CN114841021A - Method and device for correcting digital twin model, electronic device and storage medium - Google Patents
Method and device for correcting digital twin model, electronic device and storage medium Download PDFInfo
- Publication number
- CN114841021A CN114841021A CN202210776233.5A CN202210776233A CN114841021A CN 114841021 A CN114841021 A CN 114841021A CN 202210776233 A CN202210776233 A CN 202210776233A CN 114841021 A CN114841021 A CN 114841021A
- Authority
- CN
- China
- Prior art keywords
- space data
- data
- virtual space
- physical space
- reconstruction
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 48
- 238000003860 storage Methods 0.000 title claims abstract description 12
- 238000012937 correction Methods 0.000 claims abstract description 14
- 238000012549 training Methods 0.000 claims description 34
- 239000013598 vector Substances 0.000 claims description 25
- 230000015654 memory Effects 0.000 claims description 17
- 239000011159 matrix material Substances 0.000 claims description 13
- 238000012545 processing Methods 0.000 claims description 13
- 238000004590 computer program Methods 0.000 claims description 9
- 238000005457 optimization Methods 0.000 claims description 5
- 230000004048 modification Effects 0.000 description 22
- 238000012986 modification Methods 0.000 description 22
- 238000010586 diagram Methods 0.000 description 14
- 238000004364 calculation method Methods 0.000 description 11
- 230000008569 process Effects 0.000 description 10
- 238000012512 characterization method Methods 0.000 description 8
- 238000004891 communication Methods 0.000 description 8
- 230000006870 function Effects 0.000 description 6
- 238000002715 modification method Methods 0.000 description 5
- 230000009471 action Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000003745 diagnosis Methods 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 230000002123 temporal effect Effects 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000009776 industrial production Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000007493 shaping process Methods 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/04—Manufacturing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/10—Numerical modelling
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/80—Management or planning
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- General Health & Medical Sciences (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Health & Medical Sciences (AREA)
- Manufacturing & Machinery (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The embodiment of the application provides a method and a device for correcting a digital twin model, electronic equipment and a storage medium, and relates to the technical field of digital twin. Intercepting physical space data and virtual space data obtained at the current moment and before the current moment according to a preset time sequence length to obtain target physical space data and target virtual space data; inputting the target virtual space data into a pre-trained reconstruction model, and reconstructing the target virtual space data by using the reconstruction model to obtain reconstruction data in a physical space corresponding to the target virtual space data; and determining an error index according to the reconstruction data and the target physical space data, and correcting the digital twin model under the condition that the error index meets the correction condition, so that the error precision is improved, the digital twin model can be corrected in time, and the accuracy of the digital twin model is improved.
Description
Technical Field
The present application relates to the field of digital twinning technologies, and in particular, to a method and an apparatus for correcting a digital twinning model, an electronic device, and a storage medium.
Background
At present, the digital twin technology can be used for carrying out fault diagnosis and predictive maintenance on industrial equipment, the process comprises digital twin model modeling, digital twin model modification and digital twin model application, and the accuracy of the digital twin model has important significance for the normal operation of the industrial equipment, so that the modification of the digital twin model plays a crucial role.
In the prior art, the error between the digital twin model and the industrial equipment is often obtained by performing error calculation of an original data level on virtual data corresponding to the digital twin model and physical data corresponding to the industrial equipment, and whether the digital twin model is corrected is determined according to the error, but the error calculation of the original data level has the problem of poor error precision, so that the digital twin model cannot be corrected in time, and the digital twin model is inaccurate.
Disclosure of Invention
In view of this, an object of the present application is to provide a method and an apparatus for correcting a digital twin model, an electronic device, and a storage medium, so as to improve error accuracy, and further correct the digital twin model in time, thereby improving accuracy of the digital twin model.
In order to achieve the above purpose, the embodiments of the present application employ the following technical solutions:
in a first aspect, the present application provides a method for modifying a digital twin model, the method comprising:
intercepting the current time and the physical space data and the virtual space data which are obtained before the current time according to a preset time sequence length to obtain target physical space data and target virtual space data; the physical space data is operation data of each component on the industrial equipment, and the virtual space data is operation data of each component in the digital twin model corresponding to the industrial equipment;
inputting the target virtual space data into a pre-trained reconstruction model, and reconstructing the target virtual space data by using the reconstruction model to obtain reconstruction data in a physical space corresponding to the target virtual space data;
and determining an error index according to the reconstruction data and the target physical space data, and correcting the digital twin model under the condition that the error index meets a correction condition.
In an alternative embodiment, the reconstructed model is obtained by training:
acquiring a plurality of training samples with preset time sequence lengths; each training sample comprises a physical space data sample and a virtual space data sample corresponding to each moment in the preset time sequence length;
inputting all virtual space data samples into a pre-constructed reconstruction model, and reconstructing each virtual space data sample by using the reconstruction model to obtain reconstruction data in a physical space corresponding to each virtual space data sample;
calculating an error index corresponding to each training sample according to reconstruction data in a physical space corresponding to each virtual space data sample and the physical space data sample corresponding to each virtual space data sample;
and under the condition that the error index does not meet the model convergence condition, performing iterative optimization on the parameters of the reconstruction model to obtain the trained reconstruction model.
