CN114365136A

CN114365136A - Method and device for calibrating a simulation model of a production line

Info

Publication number: CN114365136A
Application number: CN201980100035.3A
Authority: CN
Inventors: 陈雪
Original assignee: Siemens AG
Current assignee: Siemens AG
Priority date: 2019-09-29
Filing date: 2019-09-29
Publication date: 2022-04-15
Also published as: WO2021056590A1

Abstract

A method for calibrating a simulation model of a production line, comprising: obtaining a simulation model (101) of a production line; converting the simulation model into a formal model, which is used for formally describing the simulation model (102); obtaining historical production data (103) generated by a production line in an actual production process; performing consistency check on the formal model and the historical production data to determine whether the product processes in the formal model are consistent with the product processes represented by the historical production data (104); and adjusting (105) the simulation model when the product process in the formal model is determined to be inconsistent with the product process represented by the historical production data. By the method, the accuracy of the simulation model and the quality of factory digitization are improved, and the method has wide universality and effectiveness.

Description

Method and device for calibrating a simulation model of a production line

Technical Field

The present disclosure relates to the field of industrial manufacturing, and more particularly, to methods, apparatus, computing devices, computer-readable storage media, and program products for calibrating a simulation model of a production line.

Background

Building simulation models for production lines in a plant is an important step in achieving plant digitization. The simulation model of the production line can perform virtual simulation on the production and manufacturing process of the product in the production line, simulate the production scene and provide reference for the decision of management or technical personnel, so that the product research and development and manufacturing efficiency of manufacturing enterprises can be improved. Generally, building a simulation model of a manufacturing line depends primarily on the structure of the manufacturing line and the manufacturing processes or process routes that produce the products in the bill of materials (BOM). Specifically, simulation modules for the respective components are first established based on the components of the production line and configuration parameters are set for the respective simulation modules, and then the operation order of the simulation modules is determined based on the production order of the products. When a simulation model is used to simulate the production of a product, appropriate inputs (e.g., the time the product entered the production line) are provided to the simulation model and a desired simulation result (e.g., the production of the production line over a period of time) may be obtained.

Disclosure of Invention

In fact, the production line simulation model constructed in the above manner tends to be an ideal model. On the one hand, in an actual factory floor, the actual process of a product typically does not strictly follow the predefined process in the manufacturing bill of materials. For example, when a product is returned to a previous machine during a manufacturing process to perform a process step, there may be some duplicative manufacturing processes. Also for example, there may be some processes that do not require a machine or workstation but take some time, such as an inspection process, which refers to a process where a manufacturing engineer performs a technical quality check, reconfirmation, or signature on a certain process step during the processing of a product. On the other hand, the configuration parameters of the simulation model are usually set to fixed values empirically, but the fixed values do not represent the actual production process well. For example, in an actual production process, the repair time ttr (time to repair) and the failure free time tbf (time Between failure) of a certain device may change according to the situation, and setting them as fixed values inevitably affects the simulation result.

Therefore, a simulation model constructed only according to the production line structure and the product procedures in the manufacturing bill of materials has ideal procedures and fixed configuration parameters, and the simulation result obtained by using the simulation model is not accurate enough, so that the decision made according to the simulation result is influenced.

A first embodiment of the present disclosure proposes a method for calibrating a simulation model of a production line, comprising: obtaining a simulation model of the production line, the simulation model being used for simulating a production process of the production line and comprising a plurality of simulation modules corresponding to components of the production line, the simulation modules being provided with corresponding configuration parameters; converting the simulation model into a formal model, wherein the formal model is used for performing formal description on the simulation model; obtaining historical production data generated by a production line in an actual production process; carrying out consistency check on the formal model and the historical production data to determine whether the product processes in the formal model are consistent with the product processes reflected by the historical production data; and adjusting the simulation model when the product process in the formal model is determined to be inconsistent with the product process embodied by the historical production data.

In the embodiment, the consistency of the product procedures of the simulation model of the production line is checked by using the historical production data generated by the production line in the factory in the actual production process, and the simulation model is adjusted by using the checking result, so that the accuracy of the simulation model is improved and is closer to the actual production process, and therefore, a more accurate simulation result can be obtained, and the digitization quality of the factory is improved. In addition, the method for calibrating the simulation model of the production line is suitable for various simulation models constructed by different modeling software, and therefore has wide universality and effectiveness.

A second embodiment of the present disclosure proposes an apparatus for calibrating a simulation model of a production line, comprising: a simulation model obtaining unit configured to obtain a simulation model of a production line, the simulation model being used for simulating a production process of the production line and including a plurality of simulation modules corresponding to respective components of the production line, the simulation modules being provided with respective configuration parameters; a simulation model conversion unit configured to convert the simulation model into a formal model, the formal model being used for formally describing the simulation model; a production data acquisition unit configured to acquire historical production data generated by the production line in an actual production process; a consistency check unit configured to perform consistency check on the formal model and the historical production data to determine whether the product processes in the formal model and the product processes embodied by the historical production data are consistent; and a simulation model adjusting unit configured to adjust the simulation model when it is determined that the product processes in the formal model and the product processes embodied by the historical production data are not consistent.

A third embodiment of the present disclosure proposes a computing device including: a processor; and a memory for storing computer-executable instructions that, when executed, cause the processor to perform the method of the first embodiment.

A fourth embodiment of the disclosure proposes a computer-readable storage medium having stored thereon computer-executable instructions for performing the method of the first embodiment.

A fifth embodiment of the disclosure proposes a computer program product, tangibly stored on a computer-readable storage medium, and comprising computer-executable instructions that, when executed, cause at least one processor to perform the method of the first embodiment.

Drawings

The features, advantages and other aspects of various embodiments of the present disclosure will become more apparent by referring to the following detailed description in conjunction with the accompanying drawings, in which several embodiments of the present disclosure are shown by way of illustration and not limitation, wherein:

FIG. 1 illustrates a method for calibrating a simulation model of a production line according to one embodiment of the present disclosure;

FIG. 2 illustrates an example production line simulation model according to one embodiment of this disclosure;

FIG. 3 illustrates a method for calibrating the production line simulation model of FIG. 2 according to one embodiment of the present disclosure;

FIG. 4 illustrates a Petri Net model constructed in accordance with the production line simulation model of FIG. 3;

FIG. 5 illustrates inspection results of a consistency check of the Petri Net model of FIG. 4 with example historical production data;

FIG. 6 shows a production line simulation model adjusted according to the inspection results of FIG. 5;

FIG. 7(a) shows the fitting results of fitting the repair time TTR of the device at the workstation 202 of FIG. 2 with a normal distribution function;

fig. 7(b) fitting results of fitting the repair time TTR of the device at the workstation 202 in fig. 2 using a poisson distribution function;

fig. 8 shows the fitting results of fitting the time without failure TBF of the device at the station 202 in fig. 2 with an exponential distribution function;

FIG. 9 illustrates an exemplary architecture of a system for calibrating simulation models of a production line according to the embodiment of FIG. 3;

FIG. 10 illustrates an apparatus for calibrating a simulation model of a production line according to another embodiment of the present disclosure; and

FIG. 11 illustrates a block diagram of a computing device for calibrating a simulation model of a production line according to one embodiment of the present disclosure.

