A molding process monitoring and control method for polymer cable materials

Abstract

The present invention discloses a molding process monitoring and control method for polymer cable materials, and relates to the field of cable molding control technology. The method comprises: reading the molding process information of polymer cable materials, establishing a molding process chain of polymer cable materials; constructing a molding node monitoring factor network of polymer cable materials, performing real-time molding process monitoring, and obtaining multiple molding process monitoring blocks of polymer cable materials; performing abnormality detection, obtaining abnormal molding process nodes and molding node abnormality detection results, inputting a molding process abnormality decision model, obtaining a node abnormality decision plan, and performing abnormal feedback control. The present invention solves the technical problem in the prior art that the molding process control accuracy is affected by the complexity of the polymer cable manufacturing process and the variability of process data, and achieves the technical effect of improving the molding process control accuracy of polymer cable materials by real-time monitoring of key nodes and dynamic optimization of process plans.

Description

Technical Field

[0001] The invention relates to the technical field of cable molding control, and in particular to a molding process monitoring and control method for polymer cable materials. Background Art

[0002] Cables are used as a medium for power and signal transmission, and their quality is directly related to the stability and safety of the power system. In particular, polymer cable materials need to meet a series of performance indicators, such as electrical properties, mechanical properties, heat resistance, etc. Therefore, monitoring and controlling the entire process is the key to ensuring cable quality.

[0003] However, due to the complexity of the cable manufacturing process, monitoring data often exhibits characteristics such as nonlinearity and multivariable coupling, which increases the difficulty of data processing and thus affects the accuracy of process control. Summary of the invention

[0004] The present application provides a molding process monitoring and control method for polymer cable materials, which is used to solve the technical problem in the prior art that the molding process control accuracy is affected by the complexity of the polymer cable manufacturing process and the variability of process data.

[0005] The first aspect of the present application provides a molding process monitoring and control method for polymer cable materials, the method comprising: reading polymer cable material molding process information according to a polymer cable material molding data interaction component; establishing a polymer cable material molding process chain according to the polymer cable material molding process information, wherein the polymer cable material molding process chain includes H polymer cable material molding process nodes, H being a positive integer greater than 1; mining node monitoring factors of the polymer cable material molding process chain according to the polymer cable material molding data interaction component to construct a polymer cable material molding node monitoring factor network; based on the polymer cable material molding node monitoring factor network The network is combined with the polymer cable material molding monitoring component to perform real-time molding process monitoring of the polymer cable material, and multiple polymer cable material molding process monitoring blocks are obtained, and each polymer cable material molding process monitoring block has a corresponding identification molding process node; according to the molding process abnormality detection integrated channel, the multiple polymer cable material molding process monitoring blocks are subjected to abnormality detection to obtain abnormal molding process nodes and molding node abnormality detection results; the abnormal molding process nodes and the molding node abnormality detection results are input into the polymer cable material molding process abnormality decision model to obtain a node abnormality decision plan, and the abnormal molding process nodes are subjected to abnormal feedback control according to the node abnormality decision plan.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0007] The molding process monitoring and control method for polymer cable materials provided in the present application relates to the field of cable molding control technology. By reading the molding process information of polymer cable materials, a molding process chain of polymer cable materials is established, and by mining node monitoring factors, a molding node monitoring factor network of polymer cable materials is constructed to perform real-time molding process monitoring and anomaly detection. According to the obtained abnormal molding process nodes and molding node anomaly detection results, a node anomaly decision plan is generated to perform anomaly feedback control, which solves the technical problem in the prior art that the molding process control accuracy is affected by the complexity of the polymer cable manufacturing process and the variability of process data, and realizes the technical effect of real-time monitoring of key nodes, dynamic optimization of process plans, and improving the molding process control accuracy of polymer cable materials. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0009] Figure 1 A schematic flow chart of a molding process monitoring and control method for polymer cable materials provided in an embodiment of the present application;

[0010] Figure 2 A schematic diagram of a process for constructing a polymer cable material molding node monitoring factor network in a molding process monitoring and control method for polymer cable materials provided in an embodiment of the present application;

[0011] Figure 3 A schematic flow chart of obtaining abnormal molding process nodes and molding node abnormality detection results in a molding process monitoring and control method for polymer cable materials provided in an embodiment of the present application. DETAILED DESCRIPTION

[0012] The present application provides a molding process monitoring and control method for polymer cable materials, which is used to solve the technical problem in the prior art that the molding process control accuracy is affected by the complexity of the polymer cable manufacturing process and the variability of process data.

