改进型概率地图航迹规划方法

Path Planning Based on Improved Probabilistic Roadmap Method

针对在线航迹规划对时间和内存要求较高,提出一种基于遗传算法的改进型概率地图算法。该方法引入分层规划的思想,在离线航迹规划时,利用遗传算法找到最优初始航迹并建立航迹走廊,以此来缩小概率地图的规划空间。概率地图方法在缩小后的空间中建立概率地图,自由线路数量大大降低,同时使得在线航迹规划搜索范围减少,时间缩短。仿真实例表明,该方案比传统PRM更能满足在线航迹规划的要求。

For the demands of limited time and high level of memory in on-line Path Planning,an improved Probabilistic Roadmap Method（PRM） which is based on Genetic Algorithm（GA） was proposed. This path planning algorithm adopted the principle of hierarchical planning.In the off-line path planning, track corridor was established based on the optimal initial paths which was generate by GA,thus reducing the programming areas of PRM.The number of uniform random samples decreased dramatically when PRM was operated in track corridor.In on-line Path planning,the areas was also reduced,which resulted in reduced programming time.Simulation results showed that this method met the requirements of on-line Path planning better than the traditional PRM.