Detailed Description

Various exemplary embodiments of the present disclosure are described in detail below with reference to the accompanying drawings. Although the exemplary methods, apparatus, and devices described below include software and/or firmware executed on hardware among other components, it should be noted that these examples are merely illustrative and should not be considered as limiting. For example, it is contemplated that any or all of the hardware, software, and firmware components could be embodied exclusively in hardware, exclusively in software, or in any combination of hardware and software. Thus, while the following describes example methods and apparatus, persons of ordinary skill in the art will readily appreciate that the examples provided are not intended to limit the manner in which the methods and apparatus may be implemented.

Furthermore, the flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of methods and systems according to various embodiments of the present disclosure. It should be noted that the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the flowchart illustrations and/or block diagrams, and combinations of blocks in the flowchart illustrations and/or block diagrams, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.

As used herein, the terms "include," "include," and similar terms are open-ended terms, i.e., "including/including but not limited to," meaning that additional content may also be included. The term "based on" is "based, at least in part, on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment," and the like.

FIG. 1 illustrates a method for calibrating a simulation model of a production line according to one embodiment of the present disclosure. A production line may be any production line in a factory used to manufacture a product, and its simulation model may be built via any suitable modeling tool or software. Referring to fig. 1, method 100 begins at step 101. In step 101, a simulation model of a production line is obtained, the simulation model being used for simulating a production process of the production line and comprising a plurality of simulation modules corresponding to respective components of the production line, the simulation modules being provided with respective configuration parameters. As mentioned above, the simulation model of the production line can perform virtual simulation on the production and manufacturing processes of the product in the production line, simulate the production scenario, and obtain the desired simulation result, so as to be used for the purposes of scheduling production, planning the design of the production line, or developing the product. The components of the production line may include, for example, manual stations and machine equipment, and the configuration parameters may include, for example, processing time or processing time of the product to be processed at the manual station or machine equipment, time of the product to be processed entering the manual station or machine equipment, time of the product to be processed leaving the manual station or machine equipment, time to repair time ttr (time to repair) and time to failure (time Between failure) of the machine equipment, waiting time of the manual station or machine equipment due to temporary shortage of raw material supply, and the like.

With continued reference to fig. 1, next, the method 100 proceeds to step 102. In step 102, the simulation model is converted into a formal model, and the formal model is used for performing formal description on the simulation model. The formal description refers to the description of the state behavior and other characteristics of the system by using a language with preset syntactic semantic definitions. In some embodiments, the formal model is a Petri net model. The Petri network model can be used for describing a discrete parallel system and is composed of a place (P), a transition (T), directed arcs and tokens (token), wherein the place (P) represents state elements, the transition represents change or event elements, the directed arcs can be changed from the place (P) to the place (P) or changed from the place (P) to the place (T), and the tokens are dynamic objects in the places (P) and can be moved from one place (P) to another place (P) through the transition.

Whereas the Petri Net model is able to describe the behavior of a discrete parallel system, it can be used to represent the processes of a product in a production line. In converting a simulation model of a production line to a formal model, transitions may be used to represent processing or handling of a product at each process, libraries may be used to represent states of a product before and after each process, tokens in a library may represent products, and directed arcs may represent tokens (products) in a library passing from one state to another via a transition (e.g., a process). For example, the process of polishing a product can be described as: the depot P1 (product state before polishing) was connected via a directed arc to transition T1 (polishing) and then to depot P2 (product state after polishing). In other embodiments, other formal models may be employed as well, provided that they are capable of abstractly describing the simulation model as a structure including states, events, and relationships between them.

In some embodiments, converting the simulation model to a formal model further comprises (not shown in fig. 1): extracting model wiring information from the simulation model, wherein the model wiring information comprises identification information and an operation sequence of a plurality of simulation modules; splitting the model wiring information into a plurality of groups of sub-line information, wherein each group of sub-line information comprises identification information and an operation sequence of each simulation module which is linearly connected from the simulation module representing the initial position of the production line to the simulation module representing the end position of the production line; and constructing a formal model according to the multiple groups of sub-line information.

In a factory floor, a production line often has a net structure, i.e., the components of the production line are in a net-like connected relationship, e.g., a production line may have two or more processes performed in parallel without interfering with each other. Accordingly, the simulation modules in the simulation model of the production line are also in a meshed connected relationship. In order to meet the requirements for building a formalized model, the mesh structure of the simulation model needs to be split into end-to-end linearly connected structures. The model wiring information may be first extracted from the simulation model and derived in a tabular form through an Application Program Interface (API) of the modeling software or a modeling language supported by the modeling software. The model wiring information is used for representing the procedures of the product in the production process. In some embodiments, simulation modules corresponding to components of the production line may be used as objects, and the model routing information includes information on four aspects, namely, an object name, an object type, a previous object, and a subsequent object, wherein the object name and the object type represent identification information of the simulation modules, and the previous object and the subsequent object represent an operation sequence of the simulation modules. Further, when the simulation module represents a machine tool or a human workstation, the model wiring information may also include the product, part, or raw material processed or processed at the machine tool or human workstation and the corresponding processing or processing time. In some embodiments, the processes (including machining processes and manual operations) implemented by each component of the production line may also be targeted in the model wiring information.

After obtaining the model routing information, the mesh structure represented by the model routing information is split into a plurality of sub-lines from end to end. For each sub-line it means for a certain product a linear non-branching process route of a certain raw material from a start position of the production line through the various processes to an end position of the production line. Starting from the simulation module representing the initial position of the production line in the model wiring information, searching for a subsequent simulation module (i.e. a subsequent object) of the simulation module according to the model wiring information, then searching for the subsequent simulation module again according to the searched subsequent simulation module, and repeating the steps until the searched simulation module represents the end position of the production line. The process route from the simulation module representing the start position of the production line to the simulation module representing the end position of the production line is a sub-line. In splitting a mesh into multiple child lines, there may be a bifurcation at a certain simulation module, i.e. there may be more than one simulation module following a certain simulation module. Alternatively, and more generally, a simulation model of a production line has a plurality of simulation modules representing starting positions. Thus, for each sub-line, only one subsequent simulation module or only one simulation module representing the starting position is selected at each bifurcation to ensure that the sub-line is linear and unbranched. Subsequent simulation modules that are not selected at the bifurcation or simulation modules representing the starting position that are not selected may be included in the further child line until all simulation modules are included in the child line.

In this way, the mesh structure represented by the model wiring information can be split into a plurality of sub-lines. Similar to the model wiring information, the sub-line information may also take the simulation module as an object, which may include information on the time when the object was found, the sub-line number, the object name, and the object type. Sub-line numbers may be used to indicate process lines, with the same sub-line number indicating the same process line. The sub-line numbers of the same group of sub-line information are the same. In each group of sub-line information, the object name and the object type represent identification information of the simulation modules, and the time for finding the object represents the operation sequence between the simulation modules on the process route represented by the sub-line.

In some embodiments, one production line may be used to produce different products, and the production sequence may be different for different products. In such embodiments, multiple sets of sub-line information need to be obtained based on both model wiring information and product type. The sub-line number in the sub-line information may also be used to indicate the product type.

