Reserch On The Coordinated Motion Of Dual-arm Robots Based On Vision

Compared with single-arm robot,dual-arm robot has th e advantages of strong applicability and high r eliability in complex tasks and changing environm ent.So dual-arm robots in industrial production,social services a nd other fields have a wide range of applications.As the key techno logy of the dual-arm robot system,the coordinated motion has becom e a hotspot.This paper investig ates the research s tatus quo of the dom estic and foreign dual arm robot,and investigates the research status quo of coordinated m ovement.Then,the NAO robot is used as the experim ental platform to study the coordinated motion of the dual-arm robots based on vision.The main contents are as follows.Firstly,the hardware and software para meters of NAO robot are introduced,and the kinematic chain of head,left arm and right arm is modeled by DH m odeling method.The kinem atics process of the three kinematic chains is analyzed,and the inverse kinematics solution of the left a nd right arm s and the derivation of the analytical solution are given.Secondly,the mathematical model of the visual system is analyzed and the cam era parameters are calibrated.Since the NAO robot head movement will m ake its m ain body unstabl e,each cam era is collected hundreds of times,and using the optimization method to get the hand-eye calibration matrix.Then,for the loose coordination environm ent,the image-based visual servo m ethod is used to realize the coordinated motion.At the same time,a mark point identification method is proposed.And then carried out experiments and simulations.The results demonstrate the feasibility of the coordinated exercise method.Finally,for the unique visual system and link structure of NAO robot,a tight c oordination scheme of two NAO robots is proposed to sim ulate a specific task.Th e relative position between two robots is determined by the calibration objects in the common field of the bottom camera of two NAOs.Use the camera on the head to identify the simulated stick tool.In addition,this paper proposed a m ethod of recognizing linear sticks with high adap tability,using improved bone refinement.Eventually,straight lines are rebuilt in the workpiece coordinate system.A method of planning is proposed for the taskor to require the tool to be perpendicular to the workpiece surface.Since NAO has only five DOF,this paper proposed a num erical solution for the unreachable spatial pose by giving up the rotation about the axis of the end-link.Eventu ally,coordination between multiple robots is achieved.And experiments and simulations were carried out using two NAOs and Matlab,and verified the feasibility of the scheme.