LED circuit board whole-process manufacturing optimization method based on Internet of Things
By constructing a manufacturing process topology map and deploying RFID tags, we have achieved full-process data collection and closed-loop management of LED circuit boards, solving the problem of the separation between production status and quality data, and improving product quality consistency and production efficiency.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- JIANGXI WELGAO ELECTRONICS CO LTD
- Filing Date
- 2026-02-04
- Publication Date
- 2026-05-12
AI Technical Summary
The lack of real-time data perception and product identification tracking throughout the LED circuit board manufacturing process leads to a disconnect between production status and quality data, making it difficult to conduct effective correlation analysis and affecting product yield, consistency, and production efficiency.
Construct a manufacturing process topology map, set up data collection points and deploy RFID tags, and monitor the status of workstations in real time through a process parameter analysis model to achieve full-process data collection and closed-loop management.
It enables accurate identification of key process steps and graded judgment of operating status, improving the production line's ability to provide early warning of anomalies and the consistency of product quality.
Smart Images
Figure CN122022178A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of LED circuit board manufacturing technology, specifically an optimization method for the entire manufacturing process of LED circuit boards based on the Internet of Things. Background Technology
[0002] In the LED circuit board manufacturing industry, achieving refined control and quality traceability of the production process is crucial. Currently, LED circuit board manufacturing processes largely rely on batch-based paper documents or simple barcodes for material and process management, resulting in isolated data between processes. Process parameters are often set based on experience, making it difficult to dynamically adapt to fluctuations in incoming materials and changes in equipment status. At the same time, key quality depends on final inspection, leading to delayed defect detection and high rework rates. Although some companies have introduced local monitoring sensors, the metal shielding effect of the aluminum substrate and the challenges of high temperatures and corrosiveness in the production environment still make it difficult to achieve real-time data perception and product identification tracking throughout the entire process. These factors cause production status and quality data to be fragmented, forming "information silos" that make it difficult to conduct effective correlation analysis to drive process improvement. Overall, the lack of a unified data foundation and adaptive optimization mechanism based on the Internet of Things severely restricts further improvements in product yield, consistency, and production efficiency.
[0003] How to achieve data-driven closed-loop optimization and precise traceability of the entire process of LED circuit boards from raw materials to finished products is a problem we need to solve. To this end, we present an IoT-based method for optimizing the entire manufacturing process of LED circuit boards. Summary of the Invention
[0004] The purpose of this invention is to provide an optimized manufacturing method for LED circuit boards based on the Internet of Things.
[0005] The objective of this invention can be achieved through the following technical solution: an optimized manufacturing process for LED circuit boards based on the Internet of Things, comprising the following steps:
[0006] Step S1: Construct a manufacturing process topology diagram based on the LED circuit board process flow, obtain the importance level of each workstation corresponding to each node in the manufacturing process topology diagram, and determine the key workstations based on the importance level.
[0007] Step S2: Set up data collection points, obtain the corresponding standard parameter thresholds according to the process monitoring parameter set for each workstation, deploy RFID tags on the carrier board of the LED circuit board, and trigger data collection through RFID tags when the carrier board flows through the data collection points of each workstation.
[0008] Step S3: Determine the current operating status of the workstation based on the collected process parameter data and standard parameter thresholds.
[0009] Furthermore, the process of constructing a manufacturing process topology map based on the LED circuit board manufacturing process flow, obtaining the importance level of each workstation corresponding to each node in the manufacturing process topology map, and determining the critical workstations based on the importance level includes:
[0010] Construct an equipment simulation model that reflects the relationship of the manufacturing process flow, obtain the process flow sequence relationship between several workstations in the target LED circuit board manufacturing production line, take each workstation as a node in the equipment simulation model, take the process flow sequence relationship between several workstations as the connection relationship between nodes, input the equipment simulation model, and obtain the manufacturing process topology diagram.
[0011] Based on the manufacturing process topology diagram, the process structure characteristics of several nodes are obtained, and the process hub coefficient of each node in the manufacturing process topology diagram is also obtained. The process structure characteristics and process hub coefficient of each node are used as evaluation indicators to obtain the importance level of each node. The importance level of each node is compared with a preset importance level threshold. If the importance level of a node is greater than or equal to the preset importance level threshold, the corresponding workstation is determined to be a critical workstation. If the importance level of a node is less than the preset importance level threshold, the corresponding workstation is determined to be a non-critical workstation. Data collection points are deployed at the location of each workstation.
