多机器人编队控制策略研究及其云实现

发布时间：2018-06-03 06:29

本文选题：多机器人系统 + 编队控制算法　；参考：《山东大学》2017年硕士论文

【摘要】：近年来,随着科学技术的发展进步,机器人技术越来越成熟,并且在社会生活中发挥了举足轻重的作用。面对复杂任务时,多机器人系统通常具备比单机器人更强的鲁棒性和工作效率,因此得到了研究人员高度重视。编队策略是多机器人协同工作的基础,它是指多机器人保持预定的队形前进,同时避开环境中的障碍物,最终抵达目标点。这一过程中涉及到多机器人定位、控制策略、路径规划以及无线通信等多方面内容。为确保机器人编队能够在未知环境中行进,需要时刻通过传感器采集大量环境信息并进行分析处理,以决定整个系统下一步的运动规划。传统机器人受自身机载设备的限制,在处理冗杂的环境信息时计算效率往往不足,影响整个系统的实时性,而云机器人概念的提出为解决这一问题提供了思路。云机器人能够将复杂的计算任务卸载到云端,利用云计算为机器人提供强大的计算能力。本文旨在将云计算与机器人编队相结合,充分发挥云端在数据运算方面的优势,降低机器人对硬件部分的依赖,提升系统的工作性能。本文的研究内容主要分为编队控制策略和云机器人系统实现两个部分。在编队控制策略部分,采取领航-跟随算法实现队形保持,利用人工势场法进行路径规划,使机器人编队具备在未知环境中工作的能力。在云机器人系统部分,为将云计算与机器人编队相结合,本文搭建了一个基于AmigoBot和中国电信天翼云主机的多机器人云平台。在这一系统中,机器人与云主机之间采取C/S构架进行编队运行。机器人作为客户端,将专注于环境信息采集和编队任务执行;云主机作为服务器,将处理机器人采集的数据并运行编队算法,为机器人编队进行运动规划。为保证网络通信的稳定性,防止数据处理过程中出现丢失等问题,机器人与云主机之间的通信采用基于TCP/IP的网络传输协议。本文首先从多机器人系统和云机器人的理论意义和现实价值出发,阐述了本课题的研究目的,并介绍了国内外关于多机器人系统和云机器人方面的研究进展,对机器人编队的一些基本概念和经典算法进行了概述。其次,根据编队需要面向未知环境的要求,综合现有的研究成果,提出了一种领航-跟随法与人工势场法相结合的编队控制策略,同时对原有算法进行了改进,解决了系统的轨迹抖动问题。之后对多机器人系统的队形形成、队形保持、路径规划等方面进行了仿真和实验。然后,介绍了实验部分将要采用的AmigoBot和天翼云主机,分别解析了二者的工作模式,概述了网络通信所采用的TCP/IP协议的基本概念和数据传输流程,并测试了 AmigoBot与云主机之间网络通信的稳定性。最后,搭建了云机器人编队系统,对云计算与机器人编队相结合的理论进行了一系列的仿真和实验,并分析了实验结果。实验结果表明该系统能够有效地将机器人的运算任务转移到云端,减轻机器人的负担,提高数据处理效率,从而证明了将云计算应用于机器人编队的可行性。
[Abstract]:In recent years, with the development and progress of science and technology, robot technology is becoming more and more mature and plays an important role in social life. In the face of complex tasks, the multi robot system is usually more robust and efficient than single robot. Therefore, the researchers attach great importance to the robot. The formation strategy is multi robot. The basis of collaborative work is that multi robots keep the scheduled formation and avoid obstacles in the environment and reach the target point. This process involves multi robot positioning, control strategy, path planning and wireless communication. In order to ensure that the robot formation can travel in the unknown environment, it takes time to make sure that the robot is able to travel in the unknown environment. A lot of environmental information is collected and analyzed by sensors to determine the motion planning of the whole system in the next step. The traditional robot is limited by its own airborne equipment, and the computation efficiency is often insufficient when dealing with the miscellaneous environmental information, which affects the real time of the whole system, and the concept of cloud robot provides a solution to this problem. The cloud robot can unload the complex computing task to the cloud, and use the cloud computing to provide the powerful computing power for the robot. This paper aims to combine the cloud computing with the robot formation, give full play to the advantages of the cloud computing in the data operation, reduce the dependence of the robot on the hardware part and improve the performance of the system. The research contents are divided into two parts: formation control strategy and cloud robot system implementation. In the formation control strategy part, the formation keeping is realized by navigation following algorithm, and the artificial potential field method is used to make path planning, so that the robot formation has the ability to work in the unknown environment. Combined with robot formation, a multi robot cloud platform based on AmigoBot and China Telecom sky wing cloud hosts is built in this paper. In this system, a C/S framework is used between the robot and the cloud host for formation and operation. As a client, the robot will focus on the environment information collection and formation task execution; the cloud host is the server. In order to ensure the stability of the network and prevent the loss of the data processing, the communication between the robot and the cloud host is based on the TCP/IP based network transmission protocol. Firstly, the multi robot system and the cloud machine are used in this paper. On the basis of the theoretical and practical value of human, this paper expounds the research aim of this subject, and introduces the research progress on the multi robot system and the cloud robot both at home and abroad, and summarizes some basic concepts and classical algorithms of the robot formation. Secondly, according to the requirements of the formation needs for the unknown environment, the existing research is integrated. As a result, a formation control strategy which combines the navigation following method and the artificial potential field method is proposed. At the same time, the original algorithm is improved and the trajectory jitter of the system is solved. After that, the formation of the formation, formation keeping and path planning of the multi robot system are simulated and experimentation. Then, the experimental part will be introduced. In order to use the AmigoBot and the sky wing cloud host, the working mode of the two people is analyzed, the basic concepts and data transmission flow of the TCP/IP protocol used in the network communication are summarized, and the stability of the network communication between the AmigoBot and the cloud host is tested. Finally, the cloud machine man formation system is built, and the cloud computing and the robot formation are formed. The theory carried out a series of simulation and experiments, and analyzed the experimental results. The experimental results show that the system can effectively transfer the computing task of the robot to the cloud end, reduce the burden of the robot and improve the efficiency of data processing, thus proving the feasibility of applying cloud computing to robot formation.
【学位授予单位】：山东大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP242

【相似文献】