CN117788203A - An improved efficient production and preparation method for cross-linked polyethylene insulated power cables - Google Patents
An improved efficient production and preparation method for cross-linked polyethylene insulated power cables Download PDFInfo
- Publication number
- CN117788203A CN117788203A CN202410217216.7A CN202410217216A CN117788203A CN 117788203 A CN117788203 A CN 117788203A CN 202410217216 A CN202410217216 A CN 202410217216A CN 117788203 A CN117788203 A CN 117788203A
- Authority
- CN
- China
- Prior art keywords
- improvement
- production
- factor
- preparation
- crosslinked polyethylene
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Processes Specially Adapted For Manufacturing Cables (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses an improved high-efficiency production and preparation method of a crosslinked polyethylene insulated power cable, which relates to the technical field of cable preparation and comprises the following steps: acquiring the production and preparation process of the crosslinked polyethylene insulated power cable, extracting working condition data in historical production and preparation, and clustering the working condition data by extracting the key performance of crosslinked polyethylene insulation; constructing an improvement factor subset for graph representation, obtaining a corresponding improvement factor sub-graph to be mapped to a related knowledge graph, obtaining local improvement measures for the mapped improvement factor sub-graph by utilizing graph convolution, splicing the improvement factor sub-graphs, and obtaining global coupling association through a graph convolution network; and aggregation is carried out through global coupling association and local improvement measures, so that full-flow collaborative improvement is carried out. The invention starts from the key performance of the crosslinked polyethylene insulating material, obtains the targeted improvement measures in the production and preparation process, reduces the defects of insulation eccentricity and the like, improves the production and preparation efficiency, and better meets the requirements of actual production.
Description
Technical Field
The invention relates to the technical field of cable preparation, in particular to an improved efficient production and preparation method of a crosslinked polyethylene insulated power cable.
Background
The power cable is used as an important carrier for transmitting electric energy from a high-voltage transformer substation to a civil transformer substation, the performance requirement on the cable is continuously improved, and the life and the quality of the power cable influence the life of people. Loss of electric energy can occur in the transmission process, and in order to reduce the loss of the electric energy in the transmission process, the direct current resistance, the dielectric loss tangent value and the like of the conductor of the cable are required to be large. At present, urban space rationalization and utilization requirements are continuously improved, such as tunnel laying, direct burial laying and the like, and the design needs to consider that maintenance is convenient and fast. The quality and the service life of the power cable are improved, and the maintenance of the power cable can be greatly reduced.
The crosslinked polyethylene insulated power cable consists of 3 parts of conductors, insulating layers and protective layers, has good insulating property, and has the advantages of light structure, high medium strength, low loss, aging resistance, high temperature resistance level, large transmission energy, simple installation and maintenance, no fall limitation in laying and the like. Because the physical and chemical properties such as the crystalline phase structure and the crosslinking degree of the crosslinked polyethylene insulated power cable can influence the macroscopic mechanical and dielectric properties of the material, the crosslinked polyethylene insulation with excellent comprehensive properties can reduce the possibility of failure of the cable in the actual operation process. Therefore, how to optimize the production process of the crosslinked polyethylene insulated power cable to improve the production efficiency and the cable performance is a problem to be solved at present.
Disclosure of Invention
The invention aims to overcome the defects in the prior art, and provides an improved efficient production and preparation method of a crosslinked polyethylene insulated power cable.
The invention provides an improved high-efficiency production and preparation method of a crosslinked polyethylene insulated power cable, which comprises the following steps:
acquiring a production and preparation process of a crosslinked polyethylene insulated power cable, extracting working condition data in historical production and preparation, extracting key performance of crosslinked polyethylene insulation, and clustering the working condition data by utilizing the key performance;
extracting improvement factors of production and preparation from the key performance corresponding class clusters, constructing a subset of the improvement factors corresponding to the key performance for graph representation, and obtaining a corresponding improvement factor subgraph;
mapping the improvement factor subgraphs to related knowledge maps, obtaining local improvement measures for the mapped improvement factor subgraphs by utilizing graph convolution, splicing the improvement factor subgraphs, and obtaining global coupling association through graph convolution network representation learning;
and polymerizing the global coupling association and the local improvement measures to carry out the whole-flow collaborative improvement of the production and preparation process of the crosslinked polyethylene insulated power cable, and replacing the existing production and preparation process.
In this scheme, draw the key performance of crosslinked polyethylene insulation, utilize the key performance will operating mode data clusters, specifically is:
acquiring working condition data produced and prepared by the crosslinked polyethylene insulated power cable in a preset history period, preprocessing the working condition data, pre-classifying the preprocessed working condition data according to a production and preparation flow, and adding a production and preparation procedure label for the working condition data;
dividing and obtaining rheological property, degassing property, scorch resistance and insulating property corresponding to the crosslinked polyethylene insulating material according to the production and preparation process, taking the key property as an initial clustering center, and carrying out preferential matching by utilizing working condition data of the production and preparation process labels corresponding to the key properties;
judging the Euclidean distance between the working condition data and the initial clustering center, attributing the working condition data to the initial clustering center closest to the working condition data to generate a class cluster, updating the clustering center through iterative clustering, and generating a final clustering result of the working condition data.