[0013] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0014] It should be noted that the terms "first", "second", etc. in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices.

[0015] Embodiment 1

[0016] like Figure 1 As shown, the present application provides a molding process monitoring and control method for polymer cable materials, the method comprising:

[0017] P10: Read the polymer cable material molding process information according to the polymer cable material molding data interaction component;

[0018] Optionally, the polymer cable material molding process information is read through the polymer cable material molding data interaction component, wherein the polymer cable material molding data interaction component is a module for collecting, transmitting and processing data generated during the polymer cable material molding process, which may be a hardware device, such as a sensor network, a data acquisition card, etc., or a software service, such as a database, an API interface, etc., for capturing real-time data during the molding process. The polymer cable material molding process information refers to process parameters related to the polymer cable material molding process, including temperature, pressure, speed, material type, formula ratio, equipment status, etc., which can be used for subsequent analysis, monitoring and control.

[0019] P20: establishing a polymer cable material molding process chain according to the polymer cable material molding process information, wherein the polymer cable material molding process chain includes H polymer cable material molding process nodes, where H is a positive integer greater than 1;

[0020] Specifically, according to the polymer cable material molding process information, a polymer cable material molding process chain is established. First, based on the polymer cable material molding process, the key steps or operations in the molding process are identified, and multiple key process nodes are determined, that is, H polymer cable material molding process nodes are determined, where H is a positive integer greater than 1. Further, the H polymer cable material molding process nodes are arranged in the logical order of the polymer cable material in the molding process to form an orderly process chain, that is, the polymer cable material molding process chain, which can provide clear guidance and basis for subsequent node monitoring, abnormality detection and feedback control.

[0021] P30: mining node monitoring factors of the polymer cable material molding process chain according to the polymer cable material molding data interaction component, and constructing a polymer cable material molding node monitoring factor network;

[0022] Further, such as Figure 2 As shown, step P30 of the embodiment of the present application also includes:

[0023] P31: collecting node monitoring records of the polymer cable material molding data interaction component according to the polymer cable material molding process chain to obtain H molding node monitoring event record sets;

[0024] P32: Perform data cleaning according to the H forming node monitoring event record sets to obtain H forming node monitoring event data sets;

[0025] P33: performing node monitoring factor analysis according to the H formed node monitoring event data sets, and constructing H formed node monitoring factor branch networks;

[0026] P34: Integrate the H molding node monitoring factor branch networks to generate the polymer cable material molding node monitoring factor network.

[0027] It should be understood that according to the polymer cable material molding data interaction component, the node monitoring factor mining of the polymer cable material molding process chain is carried out, that is, the key monitoring factors are identified to construct a polymer cable material molding node monitoring factor network.

[0028] Specifically, according to the polymer cable material molding process chain, the polymer cable material molding data interaction component is used to collect monitoring records of each molding node to generate a monitoring event record set containing H molding nodes, that is, to obtain the monitoring event record of each node in the process chain. Further, the H molding node monitoring event record set is cleaned, including data deduplication, outlier processing, missing value filling, etc., to obtain a high-quality H molding node monitoring event data set.

[0029] Furthermore, the H formed node monitoring event data sets are deeply analyzed, and the key monitoring factors of each node are identified through data mining techniques, such as association rule learning and cluster analysis. Based on the key monitoring factors, H formed node monitoring factor branch networks are constructed, and each branch network represents a set of monitoring factors of a node and their mutual relationships.

[0030] Furthermore, the H molding node monitoring factor branch networks are integrated to form a global polymer cable material molding node monitoring factor network, which can comprehensively reflect the monitoring status of each node in the molding process chain and provide data support for subsequent real-time monitoring and abnormality detection.