In an optional embodiment, the intercepting, according to a preset time sequence length, the physical space data and the virtual space data obtained at a current time and before the current time to obtain target physical space data and target virtual space data includes:
denoising physical space data and virtual space data obtained at the current moment and before the current moment to obtain denoised physical space data and virtual space data;
and intercepting the denoised physical space data and virtual space data according to the preset time sequence length to obtain target physical space data and target virtual space data.
In an optional embodiment, the inputting the target virtual space data into a previously trained reconstruction model, and reconstructing the target virtual space data by using the reconstruction model to obtain reconstructed data in a physical space corresponding to the target virtual space data includes:
inputting the target virtual space data into a coding layer of the pre-trained reconstruction model to obtain a context vector corresponding to the target virtual space data;
inputting the context vector and the target virtual space data into a decoding layer of the reconstruction model to obtain initial reconstruction data under a physical space corresponding to the target virtual space data;
and inputting the initial reconstruction data into a nonlinear projection layer of the reconstruction model to obtain the reconstruction data under the physical space corresponding to the target virtual space data.
In an alternative embodiment, the determining an error index from the reconstructed data and the target physical space data includes:
calculating a shape difference and a distortion difference between the target virtual space data and the target physical space data according to the reconstruction data and the target physical space data;
calculating the error index based on the shape difference and the distortion difference.
In an alternative embodiment, the calculating the shape difference and the distortion difference between the target virtual space data and the target physical space data according to the reconstruction data and the target physical space data includes:
calculating a shape difference and a distortion difference between the target virtual space data and the target physical space data by:
wherein,
representing the reconstruction data under the physical space corresponding to the target virtual space data,
the target physical space data is characterized,
the difference in the shape is characterized by,
the difference in distortion is characterized by the difference in distortion,
characterizing a smoothness index of 0 or greater,
representing the reconstruction data under the physical space corresponding to the target virtual space data and the regular path between the reconstruction data and the target physical space data,
and representing reconstruction data under a physical space corresponding to the target virtual space data and a regular overhead matrix of the target physical space data.
In an alternative embodiment, said calculating said error index from said shape variance and said distortion variance comprises: calculating the error index by the following formula:
wherein,
the error index is characterized in that it is,
the hyper-parameters are characterized in that,
(ii) characterizing the difference in shape,
characterizing the distortion difference.
In a second aspect, the present application provides an apparatus for modifying a digital twin model, the apparatus comprising:
the acquisition module is used for intercepting the current time and the physical space data and the virtual space data which are acquired before the current time according to the preset time sequence length to acquire target physical space data and target virtual space data; the physical space data is operation data of each component on the industrial equipment, and the virtual space data is operation data of each component in the digital twin model corresponding to the industrial equipment;
the processing module is used for inputting the target virtual space data into a pre-trained reconstruction model, reconstructing the target virtual space data by using the reconstruction model and obtaining reconstruction data under a physical space corresponding to the target virtual space data;
and the correction module is used for determining an error index according to the reconstruction data and the target physical space data and correcting the digital twin model under the condition that the error index meets a correction condition.
In a third aspect, the present application provides an electronic device comprising a processor and a memory, the memory storing a computer program executable by the processor, the processor being capable of executing the computer program to implement the method of any of the preceding embodiments.
In a fourth aspect, the present application provides a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the method according to any of the preceding embodiments.
According to the method, the device, the electronic equipment and the storage medium for correcting the digital twin model, the target physical space data and the target virtual space data are obtained by combining the time sequence information, the target virtual space data are reconstructed to obtain the reconstructed data in the physical space corresponding to the target virtual space data, and the error index is determined according to the reconstructed data and the target physical space data, so that the problem of low error precision caused by error calculation of an original data level only according to the virtual space data and the physical space data is solved, the error precision is improved, the digital twin model can be corrected in time, and the accuracy of the digital twin model is improved.
In order to make the aforementioned objects, features and advantages of the present application more comprehensible, preferred embodiments accompanied with figures are described in detail below.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are required to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained from the drawings without inventive effort.
FIG. 1 shows a schematic diagram of the temporal autocorrelation of physical data and virtual data;
FIG. 2 shows a schematic diagram without taking into account the temporal autocorrelation of physical data and virtual data;
FIG. 3 is a block diagram of an electronic device provided by an embodiment of the application;
FIG. 4 is a flow chart illustrating a method for modifying a digital twin model according to an embodiment of the present disclosure;
FIG. 5 is a schematic flow chart illustrating a modification method of a digital twin model provided by an embodiment of the present application;
FIG. 6 shows a schematic diagram of a sliding window truncated training sample;
FIG. 7 is a schematic flow chart illustrating a modification method of a digital twin model according to an embodiment of the present disclosure;
FIG. 8 is a schematic flow chart illustrating a modification method of a digital twin model provided by an embodiment of the present application;
FIG. 9 is a schematic flow chart illustrating a modification method of a digital twin model according to an embodiment of the present disclosure;
FIG. 10 is a functional block diagram of a modification apparatus for a digital twin model according to an embodiment of the present disclosure;
fig. 11 shows another functional block diagram of a modification apparatus of a digital twin model provided in an embodiment of the present application.
Icon: 100-an electronic device; 110-a memory; 120-a processor; 130-a communication module; 200-an obtaining module; 210-a processing module; 220-a correction module; 230-model training module.
Detailed Description
The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. The components of the embodiments of the present application, generally described and illustrated in the figures herein, can be arranged and designed in a wide variety of different configurations.