[0012] Furthermore, the process of setting up data collection points and obtaining the corresponding standard parameter thresholds based on the process monitoring parameter set for each workstation includes:
[0013] Based on the workstation type corresponding to each node in the manufacturing process topology diagram, a corresponding set of process monitoring parameters is configured for each workstation, and a standard parameter threshold is set for each process parameter in the set of process monitoring parameters to characterize the workstation under normal operating conditions.
[0014] Furthermore, the process of obtaining the process hub coefficients of each node in the manufacturing process topology graph includes:
[0015] Collect sample data, which is a historical full-process data set of multiple completed carrier boards in the target manufacturing line. The historical full-process data set includes historical process parameter data of each carrier board when it flows through each workstation.
[0016] The collected sample data is divided into training and test sets;
[0017] The constructed graph neural network recognition model is trained using the training set to obtain training results. The training results are then verified using the test set. If the accuracy of the training results meets expectations, the training of the graph neural network recognition model is completed. If the training results do not meet expectations, the training is repeated until the training results meet expectations or the number of training iterations reaches the preset number, thereby completing the training of the graph neural network recognition model and obtaining the process parameter analysis model.
[0018] The obtained set of process monitoring parameters is input into the trained process parameter analysis model, and the process hub coefficient corresponding to the set of process monitoring parameters is output by the process parameter analysis model.
[0019] Furthermore, RFID tags are deployed on the carrier board of the LED circuit board. The data collection process triggered by the RFID tags when the carrier board passes through the data collection points at each workstation includes:
[0020] Upon receiving an LED circuit board manufacturing task, a digital work order is generated, and a unique identifier ID is assigned to each digital work order.
[0021] Obtain the substrate material to be processed, cut it and load it onto the carrier plate, heat press the RFID tag to the non-functional area on the back of each carrier plate, and write the identifier ID into the RFID tag storage area;
[0022] During the LED circuit board manufacturing process, when the carrier board enters any workstation, the RFID reader configured at the workstation automatically reads the identifier ID stored in the RFID tag on the carrier board.
[0023] The current carrier board is identified based on the read identifier ID, and the data acquisition point of this workstation is triggered to start the acquisition of process parameters. The acquisition results are then bound to the identifier ID.
[0024] Furthermore, the process of determining the current operating status of the workstation based on the collected process parameter data and standard parameter thresholds includes:
[0025] When a carrier board with an RFID tag enters any workstation, the workstation reads the RFID tag and triggers the data acquisition point of that workstation to collect data and obtain the corresponding process parameter data of that workstation.
[0026] Based on the importance level of the workstation, different operational status determination strategies are adopted for different workstations;
[0027] For non-critical workstations, the collected process parameter data of each process parameter are directly compared with their corresponding standard parameter thresholds. If all process parameters are within their respective standard parameter threshold ranges, the non-critical workstation is determined to be operating normally. If any process parameter exceeds its standard parameter threshold range, the non-critical workstation is determined to be operating abnormally.
[0028] Furthermore, for critical workstations, the standardized deviation of each process parameter is obtained based on the standard parameter threshold and the actual measured value of the process parameter.
[0029] Based on the standardized deviation, the overall process deviation of the current key workstation is obtained;
[0030] The obtained comprehensive process deviation of the current key workstation is compared with the preset process deviation threshold:
[0031] If the overall process deviation threshold is less than or equal to the preset process deviation threshold, the critical workstation is determined to be operating normally.
[0032] If the overall process deviation threshold is greater than the preset process deviation threshold, the critical workstation is determined to be operating abnormally.
