CN-122022178-A - LED circuit board full-flow manufacturing optimization method based on Internet of things

CN122022178ACN 122022178 ACN122022178 ACN 122022178ACN-122022178-A

Abstract

The invention discloses a full-flow manufacturing optimization method of an LED circuit board based on the Internet of things, which relates to the technical field of LED circuit board manufacturing, and the method comprises the steps of constructing a manufacturing process topological graph according to the process flow of the LED circuit board, quantitatively evaluating importance of each station by combining process structural features and process pivot coefficients, and judging key stations according to importance grades; the method realizes the accurate identification of key process links in the manufacturing process of the LED circuit board, the hierarchical intelligent discrimination of the running state and the closed-loop management of full-flow data, and remarkably improves the abnormal early warning capability of the production line and the consistency of the product quality.

Inventors

CHEN XING
HE YONGQING

Assignees

江西威尔高电子股份有限公司

Dates

Publication Date: 20260512
Application Date: 20260204

Claims (7)

1. The full-flow manufacturing optimization method for the LED circuit board based on the Internet of things is characterized by comprising the following steps of: step S1, constructing a manufacturing process topological graph according to the process flow of the LED circuit board, acquiring importance levels of work stations corresponding to all nodes in the manufacturing process topological graph, and judging key work stations according to the importance levels; step S2, setting data acquisition points, acquiring corresponding standard parameter thresholds according to process monitoring parameter sets corresponding to each work station, deploying RFID labels on a carrier plate of the LED circuit board, and triggering data acquisition through the RFID labels when the carrier plate flows through the data acquisition points of each work station; And step S3, judging the running state of the current work station according to the acquired process parameter data and the standard parameter threshold value.
2. The method for optimizing the full-process manufacturing of the LED circuit board based on the Internet of things according to claim 1, wherein the process for constructing a manufacturing process topological graph according to the process flow of the LED circuit board, obtaining importance levels of work stations corresponding to all nodes in the manufacturing process topological graph, and judging key work stations according to the importance levels comprises the following steps: Constructing an equipment simulation model reflecting the relation of manufacturing process flows, acquiring the sequence relation of process flows among a plurality of work stations of an LED circuit board target manufacturing production line, taking each work station as a node of the equipment simulation model, taking the sequence relation of the process flows among the plurality of work stations as a connection relation among the nodes, and inputting the equipment simulation model to obtain a manufacturing process topological graph; And acquiring process structural features of a plurality of nodes according to the manufacturing process topological graph, acquiring process pivot coefficients of all nodes in the manufacturing process topological graph, taking the process structural features and the process pivot coefficients of all nodes as evaluation indexes, acquiring importance levels of all nodes, comparing the importance levels of all nodes with a preset importance level threshold, judging a corresponding workstation as a key workstation if the importance level of the node is greater than or equal to the preset importance level threshold, judging the corresponding workstation as a non-key workstation if the importance level of the node is less than the preset importance level threshold, and deploying a data acquisition point at the position of each workstation.
3. The method for optimizing full-flow manufacturing of the LED circuit board based on the Internet of things according to claim 2, wherein the step of setting a data acquisition point and acquiring the corresponding standard parameter threshold according to the process monitoring parameter set corresponding to each workstation comprises the following steps: According to the type of the work station corresponding to each node in the manufacturing process topological graph, configuring a corresponding process monitoring parameter set for each work station, and setting a standard parameter threshold value for representing the work station in a normal running state for each process parameter in the process monitoring parameter set.
4. The method for optimizing full-flow manufacturing of the LED circuit board based on the Internet of things according to claim 2, wherein the process of obtaining the process pivot coefficients of each node in the manufacturing process topological graph comprises the following steps: Collecting sample data, wherein the sample data are historical full-flow process data sets of a plurality of finished carrier plates of a target manufacturing production line, and the historical full-flow process data sets comprise historical process parameter data when each carrier plate flows through each work station; dividing the collected sample data into a training set and a testing set; Training the constructed graphic neural network recognition model by using a training set to obtain a training result, verifying the obtained training result by using a testing set, if the accuracy of the training result meets the expectation, completing the training of the graphic neural network recognition model, and if the training result does not meet the expectation, retraining until the training result meets the expectation or the training times reaches the preset times, thereby completing the training of the graphic neural network recognition model and obtaining a process parameter analysis model; inputting the obtained process monitoring parameter set into a process parameter analysis model which is trained, and outputting a process pivot coefficient corresponding to the process monitoring parameter set through the process parameter analysis model.
5. The method for optimizing full-process manufacturing of the LED circuit board based on the Internet of things according to claim 4, wherein the process of deploying the RFID tag on the carrier board of the LED circuit board and triggering data acquisition through the RFID tag when the carrier board flows through the data acquisition points of all the work stations comprises the following steps: After receiving the LED circuit board manufacturing task, generating digital chemical engineering sheets, and distributing unique corresponding identifiers ID for each digital chemical engineering sheet; Obtaining a substrate plate to be processed, cutting and loading the substrate plate on a carrier plate, fixing an RFID tag on a nonfunctional area on the back of each carrier plate in a hot pressing mode, and writing an identifier ID into an RFID tag storage area; in the manufacturing process of the LED circuit board, when the carrier board enters any station, an RFID reader-writer configured by the station automatically reads an identifier ID stored in an RFID tag on the carrier board; And identifying the current carrier plate based on the read identifier ID, triggering a data acquisition point of the work station to start process parameter acquisition, and binding an acquisition result with the identifier ID.
6. The method for optimizing the full-process manufacturing of the LED circuit board based on the Internet of things according to claim 5, wherein the process of judging the running state of the current workstation according to the acquired process parameter data and the standard parameter threshold comprises the following steps: When the carrier plate with the RFID tag enters any station, the station reads the RFID tag and triggers a data acquisition point of the station to acquire data, so as to obtain process parameter data corresponding to the station; According to importance levels of the work stations, different work stations are subjected to differentiated running state judging strategies; For a non-critical station, directly comparing the acquired process parameter data of each process parameter with the corresponding standard parameter threshold value, if all the process parameters are within the respective standard parameter threshold value range, judging that the non-critical station operates normally, and if any process parameter exceeds the standard parameter threshold value range, judging that the non-critical station operates abnormally.
7. The internet of things-based full-flow manufacturing optimization method for the LED circuit board, which is characterized in that for a key station, the standard deviation of each process parameter is obtained according to the standard parameter threshold value of the process parameter and the actual measured value of the process parameter; obtaining the comprehensive process deviation degree of the current key station according to the standard deviation; comparing the obtained comprehensive process deviation degree of the current key station with a preset process deviation degree threshold value: if the comprehensive process deviation threshold value is smaller than or equal to a preset process deviation threshold value, judging that the key station operates normally; And if the comprehensive process deviation threshold is greater than the preset process deviation threshold, judging that the operation of the key station is abnormal.

