Automatic feeding method and system for multiple varieties and small batch of PCBs

Abstract

The invention relates to an automatic feeding method and system for multiple varieties and small batch of PCBs. The method comprises the following steps of: quantifying factor variable data and establishing a data model by analyzing and selecting a key influencing factor based on production feeding, scrap data and product order parameter information big data; embedding the data model into an ORACLE system and obtaining a system automatic prediction model; running a model program, automatically matching calculating the historical scrap rate of the orders in the generated system automatic prediction model, and automatically obtaining order information, the number of remainders and the number of online orders; predicting the scrap rate based on the order information and historical scrap rate; based on the predicted scrap rate, predicting the number of pre-fed materials of the order; checking whether abnormal data exists in the predicted scrap rate and the number of pre-fed materials of the order, and releasing production if not. The method and system eliminate difference in manual feeding experience, improve feeding stability, improve production efficiency, improve accuracy and promote automatic lean production.

Description

technical field [0001] The invention relates to the technical field of pre-feeding production of printed circuit boards, and more specifically refers to an automatic feeding method and system for multi-variety and small-batch PCB boards. Background technique [0002] For the production of multi-variety and small batches of PCB boards, feeding is required. The existing feeding technology is to manually query the order information, according to the feeding experience, and refer to the production history records, and manually calculate the feeding quantity. The process is as follows: Manually query the ORACLE production system ;Export sales order information; determine whether it is a new order; if it is a new order, it is generally difficult to refer to the historical pre-injection quantity of other similar products; it is difficult to ask the process supervisor to determine the quantity of input; it is more difficult to find out the historical records of similar products , an...

AI Technical Summary

Problems solved by technology

[0003] However, human labor has different feeding experience and different ability to analyze and understand the difficulty information of each order, resulting in low stability of feeding effect and large difference in accuracy; moreover, there is ambiguity in product difficulty classification, and the difficulty level of products cannot be quantified. The judgment results of different feeders for the same order are different, and the volatility is not conducive to the retrospective reference of historical data; the manual calculation of the order pre-feeding quantity one by one is inefficient; the process of manually calculating and checking order information, remainder quantity, and online order quantity is time-consuming, resulting in The manpower and time investment is large, and the feeding cost is high; the above processing process is complex and lengthy, and the comprehensive feeding effect is that the feeding is highly subjective, fluctuating, low in efficiency, and there is room for improvement in accuracy

Images