CN114258551A

CN114258551A - Production modeling method, device and system

Info

Publication number: CN114258551A
Application number: CN201980099505.9A
Authority: CN
Inventors: 陈雪; 李明
Original assignee: Siemens AG
Current assignee: Siemens AG
Priority date: 2019-08-28
Filing date: 2019-08-28
Publication date: 2022-03-29
Also published as: WO2021035588A1

Abstract

A production modeling method, device and system, the method includes the following steps: s1, converting the existing production model into a map, and extracting the same subgraphs among a plurality of existing production model maps; s2, analyzing the modeling intention based on the input of the user on the modeling software, and capturing modeling events to extract an event stream based on the modeling intention; and S3, matching the sub-graph in the database based on the modeling intention event stream, and recommending a production model related to the sub-graph to the user based on the sub-graph so as to support the user to model. The provided production modeling mechanism can utilize existing production models so that production models of the same or similar domains can share some common profiles. The provided production modeling mechanism is more efficient.

Description

Production modeling method, device and system

Technical Field

The invention relates to the field of digitization, in particular to a production modeling method, a production modeling device and a production modeling system.

Background

The establishment of a production model in a factory is a key step for realizing production digitization. In order to build a production model, it is necessary to learn not only modeling instructions of modeling software but also domain knowledge of a target industry, such as an automobile industry, a food and drink industry, a discrete manufacturing industry, and even domain knowledge of a target production stage, such as in an automobile industry, a modeling target includes a press line, a welding line, a coating line, a general assembly line, or a warehouse.

A problem with existing modeling schemes is the waste of knowledge resources. On the one hand, some specific field models exist, and the learning time of engineers is greatly reduced and the modeling efficiency is improved by utilizing the existing models to support the modeling process.

Disclosure of Invention

The invention provides a production modeling method in a first aspect, which comprises the following steps: s1, converting the existing production model into a map, and extracting the same subgraphs among a plurality of existing production model maps; s2, analyzing the modeling intention based on the input of the user on the modeling software, and capturing modeling events to extract an event stream based on the modeling intention; and S3, matching the sub-graph in the database based on the modeling intention event stream, and recommending a production model related to the sub-graph to the user based on the sub-graph so as to support the user to model.

Further, the step S1 is preceded by the following steps: classifying the existing production models according to the technical field, wherein the step S3 further comprises the following steps: matching subgraphs in the database based on the modeling intent event stream and the technical field thereof.

Further, the step S2 further includes the following steps: redundant information in the user input on the modeling software other than the event stream is removed.

Further, the production modeling method further comprises the following steps: matching subgraphs in the database based on user selection.

Further, the step 1 is performed offline, and the steps S2 and S3 are performed online.

Further, the step 1 further comprises the following steps: and extracting layout information of the production model, wherein the layout information comprises a target component and attribute characteristics thereof and a connection relation between the target component, and the attribute characteristics comprise a name, a type and parameters.

A second aspect of the present invention provides a production modeling apparatus, including: a conversion device which converts an existing production model into a map; a map extraction device that extracts the same subgraph among a plurality of existing production model maps; a capturing means that analyzes the modeling intention based on an input of the user on the modeling software, and captures the modeling event to extract an event stream based on the modeling intention; a subgraph matching means that matches subgraphs in the database based on the modeling intent event stream; and the recommending device recommends the production model related to the subgraph to the user based on the subgraph so as to support the user to model.

Further, the production modeling apparatus further includes: and the classification device classifies the existing production model according to the technical field, wherein the graph matching device is also used for matching sub-graphs in the database based on the modeling intention event stream and the technical field thereof.

Further, the production modeling apparatus further includes: a cleaning device that removes redundant information other than the event stream in the user's input on the modeling software.

Further, the matching device is also used for matching the subgraph in the database based on the selection of the user.

Further, the map extracting device is further configured to extract layout information of the production model, where the layout information includes a target component and an attribute feature thereof, and a connection relationship between the target component, where the attribute feature includes a name, a type, and a parameter.