In the scheme, the improvement factors of production and preparation are extracted from the corresponding class clusters of key performance, and specifically:
acquiring a cluster corresponding to key performance to generate a working condition data subset, matching the working condition data subset with an influence index of the crosslinked polyethylene insulated power cable, and evaluating the working condition data subset according to the influence index;
according to the evaluation result, sorting and setting the extraction quantity grades of the improvement factors of the data subsets under different working conditions, wherein the larger the grade is, the more the quantity of the extracted improvement factors is, the key performance is used as a retrieval tag, and the improvement examples related to the key performance are obtained by utilizing a big data means;
acquiring working condition characteristics of different working condition data subsets, and pre-screening the retrieved improved examples according to the working condition characteristics;
extracting improvement factors involved in a pre-screening improvement example to construct an improvement factor set, carrying out principal component analysis on the improvement factor set, obtaining the contribution degree of each improvement factor, and obtaining the coupling degree of each improvement factor and key performance according to the contribution degree;
and sorting the improvement factors according to the coupling degree, and acquiring a preset number of improvement factors based on the extraction number level.
In the scheme, an improvement factor subset corresponding to key performance is constructed for graph representation, and a corresponding improvement factor subgraph is obtained, specifically:
obtaining an improvement factor subset corresponding to key performance, abstracting the improvement factor subset into an undirected graph, and defining the improvement factor as a node in the undirected graph;
calculating mutual information of different improvement factors in the improvement factor subset, calculating a maximum information coefficient between the improvement factors according to the mutual information, comparing the maximum information coefficient with a preset threshold value, and defining an edge structure between nodes in the undirected graph according to a comparison result;
and obtaining a corresponding improvement factor subgraph based on the undirected graph structure, extracting the maximum information coefficient between improvement factor nodes to construct a symmetrical matrix, replacing the maximum information coefficient smaller than a preset threshold value in the symmetrical matrix with 0, and obtaining an adjacent matrix of the improvement factor subgraph.
In the scheme, the improvement factor subgraph is mapped to the related knowledge graph, and the mapped improvement factor subgraph is rolled by the graph to obtain local improvement measures, which are specifically as follows:
acquiring a related knowledge graph of a production preparation improvement measure of the crosslinked polyethylene insulated power cable, mapping the improvement factor subgraph to the related knowledge graph, and acquiring the relation between an improvement factor node and an improvement measure entity according to the historical interaction condition of the improvement factor and the improvement measure;
performing characterization learning on the relation between the improvement factor node and the improvement measure entity by using an attention mechanism in a graph rolling network, distributing attention scores to different relations, mapping the attention scores to a low-dimensional space to calculate Euclidean distance calculation relation similarity, and reducing relation redundancy by minimizing the relation similarity;
updating the adjacency matrix of the improvement factor subgraph according to the relation between the improvement factor nodes and the improvement measure entity, and carrying out weighted aggregation on neighbor nodes in the adjacency matrix by using attention scores corresponding to the relation to acquire embedded representation of the improvement factor;
and adding the embedded representations of the different layers of improvement factors to obtain the final representation of the improvement factors, and in addition, obtaining neighbor nodes of an improvement measure entity, obtaining the final representation of the improvement measure by utilizing neighbor aggregation, taking the final representation of the improvement factors and the improvement measure as an inner product, and obtaining the local improvement measure corresponding to the improvement factor subgraph.
According to the scheme, interest vectors of the improvement factors are obtained according to historical interaction conditions of the improvement factors and improvement measures, and the similarity between the improvement factors is calculated according to the interest vectors;
and obtaining similar improvement factors to expand interaction of improvement factors-improvement measures in the related knowledge graph, and constructing potential related improvement measure nodes of the improvement factor nodes.
In the scheme, the improved factor subgraphs are spliced, and global coupling association is obtained through graph rolling network representation learning, specifically:
acquiring time sequence characteristics among the improvement factor subsets corresponding to the key performance according to the sequence of the production and preparation flow, and splicing the improvement factor subsets based on the production and preparation flow to generate an improvement factor global graph;
constructing an encoder by using a multi-layer graph convolutional neural network, learning and representing an improvement factor global graph by using the encoder, constructing an adjacent matrix to acquire the spatial relationship of adjacent improvement factors in corresponding production and preparation, and distributing different weights to neighbor nodes by using the combination of average pooling and maximum pooling;
the method comprises the steps of utilizing weights to aggregate neighbor nodes in an improvement factor global graph, updating the representation of the improvement factor nodes, importing the updated representation of the improvement factor nodes into a decoder, utilizing an RNN (network node network) to conduct feature recombination, and outputting spatial features;
and fusing the time sequence features and the space features to obtain global coupling association among all factor subgraphs.
In this scheme, the global coupling association and the local improvement measure are aggregated, specifically:
obtaining a preselected subset of local improvement measures, screening among the different subsets of local improvement measures using the global coupling association,
obtaining impact index evaluation results of different key performance working condition data subsets, obtaining key performance with highest improvement priority according to the evaluation results, and selecting local improvement measures corresponding to the highest inner product from the corresponding local improvement measure subsets;
matching the selected local improvement measures as a starting point in the residual local improvement measure subset by using a global coupling association construction constraint condition, and obtaining local improvement measures which accord with a preset similarity threshold in the residual local improvement measure subset by using similarity calculation;
and polymerizing the matched local improvement measures to generate the improved production and preparation process of the crosslinked polyethylene insulated power cable.
