CN117892817A - Knowledge graph construction method based on manufacturing full life cycle data - Google Patents
Knowledge graph construction method based on manufacturing full life cycle data Download PDFInfo
- Publication number
- CN117892817A CN117892817A CN202410299830.2A CN202410299830A CN117892817A CN 117892817 A CN117892817 A CN 117892817A CN 202410299830 A CN202410299830 A CN 202410299830A CN 117892817 A CN117892817 A CN 117892817A
- Authority
- CN
- China
- Prior art keywords
- data
- graph
- life cycle
- knowledge graph
- state
- 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
- 238000004519 manufacturing process Methods 0.000 title claims abstract description 48
- 238000010276 construction Methods 0.000 title claims abstract description 21
- 230000007704 transition Effects 0.000 claims abstract description 48
- 239000011159 matrix material Substances 0.000 claims abstract description 31
- 238000012549 training Methods 0.000 claims abstract description 20
- 238000000034 method Methods 0.000 claims description 41
- 230000008859 change Effects 0.000 claims description 30
- 239000002994 raw material Substances 0.000 claims description 17
- 238000004590 computer program Methods 0.000 claims description 7
- 238000013068 supply chain management Methods 0.000 claims description 7
- 238000003860 storage Methods 0.000 claims description 6
- 238000000605 extraction Methods 0.000 claims description 3
- 238000002372 labelling Methods 0.000 claims description 3
- 239000000758 substrate Substances 0.000 claims 1
- 239000000047 product Substances 0.000 description 15
- 238000012545 processing Methods 0.000 description 8
- 230000008569 process Effects 0.000 description 7
- 230000000875 corresponding effect Effects 0.000 description 5
- 230000000694 effects Effects 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 5
- 238000013528 artificial neural network Methods 0.000 description 4
- 230000008901 benefit Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 230000003993 interaction Effects 0.000 description 3
- 230000009471 action Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 2
- 238000013139 quantization Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000007476 Maximum Likelihood Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 239000012467 final product Substances 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 230000005855 radiation Effects 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 238000012384 transportation and delivery Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/02—Knowledge representation; Symbolic representation
- G06N5/022—Knowledge engineering; Knowledge acquisition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/02—Knowledge representation; Symbolic representation
- G06N5/027—Frames
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/04—Manufacturing
-
- 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
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Computing Systems (AREA)
- Evolutionary Computation (AREA)
- Data Mining & Analysis (AREA)
- Computational Linguistics (AREA)
- Artificial Intelligence (AREA)
- Business, Economics & Management (AREA)
- Software Systems (AREA)
- Human Resources & Organizations (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Strategic Management (AREA)
- Primary Health Care (AREA)
- Marketing (AREA)
- General Health & Medical Sciences (AREA)
- Economics (AREA)
- Health & Medical Sciences (AREA)
- Manufacturing & Machinery (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to a knowledge graph construction method based on manufacturing industry full-service full-product life cycle data, which comprises the following steps: s10, constructing a preliminary knowledge graph from top to bottom based on expert knowledge, wherein the construction of the knowledge graph defines the types of key entities and the relations between the key entities; s20, tracking and collecting complete full life cycle data of part of manufacturing products, wherein the complete full life cycle data is used as the most basic training data containing labels, and then dividing the statistical level data into a plurality of subgraphs according to time dimension for a large amount of statistical level data of the full life cycle; s30, constructing a graph Markov model, analyzing the initial state of the subgraph and the value of the final state among the states for each adjacent time segment, calculating a variation difference value as training data, and introducing the variation difference value into the graph Markov model to deduce a state transition matrix among key entities; s40, constructing a final time sequence knowledge graph by using the state transition matrix.
Description
Technical Field
The invention relates to the field of manufacturing data processing, in particular to a knowledge graph construction method based on manufacturing full life cycle data.
