A collaborative optimization method and system for intelligent manufacturing
By establishing a random forest tree learning model in the intelligent production system, comparing process parameters in real time, and automatically adjusting production equipment, the quality traceability problem of products that cannot be set with tracking codes is solved, and rapid adjustment and equipment collaborative optimization are achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- TANGSHAN XUHUA INTELLIGENT TECH CO LTD
- Filing Date
- 2025-08-08
- Publication Date
- 2026-05-26
AI Technical Summary
In existing intelligent production systems, some products cannot be equipped with tracking codes due to limitations in appearance, material, or physical characteristics, resulting in the failure of quality traceability functions, inability to effectively record and trace, and inability to quickly adjust after quality defects occur.
By collecting historical process parameters and quality inspection results of production equipment, a random forest tree learning model is established to determine the range of qualified and defective parameters. The process parameters are compared in real time and adjustment data is generated to achieve automatic adjustment of the equipment.
It enables quality traceability and rapid adjustment for all types of products, reduces batch quality problems caused by defect expansion, and improves the accuracy of equipment adjustment and the efficiency of intelligent collaborative optimization of the production process.
Smart Images
Figure CN120975314B_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of intelligent manufacturing technology, and in particular to a collaborative optimization method and system for intelligent manufacturing. Background Technology
[0002] In the current wave of transformation and upgrading in the manufacturing industry, intelligent manufacturing has become a key path to enhance enterprises' core competitiveness. Through the deep integration of advanced technologies such as the Internet of Things, big data, and artificial intelligence, the automation and informatization levels of the production process have been continuously improved, not only achieving a significant leap in production efficiency but also placing higher demands on quality control throughout the product lifecycle. Quality traceability, as a core link in ensuring product quality and safety within the intelligent manufacturing system, can accurately locate problem nodes in the production process, providing data support for quality improvement.
[0003] In existing intelligent manufacturing systems, product quality traceability primarily relies on a technology that assigns a unique identifier to each product. This unique identifier is typically attached to the product's surface or interior in the form of a barcode, QR code, or Radio Frequency Identification (RFID) tag. During production, automated equipment collects and records this unique identifier in real time, creating a complete product data archive. When a product has a quality defect, this unique identifier can be used to quickly trace back to the specific production batch, operating procedure, equipment parameters, and operator information, thereby efficiently determining the cause of the quality defect.
[0004] However, the aforementioned quality traceability scheme based on unique identifiers has significant limitations in application. The realization of its core function depends entirely on the premise that the product can stably carry the unique identifier. However, in actual production, some products have special requirements for appearance and materials from customers, such as high-precision optical components and food-grade packaging, or are limited by their own physical characteristics, such as miniaturized electronic components and easily damaged flexible material products. It is difficult to set an effective tracking code without affecting the product's performance or appearance.
[0005] Therefore, existing quality traceability systems are unable to effectively record and trace the production process of such products, resulting in an inability to quickly adjust after quality defects occur. Summary of the Invention
[0006] This application provides a collaborative optimization method and system for intelligent manufacturing, which can effectively record and track product quality during the production process, and promptly adjust production equipment after determining that product defects exist, thereby reducing the product defect rate and improving product quality.
[0007] To achieve the above objectives, this application adopts the following technical solution:
[0008] Firstly, this application provides a collaborative optimization method for intelligent manufacturing, applied at the control end, the method comprising:
[0009] The control terminal collects historical process parameters of multiple production equipment for the target product and historical quality inspection results of the quality inspection equipment for the target product.
[0010] The control terminal classifies the historical process parameters of multiple production equipment according to the historical quality inspection results to obtain a sample parameter set, which includes: sample process parameters of multiple qualified samples and sample process parameters of multiple defect types.
[0011] The control terminal inputs the sample parameter set into the random forest tree learning model to learn and obtain the qualified parameter range set corresponding to the qualified sample and the defect parameter range set corresponding to different defect types.
[0012] The control terminal collects real-time process parameters of multiple production equipment for the target product. If the real-time process parameters belong to the defect parameter range set, the adjustment data of each production equipment is obtained by comparing the real-time process parameters with the defect parameter range set.
[0013] The control unit adjusts each production device based on the adjustment data.
[0014] In some possible implementations, the real-time process parameters include real-time sub-process parameters corresponding to each of multiple production equipment, and the defect parameter range set includes defect sub-process parameter ranges corresponding to each of multiple production equipment; the real-time process parameters belonging to the defect parameter range set include: each real-time sub-process parameter within its corresponding defect sub-process parameter range.
[0015] In some possible implementations, the control terminal adjusts each production device according to the adjustment data, including:
[0016] The control terminal obtains an adjustment signal for each production device based on the adjustment data, and sends the corresponding adjustment signal to each production device to achieve adjustment of each production device.
[0017] In some possible implementations, obtaining adjustment data for each production device by comparing the real-time process parameters with the defect parameter range set includes:
[0018] The comprehensive deviation between the real-time process parameters and the defect parameter ranges corresponding to the defect parameter range set is calculated.
[0019] The dynamic adjustment coefficient is obtained by combining the comprehensive deviation value with the historical adjustment records of the production equipment;
[0020] The initial adjustment value is obtained by multiplying the comprehensive deviation value by the dynamic adjustment coefficient, and the adjustment data is obtained by using the PID algorithm based on the initial adjustment value.
[0021] In some possible implementations, the control terminal collects historical process parameters of multiple production equipment for the target product and historical quality inspection results of the quality inspection equipment for the target product, including:
[0022] The control terminal receives process data packets sent by each production device for the target product and quality data packets sent by the quality inspection device for the target product. The process data packets carry process parameters and collection times collected by the production devices, and the quality data packets carry historical quality inspection results obtained after multiple production devices process the target product and the quality inspection device performs quality inspection.
[0023] The real-time process parameters collected by each production equipment are sorted and spliced according to the collection time to obtain the historical process parameters of multiple production equipment for the target product.
