基于激光雷达的农业自主车辆导航研究

发布时间：2018-01-08 23:12

本文关键词：基于激光雷达的农业自主车辆导航研究　出处：《南京农业大学》2014年硕士论文　论文类型：学位论文

【摘要】：随着科技的发展、人民生活水平的提高、人口老龄化的加剧,农业机械化生产、精细化农业发展的要求越来越迫切。农业机械智能化作业是未来农业机械化生产的趋势,但是由于农业环境的复杂性和特殊性,智能化农业机械装备还没有真正有效的普及使用,一方面是因为很多关键技术没有突破,另一方面就是研发和维护成本偏高。为了实现农业机械的自主作业,农业车辆的自主导航是关键部分。为了实现在庄稼行中农业车辆自主行走,本文以农业机器人为平台,利用激光雷达传感器研究农业机器人在有行内距的水杉树与无行内距的冬青树中且一侧缺失一段距离的情况下的导航性能。水杉树模拟果园果树,冬青树模拟庄稼。根据激光雷达传感器获得的庄稼行信息,机器人控制系统进行导航路径计算,确定机器人的横向偏差与航向偏角,然后利用模糊控制算法实现机器人的导航控制,最后进行导航性能试验。导航性能试验时首先在走廊进行,然后在水杉树和冬青树环境下进行导航行走试验。为了实现自主车辆导航,本文完成的工作有下面几点：1.LMS219一S05型激光雷达的标定与调试通过对导航系统的标定,检验激光雷达安装后的安装误差,验证激光雷达本身的测量误差；对LMS291-S05型激光雷达进行调试与参数配置,为下一步数据获取与数据处理做准备。2.数据的采集与处理激光雷达传感器通过串口与电脑连接,利用matlab软件进行数据采集与处理程序编写,通过matlab发送数据采集和数据接收命令,对采集的数据进行处理和分析,得到满足要求的数据。3.自主车辆导航控制研究针对特定的农作物,研究相应的定位方法,通过对农作物信息提取,完成了农作物定位研究。利用最小二乘法原理进行导航路径拟合。通过模糊控制的方法实现自主车辆导航控制。4.上位机与下位机通讯上位机与下位机通过串口连接实现数据传输,在matlab软件中编写相应程序,把导航控制输入量命令从上位机发送到下位机,下位机通过导航控制输入量得到导航控制输出量,然后把导航控制输出量发送到农业机器人执行机构。5.试验及结果分析对自主车辆进行导航试验,验证其导航性能。在导航试验中,进行了室内走廊模拟试验和室外导航试验。室外导航试验时,导航行走试验以有行内距离水杉树和无行内距且一侧缺失距离为4m的冬青树为识别目标,进行相应的导航行走试验。为了验证导航系统的可行性,在不同试验场景下,机器人航向偏角为0°,农业机器人中心线与作物中心线不重合的条件下进行导航试验,最后对试验结果进行分析。试验结果表明多功能农业机器人基本能实现自主导航。当机器人初始速度为0.2m/s时,机器人航线偏角为0。、横向偏差与作物中心线重合的条件下,水杉树试验场景和冬青树试验场景的最大横向偏差分别为-28cm、-17.5cm。
[Abstract]:With the development of science and technology, the improvement of people's living standard, the increasing aging population, agricultural mechanization, fine agricultural development is more and more urgent. The intelligent agricultural machinery operation is the future trend of agricultural mechanization production, but due to the complexity and particularity of agricultural environment, agricultural machinery and equipment are not intelligent the real effective usage, one hand because there is no breakthrough in many key technologies, on the other hand is the development and maintenance of the high cost. In order to realize the autonomous operation of agricultural machinery, the autonomous navigation of agricultural vehicle is the key part. In order to realize crop line of agricultural vehicle walk independently, the agricultural robot platform, using laser radar the sensor of agricultural robot in Holly from the water from the cedar lines with no line and side missing distance under the condition of the navigation performance. Simulation of water fir orchard, holly tree crops. According to the simulation of laser radar sensor to obtain the information for the crops, the robot control system for navigation path calculation, lateral deviation and yaw determination of the robot, and then use fuzzy control algorithm to realize the navigation control of the robot, the navigation performance test. The navigation performance test first in the corridor then, walking navigation experiment in water cedar and holly environment. In order to achieve autonomous vehicle navigation, the work done in this paper are as follows: calibration and debugging through calibration of the navigation system of 1.LMS219 S05 laser radar, laser radar installation error inspection after installation, verification and measurement error of laser radar itself; the LMS291-S05 type laser radar debugging and parameter configuration, get ready for.2. data acquisition and data processing for the next data With the processing of laser radar sensor is connected to the computer through the serial port, data acquisition and processing program written by MATLAB software, receives the command sent by the MATLAB data acquisition and data processing and analysis, data collection, data satisfy.3. autonomous vehicle requirements of the car navigation control research for specific crops, positioning method and relevant research by extracting the information of crop, crop location research done. By using the least squares principle of navigation path fitting. By the fuzzy control method to realize autonomous vehicle navigation control.4. PC and the communication of upper machine and lower machine to realize data transmission through the serial port connection, write the corresponding program in MATLAB software, the navigation input control command from the PC to send to the lower machine, the machine through a navigation control input by navigation control output, and then Analysis of the navigation control output is sent to the agricultural robot actuator.5. tests and results of autonomous vehicle navigation test and verify the performance of the navigation system. The navigation test, the indoor corridor simulation and outdoor navigation test. Outdoor navigation test, navigation to line distance walking test line from cedar and no water and one missing distance for 4m holly for target identification and navigation of walking test accordingly. In order to verify the feasibility of the navigation system, in different test scenarios, the robot heading angle is 0 degrees, the agricultural robot navigation test center line and crop centerline conditions, and then the test results are analyzed. The test results show that the multi-functional agricultural robot can basically realize the autonomous navigation. When the initial velocity of the robot 0.2m/s robot, route 0. angle, lateral deviation and. Under the condition of heart line coincidence, the maximum lateral deviation of the water fir experiment scene and the holly tree test scene is -28cm, -17.5cm.

【学位授予单位】：南京农业大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN958.98;S220.3

【相似文献】