A third aspect of the present invention provides a production modeling system, comprising: a processor; and a memory coupled with the processor, the memory having instructions stored therein that, when executed by the processor, cause the electronic device to perform acts comprising: s1, converting the existing production model into a map, and extracting the same subgraphs among a plurality of existing production model maps; s2, analyzing the modeling intention based on the input of the user on the modeling software, and capturing modeling events to extract an event stream based on the modeling intention; and S3, matching the sub-graph in the database based on the modeling intention event stream, and recommending a production model related to the sub-graph to the user based on the sub-graph so as to support the user to model.

Further, the action S1 is preceded by: classifying the existing production models according to the technical field, wherein the action S3 further comprises the following steps: matching subgraphs in the database based on the modeling intent event stream and the technical field thereof.

Further, the action S2 further includes: redundant information in the user input on the modeling software other than the event stream is removed.

Further, the acts: matching subgraphs in the database based on user selection.

Further, S1 is performed offline, and the actions S2 and S3 are performed online.

Further, the action S1 further includes: and extracting layout information of the production model, wherein the layout information comprises a target component and attribute characteristics thereof and a connection relation between the target component, and the attribute characteristics comprise a name, a type and parameters.

A fourth aspect of the invention provides a computer program product tangibly stored on a computer-readable medium and comprising computer-executable instructions that, when executed, cause at least one processor to perform the method according to the first aspect of the invention.

A fifth aspect of the invention provides a computer-readable medium having stored thereon computer-executable instructions that, when executed, cause at least one processor to perform the method according to the first aspect of the invention.

The production modeling mechanism provided by the invention can utilize the existing production model, so that the production models in the same or similar fields can share some common maps. The production modeling mechanism provided by the invention has higher efficiency.

Drawings

FIG. 1 is a schematic block diagram of a production modeling apparatus according to an embodiment of the present invention;

FIG. 2 is a diagram of a graph data structure for converting a production model to a graph in production modeling according to an embodiment of the present invention;

FIG. 3 is a schematic diagram of extracting the same subgraph between profiles of different existing production models based on the profiles in production modeling according to an embodiment of the present invention;

FIG. 4 is a diagram illustrating recommendation of a modeling component to a user in production modeling according to an embodiment of the present invention;

FIG. 5 is a schematic illustration of extracting the same subgraph between the graphs of different existing production models based on the graphs in production modeling according to yet another embodiment of the present invention;

FIG. 6 is a schematic diagram of a neutron map matching and recommended production model in production modeling, according to an embodiment of the present invention.

Detailed Description

The following describes a specific embodiment of the present invention with reference to the drawings.

The invention provides a production-based modeling mechanism which can utilize the existing production model to improve the production modeling efficiency and avoid resource waste. The production modeling mechanism provided by the invention is based on modeling software and a map database.

Fig. 1 is a schematic structural diagram of a production modeling apparatus according to an embodiment of the present invention. As shown in fig. 1, the production modeling apparatus 100 includes a conversion apparatus 110, a capture apparatus 120, a matching apparatus 130, a recommendation apparatus 140, a map extraction apparatus 150, a data cleaning apparatus 160, an extraction apparatus 170, a subgraph matching apparatus 180, and a user selection apparatus 190, and a database DB. Among them, the converting means 110, the capturing means 120, the subgraph matching means 180 and the matching means 130 are executed based on an application program interface or a modeling language of modeling software, and have information interaction with the modeling software, and thus can be embedded in the modeling software. Other functional modules may be independent of the modeling software and can be identified by various programming software. The functional modules execute established functions of the functional modules on line and execute on-line functions on line according to different application scenes. The offline means that modeling is not performed by the user, and the online means that modeling is performed by the user.

