基于RRTConCon算法的船舶装配拆卸高斯采样路径规划

摘要

船舶舱室内的零部件在进行装配拆卸时,需要依据正确可行的路径进出装配体,通常采用路径规划的方法确定拆装路径.针对RRTConCon算法是采用随机采样的方法选取位姿点,在解决船舶舱室内狭窄通道的路径规划上效率不高的问题,提出了一种基于高斯采样的RRTConCon算法(RRTGaussion),采用高斯分布函数进行分区采样:在大的开阔区域设置较少的采样点,在复杂区域或狭窄通道设置较多的采样点,然后进行局部规划,找出拆装路径.通过在虚拟环境中对该算法进行仿真验证,结果表明,该算法在解决船舶舱室中狭窄通道的路径规划问题上效率高于RRTConCon算法.%The parts need to be moved in/out from the equipment in ship cabins with a proper and feasible path when the Assembly/Disassembly is implemented. Usually, motion planning is used to find the Assembly/Disassembly path. Among the existing motion planning algorithms, RRTConCon algorithm has higher efficiency than that of other algorithms and is suitable for the motion planning of ship parts' Assembly/Disassembly, but it has low efficiency with motion planning for narrow passages in ship cabins because of the random adding of free configurations. So, we proposed one kind of RRTGaussion algorithm with Caussion sampling based on RRTConCon algorithm. The algorithm used the Gaussian sampling strategy that sampled less in large open regions while more in complex regions and narrow passages , and then based on this the local planner was used to try to connect these samples with a path. The proposed algorithm was validated in virtual environment. The experiments show that in solving the motion planning problems for narrow passages in ship cabins, the proposed RRTGaussion algorithm has higher efficiency than RRTConCon algorithm.