基于HTN的巡视器动态规划方法

doi:10.12305/j.issn.1001-506X.2024.02.26

摘要/Abstract

摘要：

为提升深空探测任务中巡视器动态规划处理效率，从航天工程实践经验出发，结合分层任务网络(hierarchical task network，HTN)理论研究成果，提出巡视器领域知识建模方法以及动态规划流程。针对经典动作规划系统不能处理时态约束关系的问题，在建模中引入受限状态事件的表示方法。针对动态规划中的时态约束关系，拓展Allen区间代数的定义，提出事件、全局时间点和虚拟事件的时序逻辑判断方法。针对动态规划中存在的状态空间一致性问题，提出虚拟指令继承原有状态空间的方法。研究结果完善了HTN在航天领域的知识建模方法，解决了经典动作规划缺少时间信息约束的缺陷。所提出的虚拟事件、虚拟指令解决动态规划状态判定方法，可有效提高深空探测任务中巡视器遥操作的工作效能和地面中心进行事件处理的灵活性。

关键词: 分层任务网络, 动态规划, 虚拟指令, 遥操作

Abstract:

In order to improve the efficiency of dynamic planning processing for inspectors in deep space exploration missions, a knowledge modeling method and dynamic planning process are proposed for inspectors based on practical experience in aerospace engineering, combined with the theoretical research results of hierarchical task network (HTN). To address the issue of classical action planning systems not being able to handle temporal constraint relationships, a representation method for constrained state events is introduced in modeling. To address the temporal constraints in dynamic programming, the definition of Allen interval algebra is extended, and a temporal logic judgment method for events, global time points, and virtual events is proposed. A method for virtual instructions to inherit the original state space is proposed to address the issue of state space consistency in dynamic planning. The research results have improved the knowledge modeling method of HTN in the aerospace field, solved the defect of lack of time information constraints in classical action planning. The proposed virtual events and virtual instructions to solve dynamic planning state judgment can effectively improve the work efficiency of remote operation of the inspector in deep space exploration tasks and the flexibility of event processing in the ground center.

Key words: hierarchical task network (HTN), dynamic programming, virtual instruction, teleoperation

中图分类号:

师明, 高宇辉, 张弓. 基于HTN的巡视器动态规划方法[J]. 系统工程与电子技术, 2024, 46(2): 631-639.

Ming SHI, Yuhui GAO, Gong ZHANG. Dynamic planning method based HTN for rover[J]. Systems Engineering and Electronics, 2024, 46(2): 631-639.

