Automated production line control method and system based on high-quality datasets

Abstract

The application relates to an automatic production line control method and system based on a high-quality data set. The method comprises the following steps: acquiring industrial production line data, target parameters of products to be produced and raw material parameters; pre-processing the industrial production line data and data grouping to obtain a high-quality data set; the data grouping comprises grouping according to process parameters of each processing procedure, forward parameters before entering the processing procedure and backward parameters output by the processing procedure; each processing procedure of a target production line is taken as a target procedure, and the target parameters are disassembled to obtain target sub-parameters corresponding to each target procedure; the high-quality data set is searched according to the target procedure and the target sub-parameters to obtain forward parameters and process parameters corresponding to the target procedure; and the process parameters are called to guide the operation of the corresponding target procedure, so that the problem that effective data required by the production line is difficult to be fully extracted and causes the low utilization efficiency of industrial big data of existing enterprises can be effectively solved.

Description

Technical Field

[0001] This application relates to the field of production assembly technology, and in particular to an automated production line control method and system based on high-quality datasets. Background Technology

[0002] Industrial big data is the core engine driving the intelligent transformation of industry. By aggregating multi-dimensional data such as production and equipment, it builds a "digital mirror" that is insightful, predictable, and optimizable, propelling the manufacturing industry from experience-driven to a new stage of data-driven development.

[0003] However, due to the inconsistent quality and limited applicability of the massive amounts of data collected within enterprises and industries, it is difficult to fully extract the effective data required by the production line, resulting in low efficiency in the application of industrial big data by existing enterprises. Summary of the Invention

[0004] Therefore, it is necessary to provide an automated production line control method and system based on high-quality datasets to address the aforementioned technical problems.

[0005] Firstly, this application provides an automated production line control method based on a high-quality dataset, the method comprising:

[0006] Acquire industrial production line data, target parameters of products to be manufactured, and raw material parameters;

[0007] High-quality datasets are obtained by preprocessing and grouping industrial production line data. Data grouping includes grouping based on process parameters of each processing step in the industrial production line, forward parameters of the product before entering the processing step, and backward parameters of the product output from the processing step.

[0008] Each processing step of the target production line is taken as the target step, and the target parameters are decomposed according to the target steps to obtain the target sub-parameters corresponding to each target step.

[0009] Search for high-quality datasets based on the target process and target sub-parameters to obtain the forward parameters and process parameters of the corresponding target process;

[0010] In response to the matching of the forward parameters of the target process with the raw material parameters, the matching process parameters are invoked to guide the operation of the corresponding target process.

[0011] In one embodiment, the processing steps include multiple process flows with a time-corresponding relationship; the process parameters include equipment parameters and process parameters; the step of preprocessing the industrial production line data includes:

[0012] Filter missing values ​​in industrial production line data; missing values ​​include null values ​​and placeholders.

[0013] The alternative data is obtained by removing processing steps with missing values ​​from the industrial production line data;

[0014] Parameter values ​​that meet any of the following conditions are considered outliers: the equipment parameter is greater than the corresponding equipment's operating limit, the equipment model corresponding to the equipment parameter does not match the processing procedure, and the process parameter does not match the steps of the process flow.

[0015] Remove processing steps that contain outliers from the candidate data.

[0016] In one embodiment, the method further includes:

[0017] A digital twin model is built based on the target production line, and the equipment parameters, process parameters, and raw material parameters of each target process are input into the digital twin model to obtain the predicted finished product parameters.

[0018] The finished product parameters are decomposed according to the target process to obtain finished product sub-parameters, and then compared and analyzed with the corresponding target sub-parameters.

[0019] In response to any difference between a target sub-parameter and its corresponding finished product sub-parameter exceeding a preset range, a process error warning message is output; the process error warning message is used to indicate that the process parameters of the target process corresponding to the finished product sub-parameter are unreliable.

[0020] In one embodiment, the method further includes:

[0021] In response to the output process error prompt information, the equipment parameters of the corresponding target process in the digital twin model are gradually adjusted according to the preset step size, with the adjustment direction being to shorten the parameter difference;

[0022] If the parameter difference still exceeds the preset range, output a process parameter error prompt message; otherwise, operate the corresponding equipment according to the adjusted equipment parameters.