Background
With the development of industry 4.0 and internet of things, the manufacturing industry generates a large amount of data in the links of design, production, supply chain, after-sales service and the like. These data exist in a variety of formats and sources, such as sensor data, production logs, sales records, etc., the magnitude and variety of which present challenges to traditional data processing and utilization, as well as increasing regulatory requirements for the full life cycle of the product. The knowledge graph is used as a graph theory-based data representation method, and can organize data in the form of entities, relations and attributes so as to present the association and the connection between the data. In the manufacturing industry, the data of different life cycle stages can be correlated by constructing a knowledge graph to form a structured knowledge network. This helps to better understand the relationships between data, facilitating sharing and reuse of information. By using the knowledge graph, the manufacturing industry can realize more intelligent decision support and business process optimization.
The prior art is excellent in processing various knowledge graph application scenes, but challenges still exist when facing specific scenes of manufacturing industry full-service full-product life cycle data. In the field, information among different data sources is often not common, and circulated data is mainly statistical level data, so that the characteristic seriously influences the effective construction of a knowledge graph, and no effective solution is proposed in the prior study.
Disclosure of Invention
The invention provides a knowledge graph construction method based on full life cycle data of manufacturing industry, and aims to at least solve one of the technical problems in the prior art.
The technical scheme of the invention is a knowledge graph construction method based on manufacturing industry all-service all-product life cycle data, which comprises the following steps: s10, constructing a preliminary knowledge graph from top to bottom based on expert knowledge, wherein the construction of the knowledge graph defines the types of key entities and the relations between the key entities; s20, tracking and collecting complete full life cycle data of part of manufacturing products, wherein the complete full life cycle data is used as the most basic training data containing labels, and then counting level data of a large number of full life cycles, wherein subgraphs in a plurality of time slices in the full life cycle are divided into a plurality of subgraphs according to time dimension; s30, constructing a graph Markov model, analyzing initial state data and final state data among states of the subgraph for each adjacent time segment, calculating a variation difference value as training data, and carrying out the training data with the most basic training data containing labels so as to be carried into the graph Markov model to deduce a state transition matrix among key entities; s40, constructing a final time sequence knowledge graph by using the state transition matrix.
Further, before the step S10: before the knowledge graph is constructed, a macroscopic view of the manufacturing industry is introduced, and key entity types in a supply chain of the manufacturing industry and concepts corresponding to the key entities are determined.
Further, the step S10 includes: the key entity types are defined by industry expert knowledge and include suppliers, parts, products, and customers.
Further, the step S20 includes: and extracting the association relation of the full life cycle data among the systems by a manual labeling mode for the full life cycle data stored in the non-common system, introducing the full life cycle data into a supply chain management system, and correlating the labeled full life cycle data through the supply chain management system, wherein the full life cycle data comprises sources of raw materials, purchasing cost, arrival time, production batch, machine running time, output quality, shipment time, path and arrival time data.
Further, the step S30 includes: in a key entity, periodically collecting the statistic level data according to a preset time interval, and obtaining a plurality of change difference value data in a time segment in the time interval from the data of the subgraph, wherein the initial state data and the final state data of the subgraph are taken, and the plurality of change difference value data in the time segment are according with the following formula relation according to the knowledge graph; subgraph final state data = subgraph initial state data + change difference data + … … change difference data; and then dividing the increment state data or the decrement state data for the variation difference data.
Further, the step S30 includes: and in the preliminary knowledge graph, in the time segments of a plurality of different time nodes, data of a part of the change difference value with actual change conditions are input as a labeled data set during training of the graph Markov model, and state transition matrixes of different nodes in the knowledge graph are trained to obtain state transition matrixes of which one input is front and rear time segment state changes and output as each node in the current state.
Further, the method also comprises the following steps: s50, corresponding to the preliminary knowledge graph, in the data of the subgraph in the time segment in the time interval, a state transition matrix exists in a change mode between the initial state of the subgraph and the final state of the subgraph, and the following formula relation is met: sub-graph final state data = sub-graph initial state data x transition matrix; and under the same node state, obtaining the state transition matrix, and comparing the state transition matrix with the state transition matrix output by the graph Markov model to verify the accuracy of the graph Markov model.
The technical scheme of the invention also relates to a computer device, which comprises a memory and a processor, wherein the processor executes the computer program stored in the memory to implement the method.
The invention also relates to a computer-readable storage medium, on which computer program instructions are stored, which, when being executed by a processor, carry out the above-mentioned method.
