中文摘要：

以中型自主足球机器人为对象，研究其动态避障情况．依据传感器输入信息进行特征识别，选择距离机器人最近的为首要躲避障碍，利用产生式规则进行机器人避障控制，对远距离和近距离障碍采用不同的避障策略．并引入虚拟点以躲避碰撞．在仿真和实时环境下进行了实验，分析了障碍位置不同和输入参数不同情况下的避障情况，并与传统人工势场法进行对比，避免了传统人工势场法进行机器人避障时会出现陷入局部极小、抖动等同题．实验结果表明该方法能使机器人快速有效的躲避障碍．

英文摘要：

Taking middle-size soccer robot as object,its dynamic obstacle avoidance has been studied. According to the input information and character recognition, the nearer one is selected as the principal obstacle to avoid. Making use of production rule, the obstacle avoidance behavior of soccer robot is designed. Different avoidance strategies are devised to the remote and close obstacle. A virtue point is also proposed to avoid collision. Many experiments have been done under simulation and real game and contrast with the traditional artificial potential field method. In terms of the variety of obstacle position and input parameter, the experiment results are analyzed. The chattering and local minimal questions of traditional artificial potential field are avoided. Experiment results have shown the validity of the method proposed in this paper,and this method can make robot quickly and effectively to avoid the obstacle.