微电子组装工艺参数优化研究

发布时间：2018-04-13 10:15

本文选题：微电子组装 + 工艺参数优化　；参考：《西南交通大学》2017年硕士论文

【摘要】：工业4.0和智能制造带动了制造业的转型和升级,工业4.0的特征是多品种小批量的定制生产,根据定制产品,能快速、智能化调整和设置制造工艺和设备的参数。微组装工艺技术包含着大量的输入参数,这些参数与输出之间具有复杂的非线性关系,在实际生产过程中,往往需要大量的实验进行多次调整修正才能最终确定满足要求的工艺参数,不能智能制造。因此,通过对实际微电子组装工艺参数的分析和研究,建立微电子组装工艺专家系统,是实现微电子组装智能制造的基础。微组装技术实质上是一种高密度封装联合微焊接技术为基础的综合性工艺技术,其主要的工艺流程包括锡膏印刷、贴装、回流焊、检测和返修。本文分析和归纳出不同封装类型的组装工艺流程及其相关的工艺参数的工程经验值,主要研究微组装锡膏印刷和回流焊工艺参数的优化方法,并设计了微电子组装工艺参数专家系统。首先,给出一种析因实验设计和响应曲面中心复合设计法结合的锡膏印刷工艺参数的优化方法。通过对焊膏印刷工艺参数的研究,确定影响印刷质量的关键参数。在PCB良好且焊盘设计合理的情况下,采用5因数2水平的部分析因实验,以印刷压力、印刷速度、刮刀角度、脱模速度和脱模间距为实验因子,并以平均每块实验板的焊膏填充率作为响应结果。利用Minitab软件对实验因子进行分析,确立了影响印刷质量的主要因素。根据析因实验的结果,采用响应曲面法对印刷工艺参数进行优化,进一步确定最佳的印刷工艺参数。其次,给出一种用神经网络技术建立回流温度曲线输入输出映射模型,并用改进的遗传算法进行回流温度曲线优化的方法。先分析回流焊接工艺的过程中,确定影响回流曲线设置的主要工艺参数。再采用传统遗传算法优化的BP神经网络和改进的遗传算法优化的BP神经网络对回流曲线模型进行预测,并用Matlab软件对两种算法进行仿真分析,验证了改进的遗传算法在回流曲线优化问题上的优越性,进一步采用改进遗传算法对回流温度曲线参数设置进行优化,得出理论上最优的工艺参数组合。最后通过实际的回流焊接验证优化算法得出的最佳工艺参数的正确性。最后,采用VC6.0编程技术和SQL Server 2000数据库技术开发出微电子组装工艺参数专家系统。该系统利用规则库,对不同封装类型的组装工艺参数进行选择和设置,使用户可以在开发的软件平台上进行微组装工艺优化设置。
[Abstract]:Industry 4.0 and intelligent manufacturing drive the transformation and upgrading of manufacturing industry 4.0 is characterized by a variety of small batch customized production according to customized products can quickly and intelligently adjust and set manufacturing process and equipment parameters.The microassembly process contains a large number of input parameters, which have a complex nonlinear relationship with the output.It often needs a lot of experiments to adjust and revise many times in order to finally determine the process parameters that meet the requirements, so it can not be manufactured intelligently.Therefore, through the analysis and research of the actual microelectronic assembly process parameters, the establishment of the microelectronic assembly process expert system is the basis for the realization of intelligent microelectronic assembly manufacturing.Microassembly technology is essentially a comprehensive technology based on high-density packaging and micro-welding technology. Its main process processes include paste printing, paste, reflow welding, detection and rework.This paper analyzes and generalizes the engineering experience values of different packaging types of assembly process and related process parameters, and mainly studies the optimization methods of the process parameters of micro-assembly solder paste printing and reflow welding.An expert system for microelectronic assembly process parameters is designed.Firstly, a method of optimizing the printing process parameters of tin paste is presented, which combines the experimental design of factorial analysis with the complex design of response surface center.The key parameters affecting the printing quality are determined through the study of the printing process parameters of butt solder paste.Under the condition of good PCB and reasonable pad design, the experiment of 5 factor 2 level partial analysis is used. The experimental factors are printing pressure, printing speed, blade angle, demoulding speed and demoulding distance.The average soldering paste filling rate of each test plate was taken as the response result.The experimental factors were analyzed by Minitab software, and the main factors affecting printing quality were established.According to the results of factoring experiment, the response surface method is used to optimize the printing process parameters and to determine the best printing process parameters.Secondly, an input and output mapping model of reflux temperature curve is established by using neural network technology, and an improved genetic algorithm is used to optimize the reflux temperature curve.In the process of reflow welding, the main process parameters affecting the setting of reflux curve are determined.Then BP neural network optimized by traditional genetic algorithm and BP neural network optimized by improved genetic algorithm are used to predict the model of backflow curve, and the two algorithms are simulated and analyzed by Matlab software.The superiority of the improved genetic algorithm in the optimization of the reflux curve is verified. Further, the parameters of the reflux temperature curve are optimized by the improved genetic algorithm, and the optimal combination of the process parameters is obtained.Finally, the correctness of the optimal process parameters obtained by the optimization algorithm is verified by the actual circumfluence welding.Finally, an expert system for microelectronic assembly process parameters is developed by using VC6.0 programming technology and SQL Server 2000 database technology.The system makes use of the rule base to select and set the assembly process parameters of different packaging types so that the user can optimize the assembly process on the developed software platform.
【学位授予单位】：西南交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN405

【参考文献】