Research On Intelligent Control System Of Grab Mechanism For Image And Data Fusion

With the transformation and upgrading of China's intelligent manufacturing production,the demand for intelligent assembly lines is getting higher and higher.A large number of enterprises apply industrial robots and vision technology to automated production lines to complete the inspection,sorting and handling of workpieces.However,in the production and assembly of television,manual assembly is still the mainstay.This paper applies machine vision to the intelligent TV assembly line to achieve automatic detection and positioning,and improve production efficiency.Firstly,according to the parameters and characteristics of the TV backplane,the overall scheme of the monocular adaptive grab robot is designed and designed.The design of the rectangular coordinate robot,the adaptive manipulator transmission structure and the servo drive system are separately selected.The static finite element analysis of the adaptive manipulator structure designed and processed by NX Nastran software is carried out.By analyzing the stress and strain cloud diagram of the manipulator structure,the adaptive manipulator structure of the design meets the current practical application requirements.Secondly,a visual grab system is designed for the TV backplane assembly process,the industrial camera installation position and fixed focus shooting mode are determined,and the machine vision hardware system is selected and selected.Then analyze and calculate the internal and external parameters of the camera lens for distortion calibration.Then use Halcon for backplane appearance recognition and location analysis.The image is smoothed using image enhancement and Gaussian filtering with smaller errors to highlight the information of interest in the image and eliminate noise.Then we study the performance and difference of backplane feature contour images extracted by Prewitt edge operator,Sobel edge operator,Laplace edge operator and Canny edge operator,and analyze the high and low thresholds in Canny edge algorithm to get the best edge contour.Effect chart.The backplane image is divided into two parts by the TV backplane feature,which is detected according to the image properties and template matching method.In order to improve the system running speed,a normalized(NCC)template matching algorithm for improving the search interval is designed.Calculate the matching value to complete the appearance detection of the TV backplane.Then,according to the image processing to obtain the position information and the backboard grab control flow,a hybrid visual servo control strategy combining image and position data is proposed.After the image 2D data is input into the controller,the robot grabs the gesture and passes the robot.The absolute coordinate system in the controller converts the two-dimensional coordinates of the image into three-dimensional coordinates of the Cartesian coordinate robot to realize the adaptive crawling process of the robot.And the S-curve plan is made for the Z-axis point-to-point motion of the Cartesian coordinate robot,and the B-spline curve smoothing is performed on the X-axis and Y-axis polyline motion trajectory to avoid the impact phenomenon during the movement.Finally,build an adaptive crawling experimental platform,design the PC software and communicate with PLC and industrial cameras.Aiming at the image error,an error compensation method based on the difference between the standard template center parameter value and the incremental error value is designed to improve the grasping precision.Based on the designed adaptive crawling robot experimental platform,the comprehensive experimental analysis is carried out.