对机器人足球的决策与路径规划的研究
发布时间：2018-12-12 22:05
【摘要】： 机器人足球比赛是近年来在国际上迅速开展起来的高技术对抗活动。它是人工智能领域与机器人领域的基础研究课题，是一个极富挑战性的高技术密集型项目。本文以足球机器人系统的核心子系统——决策子系统的设计开发为背景，研究行之有效的决策推理方法。足球机器人系统是一个典型的多智能体系统，涉及机器人学、计算机视觉与模式识别、多智能体系统、轨迹规划与智能算法、自组织与自学习理论等领域。 足球机器人系统分为四个子系统——机器人子系统、视觉子系统、决策子系统和通讯子系统。其中决策子系统是整个系统的核心，应具有可自主完成知识提取，，并确定机器人协作任务的能力，使整个系统具有智能体的特征。机器人足球环境是一个具有动态性、不确定性、实时性的环境，在这样一个具有高度实时性和竞争性平台上研究路径规划也是一个很具有挑战性的课题。通过本课题的研究，得到如下的成果与结论： 1．综合国内外相关研究文献，阐述了足球机器人研究现状和主要内容。同时，详细介绍了足球机器人系统、路径规划及机器人队形确定问题的研究现状。 2．介绍了几种足球机器人路径规划的理论和算法，，如人工势场法、栅格法、可视图法以及各种人工智能方法如遗传算法、神经网络等。但这些方法在高度动态和实时的环境中的研究还不太完善，需要进一步改进。讨论它们在足球机器人系统中的应用可行性，最终确定以人工势场法为研究方向。 3．通过对足球机器人队形确定问题的研究，来阐述足球机器人路径规划中的路径冲突协调和协作问题。 4．通过一种新的开源足球机器人仿真平台robotsoccer建立仿真实验环境，然后在在平台上利用lingo编程语言编写客户端算法，实现对仿真足球机器人的控制。 本论文在对足球机器人总体系统进行介绍的基础上，分析研究了决策子系统的系统模型，提出了一种基于改进人工势场法进行动态环境的路径规划的方法。通过梯度逼近进行运动目标的搜索，解决了在实时路径规划中因环境的运动信息难以准确获取而造成的路径规划无法完成的问题。仿真结果验证了方法的有效性，能够较好的解决动态环境下特别是存在随机运动物体情况下足球机器人的路径规划问题。
[Abstract]:Robot soccer match is a high-tech counteraction that has been developed rapidly in the world in recent years. It is a basic research subject in the field of artificial intelligence and robot. It is a challenging high-tech-intensive project. Based on the design and development of the decision-making subsystem, which is the core subsystem of the soccer robot system, an effective decision reasoning method is studied in this paper. Soccer robot system is a typical multi-agent system, which involves robotics, computer vision and pattern recognition, multi-agent system, trajectory planning and intelligent algorithm, self-organization and self-learning theory and so on. Soccer robot system is divided into four subsystems-robot subsystem, vision subsystem, decision-making subsystem and communication subsystem. The decision-making subsystem is the core of the whole system, and it should have the ability to complete the knowledge extraction independently and determine the cooperative task of the robot, so that the whole system has the characteristics of the agent. Robot soccer environment is a dynamic, uncertain and real-time environment, and it is also a challenging task to study path planning on such a highly real-time and competitive platform. Through the research of this topic, the following results and conclusions are obtained: 1. The current situation and main contents of soccer robot research are expounded by synthesizing the relevant research literature at home and abroad. 2. At the same time, the research status of soccer robot system, path planning and robot formation determination is introduced in detail. 2. Several theories and algorithms of soccer robot path planning are introduced, such as artificial potential field method, grid method, visual graph method and various artificial intelligence methods such as genetic algorithm, neural network and so on. However, the research of these methods in high dynamic and real-time environment is not perfect, which needs further improvement. The feasibility of their application in the soccer robot system is discussed, and the artificial potential field method is finally chosen as the research direction. 3. The coordination and cooperation of path conflict in soccer robot path planning are discussed by studying the problem of soccer robot formation determination. 4. The simulation experiment environment is established by a new open source soccer robot simulation platform, robotsoccer, and then the client algorithm is programmed by lingo programming language on the platform to realize the control of the simulation soccer robot. Based on the introduction of the overall system of soccer robot, the system model of decision-making subsystem is analyzed and studied in this paper, and a path planning method based on the improved artificial potential field method is proposed. By using gradient approximation to search moving objects, the problem that path planning can not be completed in real-time path planning is solved because the environment motion information is difficult to obtain accurately. The simulation results show that the method is effective and can solve the path planning problem of soccer robot in dynamic environment, especially in the case of random moving objects.
【学位授予单位】：山东大学
【学位级别】：硕士
【学位授予年份】：2007
【分类号】：TP242
本文编号：2375308
[Abstract]:Robot soccer match is a high-tech counteraction that has been developed rapidly in the world in recent years. It is a basic research subject in the field of artificial intelligence and robot. It is a challenging high-tech-intensive project. Based on the design and development of the decision-making subsystem, which is the core subsystem of the soccer robot system, an effective decision reasoning method is studied in this paper. Soccer robot system is a typical multi-agent system, which involves robotics, computer vision and pattern recognition, multi-agent system, trajectory planning and intelligent algorithm, self-organization and self-learning theory and so on. Soccer robot system is divided into four subsystems-robot subsystem, vision subsystem, decision-making subsystem and communication subsystem. The decision-making subsystem is the core of the whole system, and it should have the ability to complete the knowledge extraction independently and determine the cooperative task of the robot, so that the whole system has the characteristics of the agent. Robot soccer environment is a dynamic, uncertain and real-time environment, and it is also a challenging task to study path planning on such a highly real-time and competitive platform. Through the research of this topic, the following results and conclusions are obtained: 1. The current situation and main contents of soccer robot research are expounded by synthesizing the relevant research literature at home and abroad. 2. At the same time, the research status of soccer robot system, path planning and robot formation determination is introduced in detail. 2. Several theories and algorithms of soccer robot path planning are introduced, such as artificial potential field method, grid method, visual graph method and various artificial intelligence methods such as genetic algorithm, neural network and so on. However, the research of these methods in high dynamic and real-time environment is not perfect, which needs further improvement. The feasibility of their application in the soccer robot system is discussed, and the artificial potential field method is finally chosen as the research direction. 3. The coordination and cooperation of path conflict in soccer robot path planning are discussed by studying the problem of soccer robot formation determination. 4. The simulation experiment environment is established by a new open source soccer robot simulation platform, robotsoccer, and then the client algorithm is programmed by lingo programming language on the platform to realize the control of the simulation soccer robot. Based on the introduction of the overall system of soccer robot, the system model of decision-making subsystem is analyzed and studied in this paper, and a path planning method based on the improved artificial potential field method is proposed. By using gradient approximation to search moving objects, the problem that path planning can not be completed in real-time path planning is solved because the environment motion information is difficult to obtain accurately. The simulation results show that the method is effective and can solve the path planning problem of soccer robot in dynamic environment, especially in the case of random moving objects.
【学位授予单位】：山东大学
【学位级别】：硕士
【学位授予年份】：2007
【分类号】：TP242
【引证文献】
相关期刊论文 前1条
1 胡玲;廖家平;舒军;鲁海霞;;足球机器人的路径规划方法[J];天津市经理学院学报;2010年04期
相关硕士学位论文 前1条
1 伍龙军;MiroSot足球机器人决策子系统的研究[D];西华大学;2008年
本文编号：2375308
本文链接：https://www.wllwen.com/kejilunwen/rengongzhinen/2375308.html