[0033] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention constructs a manufacturing process topology map based on the LED circuit board manufacturing process, combines process structural characteristics and process hub coefficients to quantitatively assess the importance of each workstation, and determines key workstations based on their importance levels. Through joint assessment of structural characteristics and data-driven process hub coefficients, objective identification of key process links is achieved, avoiding misjudgments or omissions caused by relying on human experience. Data collection points are set up, and corresponding standard parameter thresholds are obtained based on the process monitoring parameter set, achieving standardized and comprehensive collection of process parameters throughout the entire process, providing comprehensive data support for subsequent workstation status analysis. RFID tags are deployed on the carrier board, and data collection is triggered in real time through RFID tags. The current operating status of the workstation is determined based on the collected process parameter data and standard parameter thresholds, and the data is input into the RFID tags. Using the carrier board as a mobile data carrier, the accuracy of data association and the real-time response of the system are significantly improved. This method achieves accurate identification of key process links in the LED circuit board manufacturing process, hierarchical intelligent discrimination of operating status, and closed-loop management of the entire process data, significantly improving the production line's anomaly early warning capability and product quality consistency. Attached Figure Description
[0034] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.
[0035] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation
[0036] like Figure 1 As shown, an IoT-based method for optimizing the entire manufacturing process of LED circuit boards includes the following steps:
[0037] Step S1: Construct a manufacturing process topology diagram based on the LED circuit board process flow, obtain the importance level of each workstation corresponding to each node in the manufacturing process topology diagram, and determine the key workstations based on the importance level.
[0038] Step S2: Set up data collection points, obtain the corresponding standard parameter thresholds according to the process monitoring parameter set for each workstation, deploy RFID tags on the carrier board of the LED circuit board, and trigger data collection through RFID tags when the carrier board flows through the data collection points of each workstation.
[0039] Step S3: Determine the current operating status of the workstation based on the collected process parameter data and standard parameter thresholds.
[0040] It should be further explained that, in the specific implementation process, the process of constructing a manufacturing process topology diagram based on the LED circuit board manufacturing process flow, obtaining the importance level of each workstation corresponding to each node in the manufacturing process topology diagram, and determining the critical workstations based on the importance level includes:
[0041] Construct an equipment simulation model that reflects the relationship of the manufacturing process flow, obtain the process flow sequence relationship between several workstations in the target LED circuit board manufacturing production line, take each workstation as a node in the equipment simulation model, take the process flow sequence relationship between several workstations as the connection relationship between nodes, input the equipment simulation model, and obtain the manufacturing process topology diagram.
[0042] Based on the manufacturing process topology diagram, the process structure characteristics of several nodes are obtained. The process structure characteristics include the number of key workstations connected downstream of the nodes, process irreversibility, etc. At the same time, the process hub coefficient of each node in the manufacturing process topology diagram is obtained. The process structure characteristics and process hub coefficient of each node are used as evaluation indicators. The indicator weights of the evaluation indicators are set. The membership matrix of each node to the preset importance level is obtained through fuzzy comprehensive evaluation.
[0043] The importance level of each node is obtained based on the membership matrix and indicator weights. The importance level of each node is compared with the preset importance level threshold. If the importance level of a node is greater than or equal to the preset importance level threshold, the corresponding workstation is determined to be a critical workstation. If the importance level of a node is less than the preset importance level threshold, the corresponding workstation is determined to be a non-critical workstation. Data collection points are deployed at the location of each workstation.
[0044] The process of obtaining the importance level of each node based on the membership matrix and indicator weights includes:
[0045] The evaluation index weights and membership matrix of the evaluation index are fused by formula to obtain the fuzzy comprehensive evaluation matrix of the evaluation index. The membership degree of each node to different importance levels is obtained according to the fuzzy comprehensive evaluation matrix. The importance level with the highest membership degree for each node is selected and the importance level with the highest membership degree for each node is taken as the importance level of each node.
[0046] The formula is as follows:
[0047] ;
[0048] in, The fuzzy comprehensive evaluation matrix of the evaluation indicators is... The weights of the evaluation indicators are: For the membership matrix, " "" indicates that the elements at corresponding positions in the weight matrix and membership matrix of the evaluation index are multiplied together. The weighting parameter is used to control the balance between the weight matrix and the membership matrix in the fuzzy comprehensive evaluation matrix of the evaluation index.
[0049] It should be further explained that, in the specific implementation process, the process of setting up data collection points and obtaining the corresponding standard parameter thresholds based on the process monitoring parameter set for each workstation includes:
[0050] Based on the workstation type corresponding to each node in the manufacturing process topology diagram, configure a corresponding set of process monitoring parameters for each workstation, and set a standard parameter threshold for each process parameter in the set of process monitoring parameters to characterize the workstation under normal operating conditions.