Thus, the following detailed description of the embodiments of the present application, presented in the accompanying drawings, is not intended to limit the scope of the claimed application, but is merely representative of selected embodiments of the application. All other embodiments, which can be derived by a person skilled in the art from the embodiments of the present application without making any creative effort, shall fall within the protection scope of the present application.
It is noted that relational terms such as "first" and "second," and the like, may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
At present, the digital twin technology can be used for carrying out fault diagnosis and predictive maintenance on industrial equipment, the process comprises digital twin model modeling, digital twin model modification and digital twin model application, and the accuracy of the digital twin model has important significance for the normal operation of the industrial equipment, so that the modification of the digital twin model plays a crucial role.
In the prior art, an error between virtual data of a digital twin model and physical data of an industrial device can be calculated, so that when the error meets a preset condition, clustering error learning is performed, characteristic data of a virtual space corresponding to the digital twin model and a physical space corresponding to the industrial device are extracted, and the digital twin model is corrected. It can be understood that, in the prior art, only the virtual data of the digital twin model and the physical data of the industrial equipment are actually subjected to error calculation at the original data level, that is, the error is directly calculated through the virtual data of the digital twin model and the physical data of the industrial equipment, and on the basis of this, the following problems mainly exist in the prior art:
1. the time sequence data generated by the equipment during operation often has autocorrelation of a time sequence, and the data has certain front-back correlation along with the change of time, for example, the data presents the characteristic of monotonous change. Referring to fig. 1, it can be seen that the physical data and the virtual data may exhibit a monotone increasing characteristic with time. Therefore, in the prior art, only the error calculation of the original data level is performed, and the characteristic that data changes monotonically with time is not considered, please refer to fig. 2, although the error between the virtual space data and the physical space data in fig. 1 is smaller, but the virtual data shows a trend of decreasing after increasing with time change, obviously, autocorrelation on a time sequence is not satisfied, and when the error calculation is performed in the prior art, it is obviously impossible to consider the situation, so that the problem cannot be corrected in time by the digital twin model.
2. When the industrial equipment runs, certain correlation may exist among various operation states of various components of the industrial equipment, for example, certain dependence and limitation relation exists among shaft positions of the robot, and certain dependence relation exists between torque and current. It can be understood that the correlation is mainly reflected in that, in an overall operation process, certain correlations may exist between various components at the same time, and certain correlations may also exist between various components at different times, and obviously, the error calculation of the original data level in the prior art does not take the correlation into account.
In summary, in the prior art, the autocorrelation of data on a time sequence and the correlation of each device of each component in the operation process cannot be considered in the error calculation of the original data level, so that the problem of poor error precision exists, and the digital twin model cannot be corrected timely based on the error, so that the digital twin model is inaccurate.
Based on this, the embodiments of the present application provide a method for modifying a digital twin model to solve the above problem.
Referring to fig. 3, a block diagram of an electronic device 100 is shown, where the electronic device 100 may be a terminal device, such as a PC terminal, a mobile terminal, or the like. The electronic device 100 also needs to be in communication connection with the industrial device for acquiring the physical space data transmitted by the industrial device.
Alternatively, the industrial equipment may be equipment that is engaged in industrial production, such as industrial robots, industrial machine tools, and the like.
In a possible implementation manner, the electronic device 100 may further be provided with a digital twin model corresponding to the industrial device in advance, so that analog spatial data of the digital twin model may be directly obtained; in another possible implementation manner, the electronic device 100 may be communicatively connected with other electronic devices provided with digital twin models corresponding to industrial devices, so as to acquire analog spatial data of the digital twin models from the other electronic devices.
The electronic device 100 includes a memory 110, a processor 120, and a communication module 130. The memory 110, the processor 120 and the communication module 130 are electrically connected to each other directly or indirectly to realize data transmission or interaction. For example, the components may be electrically connected to each other via one or more communication buses or signal lines.
The memory 110 is used to store programs or data. The Memory 110 may be, but is not limited to, a Random Access Memory (RAM), a Read Only Memory (ROM), a Programmable Read-Only Memory (PROM), an Erasable Read-Only Memory (EPROM), an electrically Erasable Read-Only Memory (EEPROM), and the like.
The processor 120 is used to read/write data or programs stored in the memory and perform corresponding functions.
The communication module 130 is configured to establish a communication connection between the server and another communication terminal through the network, and to transceive data through the network.
It should be understood that the structure shown in fig. 3 is only a schematic structural diagram of the electronic device 100, and the electronic device 100 may also include more or fewer components than shown in fig. 3, or have a different configuration than shown in fig. 3. The components shown in fig. 3 may be implemented in hardware, software, or a combination thereof.
Embodiments of the present application further provide a computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, can implement the method for modifying a digital twin model provided in an embodiment of the present application.
Next, taking the electronic device 100 as an execution subject, a detailed description is given to the method for correcting the digital twin model according to the embodiment of the present application with reference to a flowchart, specifically, please refer to fig. 4, which is a flowchart of the method for correcting the digital twin model according to the embodiment of the present application, and the method includes:
step S20, intercepting the physical space data and the virtual space data obtained at the current time and before the current time according to the preset time sequence length to obtain target physical space data and target virtual space data;
the physical space data is operation data of each component on the industrial equipment, and the virtual space data is operation data of each component in a digital twin model corresponding to the industrial equipment;
optionally, the physical space data may be data obtained from the industrial equipment, including operational data of various components on the industrial equipment; the virtual space data may be data obtained from a corresponding digital twin model of the industrial plant, including operational data of the respective components.