[0024] In some possible implementations, each production device is equipped with a corresponding quality inspection process, and the method further includes:
[0025] The control terminal receives defect information sent by the first target production equipment. The defect information is the non-conforming information determined by the quality inspection process based on the target process parameters collected by the first target production equipment for the target product.
[0026] The control terminal sends a discard command to the second target production equipment. The discard command is used to instruct the discarding of the target product. The second target production equipment is the next level equipment on the production line after the first target production equipment.
[0027] In some possible implementations, the method further includes:
[0028] The control unit determines whether the first target production equipment is the first production equipment and obtains a first determination result;
[0029] If the first judgment result indicates that the first target production equipment is the first production equipment, then the target process parameters are supplemented to obtain the target sample process parameters of the target defect type.
[0030] If the first judgment result indicates that the first target production equipment is not the first production equipment, then the real-time process parameters collected from the equipment before the first target production equipment on the production line are spliced with the target process parameters and a completion operation is performed to obtain the target sample process parameters of the target defect type.
[0031] Secondly, this application provides a collaborative optimization system for intelligent production, the system comprising:
[0032] The data acquisition module is used to collect historical process parameters of multiple production equipment for the target product and historical quality inspection results of the quality inspection equipment for the target product.
[0033] The classification module is used to classify the historical process parameters of multiple production equipment according to the historical quality inspection results to obtain a sample parameter set, which includes: sample process parameters of multiple qualified samples and sample process parameters of multiple defect types.
[0034] The learning module is used to input the sample parameter set into the random forest tree learning model to learn and obtain the qualified parameter range set corresponding to the qualified sample and the defect parameter range set corresponding to different defect types.
[0035] The data acquisition module is also used to acquire real-time process parameters of multiple production equipment for the target product;
[0036] The optimization module is used to obtain adjustment data for each production equipment by comparing the real-time process parameters with the defect parameter range set if the real-time process parameters belong to the defect parameter range set; and to adjust each production equipment according to the adjustment data.
[0037] In some possible implementations, the real-time process parameters include real-time sub-process parameters corresponding to each of multiple production equipment, and the defect parameter range set includes defect sub-process parameter ranges corresponding to each of multiple production equipment; the real-time process parameters belonging to the defect parameter range set include: each real-time sub-process parameter within its corresponding defect sub-process parameter range.
[0038] In some possible implementations, the optimization module is specifically used to obtain an adjustment signal for each production device based on the adjustment data, and send the corresponding adjustment signal to each production device to achieve adjustment of each production device.
[0039] In some possible implementations, the optimization module is specifically used to calculate the comprehensive deviation value between the real-time process parameters and the defect parameter range corresponding to the defect parameter range set; obtain the dynamic adjustment coefficient based on the comprehensive deviation value and the historical adjustment record of the production equipment; multiply the comprehensive deviation value and the dynamic adjustment coefficient to obtain the initial adjustment value; and obtain the adjustment data based on the initial adjustment value through a PID algorithm.
[0040] In some possible implementations, the acquisition module is specifically used to receive process data packets sent by various production equipment for the target product and quality data packets sent by quality inspection equipment for the target product. The process data packets carry process parameters and acquisition times collected by the production equipment, and the quality data packets carry historical quality inspection results obtained after multiple production equipment processes the target product and the quality inspection equipment performs quality inspection. The process parameters collected by each production equipment are sorted and concatenated according to the acquisition times to obtain historical process parameters of multiple production equipment for the target product.
[0041] In some possible implementations, each production device is equipped with a corresponding quality inspection process, and the system also includes a receiving module and a sending module;
[0042] The receiving module is used to receive defect information sent by the first target production equipment. The defect information is information on non-compliance determined by the quality inspection process based on the target process parameters collected by the first target production equipment for the target product.
[0043] The sending module is used to send a discard instruction to the second target production equipment. The discard instruction is used to instruct the discarding of the target product. The second target production equipment is the next level equipment on the production line after the first target production equipment.
[0044] In some possible implementations, the system further includes a judgment module; the judgment module is used to determine whether the first target production equipment is the first production equipment, and obtain a first judgment result; if the first judgment result indicates that the first target production equipment is the first production equipment, then the target process parameters are supplemented to obtain target sample process parameters of the target defect type; if the first judgment result indicates that the first target production equipment is not the first production equipment, then the real-time process parameters collected from the equipment before the first target production equipment on the production line are spliced with the target process parameters and supplemented to obtain target sample process parameters of the target defect type.
[0045] Thirdly, this application provides a computing device, including a memory and a processor;
[0046] The memory stores one or more computer programs, the one or more computer programs including instructions; when the instructions are executed by the processor, the computing device performs the method as described in any one of the first aspects.
[0047] Fourthly, this application provides a computer-readable storage medium for storing a computer program for performing the method as described in any one of the first aspects.
[0048] As can be seen from the above technical solution, this application has at least the following beneficial effects:
[0049] First, the technical solution of this application overcomes the limitation of the original solution's reliance on unique identifiers. The original quality traceability solution requires products to stably carry unique identifiers, resulting in products such as high-precision optical components and miniaturized electronic components, which are difficult to code, being unable to be effectively traced and adjusted. In contrast, this technical solution manages based on the process parameters of the production equipment. By collecting historical and real-time process parameters and combining them with historical quality inspection results to establish a parameter range model, the entire process does not rely on the product's own tracking code carrying capacity. Therefore, it can cover all types of target products, including those that cannot be coded due to appearance, material, or physical characteristics, thus solving the application blind spots of the original solution.
[0050] Secondly, the technical solution of this application enables early intervention and rapid adjustment of quality defects. Existing solutions rely on traceability codes to trace the production process after a quality defect occurs, resulting in significant delays. This solution, however, pre-determines the acceptable parameter range and the defect parameter range using a random forest tree learning model. Process parameters are collected in real-time during production. Once real-time process parameters fall within the defect parameter range, adjustment data can be immediately generated through parameter comparison, and equipment adjustments can be made. This mechanism shifts the quality control node from tracing back after a defect occurs to intervening when a defect trend emerges, significantly shortening the time from problem discovery to resolution and preventing batch quality problems caused by defect escalation.
