在轨操控机器人拓邻域搜索三维路径规划

Three-dimensional Path Planning of On-orbit Manipulation Robot Based on Neighborhood Continuation Search

针对空间在轨操控机器人在桁架式空间站执行操控任务时,需要在桁架间自由穿行且时刻与空间站保持接触的任务特性,提出基于拓邻域搜索蚁群算法的三维全局路径规划方法。该方法首先建立桁架式空间站的三维栅格模型;充分考虑机器人穿行时的包络约束和操控时的接触约束,在路径规划和轨迹搜索前对被操控对象的三维栅格模型进行预处理和修正,缩减路径规划搜索空间;在轨迹搜索时,基于邻域延拓思想调节搜索步长,以提高搜索效率和路径规划速度。多次仿真实验表明,所提出的路径规划方法能提高在轨操控机器人路径规划速度与全局搜索能力,有效解决空间在轨操控机器人全局路径规划问题。

Aiming at the mission requirement of the on-orbit manipulation robots in different phases,i.e.,walking through and attaching to the space truss in walking phase and manipulation phase,a three-dimensional global path planning strategy based on the raster model and an ant colony optimization via neighborhood continuation search is proposed.In the proposed path planning strategy,a three-dimensional raster model of the international space station is built and a preprocessing of the model is conducted to reduce the potential searching space.Then an ant colony optimization based on neighborhood continuation search(ACOBNCS)is proposed to adjust the search step and improve the search speed.Multiple simulation experiment results verify the effectiveness of the path planning approach in the dimensions of time and space,and solve the difficulties in three-dimensional global path planning of on-orbit manipulation robots.