It is understood that the physical space data is actual data obtained during actual operation of the industrial equipment, and the virtual space data is virtual data obtained by performing simulation operation through a digital twin model.
Alternatively, the digital twin model may be run synchronously with its corresponding industrial equipment, in which case the physical space data and the virtual space data obtained at the same time correspond.
Under a possible implementation condition, the electronic equipment can acquire and store physical space data and virtual space data in real time, and intercept the physical space data and the virtual space data under the condition that errors need to be determined; in another possible implementation, the electronic device may collect physical space data and virtual space data from the industrial device and the virtual space data at preset time intervals.
Optionally, the preset time sequence length is a length for intercepting the physical space data and the virtual space data, for example, if the preset time sequence length is two minutes, the physical space data and the virtual space data corresponding to each moment within a time period from the current moment to two minutes from the current moment may be intercepted.
Optionally, the target physical space data is a set of physical space data corresponding to each time after the physical space data is intercepted according to a preset time sequence length, and as can be understood, the target physical space data includes physical space data corresponding to a plurality of times, and the physical space data corresponding to each time includes operation data of each component of the industrial equipment at the time.
Optionally, the target virtual space data is a set of virtual space data corresponding to each time after the virtual space data is intercepted according to a preset time sequence length, and as can be understood, the target virtual space data includes virtual space data corresponding to a plurality of times, and the virtual space data corresponding to each time includes operation data of each component in the digital twin model at the time.
Optionally, the electronic device may intercept the physical space data and the virtual space data obtained at the current time and before the current time in a sliding window manner. For example, the current time, the physical data and the virtual data obtained before the current time are placed according to a time sequence, and the target physical space data and the target virtual space data can be directly intercepted according to a preset time sequence length in a sliding window mode.
Step S21, inputting the target virtual space data into a pre-trained reconstruction model, and reconstructing the target virtual space data by using the reconstruction model to obtain reconstruction data in a physical space corresponding to the target virtual space data;
optionally, a reconstruction model trained in advance may be set in the electronic device, after the target virtual space data is obtained, the target virtual space data may be input into the reconstruction model, and the reconstruction model is used to process the target virtual space data, so as to finally obtain the reconstruction data in the physical space corresponding to the target virtual space data.
Optionally, the reconstruction data represents data obtained by reconstructing the physical space from the target virtual space data. As can be appreciated, the reconstruction data is data fed back to the physical space for the target virtual space data.
And step S22, determining an error index according to the reconstruction data and the target physical space data, and correcting the digital twin model under the condition that the error index meets the correction condition.
Optionally, the error index may characterize an error between the digital twin model and the industrial equipment, i.e. whether the feedback digital twin model can reconstruct the physical space with a higher accuracy.
Optionally, the correction condition includes the error index being greater than a first threshold, or the error index being less than a second threshold, wherein the first threshold is greater than the second threshold. It will be appreciated that the first and second thresholds are upper and lower thresholds of error, respectively. In a possible implementation manner, the first threshold and the second threshold may be values determined according to actual requirements in combination with the loss function values during training of the reconstructed model.
Optionally, the electronic device may implement the modification of the digital twin model by optimizing the digital twin model or retraining the digital twin model when the error index satisfies the modification condition. In one possible implementation, the modification of the digital twin model may be achieved through artificial intelligence techniques such as deep learning, machine vision, and the like.
Optionally, if the industrial equipment is equipment with a high precision requirement, the physical space data and the virtual space data can be acquired in real time, and the error index is calculated in real time to ensure that the digital twin model is corrected in time; if the industrial equipment is data with low precision requirement, preset time length can be set in the electronic equipment, so that the electronic equipment can calculate the error index every other preset time length.
According to the method for correcting the digital twin model, the physical space data and the virtual space data which are obtained at the current moment and before the current moment are intercepted according to the preset time sequence length, the target physical space data and the target virtual space data are obtained, then the target virtual space data are reconstructed by using the reconstruction model which is trained in advance, so that the reconstruction data under the physical space corresponding to the target virtual space data are obtained, then the error index can be determined according to the reconstruction data and the target physical space data, and whether the digital twin model is corrected or not is determined. The target physical space data and the target virtual space data are obtained by combining the time sequence information, the reconstruction data under the physical space corresponding to the target virtual space data are obtained by reconstructing the target virtual space data, and the error index is determined according to the reconstruction data and the target physical space data, so that the problem of low error precision caused by the fact that original data-level error calculation is carried out only according to the virtual space data and the physical space data is solved, the error precision is improved, the digital twin model can be corrected in time, and the accuracy of the digital twin model is improved.
Optionally, in order to achieve the acquisition of the reconstruction data, a reconstruction model needs to be trained in advance, specifically, on the basis of fig. 4, fig. 5 is another flow chart of the modification method of the digital twin model provided in the embodiment of the present application, please refer to fig. 5, the reconstruction model may be obtained through the following training steps:
step S10, obtaining a plurality of training samples with preset time sequence length;
each training sample comprises a physical space data sample and a virtual space data sample corresponding to each moment within a preset time sequence length;
optionally, the training sample is a sample for training a reconstructed model.