[0051] Furthermore, the technical solution of this application improves the accuracy and targeting of equipment adjustments. Since the defect parameter range set is obtained through machine learning model training based on historical process parameters and corresponding historical quality inspection results, each defect type has a clearly defined parameter range. Therefore, when real-time process parameters become abnormal, the specific process parameters and related production equipment causing the defect can be accurately located. Compared to the potentially fuzzy traceability in the original solution, the adjustment data in this solution is obtained through a direct comparison between real-time process parameters and defect parameter ranges. This allows for targeted adjustments to the corresponding equipment based on different defect types, improving the effectiveness of production equipment adjustments.
[0052] Finally, the technical solution of this application achieves collaborative optimization and automated control of production equipment. In this solution, the control terminal coordinates the parameter acquisition, analysis, and adjustment of multiple production devices. A parameter range system established through model learning provides a unified standard for the collaborative operation of each device. When the real-time process parameters of a certain device fall within the defective parameter range, the control terminal can generate adjustment data based on global parameter correlation analysis, ensuring that the adjustment direction of each device is consistent and avoiding new production imbalances caused by local adjustments. Simultaneously, the entire process from data acquisition to equipment adjustment is automatically completed by the control terminal, reducing errors and delays from manual judgment and improving the intelligence and collaborative optimization efficiency of the production process.
[0053] It should be understood that the descriptions of technical features, technical solutions, beneficial effects, or similar language in this application do not imply that all features and advantages can be achieved in any single embodiment. Rather, it is understood that the description of a feature or beneficial effect means that a specific technical feature, technical solution, or beneficial effect is included in at least one embodiment. Therefore, the descriptions of technical features, technical solutions, or beneficial effects in this specification do not necessarily refer to the same embodiment. Furthermore, the technical features, technical solutions, and beneficial effects described in this embodiment can be combined in any suitable manner. Those skilled in the art will understand that embodiments can be implemented without one or more specific technical features, technical solutions, or beneficial effects of a particular embodiment. In other embodiments, additional technical features and beneficial effects may be identified in specific embodiments that do not embody all embodiments. Attached Figure Description
[0054] Figure 1 A flowchart illustrating a collaborative optimization method for intelligent manufacturing provided in this application embodiment;
[0055] Figure 2 A schematic diagram of a collaborative optimization system for intelligent manufacturing provided in an embodiment of this application;
[0056] Figure 3 This is a schematic diagram of a computing device provided in an embodiment of this application. Detailed Implementation
[0057] The terms "first," "second," and "third," etc., used in this application specification and accompanying drawings are used to distinguish different objects, not to limit a specific order.
[0058] In the embodiments of this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design that is described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design. Specifically, the use of the terms "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.
[0059] To make the technical solution of this application clearer and easier to understand, the technical solution of this application will be described below with reference to the accompanying drawings.
[0060] like Figure 1 As shown, this figure is a flowchart of a collaborative optimization method for intelligent manufacturing provided in an embodiment of this application. This method can be executed by a control terminal, or it can be executed collaboratively by the control terminal and production equipment. Specifically, the method includes:
[0061] S101. The control terminal collects historical process parameters of multiple production equipment for the target product and historical quality inspection results of the quality inspection equipment for the target product.
[0062] The control terminal refers to the equipment used to control the production equipment. For example, the control terminal can be a server or a terminal device. The production equipment refers to the equipment used in the process of producing products. For example, it can be the equipment used to process existing products. There can be multiple production equipment, and each production equipment can include at least one process.
[0063] Initially, the production equipment is in standby or off state. When production is required, the control terminal sends a first start signal to the equipment. This first start signal may carry a unique quality package number and start time. Upon receiving the first start signal, the production equipment can establish a temporary database corresponding to the unique quality package number. After collecting process parameters, each piece of equipment can store these parameters sequentially in its corresponding temporary database. The temporary database corresponding to the unique quality package number stores process parameters arranged in chronological order.
[0064] Quality inspection equipment refers to equipment used to inspect the quality of products. This equipment can be located at the end of the production line to inspect the produced products and obtain inspection results. These results include pass / fail and defect types. Similarly, the control unit can send a second start signal to the quality inspection equipment. Upon receiving the second start signal, the quality inspection equipment begins inspecting the products produced by the production equipment, establishes a temporary database corresponding to the unique number of the quality package, and stores the inspection results in the temporary database. The temporary database corresponding to the unique number of the quality package can also store the inspection results.
[0065] Each production and quality inspection device can upload data from the temporary database to the control terminal, which can then process this data, such as performing alignment operations.
[0066] In some embodiments, the control terminal receives process data packets sent by each production device for the target product and quality data packets sent by the quality inspection device for the target product. The process data packets carry process parameters collected by the production devices and the collection time, and the quality data packets carry historical quality inspection results obtained after multiple production devices process the target product and the quality inspection device performs quality inspection, as well as a unique quality package number. The process parameters collected by each production device are sorted and concatenated according to the collection time to obtain the historical process parameters of multiple production devices for the target product.
[0067] Taking the production of food-grade plastic packaging bags as an example, this product cannot have a tracking code set on its surface because it needs to meet food contact standards. It involves three production equipment: equipment A, such as printing equipment; equipment B, such as laminating equipment; equipment C, such as slitting equipment; and one quality inspection equipment.
[0068] The process of transmitting process data packets is as follows:
[0069] Equipment A sends: Process parameters (printing speed 30m / min, ink temperature 50℃), acquisition time 10:00:05.
[0070] Equipment B sends: Process parameters (composite temperature 80℃, pressure 0.3MPa), acquisition time 10:00:12.
[0071] Equipment C sends: Process parameters (cutting speed 25m / min, blade distance 15cm), acquisition time 10:00:20.
[0072] Equipment A sends the following again: process parameters (printing speed 28m / min, ink temperature 52℃), acquisition time 10:00:28.