[0023] In one embodiment, the method further includes:

[0024] In response to the output process parameter error prompt information, the process parameters corresponding to the target process are removed, and the processing processes in the high-quality dataset are searched again according to the raw material parameters and equipment parameters of the target process to obtain the correspondence between multiple sets of process parameters and backward parameters;

[0025] Establish an association relationship model, and substitute the corresponding relationships into the association relationship model for training to obtain the trained association relationship model;

[0026] Substitute the target sub-parameters corresponding to the parameter differences into the trained correlation model to obtain the process correction parameters;

[0027] Replace the rejected process parameters with process correction parameters, and re-predict the finished product parameters.

[0028] In one embodiment, the method further includes:

[0029] In response to the temporal correspondence between multiple target processes in the target production line, the corresponding processing processes are searched in a high-quality dataset.

[0030] If there exists a processing step in the forward sequence whose backward parameter is the same as the forward parameter of the adjacent backward processing step, then determine whether the forward parameter of the processing step at the beginning of the time sequence matches the raw material parameter, and whether the backward parameter of the processing step at the end of the time sequence matches the target sub-parameter.

[0031] If so, then its process parameters are invoked to guide the operation of the corresponding target process.

[0032] In one embodiment, the forward parameters of the target process are matched with the raw material parameters to characterize that the raw material parameters, after being processed a preset number of times by the target production line, conform to the forward parameters of the target process.

[0033] Secondly, this application also provides an automated production line control system based on a high-quality dataset, the system comprising:

[0034] The acquisition module is used to acquire industrial production line data, target parameters of the products to be produced, and raw material parameters.

[0035] The data processing module is used to preprocess and group the acquired industrial production line data to obtain a high-quality dataset. The data grouping includes grouping based on the process parameters of each processing step in the industrial production line, the forward parameters of the product before entering the processing step, and the backward parameters of the product output from the processing step.

[0036] The parameter decomposition module is used to take each processing step of the target production line as the target step, and decompose the target parameters according to each target step to obtain the target sub-parameters of the corresponding target step.

[0037] The search module is used to search for high-quality datasets based on the target process and the corresponding target sub-parameters to obtain the forward parameters and process parameters of the target process.

[0038] The calling module is used to respond to the matching of the forward parameters of the target process with the raw material parameters, and to call the matching process parameters to guide the operation of the corresponding target process.

[0039] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the method provided in the first aspect of this application.

[0040] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the method provided in the first aspect of this application.

[0041] The aforementioned automated production line control method and system based on high-quality datasets can acquire industrial production line data, target parameters of the products to be produced, and raw material parameters. Through preprocessing and data grouping of the industrial production line data, the data is grouped into components: process parameters of each processing step in the industrial production line, forward parameters of the products entering the processing step, and backward parameters of the products output from the processing step. This creates a closely linked and clearly defined parameter group from parameter input to parameter processing and output. The target parameters of the products to be produced are then decomposed according to each processing step of the target production line, breaking them down into target sub-parameters corresponding to each target step. A high-quality dataset is then searched using the target steps and target sub-parameters to obtain the forward and process parameters of each target step. If the forward parameters match the raw material parameters, the process parameters in the high-quality dataset are determined to be usable for the target production line, and they are invoked to guide the corresponding target steps in production. This effectively solves the problem that the effective data required by the production line is difficult to extract fully, leading to low efficiency in the use of industrial big data by existing enterprises. Attached Figure Description

[0042] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 This is a flowchart illustrating the steps involved in obtaining process parameters for a target production line based on a high-quality dataset in one embodiment.

[0044] Figure 2 This is a flowchart illustrating the steps for determining process errors in one embodiment;

[0045] Figure 3 This is a flowchart illustrating the steps for determining process errors in one embodiment;

[0046] Figure 4 This is a flowchart illustrating the steps for correcting process parameters in one embodiment.

[0047] Figure 5 This is a flowchart illustrating the steps involved in invoking an associated processing step to guide the execution of a target process, as shown in one embodiment.

[0048] Figure 6 This is a block diagram of an automated production line control system based on a high-quality dataset in one embodiment. Detailed Implementation

[0049] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0050] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0051] With the rise of automated and intelligent manufacturing, more and more industrial production lines are equipped with multimodal data acquisition devices for monitoring the entire production process, such as cameras, sensors, and production data monitoring systems for each process node. This enables precise analysis and intelligent optimization of production, quality control, and energy consumption cost calculation through industrial big data, promoting the deep transformation of the manufacturing industry towards digitalization and intelligence.