[0051] For example, the process monitoring parameters for the copper plating station include the average thickness and thickness uniformity of the copper plating layer; the process monitoring parameters for the electroplating station include the corresponding electroplating time, solution temperature, and copper ion concentration; the process monitoring parameters for the etching station include the etching solution temperature and copper ion concentration; and the process monitoring parameters for the cleaning station include the cleaning solution temperature and residual ion concentration. The data acquisition point collects relevant process parameter data in real time according to the process monitoring parameter set, marks the acquisition timestamp, and sets the acquisition cycle.
[0052] It should be further explained that, in the specific implementation process, the process of obtaining the process hub coefficients of each node in the manufacturing process topology diagram includes:
[0053] Collect sample data, which is a historical full-process data set of multiple completed carrier boards in the target manufacturing line. The historical full-process data set includes historical process parameter data of each carrier board when it flows through each workstation.
[0054] The collected sample data is divided into training and test sets;
[0055] The constructed graph neural network recognition model is trained using the training set to obtain training results. The training results are then verified using the test set. If the accuracy of the training results meets expectations, the training of the graph neural network recognition model is completed. If the training results do not meet expectations, the training is repeated until the training results meet expectations or the number of training iterations reaches the preset number, thereby completing the training of the graph neural network recognition model and obtaining the process parameter analysis model.
[0056] The obtained set of process monitoring parameters is input into the trained process parameter analysis model, and the process hub coefficient corresponding to the set of process monitoring parameters is output by the process parameter analysis model.
[0057] It should be further explained that, in the specific implementation process, RFID tags are deployed on the carrier board of the LED circuit board. The process of triggering data collection through the RFID tags when the carrier board flows through the data collection points of each workstation includes:
[0058] Upon receiving an LED circuit board manufacturing task, a digital work order is generated. Each digital work order is assigned a unique identifier ID, which is composed of the production date and production line code and is used to represent the current production batch.
[0059] The production line code uses a combination of letters and numbers to identify the target manufacturing production line executing the corresponding digital work order;
[0060] Obtain the substrate material to be processed, cut it and load it onto the carrier plate, heat press RFID tags onto the non-functional area on the back of each carrier plate, and write the identifier ID into the storage area of the RFID tag, thereby realizing the identification of the carrier plate and the multiple independent small boards it carries.
[0061] During the LED circuit board manufacturing process, when the carrier board enters any workstation, the RFID reader configured at the workstation automatically reads the identifier ID stored in the RFID tag on the carrier board.
[0062] The current carrier board is identified based on the read identifier ID, and the data acquisition point of this workstation is triggered to start the acquisition of process parameters. The acquisition results are then bound to the identifier ID.
[0063] The RFID tags deployed on the carrier board serve as data carriers, used to associate and record the process parameter data of the LED circuit board at each workstation and the operating status of the workstation.
[0064] It should be further explained that, in the specific implementation process, the process of determining the current operating status of the workstation based on the collected process parameter data and standard parameter thresholds includes:
[0065] When a carrier board with an RFID tag enters any workstation, the workstation reads the RFID tag and triggers the data acquisition point of that workstation to collect data and obtain the corresponding process parameter data of that workstation.
[0066] Based on the importance level of the workstation, different operational status determination strategies are adopted for different workstations;
[0067] For non-critical workstations, the process parameter data of each collected process parameter is directly compared with its corresponding standard parameter threshold. If all process parameters are within their respective standard parameter threshold ranges, the non-critical workstation is determined to be operating normally. If any process parameter exceeds its standard parameter threshold range, the non-critical workstation is determined to be operating abnormally.