[0031] Furthermore, step P33 of the embodiment of the present application also includes:

[0032] P33-1: extracting the hth forming node monitoring event data set according to the H forming node monitoring event data sets, where h is a positive integer and h belongs to H;

[0033] P33-2: Perform monitoring element feature recognition based on the h-th forming node monitoring event data set to obtain a sample monitoring element set;

[0034] P33-3: Perform support analysis based on the sample monitoring element set to determine the support of multiple monitoring elements;

[0035] P33-4: Calculate the confidence of the sample monitoring element set according to the multiple monitoring element support degrees to obtain multiple monitoring element confidence degrees;

[0036] P33-5: Based on the monitoring element confidence constraint, the confidence selection of the sample monitoring element set is performed in combination with the multiple monitoring element confidences to determine a confident sample monitoring element set that satisfies the monitoring element confidence constraint;

[0037] P33-6: Perform decoupling optimization according to the confidence sample monitoring element set to obtain an optimized sample monitoring element set;

[0038] P33-7: Based on the optimized sample monitoring element set, construct the h-th formed node monitoring factor branch network, and add the h-th formed node monitoring factor branch network to the H formed node monitoring factor branch networks.

[0039] Specifically, the h-th forming node monitoring event data set is randomly extracted from the H forming node monitoring event data sets, where h is a positive integer and h is less than or equal to H. Further, a feature recognition algorithm, such as principal component analysis, feature selection, etc., is used to process the monitoring event data set of the h-th forming node, identify characteristic elements related to the monitoring factors, and obtain a sample monitoring element set.

[0040] Furthermore, support analysis is performed on each element in the sample monitoring element set, that is, the number of times the element appears in the data set is calculated to obtain multiple monitoring element support degrees. Furthermore, the confidence of the sample monitoring element set is calculated using the multiple monitoring element support degrees, and the confidence degree is the ratio of the support degree to the total number of elements.

[0041] Furthermore, according to actual applications or business requirements, monitoring element confidence constraints are set, and combined with the multiple monitoring element confidences, elements that meet the monitoring element confidence constraints are screened out from the sample monitoring element set to form a confident sample monitoring element set. Furthermore, decoupling optimization algorithms, such as principal component analysis and independent component analysis, are used to process the confident sample monitoring element set, remove redundant elements or elements with strong correlation, and obtain an optimized sample monitoring element set to improve the accuracy and efficiency of the monitoring factor network.

[0042] Finally, according to the optimized sample monitoring element set, multiple sample data sets are collected, and network construction algorithms, such as neural networks and graph theory algorithms, are used to train and construct the monitoring factor branch network of the hth forming node. Similarly, the monitoring factor branch networks of other forming nodes are constructed and added to the H forming node monitoring factor branch networks in turn to form a complete polymer cable material forming node monitoring factor network.

[0043] Furthermore, step P33-6 of the embodiment of the present application also includes:

[0044] P33-61: traverse the confidence sample monitoring element set to extract the first confidence sample monitoring element and the second confidence sample monitoring element;

[0045] P33-62: Perform correlation analysis based on the first confidence sample monitoring element and the second confidence sample monitoring element to obtain a monitoring element correlation coupling index;

[0046] P33-63: Determine whether the monitoring element association coupling index is less than a predetermined association coupling index;

[0047] P33-64: If the monitoring element association coupling index is less than the predetermined association coupling index, the first confidence sample monitoring element and the second confidence sample monitoring element are added to the optimized sample monitoring element set.

[0048] Optionally, the confidence sample monitoring element set is traversed, and two elements are extracted sequentially or randomly as the first confidence sample monitoring element and the second confidence sample monitoring element, and a correlation analysis is performed on them. For example, a correlation analysis algorithm, such as the Pearson correlation coefficient, the Spearman rank correlation coefficient, the mutual information, etc., can be used to calculate the degree of correlation between the first confidence sample monitoring element and the second confidence sample monitoring element to obtain the monitoring element correlation coupling index, which reflects the correlation or dependence between the two elements in the data set.

[0049] Furthermore, according to actual application requirements, a predetermined correlation coupling index is set, and it is determined whether the monitoring element correlation coupling index is less than the predetermined correlation coupling index, thereby determining whether the two confidence sample monitoring elements are too correlated or coupled. If the monitoring element correlation coupling index is less than the predetermined correlation coupling index, indicating that the correlation or coupling degree between the two confidence sample monitoring elements is low, the first confidence sample monitoring element and the second confidence sample monitoring element are simultaneously added to the optimized sample monitoring element set to improve the comprehensiveness of the sample data, thereby improving the accuracy and efficiency of the polymer cable material molding node monitoring factor network.