[0073] Equipment B sends the following again: process parameters (composite temperature 78℃, pressure 0.28MPa), acquisition time 10:00:35.
[0074] Equipment C sends the following again: process parameters (cutting speed 26m / min, blade distance 14.8cm), acquisition time 10:00:42.
[0075] The transmission process of quality data packets is as follows:
[0076] Quality inspection equipment results: For packaging bags produced in the first 3 time slots (10:00:05-10:00:20), the quality inspection result is qualified; for packaging bags produced in the last 3 time slots (10:00:28-10:00:42), the quality inspection result is unqualified, and the defect type is composite layer peeling.
[0077] After obtaining the above data, the control terminal can perform data splicing. For example, sorting the data from earliest to latest by collection time, the process parameters can be spliced sequentially to obtain multiple historical process parameters: Historical process parameter 1: 10:00:05, Equipment A: printing speed 30m / min, ink temperature 50℃; 10:00:12, Equipment B: lamination temperature 80℃, pressure 0.3MPa; 10:00:20, Equipment C: slitting speed 25m / min, blade distance 15cm; Historical process parameter 2: 10:00:28, Equipment A: printing speed 28m / min, ink temperature 52℃; 10:00:35, Equipment B: lamination temperature 78℃, pressure 0.28MPa; 10:00:42, Equipment C: slitting speed 26m / min, blade distance 14.8cm.
[0078] By using the above method, a large number of process data packages and corresponding quality data packages can be obtained, which in turn yield a large amount of tagged process data, namely, the aforementioned historical process parameters and the historical quality inspection results corresponding to the historical process parameters.
[0079] S102. The control terminal classifies the historical process parameters of multiple production equipment according to the historical quality inspection results to obtain a sample parameter set.
[0080] The sample parameter set includes: sample process parameters for multiple qualified samples and sample process parameters for multiple defect types.
[0081] For example, historical process parameter 1 is a qualified sample, and historical process parameter 2 is a defective sample, with the defect type being composite layer peeling. After obtaining a large number of historical process parameters, each historical process parameter can be classified. The defect type can be determined based on the actual scenario; in the above embodiment, the defect type could also be cutting size deviation, ink peeling, etc.
[0082] S103. The control terminal inputs the sample parameter set into the random forest tree learning model to learn and obtain the qualified parameter range set corresponding to the qualified sample and the defect parameter range set corresponding to different defect types.
[0083] In this step, the sample parameters are first divided into data sets. For example, 70% of the data is used as the training set and 30% as the validation set. The training set is used for model learning, and the validation set is used to verify the accuracy of the model.
[0084] Continuing with the above example, key features of the sample parameters are extracted. For example, key features could be: speed and temperature matching values, temperature and pressure synergy values of the laminating equipment, etc., where speed and temperature matching values can refer to the ratio of printing speed to ink temperature, and temperature and pressure synergy values can refer to the product of laminating temperature and pressure.
[0085] First, the random forest tree learning model is constructed as follows:
[0086] The number of trees is determined based on the sample size to ensure model stability while avoiding overfitting. During training each decision tree, 3-4 features (e.g., printing speed, lamination temperature, lamination pressure) are randomly selected from all process parameters as splitting criteria to improve the model's generalization ability. The maximum depth of the decision tree is set to 5 layers to avoid excessive model complexity due to excessive depth, which would reduce training and prediction efficiency and prevent the model from overfitting to the training data. The optimal splitting node is determined by comparing the splitting effect of parameter thresholds, such as whether the lamination temperature is ≥79℃, until the quality inspection results of the samples in the leaf nodes tend to be consistent.
[0087] After the model is built, it can be trained and validated, as follows:
[0088] During training, the training set is input into the model, and each decision tree independently learns the mapping relationship between parameters and quality inspection results. For example, when the printing speed is ≤30m / min and the composite pressure is ≥0.3MPa, the sample is more likely to pass. In this process, the model automatically adjusts the parameters of the decision trees to minimize the prediction error on the training set, so that each decision tree can capture the relationship between process parameters and product quality inspection results as accurately as possible.
[0089] During the validation process, the model's prediction accuracy is tested using a validation set. If the accuracy of identifying a certain defect type is below 85%, the sample size for that defect type is increased, for example, by adding three sets of process parameters for composite layer peeling. The model is then retrained until the accuracy reaches the target. Once the model achieves the desired prediction accuracy through multiple training and optimizations, the parameter range set can be extracted based on the trained model, providing crucial reference for production process optimization.
[0090] After completing model training, the parameter range set can be extracted, as follows:
[0091] In the process of extracting the qualified parameter range set, the minimum and maximum values of the parameters corresponding to all leaf nodes judged as "qualified" in the statistical model are taken as the qualified range. For example, the qualified range for printing speed is the interval between the minimum and maximum speed values among all qualified leaf nodes. By extracting the qualified parameter range set, the range of process parameters that ensure product quality during production can be clearly defined. Based on this, the qualified parameter range can be further refined and optimized in conjunction with actual production conditions. For example, the influence of factors such as the characteristics of different batches of raw materials and the degree of equipment aging on the parameters can be considered, and the parameter range can be dynamically adjusted to make the qualified parameter range set more practical and accurate.
[0092] During the extraction of the defect parameter range set, for each defect type, the parameter values of the corresponding leaf nodes are statistically analyzed separately to determine the exclusive parameter range for each defect type. For example, the composite pressure range corresponding to "composite layer peeling" is the interval between the minimum and maximum pressure values among all leaf nodes with this defect. Combining the qualified parameter range set with the defect parameter range set for analysis can form a complete production process parameter guidance system, providing a precise basis for parameter control during the production process and effectively avoiding product quality problems caused by improper parameter settings. Based on this, the production process parameter guidance system can be embedded into the enterprise production management system. By collecting real-time process parameters, the system automatically compares qualified and defective parameter ranges. Once a parameter is found to deviate from the qualified range, the system immediately issues an early warning and provides adjustment suggestions based on historical data and model analysis, realizing intelligent and precise control of the production process.
