Research On The Robot Intelligent Sorting Technology Of Firebrick

The firebrick is a kind of widely-used building materials and has broad market. In the process of firebrick production, workers need to take firebricks down from the trolley. In order to improve the efficiency of logistics, the enterprises can use robot technology to complete this work. This dissertation studies the robot intelligent sorting technology and proposes a design scheme, which will promote the application and accelerate the promotion of this technology. To realize this technology, the dissertation completes the following three tasks:The gripper of the robot is designed. Firstly, this dissertation analyzes the tasks and functions of sorting system. According to the analysis, the function and structure of the gripper are determined. And then the dissertation proposes a design scheme of the gripper. The gripper consists of three parts: space adjusting mechanism, rotating mechanism, and gripping mechanism. The space adjusting mechanism uses the ballscrew for transmission. The rotating mechanism uses right-angle reducer for transmission. The gripping mechanism uses vacuum cup to grip firebricks. This kind of gripper can grip firebricks flexibly.This dissertation uses machine vision system to obtain the location information of firebricks. Eye-in-Hand system is used in the dissertation. But when the Eye-in-Hand system works, it adopts the advantages of Eye-to-Hand system. When processing the images, the dissertation uses grayscale processing and median filtering firstly, and then the images are sharpened by Prewitt gradient operator. According to the grayscale histogram, the image is processed into binary image. Considering that there is more than one target in an image, the dissertation uses marker algorithm to split the image, which can increase computation speed. What's more, the dissertation also checks the marker result according to the size of the firebrick to increase accuracy. Since the pattern of firebrick in the image is a rectangle, Hough transformation is used to detection the sides of the rectangle. According to the sides of rectangles, we can get the location information of rectangles. This way can ensure precision of position, so that the gripper can grip firebricks successfully. In addition, the dissertation also designs hardware of the control system, and analyzes the process of program design.In this dissertation, in order to improve the efficiency of the sorting system, the sorting strategy is optimized. After analyzing the problem, two optimization goals are proposed to establish the mathematical model, which are the times of gripping and complexity of program. Pareto genetic algorithm is used to solve the multi-objective optimization model. At last, the optimization model is verified with an example. The result shows that the sorting system can work efficiently.The dissertation also points out the imperfections of the sorting system, which is subject to further study. In a word, the sorting system designed in this dissertation conforms to theoretical analysis, which can be applied and complete sorting task successfully.