外观
论文arXiv 2609.013511 个月前
面向高维 POMDP 的可扩展 Rao-Blackwellized 在线规划
原文链接：Scalable Rao-Blackwellized Online Planning for High-Dimensional POMDPs（arXiv 2609.01351 · 发布日期：2026-09-01）
Scalable Rao-Blackwellized Online Planning for High-Dimensional POMDPs
针对高维部分可观测环境中的在线规划，提出混合连续-离散信念的 RB-POMDP，并在机器人搜救任务中验证。
核心要点
- 高维状态下，采样式 POMDP 求解器因蒙特卡洛估计方差高而性能退化。
- 方法采用混合连续-离散信念表示，在树搜索中解析传播边缘化状态不确定性。
- 集成 FastSLAM 2.0 用于机器人搜救任务，并与纯采样式方法进行比较。
- 在相同计算预算下，以显著更少粒子和规划仿真获得更高累计奖励。
验证范围与局限：验证范围为集成 FastSLAM 2.0 的机器人搜救任务。
英文摘要 (Abstract)
Online planning under uncertainty remains a fundamental challenge for robotic systems operating in partially observable environments with high-dimensional state spaces. While sampling-based POMDP solvers enable approximate decision-making in large or continuous domains, their performance degrades as belief dimensionality increases due to the high variance inherent in Monte Carlo-based estimation. In this work, we extend the Rao-Blackwellized online POMDP (RB-POMDP) framework to improve its generalizability in high-dimensional settings through hybrid continuous-discrete belief representations. By analytically propagating uncertainty associated with marginalized state components during tree-based planning, the proposed approach reduces sampling-induced variance in value estimation. We demonstrate the effectiveness of this framework in a robotic search-and-rescue task by integrating it with FastSLAM 2.0. Experimental results show that the proposed planner achieves higher cumulative rewards using significantly fewer particles and planning simulations than purely sampling-based methods under equivalent computational budgets. These results suggest that structured high-dimensional robotic problems admitting tractable sufficient statistics can be effectively leveraged within the RB-POMDP framework for computationally feasible online decision-making.
作者：Lee, Jiho, Nisar Ahmed, Kyle Hollins Wray, Zachary N. Sunberg
来源
本页事实对应的公开页面，以原文为准。