After obtaining the plurality of sets of sub-line information, a formal model is constructed according to the plurality of sets of sub-line information. The formal model can embody a plurality of linear process routes respectively corresponding to each group of sub-line information in the plurality of groups of sub-line information. Each linear process route can be a process route of the same product or a process route of different kinds of products. In some embodiments, the formal model is a Petri net model, and the construction of the Petri net model can be realized by using algorithms such as Inductive Miner, Alpha Miner and the like.

With continued reference to FIG. 1, after step 102, the method proceeds to step 103 where historical production data generated by the production line during the actual production process is obtained. The purpose of this step is to obtain information on each product and the process steps that the product has gone through during the actual production process and the corresponding time stamp. A large amount of production data (also referred to as process data) is generated during the actual production process of a production line. The production data may include data indicative of a complete production process of a machine through which a product or its raw material is circulated, a manual station, a time of entering or leaving the machine or manual station, a process completed at the machine or manual station, a time of occurrence of a malfunction of the machine, a time of elimination of a malfunction of the machine, and the like. These historical production data can truly represent the actual production process of the product on the production line and can therefore be used to verify or verify the product process in the formal model of the simulation model. Historical production data may be obtained for any period of time that the production line is in the same state (e.g., the production line is not altered), such as for a week or a month.

In some embodiments, the historical production data may be pre-processed. Production data may be stored in different formats and the data server that stores the data may not be on the same physical machine, possibly even a handwritten log. Thus, the disparate data formats and the distributed storage of data contribute to the heterogeneity of these historical production data. Therefore, after obtaining historical production data generated by the production line in the actual production process, it is first necessary to integrate heterogeneous production data according to certain rules to form an efficient and centralized historical production data set, and convert the historical production data set into data targeted for a product or its raw materials, such as names of various processes that a certain raw material passes through, passing time, and the like. In addition, because some missing data or non-target data may exist in the production data, the integrated historical production data can be cleaned to remove wrong or unreasonable data and supplement the missing data.

It should be noted that although step 103 is performed after step 102 in the above description, step 103 may be performed before

step

101 or 102, or may be performed in synchronization with

steps

101 and 102.

Next, the method 100 proceeds to step 104 where a consistency check is performed on the formal model and the historical production data to determine whether the product processes in the formal model and the product processes embodied by the historical production data are consistent. As described previously, the formal model includes a plurality of linear process routes corresponding to each of the plurality of sets of sub-line information, which can embody the processes of the product in the simulation model, and the processes of the product in the actual production process can be obtained through the integration and cleaning of the historical production data. By comparing the product procedures in the formal model with the actual product procedures, it can be determined whether the product procedures in the simulation model are accurate. Since the historical production data is production data in a certain period of time, it usually includes multiple sets of production data (i.e. production data of multiple products) of the same product, and each set of production data can represent a complete process of the product. Therefore, each set of production data of the product can be subjected to consistency check with the formal model, and whether the product processes in the formal model are consistent with the product processes reflected by the historical production data can be determined according to preset rules and based on a plurality of check results of the consistency check of the plurality of sets of production data and the formal model. For example, a threshold ratio, such as X%, may be set, and when the inspection result exceeding the threshold ratio in the inspection results of the consistency check of the plurality of sets of production data with the formal model indicates that there is a certain process that does not exist in the formal model in the production data, it may be considered that the product process in the formal model does not coincide with the product process represented by the historical production data.

In some embodiments, consistency checking the formal model and the historical production data further comprises: and replaying the historical production data on the formal model, wherein the replay is used for determining whether a product process which is not embodied by the historical production data exists in the formal model or whether the historical production data embodies a product process which is not embodied in the formal model. Replay is one such process: taking historical production data and a formal model as input, executing the historical production data on the formal model again, if the trace (product process) is completely fitted with the model, meaning that the trace can completely occur according to the action sequence executed by the model; if the trace and the model are not completely fitted, the replay is stopped when the process that the trace and the model do not accord is replayed. By replaying the historical production data, it can be determined whether a product process not reflected by the historical production data exists in the formal model or whether the historical production data reflects a product process not present in the formal model.

The method 110 then proceeds to step 105 where the simulation model is adjusted when it is determined that the product sequence in the formal model and the product sequence embodied in the historical production data are not consistent. The simulation model is adjusted by using the consistency check result of the formal model and the historical production data, so that the product process in the simulation model is closer to the product process in the actual production process. Because the product process is embodied in the simulation model by the simulation module and the position of the simulation module, the adjustment of the simulation model may include the addition, deletion, and/or modification of the simulation module in the simulation model.

In some embodiments, step 105 further comprises (not shown in fig. 1): determining differences between product processes in the formal model and product processes embodied by historical production data; determining product procedures that need to be added to and/or deleted from the simulation model based on the differences; and adjusting the simulation model based on the determined product procedures that need to be added and/or deleted.

When it is determined that the product process in the formal model is inconsistent with the product process represented by the historical production data, the checking result of the consistency check may include a difference between the product process in the formal model and the product process represented by the historical production data, which may include the product process in which the formal model is inconsistent with the historical production data and a location where the product process is located in the entire process flow.

Due to the complexity of the actual production process, the production processes represented by the historical production data but not represented in the formal model are not necessarily missing processes in the simulation model (e.g., there is an error in the actual production process), and the production processes represented by the formal model but not represented by the historical production data are not necessarily redundant processes in the simulation model (e.g., a certain product has a low rework rate). Therefore, it is desirable to determine product procedures that need to be added to and/or deleted from the simulation model based on the determined differences. This process may be performed automatically by software or selected empirically by a technician.

After determining the product sequence that needs to be added to and/or removed from the simulation model, the simulation model may be adjusted according to where the sequence is located throughout the process flow. In some embodiments, the simulation model may be automatically adjusted. Specifically, when a process needs to be added, the process to be added or the machine equipment/manual workstation corresponding to the process may be used as an object, and related information such as an object name, an object type, a previous object, a subsequent object and the like may be generated, where the object name may be the name of the process to be added or the machine equipment/manual workstation corresponding to the process, the object type may be obtained through a preset lookup table, and the previous object and the subsequent object may be determined according to a position where the process to be added is located in the entire process flow. And then, modifying the model wiring information of the simulation model according to the information so as to update the model wiring information. When a process needs to be deleted, the name of the process to be deleted or the name of the machine equipment/manual station corresponding to the process can be used as an object name, the object is searched from the model wiring information of the simulation model, and the related information of the object is deleted from the model wiring information, so that the model wiring information is updated. The updated simulation model of the production line may be automatically generated based on the updated model wiring information. In some embodiments, the simulation model may also be adjusted manually by the modeling software. Therefore, wiring calibration of the simulation model of the production line is achieved.

In some embodiments, method 100 further comprises (not shown in fig. 1): obtaining historical data of configuration parameters from historical production data; determining the frequency distribution characteristics of the configuration parameters in the actual production process based on the historical data of the configuration parameters; and setting the frequency distribution characteristics of the configuration parameters in the adjusted simulation model.

The historical production data of the production line can reflect the production process of the product, and also comprises actual data of configuration parameters of each simulation module in the simulation model. The configuration parameters may include, for example, the processing time or processing time of the product to be processed at the manual station or machine equipment, the time the product to be processed enters the manual station or machine equipment, the time the product to be processed leaves the manual station or machine equipment, the repair time TTR and the time without failure TBF of the machine equipment, the waiting time of the manual station or machine equipment, etc. In order to bring the simulation model closer to the actual production line, calibration of the configuration parameters is also possible.

