Study On Tactile-Based Security Technology Of Cooperative Robots

Robots are playing an increasingly important role in human daily life such as industrial manufacturing and medical services with the maturing of robot technology.Interactions between human and robots are becoming more and more common.Due to the inertia of traditional industrial robots and the lack of safety protection measures,it is necessary to isolate robots with safety barriers during working process,which cannot meet the requirements of flexible manufacturing production and interactive safety requirements.Collaborative robots has been emerging and attract people's attention.When cooperative robots work with people on the same production line,it is inevitable that collisions will occur and harm their human coworkers.Therefore,the problem of human-robot interaction security needs to be solved urgently.In this paper,the safety technology of cooperative robots is studied based on a three-degree-of-freedom light robotic arm SoftArm III.The main research contents are collision detection and safety control of the cooperative robot.The aim of this study is to detect collisions in real time and to identify the type,position and direction of the collision force.According to the feedback information of collision force,the security control strategy is adopted to improve the safety of human-robot collaboration.The main work of this paper are as follows:In order to detect collision information,a wearable tactile sensing device is designed based on hierarchical structure.The tactile units of the tactile sensing device are calibrated statically.The mathematical models of each tactile unit are obtained based on the least square method.The contact recognition of tactile array in curved and in plane states are carried out respectively to verify the effectiveness of the tactile sensing device.The tactile sensing device is applied to the prototype of 3-DOF robotic arm.The kinematic model of the reference tactile unit is built in the frame of the robotic arm.The position and direction of the reference tactile unit in the base coordinate system of the manipulator is solved using D-H method.The position and direction of the rest tactile units and the pressure direction in normal are obtained through coordinate transformation based on the reference tactile unit.Based on this approach,a tactile-based collision force recognition algorithm and three security control strategies are proposed for the safe control of the robotic arm.The experimental platform of the manipulator safety control is built.The softwares of tactile information feedback and collision force classification are programmed in the upper computer.The softwares of motion control and security control are programmed in the lower computer..During the motion of the robotic arm,the safety control experiments are carried out for two different collision forces,such as emergency stop,the original path return,and the direction of the collision force.The blunt force and the sharp force collision are different in the response speed of the safety control strategy.The experimental results show that the algorithm of collision force recognition can recognize the collision force class correctly,and the three security control strategies can reduce the collision force,reduce the degree of collision injury,and improve the safety of human-robot interaction.At the same time,the experimental results of three groups of different collision forces are compared and analyzed.The results show that the response measures along the collision force direction can leave the collision point faster and have better safety control effect.It validates the feasibility and rationality of the algorithm of collision force recognition and the safety control strategy.Three safety control strategies of emergency stop,return along the old path,and return along the collision force direction are tested for the blunt force and the sharp force collisions in the situation of the robotic arm moving.The results shows that the collision force recognition algorithm can identify the collision force category.The three security control strategies can reduce the collision force,so as to reduce the impact damage and improve the security of human-robot interaction.At the same time,the experimental results of three groups of different collision forces are compared and analyzed.The results show that the strategy of return along the collision force direction can enable the robotic arm to leave the collision point faster and do less harm to the robot and its human coworker.It proves the feasibility and rationality of the algorithm of collision force recognition and the safety control strategy.