According to some embodiments of the present invention,
The beneficial effects of the invention are as follows.
By constructing a knowledge graph from full lifecycle data of the manufacturing industry, the present method provides a deeper understanding of the data; this deep understanding helps reveal patterns and relationships hidden in complex data sets, providing a decision maker with more rich information; the method provides a processing mode for the generated data of the statistic level, and the complex statistic data is structured to enable the generated data to be easier to operate and analyze, so that the application range of the knowledge graph construction method is greatly expanded; the structured data is more suitable for the application of various advanced analysis technologies and tools, and can be converted into actual business insights and action guidelines by combining the existing knowledge graph application technology, so that the structured data has remarkable effect on improving the overall operation efficiency and competitiveness of enterprises.
Further, additional aspects and advantages of the invention will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the invention.
Drawings
Fig. 1 is a general flow chart of a knowledge graph construction method based on manufacturing full-service full-product lifecycle data, according to an embodiment of the present invention.
Fig. 2 is a flowchart of an implementation of a knowledge graph construction method based on manufacturing full-service full-product lifecycle data, according to an embodiment of the present invention.
Fig. 3 is a schematic diagram of a first example illustration of an application on a supplier-supply-parts-processing-line, in accordance with an embodiment of the present invention.
Fig. 4 is a schematic diagram illustrating a second example of an application on a supplier-supply-parts-processing-line, according to an embodiment of the present invention.
Fig. 5 is a schematic diagram illustrating a third example of an application on a supplier-supply-parts-processing-line, according to an embodiment of the present invention.
Detailed Description
The conception, specific structure, and technical effects produced by the present invention will be clearly and completely described below with reference to the embodiments and the drawings to fully understand the objects, aspects, and effects of the present invention.
It should be noted that, unless otherwise specified, when a feature is referred to as being "fixed" or "connected" to another feature, it may be directly or indirectly fixed or connected to the other feature. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. Furthermore, unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in the description presented herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and/or" as used herein includes any combination of one or more of the associated listed items.
It should be understood that although the terms first, second, third, etc. may be used in this disclosure to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element of the same type from another. For example, a first element could also be termed a second element, and, similarly, a second element could also be termed a first element, without departing from the scope of the present disclosure. The use of any and all examples, or exemplary language (e.g., "such as") provided herein, is intended merely to better illuminate embodiments of the invention and does not pose a limitation on the scope of the invention unless otherwise claimed.
Referring to fig. 1 to 5, in some embodiments, the present invention discloses a knowledge graph construction method based on manufacturing industry all-business all-product lifecycle data, the method comprising the steps of:
s10, constructing a preliminary knowledge graph from top to bottom based on expert knowledge, wherein the construction of the knowledge graph defines the types of key entities and the relations among the key entities.
S20, tracking and collecting a small amount of complete full life cycle data of the manufacturing industry products, wherein the complete full life cycle data is used as the most basic training data containing labels, and then dividing the statistical level data into a plurality of subgraphs according to the time dimension for a large amount of statistical level data of the full life cycle.
S30, constructing a graph Markov model, analyzing the initial state of the subgraph and the value of the final state among the states for each adjacent time segment, calculating a variation difference value as training data, and taking the variation difference value and the most basic training data containing labels into the graph Markov model to deduce a state transition matrix among key entities.
S40, constructing a final time sequence knowledge graph by using the state transition matrix.
In a step S10 of the method,
S10, constructing a preliminary knowledge graph from top to bottom based on expert knowledge, wherein the construction of the knowledge graph defines the types of key entities and the relations among the key entities. This step focuses on defining the basic relationships between the entity classes. This step focuses on building the necessary triplet types, identifying possible associations between entities, but does not involve specific entity interaction results.
Further, the supply chain of the manufacturing industry often involves huge data, and before the knowledge graph is constructed, a macroscopic view of the manufacturing industry is introduced to determine the types of key entities in the supply chain of the manufacturing industry and concepts corresponding to the key entities. To ensure that the graph structure accurately reflects the core features and processes of the industry, it is often necessary to cooperate with industry experts, by means of their abundant industry knowledge and experience, to define the key entity types, including suppliers, parts, products and customers, by the knowledge of the industry experts.