After obtaining the historical production data, the historical data of the configuration parameters can be obtained by performing simple calculations on some of the historical production data. For example, when the repair time TTR and the non-failure time TBF of a certain machine device are concerned, a plurality of failure occurrence times and a plurality of corresponding failure elimination times of the machine device may be obtained in historical production data, the plurality of historical repair times TTR of the machine device may be obtained by subtracting the corresponding failure occurrence time from each failure elimination time, and the plurality of historical non-failure times TBF of the machine device may be obtained by subtracting the last failure elimination time from each failure occurrence time. In other embodiments, historical data for other configuration parameters (e.g., wait time at a machine or human workstation, time to enter or leave a machine or human workstation, etc.) may also be obtained.

Subsequently, a frequency distribution characteristic of the configuration parameter during the actual production process may be determined based on historical data of the configuration parameter derived from historical production data. In some embodiments, determining the frequency distribution characteristic of the configuration parameter in the actual production process may further include: fitting the historical data of the configuration parameters by using different distribution functions; and determining the distribution function with the highest goodness of fit as the frequency distribution characteristic of the configuration parameter. Common distribution functions include, for example, binomial distribution functions, laplacian distribution functions, gamma distribution functions, exponential distribution functions, poisson distribution functions, normal distribution functions, and the like. The historical data of the configuration parameters can be respectively fitted with some common distribution functions, and the parameter values of the distribution functions are obtained through calculation in the fitting process. The distribution function with the highest goodness-of-fit may be determined, for example, by a test such as a K-S test, a chi-squared test, or the like.

After obtaining the frequency distribution characteristics of the configuration parameters in the actual production process, the frequency distribution characteristics of the configuration parameters, i.e. the distribution function with the highest goodness of fit and the parameter values of the function, can be set in the wiring-calibrated simulation model automatically or manually through a modeling tool.

In some embodiments, method 100 further comprises (not shown in fig. 1): verifying the adjusted simulation model by using the frequency distribution characteristics of the configuration parameters; and further adjusting the range of the configuration parameter based on the verification result. To improve the robustness of the simulation model, the adjusted simulation model may also be verified after both the wiring and configuration parameters of the simulation model are calibrated. A plurality of random values across the distribution function may be generated as values of the configuration parameters according to the distribution function to which the configuration parameters conform by using a technique such as a monte carlo method, and the simulation verification of the production line by using the simulation model may be performed to obtain a verification result, where the verification result may include a relationship between a change in the configuration parameters and a change in the simulation result. The verification result may indicate that the simulation model is sensitive to a change of a certain configuration parameter, that is, a change of a certain configuration parameter may cause a large change of the simulation result, in which case, the value range of the configuration parameter may be further adjusted, so that the simulation model has higher robustness.

The consistency of product procedures is checked on the simulation model of the production line by using the historical production data generated by the production line in a factory in the actual production process, and the simulation model is adjusted by using the checking result, so that the accuracy of the simulation model is improved, the simulation model is closer to the actual production process, the more accurate simulation result can be obtained, and the digitization quality of the factory is improved. In addition, the method for calibrating the simulation model of the production line is suitable for various simulation models constructed by different modeling software, and therefore has wide universality and effectiveness.

In addition, in the prior art, the difference between the two outputs is usually reduced by comparing the outputs of the simulation model and the actual production line under the same input and adjusting a certain configuration parameter in the simulation model according to experience, which is not only time-consuming and labor-consuming, but also can cause different configuration parameters to be adjusted each time. According to the method for calibrating the simulation model of the production line, the frequency distribution characteristics of the configuration parameters are obtained by utilizing the historical production data generated in the actual production process, so that more accurate configuration parameters can be set in the simulation model, and the accuracy and the effectiveness of the simulation model are further improved.

The method for calibrating a simulation model of a production line shown in FIG. 1 is described below with reference to a specific embodiment. FIGS. 2-8 illustrate the method for calibrating a production line simulation model, as an example.

FIG. 2 illustrates an example production line simulation model 200 according to one embodiment of this disclosure. For simplicity, in the present embodiment, the actual production line to which the simulation model 200 corresponds is used to produce only one type of product. In simulation model 200,

simulation modules

201 and 204, referred to as sources, represent starting locations in a production line, for example, a supply of raw materials or a machine that produces parts. The

simulation modules

202, 203, 205, and 207 represent stations in a production line at each of which a process is performed on raw materials or parts. The workstation may be a robotic device or a manual workstation. The simulation module 206 represents an assembly station in a production line where parts are assembled. The assembly station may also be a robotic device or a manual station. The simulation module 208, referred to as an exit, represents an end location in the production line, e.g., where the product leaves the production line. The connecting lines with arrows between the simulation modules represent the flowing direction of the materials. As can be seen from simulation model 200, the raw materials or parts produced at simulation module 201 are delivered to

workstations

202 and 203 in sequence, and the raw materials or parts produced at simulation module 204 are delivered to workstation 205. After passing through the processes at the

workstations

202, 203 and 205, respectively, the two parts produced are transported to the assembly station 206 and then exit the production line via the processes at the assembly station 206 and the workstation 207, respectively, to the simulation module 208.

Referring now to FIG. 3, FIG. 3 illustrates a method for calibrating a simulation model of the production line of FIG. 2 according to one embodiment of the present disclosure. First, in step 301 of method 300, a simulation model of a production line is obtained. In the present embodiment, the simulation model of the production line is the production line simulation model 200 of the example shown in fig. 2. Next, in step 302, model wiring information is extracted from the simulation model, the model wiring information including identification information and an order of operation for the plurality of simulation modules. In this embodiment, the wiring information is saved as a.csv file. The extracted model wiring information according to the present embodiment is shown in table 1 below.

Object name

Object type

Previous object

Subsequent object

Source 201	Source	Is free of	Workstation 202
Workstation 202	Work station	Source	201	Workstation 203
Workstation 203	Work station	Workstation	202	Assembly station 206
Assembly station 206	Assembly station	Workstation	203	Workstation 207
Workstation 205	Work station	Source	204	Assembly station 206
Source 204	Source	Is free of	Workstation 205
Connecting wire	Connecting wire	Source	201	Workstation 202
Connecting wire 1	Connecting wire	Workstation	202	Workstation 203
Connecting wire 2	Connecting wire	Source	204	Workstation 205
Connecting wire 3	Connecting wire	Workstation	203	Assembly station
Connecting wire 4	Connecting wire	Workstation	205	Assembly station
Workstation
207	Work station	Assembly station	An outlet 208
An outlet 208	An outlet	Workstation	207	Is free of
Connecting wire 5	Connecting wire	Assembly station	206	Workstation 207
Connecting wire 6	Connecting wire	Workstation	207	An outlet 208

TABLE 1

As seen from Table 1, with each simulation module in simulation model 200 as an object, the model routing information includes information on four aspects, namely, an object name, an object type, a previous object, and a subsequent object, wherein the object name and the object type represent identification information of the simulation module, and the previous object and the subsequent object represent an operation order of the simulation module. In other embodiments, processing time or processing time at each workstation or assembly station may also be included in the model wiring information.

