自变量对预投料精准度影响研究

Studying on how parameter influence the accuracy of material preparation

随着工业4．0时代的到来，如何实现印制线路板的“智造”，是大家迫切想知道并想实现的课题。特别是样板生产以及型号众多的中小批量的印制线路板企业，预投料的量通常由人依靠其工作经验进行控制，不仅工作量大，而且精度不高导致浪费或补投。如果可以使得产品最终产出最大程度地等于或接近客户订单需求量，就可以实现企业效益的最大化。本文的研究内容是如何利用统计学原理、通过设定恰当的变量并固化到计算机系统中实现这个目标，所以变量种类的选择、变量维度的选取，对模型的精确度起到至关重要的作用；当然过多的变量会使得数据分析量变大，系统运算复杂，反过来会影响最终结果的有效性，恰当地平衡这个关系也是研究的重点之一。

As the coming age of the industry 4.0, it is argent problem how to manufacture the PCB smartly, especially for sample manufacturing and the high-mix ＆ low-volume PCB factory. In these production processes, due to lots of different production types, material preparation and cutting are normally controlled by people＇s experience, which is not only time consuming but also has low accuracy. If the output results equal to the orders, then the enterprise can realize the maximum of benefits. In this paper, we study on how to meet the target by computer model calculation by setting proper parameters based on statistic principle mathematics model. The choose of parameter types and dimension are very significant to the model accuracy. However, excessive parameters will lead to large amounts of data analysis and complicated system calculation, which may affect the operating speed and results. Therefore it is one of important research contents how to balance the parameter setting and the running speed of the system.