Font Size: a A A

Research On Key Technologies Of High Speed And High Precision SMT Machine Mounting Production

Posted on:2019-07-21Degree:DoctorType:Dissertation
Country:ChinaCandidate:Y H HuangFull Text:PDF
GTID:1368330575478834Subject:Mechanical engineering
Abstract/Summary:PDF Full Text Request
The rapid development of precision electronics manufacturing technology and increasingly fierce market competition have brought serious challenges for surface-mount technology of high speed and high precision SMT machines.Different from mature mounter technology in developed countries,domestic machines are characterized by poor performance,slow speed and poor accuracy compared with the imported machine.As for how to realize the fast and accurate positioning of the precision electronic surface mount manufacturing,shorten the mounting time and increase the production efficiency,and help domestic high-speed high-precision mounter machine in the current market environment to obtain more development,the key technologies of high-speed high-precision surface mount manufacturing have brought us a new research idea.This paper introduces the research background and significance of surface mount manufacturing of high speed and high precision mount machine as well as its related key technologies,analyzes the current research status of surface mount technology for high speed and high precision mount machine at home and abroad,and summarizes the shortcomings and limitations of the current research on the precise positioning and mounting optimization of high speed and high precision surface mounting production.Based on this,the main contents of this paper are as follows:1.In view of the high-precision and high-speed surface mount production requirements of SMT machines,a high-speed and high-precision SMT machine system is presented for automatic,reliable and fast surface mount production,which is composed by the mechanical system,control system,software system and automatic optical detection system.This paper focuses on the establishment of the automatic optical detection system,to achieve equipment calibration,reference point positioning,defect identification,alignment and posture correction.The automatic optical inspection system can meet the high-speed high-precision mount machine positioning requirements with mounting speed of 30,000/hour and the accuracy of 20um.2.Aiming at the characteristics of cross-shaped fiducial image of PCB board mounted by high-speed and high-precision placement machine,a geometric feature recognition and location algorithm is proposed.The localization algorithm mainly includes two parts:pattern training and real-time search.Experimental results show that,compared with the traditional location algorithm of edge algorithm,the proposed algorithm not only has good adaptability to the linear change of gray and non-linear gray-scale,but also supports the change of scale and angle.The positioning deviation does not exceed 0.5pixel with good robustness,which is suitable for the PCB board mounting of the cross-shaped mark point.3.Aiming at the present big precision error and poor adaptability of circular mark benchmark fast positioning for high-precision and high-precision SMD components(such as 0402,0603 and so on)mounted with PCB,an improved Cramer-Rao lower bound algorithm is proposed.The algorithm uses the Otsu threshold method to realize the preliminary segmentation based on image features.By introducing a Gaussian kernel into the canny operator,a set of dom inant sets of edges is extracted on the multi-scale space,and then the improved Cramer-Rao lower bound operator is used to dominate point set for a circle fitting,so as to accurately locate the mark point.The experimental results show that this proposed method is of good robustness.Compared with LSF,AF and the Hough circle detection,the speed is at least 1.3 times that of LSF and AF,and 2.6 times that of the Hough circle detection method.4.Aiming at the interfering information and high precision&speed requirements of circular mark benchmark fast positioning for high speed and high precision chip components(ADS7841E,MP2359DJ)mounted with PCB,an arc projection algorithm is proposed.The arc projection algorithm first make ruler search,projection difference and differential smoothing,and then use iterative two-point K-means clustering for denoising,and finally finish the Ransac circle fitting.The experimental results show that the proposed algorithm can detect sub-pixel Mark points with fast speed,robustness and high accuracy.The effectiveness of the method is verified by engineering application.5.In order to solve the multi-optimization mutual coupling process optimization for mounting production with multiple ADS7841E,MP2359DJ chip components and high nozzle requirement,a multi-objective hierarchical optimization with high matching requirement nozzle optimization as the main level is proposed to minimize the number of exchange times and matching.Then the linear programming is used for solution.The second level is to solve the optimization of mounting sequence and the feeder layout and feeding optimization in parallel.The mounting sequence optimization aims to minimize the mounting distance and is solved by an improved leapfrog algorithm to construct the initial solution population with improved nearest neighbor method.The feeder layout scheme is constructed by voting integral method.Then the feeding optimization is carried out to maximum the synchronous feeding and solved by the heuristic greedy algorithm.The experimental results show that this method has obvious performance optimization for PCB mounting with high nozzle requirements.The effectiveness of the method is verified by engineering application.6.As for the multi-optimization mutual coupling process optimization for mounting production with multiple 1210 or1812 resistor components and the relatively low nozzle requirement,a fast multi-objective hierarchical optimization method with low matching requirement for parallel nozzle and placement sequence optimization is proposed.The first main level model to minimize the number of nozzle exchange times and the mounting distance is established and solved by an improved leapfrog algorithm.At the same time,the leapfrog method is used to design the feeder layout scheme with voting integral method.Then the feeding optimization is carried out to maximum the synchronous feeding and solved by the leapfrog method.The experimental results show that this method has better performance than that of Section 5.The effectiveness of the method is verified by engineering application.
Keywords/Search Tags:high speed and high precision SMT machine, automatic optic inspection, mark point precise positioning, improved Gramer-Rao operator, circle fitting, arc projection algorithm, hierarchical multi-obj ective optimization
PDF Full Text Request
Related items