[0050] Furthermore, step P33-6 of the embodiment of the present application also includes:

[0051] P33-64a: If the monitoring element association coupling index is greater than / equal to the predetermined association coupling index, the first confidence sample monitoring element and the second confidence sample monitoring element are fused to generate a decoupled confidence sample monitoring element, and the decoupled confidence sample monitoring element is added to the optimized sample monitoring element set.

[0052] In a possible embodiment of the present application, if the monitoring element association coupling index is greater than or equal to the predetermined association coupling index, indicating that the correlation or coupling degree between the two confidence sample monitoring elements is high, the first confidence sample monitoring element and the second confidence sample monitoring element are fused by weighted averaging or the like to generate a new decoupled confidence sample monitoring element, and the new decoupled confidence sample monitoring element is added to the optimized sample monitoring element set to replace the original two associated elements, thereby reducing the redundancy and complexity of highly correlated elements in the polymer cable material molding node monitoring factor network while retaining key information.

[0053] P40: Based on the polymer cable material molding node monitoring factor network, combined with the polymer cable material molding monitoring component, real-time molding process monitoring of the polymer cable material is performed to obtain multiple polymer cable material molding process monitoring blocks, and each polymer cable material molding process monitoring block has a corresponding identification molding process node;

[0054] Specifically, through the polymer cable material molding node monitoring factor network, combined with the polymer cable material molding monitoring component, the molding process data of the polymer cable material is collected and processed in real time, and during the monitoring process, the molding process of the polymer cable material is divided into multiple monitoring blocks, each polymer cable material molding process monitoring block corresponds to a specific molding process node, that is, a key link or potential failure point in the molding process, and each monitoring block is assigned a unique node identifier for identifying its corresponding molding process node for rapid positioning in subsequent abnormality detection or fault diagnosis.

[0055] P50: performing abnormality detection on the plurality of polymer cable material molding process monitoring blocks according to the molding process abnormality detection integrated channel, and obtaining abnormal molding process nodes and molding node abnormality detection results;

[0056] Further, such as Figure 3 As shown, step P50 of the embodiment of the present application also includes:

[0057] P51: Perform deep reinforcement learning based on the polymer cable material molding node monitoring factor network to build the molding process abnormality detection integrated channel;

[0058] P52: extracting a first polymer cable material molding process monitoring block according to the plurality of polymer cable material molding process monitoring blocks, and matching a first corresponding identification molding process node, wherein the first polymer cable material molding process monitoring block includes any one polymer cable material molding process monitoring block among the plurality of polymer cable material molding process monitoring blocks;

[0059] P53: activating a matching molding node abnormality detection branch in the molding process abnormality detection integrated channel according to the first corresponding identification molding process node;

[0060] P54: Input the first polymer cable material molding process monitoring block into the matching molding node abnormality detection branch to obtain the first monitoring block node abnormality identification data and the first monitoring block node abnormality evaluation index;

[0061] P55: input the first corresponding identification molding process node into the polymer cable material molding node abnormal constraint table to match the first molding node abnormal constraint;

[0062] P56: Determine whether the first monitoring block node abnormality assessment index satisfies the first forming node abnormality constraint;

[0063] P57: If the first monitoring block node abnormality assessment index does not meet the first molding node abnormality constraint, the first monitoring block node abnormality identification data is added to the molding node abnormality detection result, and the first corresponding identification molding process node is added to the abnormal molding process node.

[0064] It should be understood that the molding process anomaly detection integrated channel is used to perform anomaly detection on the multiple polymer cable material molding process monitoring blocks respectively. Specifically, firstly, deep reinforcement learning is performed on the polymer cable material molding node monitoring factor network, that is, multiple anomaly detection branches are trained for multiple molding node monitoring factor branch networks, and thus the molding process anomaly detection integrated channel is formed.

