Abstract: Aiming at the path planning in unknown static environment，a new mobile robot path planning approach is presented which combines of rolling windows and behavioral dynamics.Firstly，subgoal choosing is optimized by using heuristic function according to the window’s information，it is local environment’s information，which is obtained by camera.Then navigation behavior is divided into GO TO TARGET behavior and OBSTACLE AVOIDANCE behavior，their behaviorial states and behaviorial modes dynamics models are established，and then one autonomous local path planning online is implemented using navigation behavior dynamics model in a single window.Connecting a series of local paths according to continuity，the robot accomplishes one global planning ultimately.The approach has a simple principle and low calculating burden，smooth path and high applied value.The approach’s efficiency and adaptability is demonstrated by example simulating.

Key words: mobile robot, rolling windows, behavioral dynamics, path planning, subgoal

摘要： 针对存在静态障碍物的未知环境下移动机器人路径规划问题，提出运用行为动力学与滚动窗口相结合进行路径规划的方法。首先根据所获得的窗口（局部环境）信息，采用启发式函数进行局部子目标优化选择；然后将路径规划问题即导航行为分解为趋于目标行为和避障行为，并对这两种行为分别建立了行为状态和行为模式动力学模型；在此基础上，以窗口为单位，利用导航行为动力学模型进行在线自主路径规划；将一系列窗口中的规划轨迹按照连续性条件首尾相接，最终完成了一条全局规划任务。该方法原理简单，计算量小，规划路径光滑，具有较强的实际应用价值。通过计算机实例仿真验证了该方法的有效性和适应性。

关键词: 移动机器人, 滚动窗口, 动力学模型, 路径规划, 子目标