The invention discloses an improved high-efficiency production and preparation method of a crosslinked polyethylene insulated power cable, which comprises the following steps: acquiring the production and preparation process of the crosslinked polyethylene insulated power cable, extracting working condition data in historical production and preparation, and clustering the working condition data by extracting the key performance of crosslinked polyethylene insulation; constructing an improvement factor subset corresponding to the key performance for graph representation, and obtaining a corresponding improvement factor subgraph; mapping the improvement factor subgraphs to related knowledge maps, obtaining local improvement measures for the mapped improvement factor subgraphs by utilizing graph convolution, splicing the improvement factor subgraphs, and obtaining global coupling association through a graph convolution network; and aggregation is carried out through global coupling association and local improvement measures, so that full-flow collaborative improvement is carried out. The invention starts from the key performance of the crosslinked polyethylene insulating material, obtains the targeted improvement measures in the production and preparation process, reduces the defects of insulation eccentricity and the like, improves the production and preparation efficiency, and better meets the requirements of actual production.
Drawings
In order to more clearly illustrate the technical solutions of embodiments or examples of the present invention, the drawings that are required to be used in the embodiments or examples of the present invention will be briefly described below, and it is apparent that the drawings in the following description are only some embodiments of the present invention, and other drawings may be obtained according to the drawings without inventive efforts for those skilled in the art.
FIG. 1 shows a flow chart of a method of efficiently producing and preparing an improved crosslinked polyethylene insulated power cable of the present invention;
FIG. 2 shows a flow chart of the present invention for obtaining local improvement measures;
FIG. 3 illustrates a flow chart for acquiring global coupling associations in accordance with the present invention;
fig. 4 shows a block diagram of an improved high efficiency production and preparation system for crosslinked polyethylene insulated power cables according to the invention.
The achievement of the objects, functional features and advantages of the present invention will be further described with reference to the accompanying drawings, in conjunction with the embodiments.
Detailed Description
In order that the above-recited objects, features and advantages of the present invention will be more clearly understood, a more particular description of the invention will be rendered by reference to the appended drawings and appended detailed description. It should be noted that, in the case of no conflict, the embodiments of the present application and the features in the embodiments may be combined with each other.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, however, the present invention may be practiced in other ways than those described herein, and therefore the scope of the present invention is not limited to the specific embodiments disclosed below.
FIG. 1 shows a flow chart of a method of efficiently producing and preparing an improved crosslinked polyethylene insulated power cable of the present invention;
as shown in fig. 1, a first aspect of the present invention provides an improved efficient production and preparation method of a crosslinked polyethylene insulated power cable, comprising:
s102, acquiring a production and preparation process of a crosslinked polyethylene insulated power cable, extracting working condition data in historical production and preparation, extracting key performance of crosslinked polyethylene insulation, and clustering the working condition data by utilizing the key performance;
s104, extracting improvement factors of production and preparation from the key performance corresponding class clusters, constructing a subset of the improvement factors corresponding to the key performance for graph representation, and obtaining a corresponding improvement factor subgraph;
s106, mapping the improvement factor subgraphs to related knowledge maps, obtaining local improvement measures for the mapped improvement factor subgraphs by utilizing graph convolution, splicing the improvement factor subgraphs, and obtaining global coupling association through graph convolution network representation learning;
and S108, polymerizing the global coupling association and the local improvement measures, and carrying out the whole-flow collaborative improvement of the production and preparation process of the crosslinked polyethylene insulated power cable to replace the existing production and preparation process.
The production process of the crosslinked polyethylene insulated power cable is approximately as follows: ethylene is polymerized into a low-density polyethylene base material by initiating free radical reaction by an initiator under the conditions of high temperature and high pressure, and the base material is introduced with a crosslinking agent and an antioxidant in a compounding process to form a crosslinked polyethylene insulating material; the insulating material is extruded and formed, and then is subjected to crosslinking reaction to form cable insulation, and the cable insulation is processed into a finished cable after removing crosslinking byproducts in the degassing process.
Acquiring working condition data produced and prepared by the crosslinked polyethylene insulated power cable in a preset history period, preprocessing the working condition data such as data cleaning, pre-classifying the preprocessed working condition data according to a production and preparation flow, and adding a production and preparation procedure label for the working condition data; dividing and obtaining rheological property, degassing property, scorch resistance and insulating property corresponding to the crosslinked polyethylene insulating material according to the production and preparation process, taking the key property as an initial clustering center, and carrying out preferential matching by utilizing working condition data of the production and preparation process labels corresponding to the key properties; judging the Euclidean distance between the working condition data and the initial clustering center, attributing the working condition data to the initial clustering center closest to the working condition data to generate a class cluster, updating the clustering center in iterative clustering by using average taking operation, and generating a final clustering result of the working condition data.
