跳转至

文章背景与核心概要

分层任务网络(HTN)规划近年来得到了蓬勃发展,然而其对数值推理的原生支持依然严重受限。本文深入研究了数值型完全有序 HTN(TOHTN)规划,并展示了如何利用可满足性模理论(SMT)自然地扩展标准的基于 SAT 的编码,以适应数值流畅量(numeric fluents)。

为了建立评估的基础标准,作者引入了一个专门针对数值型 TOHTN 规划的新基准测试集。其实验结果表明,这种直接的基于 SMT 的方法构成了一个极具竞争力的基准(baseline),为自动化规划中更具表现力的方法铺平了道路。


基于 SMT 的 HTN-SAT 编码实现数值型完全有序 HTN 规划

摘要与总结

Hierarchical Task Network (HTN) planning has experienced substantial growth, yet native support for numerical reasoning remains severely restricted. This paper investigates numerical Totally-Ordered HTN (TOHTN) planning and demonstrates how standard SAT-based encodings can be naturally expanded using Satisfiability Modulo Theories (SMT) to accommodate numeric fluents.

To establish a foundational standard for evaluation, the authors introduce a new benchmark suite tailored for numerical TOHTN planning. Their experimental findings indicate that this straightforward SMT-based approach serves as a highly competitive baseline, paving the way for significantly more expressive methodologies in automated planning.

分层任务网络(HTN)规划近年来得到了蓬勃发展,然而其对数值推理的原生支持依然严重受限。本文深入研究了数值型完全有序 HTN(TOHTN)规划,并展示了如何利用可满足性模理论(SMT)自然地扩展标准的基于 SAT 的编码,以适应数值流畅量。

为了建立评估的基础标准,作者引入了一个专门针对数值型 TOHTN 规划的新基准测试集。其实验结果表明,这种直接的基于 SMT 的方法构成了一个极具竞争力的基准,为自动化规划中更具表现力的方法铺平了道路。


元数据与出版详情

Metadata & Publication Details

  • arXiv ID: arXiv:2609.03938 [cs.AI]

  • 学科分类: 人工智能 (cs.AI)

    • Subject: Artificial Intelligence (cs.AI)
  • 作者团队:

  • Gaspard Quenard
  • Takudzwa Togarepi
  • Damien Pellier
  • Humbert Fiorino

    • Authors:
    • Gaspard Quenard
    • Takudzwa Togarepi
    • Damien Pellier
    • Humbert Fiorino
  • 提交日期: 2026年9月3日

    • Submission Date: September 3, 2026
  • 期刊引用: 第9届 ICAPS 分层规划研讨会论文集(HPlan 2026),第 32–36 页

    • Journal Reference: Proceedings of the 9th ICAPS Workshop on Hierarchical Planning (HPlan 2026) (pp. 32–36)

摘要原文

Abstract

While HTN planning has received significant attention in recent years, support for numerical reasoning remains very limited. In this paper, we investigate numerical Totally-Ordered HTN (TOHTN) planning and show how standard SAT-based encodings can be naturally extended with SMT to handle numeric fluents. In addition, we introduce a benchmark suite for numerical TOHTN planning, providing a first common basis for evaluation in this setting. Experimental results show that this simple encoding already constitutes a competitive baseline. This work opens the way to more expressive approaches to HTN planning.

(参见前文中文摘要)


全文与访问链接


相关资源与工具

Associated Resources & Tools

  • 外部引用: Google Scholar | Semantic Scholar | NASA ADS

  • 代码与目录: 可通过 CatalystX、Hugging Face 和 DagsHub 等社区平台访问(实时追踪请参考源平台上的 arXivLabs 集成面板)。

    • Code & Catalogs: Accessible via community platforms like CatalystX, Hugging Face, and DagsHub (refer to arXivLabs integration panels on the source platform for live tracking).