The entity definitions in this step are used to construct a preliminary knowledge-graph framework. This framework contains relationships between entities such as "vendor-offer-part" or "product-include-part". The key to this stage is to build a structured view that maps the basic elements and their interactions in the manufacturing supply chain, but does not involve specific operational data or complex interaction patterns.
In a step S20 of the method,
S20, tracking and collecting complete full life cycle data of part of manufacturing products, wherein the complete full life cycle data is used as the most basic training data containing labels, and then counting level data of a large number of full life cycles, wherein sub-images in a plurality of time slices in the full life cycle are divided into a plurality of sub-images according to time dimension.
Complete tracking and collection of full lifecycle data of manufacturing products during the supply chain is an extremely difficult comprehensive task involving the application of cross-sector collaboration and technology and the enormous cost of tracking for all products. The full life cycle of a manufacturing product encompasses the purchase from raw materials, the flow of product manufacture, the supply chain link, and the delivery of the final product to the customer.
In this step, the supply chain management system of the manufacturing industry and the assistance of the practitioner are required, and these data are stored in different systems of the manufacturing industry, respectively, but are not shared between the systems. And extracting the association relation of the full life cycle data among the systems by using a manual labeling mode for the full life cycle data stored in the non-common system, introducing the full life cycle data into a supply chain management system, and associating the labeled full life cycle data through the supply chain management system. The full life cycle data includes source of raw materials, procurement costs, arrival time, production lot, machine run time, yield quality, shipment time, route path, and arrival time data.
In the method steps S30 and S40,
S30, constructing a graph Markov model, analyzing initial state data and final state data among states of the subgraph for each adjacent time segment, calculating a variation difference value as training data, and taking the variation difference value and the most basic training data containing labels into the graph Markov model to derive a state transition matrix among key entities.
Further, in the key entity, the statistics level data are collected periodically at preset time intervals, and such a data collection arrangement ensures continuity and consistency of data among the data of the subgraph in the time segments within the time intervals. The data collected for each time segment reflects the particular data state at that time, and when comparing the data for adjacent time segments, the observed changes reveal transitions between states. In order to effectively process and analyze these large amounts of full life cycle statistics, dividing them into multiple sub-graphs according to the time dimension becomes an efficient method. For each pair of adjacent time segments, a set of critical data points can be obtained by analyzing and calculating the data of their initial state and the data of the final state, as well as the data of the difference in variation between the two states. These data points reflect not only changes in the production process, but also information in many ways, such as fluctuations in market demand, changes in supply chain efficiency, etc.
As shown with reference to the embodiment of fig. 3-5, this embodiment maps the relationship between the following entities: supplier-supply-parts-processing-production line. Between the entities in fig. 3, it is difficult to directly collect data because of the actual raw material flow direction, but because the information is not common between different raw material factories or different production departments; in fig. 4, since statistical data are used between different raw material factories or different production departments, the physical relationship between part of raw material factories and processing procedures is omitted. In this embodiment, compared with the accurate and complete production data record, a large amount of information is missing, and the construction of the knowledge graph is directly affected.
To solve this problem, it is necessary to implement restoration of data at a statistical level to complete production data. Further, taking the sub-graph initial state-data and the sub-graph final state data, and obtaining a plurality of variation difference-data in the time segment according to the knowledge graph and the following formula relation.
Sub-picture final state data = sub-picture initial state data + change difference data + … … change difference data.
And then dividing the increment state data or the decrement state data for the variation difference data.
Taking the period from raw material purchase to processing as an example, the change in a raw material between adjacent time slices can be represented as , i.e., raw material stock for the last time slice + newly added in that time slice-consumed in that time slice. In the above formula, the change difference data of the raw material purchase data amount is divided into the data of the increment state, and the data amount consumed by the raw material production line is divided into the data of the decrement state. In practice, the newly added part and the consumed part follow a certain rule, which corresponds to the association relationship between the entities in the knowledge graph. The key of the extraction of the state transition matrix is to understand and quantify the state change condition of each entity in the data under different time nodes. The operation process of the part is similar to the condition of the association relation between the nodes by means of expert knowledge quantization when the knowledge graph is constructed, and a plurality of quantifiable indexes are required to be specifically defined for different stages to represent states, such as the quantity of raw materials in the raw material processing period, the quantity of subsequent orders and the like. Then, the change of the state of each entity between adjacent time segments is calculated as the change degree of the state in the time segment, which can be realized by calculating the difference value, such as the stock quantity of the next time segment minus the stock quantity of the previous time segment, and the obtained difference value reflects the change of the stock of the raw materials in the time segment.