The method comprises the steps of obtaining a cluster corresponding to key performance to generate a working condition data subset, matching the working condition data subset with influence indexes such as forming quality, production efficiency, processability and impurity defect condition of the crosslinked polyethylene insulated power cable, and evaluating the working condition data subset according to the influence indexes; according to the evaluation result, sorting and setting the extraction quantity grades of the improvement factors of the data subsets under different working conditions, wherein the larger the grade is, the more the quantity of the extracted improvement factors is, the key performance is used as a retrieval tag, and the improvement examples related to the key performance are obtained by utilizing a big data means; acquiring working condition characteristics of different working condition data subsets, and pre-screening the retrieved improved examples according to the working condition characteristics; extracting improvement factors involved in pre-screening improvement examples to construct an improvement factor set, carrying out principal component analysis on the improvement factor set, and preferably calculating the correlation between the improvement factors through pearson correlation coefficients to construct the correlation between the improvement factorsCovariance matrix of (2)Obtaining covariance matrix->Characteristic value of +.>Obtaining feature vectors corresponding to the feature values, sorting the feature values, and calculating contribution degree by taking the feature vector with the largest feature value as a main component> P represents the total number of eigenvalues, +.>Characteristic values of the r-th and k-th characteristic vectors are respectively represented; obtaining the coupling degree of each improvement factor and key performance according to the contribution degree>；Wherein n represents the number of factors calculating the coupling, +.>The contribution degree of the ith and jth improvement factors is represented.
The improved factors are ordered according to the coupling degree, and a preset number of improved factors are obtained based on the extraction quantity grade, for example, the rheological property of the crosslinked polyethylene insulating material is mainly determined by the rheological property of the low-density polyethylene base material, and the predicted temperature, pressure, initiator, regulator and other various design and operation variables are taken as the improved factors to realize the regulation and control of the molecular chain structure. The degassing performance of the crosslinked polyethylene insulating material is closely related to that of the crosslinking agent, and an improvement factor is obtained based on the association of the crosslinking agent compounding and the aggregation state structure of the low-density polyethylene base material. The scorch resistance of the crosslinked polyethylene insulating material refers to the ability of inhibiting premature crosslinking and gel formation in the extrusion process, and the key of improving the scorch resistance is to improve the acting efficiency of an antioxidant, improve a cable insulating material compounding system and a compounding process and obtain improvement factors based on an antioxidant and crosslinking coupling mechanism; the key indexes of the electrical insulation performance of the crosslinked polyethylene insulation comprise dielectric loss, dielectric constant, conductivity, breakdown field strength and the like, and improvement factors are obtained based on the association of the electrical branches and the cable insulation material chain structure.
It should be noted that, obtaining an improvement factor subset corresponding to the key performance, abstracting the improvement factor subset into an undirected graph, and defining the improvement factor as a node in the undirected graph; calculating mutual information of different improvement factors in the improvement factor subset, calculating a maximum information coefficient between the improvement factors according to the mutual information, comparing the maximum information coefficient with a preset threshold value, and defining an edge structure between nodes in the undirected graph according to a comparison result; and obtaining a corresponding improvement factor subgraph based on the undirected graph structure so as to reduce the calculated amount, extracting the maximum information coefficient between the improvement factor nodes to construct a symmetrical matrix, replacing the maximum information coefficient smaller than a preset threshold value in the symmetrical matrix with 0, and obtaining an adjacent matrix of the improvement factor subgraph.
Fig. 2 shows a flow chart of the inventive approach to achieving local improvement.
According to the embodiment of the invention, the improvement factor subgraph is mapped to the related knowledge graph, and the mapped improvement factor subgraph is rolled by the graph to obtain local improvement measures, which are specifically as follows:
s202, acquiring a related knowledge graph of a production preparation improvement measure of a crosslinked polyethylene insulated power cable, mapping the improvement factor subgraph to the related knowledge graph, and acquiring the relation between an improvement factor node and an improvement measure entity according to the historical interaction condition of the improvement factor and the improvement measure;
s204, performing characterization learning on the relation between the improvement factor node and the improvement measure entity by using an attention mechanism in the graph convolution network, distributing attention scores to different relations, mapping the attention scores to a low-dimensional space to calculate Euclidean distance calculation relation similarity, and reducing relation redundancy by minimizing the relation similarity;
s206, updating the adjacent matrix of the improvement factor subgraph according to the relation between the improvement factor node and the improvement measure entity, and carrying out weighted aggregation on neighbor nodes in the adjacent matrix by using attention scores corresponding to the relation to acquire an embedded representation of the improvement factor;
and S208, adding the embedded representations of the improvement factors of different layers to obtain the final representation of the improvement factors, and in addition, obtaining neighbor nodes of an improvement measure entity, obtaining the final representation of the improvement measure by utilizing neighbor aggregation, taking the final representation of the improvement factors and the improvement measure as an inner product, and obtaining the local improvement measure corresponding to the improvement factor subgraph.
It should be noted that, the improvement factor subgraph is mapped to the related knowledge graph, and the improvement factor-improvement measure interaction graph and the improvement measure knowledge graph are processed by using a relationship-based attention mechanism, and the importance degree of the different relationships is different, so that the important relationships match with larger attention scores. The different relationships characterize the "preference" information of the improvement factors for the improvement measures, and when the different relationships obtain similar improvement measures, the preference of the improvement factors can be deduced from the preference of another improvement factor, and the relationship redundancy is reduced by introducing the minimized relationship similarity, so that the calculation amount is reduced. Obtaining interest vectors of the improvement factors according to historical interaction conditions of the improvement factors and improvement measures, and calculating similarity among the improvement factors according to the interest vectors by using a cosine similarity or Euclidean distance method, wherein the cosine similarity is，Respectively represent improvement factors->Factor of improvement->: european distance similarity is->The method comprises the steps of carrying out a first treatment on the surface of the When the similarity is larger than a preset similarity threshold, the similarity is regarded as a similarity improvement factor; and the similar improvement factors are obtained to expand interaction of improvement factors and improvement measures in the related knowledge graph, and potential related improvement measure nodes of the improvement factor nodes are constructed, so that potential association loss in the original data is made up to a certain extent, and the data sparseness problem is relieved.
