Research On Key Technology Of Delta Robot Dynamic Grabbing System Based On Machine Vision

As the core carrier of intelligent manufacturing,industrial robots are increasingly used in various highly repetitive process production lines.In order to further improve the flexibility and automation of the robot,so that the robot can be used in more experimental scenarios,the industrial robot dynamic grasping system based on machine vision has become one of the research hotspots in the field.In this paper,based on the existing technologies of machine vision and Delta robot,the dynamic grasping control system of Delta robot based on machine vision is studied by combining the requirements of dynamic target grasping operation in the field of food internal packaging.The main work contents are as follows:(1)According to the requirements of the project,the composition and structure of the whole robot dynamic grasping system were analyzed and studied,and the hardware selection in the visual system was completed.By analyzing the advantages and disadvantages of the new soft contact gripper,the practical working scenarios which can be applied are discussed,and the corresponding control scheme is designed.(2)Firstly,the camera was calibrated,then the conversion relationship between each coordinate system was analyzed,and the parameters of the camera were obtained by Matlab simulation.Then,the principle of image processing technology is analyzed,and the actual effect of existing image preprocessing related methods is compared through experiments.Finally,the method based on template matching is used to realize the recognition and positioning of the workpiece.By establishing the model of the target object,the similarity measure method is used to compare the template with all the objects in the image to realize the recognition of the target object.(3)Achieve the tracking and grasping of dynamic targets.Firstly,the conveyor belt is calibrated,and then the dynamic position of the target object is obtained by combining the offset of the conveyor belt and the feedback value of the encoder.The method based on position prediction is used to grasp the target object dynamically,and the actual grasping position of the robot is obtained by analyzing the trajectory planning of the robot.At the same time,the influence of position feedback on the grasping speed of robot in the situation of low grasping precision is analyzed through experiments.(4)A method for optimizing the grasping order of target objects in the process of dynamic grasping of robots is proposed.First,calculate the time when each target object goes out of the robot's working interval,and sort the target object for the first time according to the size of time.Then,based on the first sorting,the complete image is divided into regions,the optimal path of all the objects in each region is calculated,and the target object is sorted again by combining all regions to obtain the final capture order.Finally,an experimental platform was built to verify the algorithm,which proved that the algorithm could significantly improve the grasping efficiency of the robot.