【作者】 张雷；

【导师】 冯晓云；

【作者基本信息】 西南交通大学 ， 电力电子与电力传动， 2007， 硕士

【摘要】 机器人技术的发展彰显一个国家科技水平的高低。在机器人关键技术的发展过程中，提高个体机器人的智能程度和群体机器人之间的协调与协作成为研究的热点。因此，涉及机器人自主规划、多机器人协作的路径规划问题与编队问题成为机器人研究的重要课题，无论是工业机器人还是移动机器人的设计都与此息息相关。对提高机器人的智能化水平及加快机器人的实用化进程具有重要的理论研究意义和实用价值。论文分析机器人路径规划所要解决的基本问题与相关技术，由于机器人路径规划具有复杂性、随机性、多约束、多目标等复杂系统的特点，在现有的研究基础上，本文提出面向机器人的全局规划思路，设计单个机器人路径规划方法，将机器人所处的客观环境知识与遗传算法融合，改变了原有的编码、种群初始化方法，设计新的进化操作算子。通过仿真验证了算法的可行性，同时对于算法的参数影响进行了分析和研究。在此基础上，论文进一步分析多机器人的路径规划特点，引入协同进化概念，设计一种新的基于共生机制的遗传算法。该方法依据协同进化论的理论，充分发挥遗传算法的本质并行性，通过种群划分和信息交换的协作，实现了多机器人的全局路径规划。关于机器人编队问题，论文从队形形成和队形控制两个层次分析问题，设计并实现了一种多机器人在执行协作任务时的协调控制方法。采用基于Leader的队形控制方法与队形动态调整相结合的策略，增强多机器人队列处理突发事件的能力，同时结合机器人角色转换方法，使得机器人协作任务得以更好地完成。本文分析了智能机器人仿真系统的功能结构，设计用户交互界面和功能模块，开发用于路径规划、编队问题研究的机器人仿真软件，并对本文中提出的路径规划中的算法和队列控制算法进行了仿真验证。更多还原

【Abstract】 The development of robot technology represents the level of a nation’s science and technology. In the process of development in the robot key technology area, single robot’s intelligence and multi-robots’ cooperation and coordination become the research focus. So Path-planning and Formation which refer to the robot self-planning and multi-robot cooperation are the two key points of the robot research, which is closely linked to the design of the mobile robot and industrial robot. It’s very worthwhile for the improvement of the robot’s intelligence and accelerating the robot practicality in the theoretic research and application.The thesis analyzes the basic problems and associated techniques to be solved in the path-planning. According to the path-planning problem’s characters, such as complexity, randomicity, multi-restriction and multi-objective, it proposes a global planning theory on robot-orientation based the existing methods. In the designing of the Single-Robot Path-Planning, the path-planning background and basic genetic algorithms (GA) adopted in single-robot are combined, thus the GA based on background technology is formed. The algorithms changes the original code and population initialization method, according to path-planning, a new operator is designed, which is realized in this thesis, and simulation results prove the validation of this method, at the same time, the influence of algorithmic parameters is analyzed and researched.Based on the single-robot path-planning method, the character of multi-robot’s path-planning is analyzed, and co-evolutionism is introduced, finally a global path-planning way is proposed in the thesis. This method fully exerts the essence parallel of the GA on the theory of co-evolution, therefore realizes global path-planning of multi-robots by the cooperation of population classification and information exchange.In the research of multi-robots formation, this thesis analyzes the formation and control of robots’ team. It realizes a coordination control method during the executing of the collaboration task. It adopts a team control strategy which is based on the combination of the "Leader" control method and team dynamic adjustment, thus it can improve the capability how the multi-robot team deal with the emergency. On the base of this method, and additionally by combination of robot’s role conversion, multi-robots can accomplish the cooperating task much well.The thesis analyzes the function structure of the intelligent robot simulation system, designs the simulation software which provides the user mutual interface and function models for path-planning and formation research, more over proves the methods of path-planning and formation by simulation.更多还原