FIG. 3 illustrates a flow chart of the present invention for acquiring global coupling associations.
According to the embodiment of the invention, the improved factor subgraphs are spliced, and global coupling association is obtained through graph rolling network representation learning, specifically:
s302, acquiring time sequence characteristics among the improvement factor subsets corresponding to key performances according to the sequence of the production and preparation flow, and splicing the improvement factor subsets based on the production and preparation flow to generate an improvement factor global graph;
s304, constructing an encoder by using a multi-layer graph convolutional neural network, learning and representing an improvement factor global graph by using the encoder, constructing an adjacent matrix to acquire the spatial relationship of adjacent improvement factors in corresponding production and preparation, and distributing different weights to neighbor nodes by using the combination of average pooling and maximum pooling;
s306, carrying out aggregation of neighbor nodes in the improvement factor global graph by using weights, updating the representation of the improvement factor nodes, importing the updated representation of the improvement factor nodes into a decoder, carrying out feature recombination by using an RNN network, and outputting spatial features;
s308, fusing the time sequence features and the space features to obtain global coupling association among all factor subgraphs.
It should be noted that, the RNN network is used to construct the decoder, and the cyclic convolution is used to replace the normal convolution, so as to increase the interactivity between features and improve the performance of the decoder. Acquiring a preselected local improvement measure subset, screening in different local improvement measure subsets by utilizing the global coupling association, acquiring impact index evaluation results of different key performance working condition data subsets, acquiring key performance with highest improvement priority according to the evaluation results, and selecting a local improvement measure corresponding to the highest inner product from the corresponding local improvement measure subset; matching the selected local improvement measures as a starting point in the residual local improvement measure subset by using a global coupling association construction constraint condition, and obtaining local improvement measures which accord with a preset similarity threshold in the residual local improvement measure subset by using similarity calculation; and polymerizing the matched local improvement measures to generate the improved production and preparation process of the crosslinked polyethylene insulated power cable.
The quality detection information of the crosslinked polyethylene insulated power cable is obtained, and the defect power cable is obtained through extraction; extracting defect information based on the defect power cable, acquiring a performance grade dividing threshold according to the influence index, performing performance evaluation according to the defect information to obtain performance grade dividing information, and storing the performance grade dividing information into a preset database; performing defect tracing according to the influence index deviation to obtain defect analysis information; judging whether the defective product can be processed secondarily according to the defect analysis information, analyzing a corresponding secondary processing scheme, and predicting the performance of the power cable after secondary processing; retrieving and acquiring a utilization standard in a preset database by utilizing big data, and analyzing whether the utilization standard is reached after secondary processing; if the defect product is obtained, the secondary processing scheme is stored in a preset database, so that the subsequent extraction and use are facilitated, the utilization rate of the defect product is improved, and the economic loss is reduced.
Fig. 4 shows a block diagram of an improved high efficiency production and preparation system for crosslinked polyethylene insulated power cables according to the invention.
The second aspect of the present invention also provides an improved efficient production and preparation system 4 for crosslinked polyethylene insulated power cables, the system comprising: a memory 41, a processor 42, said memory comprising therein an improved efficient production preparation method program for a crosslinked polyethylene insulated power cable, said improved efficient production preparation method program for a crosslinked polyethylene insulated power cable, when executed by said processor, realizing the steps of:
acquiring a production and preparation process of a crosslinked polyethylene insulated power cable, extracting working condition data in historical production and preparation, extracting key performance of crosslinked polyethylene insulation, and clustering the working condition data by utilizing the key performance;
extracting improvement factors of production and preparation from the key performance corresponding class clusters, constructing a subset of the improvement factors corresponding to the key performance for graph representation, and obtaining a corresponding improvement factor subgraph;
mapping the improvement factor subgraphs to related knowledge maps, obtaining local improvement measures for the mapped improvement factor subgraphs by utilizing graph convolution, splicing the improvement factor subgraphs, and obtaining global coupling association through graph convolution network representation learning;
and polymerizing the global coupling association and the local improvement measures to carry out the whole-flow collaborative improvement of the production and preparation process of the crosslinked polyethylene insulated power cable, and replacing the existing production and preparation process.
The third aspect of the present invention also provides a computer-readable storage medium including therein an improved efficient production preparation method program of a crosslinked polyethylene insulated power cable, which when executed by a processor, implements the steps of the improved efficient production preparation method of a crosslinked polyethylene insulated power cable as described in any of the above.
In the several embodiments provided in this application, it should be understood that the disclosed apparatus and method may be implemented in other ways. The above described device embodiments are only illustrative, e.g. the division of the units is only one logical function division, and there may be other divisions in practice, such as: multiple units or components may be combined or may be integrated into another system, or some features may be omitted, or not performed. In addition, the various components shown or discussed may be coupled or directly coupled or communicatively coupled to each other via some interface, whether indirectly coupled or communicatively coupled to devices or units, whether electrically, mechanically, or otherwise.