And in the preliminary knowledge graph, in the time segments of a plurality of different time nodes, taking part of the change difference data with actual change conditions as a labeled dataset in the training of the Markov model, and training state transition matrixes of different nodes in the knowledge graph to obtain state transition matrixes of which one input is the state change of the front time segment and the back time segment and output as each node in the current state. The state transition matrix contains the state change of the quantization index between the current node and other nodes. Therefore, the statistical level data can be converted into the relation among different nodes, so that the construction of the knowledge graph is realized.
By taking the thought of the Markov chain into account, transition equations of actual states necessarily exist between different observation time slices, so that a semi-supervised graph Markov model can be constructed to acquire state transition relations in the transition equations, and attention mechanisms are introduced to capture the influence of each entity on other entities in the radiation range of the entity. In the timing diagram, for a particular entity, a transition of that entity to all neighboring entities occurs between two timestamps. This phenomenon conceals a corresponding state transition matrix for the entity at a later time, assuming that the scope of influence of each entity is a specific sequence of entities, i.e. at each instant the entity should have a probability transition matrix to other entities within the sequence of entities. Therefore, the influence of one entity at the next moment is diffused to all the adjacent entities, and the statistical data value of the next moment can be obtained by superposing the probability transition matrixes of all the entities and the original statistical data. Notably, the current task is to find the best transition probability matrix based on the statistical data values at the current time and the next time. The goal of the actual task is therefore to train a state transition matrix of a graph markov model containing time-series states such that the maximum likelihood estimate of the state transition process is maximized over time.
The advantage of the graph Markov chain algorithm in the selection of the algorithm is that the algorithm combines the advantages of statistical relationship learning and graph neural network, and is effectively used for semi-supervised object classification. The algorithm simulates the joint distribution of object labels through a conditional random field and uses a variational algorithm for efficient training. The graph Markov chain algorithm simultaneously comprises two functionally different graph neural networks, namely a network represented by a learning object and a network of local label dependence of a model, and the two networks are alternately trained, so that the limitations of the existing statistical relationship method and the limitation of the graph neural network are effectively solved, and the learning of edges of different relationships is more focused, so that the method is more suitable for the current situation.
Compared with the existing research and manufacturing industry data trend research, the method has the advantages that overall statistical data are used as input of a graph neural network, all data are subjected to feature extraction, a subgraph containing association relations among spaces is generated, and the method thinking usually focuses on future relation change trends. By constructing a knowledge graph from full lifecycle data of the manufacturing industry, the present method provides a deeper understanding of the data; this deep understanding helps reveal patterns and relationships hidden in complex data sets, providing a decision maker with more rich information; the method provides a processing mode for the generated data of the statistic level, and the complex statistic data is structured to enable the generated data to be easier to operate and analyze, so that the application range of the knowledge graph construction method is greatly expanded; the structured data is more suitable for the application of various advanced analysis technologies and tools, and can be converted into actual business insights and action guidelines by combining the existing knowledge graph application technology, so that the structured data has remarkable effect on improving the overall operation efficiency and competitiveness of enterprises.
With continued reference to fig. 2, in a further step S50,
S50, corresponding to the preliminary knowledge graph, in the data of the subgraph in the time segment in the time interval, a state transition matrix exists in a change mode between the initial state of the subgraph and the final state of the subgraph, and the following formula relation is met:
subgraph final state data = subgraph initial state data x transition matrix
With continued reference to the above embodiments of fig. 3 to 5 regarding raw material entities, the newly added portion and the consumed portion actually follow a certain rule, which corresponds to the association relationship between entities in the knowledge graph. One can consider that there is a state transition matrix, , for the change between the two states to/> . The statistical value change of the statistical data is modified for the transition between states through the transition, and the transition matrix contains specific transition conditions of the current node and the front and rear nodes, and the transition conditions can be used as relations in the knowledge graph, so that the knowledge graph construction based on the statistical level data is realized; and meanwhile, the state transition matrix output by the graph Markov model can be obtained and compared with the state transition matrix output by the graph Markov model under the same node state, and the accuracy of the graph Markov model is verified.