[0065] Further, from the plurality of polymer cable material molding process monitoring blocks, any one polymer cable material molding process monitoring block is randomly extracted as the first polymer cable material molding process monitoring block, and its corresponding first corresponding identification molding process node is determined. Further, according to the first corresponding identification molding process node, the corresponding matching molding node abnormality detection branch in the molding process abnormality detection integrated channel is activated, and the first polymer cable material molding process monitoring block is input into the activated matching molding node abnormality detection branch, and the first monitoring block node abnormality identification data and the first monitoring block node abnormality evaluation index are obtained through abnormality detection.

[0066] Furthermore, based on the historical experience in polymer cable material molding, the abnormal threshold of each molding node is obtained, that is, the maximum allowable value of the parameter abnormality, an abnormal constraint table of the polymer cable material molding node is constructed, and the first corresponding identified molding process node is input into the abnormal constraint table of the polymer cable material molding node to match the first molding node abnormal constraint, that is, the allowable range of the parameter abnormality of the first molding node.

[0067] Further, it is determined whether the first monitoring block node abnormality assessment index satisfies the first forming node abnormality constraint. If not, it indicates that the node is abnormal. Then, the first monitoring block node abnormality identification data is added to the forming node abnormality detection result, and the first corresponding identification forming process node is added to the abnormal forming process node as a reference for subsequent abnormality correction.

[0068] Furthermore, step P51 of the embodiment of the present application further includes:

[0069] P51-1: Collect abnormal detection records according to the h-th forming node monitoring factor branch network in the polymer cable material forming node monitoring factor network, and obtain the h-th forming node monitoring block sample set, as well as the h-th monitoring block node abnormality identification sample set and the h-th monitoring block node abnormality assessment sample set corresponding to the h-th forming node monitoring block sample set;

[0070] P51-2: Perform supervised learning on the h-th forming node monitoring block sample set and the h-th monitoring block node abnormality identification sample set according to the deep reinforcement learning function, and construct an h-th forming node abnormality identification model that meets the deep reinforcement learning constraints;

[0071] P51-3: Perform supervised learning on the h-th monitoring block node abnormality identification sample set and the h-th monitoring block node abnormality assessment sample set according to the deep reinforcement learning function, and construct an h-th forming node abnormality assessment model that meets the deep reinforcement learning constraints;

[0072] P51-4: Connect the h-th forming node abnormality recognition model and the h-th forming node abnormality assessment model to generate an h-th forming node abnormality detection branch;

[0073] P51-5: Add the h-th molding node abnormality detection branch to the molding process abnormality detection integrated channel.

[0074] Wherein, the deep reinforcement learning function is:

[0075] ;

[0076] Among them, ILI represents the deep reinforcement learning index, LIA represents the test output accuracy coefficient, LIL represents the test output error coefficient, IAW represents the predetermined output accuracy weight, ILW represents the predetermined output error weight, and the sum of IAW and ILW is 1.

[0077] Optionally, anomaly detection records are collected for the h-th forming node monitoring factor branch network within the polymer cable material forming node monitoring factor network, and based on the historical data of the branch network, the h-th forming node monitoring block sample set and the corresponding h-th monitoring block node anomaly identification sample set and h-th monitoring block node anomaly evaluation sample set are collected.

[0078] Furthermore, a deep reinforcement learning function is used to perform supervised learning on the h-th forming node monitoring block sample set and the h-th monitoring block node abnormality identification sample set, wherein the deep reinforcement learning function is: ; Among them, ILI represents the deep reinforcement learning index, LIA represents the test output accuracy coefficient, LIL represents the test output error coefficient, IAW represents the predetermined output accuracy weight, ILW represents the predetermined output error weight, and the sum of IAW and ILW is 1. Through multiple iterative training, an h-th forming node abnormality recognition model that meets the deep reinforcement learning constraints is obtained, which can be used to identify the abnormal state of the corresponding monitoring area.

[0079] Similarly, the deep reinforcement learning function is used to perform supervised learning on the h-th monitoring block node anomaly identification sample set and the h-th monitoring block node anomaly assessment sample set, and an h-th forming node anomaly assessment model that meets the deep reinforcement learning constraints is constructed, which can be used to quantitatively evaluate the identified anomalies and provide an indicator of the degree or severity of the anomaly.

[0080] Furthermore, the h-th molding node abnormality recognition model and the h-th molding node abnormality assessment model are connected to form a complete h-th molding node abnormality detection branch. Similarly, the abnormality detection branches of other molding nodes are trained with reference to the above method, that is, the H molding node abnormality detection branches corresponding to the H polymer cable material molding process nodes, and all are added to the molding process abnormality detection integrated channel for molding process abnormality detection.

