Barcode abnormity detection method based on artificial intelligence

Abstract

The invention discloses a barcode abnormity detection method based on artificial intelligence without pre-configuration. The method comprises the following steps: (1) when label data are less, judgingwhether a bar code obtained by code scanning is abnormal by adopting a basic statistical rule, and adding the bar code conforming to the basic statistical rule into a normal label data group and adding the bar code not conforming to the basic statistical rule into an abnormal label data group; 2) when the label data reaches an initial set value, predicting whether the bar code obtained by code scanning is abnormal by adopting an abnormality detection algorithm, and manually confirming whether the bar code is added into an abnormal label data set or a normal label data set; and (3) when the label data reaches a certain quantity, judging whether the input of the bar code is abnormal or not by adopting a classification prediction algorithm for the input bar code, performing manual confirmation, adding an abnormal label data set if the detected suspected abnormal bar code is determined to be abnormal, and otherwise, adding a normal label data set. The method can be used for abnormal judgment of various bar codes.

Description

technical field [0001] The invention relates to the field of battery production and manufacturing, in particular to the field of lithium battery production and manufacturing, aiming at a method for judging the abnormality of the battery barcode in the scanning link of the composition and capacity production process. Background technique [0002] At present, in the battery production and manufacturing process, various battery manufacturers need to scan the battery barcode in order to trace the manufacturing process and control the quality in many manufacturing links. After the production process is completed, upload the relevant data together with the battery barcode to various systems, these systems include BMIS / MES / ERP / MRP, etc. Once these data are uploaded to different systems, it will be very difficult to change if the barcode is found to be abnormal . Therefore, it is very important to judge whether the input of the battery barcode (barcode for short) is correct or not ...

AI Technical Summary

Problems solved by technology

At present, it is generally adopted to set some rules to verify whether the input of the battery barcode is abnormal. These rules include judging whether the length of the barcode is a fixed length, or judging the beginning and end of the barcode, which is more complicated such as setting a regular expression, but, Due to the independent operation of production equipment and the changeable barcodes due to different products, this method was abandoned in practice because it was too complicated

Images