arXiv 2503.069952025-03-10cs.RO

基于物理信息神经网络预测控制的四足机器人运动控制

Physics-informed Neural Network Predictive Control for Quadruped Locomotion

School of Intelligent Systems Engineering, Shenzhen Campus of Sun Yat-sen University（中山大学深圳校区智能系统工程学院）；General Embodied AI Center of Sun Yat-sen University（中山大学通用具身人工智能中心）；Guangdong Provincial Key Laboratory of Fire Science and Intelligent Emergency Technology（广东省火灾科学与智能应急技术重点实验室）；College of Engineering, Peking University（北京大学工学院）；Robotics and Autonomous Systems Thrust, The Hong Kong University of Science and Technology (Guangzhou)（香港科技大学（广州）机器人与自主系统部）；George W. Woodruff School of Mechanical Engineering, Georgia Institute of Technology（佐治亚理工学院乔治·W. 伍德拉夫机械工程学院）

Haolin Li, Yikang Chai, Bailin Lv, Lecheng Ruan, Hang Zhao, Ye Zhao, Jianwen Luo

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

本文提出OPI-PINNPC框架，将在线负载辨识与物理信息神经网络结合并嵌入模型预测控制，在25-100 kg负载下实现四足机器人跟踪精度提升35%且收敛更快。

摘要：

本研究引入了一个统一的控制框架，以解决在未知负载下实现四足机器人精确运动控制的挑战，该框架被命名为基于在线负载辨识的物理信息神经网络预测控制（OPI-PINNPC）。通过将在线负载辨识与物理信息神经网络（PINNs）相结合，我们的方法将辨识出的质量参数直接嵌入神经网络的损失函数中，从而在适应变化负载条件的同时确保物理一致性。受物理约束的神经表示作为非线性模型预测控制器中的高效代理模型，使得尽管腿足运动动力学复杂，仍能实现实时优化。在我们的四足机器人平台上的实验验证表明，在25-100 kg的多种负载条件下，位置和姿态跟踪精度提高了35%，且与先前的自适应控制方法相比收敛速度显著加快。我们的框架提供了一种自适应解决方案，在不牺牲计算效率的情况下维持变负载条件下的运动性能。

英文摘要：

This study introduces a unified control framework that addresses the challenge of precise quadruped locomotion with unknown payloads, named as online payload identification-based physics-informed neural network predictive control (OPI-PINNPC). By integrating online payload identification with physics-informed neural networks (PINNs), our approach embeds identified mass parameters directly into the neural network's loss function, ensuring physical consistency while adapting to changing load conditions. The physics-constrained neural representation serves as an efficient surrogate model within our nonlinear model predictive controller, enabling real-time optimization despite the complex dynamics of legged locomotion. Experimental validation on our quadruped robot platform demonstrates 35% improvement in position and orientation tracking accuracy across diverse payload conditions (25-100 kg), with substantially faster convergence compared to previous adaptive control methods. Our framework provides a adaptive solution for maintaining locomotion performance under variable payload conditions without sacrificing computational efficiency.

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。