In some embodiments, the invention also discloses a computer device comprising a memory and a processor, which executes a computer program stored in the memory to implement the method as described above. Also disclosed is a computer readable storage medium having stored thereon program instructions which when executed by a processor perform a method as described above.
It should be appreciated that the method steps in embodiments of the present invention may be implemented or carried out by computer hardware, a combination of hardware and software, or by computer instructions stored in non-transitory computer-readable memory. The method may use standard programming techniques. Each program may be implemented in a high level procedural or object oriented programming language to communicate with a computer system. However, the program(s) can be implemented in assembly or machine language, if desired. In any case, the language may be a compiled or interpreted language. Furthermore, the program can be run on a programmed application specific integrated circuit for this purpose.
Furthermore, the operations of the processes described herein may be performed in any suitable order unless otherwise indicated herein or otherwise clearly contradicted by context. The processes (or variations and/or combinations thereof) described herein may be performed under control of one or more computer systems configured with executable instructions, and may be implemented as code (e.g., executable instructions, one or more computer programs, or one or more applications), by hardware, or combinations thereof, collectively executing on one or more processors. The computer program includes a plurality of instructions executable by one or more processors.
Further, the method may be implemented in any type of computing platform operatively connected to a suitable computing platform, including, but not limited to, a personal computer, mini-computer, mainframe, workstation, network or distributed computing environment, separate or integrated computer platform, or in communication with a charged particle tool or other imaging device, and so forth. Aspects of the invention may be implemented in machine-readable code stored on a non-transitory storage medium or device, whether removable or integrated into a computing platform, such as a hard disk, optical read and/or write storage medium, RAM, ROM, etc., such that it is readable by a programmable computer, which when read by a computer, is operable to configure and operate the computer to perform the processes described herein. Further, the machine readable code, or portions thereof, may be transmitted over a wired or wireless network. When such media includes instructions or programs that, in conjunction with a microprocessor or other data processor, implement the steps described above, the invention described herein includes these and other different types of non-transitory computer-readable storage media. The invention may also include the computer itself when programmed according to the methods and techniques of the present invention.
The computer program can be applied to the input data to perform the functions described herein, thereby converting the input data to generate output data that is stored to the non-volatile memory. The output information may also be applied to one or more output devices such as a display. In a preferred embodiment of the invention, the transformed data represents physical and tangible objects, including specific visual depictions of physical and tangible objects produced on a display.
The present invention is not limited to the above embodiments, but can be modified, equivalent, improved, etc. by the same means to achieve the technical effects of the present invention, which are included in the spirit and principle of the present invention. Various modifications and variations are possible in the technical solution and/or in the embodiments within the scope of the invention.
Claims (9)
1. The knowledge graph construction method based on the manufacturing industry full-service full-product life cycle data is characterized by comprising the following steps of:
S10, constructing a preliminary knowledge graph from top to bottom based on expert knowledge, wherein the construction of the knowledge graph defines the types of key entities and the relations between the key entities;
S20, tracking and collecting complete full life cycle data of part of manufacturing products, wherein the complete full life cycle data is used as basic training data containing labels, and then statistics level data of the full life cycle are obtained, wherein subgraphs in a plurality of time slices in the full life cycle are divided into a plurality of subgraphs according to a time dimension;
s30, constructing a graph Markov model, analyzing initial state data and final state data among states of the subgraph for each adjacent time segment, calculating a variation difference value as training data, and carrying the variation difference value into the graph Markov model together with basic training data containing labels so as to derive a state transition matrix among key entities;
s40, constructing a final time sequence knowledge graph by using the state transition matrix.
2. The method according to claim 1, characterized in that, before said step S10:
before the knowledge graph is constructed, a macroscopic view of the manufacturing industry is introduced, and key entity types in a supply chain of the manufacturing industry and concepts corresponding to the key entities are determined.
