【作者】 韩金华；

【导师】 孟庆鑫；

【作者基本信息】 哈尔滨工程大学 ， 机械设计及理论， 2009， 博士

【摘要】 护士助手机器人是一种在医院环境中工作、自主导航的室内移动服务机器人。可以替代医护人员进入传染病病房,执行运送药品、食品、生活用品及喷雾消毒等医护任务,在提高病房的自动化水平,保护医护人员,避免交叉感染等方面具有重要应用价值和前景。本课题来源于黑龙江省自然科学基金重点项目“护士助手机器人基础技术研究”,结合课题要求,本文旨在设计具有自主导航功能的护士助手机器人试验样机,并对其涉及的关键技术进行研究。论文综述了护士助手机器人的国内外研究和发展现状,并对移动机器人涉及的主要技术进行了概述。针对护士助手机器人的性能和技术要求,提出了一种护士助手机器人的总体方案,包括机械本体结构、驱动方式、传感器系统和控制系统。护士助手机器人采用两轮差速驱动的驱动形式,转向灵活。设计了护士助手机器人控制系统,采用双层控制结构,执行级由DSP TMS320F2812和ATmegal6单片机来完成,组织级则由PC104来实现,组织级与执行级的通讯采用CAN总线来完成。针对护士助手机器人的特点,设计超声波传感器模块的硬件和软件,构建超声波传感器模块的TWI通信网络,研究其通讯协议及其接口设置,进行超声波传感器的测距实验,并利用磁传感器进行室内环境绝对定位实验研究。对护士助手机器人模糊控制策略进行研究,设计模糊控制器,利用MOBOTSIM软件对所设计的模糊控制器的进行避障,沿墙行走和寻求目标等仿真实验。研究护士助手机器人的自主导航定位方法,建立差速驱动运动学模型,基于该模型在已知的结构环境中构建完整的导航系统。研究相对定位的航位推算法,对航位推算法在护士助手机器人导航中的应用进行实例分析。建立实验环境的拓扑地图,对基于拓扑地图的运动算法进行分析。针对护士助手机器人运动过程中避障的实际问题,对超声波传感器的测障方法进行研究,提出基于超声波传感器测距避障的栅格地图法,实现机器人的自主避障。针对护士助手机器人路径规划中的具体实现问题进行研究,提出基于遗传模拟退火算法的路径规划方法,并对算法的各个参数进行了研究和设置;建立了基于神经网络模型的机器人工作空间,并以此为基础描述机器人静态工作环境,利用MATLAB软件对基于标准遗传算法和遗传模拟退火算法的静态路径规划分别进行了仿真实验,并对仿真实验结果进行分析。更多还原

【Abstract】 Nurse assistant robot is a kind of mobile service robot, and it is operating in hospitals and automatic navigation. It can deliver all kinds of things for the patients and spray disinfection in infection sickroom. It can be applied to promote ’ hospital automation, protect medical care personnel from cross contamination, it has important application value and prospect. The research of this paper is supported by project of "Key Technologies of Nurse Assistant Robot", funded by the "Hei Longjiang Natural Science Fund". This paper is aimed to develop a automatic navigation nurse assistant robot test prototype and research the key technology.The paper overviews the research and developing situation of nurse assistant robot, at home and abroad, and summarizes the main technology of the mobile robot.Considering the performance and the technology demand of the nurse assistant robot, the robot’s whole scheme is presented, including mechanical structure, sensor system and control system. Differential driving type is used, and nurse assistant robot steering system is flexible. Based on the hierarchy control structure theory, the control system is designed. The execution level is composed of DSP TMS320F2812 and ATmegal6, and the organizational level is PC104. The organizational level communicates with the execution level by CAN bus.The hardware and software of the ultrasonic sensor are designed. The TWIcommunication network is constructed, research on the communication protocolsand the interface set. The ranging experiment of the ultrasonic sensor is tested.The method of absolute positioning in indoor based on magnetic sensor isanalyzed.Fuzzy control arithmetic is. researched. The fuzzy control system of nurseassistant robot is developed; fuzzy control arithmetic is simulated by Mobotsimsoft. The method of the nurse assistant robot automatic navigation is researched. The kinematics model of the differential driving robot are theorized and based on the model, the robot navigation system is established in the known environment. The dead reckoning algorithm used to robot relative positioning is developed, and its application in nurse assistant robot is analyzed. The topological map of the experiment environment is constructed, and the motion algorithm based on the topological map is researched. In order to avoid obstacle, the method of ultrasonic obstacle measurement is researched. The grid map method based on the ultrasonic obstacle measurement is proposed, and the nurse assistant robot could avoid obstacle autonomously.For the actual problem of the nurse assistant robot path planning, a path planning method which is based on the genetic simulated annealing algorithm is developed for the robot. The characteristic and steps are introduced with the parameters of the algorithm being studied and set. The neural network model of environment information is established, and the robot static working environment based on the model is described. Then the simulation of the genetic algorithm and the genetic simulated annealing algorithm is carried out in MATLAB and the simulation results are analyzed.更多还原