【作者】 张大为；

【导师】 夏田；

【作者基本信息】 陕西科技大学 ， 机械工程， 2022， 硕士

【摘要】 随着人口老龄化加剧,居家老人的服务问题逐渐被重视,服务机器人作为机器人领域的重要分支,可在一定程度上缓解该问题。但现有的服务机器人存在定位与建图效果差、自主移动感知范围小、功能单一等不足,不能满足服务老人的需求。本文利用机器人学相关技术,研制了一台能够构建高质量环境地图、定位准确、自主移动、向老人投递物品的服务机器人,并对其中的重要算法与功能进行改进,具体研究内容如下:(1)结构模型与运动模型的建立。分析服务机器人的应用场景及工作特性,明确其应具备的功能及控制要求,制定实现方案。利用UG软件与数学分析法分别建立其结构、运动学与里程计模型。(2)控制系统构建。利用分布式控制技术,构建控制系统的硬件框架;采用机器人操作系统(ROS)作为控制系统的软件开发平台。以向居家老人服务的场景为例,分析服务机器人的通用工作流程,研究并设计各层级的控制程序,实现服务机器人的软件控制。(3)关键算法研究。研究实时定位与地图构建(SLAM)算法,针对原始算法存在的里程计精度低与激光雷达数据误差问题,提出基于多传感器融合技术的改进方案,有效改善了 SLAM算法的定位与建图效果。利用代价地图的图层插件融合超声波传感器完成自主移动时的盲区补偿,实现服务机器人的高精度自主移动。研究视觉标定与识别算法的原理,利用ORB图像特征匹配算法实现物品的准确识别与抓取。(4)仿真与实验验证。搭建服务机器人的实验样机,在Gazebo仿真环境内完成服务机器人主体的功能仿真,验证SLAM与自主移动功能的可行性。利用实验样机在室内复杂环境中开展SLAM实验,对比原始算法与改进算法的定位精度和建图效果,结果说明改进后的SLAM算法定位更准确,构建的环境地图效果更好。通过自主移动实验,说明服务机器人能够实现高精度的自主移动,可借助超声波传感器有效避开盲区障碍。最后进行整体功能实验,验证服务机器人是否能完成期望功能。本文以服务机器人为研究对象,分别对其结构与运动模型、控制系统、核心算法进行了详细研究,针对不足之处提出相关改进,并搭建实验样机。通过仿真分析与样机实验,说明研究的服务机器人能够完成期望功能,所提出的改进算法和优化方案有效解决了对应的技术问题,为后续研究提供了理论依据和实践基础。更多还原

【Abstract】 As the population ages,the problem of serving the elderly at home is gradually being emphasized,and service robots,as an important branch in the field of robotics,can alleviate the problem to a certain extent.However,existing service robots suffer from poor positioning and map building,small autonomous mobile sensing range,and single function,which cannot meet the demand of serving the elderly.Therefore,in this paper,a service robot capable of constructing highquality environmental maps,accurate positioning,autonomous movement,and delivering items to the elderly is developed using robotics-related technologies,and the important algorithms and functions of which are improved as follows.(1)Establishment of structure model and motion model.Analyze the application scenario and working characteristics of the service robot,clarify the functions and control requirements it should have,and formulate the realization plan.Use UG software and mathematical analysis to establish its structure,kinematics and odometry models respectively.(2)Control system construction.The hardware framework of the control system is constructed by using distributed control technology;the robot operating system(ROS)is adopted as the software development platform of the control system.Taking the scenario of service to the elderly at home as an example,we analyze the general workflow of the service robot,study and design the control program at each level,and realize the software control of the service robot.(3)Key algorithm research.Research on real-time localization and map building(SLAM)algorithm,for the original algorithm exists low odometer accuracy and LIDAR data error problem,proposed an improvement scheme based on multi-sensor fusion technology,effectively improve the localization and map building effect of SLAM algorithm.The layer plug-in of the cost map is used to fuse ultrasonic sensors to complete the blind area compensation when moving autonomously and realize the high-precision autonomous movement of the service robot.The principle of visual calibration and recognition algorithm is studied,and the ORB image feature matching algorithm is used to realize the accurate recognition and grasping of items.(4)Simulation and experimental verification.The experimental prototype of the service robot is built,and the functional simulation of the main body of the service robot is completed within the Gazebo simulation environment to verify the feasibility of SLAM and autonomous mobility functions.The experimental prototype is used to conduct SLAM experiments in a complex indoor environment to compare the localization accuracy and map building effect of the original algorithm and the improved algorithm,and the results show that the improved SLAM algorithm has more accurate localization and better environmental map building effect.The autonomous movement experiment illustrates that the service robot can achieve high-precision autonomous movement and can effectively avoid blind obstacles with the help of ultrasonic sensors.Finally,overall functional experiments are conducted to verify whether the service robot can accomplish the desired functions.This paper takes the service robot as the research object,conducts a detailed study on its structure and motion model,control system,and core algorithm,proposes relevant improvements for the shortcomings,and builds an experimental prototype.Through simulation analysis and prototype experiments,it is shown that the studied service robot can accomplish the desired functions,and the proposed improvement algorithm and optimization scheme effectively solve the corresponding technical problems,which provides a theoretical basis and practical foundation for the subsequent research.更多还原