3. The method according to claim 1, wherein the step S10 includes:
The key entity types are defined by industry expert knowledge and include suppliers, parts, products, and customers.
4. The method according to claim 1, wherein the step S20 includes:
For the full life cycle data stored in the non-common system, the association relation extraction of the full life cycle data among the systems is realized by a manual labeling mode, a supply chain management system is introduced, the labeled full life cycle data is associated by the supply chain management system,
The full life cycle data comprises raw material sources, purchasing cost, arrival time, production batch, machine running time, output quality, shipment time, path and arrival time data.
5. The method according to claim 1, wherein the step S30 includes:
in the key entity, the statistical level data is collected periodically according to a preset time interval, in the data of the subgraph in the time segment in the time interval,
Taking the sub-graph initial state data and the sub-graph final state data, and obtaining a plurality of change difference data in the time segment according to the knowledge graph and the following formula relation;
Subgraph final state data = subgraph initial state data + change difference data + … … change difference data;
then dividing the increment state data or the decrement state data for the change difference value data;
And in the preliminary knowledge graph, in the time segments of a plurality of different time nodes, data of a part of the change difference value with actual change conditions are input as a labeled data set during training of the graph Markov model, and state transition matrixes of different nodes in the knowledge graph are trained to obtain state transition matrixes of which one input is front and rear time segment state changes and output as each node in the current state.
6. The method of claim 5, further comprising the step of:
s50, corresponding to the preliminary knowledge graph, in the data of the subgraph in the time segment in the time interval, a state transition matrix exists in a change mode between the initial state of the subgraph and the final state of the subgraph, and the following formula relation is met:
subgraph final state data = subgraph initial state data x transition matrix
And under the same node state, obtaining the state transition matrix, and comparing the state transition matrix with the state transition matrix output by the graph Markov model to verify the accuracy of the graph Markov model.
7. The method of claim 1, wherein the step of determining the position of the substrate comprises,
The graph Markov model is a semi-supervised graph Markov model.
8. A computer device comprising a memory and a processor, wherein the processor implements the method of any of claims 1 to 7 when executing a computer program stored in the memory.
9. A computer readable storage medium having stored thereon program instructions which, when executed by a processor, implement the method of any of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410299830.2A CN117892817B (en) | 2024-03-15 | 2024-03-15 | Knowledge graph construction method based on manufacturing full life cycle data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410299830.2A CN117892817B (en) | 2024-03-15 | 2024-03-15 | Knowledge graph construction method based on manufacturing full life cycle data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117892817A true CN117892817A (en) | 2024-04-16 |
| CN117892817B CN117892817B (en) | 2024-07-02 |
Family
ID=90643284
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410299830.2A Active CN117892817B (en) | 2024-03-15 | 2024-03-15 | Knowledge graph construction method based on manufacturing full life cycle data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117892817B (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040153430A1 (en) * | 2002-09-24 | 2004-08-05 | Saed Sayad | Method and apparatus for data analysis |
| CN110674310A (en) * | 2019-09-04 | 2020-01-10 | 东华大学 | Knowledge graph-based industrial Internet of things identification method |
| US20210216887A1 (en) * | 2020-01-14 | 2021-07-15 | Nec Laboratories America, Inc. | Knowledge graph alignment with entity expansion policy network |
| CN116702891A (en) * | 2023-03-20 | 2023-09-05 | 南方科技大学 | Fault tracing method and terminal based on automobile manufacturing industry |
| CN116805010A (en) * | 2023-05-18 | 2023-09-26 | 西安理工大学 | Multi-data chain integration and fusion knowledge graph construction method oriented to equipment manufacturing |
| CN116993306A (en) * | 2023-06-15 | 2023-11-03 | 中国科学院自动化研究所 | Knowledge graph-based construction method and device of network collaborative manufacturing system |
