An interpretable neural network-based to dynamic parameter identification for collaborative robots

doi:10.1016/j.aei.2025.103962

An interpretable neural network-based to dynamic parameter identification for collaborative robots

2025-10-22

0

PRE

AI

DOI：10.1016/j.aei.2025.103962

Original

Original request for help

Save

Abstract

En 中文

• Robot dynamics network (RDN) integrates interpretable hyperparameters with dynamic parameters, enhancing accuracy and generalizability. • The integrated joint friction moment model takes into account more nonlinear errors, which improves the robot modeling accuracy. • Cross validation shows RDN’s superior performance over traditional and neural network-based identification methods.

An interpretable neural network-based to dynamic parameter identification for collaborative robots

Abstract

Journal

Organization

Cited Papers