[0093] S104. The control terminal collects real-time process parameters of multiple production equipment for the target product. If the real-time process parameters belong to the defect parameter range set, the adjustment data of each production equipment is obtained by comparing the real-time process parameters with the defect parameter range set.
[0094] After obtaining the set of qualified parameters and the set of defective parameters, the control terminal can judge the real-time process parameters collected by the production equipment for the target product based on the set of qualified parameters and the set of defective parameters, and thus determine whether there are defects.
[0095] The control terminal can collect real-time process parameters of multiple production equipment for the target product. If the real-time process parameters belong to the defect parameter range set, the adjustment data of each production equipment can be obtained by comparing the real-time process parameters with the defect parameter range set.
[0096] The real-time process parameters include real-time sub-process parameters corresponding to each of the multiple production equipment, and the defect parameter range set includes defect sub-process parameter ranges corresponding to each of the multiple production equipment; the real-time process parameters belonging to the defect parameter range set can mean that each real-time sub-process parameter is within its corresponding defect sub-process parameter range.
[0097] In some embodiments, obtaining adjustment data for each production equipment by comparing the real-time process parameters with the defect parameter range set can be achieved by: first calculating the comprehensive deviation value between the real-time process parameters and the corresponding defect parameter range in the defect parameter range set; obtaining a dynamic adjustment coefficient based on the comprehensive deviation value and the historical adjustment records of the production equipment; multiplying the comprehensive deviation value by the dynamic adjustment coefficient to obtain an initial adjustment value; and obtaining adjustment data based on the initial adjustment value using a PID algorithm.
[0098] The following example illustrates this.
[0099] The real-time process parameters can be: printing speed 33m / min, ink temperature 47℃, lamination temperature 76℃, lamination pressure 0.26MPa, slitting speed 27m / min, and blade distance 14.7cm. The defect type corresponding to these real-time process parameters is lamination layer peeling. Specifically, through parameter comparison, it is determined that the lamination temperature of 76℃ belongs to the defect sub-process parameter range (74-80℃) corresponding to lamination layer bubbles, and the lamination pressure of 0.26MPa belongs to the defect sub-process parameter range (0.27-0.28MPa) corresponding to lamination layer peeling defects.
[0100] For each parameter falling within the range of the defective sub-process parameters, calculate its deviation from the boundary of the defective sub-process parameter range, and obtain the comprehensive deviation value through weighted summation. See the following formula:
[0101]
[0102] in, Indicates the first The single-parameter deviation value of each process parameter is used to reflect the degree to which a single parameter deviates from the center of the defect range. Indicates the first The real-time value of each process parameter, i.e., the real-time process parameter. Indicates the first The center value of the defect range corresponding to each process parameter Indicates the first The defect range half-width corresponding to each process parameter.
[0103]
[0104] in, This represents the overall deviation value, used to reflect the degree of deviation of all abnormal parameters. This represents the total number of process parameters that fall within the range of defective sub-process parameters. Indicates the first The influence weight of each process parameter.
[0105] Based on historical adjustment records of production equipment, the relationship between different parameter adjustments and the final quality improvement effect is analyzed to generate dynamic adjustment coefficients. The formula for calculating the dynamic adjustment coefficients is:
[0106]
[0107] in, Indicates the dynamic adjustment coefficient. This represents the base coefficient, which can be a preset adjustment benchmark value based on the defect type. For example, the base coefficient for a composite layer defect is set to 1.2. This represents the historical effect correction factor, which can be a correction value obtained based on historical adjustment effects, such as when the historical adjustment effect is 80%. , This represents the equipment status factor, a correction value used to reflect the current operating status of the equipment. For example, it is set to 1.1 when efficiency decreases after continuous operation.
[0108] The initial adjustment value can be calculated using the following formula:
[0109]
[0110] in, This represents the initial adjustment value.
[0111] The process of obtaining adjustment data from the initial adjustment value using the PID algorithm is as follows:
[0112]
[0113] in, This represents the final adjustment data, such as the specific adjustment amount. This represents the proportionality coefficient. Represents the integral coefficient. Denotes the differential coefficient. This represents the current error, i.e., the initial adjustment value. This represents the integral term of the error, the cumulative error value from the initial time to the current time. This represents the differential term of the error, which is the rate of change of the current error compared to the error at the previous time step.
[0114] Through iterative adjustment by the PID controller, the parameters gradually approach the center of the acceptable range, forming a stable closed-loop control. This adjustment mechanism avoids the oscillation problem caused by simple proportional control and also solves the slow response of integral control, making it suitable for scenarios with high parameter stability requirements.
[0115] S105. The control terminal adjusts each production device according to the adjustment data.
[0116] After receiving the adjustment data, the control unit can adjust each production device based on that data.
[0117] The control terminal can adjust each production device based on the adjustment data by obtaining an adjustment signal for each production device and sending the corresponding adjustment signal to each production device. Alternatively, if a production device does not require adjustment, no adjustment signal may be sent to that device.
[0118] In some embodiments, each production device is configured with a corresponding quality inspection process. The control terminal receives defect information sent by a first target production device. The defect information is non-conforming information determined by the quality inspection process based on the target process parameters collected by the first target production device for the target product. The control terminal sends a discard instruction to a second target production device. The discard instruction is used to instruct the discarding of the target product. The second target production device is the next-level device on the production line after the first target production device.
[0119] In this embodiment, after the first target production equipment detects a non-conforming product, the control terminal immediately instructs the next-level equipment, i.e., the second target production equipment, to discard the product. This directly avoids the downstream processes from continuously processing non-conforming products, reducing the ineffective input of resources such as raw materials, energy, and equipment operating time from the source, and lowering the resource waste rate in the production process. By intercepting non-conforming products from entering the next level of process, downstream equipment does not need to consume production time for non-conforming products, and can concentrate all production capacity on processing qualified products, reducing the proportion of ineffective labor, shortening the production cycle of a unit qualified product, and improving the effective output efficiency of the production line. Timely interception at the process node where non-conforming products are generated can prevent defects from further spreading in subsequent processes or causing secondary defects, locking quality problems within a single process, simplifying the problem tracing and root cause location process, accelerating the formation of a quality control closed loop of detection, interception, and adjustment, and reducing batch quality risks. This solution achieves non-conforming product interception through instruction transmission between equipment, without relying on the physical identification of the product itself. It is applicable to all product production scenarios where identification cannot be set or the identification is easily damaged, ensuring that quality control covers all types of products and eliminating the application blind spots of traditional traceability solutions.