Next, the method 300 proceeds to step 303, where a plurality of sets of sub-line information are obtained from the model wiring information, each set of sub-line information including identification information and an operation order of simulation modules linearly connected from a simulation module representing a start position of the production line to a simulation module representing an end position of the production line. In the present embodiment, as shown in Table 1, the simulation model 200 has two

sources

201 and 204, which are simulation modules representing the starting position of the production line. Starting from the source 201, the subsequent object of the source 201 is found from table 1 as the workstation 202, and then the subsequent object of the workstation 202 is continuously found from table 1 as the workstation 203, the subsequent object of the workstation 203 is the assembly station 206, the subsequent object of the assembly station 206 is the workstation 207, and the subsequent object of the workstation 207 is the exit 208, which represents the end position of the production line, so that a linear non-branched process route representing the process from the source 201 to the exit 208 through the processes at the workstation 202, the workstation 203, the assembly station 206 and the workstation 207 is obtained. In a similar manner, looking up subsequent objects in Table 1 starting from another source 204 of simulation model 200, another process route can be obtained that represents a linear non-branching from source 204 through various processes at workstation 205, assembly station 206, and workstation 207 to exit 208. In this embodiment, two linear unbranched sub-lines may be obtained, with corresponding sub-line information as shown in table 2 below. In this embodiment, the obtained sub-line information is also saved as a.csv file.

Time	Number of sub-line	Object name	Object type
9/17/2019 18:00	1	Source 201	Source
9/17/2019 18:05	1	Workstation 202	Work station
9/17/2019 18:10	1	Workstation 203	Work station
9/17/2019 18:15	1	Assembly station 206	Assembly station
9/17/2019 18:20	1	Workstation 207	Work station
9/17/2019 18:25	1	An outlet 208	An outlet
9/17/2019 18:30	2	Source 204	Source
9/17/2019 18:35	2	Workstation 205	Work station
9/17/2019 18:40	2	Assembly station 206	Assembly station
9/17/2019 18:45	2	Workstation 207	Work station
9/17/2019 18:50	2	An outlet 208	An outlet

TABLE 2

As seen from table 2, for each of the two sub-lines, the time, the sub-line number, the object name, and the object type of the object to be found are recorded from top to bottom according to the sequence of the found object, the object name and the object type represent identification information of the simulation module, and the time of the found object represents the operation sequence between the simulation modules located on the process route represented by the sub-line.

With continued reference to FIG. 3, in a next step 304, a formal model is constructed from the obtained sets of sub-line information. In this embodiment, the formal model is a Petri net model, and the Petri net model may be constructed by using a process mining algorithm such as Inductive Miner and Alpha Miner. Before the Petri net model is constructed, the file format of the file with the multiple groups of sub-line information can be converted into a required file format. In this embodiment, a file of the csv file format having a plurality of sets of sub-line information is converted into a file of the XES file format. FIG. 4 shows a Petri Net model constructed according to the production line simulation model of FIG. 3. In the Petri Net model 400 shown in FIG. 4, circles represent places (P), boxes represent transitions (T), arrows represent directed arcs, and "·" in the circles represent tokens. Transitions represent processing or handling of a product at each process, the bins represent states of the product before and after each process, tokens represent products, and directed arcs represent tokens (products) in the bins from one state to another via a transition (e.g., a process). With reference to fig. 2 and 4, transition 401 corresponds to the creation of a raw material or component at source 201, transition 402 corresponds to a process at workstation 202, transition 403 corresponds to a process at workstation 203, transition 404 corresponds to the creation of a raw material or component at source 204, transition 405 corresponds to a process at workstation 205, transition 406 corresponds to a process at assembly station 406, transition 407 corresponds to a process at workstation 407, and transition 408 corresponds to exiting the production line at exit 408.

Returning to FIG. 3, after step 301-. The log file that records the historical production data for the production line may be obtained from the memory of the different machine equipment, manually recorded log records, in the memory of the controller, etc. Next, in step 306, the historical production data is integrated and data-cleaned, and the cleaned historical production data is converted into data targeting the product or its raw material according to the product information, that is, the name of each process that the product or raw material passes through in the actual production process, and the start time and end time of each process. In the present embodiment, since the production line of the example is used only for producing one type of product, it may not be necessary to record product information. A part of the historical production data after cleaning according to the present embodiment is shown in table 3 below. In this embodiment, the historical production data obtained is also saved as a.csv file.

Example numbering	Procedure (ii)	Starting time	End time
1	Source 201	8/17/2019 18:00	8/17/2019 18:00
1	Workstation 202	8/17/2019 18:05	8/17/2019 18:10
1	Workstation 203	8/17/2019 18:10	8/17/2019 18:20
1	Assembly station 206	8/17/2019 18:15	8/17/2019 18:25
1	Workstation 207	8/17/2019 18:25	8/17/2019 18:35
1	Manual inspection	8/17/2019 18:25	8/17/2019 18:37
1	An outlet 208	8/17/2019 18:37	8/17/2019 18:37
2	Source 204	8/17/2019 18:00	8/17/2019 18:00
2	Workstation 205	8/17/2019 18:00	8/17/2019 18:15
2	Assembly station 206	8/17/2019 18:15	8/17/2019 18:25

2	Workstation 207	8/17/2019 18:25	8/17/2019 18:35
2	Manual inspection	8/17/2019 18:25	8/17/2019 18:37
2	An outlet 208	8/17/2019 18:37	8/17/2019 18:37
3	Source 201	8/18/2019 18:00	8/18/2019 18:00
3	Workstation 202	8/18/2019 18:05	8/18/2019 18:10
3	Workstation 203	8/18/2019 18:10	8/18/2019 18:20
3	Assembly station 206	8/18/2019 18:15	8/18/2019 18:25
3	Workstation 207	8/18/2019 18:25	8/18/2019 18:35
3	Manual inspection	8/18/2019 18:25	8/18/2019 18:37
3	An outlet 208	8/18/2019 18:37	8/18/2019 18:37
4	Source 204	8/18/2019 18:00	8/18/2019 18:00
4	Workstation 205	8/18/2019 18:00	8/18/2019 18:15
4	Assembly station 206	8/18/2019 18:15	8/18/2019 18:25
4	Workstation 207	8/18/2019 18:25	8/18/2019 18:35
4	Manual inspection	8/18/2019 18:25	8/18/2019 18:37
4	An outlet 208	8/18/2019 18:37	8/18/2019 18:37

TABLE 3

In table 3, the actual production process of the product on the production line to be calibrated is shown, the whole process of the product from the source of the raw material or component to the production process through the various processes and leaving the production line and the corresponding time stamp. In the present embodiment, only an example of 4 sets of production data is shown in table 3 for convenience of explanation, but in other embodiments, any set of production data may be included.