[0068] For critical workstations, each process parameter that needs to be monitored is labeled and denoted as j, where j = 1, 2, ..., m;
[0069] The standardized deviation of each process parameter is obtained by comparing the standard parameter threshold with the actual measured value of the process parameter, and is denoted as . ,Right now:
[0070] ;
[0071] in, This represents the standardized deviation of the j-th process parameter. This represents the actual measured value of the j-th process parameter. This represents the historical average value of the j-th process parameter. This represents the historical standard deviation of the j-th process parameter;
[0072] Based on the standardized deviation, the overall process deviation of the current critical workstation is obtained, denoted as . ,Right now:
[0073] ;
[0074] in, This indicates the overall process deviation of the current key workstation. This indicates the number of process parameters for this critical workstation;
[0075] The obtained comprehensive process deviation of the current key workstation is compared with the preset process deviation threshold:
[0076] like The critical workstation was determined to be operating normally.
[0077] like The critical workstation was determined to be malfunctioning.
[0078] in, This is the preset process deviation threshold;
[0079] For example:
[0080] After the copper plating process is completed on the substrate, the data acquisition process of the copper plating station is triggered. A thickness measuring instrument deployed at this data acquisition point performs multi-point sampling. Based on the data from all sampling points, the average thickness and thickness uniformity of the overall copper plating layer on the substrate are calculated, denoted as follows: , ,Right now:
[0081] ;
[0082] in, This represents the average thickness of the copper plating layer, where N is the total number of sampling points. Let be the thickness value at the i-th sampling point. The standard deviation of the copper plating layer thickness;
[0083] If the copper plating station is a critical station, it includes and Two process parameters;
[0084] Calculate the standardized deviation of each process parameter, denoted as . , ,Right now:
[0085] , ;
[0086] in, The standardized deviation of the average thickness. This represents the actual measured value of the collected average thickness. This represents the historical average thickness. This represents the historical standard deviation of the average thickness. The standardized deviation representing thickness uniformity The actual measured value representing the uniformity of the collected thickness. This represents the historical average value of thickness uniformity. Historical standard deviation representing thickness uniformity;
[0087] Based on the standardized deviation, the overall process deviation of the current copper plating station is obtained, denoted as D, i.e.:
[0088] ;
[0089] The calculated overall process deviation is compared with the preset process deviation threshold:
[0090] like The copper plating station was determined to be operating normally.
[0091] like The copper plating station was determined to be malfunctioning.
[0092] After determining the current workstation's operating status, the identifier ID, the workstation's operating status determination result, the timestamp, and whether it is a critical workstation (or, if it is, the corresponding comprehensive process deviation value) are encapsulated. The data is recorded in the format of "Identifier ID - Timestamp - Workstation Operating Status - Whether it is a Critical Workstation - Comprehensive Process Deviation" to obtain a standard formatted dataset. The dataset is then sorted in ascending order according to the collection timestamp and input into the RFID tag.
[0093] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any modifications or equivalent substitutions made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.
Claims
1. A method for optimizing the entire manufacturing process of LED circuit boards based on the Internet of Things, characterized in that, Includes the following steps: Step S1: Construct a manufacturing process topology diagram based on the LED circuit board process flow, obtain the importance level of each workstation corresponding to each node in the manufacturing process topology diagram, and determine the key workstations based on the importance level. Step S2: Set up data collection points, obtain the corresponding standard parameter thresholds according to the process monitoring parameter set for each workstation, deploy RFID tags on the carrier board of the LED circuit board, and trigger data collection through RFID tags when the carrier board flows through the data collection points of each workstation. Step S3: Determine the current operating status of the workstation based on the collected process parameter data and standard parameter thresholds.
2. The method for optimizing the entire manufacturing process of LED circuit boards based on the Internet of Things according to claim 1, characterized in that, The process of constructing a manufacturing process topology diagram based on the LED circuit board manufacturing process flow, obtaining the importance level of each workstation corresponding to each node in the manufacturing process topology diagram, and determining the critical workstations based on the importance level includes: Construct an equipment simulation model that reflects the relationship of the manufacturing process flow, obtain the process flow sequence relationship between several workstations in the target LED circuit board manufacturing production line, take each workstation as a node in the equipment simulation model, take the process flow sequence relationship between several workstations as the connection relationship between nodes, input the equipment simulation model, and obtain the manufacturing process topology diagram. Based on the manufacturing process topology diagram, the process structure characteristics of several nodes are obtained, and the process hub coefficient of each node in the manufacturing process topology diagram is also obtained. The process structure characteristics and process hub coefficient of each node are used as evaluation indicators to obtain the importance level of each node. The importance level of each node is compared with a preset importance level threshold. If the importance level of a node is greater than or equal to the preset importance level threshold, the corresponding workstation is determined to be a critical workstation. If the importance level of a node is less than the preset importance level threshold, the corresponding workstation is determined to be a non-critical workstation. Data collection points are deployed at the location of each workstation.
