【摘要】 本文提出了智能系统中的二层路径规划子系统,其中的全局路径规划层接收来自任务规划子系统的目标序列,针对已知的静态障碍环境信息,采用基于 Hopfield 人工神经网络(HANN)算法的新型路径规划方法,得出点到点的直线路径(预直线路径)序列,并进行了仿真。局部路径规划与跟踪层接收来自全局路径规划层的预直线路径,考虑全局路径规划层未知的静态和动态障碍,采用势场法来进行局部路径规划与跟踪,并进行了仿真。更多还原

【Abstract】 In this paper,path planning sub-system of two level architecture is presented,which consists ofglobal path planning(GPP)level and local path planning and following(LPPF)level.GPP level re-ceives goal sequences from mission planning sub-system,planning global collision-free and straightpath according to known environment map information,where new type GPP method based on Hop-field Artificial Neural Network algorithm is proposed,and so is the simulation.LPPF level receivesPrepared Straight Path(PSP)from GPP level,planning and following local collision-free path ac-cording to PSP and unknown static or dynamic obstacles in GPP level,where LPPE method based onpotential field is proposed,and so is the simple simulation.更多还原