[0052] On the other hand, due to varying production precision requirements and the personalized customization by clients, companies often need to make adjustments to various processes during production. This makes it increasingly difficult to guide the operation of a modified production line simply by analyzing historical data. Many companies, when modifying production lines to manufacture new products, have to rely on engineers' experience and equipment supplier parameters for trial-and-error debugging, investing significant time and resources to re-collect data and build models. This process is not only lengthy and costly, but also leads to yield fluctuations, capacity losses, and material waste due to instability during debugging, severely restricting companies' ability to respond to market changes and achieve flexible manufacturing.

[0053] The automated production line control method based on high-quality datasets provided in this application embodiment, such as... Figure 1 As shown, the method includes the following steps S202 to S210. Wherein:

[0054] Step S202: Obtain industrial production line data, target parameters of the product to be produced, and raw material parameters.

[0055] Industrial production line data can include internal production data parameters, publicly available data parameters for various production equipment and processes within the industry, and data parameters provided to enterprises by equipment suppliers during equipment commissioning and industrial application.

[0056] Furthermore, the data parameters may include historical production data, process parameters at each processing step, and product parameters recorded by multimodal sensors.

[0057] On the other hand, raw material parameters can include physical, chemical, and performance parameters corresponding to different production and processing directions, such as type, size, and strength.

[0058] Step S204: Preprocess and group the industrial production line data to obtain a high-quality dataset; the data grouping includes grouping based on the process parameters of each processing step in the industrial production line, the forward parameters of the product before entering the processing step, and the backward parameters of the product output from the processing step.

[0059] Data preprocessing may include removing missing and outlier data.

[0060] Specifically, in the process of industrial production, since the products to be produced may include multiple components processed by different processing steps, the industrial production line data is divided by processing steps to achieve the segmentation of the production line process flow, ensuring the searchability and applicability of high-quality datasets during the transformation of the target production line.

[0061] Furthermore, since the raw materials required for each processing step are different, and the forward parameters of the processed products of some processing steps may be the product outputs of other processing steps, i.e. backward product parameters, by determining the process parameters, forward parameters and backward parameters through the processing steps, it is possible to effectively establish the correspondence between processes, between processes and products to be produced, and between processes and raw materials after acquiring data. This enables the processing steps extracted from high-quality datasets to be fully adapted to the production line that needs to be applied.

[0062] Step S206: Take each processing step of the target production line as the target step, and decompose the target parameters according to the target steps to obtain the target sub-parameters corresponding to each target step.

[0063] Specifically, by breaking down the target parameters through the target process, we can obtain the target sub-parameters for each target process. This not only determines the tasks of each target process, but also enables us to search high-quality datasets using the target process as a node. This avoids situations where data is difficult to match if we search directly through the production line, or where searching by equipment is difficult to adapt to the production line requirements.

[0064] Step S208: Search for a high-quality dataset based on the target process and target sub-parameters to obtain the forward parameters and process parameters of the corresponding target process.

[0065] Specifically, given the target sub-parameters, which are equivalent to the backward parameters of the established processing steps, the forward parameters and process parameters of the target steps can be obtained by searching a high-quality dataset for processing steps identical to the target steps and the target sub-parameters.

[0066] Specifically, it is the same as the target process, that is, the steps of the process flow are the same, and the required equipment is the same.

[0067] Step S210: In response to the matching of the forward parameters of the target process with the raw material parameters, the matching process parameters are invoked to guide the operation of the corresponding target process.

[0068] Among them, the forward parameters of the target process are matched with the raw material parameters, that is, all or part of the raw material parameters can be the same as the forward parameters, or the raw materials, after being processed by the target production line through a preset number of times, can be all or part of the same as the forward parameters.

