【作者】 王火亮；

【导师】 孙守迁；

【作者基本信息】 浙江大学 ， 计算机应用， 2002， 硕士

【摘要】 近年来，家用的自主式移动机器人研究和设计成为了许多人关注的焦点，机器人的研究和设计领域已经不仅仅限于军用领域中。许多应用于军用机器人上的成功技术已经使用于民用的机器人中。如利用各种传感器来采集环境信息并自动引航等先进技术。 对于移动机器人来说，机器人自动导航和定位是两个重要的问题，是实现机器人智能性的最重要的核心技术。它主要包括：环境信息获取功能，环境信息理解和建模功能以及实现自动导航的路径规划算法功能组成。其中环境信息理解和建模功能是核心。因此移动机器人导航系统研究就是针对所处环境选择相应的传感器，然后根据传感器信息特征采用相应的环境模型，最后建立面向该环境模型的路径规划控制算法。但是目前从理论上还没有很好的方法来解决机器人自动导航和定位问题，这是一个有待从原理去探索的问题。 对环境的感知是自主式移动机器人研究的关键技术之一。环境信息的描述是实现对机器人自主式导航的算法基础。环境信息采集和建模描述的精确度应该以实现对机器人导航的可靠性为依据。强制要求传感器对环境信息采集和描述的精确度会影响传感器的信息采集速率，增加算法复杂度和计算量，从而影响算法的实时性、可靠性和鲁棒性。虽然现在对环境信息的建模方法有多种，其中比较成功和有效的方法主要有势场法和栅格法两种。1986年，0.Khatib提出了人工势场法，后来由C.W.Warren在1989和1990年又加以了改进。本文考虑家用自主智能吸尘机器人的使用环境和其设计的指标要求，以提高它的工作效率为主要目的的前提下，本文采用超声波测距传感器来采集环境信息和基于栅格法的环境信息建模。 本文由七个部分组成：第一章首先介绍了本文工作的背景，然后介绍了智能机器人导航系统实现的三个组成部分的关系；第二章详细描述了超声波传感器的相关特性和当前两种测距技术——有源测距技术和无源测距技术，同时详细的分析了超声波传感器的测距原理和在智能吸尘机器人中选用超声波传感器的因素分析；第三章主要介绍了两种主要的建模方法——栅格法和势场法；第四章详细的描述了基于栅格法的区域扫描路径规划导航算法的实现；第五章详细的描述了基于超声波测距的沿边走路径规划导航算法的实现；第六章对智能吸尘机器人的 总体设计框架和功能模块的设计进行了详细的描述，并结合基于超声波测距的沿 边走路径规划的导航算法给出了应用实例的程序模拟；第七章对本论文的工作进 行了总结，同时结合自己在参加智能吸尘机器人项目中所积累的经验和知识对智 能吸尘机器人的发展进行了粗略的展望。更多还原

【Abstract】 ABSTRACT Research and design of home-used mobile robot has become a hot field last years. They have not been limited in military field. Many technologies, which had been used in the military field successfully, have been used in the field of home-used robot, for example technology of collection environment information by sensor and technology of autonomous navigation and location. Autonomous navigation and location is key technology to implement intelligence of mobile robot in the field of mobile robot. It includes three technologies-collection environment information, environment information modeling and algorithm of path planning to implement autonomous navigation and location. Environment information modeling is key technology. So research of mobile robot autonomous navigation includes three aspects. Firstly, we collect the environment information. Then, we use the method of environment information modeling to depict the environment information. Finally we design the algorithm of path planning to implement autonomous navigation and location function. But we have not a perfect theory to resolve it now. Environment information modeling is one of key technologies in the field of mobile robot. Depiction of environment information is foundation to implement autonomous navigation and location function. Definition of collection environment information is depended on reliability of implementation autonomous navigation and location. If we abuse to demand it, it will impact ability of real time, reliability and robust ability. We have two method of environment information modeling-method of grid and method of potential field now. They have been used in field of mobile robot successfully. The paper is organized in seven parts: In the first part, the background of this paper is present first. Then, the relationship of three technologies to implement autonomous navigation and location is depicted in detail. In the second part, specific of ultrasonic sensor and two range technologies are depicted in detail first. Then, the fundamental of ultrasonic sensor ranging is depicted in detail. In the third part, method of grid and method of potential field is depicted in detail. In the fourth part, algorithm based on method of grid area dismember is depicted in detail. In the fifth part, algorithm of along wall is depicted. In the sixth part, how to design intelligent cleaning robot is depicted. Then, by the example of intelligent cleaning robot, algorithm of along wall is depicted in detail. In the seventh part, the outlook of intelligent cleaning robot is depicted.更多还原