Optionally, the electronic device may first obtain physical space data corresponding to the industrial device and virtual space data corresponding to the twin model within a preset time period, and then obtain the training sample according to a preset time sequence length. It is understood that the physical space data is operation data of each component on the industrial equipment, and the virtual space data is operation data of each component in the digital twin model corresponding to the industrial equipment.
Optionally, the preset time period may be set in the electronic device in advance according to the operation cycle of the industrial device, for example, the preset time period may be a time period corresponding to one operation cycle of the industrial device, or a time period corresponding to a multiple of the operation cycle of the industrial device.
Optionally, the electronic device may intercept, in a sliding window manner, the physical space data corresponding to the industrial device and the virtual space data corresponding to the twin model within a preset time period according to the preset time sequence length.
In an example, please refer to fig. 6, the physical space data and the virtual space data may be corresponded according to different moments, and the physical space data and the virtual space data are simultaneously intercepted according to a preset time sequence length in a sliding window manner to obtain a plurality of windows, where sample data corresponding to each window is a training sample with the preset time sequence length, and the training sample includes physical space data samples and virtual space data samples corresponding to each moment.
It can be understood that, for one physical space data sample, the physical space data corresponding to each time in the preset time sequence length is included, and the physical space data corresponding to each time includes the operation data of each component of the industrial equipment at that time; for one virtual space data sample, virtual space data corresponding to each time in a preset time sequence length is contained, and the virtual space data corresponding to each time comprises operation data of each component in the digital twin model at the time.
Optionally, in order to ensure the training precision, the electronic device may further perform screening, for example, denoising processing on the physical space data and the virtual space data in the preset time period before the physical space data and the virtual space data in the preset time period are intercepted, so as to intercept the physical space data and the virtual space data in the preset time period after denoising.
Step S11, inputting all virtual space data samples into a pre-constructed reconstruction model, and reconstructing each virtual space data sample by using the reconstruction model to obtain reconstruction data in a physical space corresponding to each virtual space data sample;
optionally, the reconstruction model may include an encoding layer, a decoding layer, and a nonlinear projection layer, and after all the virtual space data samples are input into the pre-constructed reconstruction model, each virtual space data sample may be processed by respectively using the encoding layer, the decoding layer, and the nonlinear projection layer in the reconstruction model in sequence, so as to obtain reconstruction data in a physical space corresponding to each virtual space data sample.
Specifically, the encoding layer may be an LSTM (Long Short-Term Memory) unit, which is configured to learn overall characteristics of the virtual space data sample from a plurality of data in the virtual space data sample as a whole.
Optionally, the coding layer may process the virtual space data corresponding to each time in the virtual space data samples, and update the hidden state of the virtual space data samples, that is, obtain the hidden state of the next time by combining the hidden state of the previous time and the virtual space data of the current time. On the basis, the hidden state output by the encoder at the time t can be expressed as the following formula:
wherein,
the hidden state at the time t output by the encoder is represented,
the hidden state at the t-1 moment output by the encoder is represented,
and representing virtual space data corresponding to the t moment.
Optionally, processing the virtual space data samples by the encoder may obtain a context vector of the virtual space data samples at a last time corresponding to the virtual space data in the virtual space data samples, where the context vector may be expressed by the following formula:
wherein,
the context vector is characterized in that it is,
characterizing a last instance of time corresponding to virtual space data in the virtual space data samples,
characterizing the output of the encoder
The hidden state at the moment in time,
characterization of
Virtual space data corresponding to the moment.
It can be understood that after all the virtual space data samples are input into the coding layer, a context vector corresponding to each data sample can be obtained, and the context vector can reflect the relationship between the virtual space data at each time included in the virtual space data sample and the whole virtual space data at each time.
Optionally, the decoding layer may also be an LSTM unit for reconstructing data of the virtual space data samples in the physical space in combination with the context vectors and the virtual space data samples.
In one example, the decoding layer may employ a teacher-shaping mode when decoding, in which a hidden state at a next time instant may be obtained in combination with a hidden state at a previous time instant, virtual spatial data at a previous time instant, and a context vector. On the basis, the hidden state output by the decoder at time t can be expressed as the following formula:
wherein,
characterizing the output of the decoder
The hidden state at the moment in time,
characterizing the output of the decoder
The hidden state at the moment in time,
characterization of
The virtual space data corresponding to the time of day,
a context vector is characterized.
In another example, the decoder may not use the teacher-forcing mode in decoding, in which case the hidden state at the next time may be obtained by combining the hidden state at the previous time, the reconstructed data in the physical space corresponding to the virtual space data at the previous time, and the context vector. On the basis, the hidden state output by the decoder at time t can be expressed as the following formula:
wherein,
characterizing the output of the decoder
The hidden state of the moment of time,
characterizing the output of the decoder
The hidden state at the moment in time,
characterization of
Reconstruction data in physical space corresponding to virtual space data at a time,
a context vector is characterized.
In this embodiment, after each virtual space data sample and the corresponding context vector thereof are input to a decoding layer of the reconstruction model, initial reconstruction data in a physical space corresponding to each virtual space data sample can be finally obtained through processing of the decoding layer. At this time, for any initial reconstruction data, the data dimension may be different from the data dimension of the corresponding virtual space data sample.
