基于传感器的操作臂惯性参数在线识别神经网络模型

机器人惯性参数识别是机器人精确建模以及机器人控制和仿真的关键问题之一。机器人腕力传感器的接入将影响机器人系统的动力学特性 ,同时腕力传感器的输出也真实地反映了机器人的力作用和机器人末端的动力学特性。本文基于腕力传感器的输出信号 ,对在线识别机器人操作臂末端的惯性参数的方法进行了分析和研究 ,并建立了惯性参数在线识别的神经网络模型 ,网络学习后其权值即为辨识的惯性参数。

On-line Identification of Inertial Parameters of the End-effector of a Manipulator Based on Sensors

Inertial parameters identification, which plays an in important role in robotic mechanics, is an important base of robot modeling and controlling. The introduction of wrist force sensor to a manipulator not only changes the dynamic parameters of the end-effector of the manipulator, but also provides many useful messages of acting force and dynamic characteristics of the end-effector. In this paper, on-line dynamic parameter identification based on 6-dimensional wrist force sensor is studied and investigated, and a neural network model of the identification is established, the net's power after learning is the inertial parameters.