The user executes modeling based on one modeling software 200, the production modeling device 100 calls at least one existing production model according to the user requirement, the conversion device 110 converts the existing production model into a map, the capture device 120 captures the input action of the user in the modeling software 200 and extracts a modeling intention by using the extraction device 170, then the matching device 180 matches a subgraph based on the modeling intention and recommends a production model related to the subgraph in the recommendation device 140, and finally recommends a final model through the matching device 130. Wherein the database DB includes a first database DB₁And a second database DB₂First database DB₁For storing maps G₁、G ₂……G _nA second database DB₂For storing subgraphs S₁、S ₂……S _n。

In a first aspect, the invention provides a production modeling method.

First, S1 is executed, the converting means 110 converts the existing production model into a map, and the map extracting means 150 extracts the same subgraph between different existing production models based on the map. The step 1 further comprises the following steps: and extracting layout information of the production model, wherein the layout information comprises a target component and attribute characteristics thereof and a connection relation between the target component, and the attribute characteristics comprise a name, a type and parameters.

Specifically, the existing production model refers to the production model which is already modeled, and all the existing production models M are used₁、M ₂……M _n. The conversion device 110 calls at least one existing production model, and extracts layout information of the production model, including target components and their attribute features, and connection relationships between the components. The attribute characteristics include nameScale, type, parameters, etc. Illustratively, the layout information for these production models is saved in the csv format for later use.

TABLE 1 production model layout information List

Serial number	Name of object	Object type	Front assembly	The latter assembly	Material
1	Pipe	*.Fluids.Pipe	A_mixer	Pipe3	A1
2	Line	*.MaterialFlow.line	Capping	Pipe2	B1
3	Connector	*.MaterialFlow.connector	Pipe4	Tank	C1

The table above is a list of production model layout information, where the layout information includes serial number, target name, target type, previous component, next component, and material. Specifically, a target component having a target name of "Pipe" has a target type of fluid, Pipe, the former component of which is "a mixer", the latter component of which is "Pipe 3", and the target component of which is "Pipe" is made of "a 1". Line (carousel), the former being "Capping", the latter being "Pipe 2", the target component "Line" being of the material "B1". The target component with the target name "Connector" has the target type ". MaterialFlow. Connector", the former component is "Pipe 4", the latter component is "Tank" (buffer node), and the material of the target component "Connector" is "C1".

The conversion device 110 is used for converting the layout information of the production model into a map. FIG. 2 is a diagram of a graph data structure for converting a production model into a graph in production modeling, according to an embodiment of the present invention. As shown in FIG. 2, a "component type" is a target node, representing a component. "name" and "processing time" are feature nodes, representing the feature attributes of the component. One of the component types has a value of "assembly", which has a name parameter having a value of "part m automatic assembly machine", and a value of "processing time 1 minute". Another component type value is a single step process with a name parameter having a value of "initial test", and a component type also having a processing time having a value of "3 minutes". Is there a In addition, the relationships of the target nodes include connection relationships, and may also be described as parameters. For example, the relationship between the two component types in FIG. 2 is "having a previous component" and "having a subsequent component".

The existing production model is converted into a map and stored in a first database DB1, wherein the first database DB1 is used for storing a map of the production model, and a plurality of maps, such as a first map G, are stored therein₁First map G₂… … nth map G_n。

The map extraction means 150 then extracts the same subgraph among a plurality of existing production model maps based on the map. A subgraph (subgraph) graph extraction algorithm is prestored in the graph extraction device 150 to search for the same subgraph among a plurality of production model graphs. As shown in FIG. 3, a first map G₁For the first production model M₁Having node a, node b, node c, node d and node e. Second map G₂For the second production model M₂Having node a, node b, node c, node d and node f. The third map G3 is a third production model M₃Having node a, node b, node c, node d, node e, and node f. Therefore, the first map G is extracted by a pattern extraction algorithm of a pre-stored subgraph (subgraph) in the map extraction device 150₁Second map G₂And a third map G₃The subgraph with the same relationship is subgraph S₁。

It should be noted that, a plurality of existing production models are subjected to iterative extraction of the same sub-graphs by the graph extraction device 150 to obtain a plurality of sub-graphs, and each sub-graph extracted for a different production model is stored in the second database DB₂。

According to a preferred embodiment of the present invention, the conversion device 110 and the atlas extraction device 150 are performed online to prepare data needed by the user to reproduce modeling at ordinary times, so as to further improve efficiency.

