Research On PCB Drilling Process And Parameter Optimization Based On MATLAB

Normally, multi-layer PCB includes inner layers and outer layers. To achieve conductive interconnection between the inner and outer layers, the industry generally uses drilling process and hole wall copper plating technology.There are many parameters of PCB drilling process, if related parameters are not appropriately, it will cause quality defects. However, the industry for the determination of the drilling process parameters, almost depends on the engineers understanding of the past production status and experience, not form a scientific and systematic solution. Therefore, this thesis is committed to research and optimize parameters of drilling process; solve the quality problem of the precision of hole position in drilling process and improve production efficiency.In this thesis, based on drilling equipment and technology of Kunshan WUS Electronics Co., Ltd.(hole diameter: 0.35 mm, aspect ratio: 9:1), the primary contents and results are as follows:(1) Design orthogonal experiment with spindle speed, in feed, out feed, bit life, lifted height of spindle; analysis of CpK data by Minitab software, the result shows the key factors affecting the precision of hole position are spindle speed, bit life and in feed.(2) According to the drilling process parameters are complex and interactive, it proposes BP neural network method to forecast precision of hole position. Using MATLAB to train and test these samples come from orthogonal experiment, the result shows network simulation values correspond to target value.(3) Using the trained BP neural network to optimize drilling parameters, it shows the best technological parameters: 150 KRPM for spindle speed, 1800 holes for bit life, 115 IPM for in feed; According to the these parameters to produce PCBs, precision of hole position can meet the CpK target.In this thesis, research methods based on the experiments of actual production and artificial neural networks, breaking the bottleneck of traditional mathematical model which is difficult to establish, and develop a new path for the study of PCB drilling technology.