The units described above as separate components may or may not be physically separate, and components shown as units may or may not be physical units; can be located in one place or distributed to a plurality of network units; some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional unit in each embodiment of the present invention may be integrated in one processing unit, or each unit may be separately used as one unit, or two or more units may be integrated in one unit; the integrated units may be implemented in hardware or in hardware plus software functional units.
Those of ordinary skill in the art will appreciate that: all or part of the steps for implementing the above method embodiments may be implemented by hardware related to program instructions, and the foregoing program may be stored in a computer readable storage medium, where the program, when executed, performs steps including the above method embodiments; and the aforementioned storage medium includes: a mobile storage device, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk or an optical disk, or the like, which can store program codes.
Alternatively, the above-described integrated units of the present invention may be stored in a computer-readable storage medium if implemented in the form of software functional modules and sold or used as separate products. Based on such understanding, the technical solutions of the embodiments of the present invention may be embodied in essence or a part contributing to the prior art in the form of a software product stored in a storage medium, including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in the embodiments of the present invention. And the aforementioned storage medium includes: a removable storage device, ROM, RAM, magnetic or optical disk, or other medium capable of storing program code.
Claims (8)
1. An improved efficient production and preparation method of a crosslinked polyethylene insulated power cable is characterized by comprising the following steps:
acquiring a production and preparation process of a crosslinked polyethylene insulated power cable, extracting working condition data in historical production and preparation, extracting key performance of crosslinked polyethylene insulation, and clustering the working condition data by utilizing the key performance;
extracting improvement factors of production and preparation from the key performance corresponding class clusters, constructing a subset of the improvement factors corresponding to the key performance for graph representation, and obtaining a corresponding improvement factor subgraph;
mapping the improvement factor subgraphs to related knowledge maps, obtaining local improvement measures for the mapped improvement factor subgraphs by utilizing graph convolution, splicing the improvement factor subgraphs, and obtaining global coupling association through graph convolution network representation learning;
and polymerizing the global coupling association and the local improvement measures to carry out the whole-flow collaborative improvement of the production and preparation process of the crosslinked polyethylene insulated power cable, and replacing the existing production and preparation process.
2. The improved efficient production and preparation method of the crosslinked polyethylene insulated power cable according to claim 1, wherein key performance of crosslinked polyethylene insulation is extracted, and the working condition data are clustered by utilizing the key performance, specifically:
acquiring working condition data produced and prepared by the crosslinked polyethylene insulated power cable in a preset history period, preprocessing the working condition data, pre-classifying the preprocessed working condition data according to a production and preparation flow, and adding a production and preparation procedure label for the working condition data;
dividing and obtaining rheological property, degassing property, scorch resistance and insulating property corresponding to the crosslinked polyethylene insulating material according to the production and preparation process, taking the key property as an initial clustering center, and carrying out preferential matching by utilizing working condition data of the production and preparation process labels corresponding to the key properties;
judging the Euclidean distance between the working condition data and the initial clustering center, attributing the working condition data to the initial clustering center closest to the working condition data to generate a class cluster, updating the clustering center through iterative clustering, and generating a final clustering result of the working condition data.
3. The method for efficiently producing and preparing the improved crosslinked polyethylene insulated power cable according to claim 1, wherein the improvement factors of production and preparation are extracted from the key performance corresponding clusters, specifically:
acquiring a cluster corresponding to key performance to generate a working condition data subset, matching the working condition data subset with an influence index of the crosslinked polyethylene insulated power cable, and evaluating the working condition data subset according to the influence index;
according to the evaluation result, sorting and setting the extraction quantity grades of the improvement factors of the data subsets under different working conditions, wherein the larger the grade is, the more the quantity of the extracted improvement factors is, the key performance is used as a retrieval tag, and the improvement examples related to the key performance are obtained by utilizing a big data means;
acquiring working condition characteristics of different working condition data subsets, and pre-screening the retrieved improved examples according to the working condition characteristics;
extracting improvement factors involved in a pre-screening improvement example to construct an improvement factor set, carrying out principal component analysis on the improvement factor set, obtaining the contribution degree of each improvement factor, and obtaining the coupling degree of each improvement factor and key performance according to the contribution degree;
and sorting the improvement factors according to the coupling degree, and acquiring a preset number of improvement factors based on the extraction number level.
4. The method for efficiently producing and preparing the improved crosslinked polyethylene insulated power cable according to claim 1, wherein a subset of improvement factors corresponding to key performances is constructed for graph representation, and a corresponding improvement factor subgraph is obtained, specifically:
obtaining an improvement factor subset corresponding to key performance, abstracting the improvement factor subset into an undirected graph, and defining the improvement factor as a node in the undirected graph;
calculating mutual information of different improvement factors in the improvement factor subset, calculating a maximum information coefficient between the improvement factors according to the mutual information, comparing the maximum information coefficient with a preset threshold value, and defining an edge structure between nodes in the undirected graph according to a comparison result;
and obtaining a corresponding improvement factor subgraph based on the undirected graph structure, extracting the maximum information coefficient between improvement factor nodes to construct a symmetrical matrix, replacing the maximum information coefficient smaller than a preset threshold value in the symmetrical matrix with 0, and obtaining an adjacent matrix of the improvement factor subgraph.