[0120] In some embodiments, the method further includes: the control terminal determining whether the first target production equipment is the first production equipment, and obtaining a first determination result; if the first determination result indicates that the first target production equipment is the first production equipment, then performing a completion operation on the target process parameters to obtain target sample process parameters of the target defect type; if the first determination result indicates that the first target production equipment is not the first production equipment, then performing a completion operation after splicing the real-time process parameters collected from the equipment before the first target production equipment on the production line with the target process parameters to obtain target sample process parameters of the target defect type.
[0121] If the control unit determines that a product is defective through the quality inspection process of the first target production equipment, it will stop the production equipment from processing or handling the defective product. This prevents subsequent production equipment from collecting relevant data. To ensure data length consistency, data alignment can be performed, for example, by padding with zeros. After data length consistency is achieved, this data can be used for targeted model training.
[0122] In this embodiment, when the first target production equipment is the first piece of equipment, a complete production process data chain cannot be formed solely through its target process parameters. The completion operation can fill in the gaps in subsequent process parameters, such as supplementing default initial parameters or subsequent benchmark parameters for similar qualified products. When the first target production equipment is not the first piece of equipment, splicing the parameters of preceding equipment before completion integrates the entire process data. In both cases, the sample parameter set for the target defect type contains complete production process information, ensuring uniform data length and avoiding invalid samples due to missing data.
[0123] Based on the above description, the technical solution of this application overcomes the limitation of the original solution's reliance on unique identifiers. The original quality traceability solution requires products to stably carry unique identifiers, resulting in products such as high-precision optical components and miniaturized electronic components, which are difficult to code, being unable to be effectively traced and adjusted. In contrast, this technical solution manages based on the process parameters of the production equipment. By collecting historical and real-time process parameters and combining them with historical quality inspection results to establish a parameter range model, the entire process does not rely on the product's own tracking code carrying capacity. Therefore, it can cover all types of target products, including those that cannot be coded due to appearance, material, or physical characteristics, thus solving the application blind spots of the original solution.
[0124] Secondly, the technical solution of this application enables early intervention and rapid adjustment of quality defects. Existing solutions rely on traceability codes to trace the production process after a quality defect occurs, resulting in significant delays. This solution, however, pre-determines the acceptable parameter range and the defect parameter range using a random forest tree learning model. Process parameters are collected in real-time during production. Once real-time process parameters fall within the defect parameter range, adjustment data can be immediately generated through parameter comparison, and equipment adjustments can be made. This mechanism shifts the quality control node from tracing back after a defect occurs to intervening when a defect trend emerges, significantly shortening the time from problem discovery to resolution and preventing batch quality problems caused by defect escalation.
[0125] Furthermore, the technical solution of this application improves the accuracy and targeting of equipment adjustments. Since the defect parameter range set is obtained through machine learning model training based on historical process parameters and corresponding historical quality inspection results, each defect type has a clearly defined parameter range. Therefore, when real-time process parameters become abnormal, the specific process parameters and related production equipment causing the defect can be accurately located. Compared to the potentially fuzzy traceability in the original solution, the adjustment data in this solution is obtained through a direct comparison between real-time process parameters and defect parameter ranges. This allows for targeted adjustments to the corresponding equipment based on different defect types, improving the effectiveness of production equipment adjustments.
[0126] Finally, the technical solution of this application achieves collaborative optimization and automated control of production equipment. In this solution, the control terminal coordinates the parameter acquisition, analysis, and adjustment of multiple production devices. A parameter range system established through model learning provides a unified standard for the collaborative operation of each device. When the real-time process parameters of a certain device fall within the defective parameter range, the control terminal can generate adjustment data based on global parameter correlation analysis, ensuring that the adjustment direction of each device is consistent and avoiding new production imbalances caused by local adjustments. Simultaneously, the entire process from data acquisition to equipment adjustment is automatically completed by the control terminal, reducing errors and delays from manual judgment and improving the intelligence and collaborative optimization efficiency of the production process.
[0127] The above text combined Figure 1 The collaborative optimization method for intelligent production provided in the embodiments of this application has been described in detail. The system and equipment provided in the embodiments of this application will be described below with reference to the accompanying drawings.
[0128] like Figure 2 As shown in the figure, this is a schematic diagram of a collaborative optimization system for intelligent manufacturing provided in an embodiment of this application. The system includes:
[0129] The data acquisition module 201 is used to acquire historical process parameters of multiple production equipment for the target product and historical quality inspection results of the quality inspection equipment for the target product;
[0130] The classification module 202 is used to classify the historical process parameters of multiple production equipment according to the historical quality inspection results to obtain a sample parameter set, which includes: sample process parameters of multiple qualified samples and sample process parameters of multiple defect types.
[0131] Learning module 203 is used to input the sample parameter set into the random forest tree learning model to learn and obtain the qualified parameter range set corresponding to the qualified sample and the defect parameter range set corresponding to different defect types.
[0132] The acquisition module 201 is also used to acquire real-time process parameters of multiple production equipment for the target product;
[0133] The optimization module 204 is used to obtain adjustment data for each production equipment by comparing the real-time process parameters with the defect parameter range set if the real-time process parameters belong to the defect parameter range set; and to adjust each production equipment according to the adjustment data.
[0134] In some possible implementations, the real-time process parameters include real-time sub-process parameters corresponding to each of multiple production equipment, and the defect parameter range set includes defect sub-process parameter ranges corresponding to each of multiple production equipment; the real-time process parameters belonging to the defect parameter range set include: each real-time sub-process parameter within its corresponding defect sub-process parameter range.
