Research On 7-DOF Collaborative Robot Compliant Drag And Collision Detection Based On Joint Force Feedback

With the application scene of cooperative robots gradually expanding to complex and changeable fields such as medical rehabilitation,aerospace and 3C manufacturing,new challenges are also posed to robot performance.In addition to requiring robots to realize Cartesian space direct drag teaching,there is also a huge demand for robot space sharing,space drag obstacle avoidance and high-performance safe collision detection functions with better flexibility.The realization of the above-mentioned cooperative goal requires robots to have redundant degrees of freedom while efficiently completing a large number of dynamic operations.Considering that currently widely used cooperative robots such as UR are mostly designed with six degrees of freedom,and their control systems are only open to users and cannot be developed with bottom-level programming,it is difficult to meet the current demand for high real-time cooperation and obstacle avoidance.Therefore,how to improve the dragging flexibility and safety of cooperative robots in complex applications has become a key issue in the development of cooperative robots.Based on this,on the basis of self-designed and built 7-DOF lightweight robot control system,and based on joint force feedback information,this paper makes in-depth research on two key technologies of robot compliant dragging and safety collision detection,which lays a foundation for cooperative robots to deal with complex application scenarios.This article mainly includes the following aspects:First of all,in view of the real-time requirements and insufficient degrees of freedom for cooperative robots in current applications,this paper independently develops a seven-axis cooperative robot system with redundant degrees of freedom,and designs and implements the software and hardware systems.VC-based human-computer interaction control is adopted,GMAS controller compiles the underlying kinematics and dynamics control algorithm,and software control scheme of high-speed EtherCAT bus communication.The design and construction of the entire hardware system for selection,design,assembly and debugging are completed,which improves the control efficiency and provides an experimental platform for the realization of subsequent human-computer interaction algorithms.Secondly,aiming at the application requirements of compliant dragging and flexible obstacle avoidance in the joint space of the robot,a zero-force control algorithm based on position control is adopted to solve the dragging problems in the joint space and Cartesian space of the 7-DOF robot.Based on the designed configuration and parameters of the robot,the robot kinematics and dynamics modeling,analysis and parameter identification are completed,which provides theoretical and model basis for the human-computer interaction control algorithm.According to the design characteristics of the integrated torque sensor in the robot joint,the joint space and Cartesian space drag controllers are designed by using the method of combining zero force control and admittance control based on position control.Based on the principle of minimum acceleration speed,the comparative experiments show that under the normal drag speed of the robot,when the admittance coefficient value is 1/4500,the drag compliance is relatively good.Thirdly,in order to meet the safety requirements of man-machine cooperative work,the collision detection technology is studied by using the observed external moment method based on inverse dynamics.The method realizes the design of a joint collision detection controller based on a seven-degree-of-freedom robot dynamic model and a joint torque sensor feedback signal.The controller makes a difference between the real-time torque value collected by the torque sensor and the torque estimation value of the dynamic model to obtain an external torque estimation value.Aiming at the error problem caused by noise of the torque acquisition value,Kalman filtering is adopted to filter the external torque estimation value,and the joint collision threshold value estimation is further obtained based on the filtering result.In addition,in order to reduce the danger after human-machine collision,three collision stopping strategies are proposed.Through experiments,collision detection and response stopping simulation experiments are carried out on one of the common strategies,which verify the effectiveness of the methods and strategies.Finally,a 7-DOF robot algorithm verification platform is built,and joint space and Cartesian space drag experiments are carried out to verify the effectiveness of the controller design.Meanwhile,collision detection experiments are carried out in the robot operation process,and a collision stop strategy is added to verify the collision detection accuracy and practicability of the strategy.The experimental results show that the forcefree control method based on position control achieves the space obstacle avoidance target in the robot joint space drag,the Cartesian space drag error can be kept within 5mm,which meets the requirements of direct drag teaching.The collision detection method based on inverse dynamic external force estimation optimizes the average collision threshold to 5.6%of the joint maximum torque value,improves the collision detection accuracy,and realizes the expected collision and safe stop effects at the same time,which has great research significance.