Multi-Purpose Mobile Robot for Precision Farming

本研究的目的是开发农业生产自动化系统。我们的目标是建立一种农业移动机器人的技术，它能够像熟练工人一样质量地进行农业工作。研制的机器人可以在田间独自行走，并在行走过程中自动获取与土壤和作物相关的田间环境信息。1)模型车的设计与制造研制了满足移动机器人使用要求的模型车。2)地磁方向传感器导航系统的研制为了仅利用地磁方向传感器获得的定位数据对机器人进行精确控制，研制了地磁方向传感器导航系统。利用人工神经网络建立了能够补偿机器人周围地磁偏差的精化方法。3)开发…基于机器视觉的定位系统在导航系统中的应用为实时了解机器人的位置，开发了基于机器视觉的定位系统。测量原理是三角测量，通过在场地上设置两个视觉传感器，自动跟踪机器人的运动并测量距离场地基线的角度。建立了基于定位系统和GDS的导航系统，并通过田间试验对机器人的性能进行了研究。4)玉米抗性和作物生长传感器的研制开发了一种能够自动测量玉米抗性的传感器系统，用于机器人对田间土壤紧实度的评价。由于安装在移动机器人上的传感器系统可以同时获取玉米抗性和位置，因此可以创建玉米抗性空间地图，并在计算机屏幕上评估田间土壤的紧实度。此外，为了评价田间作物生长的变异性，研制了一种基于机器视觉的作物生长传感器。该传感器可以使用遗传算法创建的模糊逻辑来分割农作物和杂草区域。5)通信系统的开发为利用玉米抗性、作物生长等田间信息进行作物管理，建立了移动机器人与基站之间的通信系统。较少

The purpose of the study is to develop automation system for agricultural production. We aim to build technology of an agricultural mobile robot, which is able to conduct agricultural work in the same quality with a skilled worker. The developed robot can travel alone in a field, and automatically acquire the field environmental information related to soil and crop during the travel. The results obtained in the research project are follows :1)Design and Fabrication of Model VehicleThe model vehicle, which was satisfied with requirements for utilizing a mobile robot, was developed. The specifications and basic performance were studied.2)Development of Navigation System Using a Geomagnetic Direction Sensor (GDS)To precisely control the robot using only orientation data obtained by the GDS,the navigation system using the GDS was developed. The refining method, which was able to compensate the geomagnetic warp surrounding the robot, was built using an Artificial Neural Network.3)Developmen … More t of Navigation System Using a Positioning System Based on Machine VisionThe positioning system based on machine vision was developed to know the robot position in real time. The measurement principle is triangulation by setting the two vision sensors on the field that can automatically track the robot movement and measure the angle from the base line of the field. The navigation system using the positioning system and the GDS was build, and the robot performance was investigated by field tests.4)Sensor Development of Corn Resistance and Crop GrowthThe sensor system, which could automatically measure corn resistance, was developed to evaluate soil compaction of a field by the robot. Because the sensor system attached on the mobile robot can acquire the corn resistance and the position at the same time, the spatial map of the corn resistance can be created, and the soil compaction of the field can be evaluated on a computer screen. In addition, to evaluate the crop growth variability on a field, a crop growth sensor based on machine vision was developed. The sensor can segment crop and weed areas using a Fuzzy Logic created by a Genetic Algorithm. Then, the sensor can individually predict the crop growth such as the height and the width.5)Development of Communication SystemTo utilize the field information such as corn resistance and crop growth for crop management, the communication system between the mobile robot and the base station was built. Less