[0135] In some possible implementations, the optimization module 204 is specifically used to obtain an adjustment signal for each production device based on the adjustment data, and send the corresponding adjustment signal to each production device to achieve adjustment of each production device.
[0136] In some possible implementations, the optimization module 204 is specifically used to calculate the comprehensive deviation value between the real-time process parameters and the defect parameter range corresponding to the defect parameter range set; obtain the dynamic adjustment coefficient based on the comprehensive deviation value and the historical adjustment record of the production equipment; multiply the comprehensive deviation value and the dynamic adjustment coefficient to obtain the initial adjustment value; and obtain the adjustment data based on the initial adjustment value through a PID algorithm.
[0137] In some possible implementations, the acquisition module 201 is specifically used to receive process data packets sent by various production equipment for the target product and quality data packets sent by quality inspection equipment for the target product. The process data packets carry process parameters and acquisition time collected by the production equipment, and the quality data packets carry historical quality inspection results obtained after multiple production equipment processes the target product and the quality inspection equipment performs quality inspection. The real-time process parameters collected by each production equipment are sorted and spliced according to the acquisition time to obtain historical process parameters of multiple production equipment for the target product.
[0138] In some possible implementations, each production device is equipped with a corresponding quality inspection process, and the system also includes a receiving module and a sending module;
[0139] The receiving module is used to receive defect information sent by the first target production equipment. The defect information is the non-conforming information determined by the quality inspection process based on the target process parameters collected by the first target production equipment for the target product.
[0140] The sending module is used to send a discard instruction to the second target production equipment. The discard instruction is used to instruct the discarding of the target product. The second target production equipment is the next level equipment on the production line after the first target production equipment.
[0141] In some possible implementations, the system further includes a judgment module; the judgment module is used to determine whether the first target production equipment is the first production equipment, and obtain a first judgment result; if the first judgment result indicates that the first target production equipment is the first production equipment, then the target process parameters are supplemented to obtain target sample process parameters of the target defect type; if the first judgment result indicates that the first target production equipment is not the first production equipment, then the real-time process parameters collected from the equipment before the first target production equipment on the production line are spliced with the target process parameters and supplemented to obtain target sample process parameters of the target defect type.
[0142] The intelligent production collaborative optimization system according to the embodiments of this application can correspond to the execution of the methods described in the embodiments of this application, and the other operations and / or functions of each module / unit of the intelligent production collaborative optimization system are respectively for implementing Figure 1 For the sake of brevity, the corresponding processes of each method in the illustrated embodiments will not be described in detail here.
[0143] This application also provides a computing device. For example... Figure 3 As shown in the figure, this is a schematic diagram of a computing device provided in an embodiment of this application. The computing device 700 includes a bus 701, a processor 702, a communication interface 703, and a memory 704. The processor 702, the memory 704, and the communication interface 703 communicate with each other via the bus 701.
[0144] The 701 bus can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of representation, Figure 3 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.
[0145] The processor 702 can be any one or more of the following processors: central processing unit (CPU), graphics processing unit (GPU), microprocessor (MP), or digital signal processor (DSP).
[0146] The communication interface 703 is used for communication with external devices.
[0147] Memory 704 may include volatile memory, such as random access memory (RAM). Memory 704 may also include non-volatile memory, such as read-only memory (ROM), flash memory, hard disk drive (HDD), or solid state drive (SSD).
[0148] The memory 704 stores executable code, and the processor 702 executes the executable code to perform the aforementioned collaborative optimization method for intelligent manufacturing.
[0149] Specifically, in achieving Figure 2 In the case of the illustrated embodiment, and Figure 2 When the modules or units of the intelligent manufacturing collaborative optimization system described in the embodiments are implemented through software, the following steps are executed: Figure 2 The software or program code required for the functions of each module / unit can be partially or entirely stored in memory 704. Processor 702 executes the program code corresponding to each unit stored in memory 704, and executes the aforementioned collaborative optimization method for intelligent manufacturing.
[0150] This application also provides a computer-readable storage medium. The computer-readable storage medium can be any available medium that a computing device can store, or a data storage device such as a data center containing one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state drive). The computer-readable storage medium includes instructions that instruct the computing device to execute the aforementioned collaborative optimization method for intelligent manufacturing.
[0151] This application also provides a computer program product comprising one or more computer instructions. When the computer instructions are loaded and executed on a computing device, all or part of the processes or functions described in this application are generated.
[0152] The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions may be transmitted from one website, computer, or data center to another website, computer, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) means.
[0153] When the computer program product is executed by a computer, the computer executes any of the aforementioned collaborative optimization methods for intelligent manufacturing. The computer program product can be a software installation package; when any of the aforementioned collaborative optimization methods for intelligent manufacturing is required, the computer program product can be downloaded and executed on the computer.
[0154] The descriptions of the processes or structures corresponding to the above figures each have their own emphasis. For parts of a process or structure that are not described in detail, please refer to the relevant descriptions of other processes or structures.
[0155] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions within the technical scope disclosed in this application should be covered within the scope of protection of this application.