Description

LED circuit board full-flow manufacturing optimization method based on Internet of things Technical Field The invention relates to the technical field of LED circuit board manufacturing, in particular to a full-flow manufacturing optimization method of an LED circuit board based on the Internet of things. Background In the field of LED circuit board manufacturing, the realization of fine management and quality tracing of a production process is very important, the current LED circuit board manufacturing process is mainly based on batch-based paper documents or simple bar codes for material and process management, so that data among various processes are isolated, technological parameters are often empirically set and are difficult to dynamically adapt to incoming material fluctuation and equipment state change, meanwhile, the key quality depends on final detection, defects are delayed in discovery, the reworking rate is high, although partial enterprises introduce local monitoring sensors, the realization of whole process real-time data perception and product identity tracing still has difficulty due to the metal shielding effect of an aluminum substrate and high-temperature and corrosiveness challenges in a production environment, the production state and quality data are mutually split due to the factors, so that 'information island' is difficult to carry out effective correlation analysis to drive process improvement, and in general, a unified data base and self-adaptive optimization mechanism based on the Internet of things are lacked, so that the further improvement of the product yield, the consistency and the production energy efficiency are seriously restricted; How to realize the whole process from raw materials to finished products of an LED circuit board, and the closed-loop optimization and the accurate traceability of data driving are problems to be solved, and therefore, the whole process manufacturing optimization method of the LED circuit board based on the Internet of things is provided. Disclosure of Invention The invention aims to provide an LED circuit board full-flow manufacturing optimization method based on the Internet of things. The invention aims at realizing the technical scheme that the full-flow manufacturing optimization method of the LED circuit board based on the Internet of things comprises the following steps of: step S1, constructing a manufacturing process topological graph according to the process flow of the LED circuit board, acquiring importance levels of work stations corresponding to all nodes in the manufacturing process topological graph, and judging key work stations according to the importance levels; step S2, setting data acquisition points, acquiring corresponding standard parameter thresholds according to process monitoring parameter sets corresponding to each work station, deploying RFID labels on a carrier plate of the LED circuit board, and triggering data acquisition through the RFID labels when the carrier plate flows through the data acquisition points of each work station; And step S3, judging the running state of the current work station according to the acquired process parameter data and the standard parameter threshold value. Further, the process of constructing a manufacturing process topological graph according to the process flow of the LED circuit board, obtaining importance levels of work stations corresponding to each node in the manufacturing process topological graph, and judging key work stations according to the importance levels comprises the following steps: Constructing an equipment simulation model reflecting the relation of manufacturing process flows, acquiring the sequence relation of process flows among a plurality of work stations of an LED circuit board target manufacturing production line, taking each work station as a node of the equipment simulation model, taking the sequence relation of the process flows among the plurality of work stations as a connection relation among the nodes, and inputting the equipment simulation model to obtain a manufacturing process topological graph; And acquiring process structural features of a plurality of nodes according to the manufacturing process topological graph, acquiring process pivot coefficients of all nodes in the manufacturing process topological graph, taking the process structural features and the process pivot coefficients of all nodes as evaluation indexes, acquiring importance levels of all nodes, comparing the importance levels of all nodes with a preset importance level threshold, judging a corresponding workstation as a key workstation if the importance level of the node is greater than or equal to the preset importance level threshold, judging the corresponding workstation as a non-key workstation if the importance level of the node is less than the preset importance level threshold, and deploying a data acquisition point at the position of each workstation. Further, the pr