5. The efficient production and preparation method of the improved crosslinked polyethylene insulated power cable according to claim 1, wherein the improvement factor subgraphs are mapped to related knowledge maps, and the mapped improvement factor subgraphs are rolled by the graphs to obtain local improvement measures, specifically:
acquiring a related knowledge graph of a production preparation improvement measure of the crosslinked polyethylene insulated power cable, mapping the improvement factor subgraph to the related knowledge graph, and acquiring the relation between an improvement factor node and an improvement measure entity according to the historical interaction condition of the improvement factor and the improvement measure;
performing characterization learning on the relation between the improvement factor node and the improvement measure entity by using an attention mechanism in a graph rolling network, distributing attention scores to different relations, mapping the attention scores to a low-dimensional space to calculate Euclidean distance calculation relation similarity, and reducing relation redundancy by minimizing the relation similarity;
updating the adjacency matrix of the improvement factor subgraph according to the relation between the improvement factor nodes and the improvement measure entity, and carrying out weighted aggregation on neighbor nodes in the adjacency matrix by using attention scores corresponding to the relation to acquire embedded representation of the improvement factor;
and adding the embedded representations of the different layers of improvement factors to obtain the final representation of the improvement factors, and in addition, obtaining neighbor nodes of an improvement measure entity, obtaining the final representation of the improvement measure by utilizing neighbor aggregation, taking the final representation of the improvement factors and the improvement measure as an inner product, and obtaining the local improvement measure corresponding to the improvement factor subgraph.
6. The method for efficient production and preparation of an improved crosslinked polyethylene insulated power cable according to claim 5, wherein interest vectors of improvement factors are obtained according to historical interaction conditions of the improvement factors and improvement measures, and similarity between the improvement factors is calculated according to the interest vectors;
and obtaining similar improvement factors to expand interaction of improvement factors-improvement measures in the related knowledge graph, and constructing potential related improvement measure nodes of the improvement factor nodes.
7. The method for efficiently producing and preparing the improved crosslinked polyethylene insulated power cable according to claim 1, wherein the improvement factor subgraphs are spliced, and global coupling association is obtained through graph rolling network representation learning, specifically:
acquiring time sequence characteristics among the improvement factor subsets corresponding to the key performance according to the sequence of the production and preparation flow, and splicing the improvement factor subsets based on the production and preparation flow to generate an improvement factor global graph;
constructing an encoder by using a multi-layer graph convolutional neural network, learning and representing an improvement factor global graph by using the encoder, constructing an adjacent matrix to acquire the spatial relationship of adjacent improvement factors in corresponding production and preparation, and distributing different weights to neighbor nodes by using the combination of average pooling and maximum pooling;
the method comprises the steps of utilizing weights to aggregate neighbor nodes in an improvement factor global graph, updating the representation of the improvement factor nodes, importing the updated representation of the improvement factor nodes into a decoder, utilizing an RNN (network node network) to conduct feature recombination, and outputting spatial features;
and fusing the time sequence features and the space features to obtain global coupling association among all factor subgraphs.
8. The method for efficient production and preparation of an improved crosslinked polyethylene insulated power cable according to claim 1, characterized in that the global coupling correlation and local improvement measures are polymerized, in particular:
obtaining a preselected subset of local improvement measures, screening among the different subsets of local improvement measures using the global coupling association,
obtaining impact index evaluation results of different key performance working condition data subsets, obtaining key performance with highest improvement priority according to the evaluation results, and selecting local improvement measures corresponding to the highest inner product from the corresponding local improvement measure subsets;
matching the selected local improvement measures as a starting point in the residual local improvement measure subset by using a global coupling association construction constraint condition, and obtaining local improvement measures which accord with a preset similarity threshold in the residual local improvement measure subset by using similarity calculation;
and polymerizing the matched local improvement measures to generate the improved production and preparation process of the crosslinked polyethylene insulated power cable.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410217216.7A CN117788203B (en) | 2024-02-28 | 2024-02-28 | An improved method for efficiently producing cross-linked polyethylene insulated power cable |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410217216.7A CN117788203B (en) | 2024-02-28 | 2024-02-28 | An improved method for efficiently producing cross-linked polyethylene insulated power cable |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117788203A true CN117788203A (en) | 2024-03-29 |
| CN117788203B CN117788203B (en) | 2024-05-10 |
Family
ID=90383775