Then, in step 307 of the method 300, a consistency check is performed on the formal model and the historical production data to determine whether the product processes in the formal model and the product processes embodied in the historical production data are consistent. Before the consistency check, the file format of the file in which the washed historical production data is stored may be converted into a required file format. In this embodiment, the file in the csv file format, which stores the washed history production data, is converted into a file in the XES file format. In this embodiment, the historical production data is replayed on the Petri net model of fig. 4 to determine whether there are product processes in the Petri net model that are not reflected by the historical production data or whether the historical production data reflects product processes that are not reflected in the Petri net model. Referring to fig. 4 and table 3, taking example 1 of the production data as an example, the procedure of the product in example 1 was performed again on the Petri net model. It can be seen that the next process in example 1 is a manually inspected process when reentering transition 407 (corresponding to a process at workstation 207) on the Petri Net model, whereas there is no next process in the Petri Net model, the product leaves the production line directly at transition 408. The same results were obtained by continuing to reenact the product sequence in examples 2-4. Therefore, the product procedures in the Petri net model can be determined to be inconsistent with the product procedures reflected by the historical production data.

In step 308, differences between the product process in the formal model and the product process embodied by the historical production data are determined. As described above, in the present embodiment, after the processes in examples 1 to 4 are replayed on the Petri net model using the product, it can be determined that the product processes in the Petri net model are inconsistent with the product processes embodied in the historical production data. It is also determined from the replay results in examples 1-4 that the difference between the product sequence in the Petri Net model and the product sequence represented by the historical production data is that there is also a sequence of manual inspection in the actual production process between transition 407 (corresponding to the sequence at workstation 207 in the simulation model) and transition 408 (corresponding to exit 208 in the simulation model). In other embodiments, there may be situations where the process steps in different instances differ from the process steps in the formal model. Fig. 5 shows the checking result of the consistency check of the Petri net model of fig. 4 with the example historical production data. The dashed boxes in fig. 5 represent the differences between the production process in the Petri net model and the production data exemplified in table 3.

Next, in step 309, product procedures that need to be added to and/or deleted from the simulation model are determined based on the determined differences. In this embodiment, the manual inspection before the product leaves the production line belongs to a process that must be missing in the simulation model during the actual production process. Therefore, it is automatically determined that a process requiring manual inspection is added to the simulation model of fig. 2. In other embodiments, product procedures that need to be added to and/or deleted from the simulation model may be selected by an engineer.

Thereafter, in step 310, model wiring information for the simulation model is updated. Specifically, the process of manual inspection missing in the simulation model is used as an object, and an object name (manual inspection), an object type (manual inspection), a previous object (workstation 207), and a subsequent object (exit 208) are generated and added to the model wiring information of the simulation model in table 1. A portion of the updated model routing information is shown below in table 4, where the bold-faced portions represent modified or added content.

Object name	Object type	Previous object	Subsequent object
…	…	…	…
Workstation 207	Work station	Assembly station	Manual inspection 209
Manual inspection 209	Manual inspection	Workstation	207	An outlet 208

An outlet 208	An outlet	Manual inspection 209	Is free of
…	…	…	…
Connecting wire 6	Connecting wire	Workstation	207	Manual inspection 209
Connecting wire 7	Connecting wire	Manual inspection 209	An outlet 208

TABLE 4

In step 311, a new simulation model is generated based on the updated model routing information. In the present embodiment, a new simulation model is generated by an automatic model generation algorithm. FIG. 6 shows a production line simulation model adjusted according to the inspection results of FIG. 5. As shown in FIG. 6, a simulation module 209 is added between

simulation modules

207 and 208 of simulation model 600, which represents a process with a manual inspection 209 added between the process at workstation 207 and exit 208. In other embodiments, the simulation model may also be adjusted by an engineer through a modeling tool. Therefore, the process of carrying out wiring calibration on the simulation model of the production line by using the historical production data of the production line is realized.

Returning to FIG. 3, on the other hand, historical production data of the production line after integration and cleaning in step 306 may also be used to calibrate configuration parameters of simulation modules in the simulation model. In this embodiment, the historical production data after cleaning may also include the time to failure and the time to failure to clear the equipment at each workstation or assembly station. In step 312, historical data for the configuration parameters is obtained from the purged historical production data. In this embodiment, attention needs to be paid to the repair time TTR and the time without failure TBF of the device at the workstation 202 in the simulation model of fig. 2. A plurality of times of occurrence of faults and a corresponding plurality of times of elimination of faults of the equipment in the actual production process at the workstation 202 may be obtained from the historical production data. And subtracting the corresponding failure occurrence time from each failure elimination time to obtain a plurality of historical restoration times TTR of the equipment, and subtracting the previous failure elimination time from each failure occurrence time to obtain a plurality of historical non-failure times TBF of the equipment. In other embodiments, historical production data may be purged based on the configuration parameters of interest to obtain historical data associated with the configuration parameters.

Next, in step 313, the historical data of the configuration parameters is fitted with different distribution functions. In this embodiment, a plurality of historical restoration times TTR and a plurality of historical non-failure times TBFs are processed, and frequency histograms of the plurality of historical restoration times TTR and the plurality of historical non-failure times TBFs are obtained, respectively. Then, a common distribution function which can be used for fitting is roughly determined according to the frequency histogram, and parameter values of the distribution function are calculated in the fitting process. In this embodiment, the distribution function selected for the historical time to repair TTR is a normal distribution function and a poisson distribution function, and the distribution function selected for the historical time to fail TBF is an exponential distribution function. Fig. 7(a) shows the fitting result of fitting the repair time TTR of the device at the station 202 in fig. 2 with a normal distribution function, and fig. 7(b) shows the fitting result of fitting the repair time TTR of the device at the station 202 in fig. 2 with a poisson distribution function; fig. 8 shows the fitting results of fitting the failure-free time TBF of the device at the station 202 in fig. 2 with an exponential distribution function.

Returning to fig. 3, in step 314, the distribution function with the highest goodness of fit is determined as the frequency distribution characteristic of the configuration parameters. In this embodiment, a fitting of the repair time TTR to the normal distribution function and the poisson distribution function is checked by using a K-S test method, and the distribution function with the highest goodness of fit is found to be the normal distribution function. Therefore, the normal distribution function is used as the frequency distribution characteristic of the repair time TTR. In addition, since only the failure time-free TBF and the exponential distribution function are fitted in step 313, the exponential distribution function is directly used as the frequency distribution characteristic of the failure time-free TBF without checking the failure time-free TBF and the exponential distribution function.

The frequency distribution characteristics of the configuration parameters are then set in the wiring-calibrated simulation model in step 315. In the present embodiment, the setting of the frequency distribution characteristics of the configuration parameters may be performed by an automatic model generation algorithm. Therefore, the process of calibrating the configuration parameters of the simulation model of the production line by using the historical production data of the production line is realized. Finally, in step 316, the adjusted simulation model is verified using the frequency distribution characteristics of the configuration parameters, and the range of the configuration parameters is further adjusted based on the verification result. In the present embodiment, the verification is performed using the monte carlo method. Thousands of random values of the cross-distribution function are generated according to the frequency distribution characteristics (normal distribution) of the repair time TTR and the frequency distribution characteristics (exponential distribution) of the TBF without the fault time respectively to serve as test data of the repair time TTR and the TBF without the fault time. And simulating the adjusted simulation model, determining the more sensitive configuration parameters of the simulation model, and further adjusting the numerical range of the configuration parameters to increase the robustness of the simulation model.

In addition, the frequency distribution characteristics of the configuration parameters are obtained by utilizing historical production data generated in the actual production process, so that more accurate configuration parameters can be set in the simulation model, and the accuracy and the effectiveness of the simulation model are further improved.