Optionally, in order to unify the initial reconstruction data with the data dimensions of the virtual space data sample corresponding to the initial reconstruction data, each initial reconstruction data may be input into the nonlinear projection layer for processing, so as to obtain the reconstruction data in the physical space corresponding to each virtual space data sample. It can be understood that, for any reconstruction data, the reconstruction data includes reconstruction data in a physical space corresponding to virtual space data at each time in a preset time sequence length, and a data dimension of the reconstruction data is also the same as a data dimension of a corresponding physical space data sample.
Step S12, calculating an error index corresponding to each training sample according to the reconstruction data in the physical space corresponding to each virtual space data sample and the physical space data sample corresponding to each virtual space data sample;
optionally, a shape difference and a distortion difference between each virtual space data sample and its corresponding physical space data sample may be calculated according to the reconstruction data in the physical space corresponding to each virtual space data sample and the physical space data sample corresponding to each virtual space sample, respectively, and then an error index corresponding to each training sample is calculated according to the shape difference and the distortion difference.
Alternatively, the shape difference and distortion difference may be calculated by constructing a warping path between the reconstructed data and the corresponding physical space sample data.
Specifically, for any physical space sample data, if the physical space sample data includes physical space data of n components at multiple time instants, it can be understood that the corresponding reconstruction data also includes data of n components at multiple time instants, and on this basis, one physical space sample data may be constructed
And defining the starting point and the end point of a planned path between the reconstruction data and the physical space data sample as the coordinates (0, 0) at the upper left corner and the coordinates (n, n) at the lower right corner of the regular matrix respectively, and defining that the regular matrix can only move along the direction of the lower right corner, the lower right corner or the lower right corner at each time.
On this basis, the error index can be calculated by the following formula:
wherein,
the error index is characterized by the fact that,
the hyper-parameters are characterized in that,
the difference in the shape is characterized by,
the distortion difference was characterized.
On the basis of the formula, the Bellman equation is combined to obtain:
wherein,
representing the reconstruction data in the physical space corresponding to the virtual space data sample,
the physical space data samples are characterized by a characterization,
representing the reconstruction data in the physical space corresponding to the virtual space data sample, and performing dynamic time warping calculation on the physical space data sample,
the difference in the shape is characterized by,
characterizing a smoothness index of 0 or greater,
representing reconstructed data in a physical space corresponding to the virtual space data samples, and regular paths between the reconstructed data and the physical space data samples,
and representing reconstruction data under a physical space corresponding to the target virtual space data and a regular overhead matrix of the target physical space data.
Alternatively, the warp variance can be calculated by calculating the variance between the optimal regular path and the penalty square, and then the warp variance can be obtained by the following equation:
wherein,
the difference in distortion is characterized by the difference in distortion,
representing the optimal regular path between the reconstruction data in the physical space corresponding to the virtual space data sample and the physical space data sample,
is one
Each element value in the penalty matrix characterizes the distance of a point at the element position to the diagonal of the matrix.
Alternatively,
each element value in the square matrix can be obtained by the following formula:
wherein,
characterization of
The row information in the square matrix is,
characterization of
The information of the columns in the square matrix,
characterization of
Element positions in the square matrix. In summary, in combination with the shape difference and the distortion difference, the error index can be finally determined by the following formula:
in this embodiment, the error index needs to be calculated according to each reconstruction data and by combining the corresponding physical space data sample, so as to obtain the error index corresponding to each training sample.
And step S13, performing iterative optimization on the parameters of the reconstruction model under the condition that the error index does not meet the model convergence condition to obtain the trained reconstruction model.
Optionally, the convergence condition is a condition for determining to stop performing iterative optimization on the reconstructed model, and in a possible implementation manner, the convergence condition may be that the number of times that the error index corresponding to each training sample is continuously consistent in multiple training reaches a preset number of times.
Optionally, the preset number is a number preset in the electronic device according to actual requirements.
Optionally, if the error index does not satisfy the model convergence condition, iteratively optimizing the parameters of the reconstructed model based on the error index corresponding to each sample, and stopping training until the number of times that the error index corresponding to each training sample is continuously consistent in multiple times of training reaches a preset number of times, so as to obtain the trained reconstructed model.
Optionally, in the process of practical application, in order to improve the accuracy of the error, the physical space data and the virtual space data obtained at and before the current time may be first filtered, and then the filtered physical space data and the filtered virtual space data may be intercepted. Specifically, on the basis of fig. 4, fig. 7 is another schematic flow chart of a method for correcting a digital twin model according to an embodiment of the present application, please refer to fig. 7, where the step S20 may be further obtained by:
step S20-1, denoising the physical space data and the virtual space data obtained at the current moment and before the current moment, and obtaining the denoised physical space data and virtual space data;
and step S20-2, intercepting the denoised physical space data and virtual space data according to a preset time sequence length to obtain target physical space data and target virtual space data.
In this embodiment, the obtained physical space data and the virtual space data may be denoised, so that the physical space data and the virtual space data obtained at the current time and before the current time are cleaned, and then the denoised physical space data and the denoised virtual space data may be intercepted according to a preset time sequence length, so as to obtain the target physical space data and the target virtual space data.
Optionally, the pre-trained reconstruction model may include an encoding layer, a decoding layer, and a nonlinear projection layer, and after the target virtual space data is input into the pre-trained reconstruction model, the target virtual space data may be processed by the encoding layer, the decoding layer, and the nonlinear projection layer, respectively, so as to finally obtain the reconstruction data in the physical space corresponding to the target virtual space data.
