Robot Compliance Control Base On Force And Vision

The force and vision of robot is an important part of acquisition and feedback in robot control.With the increasing complexity of robot tasks,the research of multi-sensor data fusion is more and more important.With the increasing complexity of robot tasks,the research of multi-sensor data fusion is more and more important.In this paper,the force and vision of the robot in the control process are studied,and the research on the compliance control is completed on this basis.This article first introduces the various parameters of the robot,according to the Denavit Hartenberg method to establish the kinematics model of the robot,by analytic method solve inverse kinematics equations of the robot,And in Matlab to write kinematics forward and inverse solution program.Proof the forward and inverse kinematics solutions of the robot using the Robotics Toolbox,The simulation results show that the kinematics equation is solved correctly.The camera imaging principle in the vision system is introduced,the relationship between each coordinate system in the process of vision acquisition is explained,the camera calibration is completed,the camera's internal parameters are obtained,the robot and the camera are calibrated by hand and eye,and the camera's external parameter matrix is obtained through a number of experiments.The mechanical model of robot terminal contact with the environment is established,and the classical impedance control in robot compliance control is studied.By contrast,based on the location of the impedance control as a topic in robot compliance control strategy,based on the proposed according to different scenarios,force control based on kalman filtering and adaptive impedance control,two kinds of control strategy for simulation and confirmed the feasibility of the control strategies in their respective application scenarios.The image acquired by the camera is denoised and compared in a variety of binarization threshold solution methods to obtain the calculation formula suitable for the scene.The processed image is segmated to obtain the target edge contour,and the target contour is expressed in the camera coordinate system by solving the PnP problem.The target contour through transition to the robot base coordinate system is derived,said introduced with Kuka tool coordinate system calibration method,provided by the company of robots in the Solidwork,tools and the target workpiece at the end of the 3 d model establishment and processing,import Simscape Multibody in complete robot visual force control simulation,the simulation of the force feedback data for processing the classical impedance control based on kalman filter in the trajectory tracking with force feedback control force.The results show that the latter has better control effect.