| CN117474091A (en) * | 2023-12-07 | 2024-01-30 | 中国农业银行股份有限公司 | Knowledge graph construction method, device, equipment and storage medium |
| CN117573880A (en) * | 2023-10-31 | 2024-02-20 | 北京科技大学 | Rolling process data element model and data space construction method and system |
-
2024
- 2024-03-15 CN CN202410299830.2A patent/CN117892817B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040153430A1 (en) * | 2002-09-24 | 2004-08-05 | Saed Sayad | Method and apparatus for data analysis |
| CN110674310A (en) * | 2019-09-04 | 2020-01-10 | 东华大学 | Knowledge graph-based industrial Internet of things identification method |
| US20210216887A1 (en) * | 2020-01-14 | 2021-07-15 | Nec Laboratories America, Inc. | Knowledge graph alignment with entity expansion policy network |
| CN116702891A (en) * | 2023-03-20 | 2023-09-05 | 南方科技大学 | Fault tracing method and terminal based on automobile manufacturing industry |
| CN116805010A (en) * | 2023-05-18 | 2023-09-26 | 西安理工大学 | Multi-data chain integration and fusion knowledge graph construction method oriented to equipment manufacturing |
| CN116993306A (en) * | 2023-06-15 | 2023-11-03 | 中国科学院自动化研究所 | Knowledge graph-based construction method and device of network collaborative manufacturing system |
| CN117573880A (en) * | 2023-10-31 | 2024-02-20 | 北京科技大学 | Rolling process data element model and data space construction method and system |
| CN117474091A (en) * | 2023-12-07 | 2024-01-30 | 中国农业银行股份有限公司 | Knowledge graph construction method, device, equipment and storage medium |
Non-Patent Citations (2)
| Title |
|---|
| HUIHUI HAN ET AL: ""Construction and Evolution of Fault Diagnosis Knowledge Graph in Industrial Process"", 《IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT》, vol. 71, 19 August 2022 (2022-08-19), pages 1 - 14 * |
| JIXIAO ZHANG ET AL: ""Hyper-Relational Knowledge Graph Neural Network for Next POI"", 《ARXIV:2311.16683V1》, 28 November 2023 (2023-11-28), pages 1 - 9 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117892817B (en) | 2024-07-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Trapero et al. | Analysis of judgmental adjustments in the presence of promotions | |
| JP6009864B2 (en) | Company evaluation system, company evaluation method and company evaluation program | |
| US8600709B2 (en) | Adaptive analytics multidimensional processing system | |
| CN107464025B (en) | Supply chain demand prediction and intelligent decision method based on big data | |
| EP2273431B1 (en) | Model determination system | |
| CN109409561B (en) | Construction method of multi-time scale time sequence collaborative prediction model | |
| Palacín et al. | Robust integrated production-maintenance scheduling for an evaporation network | |
| Vrabič et al. | Computational mechanics approach to managing complexity in manufacturing systems | |
| Wen et al. | A dual energy benchmarking methodology for energy-efficient production planning and operation of discrete manufacturing systems using data mining techniques | |
| CN103186713A (en) | Method and web application for OEE-analysis | |
| CN117291655B (en) | Consumer life cycle operation analysis method based on entity and network collaborative mapping | |
| CN113127538B (en) | High-precision spare part demand prediction method | |
| CN117952342A (en) | Construction scheme generation method and system integrating process knowledge base and space constraint | |
| CN117892817B (en) | Knowledge graph construction method based on manufacturing full life cycle data | |
| CN112199395A (en) | Artificial intelligence analysis method and system | |
| Yılmaz | Generative adversarial network for load data generation: Türkiye energy market case | |
| CN111768282B (en) | Data analysis method, device, equipment and storage medium | |
| Odrekhivskyі et al. | Problems of Construction of Smart Innovative Enterprises | |
| Ferreira et al. | Application of MCDM method for technologies selection to support energy management | |
| CN103970974A (en) | Defect-category-based security risk assessment method | |
| Šitova et al. | A concept of simulation-based SC performance analysis using SCOR metrics | |
| CN112632468A (en) | GDP prediction method based on integrated algorithm and enterprise electricity utilization data | |
| Ji et al. | Modeling of image, video and text fusion quality data packet system for aerospace complex products based on business intelligence | |
| CN118395135B (en) | Data preprocessing system and method for supply chain planning system | |
| TWI550531B (en) | Enterprise resource planning performance evaluation system and method |
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 |