Optimization Of Microelectronic Assembly Process Parameters

Industrial 4.0 and intelligent manufacturing led to the transformation and upgrading of the manufacturing industry.Industrial 4.0 is characterized by a variety of small quantities of custom production,which can be quickly and intelligently adjust the manufacturing process and equipment parameters according to custom products.Microelectronics assembly technology contains a large number of input parameters,and these parameters and the output has a complex nonlinear relationship between them.In the actual production process,the process parameters to meet the requirement often need to be adjusted by a large number of experiments to determine,which led to not intelligent manufacturing.Therefore,the establishment of microelectronic assembly technology expert system is fundamental to realize the intelligent manufacturing of microelectronic assembly by analyzing and researching the actual microelectronic assembly process parameters.Microelectronic assembly technology is virtually a comprehensive process technology based on a kind of high density packaging integrated micro welding technology.The main processes of microelectronic assembly technology include solder paste printing,patching,reflow soldering,inspection and rework.Different package types of micro-assembly process and its associated process parameters of engineering experience values are analyzed and summarized in this paper.In addition,the optimization method of micro-assembly solder paste printing and reflow process parameters was studied and the expert system of microelectronic assembly process parameters was designed in this paper.First of all,a method of optimizing the printing process parameters combined with experimental design and the response surface complex design method is given.The key parameters affecting the quality of printing are determined through the study of the solder paste printing.Based on the good of the PCB and rationalization of pad design,this paper uses the fractional design of five factors and two levels.In this experiment,the printing pressure,the printing speed,the blade angle,the stripping speed and the stripping distance are regarded as experiment factors,and the average paste filling rate of each experimental PCB is regarded as experimental factor.The Minitab software is used analyze the experimental factors,establishing the main factors affecting the printing quality.According to the results of the factorial experiment,the response surface methodology was used to optimize the printing process parameters in order to determine the optimal range of printing parameters.In addition,a method of establishing a reflow temperature curve input and output mapping model with neural network technology is proposed,and a modified genetic algorithm is used to optimize the reflow temperature curve.In the process of reflow soldering the main process parameters affecting the reflow profile are determined.The BP neural network optimized by traditional genetic algorithm and the BP neural network optimized by improved genetic algorithm are used to predict the reflow curve model.The two algorithms are simulated and analyzed by Matlab software,and the superiority of the improved genetic algorithm on the optimization of the reflow curve is verified.The genetic improved genetic algorithm is used to optimize the parameters of the reflow temperaturecurve,and the theoretical optimal combination of process parameters is obtained.Finally,the correctness of the optimal process parameters is verified by the actual reflow soldering.Finally,the expert system for microelectronic assembly process using VC6.0 programming technology and SQL Server 2000 is developed.The system use the Rule Repository to select and set up the assembly process parameters of different packing types,which makes the user can optimize the microelectronics assembly process on the software platform.