scenes, several tests were conducted under the condition that the heading angle of agriculture robot was sent to 0° and the initial center line of the agriculture did not coincide with the center line of the crop line. Finally, the test results were analyzed. Results show that multifunctional agriculture robot can achieve autonomous navigation basically. With the 0.2m/s velocity of the robot, the maximum lateral deviation of metasequoias test scenarios and the evergreen trees test scenarios are 28 cm and 17.5 cm respectively on condition that the heading angle was set to 0° and the lateral deviation coincided with the crops center line.