In addition, plant models are made up of individual nodes, such as a node representing a workstation. For example, although there are a plurality of production lines in different automobile factories, the production lines are different from one another in the same type of automobile factory, and although the production lines are different from one another, the differences are not great.

Then, step S2 is executed, and the capturing device 120 analyzes the modeling intention based on the input of the user on the modeling software 200, and captures the modeling event to extract an event stream based on the modeling intention. The goal of the capture device 120 is to obtain modeling information for the user, including which components are needed for the ongoing modeling operation, and how these components are connected.

Specifically, the capturing device 120 captures modeling software-based input on a personal computer, wherein the input corresponding modeling information includes an event and a modeling software interface screen shot when the event occurs, wherein the event is a command for operating the modeling software to the personal computer in response to a mouse action, a keyboard action or a touch screen action. The capture device 120 decomposes the modeling intention into a plurality of times according to the time and the positioning of the screen shot of the modeling software interface by the input of the personal computer on the modeling software, extracts and generates an event record table, and finally generates an event stream based on the modeling intention.

Since the existing production model necessarily includes many repetitive operations or back-and-forth correction operations in the process of manual modeling, these unnecessary operations become noise data, which is redundant information. It is therefore necessary to perform a removal operation on redundant information of the event stream. Therefore, step S2 further includes the steps of: the cleaning device 160 removes redundant information of the user input on the modeling software except for the event stream, and specifically, the cleaning device 160 performs some data cleaning operation, such as deleting noise information, missing data, or combining related data. Then, the extraction means 170 extracts the modeling intention of the engineer on the modeling software 200, that is, the model map of the engineer on the interface of the modeling software 200, using the process extraction algorithm.

Finally, step S3 is executed, the sub-graph matching device 180 matches the sub-graph in the database based on the modeling intention event stream, and the recommending device 140 recommends the production model related to the sub-graph to the user based on the sub-graph to support the user to model.

Specifically, the input of the subgraph matching device 180 is the modeling intention event stream and the second database DB₂Compares the model graph of the engineer modeling the interface of the software 200 with the second database DB₂And identifies whether the graph being modeled is any second database DB₂Subfigure (1). As shown in FIG. 4, if the user drags two nodes a and b in modeling software 200, sub-graph S is the same sub-graph₁Is extracted and stored in the second database DB₂At this time, the subgraph S will be automatically matched₁Then, the recommending means 140 recommends the node c and the node d on the user interface of the modeling software 200. Finally, if the user is satisfied with the recommended nodes c and d, the matching means 130 will present automatically a sub-graph S with nodes a, b, c and d in the modeling software 200 interface after the user has entered confirmation information₁。

Specifically, the step S1 is preceded by the following steps: classifying the existing production models according to the technical field, wherein the step S3 further comprises the following steps: matching subgraphs in the database based on the modeling intent event stream and the technical field thereof. For example, although there are a plurality of production lines in different automobile factories, the production lines are different from each other for the same type of automobile factories, and the differences are not large although the differences are necessarily different, so that the method can be divided into a field, and once a user performs modeling on one production line of a soda filling factory by using a modeling intention event stream on modeling software, the sub-graphs of the same type of production lines of the soda filling factory are matched in a database immediately.

