针对舰载机甲板路径规划问题,在Informed-RRT*(informed rapidly-exploring random tree)的椭圆采样基础上,提出使用正态分布方式采样的IN-RRT*(informed normal-RRT*)算法.首先,针对舰载机与运动场景建模,定义舰载机运动约束和避障策略;其次,将正态分布采样策略与椭圆采样相结合,获取优质高效采样点;引入人工势场法,自适应调节随机树的搜索步长值;使用向心Catmull-Rom样条插值法对路径进行平滑优化处理;提出针对动态障碍改进的动态窗口法,实现局部动态避障.最后,运用甲板平面环境实验检验算法性能.结果表明,IN-RRT*算法能显著优化搜索时间和搜索路径质量,可应对动态场景规划出合理可行的平滑路径.
Aircraft Deck Flat Path Planning Based on Improved Informed-RRT*Algorithm
In order to solve the problem of aircraft deck path planning,based on the elliptic sampling of informed rapidly-exploring random tree(Informed-RRT*),an informed normal-RRT*(IN-RRT*)algorithm using normal distribution sampling was proposed.Firstly,the carrier-based aircraft and the motion scene were modeled,the motion constraints and obstacle avoidance strategies of the carrier-based aircraft were defined,then the normally distributed sampling strategy was combined with elliptic sampling to obtain high-quality and efficient sampling points,the artificial potential field method was introduced to adaptively adjust the search step value of the random tree,the centripetal Catmull-Rom spline interpolation method was used to smooth and optimize the path,and the dynamic window method improved for dynamic obstacles was proposed to realize local dynamic obstacle avoidance.Finally,the performance of the algorithm was tested by experiments in the flat environment of the deck.The results show that the IN-RRT*algorithm can significantly optimize the search time and search path quality,and can plan a reasonable and feasible smooth path for dynamic scenes.
万方数据