Claims
1. A collaborative optimization method for intelligent manufacturing, characterized in that, Applied to the control terminal, the method includes: The control terminal collects historical process parameters of multiple production equipment for the target product and historical quality inspection results of the quality inspection equipment for the target product. The control terminal classifies the historical process parameters of multiple production equipment according to the historical quality inspection results to obtain a sample parameter set, which includes: sample process parameters of multiple qualified samples and sample process parameters of multiple defect types. The control terminal inputs the sample parameter set into the random forest tree learning model to learn and obtain the qualified parameter range set corresponding to the qualified sample and the defect parameter range set corresponding to different defect types. The control terminal collects real-time process parameters of multiple production equipment for the target product. If the real-time process parameters belong to the defect parameter range set, the adjustment data of each production equipment is obtained by comparing the real-time process parameters with the defect parameter range set. The control terminal adjusts each production device according to the adjustment data; The step of obtaining adjustment data for each production device by comparing the real-time process parameters with the defect parameter range set includes: The comprehensive deviation between the real-time process parameters and the defect parameter ranges corresponding to the defect parameter range set is calculated. The dynamic adjustment coefficient is obtained by combining the comprehensive deviation value with the historical adjustment records of the production equipment. The initial adjustment value is obtained by multiplying the comprehensive deviation value by the dynamic adjustment coefficient, and the adjustment data is obtained by using the PID algorithm based on the initial adjustment value; The formula for calculating the dynamic adjustment coefficient is as follows: Indicates the dynamic adjustment coefficient. Indicates the base coefficient. Indicates the historical effect correction factor. Indicates the device status factor; The initial adjustment value is calculated using the following formula: in, Indicates the initial adjustment value; This represents the overall deviation value. This represents the total number of process parameters that fall within the range of defective sub-process parameters. Indicates the first The influence weight of each process parameter; Indicates the first The single-parameter deviation value of each process parameter. Indicates the first Real-time values of each process parameter Indicates the first The center value of the defect range corresponding to each process parameter Indicates the first The defect range half-width corresponding to each process parameter.
2. The method according to claim 1, characterized in that, The real-time process parameters include real-time sub-process parameters corresponding to each of the multiple production equipment, and the defect parameter range set includes the defect sub-process parameter ranges corresponding to each of the multiple production equipment. The real-time process parameters belong to the defect parameter range set, including: each real-time sub-process parameter within its corresponding defect sub-process parameter range.
3. The method according to claim 1, characterized in that, The control terminal adjusts each production device according to the adjustment data, including: The control terminal obtains an adjustment signal for each production device based on the adjustment data, and sends the corresponding adjustment signal to each production device to achieve adjustment of each production device.
4. The method according to claim 1, characterized in that, The control terminal collects historical process parameters of multiple production equipment for the target product and historical quality inspection results of the quality inspection equipment for the target product, including: The control terminal receives process data packets sent by each production device for the target product and quality data packets sent by the quality inspection device for the target product. The process data packets carry process parameters and collection times collected by the production devices, and the quality data packets carry historical quality inspection results obtained after multiple production devices process the target product and the quality inspection device performs quality inspection. The real-time process parameters collected by each production equipment are sorted and spliced according to the collection time to obtain the historical process parameters of multiple production equipment for the target product.
5. The method according to claim 4, characterized in that, Each production piece of equipment is equipped with a corresponding quality inspection process, and the method further includes: The control terminal receives defect information sent by the first target production equipment. The defect information is the non-conforming information determined by the quality inspection process based on the target process parameters collected by the first target production equipment for the target product. The control terminal sends a discard command to the second target production equipment. The discard command is used to instruct the discarding of the target product. The second target production equipment is the next level equipment on the production line after the first target production equipment.
6. The method according to claim 5, characterized in that, The method further includes: The control unit determines whether the first target production equipment is the first production equipment and obtains a first determination result; If the first judgment result indicates that the first target production equipment is the first production equipment, then the target process parameters are supplemented to obtain the target sample process parameters of the target defect type. If the first judgment result indicates that the first target production equipment is not the first production equipment, then the real-time process parameters collected from the equipment before the first target production equipment on the production line are spliced with the target process parameters and a completion operation is performed to obtain the target sample process parameters of the target defect type.
7. A collaborative optimization system for intelligent production, characterized in that, The system includes: The data acquisition module is used to collect historical process parameters of multiple production equipment for the target product and historical quality inspection results of the quality inspection equipment for the target product. The classification module is used to classify the historical process parameters of multiple production equipment according to the historical quality inspection results to obtain a sample parameter set, which includes: sample process parameters of multiple qualified samples and sample process parameters of multiple defect types. The learning module is used to input the sample parameter set into the random forest tree learning model to learn and obtain the qualified parameter range set corresponding to the qualified sample and the defect parameter range set corresponding to different defect types. The data acquisition module is also used to acquire real-time process parameters of multiple production equipment for the target product; An optimization module is used to, if the real-time process parameters belong to the defect parameter range set, obtain adjustment data for each production equipment by comparing the real-time process parameters with the defect parameter range set; and adjust each production equipment according to the adjustment data. The step of obtaining the adjustment data for each production equipment by comparing the real-time process parameters with the defect parameter range set includes: calculating a comprehensive deviation value between the real-time process parameters and the corresponding defect parameter range in the defect parameter range set; obtaining a dynamic adjustment coefficient based on the comprehensive deviation value and the historical adjustment records of the production equipment; multiplying the comprehensive deviation value by the dynamic adjustment coefficient to obtain an initial adjustment value; and obtaining adjustment data based on the initial adjustment value using a PID algorithm. The formula for calculating the dynamic adjustment coefficient is as follows: Indicates the dynamic adjustment coefficient. Indicates the base coefficient. Indicates the historical effect correction factor. Indicates the device status factor; The initial adjustment value is calculated using the following formula: in, Indicates the initial adjustment value; This represents the overall deviation value. This represents the total number of process parameters that fall within the range of defective sub-process parameters. Indicates the first The influence weight of each process parameter; Indicates the first The single-parameter deviation value of each process parameter. Indicates the first Real-time values of each process parameter Indicates the first The center value of the defect range corresponding to each process parameter Indicates the first The defect range half-width corresponding to each process parameter.
8. The system according to claim 7, characterized in that, The real-time process parameters include real-time sub-process parameters corresponding to each of the multiple production equipment, and the defect parameter range set includes the defect sub-process parameter ranges corresponding to each of the multiple production equipment. The real-time process parameters belong to the defect parameter range set, including: each real-time sub-process parameter within its corresponding defect sub-process parameter range.
9. The system according to claim 7, characterized in that, The optimization module is specifically used to obtain an adjustment signal for each production device based on the adjustment data, and send the corresponding adjustment signal to each production device to achieve adjustment of each production device.