[0081] P60: Input the abnormal molding process node and the abnormal molding node detection result into the polymer cable material molding process abnormality decision model to obtain a node abnormality decision plan, and perform abnormal feedback control on the abnormal molding process node according to the node abnormality decision plan.

[0082] It should be understood that the abnormal molding process nodes and molding node abnormal detection results are passed as input to the polymer cable material molding process abnormality decision model, and the polymer cable material molding process abnormality decision model makes process optimization decisions and generates a node abnormality decision plan.

[0083] Among them, the polymer cable material molding process abnormality decision model can be constructed by collecting sample process abnormality optimization data and combining machine learning algorithms or other decision support technologies. In addition, the node abnormality decision plan may involve the adjustment of process parameters, the maintenance or replacement of equipment, the improvement of operation procedures, etc. According to the node abnormality decision plan, abnormal feedback control is implemented on the abnormal molding process nodes to correct or adjust the process parameters, prevent the problem from further deteriorating, and ensure the stability of the molding process and product quality.

[0084] In summary, the embodiments of the present application have at least the following technical effects:

[0085] The present application establishes a polymer cable material molding process chain by reading the polymer cable material molding process information, and constructs a polymer cable material molding node monitoring factor network through node monitoring factor mining, so as to perform real-time molding process monitoring and anomaly detection, and generate a node abnormality decision plan for abnormal feedback control based on the obtained abnormal molding process nodes and molding node abnormality detection results.

[0086] The technical effect of real-time monitoring of key nodes, dynamic optimization of process plans and improvement of the control accuracy of the molding process of polymer cable materials has been achieved.

[0087] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0088] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

