arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2203.00975cs.ROcs.LG
收藏

L4KDE:面向运动动力学树扩展的学习方法

L4KDE: Learning for KinoDynamic Tree Expansion

The University of Sydney(悉尼大学);University of Utah(犹他大学);NVIDIA(英伟达)

Tin Lai, Weiming Zhi, Tucker Hermans, Fabio Ramos

AI总结:

本文提出 L4KDE 方法,利用神经网络批量预测状态间转移成本以改进运动动力学树扩展中的节点选择,在保持渐近最优性的同时显著提升规划性能。

摘要:

我们提出了面向运动动力学规划的运动动力学树扩展学习方法(L4KDE)。基于树的规划方法,例如快速探索随机树(RRT),是在连续状态空间运动规划中寻找全局最优计划的主导方法。这些方法的核心是树扩展,即将新节点添加到不断扩展的树中的过程。我们研究基于树的规划的运动动力学变体,其中系统动力学和运动学约束是已知的。为了快速选择用于连接新采样坐标的节点,现有方法通常无法通过优化找到转换到采样坐标成本较低的节点。相反,它们使用坐标之间的欧几里得距离等度量作为选择候选节点以连接到搜索树的启发式方法。我们提出 L4KDE 来解决这一问题。L4KDE 使用神经网络预测查询状态之间的转移成本,可以批量高效计算,与常用的启发式方法相比,能够提供更高质量的转移成本估计,同时保持几乎必然的渐近最优性保证。我们通过实验证明,L4KDE 在各种具有挑战性的系统动力学上带来了显著的性能提升,具备在同一模型类别的不同实例之间泛化的能力,并且能够与一系列现代基于树的运动规划器协同工作。

英文摘要:

We present the Learning for KinoDynamic Tree Expansion (L4KDE) method for kinodynamic planning. Tree-based planning approaches, such as rapidly exploring random tree (RRT), are the dominant approach to finding globally optimal plans in continuous state-space motion planning. Central to these approaches is tree-expansion, the procedure in which new nodes are added into an ever-expanding tree. We study the kinodynamic variants of tree-based planning, where we have known system dynamics and kinematic constraints. In the interest of quickly selecting nodes to connect newly sampled coordinates, existing methods typically cannot optimise to find nodes that have low cost to transition to sampled coordinates. Instead, they use metrics like Euclidean distance between coordinates as a heuristic for selecting candidate nodes to connect to the search tree. We propose L4KDE to address this issue. L4KDE uses a neural network to predict transition costs between queried states, which can be efficiently computed in batch, providing much higher quality estimates of transition cost compared to commonly used heuristics while maintaining almost-surely asymptotic optimality guarantee. We empirically demonstrate the significant performance improvement provided by L4KDE on a variety of challenging system dynamics, with the ability to generalise across different instances of the same model class, and in conjunction with a suite of modern tree-based motion planners.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