图/表 7

图1

图2

表1

表2

表3

表4

表5

参考文献 28

1	CESTA A , CORTELLESSA G , DENIS M , et al. Mexar2: AI solves mission planner problems[J]. IEEE Intelligent Systems, 2007, 22 (4): 12- 19. doi: 10.1109/MIS.2007.75
2	FRATINI S, CESTA A. The APSI framework: a platform for timeline synthesis[C]//Proc. of the Workshop on Planning and Scheduling with Timelines, 2012.
3	CESTA A , CORTELLESSA G , FRATINI S , et al. MrSPOCK-steps in developing an end-to-end space application[J]. Computational Intelligence, 2011, 27 (1): 83- 102. doi: 10.1111/j.1467-8640.2010.00373.x
4	CESTA A, CORTELLESSA G, FRATINI S, et al. MrSPOCK: a long-term planning tool for MARS express[C]//Proc. of the 6th International Workshop on Planning and Scheduling for Space, 2009: 83-102.
5	JOHNSTON M D. Spike: AI scheduling for hubble space telescope after 18 months of orbital operations[C]//Proc. of the Working Notes AAAI Spring Symposium on Practical Approaches to Scheduling and Planning, 1992.
6	BERNARD D, DORAIS G, GAMBLE E, et al. Spacecraft autonomy flight experience: the DS1 remote agent experiment[R]. Washington: American Institute of Aeronautics and Astronautics, 1999.
7	MUSCETTOLA N. HSTS: integrating planning and scheduling[R]. Pittsburgh: Carnegie Mellon University Robotics Institute, 1993.
8	SHERWOOD R, GOVINDJEE A, YAN D, et al. Using aspen to automate EO-1 activity planning[C]//Proc. of the IEEE Aerospace Conference, 1998.
9	SMITH B, SHERWOOD R, GOVINDJEE A, et al. Representing spacecraft mission planning knowledge in ASPEN[C]//Proc. of the Artificial Intelligence Planning Systems Workshop on Knowledge Acquisition, 1998.
10	BEDRAX-WEISS T, MCGANN C, BACHMANN A, et al. EUROPA2: user and contributor guide[R]. Washington: National Aeronautics and Space Administration Ames Research Center, 2005.
11	BERNARDINI S, SMITH D. Towards search control via dependency graphs in Europa2[C]//Proc. of the 19th International Conference on Automated Planning and Scheduling, 2009.
12	BERNARDINI S, SMITH D E. Developing domain-indepen-dent search control for Europa2[C]//Proc. of the International Conference on Automated Planning and Scheduling, 2007.
13	AI-CHANG M , BRESINA J , CHAREST L , et al. Mapgen: mixed-initiative planning and scheduling for the Mars exploration rover mission[J]. IEEE Intelligent Systems, 2004, 19 (1): 8- 12. doi: 10.1109/MIS.2004.1265878
14	BRESINA J L, MORRIS P H. Mission operations planning: beyond MAPGEN[C]//Proc. of the IEEE 2nd International Conference on Space Mission Challenges for Information Technology, 2006.
15	孙泽洲, 张廷新, 张熇, 等. 嫦娥三号探测器的技术设计与成就[J]. 中国科学(技术科学), 2014, 44 (4): 331- 343.
	SUN Z Z , ZHANG T X , ZHANG H , et al. The technical design and achievements of Chang'E-3 probe[J]. Scientia Sinica(Technologica), 2014, 44 (4): 331- 343.
16	张辉, 卢皓, 于天一, 等. "祝融号"火星车遥操作技术[J]. 深空探测学报(中英文), 2021, 8 (6): 582- 591.
	ZHANG H , LU H , YU T Y , et al. Teleoperation technology of Zhurong Mars rover[J]. Journal of Deep Space Exploration, 2021, 8 (6): 582- 591.
17	高宇辉, 师明, 蔡敦波, 等. 一种通用型遥操作任务智能规划方法研究与应用[J]. 深空探测学报(中英文), 2021, 8 (2): 140- 146.
	GAO Y H , SHI M , CAI D B , et al. Research and application of a general teleoperation task intelligent planning method[J]. Journal of Deep Space Exploration, 2021, 8 (2): 140- 146.
18	SACERDOTI E D. The nonlinear nature of plans[R]. California: Stanford University Institute, 1975.
19	YANG Q . Formalizing planning knowledge for hierarchical planning[J]. Computational Intelligence, 1990, 6 (1): 12- 24. doi: 10.1111/j.1467-8640.1990.tb00126.x
20	KAMBHAMPATI S , HENDLER J A . A validation-structure-based theory of plan modification and reuse[J]. Artificial Intelligence, 1992, 55 (2-3): 193- 258. doi: 10.1016/0004-3702(92)90056-4
21	EROL K, HENDLER J A, NAU D S. UMCP: a sound and complete procedure for hierarchical task-network planning[C]//Proc. of the International Conference on Artificial Intelligence Planning Systems, 1994: 249-254.
22	EROL K , NAU D S , SUBRAHMANIAN V S . Complexity, decidability and undecidability results for domain-independent planning[J]. Artificial Intelligence, 1995, 76 (1-2): 75- 88. doi: 10.1016/0004-3702(94)00080-K
23	NAU D , AU T C , ILGHAMI O , et al. SHOP2: an HTN planning system[J]. Journal of Artificial Intelligence Research, 2003, 12 (20): 379- 404.
24	LI M L , WANG H W , QI C , et al. Handling temporal constraints with preferences in HTN planning for emergency decision-making[J]. Journal of Intelligent & Fuzzy Systems, 2016, 30 (4): 1881- 1891.
25	HUNSBERGER L . Efficient execution of dynamically controllable simple temporal networks with uncertainty[J]. Acta Informatica, 2016, 53 (2): 89- 147. doi: 10.1007/s00236-015-0227-0
26	TATE A , DALTON J , LEVINE J . O-Plan: a web-based AI planning agent[J]. New Directions in AI Planning, 1996, (31): 247- 262.
27	CASTILLO L , FDEZ-OLIVARES J , GARCıA-PEREZ O , et al. SIADEX: an integrated planning framework for crisis action planning[J]. AI Communications, 2005, 18 (4): 257- 268.
28	ALLEN J F . Towards a general theory of action and time[J]. Artificial Intelligence, 1984, 23 (2): 123- 154. doi: 10.1016/0004-3702(84)90008-0