FIG. 9 illustrates an exemplary architecture of a system for calibrating simulation models of a production line according to the embodiment of FIG. 3. In the system architecture 900, three parts, namely modeling software 901, calibration software 902 and historical production log 903, are mainly included. The calibration software 902 performs the method of FIG. 3 for calibrating a simulation model of a production line. The model information processing module 901 in the calibration software 902 is configured to obtain a simulation model 9011 of the production line from the modeling software 901, extract model wiring information from the simulation model, split the model wiring information into a plurality of sets of sub-line information, and provide the plurality of sets of sub-line information to the wiring and parameter calibration module 9025. In addition, the historical production data preprocessing module 902 in the calibration software 902 obtains the historical production data of the production line in the actual production process from the historical production log 903, performs preprocessing such as integration and cleaning on the historical production data, and provides the preprocessed historical production data to the wiring and parameter calibration module 9025. The historical production log 903 may be obtained from a manual records log 9031, databases 9032 at the various devices, and a manufacturing execution system MES 9033.

The wiring and parameter calibration module 9025, on the one hand, constructs a formal model according to the obtained multiple sets of sub-line information, and performs consistency check on the formal model and the preprocessed historical production data to calibrate the wiring of the simulation model. On the other hand, the wiring and parameter calibration module 9025 further obtains historical data of the configuration parameters from the preprocessed historical production data, and determines the frequency distribution characteristics of the configuration parameters in the actual production process. The wiring and parameter calibration module 9025 provides the consistency check results and the frequency distribution characteristics of the configuration parameters to the automatic model generation module 9022, thereby generating a wiring and parameter calibrated simulation model. The model verification module 9023 is configured to verify the calibrated simulation model.

FIG. 10 illustrates an apparatus for calibrating a simulation model of a production line according to one embodiment of the present disclosure. Referring to fig. 10, the apparatus 1000 includes a simulation model acquisition unit 1001, a simulation model conversion unit 1002, a production data acquisition unit 1003, a consistency check unit 1004, and a simulation model adjustment unit 1005. The simulation model obtaining unit is configured to obtain a simulation model of the production line, the simulation model being used for simulating a production process of the production line and including a plurality of simulation modules corresponding to respective components of the production line, the simulation modules being provided with respective configuration parameters. The simulation model conversion unit 1002 is configured to convert the simulation model into a formal model, which is used to formally describe the simulation model. The production data acquisition unit 1003 is configured to acquire historical production data generated by the production line during actual production. The consistency check unit 1004 is configured to perform consistency check on the formal model and the historical production data to determine whether the product processes in the formal model and the product processes embodied by the historical production data are consistent. The simulation model adjustment unit 1005 is configured to adjust the simulation model when it is determined that the product process in the formal model and the product process embodied by the historical production data are not consistent. The units in fig. 10 may be implemented by software, hardware (e.g., integrated circuit, FPGA, etc.), or a combination of software and hardware.

In some embodiments, the simulation model adjustment unit 1005 is further configured to: determining differences between product processes in the formal model and product processes embodied by historical production data; and determining product procedures that need to be added to and/or deleted from the simulation model based on the differences; and adjusting the simulation model based on the determined product procedures that need to be added and/or deleted.

In some embodiments, the plurality of simulation modules are meshed in the simulation model, and the simulation model conversion unit 1002 is further configured to: extracting model wiring information from the simulation model, wherein the model wiring information comprises identification information and an operation sequence of a plurality of simulation modules; splitting the model wiring information into a plurality of groups of sub-line information, wherein each group of sub-line information comprises identification information and an operation sequence of each simulation module which is linearly connected from the simulation module representing the initial position of the production line to the simulation module representing the end position of the production line; and constructing a formal model according to the multiple groups of sub-line information.

In some embodiments, the consistency check unit 1004 is further configured to: and replaying the historical production data on the formal model, wherein the replay is used for determining whether the product processes which are not embodied by the historical production data exist in the formal model or whether the product processes which are not embodied by the historical production data exist in the formal model.

In some embodiments, the apparatus 1000 further comprises a historical data acquisition unit (not shown in fig. 10) configured to obtain historical data of the configuration parameters from the historical production data.

In some embodiments, the apparatus 1000 further comprises a distribution characteristic determination unit (not shown in fig. 10) configured to determine a frequency distribution characteristic of the configuration parameter in the actual production process based on historical data of the configuration parameter.

In some embodiments, the apparatus 1000 further comprises a configuration parameter setting unit (not shown in fig. 10) configured to set a frequency distribution characteristic of the configuration parameter in the adjusted simulation model.

In some embodiments, the distribution feature determination unit is further configured to: fitting the historical data of the configuration parameters by using different distribution functions; and determining the distribution function with the highest goodness of fit as the frequency distribution characteristic of the configuration parameter.

In some embodiments, the apparatus 1000 further comprises a configuration parameter verification unit (not shown in fig. 10) configured to: verifying the adjusted simulation model by using the frequency distribution characteristics of the configuration parameters; and further adjusting the range of the configuration parameter based on the verification result.

In some embodiments, the formal model is a Petri net model.

In some embodiments, the apparatus 1000 further comprises a production data processing unit (not shown in FIG. 10) configured to pre-process the historical production data prior to consistency checking the formal model and the historical production data.

FIG. 11 illustrates a block diagram of a computing device 1100 for detecting mechanical device components, according to one embodiment of the present disclosure. As can be seen in fig. 11, a computing device 1100 for detecting mechanical device components includes a processor 1101 and a memory 1102 coupled to the processor 1101. The memory 1102 is for storing computer-executable instructions that, when executed, cause the processor 1101 to perform the methods in the above embodiments.

Further, alternatively, the above-described method can be implemented by a computer-readable storage medium. Computer readable storage media has computer readable program instructions embodied thereon for performing the various embodiments of the disclosure. The computer readable storage medium may be a tangible device that can hold and store the instructions for use by the instruction execution device. The computer readable storage medium may be, for example, but not limited to, an electronic memory device, a magnetic memory device, an optical memory device, an electromagnetic memory device, a semiconductor memory device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the computer readable storage medium would include the following: a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a Static Random Access Memory (SRAM), a portable compact disc read-only memory (CD-ROM), a Digital Versatile Disc (DVD), a memory stick, a floppy disk, a mechanical coding device, such as punch cards or in-groove projection structures having instructions stored thereon, and any suitable combination of the foregoing. Computer-readable storage media as used herein is not to be construed as transitory signals per se, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through a waveguide or other transmission medium (e.g., optical pulses through a fiber optic cable), or electrical signals transmitted through electrical wires.

Thus, in another embodiment, the present disclosure proposes a computer-readable storage medium having stored thereon computer-executable instructions for performing the methods in the various embodiments of the present disclosure.

In another embodiment, the present disclosure proposes a computer program product, tangibly stored on a computer-readable storage medium, and comprising computer-executable instructions that, when executed, cause at least one processor to perform the method in the various embodiments of the present disclosure.

