面向空间机器人辅助操作的任务规划方法研究  

Research on Task Planning Methods for Space Robot Assisted Operation

作　　者：房国强 常海涛[1] 刘星 刘正雄[1] 黄攀峰[1] Fang Guoqiang;Chang Haitao;Liu Xing;Liu Zhengxiong;Huang Panfeng(College of Astronautics,Northwestern Polytechnical University,Xi'an 710072,China)

机构地区：[1]西北工业大学航天学院,陕西西安710072

出　　处：《系统仿真学报》2025年第2期325-334,共10页Journal of System Simulation

基　　金：国家自然科学基金(62273280,62103334,92370123,62103337);中央高校基本科研业务(D5000230139)。

摘　　要：针对空间机器人辅助操作时任务流程复杂、任务约束众多,导致任务规划困难的情况,提出了一种快速前向搜索算法和分层网络算法相结合的任务规划方法。将空间辅助操作任务规划过程分为任务规划和重规划。基于操作代价的快速前向搜索任务规划方法,得到操作代价最小的动作执行序列;基于分层网络的任务自适应重规划方法,根据问题排列移动、抓取、释放的修正补偿优先级对问题修正补偿。针对空间机器人在轨辅助加注任务进行了对比仿真,得到了操作代价最小的动作序列,并采用重规划算法解决了任务执行过程中环境信息变化导致的错误,验证了算法的有效性,为空间辅助操作提供了一种有效的任务规划方法。In view of the difficulties caused by the complicated task process and numerous task constraints during the space robot assisted operation,a task planning method combining fast forward search algorithm and hierarchical network algorithm is proposed,in which the task planning process is divided into task planning and replanning.Based on the fast forward search task planning method of operation cost,the execution sequence of actions with minimum operation cost is obtained.The task adaptive replanning method based on hierarchical network corrects and compensates the problems according to the priority of compensation for the movement,grab and release actions.A comparative simulation is carried out for the space robot's on-orbit auxiliary filling task,and the action sequence with the lowest operational cost is obtained.The error caused by the change of environmental information during the task execution is solved by using the replanning algorithm,which verifies the effectiveness of the algorithm and provides an effective task planning algorithm for the space auxiliary operation.

关 键 词：空间机器人 任务规划 重规划 快速前向搜索 分层任务网络 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]