[0089] This specification and the drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. A molding process monitoring and control method for polymer cable materials, characterized in that: The method comprises: According to the polymer cable material molding data interaction component, the polymer cable material molding process information is read; According to the polymer cable material molding process information, a polymer cable material molding process chain is established, wherein the polymer cable material molding process chain includes H polymer cable material molding process nodes, and H is a positive integer greater than 1; According to the polymer cable material molding data interaction component, the node monitoring factor mining of the polymer cable material molding process chain is carried out to construct a polymer cable material molding node monitoring factor network; Based on the polymer cable material molding node monitoring factor network, the polymer cable material molding monitoring component is combined to perform real-time molding process monitoring of the polymer cable material, and a plurality of polymer cable material molding process monitoring blocks are obtained, and each polymer cable material molding process monitoring block has a corresponding identification molding process node; Performing abnormality detection on the plurality of polymer cable material molding process monitoring blocks according to the molding process abnormality detection integrated channel to obtain abnormal molding process nodes and molding node abnormality detection results; Inputting the abnormal molding process node and the abnormal molding node detection result into the polymer cable material molding process abnormality decision model to obtain a node abnormality decision plan, and performing abnormal feedback control on the abnormal molding process node according to the node abnormality decision plan; Among them, according to the polymer cable material molding data interaction component, the node monitoring factor mining of the polymer cable material molding process chain is carried out to construct a polymer cable material molding node monitoring factor network, including: According to the polymer cable material molding process chain, the polymer cable material molding data interaction component performs node monitoring record collection to obtain H molding node monitoring event record sets; Perform data cleaning according to the H forming node monitoring event record sets to obtain H forming node monitoring event data sets; Perform node monitoring factor analysis based on the H formed node monitoring event data sets to construct H formed node monitoring factor branch networks; Integrate the H molding node monitoring factor branch networks to generate the polymer cable material molding node monitoring factor network; Node monitoring factor analysis is performed according to the H formed node monitoring event data sets, and H formed node monitoring factor branch networks are constructed, including: Extracting an h-th forming node monitoring event data set according to the H forming node monitoring event data sets, wherein h is a positive integer and h belongs to H; Performing monitoring element feature recognition according to the h-th forming node monitoring event data set to obtain a sample monitoring element set; Perform support analysis based on the sample monitoring element set to determine the support of multiple monitoring elements; Calculating the confidence of the sample monitoring element set according to the multiple monitoring element support degrees to obtain multiple monitoring element confidence degrees; Based on the monitoring element confidence constraint, the sample monitoring element set is selected in confidence in combination with the multiple monitoring element confidences to determine a confident sample monitoring element set that satisfies the monitoring element confidence constraint; Performing decoupling optimization according to the confidence sample monitoring element set to obtain an optimized sample monitoring element set; According to the optimized sample monitoring element set, constructing the h-th formed node monitoring factor branch network, and adding the h-th formed node monitoring factor branch network to the H formed node monitoring factor branch networks; Performing decoupling optimization according to the confidence sample monitoring element set to obtain an optimized sample monitoring element set includes: Traversing the confidence sample monitoring element set to extract a first confidence sample monitoring element and a second confidence sample monitoring element; Performing correlation analysis based on the first confidence sample monitoring element and the second confidence sample monitoring element to obtain a monitoring element correlation coupling index; Determining whether the monitoring element association coupling index is less than a predetermined association coupling index; If the monitoring element association coupling index is less than the predetermined association coupling index, adding the first confidence sample monitoring element and the second confidence sample monitoring element to the optimized sample monitoring element set; If the monitoring element association coupling index is greater than / equal to the predetermined association coupling index, fusing the first confidence sample monitoring element and the second confidence sample monitoring element to generate a decoupled confidence sample monitoring element, and adding the decoupled confidence sample monitoring element to the optimized sample monitoring element set; According to the molding process abnormality detection integrated channel, the plurality of polymer cable material molding process monitoring blocks are subjected to abnormality detection to obtain abnormal molding process nodes and molding node abnormality detection results, including: Conduct deep reinforcement learning based on the polymer cable material molding node monitoring factor network to build the molding process abnormality detection integrated channel; Extracting a first polymer cable material molding process monitoring block according to the plurality of polymer cable material molding process monitoring blocks, and matching a first corresponding identification molding process node, wherein the first polymer cable material molding process monitoring block includes any one polymer cable material molding process monitoring block among the plurality of polymer cable material molding process monitoring blocks; activating a matching molding node anomaly detection branch in the molding process anomaly detection integrated channel according to the first corresponding identification molding process node; Input the first polymer cable material molding process monitoring block into the matching molding node abnormality detection branch to obtain the first monitoring block node abnormality identification data and the first monitoring block node abnormality evaluation index; Inputting the first corresponding identification molding process node into the polymer cable material molding node abnormal constraint table to match the first molding node abnormal constraint; Determine whether the first monitoring block node abnormality assessment index satisfies the first forming node abnormality constraint; If the first monitoring block node abnormality evaluation index does not satisfy the first forming node abnormality constraint, the first monitoring block node abnormality identification data is added to the forming node abnormality detection result, and the first corresponding identification forming process node is added to the abnormal forming process node; According to the polymer cable material molding node monitoring factor network, deep reinforcement learning is performed to build the molding process abnormality detection integrated channel, including: According to the h-th forming node monitoring factor branch network in the polymer cable material forming node monitoring factor network, abnormality detection records are collected to obtain the h-th forming node monitoring block sample set, and the h-th monitoring block node abnormality identification sample set and the h-th monitoring block node abnormality evaluation sample set corresponding to the h-th forming node monitoring block sample set; Performing supervised learning on the h-th forming node monitoring block sample set and the h-th monitoring block node abnormality recognition sample set according to the deep reinforcement learning function, and constructing an h-th forming node abnormality recognition model that meets the deep reinforcement learning constraints; Performing supervised learning on the h-th monitoring block node abnormality identification sample set and the h-th monitoring block node abnormality assessment sample set according to the deep reinforcement learning function, and constructing the h-th forming node abnormality assessment model that meets the deep reinforcement learning constraints; Connecting the h-th forming node anomaly recognition model and the h-th forming node anomaly assessment model to generate an h-th forming node anomaly detection branch; Adding the hth molding node abnormality detection branch to the molding process abnormality detection integrated channel; The deep reinforcement learning function is: ； Among them, ILI represents the deep reinforcement learning index, LIA represents the test output accuracy coefficient, LIL represents the test output error coefficient, IAW represents the predetermined output accuracy weight, ILW represents the predetermined output error weight, and the sum of IAW and ILW is 1.

Images