Specifically, on the basis of fig. 4, fig. 8 is another schematic flow chart of a method for correcting a digital twin model according to an embodiment of the present application, please refer to fig. 8, where the step S21 may be further obtained by:
step S21-1, inputting the target virtual space data into a coding layer of a pre-trained reconstruction model to obtain a context vector corresponding to the target virtual space data;
optionally, the target virtual space data is input into a pre-trained reconstruction model, the target virtual space data may be processed through an encoding layer of the reconstruction model, a hidden state at a previous time, virtual space data at a previous time, and a context vector are combined to obtain a hidden state at a next time, and finally a context vector corresponding to the target virtual space data is obtained, and the context vector may reflect a relationship between the entire virtual space data at each time included in the target virtual space data.
Step S21-2, inputting the context vector and the target virtual space data into a decoding layer of a reconstruction model to obtain initial reconstruction data under a physical space corresponding to the target virtual space data;
optionally, the obtained context vector and the target virtual space data may be input to a decoding layer of the reconstruction model, and the hidden state at the next time is obtained by combining the hidden state at the previous time, the virtual space data at the previous time, and the context vector, or the hidden state at the next time is obtained by combining the hidden state at the previous time, the reconstruction data in the physical space corresponding to the virtual space data at the previous time, and the context vector, so that the initial reconstruction data in the physical space corresponding to the target virtual space data is finally obtained.
And step S21-3, inputting the initial reconstruction data into a nonlinear projection layer of the reconstruction model to obtain the reconstruction data under the physical space corresponding to the target virtual space data.
In this embodiment, the initial reconstruction data may be processed by using the nonlinear projection layer, so as to obtain the reconstruction data in the physical space corresponding to the target virtual space data, which has the same dimension as the target virtual space data.
As can be understood, the reconstruction data at this time is a set of reconstruction data of the virtual space data corresponding to each time in the target virtual space data in the physical space.
According to the method for correcting the digital twin model, the virtual space data corresponding to each moment in the target virtual space data are taken as a whole through the coding layer, the decoding layer and the nonlinear projection layer, the overall characteristics of the virtual space data are determined, reconstruction is carried out according to the target virtual space data, reconstruction data in a physical space corresponding to the target virtual space data are obtained, and on the basis that the autocorrelation of the virtual space data on a time sequence and the correlation among various components and various operating states are considered, the data in the physical space corresponding to the target virtual space data are reconstructed as much as possible through the reconstruction model, so that the precision of an error index can be improved when the error index is determined according to the reconstruction data and the target physical space data.
Optionally, on the basis of fig. 4, fig. 9 is another schematic flow chart of a method for correcting a digital twin model provided in the embodiment of the present application, please refer to fig. 9, where the determining of the error index according to the reconstructed data and the target physical space data in the step S22 may be further implemented by:
step S22-1, calculating the shape difference and distortion difference between the target virtual space data and the target physical space data according to the reconstruction data and the target physical space data;
alternatively, the shape difference and the distortion difference may be calculated by the following formulas:
wherein,
representing the reconstruction data under the physical space corresponding to the target virtual space data,
the target physical space data is characterized,
the difference in the shape is characterized by,
the difference in distortion is characterized by the difference in distortion,
characterizing a smoothness index of 0 or greater,
representing the reconstruction data under the physical space corresponding to the target virtual space data and the regular path between the reconstruction data and the target physical space data,
and representing reconstruction data under a physical space corresponding to the target virtual space data and a regular overhead matrix of the target physical space data.
Step S22-2, calculating an error index according to the shape difference and the distortion difference.
In this embodiment, the error index can be calculated by the following equation:
wherein,
the error index is characterized by the fact that,
the hyper-parameters are characterized in that,
the difference in the shape is characterized by,
the distortion difference was characterized.
Alternatively, in the case where the shape difference and the distortion difference are characterized by the above formulas, the error index may also be obtained by the following formulas:
it is understood that after the error index is calculated, whether to modify the digital twin model may be determined according to the error index and the modification condition.
In order to perform the corresponding steps in the above embodiments and various possible manners, an implementation manner of the modification apparatus of the digital twin model is given below. Further, referring to fig. 10, fig. 10 is a functional block diagram of a modification apparatus of a digital twin model according to an embodiment of the present application. It should be noted that the basic principle and the generated technical effect of the modification apparatus of the digital twin model provided in the present embodiment are the same as those of the above embodiments, and for the sake of brief description, no part of the present embodiment is mentioned, and corresponding contents in the above embodiments may be referred to. The device for correcting the digital twin model comprises: an acquisition module 200, a processing module 210, and a modification module 220.
The acquiring module 200 is configured to intercept physical space data and virtual space data obtained at a current time and before the current time according to a preset time sequence length, so as to obtain target physical space data and target virtual space data; the physical space data is the operation data of each component on the industrial equipment, and the virtual space data is the operation data of each component in the digital twin model corresponding to the industrial equipment;
it is understood that the obtaining module 200 may be configured to perform the step S20;
the processing module 210 is configured to input target virtual space data into a pre-trained reconstruction model, and reconstruct the target virtual space data by using the reconstruction model to obtain reconstruction data in a physical space corresponding to the target virtual space data;
it is understood that the processing module 210 can be configured to execute the above step S21;
the correcting module 220 is configured to determine an error index according to the reconstruction data and the target physical space data, and correct the digital twin model when the error index satisfies a correction condition.