In general, the various example embodiments of this disclosure may be implemented in hardware or special purpose circuits, software, firmware, logic or any combination thereof. Certain aspects may be implemented in hardware, while other aspects may be implemented in firmware or software which may be executed by a controller, microprocessor or other computing device. While aspects of embodiments of the disclosure have been illustrated or described as block diagrams, flow charts, or using some other pictorial representation, it is well understood that the blocks, apparatus, systems, techniques or methods described herein may be implemented in, as non-limiting examples, hardware, software, firmware, special purpose circuits or logic, general purpose hardware or controller or other computing devices, or some combination thereof.

Computer-readable program instructions or computer program products for executing the embodiments of the present disclosure can also be stored in the cloud, and when a call is needed, a user can access the computer-readable program instructions stored in the cloud for executing one embodiment of the present disclosure through a mobile internet, a fixed network, or other networks, so as to implement the technical solutions disclosed according to the embodiments of the present disclosure.

While embodiments of the present disclosure have been described with reference to several particular embodiments, it should be understood that embodiments of the present disclosure are not limited to the particular embodiments disclosed. The embodiments of the disclosure are intended to cover various modifications and equivalent arrangements included within the spirit and scope of the appended claims. The scope of the claims is to be accorded the broadest interpretation so as to encompass all such modifications and equivalent structures and functions.

Claims

A method for calibrating a simulation model of a production line, comprising:

obtaining a simulation model of a production line, wherein the simulation model is used for simulating the production process of the production line and comprises a plurality of simulation modules corresponding to all components of the production line, and the simulation modules are provided with corresponding configuration parameters;

converting the simulation model into a formal model, wherein the formal model is used for performing formal description on the simulation model;

obtaining historical production data generated by the production line in the actual production process;

carrying out consistency check on the formal model and the historical production data to determine whether the product processes in the formal model are consistent with the product processes embodied by the historical production data; and

and when the product process in the formal model is determined to be inconsistent with the product process embodied by the historical production data, adjusting the simulation model.
The method of claim 1, wherein when it is determined that the product sequence in the formal model and the product sequence embodied by the historical production data are inconsistent, adjusting the simulation model comprises:

determining differences between product processes in the formal model and product processes embodied by the historical production data;

determining product procedures that need to be added to and/or deleted from the simulation model based on the differences; and

adjusting the simulation model based on the determined product procedures that need to be added and/or deleted.
The method of claim 1, wherein the plurality of simulation modules are meshed in the simulation model, and converting the simulation model to a formal model comprises:

extracting model wiring information from the simulation model, the model wiring information including identification information and an operation order of the plurality of simulation modules;

obtaining a plurality of groups of sub-line information according to the model wiring information, wherein each group of sub-line information comprises identification information and an operation sequence of each simulation module which is linearly connected from the simulation module representing the initial position of the production line to the simulation module representing the end position of the production line; and

and constructing the formal model according to the plurality of groups of sub-line information.
The method of claim 1, wherein consistency checking the formal model and the historical production data further comprises:

and replaying the historical production data on the formal model, wherein the replay is used for determining whether the formal model has the product processes which are not embodied by the historical production data or whether the historical production data embodies the product processes which are not embodied by the formal model.
The method of claim 1, further comprising:

obtaining historical data of the configuration parameters from the historical production data;

determining a frequency distribution characteristic of the configuration parameter in the actual production process based on the historical data of the configuration parameter; and

setting the frequency distribution characteristics of the configuration parameters in the adjusted simulation model.
The method of claim 5, wherein determining a frequency distribution profile of the configuration parameter during the actual production process based on the historical data of the configuration parameter further comprises:

fitting the historical data of the configuration parameters using different distribution functions; and

and determining the distribution function with the highest goodness of fit as the frequency distribution characteristic of the configuration parameters.
The method of claim 5, further comprising:

verifying the adjusted simulation model using the frequency distribution characteristics of the configuration parameters; and

further adjusting the range of the configuration parameter based on the verification result.
The method of claim 1, wherein the formal model is a Petri net model.
The method of claim 1, further comprising: preprocessing the historical production data prior to consistency checking the formal model and the historical production data.
Apparatus for calibrating a simulation model of a production line, comprising:

a simulation model obtaining unit configured to obtain a simulation model of a production line, the simulation model being used for simulating a production process of the production line and including a plurality of simulation modules corresponding to respective components of the production line, the simulation modules being provided with respective configuration parameters;

a simulation model conversion unit configured to convert the simulation model into a formal model, the formal model being used for formally describing the simulation model;

a production data acquisition unit configured to acquire historical production data generated by the production line in an actual production process;

a consistency check unit configured to perform consistency check on the formal model and the historical production data to determine whether a product process in the formal model and a product process embodied by the historical production data are consistent; and

a simulation model adjustment unit configured to adjust the simulation model when it is determined that the product processes in the formal model and the product processes embodied by the historical production data are not consistent.
The apparatus of claim 10, wherein the simulation model adjustment unit is further configured to:

determining differences between product processes in the formal model and product processes embodied by the historical production data; and

determining product procedures that need to be added to and/or deleted from the simulation model based on the differences; and

adjusting the simulation model based on the determined product procedures that need to be added and/or deleted.
The apparatus of claim 10, wherein the plurality of simulation modules are meshed in the simulation model, and the simulation model conversion unit is further configured to:

extracting model wiring information from the simulation model, the model wiring information including identification information and an operation order of the plurality of simulation modules;

splitting the model wiring information into a plurality of groups of sub-line information, wherein each group of sub-line information comprises identification information and an operation sequence of each simulation module which is linearly connected from a simulation module representing the starting position of the production line to a simulation module representing the ending position of the production line; and

and constructing the formal model according to the plurality of groups of sub-line information.
The apparatus of claim 10, wherein the consistency check unit is further configured to:

and replaying the historical production data on the formal model, wherein the replay is used for determining whether the formal model has the product processes which are not embodied by the historical production data or whether the historical production data embodies the product processes which are not embodied by the formal model.
The apparatus of claim 10, further comprising:

a historical data acquisition unit configured to obtain historical data of the configuration parameters from the historical production data;

a distribution characteristic determination unit configured to determine a frequency distribution characteristic of the configuration parameter in the actual production process based on the historical data of the configuration parameter; and

a configuration parameter setting unit configured to set the frequency distribution characteristics of the configuration parameters in the adjusted simulation model.
The apparatus of claim 14, wherein the distribution feature determination unit is further configured to:

fitting the historical data of the configuration parameters using different distribution functions; and

and determining the distribution function with the highest goodness of fit as the frequency distribution characteristic of the configuration parameters.
The apparatus of claim 14, further comprising a configuration parameter verification unit configured to:

verifying the adjusted simulation model using the frequency distribution characteristics of the configuration parameters; and

further adjusting the range of the configuration parameter based on the verification result.
The apparatus of claim 10, wherein the formal model is a Petri net model.
The apparatus of claim 10, further comprising a production data processing unit configured to: preprocessing the historical production data prior to consistency checking the formal model and the historical production data.
A computing device, comprising:

a processor; and

a memory for storing computer-executable instructions that, when executed, cause the processor to perform the method of any of claims 1-9.
A computer-readable storage medium having computer-executable instructions stored thereon for performing the method of any one of claims 1-9.
A computer program product, tangibly stored on a computer-readable storage medium, and comprising computer-executable instructions that, when executed, cause at least one processor to perform the method of any one of claims 1-9.