[0069] The aforementioned automated production line control method based on high-quality datasets can acquire industrial production line data, target parameters of the products to be produced, and raw material parameters. Through preprocessing and data grouping of the industrial production line data, the method integrates the data by grouping it into components: process parameters of each processing step in the industrial production line, forward parameters of the products entering the processing step, and backward parameters of the products output from the processing step. This creates a closely linked and clearly defined parameter group from parameter input to parameter processing and output. Then, the target parameters of the products to be produced are decomposed according to each processing step of the target production line, breaking them down into target sub-parameters corresponding to each target step. Next, a high-quality dataset is searched using the target steps and target sub-parameters to obtain the forward and process parameters of each target step. If the forward parameters match the raw material parameters, the process parameters in the high-quality dataset are determined to be usable for the target production line, and they are invoked to guide the corresponding target steps in production. This effectively solves the problem that the effective data required by the production line is difficult to extract fully, leading to low efficiency in the use of industrial big data by existing enterprises.

[0070] In one embodiment, the processing steps include multiple process flows with a time-corresponding relationship; the process parameters include equipment parameters and process parameters; the preprocessing steps for the industrial production line data include the following steps S2042 to S2048. Wherein:

[0071] Step S2042: Filter out missing values ​​in the industrial production line data; missing values ​​include null values ​​and placeholders.

[0072] Step S2044: Remove processing steps containing missing values ​​from the industrial production line data to obtain candidate data.

[0073] Among these steps, the process of removing a processing step involves removing all data corresponding to that processing step to minimize the incompatibility of other data parameters due to missing data, thereby enhancing data stability and usability.

[0074] Step S2046: Parameter values ​​that meet one of the following conditions are judged as abnormal values: the equipment parameter is greater than the corresponding equipment operating limit value, the equipment model corresponding to the equipment parameter does not match the processing procedure, and the process parameter does not match the steps of the process flow.

[0075] Specifically, during the data collection process, data collection and entry errors may occur. Therefore, it is necessary to remove data parameters that are obviously not part of the processing procedure, such as data exceeding the limit, incorrect model entry, or process parameters that do not match the steps of the process flow.

[0076] Among them, process parameters are used to characterize the parameters required for each process step during the execution of the process, such as cutting size and cooling time.

[0077] Step S2048: Remove processing steps that contain outliers from the candidate data.

[0078] In one embodiment, such as Figure 2 As shown, the method further includes the following steps S302 to S306. Wherein:

[0079] Step S302: Construct a digital twin model based on the target production line, and input the equipment parameters, process parameters, and raw material parameters of each target process into the digital twin model to obtain the predicted finished product parameters.

[0080] Specifically, constructing a digital twin model of a target production line can be achieved by fully simulating the production architecture, sensor deployment, control system integration, data communication network, and modeling logical behavior rules, as well as through simulation analysis and visualization applications. For details, refer to existing methods for constructing digital twin models of production lines, which will not be elaborated here.

[0081] Step S304: Decompose the finished product parameters according to the target process to obtain finished product sub-parameters, and compare and analyze them with the corresponding target sub-parameters.

[0082] Step S306: In response to any target sub-parameter having a parameter difference exceeding a preset range with the corresponding finished product sub-parameter, output process error prompt information; the process error prompt information is used to indicate that the process parameters of the target process corresponding to the finished product sub-parameter are unreliable.

[0083] Specifically, even with data preprocessing, it is still possible to miss process parameters of processing steps that do not match the target process. Therefore, it is necessary to use a digital twin model to simulate whether the final finished product sub-parameters match the target sub-parameters of the product to be produced. If they do not match, it indicates that the process parameters of the corresponding processing step are unreliable and further parameter analysis is required.

[0084] In another embodiment, the process parameters corresponding to the mismatched processing steps can be directly eliminated, and other process parameters that meet the conditions can be selected from the high-quality dataset and verified. If they are still mismatched and the error is greater than the preset threshold, it indicates that there is a problem in the construction process of the digital twin model, and further investigation of the problem source of the digital twin model is needed.

[0085] In one embodiment, such as Figure 3 As shown, the method further includes the following steps S402 to S404. Wherein:

[0086] In step S402, in response to the output process error prompt information, the equipment parameters of the corresponding target process in the digital twin model are gradually adjusted according to the preset step size, with the adjustment direction being to shorten the parameter difference.

[0087] The preset step size can be set based on the error range of the parameter difference value.

[0088] Specifically, if the error range of the process error is smaller than the preset error range, and the digital twin model is constructed correctly, the predicted finished product sub-parameters can be gradually brought closer to the target sub-parameters by adjusting the parameter difference.

