# h-NESDF：面向机器人运动规划的无限可微符号距离函数

> 原文链接：https://robotics.hellobike.com/papers/2026-11-04-2611-03944-dd636fd6
> 频道：papers | 更新日期：2026-11-04 | 标签：auto, arxiv
> 核心摘要：针对欧氏距离不可微导致运动规划梯度计算失效问题，提出h-NESDF算法。该方法引入平滑参数h实现集合间距离无限可微，支持GPU并行加速，在保持几何精度的同时为基于梯度的机器人控制与规划提供解析导数。

# h-NESDF：面向机器人运动规划的无限可微符号距离函数

::: v-pre
> 本文由 AI 根据公开内容整理，可能有误，请以原文为准：[h-NESDF: An Infinitely Differentiable Signed Distance Between Two Sets with Robotic Applications](https://arxiv.org/abs/2611.03944)（arXiv 2611.03944）

## 核心内容

arXiv论文提出h-NESDF（Normal Euclidean Signed Distance Function）算法，旨在解决机器人运动规划与控制中因欧氏距离不可微导致的梯度优化难题。该研究定义了适用于一般非凸集合的NESDF度量，并进一步通过引入平滑参数h构建广义形式，确保当h>0时函数具备无限阶可微性，且在h趋近0时收敛于标准NESDF，从数学层面消除了距离计算中的奇点障碍。

技术实现上，h-NESDF被设计为高度适配GPU并行计算架构，能够高效处理复杂几何体间的实时距离查询与梯度回传。仿真实验验证了其在运动规划任务中的有效性，表明该度量可直接嵌入基于梯度的轨迹优化器或避障控制器中，替代传统需数值近似或次梯度处理的距离函数，显著提升了依赖解析导数的机器人算法在非标环境下的收敛稳定性与计算效率。

## 关键看点

- **无限可微性**：引入平滑参数h使符号距离函数在h>0时处处无限可微，彻底解决欧氏距离在非光滑点不可导导致的梯度优化中断问题
- **GPU并行加速**：算法结构原生适配GPU并行计算，支持大规模点云或网格模型间的实时距离与梯度计算，满足高频控制回路时效要求
- **非凸集兼容**：NESDF定义放宽凸性假设，仅需轻度正则条件即可处理通用非凸几何体，扩展了可微距离在复杂机器人本体与环境建模中的适用范围
- **渐近一致性**：h-NESDF在平滑参数h趋近0时严格收敛至标准NESDF，保证算法在提升可微性的同时不牺牲原始几何距离的物理语义精度

## 工程与产业观察

该工作填补了机器人几何感知与梯度优化之间的数学断层，为端到端可微规划、强化学习碰撞惩罚及基于优化的MPC提供了严谨且高效的距离原语。其GPU友好特性使其具备部署潜力，有望成为下一代具身智能系统中连接物理约束与神经网络优化的标准基础设施组件。

## 英文摘要 (Abstract)

Computing distances between sets is important in many robotic applications, such as motion planning and control. In many applications, derivatives of the distance are needed as well. It is well known that the Euclidean distance is not always differentiable, which create issues to algorithms that need a derivative. In this paper, we introduce NESDF, the Normal Euclidean Signed Distance Function between two general (with mild regularity assumptions but possibly non-convex) sets. This metric is still not differentiable, but it allows us to introduce a generalization, $h$-NESDF, which has a smoothing parameter $h$ such that the metric is infinitely differentiable for all $h>0$ and that reduces to the NESDF when $h$ goes to $0$. The metric is amenable to GPU parallelization, and thus can be computed efficiently. We show extensive simulation results to evaluate its viability in motion planning.

作者：Vinicius Mariano Gon\c{c}alves, Pedro Antonio de Souza Campos Rodrigues
:::

## 原始参考来源 (Sources)

- [arXiv 2611.03944：h-NESDF: An Infinitely Differentiable Signed Distance Between Two Sets with Robotic Applications](https://arxiv.org/abs/2611.03944)