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410217216.7A Active CN117788203B (en) | 2024-02-28 | 2024-02-28 | An improved method for efficiently producing cross-linked polyethylene insulated power cable |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117788203B (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20190057145A1 (en) * | 2017-08-17 | 2019-02-21 | International Business Machines Corporation | Interactive information retrieval using knowledge graphs |
| US20220229400A1 (en) * | 2021-01-18 | 2022-07-21 | Siemens Aktiengesellschaft | Industrial device and method for building and/or processing a knowledge graph |
| WO2023273182A1 (en) * | 2021-06-29 | 2023-01-05 | 北京邮电大学 | Multi-source knowledge graph fusion-oriented entity alignment method and apparatus, and system |
| WO2023030513A1 (en) * | 2021-09-05 | 2023-03-09 | 汉熵通信有限公司 | Internet of things system |
| WO2023065545A1 (en) * | 2021-10-19 | 2023-04-27 | 平安科技（深圳）有限公司 | Risk prediction method and apparatus, and device and storage medium |
| CN116300691A (en) * | 2023-05-25 | 2023-06-23 | 深圳市正和楚基科技有限公司 | A state monitoring method and system for multi-axis linkage numerical control machining |
| CN116628220A (en) * | 2023-05-11 | 2023-08-22 | 国能长源汉川发电有限公司 | Method and system for generating electrical work tickets based on relational graph convolutional neural network |
| CN117455037A (en) * | 2023-10-13 | 2024-01-26 | 浙江工业大学 | A method for predicting the remaining time of production business processes based on multi-modal case knowledge graphs |
-
2024
- 2024-02-28 CN CN202410217216.7A patent/CN117788203B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20190057145A1 (en) * | 2017-08-17 | 2019-02-21 | International Business Machines Corporation | Interactive information retrieval using knowledge graphs |
| US20220229400A1 (en) * | 2021-01-18 | 2022-07-21 | Siemens Aktiengesellschaft | Industrial device and method for building and/or processing a knowledge graph |
| WO2023273182A1 (en) * | 2021-06-29 | 2023-01-05 | 北京邮电大学 | Multi-source knowledge graph fusion-oriented entity alignment method and apparatus, and system |
| WO2023030513A1 (en) * | 2021-09-05 | 2023-03-09 | 汉熵通信有限公司 | Internet of things system |
| WO2023065545A1 (en) * | 2021-10-19 | 2023-04-27 | 平安科技（深圳）有限公司 | Risk prediction method and apparatus, and device and storage medium |
| CN116628220A (en) * | 2023-05-11 | 2023-08-22 | 国能长源汉川发电有限公司 | Method and system for generating electrical work tickets based on relational graph convolutional neural network |
| CN116300691A (en) * | 2023-05-25 | 2023-06-23 | 深圳市正和楚基科技有限公司 | A state monitoring method and system for multi-axis linkage numerical control machining |
| CN117455037A (en) * | 2023-10-13 | 2024-01-26 | 浙江工业大学 | A method for predicting the remaining time of production business processes based on multi-modal case knowledge graphs |
Non-Patent Citations (1)
| Title |
|---|
| 官赛萍;靳小龙;贾岩涛;王元卓;程学旗;: "面向知识图谱的知识推理研究进展", 软件学报, no. 10, 8 February 2018 (2018-02-08) * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117788203B (en) | 2024-05-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115409208A (en) | Method and device for selection of repair scheme for stripped cable based on risk assessment | |
| CN112561251B (en) | Distribution network abnormal point detection method, device, computer equipment and storage medium | |
| CN118332424B (en) | A small sample cross-domain classification prediction method for electromagnetic compatibility defects in battery simulators | |
| CN115459782A (en) | Industrial Internet of things high-frequency data compression method based on time sequence segmentation and clustering | |
| CN116340586B (en) | Cloud platform-based automobile fault management method, system and storage medium | |
| CN108845227A (en) | The method that a kind of pair of high-tension cable carries out fault pre-alarming | |
| CN107784393A (en) | A kind of the defects of transmission line of electricity Forecasting Methodology and device | |
| CN107784392A (en) | A kind of the defects of transmission line of electricity based on machine learning Forecasting Methodology and device | |
| CN115828984B (en) | GIS internal intermittent discharge mode identification system, method, electronic equipment and medium based on SSA and RF | |
| CN115935807A (en) | Diagnostic model training method based on graph Markov neural network | |
| CN117788203B (en) | An improved method for efficiently producing cross-linked polyethylene insulated power cable | |
| CN117708686A (en) | Fault diagnosis method and device for road illumination power distribution system | |
| CN112327190A (en) | A method for identifying the state of health of an energy storage battery | |
| CN119293564B (en) | Power supply safety state monitoring and early warning method and system for liquid cooling cable | |
| CN113629861A (en) | Slope monitoring method, slope monitoring device, computer device, and storage medium | |
| CN114154037B (en) | Unbalanced trademark retrieval method based on depth hash of aggregation similarity | |
| CN118035801A (en) | Partial discharge signal localization method based on multi-scale detection self-attention network | |
| CN115114991B (en) | Electrical main equipment wiring design method, device, equipment and medium | |
| Nie et al. | A Hierarchical Fault Diagnosis Model for Power Transformers | |
| CN120910771B (en) | Data Analysis-Based Monitoring Method and Device for Energy Storage Cabinets | |
| Liu et al. | Transformer fault diagnosis model based on iterative nearest neighbor interpolation and ensemble learning | |
| CN120045530B (en) | Question answering method for electric power intelligent question answering system | |
| CN119646592B (en) | KFWAdaBoost-based transformer fault detection method, KFWAdaBoost-based transformer fault detection device, KFWAdaBoost-based transformer fault detection equipment and medium | |
| CN119805118B (en) | Partial discharge mode identification method and system of GIS equipment, terminal equipment and storage medium | |
| CN116365520B (en) | Power transmission line equipment risk prediction method and device and computer equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right |
Denomination of invention: An improved efficient production method for cross-linked polyethylene insulated power cables Granted publication date: 20240510 Pledgee: Xi'an Caijin Financing Guarantee Co.,Ltd. Pledgor: Xi'an Hualian Power Cable Co.,Ltd. Registration number: Y2024980024616 |
|
| PE01 | Entry into force of the registration of the contract for pledge of patent right |