[0089] In step S404, if the parameter difference still exceeds the preset range, output a process parameter error prompt message; otherwise, operate the corresponding equipment according to the adjusted equipment parameters.

[0090] In one embodiment, both the gradual approach and stabilization of the finished product sub-parameter towards the target sub-parameter during parameter difference adjustment, and the initial approach and subsequent deviation of the finished product sub-parameter towards the target sub-parameter, indicate problems with both equipment and process parameters. If the parameter difference remains constant during adjustment, it indicates a problem with the process parameters.

[0091] In one embodiment, such as Figure 4 As shown, the method further includes the following steps S502 to S508. Wherein:

[0092] In step S502, in response to the output process parameter error prompt information, the process parameters corresponding to the target process are removed, and the processing processes in the high-quality dataset are searched again according to the raw material parameters and equipment parameters of the target process to obtain the correspondence between multiple sets of process parameters and backward parameters.

[0093] Specifically, since process parameters involve multiple parameters in the process steps, it is necessary to use a correlation model, combined with the advantage of the massive amount of industrial big data, to gradually learn the correspondence between each process parameter and the backward parameters, and then identify the problematic variables in the process parameters.

[0094] Step S504: Establish an association relationship model and substitute the corresponding relationships into the association relationship model for training to obtain the trained association relationship model.

[0095] In one embodiment, the data in the correspondence can be divided into a training set and a validation set. The training set is used to train the association model, and the validation set is used to validate the trained association model.

[0096] Step S506: Substitute the target sub-parameters corresponding to the parameter difference into the trained association model to obtain the process correction parameters.

[0097] Step S508: Replace the rejected process parameters with process correction parameters and re-predict the finished product parameters.

[0098] In one embodiment, if the finished product parameter prediction eliminates parameter discrepancies, the process correction parameters are used to replace the corresponding processing steps in the high-quality dataset, and the target production line to be adapted is marked.

[0099] In one embodiment, such as Figure 5 As shown, the method further includes the following steps S602 to S606. Wherein:

[0100] Step S602: In response to the temporal correspondence between multiple target processes in the target production line, search for the corresponding processing process in the high-quality dataset.

[0101] Step S604: If there exists a processing step in the forward sequence whose backward parameter is the same as the forward parameter of the adjacent backward sequence processing step, then determine whether the forward parameter of the processing step at the beginning of the time sequence correspondence matches the raw material parameter, and whether the backward parameter of the processing step at the end of the time sequence correspondence is the same as the target sub-parameter.

[0102] Specifically, when searching for high-quality datasets based on target processes, it may be impossible to find the forward parameters and process parameters corresponding to the target processes in the high-quality datasets based on the target processes and target sub-parameters. One reason is that the backward parameters of the target processes in the preceding time series (located in the forward time series) are the same as the forward parameters of the following time series (adjacent backward time series). That is, the two target processes are related target processes with a time series relationship. Therefore, the high-quality dataset can be searched by treating the related target processes as a whole and determining whether the forward parameters of the processing process at the beginning of the time series correspondence match the raw material parameters, and whether the backward parameters of the processing process at the end of the time series correspondence are the same as the target sub-parameters.

[0103] Step S606: If so, call its process parameters to guide the corresponding target process operation.

[0104] Specifically, if the forward parameter of the first processing step in the time-series correspondence matches the raw material parameter, and the backward parameter of the last processing step in the time-series correspondence is the same as the target sub-parameter, it means that the associated target process can find the same time-series correspondence in the high-quality dataset. Therefore, the process parameters of the time-series correspondence in the high-quality dataset can be used in the associated target process.

[0105] In one embodiment, the forward parameters of the target process are matched with the raw material parameters to characterize that the raw material parameters, after being processed a preset number of times by the target production line, conform to the forward parameters of the target process.

[0106] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0107] Based on the same inventive concept, this application also provides an automated production line control system based on high-quality datasets for implementing the automated production line control method based on high-quality datasets described above. The solution provided by this system is similar to the implementation scheme described in the above method. Therefore, the specific limitations of one or more embodiments of the automated production line control system based on high-quality datasets provided below can be found in the limitations of the automated production line control method based on high-quality datasets described above, and will not be repeated here.

[0108] To further illustrate the solution of this application, a specific example is provided below:

[0109] The automotive parts welding production line is used to produce Type A car frames. Now it has received a customized order to produce Type B car frames with similar structure but different dimensions and material thickness.

