用约束方法改进机器人感知与规划，确保不越界、不碰撞。
Constrained Stein Variational Gradient Descent for Robot Perception, Planning, and Identification
- 将约束优化融入Stein变分梯度下降算法，支持任意约束类型。
- 生成满足精确避障、表面对齐等约束的多解分布。
- 适用于运动规划、关节姿态、点云物体位姿等任务。
机器人领域的许多核心问题可建模为约束优化问题。系统常存在不确定性，或需识别多个高质量可行解。为此，我们提出两种新框架，将约束优化原理应用于变分推断中的Stein变分梯度下降算法。该通用框架支持多种约束优化器，可处理任意约束。实验表明，该方法能学习近似分布且严格遵守约束。具体包括：精确避障的机器人运动规划分布、SE(3)流形上满足表面对齐约束的机械臂关节角分布，以及基于点云、满足表面对齐约束的对象位姿分布。
原文摘要 · Abstract (English)
Many core problems in robotics can be framed as constrained optimization problems. Often on these problems, the robotic system has uncertainty, or it would be advantageous to identify multiple high quality feasible solutions. To enable this, we present two novel frameworks for applying principles of constrained optimization to the new variational inference algorithm Stein variational gradient descent. Our general framework supports multiple types of constrained optimizers and can handle arbitrary constraints. We demonstrate on a variety of problems that we are able to learn to approximate distributions without violating constraints. Specifically, we show that we can build distributions of: robot motion plans that exactly avoid collisions, robot arm joint angles on the SE(3) manifold with exact table placement constraints, and object poses from point clouds with table placement constraints.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。