USB (Type-C) Soldering Machine Automatic Cable Module Design And Implementation

In order to realize industrial automation,"replacing workers by machine" is a popular research topic recently.The automation of soldering is especially imperative due to poisonous exhaust gas produced in the processing.With the rapidly development of electronics,data cables have become everyday consumables.In the soldering process of data cables,the arrangement of core lines in order is a remaining problem which restricts the automation.This thesis focused on this problem and studied USB cables and Type-C cables which represent two kinds of date cables.Mechanism was designed to fulfill this task,utilizing color image segmentation algorithms proposed in this thesis.Further experiments were carried out to verify the proposed method.The general USB cable with four core lines was the first study object.In order to disperse and array the core lines automatically,two different schemes,"cut" scheme and "stir"scheme,were proposed.Then experimental platforms and motion control programs were designed to test the feasibility of these two different schemes.The "stir" scheme was proved to be feasible.The free ends of some core lines were overlapped while Type-C cables were dispersed by the USB experimental platform,because of the Type-C cable has more core lines(the number ranges from 12 to 15).The "comb" mechanism was designed and combined to "stir" scheme to solve this problem.It was proved feasible in the following experiments carried out on the improved experimental platform.The "untwist" mechanism was further designed to improve the dispersion.And a ball twisting component was designed to avoid the harm to core lines in untwisting.In order to arrange the core lines efficiently,a vision system was set up to identify different core lines by color.After appropriate camera and lens were selected and the low angle shadowless light was proved to be more competitive than other lights,ideal images for algorithm testing were acquired.The acquired USB image was split in RGB and HSV color spaces.Then a well-directed threshold segmentation algorithm was designed according to the channel components,which was proved to be effective.Different algorithms of color image segmenting such as multichannel threshold segmentation,fuzzy c-means clustering,edge detection by Canny edge operator,edge detection by Sobel edge operator,edge detection by Scharr edge operator and watershed algorithm were tested on Type-C images,but no good result could be acquired.An improved multichannel threshold segmentation algorithm which coupled exponential enhancement with thresholding was proposed in this thesis.Utilizing the newly proposed algorithm,good segmenting results were acquired.