[0110] The system automatically cleans the data acquired from industrial big data, removing entire batch records with missing data, as well as current parameter records that were entered due to operational errors and were clearly outside the working range of the welding machine. Then, taking each welding station (such as the left-side spot welding station and the base plate arc welding station) as a unit, its corresponding welding parameters, incoming workpiece dimensions, and post-weld workpiece dimensions and quality are packaged into structured welding process packages and stored in a high-quality dataset.

[0111] Input the target parameters for the Type B chassis: the final 3D CAD model dimensions and the welding strength requirements at various locations. Based on the production line layout, the system breaks down the Type B chassis model by workstation, obtaining the target sub-parameters required for each workstation. For example, for the left side circumference spot welding workstation, the target sub-parameter is: after completing 20 welds, the dimensional tolerance of the key assembly holes in the left side circumference component must be within ±0.5mm. Using the left side circumference spot welding workstation and the aforementioned target sub-parameters as conditions, the system searches for high-quality datasets. During this process, it may not find a record that perfectly matches the Type B chassis dimensions, but it does find a record for another similar structure, the Type C chassis, whose backward parameters (post-welding dimensional accuracy) are highly consistent with the requirements of the Type B chassis. The system extracts the forward parameters (the incoming dimensions of the Type C chassis when entering the workstation) and process parameters (welding robot trajectory, spot welding parameters) from this record.

[0112] The system inputs the actual raw materials (new sheet metal dimensions) of the B-type chassis and the searched process parameters into the digital twin model of the welding production line. The model simulates the welding process and outputs the "predicted dimensions of the B-type chassis after welding." Comparison revealed that the deformation caused by several weld points exceeded the allowable tolerance range of the target sub-parameters (parameter difference). The system first attempted to automatically adjust the equipment parameters: fine-tuning the welding torch pressure and welding time (preset step size) in the digital twin. Simulation showed that the deformation decreased after adjustment but still did not meet the requirements. The system determined that the process parameters might be problematic and initiated a self-correcting process: finding all historical records using the same sheet metal and welding robot from a high-quality dataset, analyzing the relationship between the welding sequence (process parameters) and the amount of welding deformation (backward parameters), and training a predictive model. Based on the required accuracy of the B-type chassis (target sub-parameters), this model derived an optimized welding sequence suggestion (e.g., welding the middle first and then the ends, instead of the original sequence). This new sequence was then updated into the process plan as a "process correction parameter."

[0113] Secondly, such as Figure 6 As shown, this application also provides an automated production line control system 700 based on a high-quality dataset, the system comprising:

[0114] The acquisition module 701 is used to acquire industrial production line data, target parameters of the product to be produced, and raw material parameters.

[0115] The data processing module 702 is used to preprocess and group the acquired industrial production line data to obtain a high-quality dataset. The data grouping includes grouping based on the process parameters of each processing step in the industrial production line, the forward parameters of the product before entering the processing step, and the backward parameters of the product output from the processing step.

[0116] The parameter decomposition module 703 is used to take each processing step of the target production line as the target step, and decompose the target parameters according to each target step to obtain the target sub-parameters of the corresponding target steps.

[0117] Search module 704 is used to search for high-quality datasets based on the target process and the corresponding target sub-parameters to obtain the forward parameters and process parameters of the corresponding target process.

[0118] Module 705 is invoked in response to the matching of the forward parameters and raw material parameters of the target process, and the matching process parameters are invoked to guide the operation of the corresponding target process.

[0119] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the method provided in the first aspect of this application.

[0120] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method provided in the first aspect of this application.

[0121] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the method provided in the first aspect of this application.

[0122] The modules in the aforementioned automated production line control system based on high-quality datasets can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the computer device's memory as software, so that the processor can call and execute the corresponding operations of each module.