Specifically, according to a preferred embodiment of the present invention, the layout information of the production model of the conversion means 110 is converted into a map. As shown in FIG. 5, one of the production models was converted into a third map G3, and the other production model was converted into a map G4. The third graph G3 comprises six fluid Source nodes, seven Tank nodes, a mixer node, two SingleProc nodes, three line nodes, a Portector node, an Assembly node and a Source node. The fourth graph G4 includes two FluidSource nodes, two Tank nodes, one mixer node, one Portector node, two SingleProc nodes, one Assembly node, and one Drain node. Although the third graph G3 and the fourth graph G4 have the same nodes, the connection relationships between the nodes are not completely the same, although they have the same parts. The map extracting means 150 extracts the same subgraph between the two production models based on the third map G3 and the fourth map G4, and as shown in fig. 5, the subgraph in which the third map G3 and the fourth map G4 are the same is a second subgraph S2 including two FluidSource nodes, two Tank nodes, one mixer node, one portationer node, and one Assembly node. The FluidSource and the Tank are node types, the FluidSource is an inlet, the Tank is a buffer node area, the Assembly is an Assembly type, the Portioner is a split type, and the mixer is a mixing and stirring device.

The capture device 120 analyzes the modeling intent based on the user's input on the modeling software 200 to obtain the user's modeling information, including which components are needed for the ongoing modeling operation, with connections between these components. As shown in fig. 6, the first input f1 of the user on the modeling software 200 is a fluudsource node, the second input f1 is a sink node continuously established on the fluudsource node, the third input f3 is another sink node continuously established on the fluudsource node, and the fourth input f4 is another sink node continuously established on the fluudsource node. The user sequentially sends a first input f1, a second input f2, a third input f3 and a fourth input f4 on the modeling software 200 to the matching device 180, the matching device 180 sequentially scores recommendation values of the first input f1, the second input f2, the third input f3 and the fourth input f4, and a subgraph is pushed as long as the matching recommendation value of a certain input and the subgraph is greater than a recommendation value threshold. For example, for the second sub-graph S2, the matching device 180 recommends the value d1 of 0.077 for the first input f1, the matching device 180 recommends the value d2 of 0.231 for the second input f2, the matching device 180 recommends the value d3 of 0.308 for the third input f3, the matching device 180 recommends the value d4 of 0.462 for the fourth input f4, when the matching device 180 obtains the recommended value d4 of 0.462, it is determined that d4 is greater than the recommended threshold of sub-graph S2, the sub-graph S2 is pushed, the sub-graph S2 is retrieved from the second database DB2, and finally, the production model associated with the second sub-graph S2 is recommended to the user in the recommending device 140. The second database DB2 has therein a second sub-graph S2, a third sub-graph S3, a fourth sub-graph S4, and so on.

Further, the user selection means 190 provides a selection interface to the user, and the matching means 180 matches the subgraph in the database based on the user's selection.

Preferably, the step 1 is performed offline, and the steps S2 and S3 are performed online.

Further, the acts: matching subgraphs in the database based on user selection.

A fifth aspect of the invention provides a computer-readable medium having stored thereon computer-executable instructions that, when executed, cause at least one processor to perform the method according to the first aspect of the invention

The help files and instantiations provided by the modeling software can only help the user to know the basic operation of the modeling software. To build a more efficient production model, the user needs to know the domain knowledge of many target industries. The modeling mechanism provided by the invention can provide a model map in the target industry field, and is more effective, smoother and more convenient.

The production modeling mechanism provided by the invention can utilize the existing production model, so that the production models in the same or similar fields can share some common maps. The present invention provides a way to efficiently extract values in these models, applying these extracted knowledge in guiding the modeling process.

While the present invention has been described in detail with reference to the preferred embodiments, it should be understood that the above description should not be taken as limiting the invention. Various modifications and alterations to this invention will become apparent to those skilled in the art upon reading the foregoing description. Accordingly, the scope of the invention should be determined from the following claims. Furthermore, any reference signs in the claims shall not be construed as limiting the claim concerned; the word "comprising" does not exclude the presence of other devices or steps than those listed in a claim or the specification; the terms "first," "second," and the like are used merely to denote names, and do not denote any particular order.

