移动机器人系统作为一个典型的多智能体系统，它为智能理论研究及多种技术的集 成应用提供了良好的实验平台，移动机器人在运动过程中能够根据周围环境的变化而自 主的采取相应措施的能力也越来越强，因此，作为机器人智能的一个重要因素——路径 规划的研究，就显得十分重要。 本课题主要研究的内容就是以足球机器人为研究平台，在足球比赛实际过程以及特 殊状态环境中，利用改进型人工免疫算法完成足球机器人的寻找目标，实现避障等动作， 解决了传统算法在路径规划过程中可能出现的易陷入局部最优解以及死锁问题。 机器人的路径规划是一个很复杂的问题，不仅要寻求一条无碰撞的最短路径，而且 还要求该路径尽可能平滑并满足一定的安全性。本文首先介绍足球机器人的硬件结构组 成及其软件平台，并重点讨论了动力系统的改造，随后，介绍了机器人路径规划的传统 算法以及智能规划算法，在以上工作的基础上分析了目前各种路径规划算法的优缺点， 由此提出了一种基于改进型人工免疫算法的移动机器人路径规划。 针对静态路径规划，使用栅格法建立空间数学模型，并利用人工势场法建立人工免 疫算法所需要的初始路径种群，优化了变异算子，提出了新的亲和度函数，并引入了免 疫接种，交叉与变异等免疫机制，使得较优个体能较早的生成，从而保证了算法的收敛 速度和种群的多样性，防止了早熟收敛，提高了全局搜索能力。仿真结果表明，该方法 在静态环境下有良好的适应能力，能有效地保证路径的规划效率并改善路径规划的质 量。 针对动态路径规划，利用二维笛卡尔栅格建立移动机器人的动态环境模型，利用人 工势场算法进行种群的初始优化，通过改进静态环境下免疫算法中的亲和度函数，来对 个体进行评估。通过仿真实验对比，表明了该算法能较好的适用于动态复杂环境下移动 机器人的路径规划。 最后，在实际的足球机器人系统平台上进行实验，在实际的足球比赛环境中对改进 型人工免疫算法路径规划进行验证。实验表明：足球机器人在各种状态环境下所表现出 的优良避障和寻找目标效果，表明了该算法所规划路径的优越性，圆满地实现了论文的 设计目标。由于对足球机器人软件和硬件平台中动力系统两方面的改进，使我校足球机器人团队在历年的全国大赛上取得了优异的成绩。

Mobile robot system as a typical multi-agent system, it is provides a good experimental platform of the intelligent theory and integration of multiple technologies applications . Therefore, as an important factor in intelligent robots - path planning is very important. The main research topic is the content of the soccer robot research platform, The actual process and special state environment in soccer, complete soccer robot to find the target, obstacle actions such as avoidance. Solve the path planning process may be easy to fall into local optimal solution and the deadlock in traditional algorithm. Robot path planning is a very complex issue, not only to find a collision-free shortest path, but also requires the path as smooth as possible and to meet certain security. This article use the soccer robot subject group of Xian university of science and technology as research platform, First introduced the composition and structure and software platform of soccer robot ,and focused on the transformation of power system, Second, describes the robot path planning traditional algorithm and intelligent planning algorithm, Based on the analysis of the current advantages and disadvantages of various path planning algorithm, Thus proposed a path planning algorithm based on modified artificial immune mobile robot. For static path planning, using the grid method to establish the space mathematical model, and then using the artificial potential field to establish the initial path population of artificial immune algorithm, optimize the mutation operator, proposed a new affinity function, introduced vaccination, the immune mechanisms such as crossover and mutation, allowed better individuals to an earlier generation, Thus ensuring the convergence speed and diversity of the population, prevent premature convergence, improve the global search capability, Simulation results show that the method under the static environment have a good ability, can effectively guarantee the efficiency and improve the quality of path planning. For dynamic path planning, using Two-dimensional Cartesian grid established mobile robot dynamic model of the environment, and then using artificial potential field algorithm to optimize the initial population, By fitness function of the immune algorithm in a improving static environment to assess individual. Compared by simulation, show that this algorithm for the dynamic complexity environment of mobile robot path planning. At last, in a real soccer robot system experiment platform and in a real football environment to authentication the improved artificial immune algorithm for path planning. Experiments show that: soccer robots in a variety of state conditions have shown a good look for obstacle avoidance and target effects, have demonstrated the superiority of the path planning algorithm, success in achieving the design goal of paper. Because the improvements of power systems in both software and hardware platform of the soccer robot, so that our school's robot soccer team in the national competition over the years has made outstanding achievements.