[0123] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0124] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0125] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. An automated production line control method based on high-quality datasets, characterized in that, The method includes: Acquire industrial production line data, target parameters of products to be manufactured, and raw material parameters; The industrial production line data is preprocessed and grouped to obtain a high-quality dataset; the data grouping includes grouping based on the process parameters of each processing step in the industrial production line, the forward parameters of the product before entering the processing step, and the backward parameters of the product output from the processing step. Each processing step of the target production line is taken as the target step, and the target parameters are decomposed according to the target steps to obtain the target sub-parameters corresponding to each target step; Search the high-quality dataset based on the target process and the target sub-parameters to obtain the forward parameters and process parameters corresponding to the target process; In response to a match between the forward parameters of the target process and the raw material parameters, the matching process parameters are invoked to guide the operation of the corresponding target process; In response to the existence of a temporal correspondence between multiple target processes of the target production line, the corresponding processing process is searched in the high-quality dataset; If there exists a backward parameter of the processing step in the forward sequence that is the same as the forward parameter of the processing step in the adjacent backward sequence, then determine whether the forward parameter of the processing step at the beginning of the sequence correspondence matches the raw material parameter, and whether the backward parameter of the processing step at the end of the sequence correspondence is the same as the target sub-parameter. If so, the process parameters described therein are invoked to guide the operation of the corresponding target process.

2. The method according to claim 1, characterized in that, The processing steps include multiple process flows with a time-corresponding relationship; The process parameters include equipment parameters and process parameters; the step of preprocessing the industrial production line data includes: Filter out missing values ​​in the industrial production line data; the missing values ​​include null values ​​and placeholders. The processing steps containing the missing values ​​in the industrial production line data are removed to obtain alternative data; A parameter value that meets one of the following conditions is judged as an outlier: the equipment parameter is greater than the limit value of the corresponding equipment operation, the equipment model corresponding to the equipment parameter does not match the processing procedure, and the process parameter does not match the steps of the process flow. The processing steps that contain the outlier values ​​in the candidate data are removed.

3. The method according to claim 2, characterized in that, The method further includes: A digital twin model is constructed based on the target production line, and the equipment parameters and process parameters of each target process, as well as the raw material parameters, are input into the digital twin model to obtain the predicted finished product parameters. The finished product parameters are decomposed according to the target process to obtain finished product sub-parameters, and then compared and analyzed with the corresponding target sub-parameters. In response to any difference between the target sub-parameter and the corresponding finished product sub-parameter exceeding a preset range, a process error prompt message is output; the process error prompt message is used to indicate that the process parameters of the target process corresponding to the finished product sub-parameter are unreliable.

4. The method according to claim 3, characterized in that, The method further includes: In response to the output of the process error prompt information, the equipment parameters of the corresponding target process in the digital twin model are gradually adjusted according to a preset step size, with the adjustment direction being to shorten the parameter difference; If the parameter difference still exceeds the preset range, a process parameter error prompt message will be output; otherwise, the corresponding equipment will be operated according to the adjusted equipment parameters.

5. The method according to claim 4, characterized in that, The method further includes: In response to the output of the process parameter error prompt information, the process parameters corresponding to the target process are removed, and the processing processes in the high-quality dataset are searched again according to the raw material parameters and equipment parameters of the target process to obtain the correspondence between multiple sets of process parameters and backward parameters; Establish an association relationship model, and substitute the corresponding relationship into the association relationship model for training to obtain a trained association relationship model; Substitute the target sub-parameters corresponding to the parameter difference into the trained association model to obtain the process correction parameters; The eliminated process parameters are replaced with the process correction parameters, and the finished product parameters are re-predicted.

6. The method according to claim 1, characterized in that, The forward parameters of the target process are matched with the raw material parameters to indicate that the raw material parameters, after being processed a preset number of times by the target production line, conform to the forward parameters of the target process.

7. An automated production line control system based on a high-quality dataset, characterized in that, The system is implemented based on the method described in any one of claims 1 to 6; the system comprises: The acquisition module is used to acquire industrial production line data, target parameters of the products to be produced, and raw material parameters. The data processing module is used to preprocess and group the acquired industrial production line data to obtain a high-quality dataset. The data grouping includes grouping based on the process parameters of each processing step in the industrial production line, the forward parameters of the product before entering the processing step, and the backward parameters of the product output from the processing step. The parameter decomposition module is used to take each processing step of the target production line as the target step, and decompose the target parameters according to each target step to obtain the target sub-parameters corresponding to the target step. The search module is used to search the high-quality dataset based on the target process and the corresponding target sub-parameters to obtain the forward parameters and process parameters corresponding to the target process. The calling module is used to call the matching process parameters to guide the operation of the corresponding target process in response to the matching of the forward parameters of the target process and the raw material parameters.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images