3. The method for optimizing the entire manufacturing process of LED circuit boards based on the Internet of Things according to claim 2, characterized in that, The process of setting up data collection points and obtaining the corresponding standard parameter thresholds based on the process monitoring parameter set for each workstation includes: Based on the workstation type corresponding to each node in the manufacturing process topology diagram, a corresponding set of process monitoring parameters is configured for each workstation, and a standard parameter threshold is set for each process parameter in the set of process monitoring parameters to characterize the workstation under normal operating conditions.
4. The method for optimizing the entire manufacturing process of LED circuit boards based on the Internet of Things according to claim 2, characterized in that, The process of obtaining the process hub coefficients of each node in the manufacturing process topology graph includes: Collect sample data, which is a historical full-process data set of multiple completed carrier boards in the target manufacturing line. The historical full-process data set includes historical process parameter data of each carrier board when it flows through each workstation. The collected sample data is divided into training and test sets; The constructed graph neural network recognition model is trained using the training set to obtain training results. The training results are then verified using the test set. If the accuracy of the training results meets expectations, the training of the graph neural network recognition model is completed. If the training results do not meet expectations, the training is repeated until the training results meet expectations or the number of training iterations reaches the preset number, thereby completing the training of the graph neural network recognition model and obtaining the process parameter analysis model. The obtained set of process monitoring parameters is input into the trained process parameter analysis model, and the process hub coefficient corresponding to the set of process monitoring parameters is output by the process parameter analysis model.
5. The method for optimizing the entire manufacturing process of LED circuit boards based on the Internet of Things according to claim 4, characterized in that, The process of deploying RFID tags on the carrier board of the LED circuit board, and triggering data collection via RFID tags when the carrier board flows through the data collection points of each workstation, includes: Upon receiving an LED circuit board manufacturing task, a digital work order is generated, and a unique identifier ID is assigned to each digital work order. Obtain the substrate material to be processed, cut it and load it onto the carrier plate, heat press the RFID tag to the non-functional area on the back of each carrier plate, and write the identifier ID into the RFID tag storage area; During the LED circuit board manufacturing process, when the carrier board enters any workstation, the RFID reader configured at the workstation automatically reads the identifier ID stored in the RFID tag on the carrier board. The current carrier board is identified based on the read identifier ID, and the data acquisition point of this workstation is triggered to start the acquisition of process parameters. The acquisition results are then bound to the identifier ID.
6. The method for optimizing the entire manufacturing process of LED circuit boards based on the Internet of Things according to claim 5, characterized in that, The process of determining the current operating status of the workstation based on the collected process parameter data and standard parameter thresholds includes: When a carrier board with an RFID tag enters any workstation, the workstation reads the RFID tag and triggers the data acquisition point of that workstation to collect data and obtain the corresponding process parameter data of that workstation. Based on the importance level of the workstation, different operational status determination strategies are adopted for different workstations; For non-critical workstations, the collected process parameter data of each process parameter are directly compared with their corresponding standard parameter thresholds. If all process parameters are within their respective standard parameter threshold ranges, the non-critical workstation is determined to be operating normally. If any process parameter exceeds its standard parameter threshold range, the non-critical workstation is determined to be operating abnormally.
7. The method for optimizing the entire manufacturing process of LED circuit boards based on the Internet of Things according to claim 6, characterized in that, For critical workstations, the standardized deviation of each process parameter is obtained based on the standard parameter threshold and the actual measured value of the process parameter. Based on the standardized deviation, the overall process deviation of the current key workstation is obtained; The obtained comprehensive process deviation of the current key workstation is compared with the preset process deviation threshold: If the overall process deviation threshold is less than or equal to the preset process deviation threshold, the critical workstation is determined to be operating normally. If the overall process deviation threshold is greater than the preset process deviation threshold, the critical workstation is determined to be operating abnormally.