Robot contact force detecting method based on torque observation and friction identification

Abstract

The invention relates to a robot contact force detecting method based on torque observation and friction identification. A robot kinetic model and a generalized momentum are adopted to establish a joint torque observer for detecting joint torque change caused by a tail end contact force; a mode of combining an exponential friction model and a sine cosine function is adopted to describe the joint friction effect of a robot provided with a harmonic gear driving structure, a joint torque observation quantity is utilized to identify a joint friction parameter when an end effector of the robot has no contact force effect, and a joint friction torque is calculated according to the identified friction parameter; the torque observation quantity in a theoretical operating state of the robot sis regarded as a system model error, and an accurate joint contact torque observation quantity is finally obtained after the influences of friction and the system model error are considered, and finally a tail end contact force is calculated through a force Jacobian matrix of the robot.

Description

technical field [0001] The invention relates to a robot contact force detection method based on torque observation and friction identification, which utilizes robot dynamic equations and generalized momentum to construct a moment observer to detect the robot end contact force. Background technique [0002] With the transformation of the manufacturing industry from "automation" and "unmanned" to "intelligence" and "collaboration", more and more production fields require workers and robots to work together in a shared workspace. In the human-machine fusion production mode, more and more robots need to be in contact with the surrounding environment, and the contact force needs to be strictly controlled. Therefore, it is first necessary to detect the contact force on the end effector during the operation of the robot. Traditional contact force detection is mainly based on external force sensors, which not only increases the cost, but also makes the robot control system more comp...

AI Technical Summary

Problems solved by technology

The method proposed in this patent cannot compensate for robot joint friction errors and system model errors. It is theoretically feasible, but needs to be improved in practical applications.

Images