状态受限事件	动作集合	状态集合	状态转移条件	动作持续时间/s
遥测模式I₀	定向天线设置A_{0_1}	[0, 1]	确定遥测周期	10
遥测模式I₀	遥测周期修改A_{0_2}	[0, 1]	周期修改为5 s	300
移动盲走模式I₁	注入模式表A_{1_1}	[0, 1]	器上FLASH正常	60
	陀螺零偏修正A_{1_2}	[0, 1]	阈值已设置	2
	移动状态设置A_{1_3}	[0, 1]	电量满足移动条件	2
	火星车移动A_{1_4}	[0, 1]	GNC已获取控制权	620
	移动转向轮回零A_{1_5}	[0, 1]	-	2
	移动断电A_{1_6}	[0, 1]	-	60
	模式结束A_{1_7}	[0, 1]	-	60
遥测模式I₀	遥测周期复位A_{0_3}	[0, 1]	周期修改为300 s	10

动作序列	内容	时长/s	参数	指令序列
火星车移动	连续运动里程监测	100	default	XL_{1_4_1}
	盲走指令	2	2次	XL_{1_4_2}
	原地转弯	10	default	XL_{1_4_3}
	蟹行控制	30	default	XL_{1_4_4}
	盲走移动	478	default	XL_{1_4_5}

状态受限事件	动作集合	状态集合	状态转移条件	动作持续时间/s
通信模式I₂	定向天线机构自主控温使能A_{2_1}	[0, 1]	预报文件中有通信时间窗口	300
	接收机打开A_{2_2}	[0, 1]	-	1 500
	对地通信A_{2_3}	[0, 1]	-	1 680
	定向天线机构自主控温禁能、加热回路断开A_{2_4}	[0, 1]	-	10

[1]	杨清清, 高盈盈, 郭玙, 夏博远, 杨克巍. 基于深度强化学习的海战场目标搜寻路径规划[J]. 系统工程与电子技术, 2022, 44(11): 3486-3495.
[2]	郭建国, 苏亚鲁. 高超飞行器自适应动态规划的控制系统设计[J]. 系统工程与电子技术, 2021, 43(6): 1628-1635.
[3]	周鑫, 王维平, 朱一凡, 王涛, 井田. 基于顺次分配机制的无人装备体系架构方案空间搜索方法[J]. 系统工程与电子技术, 2021, 43(11): 3211-3219.
[4]	吴文海, 郭晓峰, 周思羽. 基于NURBS和GOBL-ACDE的航迹规划算法[J]. 系统工程与电子技术, 2020, 42(5): 1073-1082.
[5]	刘正雄, 司继康, 陈刚, 黄攀峰. 面向遥操作手眼协调的虚拟仿真交互控制方法[J]. 系统工程与电子技术, 2020, 42(5): 1146-1151.
[6]	张袁鹏, 郑岱堃, 李昕哲, 孙永健. 基于隐马尔可夫模型的动态规划检测前跟踪算法[J]. 系统工程与电子技术, 2019, 41(11): 2479-2487.
[7]	钟雷, 李勇, 牟之英, 程伟, 李浩彬. 未知强杂波下基于DP-TBD的雷达弱目标检测[J]. 系统工程与电子技术, 2019, 41(1): 43-49.
[8]	向云武, 章文毅, 田妙苗. 卫星数据传输全流程调度及优化算法[J]. 系统工程与电子技术, 2018, 40(6): 1288-1293.
[9]	陈建勇, 陈长康, 孙明军. 连续搜索路径的最优化计算[J]. 系统工程与电子技术, 2018, 40(5): 1155-1159.
[10]	颜培远, 刘曙, 王君. 基于动态规划遗传算法的防空部署优化模型[J]. 系统工程与电子技术, 2018, 40(10): 2249-2255.
[11]	郝振兴, 罗继勋, 胡朝晖. 空基双探测端分布式定位系统的航迹控制[J]. 系统工程与电子技术, 2016, 38(8): 1886-1891.
[12]	刘正雄, 黄攀峰台健生. 空间遥操作预测仿真快速图形碰撞检测算法[J]. 系统工程与电子技术, 2016, 38(7): 1690-1696.
[13]	王硕, 张奕群. 检测低信噪比运动目标的改进DP算法[J]. 系统工程与电子技术, 2016, 38(10): 2244-2251.
[14]	付雄，王义波，朱鑫鑫，韩京宇. 数据网格中QoS感知的副本放置方法[J]. 系统工程与电子技术, 2014, 36(4): 784-788.
[15]	蔡文新, 方洋旺, 李锐, 伍友利. 离散Markov跳变线性系统最优控制[J]. Journal of Systems Engineering and Electronics, 2012, 34(7): 1458-1462.