Claims

The production modeling method comprises the following steps:

s1, converting the existing production model into a map, and extracting the same subgraphs among a plurality of existing production model maps;

s2, analyzing the modeling intention based on the input of the user on the modeling software, and capturing modeling events to extract an event stream based on the modeling intention;

and S3, matching the sub-graph in the database based on the modeling intention event stream, and recommending a production model related to the sub-graph to the user based on the sub-graph so as to support the user to model.
The production modeling method of claim 1, further comprising, before said step S1, the steps of:

the existing production models are classified according to the technical field,

wherein the step S3 further includes the following steps:

matching subgraphs in the database based on the modeling intent event stream and the technical field thereof.
The production modeling method of claim 1, wherein said step S2 further comprises the steps of:

redundant information in the user input on the modeling software other than the event stream is removed.
The production modeling method of claim 1, further comprising the steps of: matching subgraphs in the database based on user selection.
The production modeling method of claim 1, wherein said step 1 is performed off-line, and said steps S2 and S3 are performed on-line.
The production modeling method of claim 1, wherein the step 1 further comprises the steps of:

and extracting layout information of the production model, wherein the layout information comprises a target component and attribute characteristics thereof and a connection relation between the target component, and the attribute characteristics comprise a name, a type and parameters.
Production modeling apparatus, comprising, among others:

a conversion device which converts an existing production model into a map;

a map extraction device that extracts the same subgraph among a plurality of existing production model maps;

a capturing device which analyzes the modeling intention based on the input of the user on the modeling software, and captures the modeling event to extract an event stream based on the modeling intention;

a subgraph matching means that matches subgraphs in the database based on the modeling intent event stream;

and the recommending device recommends the production model related to the subgraph to the user based on the subgraph so as to support the user to model.
The production modeling apparatus of claim 7, further comprising:

a classification device which classifies the existing production models according to the technical field,

wherein the graph matching means is further for matching sub-graphs in the database based on the modeling intent event stream and the technical field thereof.
The production modeling apparatus of claim 7, further comprising:

a data cleansing means that removes redundant information other than the event stream in the user's input on the modeling software.
The production modeling apparatus of claim 7, wherein the matching means is further configured to match sub-graphs in the database based on a user's selection.
The production modeling apparatus of claim 7, wherein the map extracting apparatus is further configured to extract layout information of the production model, the layout information including a target component and its attribute feature and a connection relationship between the target component, wherein the attribute feature includes a name, a type and a parameter.
A production modeling system, comprising:

a processor; and

a memory coupled with the processor, the memory having instructions stored therein that, when executed by the processor, cause the electronic device to perform acts comprising:

s1, converting the existing production model into a map, and extracting the same subgraphs among a plurality of existing production model maps;

s2, analyzing the modeling intention based on the input of the user on the modeling software, and capturing modeling events to extract an event stream based on the modeling intention;

and S3, matching the sub-graph in the database based on the modeling intention event stream, and recommending a production model related to the sub-graph to the user based on the sub-graph so as to support the user to model.
The production modeling system of claim 12, wherein said act S1 is preceded by:

the existing production models are classified according to the technical field,

wherein the act S3 further comprises the steps of:

matching subgraphs in the database based on the modeling intent event stream and the technical field thereof.
The production modeling system of claim 12, wherein said act S2 further comprises:

redundant information in the user input on the modeling software other than the event stream is removed.
The production modeling system of claim 12, wherein the actions: matching subgraphs in the database based on user selection.
The production modeling system of claim 12, wherein the act S1 is performed off-line, and the acts S2 and S3 are performed on-line.
The production modeling system of claim 12, wherein said act S1 further comprises:

and extracting layout information of the production model, wherein the layout information comprises a target component and attribute characteristics thereof and a connection relation between the target component, and the attribute characteristics comprise a name, a type and parameters.
A computer program product tangibly stored on a computer-readable medium and comprising computer-executable instructions that, when executed, cause at least one processor to perform the method of any one of claims 1 to 6.
A computer-readable medium having stored thereon computer-executable instructions that, when executed, cause at least one processor to perform the method of any one of claims 1 to 6.