| Wires and connectors are the basic components for connecting circuit modules of electrical equipment.In the process of circuit assembly,it is necessary to take out the wire of specific specification from the wires group,and then insert the metal terminals of the two ends of the wire into the corresponding connector jack.In order to improve efficiency,this process is carried out on a separate work panel,which has specific requirements for the routing of wires.The production line mainly relies on manual operation,which has the disadvantages of low efficiency,high labor cost and monotonous work at present.Therefore,it is necessary to develop a full-automatic solution that can replace human and realize the wires positioning,grabbing and plugging terminals.There are many challenges to realize the robot automatic wiring because of the flexibility of wires,the density of wires harness and the complexity of connection.Started from the requirements of the production line,the key algorithms of visual processing for the identification and positioning of connector jacks,wire grabbing and terminal posture detection are designed in this study based on industrial robots and machine vision technology,and they are verifies on the prototype of the automatic wiring system of industrial robot.The main research works of the paper are as follows:(1)The design of the executive mechanisms and visual imaging scheme of the wiring system.Based on the analysis of the key steps in the process of wiring,an integrated solution for automatic wiring is proposed,and the coordinate transformation model,the hand-eye geometric model of the actuator and the moving platform are established according to the design scheme,and the calibration method of each module is given;(2)Connector target identification and jack positioning.In view of the interference caused by the imaging perspective and the clutter background,a connector recognition method based on cumulative quantized gradient orient features is proposed,which improves the recognition stability.On this basis,in order to solve the problem that it is difficult to locate multiple jack centers at the same time due to the dense arrangement of jacks,a multi jacks location algorithm based on the histogram feature of scan line is proposed to provide accurate target position information for the wiring-guide mechanism;(3)The end of wire detection and spatial pose estimation based on multi-view stereo vision technology.In the process of wires grabbing,the follow-up camera is used to collect the multi view image of wires,and the segmentation and edge extraction of the target area of the wire end are realized based on the color and shape features.Afterwards a matching method based on the local gradient orient histogram is proposed to solve the difficulty caused by the lack of texture features of the wire object to the stereo matching,and the spatial pose estimation of the wire is realized in order to guide the gripper to grasp the wire.After the wire grabbing,multiple images of the wire terminal are collected by the fixed camera to estimate the relative pose relationship between the terminal and the gripper,so as to determine the validity of grabbing and provide necessary conditions for the subsequent wire-jack assembly;(4)The construction of the experimental platform and the test and the verification of related algorithms.According to the design scheme the experimental platform is set up and the software of the wiring system is developed,and then the system model is calibrated and the accuracy is verified.After that,the positioning algorithm of connector jack,the algorithm of wire grabbing and the algorithm of terminal pose detection are verified based on the experimental platform.Through the above works,a wiring prototype system based on industrial robot has been designed,and the design of the visual positioning and detection algorithm in several key wiring steps is realized,and they are verified through the prototype.The works in this paper is expected to provide a feasible solution to realize wiring automatically for production line. |
PDF Full Text Request