It is understood that the modification module 220 may be configured to perform the step S22.
Alternatively, fig. 11 is another functional block diagram of a modification apparatus of a digital twin model according to an embodiment of the present application, and referring to fig. 11, the modification apparatus of the digital twin model may further include a model training module 230.
The model training module 230 is configured to obtain a plurality of training samples with preset time sequence lengths; each training sample comprises a physical space data sample and a virtual space data sample corresponding to each moment within a preset time sequence length; inputting all the virtual space data samples into a pre-constructed reconstruction model, and reconstructing each virtual space data sample by using the reconstruction model to obtain reconstruction data in a physical space corresponding to each virtual space data sample; calculating an error index corresponding to each training sample according to the reconstruction data in the physical space corresponding to each virtual space data sample and the physical space data sample corresponding to each virtual space data sample; and under the condition that the error index does not meet the convergence condition of the model, carrying out iterative optimization on the parameters of the reconstruction model to obtain the trained reconstruction model.
It is understood that the model training module 230 can be used to perform the steps S10-S13.
Optionally, the obtaining module 200 is further configured to denoise physical space data and virtual space data obtained at and before the current time, and obtain the denoised physical space data and virtual space data; and intercepting the denoised physical space data and virtual space data according to a preset time sequence length to obtain target physical space data and target virtual space data.
It is understood that the obtaining module 200 may be further configured to perform the steps S20-1 to S20-2.
Optionally, the processing module 210 is further configured to input the target virtual space data into a coding layer of a pre-trained reconstruction model, so as to obtain a context vector corresponding to the target virtual space data; inputting the context vector and the target virtual space data into a decoding layer of a reconstruction model to obtain initial reconstruction data in a physical space corresponding to the target virtual space data; and inputting the initial reconstruction data into a nonlinear projection layer of the reconstruction model to obtain the reconstruction data under the physical space corresponding to the target virtual space data.
It is understood that the processing module 210 can also be used to perform the steps S21-1 to S21-3.
Optionally, the modification module 220 is further configured to calculate a shape difference and a distortion difference between the target virtual space data and the target physical space data according to the reconstruction data and the target physical space data; an error index is calculated from the shape difference and the distortion difference.
It is understood that the modification module 220 can be used for executing the steps S22-1 to S22-2.
Optionally, the modification module 220 is further configured to calculate a shape difference and a distortion difference between the target virtual space data and the target physical space data by the following formulas:
wherein,
representing the reconstruction data under the physical space corresponding to the target virtual space data,
the target physical space data is characterized,
the difference in the shape is characterized by,
the difference in distortion is characterized by the difference in distortion,
characterizing a smoothness index of 0 or greater,
representing the reconstruction data under the physical space corresponding to the target virtual space data and the regular path between the reconstruction data and the target physical space data,
and representing reconstruction data under a physical space corresponding to the target virtual space data and a regular overhead matrix of the target physical space data.
Optionally, the modifying module 220 is further configured to calculate an error index by the following formula:
wherein,
the error index is characterized by the fact that,
characterization of SupervisoryThe parameters are set to be in a predetermined range,
the difference in the shape is characterized by,
the distortion difference was characterized.
According to the correction device of the digital twin model, the acquisition module intercepts physical space data and virtual space data obtained at the current moment and before the current moment according to the preset time sequence length to obtain target physical space data and target virtual space data; the physical space data is the operation data of each component on the industrial equipment, and the virtual space data is the operation data of each component in the digital twin model corresponding to the industrial equipment; inputting the target virtual space data into a pre-trained reconstruction model through a processing module, and reconstructing the target virtual space data by using the reconstruction model to obtain reconstruction data under a physical space corresponding to the target virtual space data; and determining an error index according to the reconstruction data and the target physical space data through a correction module, and correcting the digital twin model under the condition that the error index meets the correction condition. The target physical space data and the target virtual space data are obtained by combining the time sequence information, the reconstruction data under the physical space corresponding to the target virtual space data are obtained by reconstructing the target virtual space data, and the error index is determined according to the reconstruction data and the target physical space data, so that the problem of low error precision caused by the fact that original data-level error calculation is carried out only according to the virtual space data and the physical space data is solved, the error precision is improved, the digital twin model can be corrected in time, and the accuracy of the digital twin model is improved.
Alternatively, the modules may be stored in the memory shown in fig. 3 in the form of software or Firmware (Firmware) or be fixed in an Operating System (OS) of the electronic device, and may be executed by the processor in fig. 3. Meanwhile, data, codes of programs, and the like required to execute the above modules may be stored in the memory.
In the embodiments provided in the present application, it should be understood that the disclosed apparatus and method can be implemented in other ways. The apparatus embodiments described above are merely illustrative, and for example, the flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, 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 block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, 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.
In addition, functional modules in the embodiments of the present application may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part.
The functions, if implemented in the form of software functional modules and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present application or portions thereof that substantially contribute to the prior art may be embodied in the form of a software product stored in a storage medium and including instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
The above description is only a preferred embodiment of the present application and is not intended to limit the present application, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, improvement and the like made within the spirit and principle of the present application shall be included in the protection scope of the present application.
Claims (10)
1. A method of modifying a digital twin model, the method comprising:
intercepting the current time and the physical space data and the virtual space data which are obtained before the current time according to a preset time sequence length to obtain target physical space data and target virtual space data; the physical space data is operation data of each component on the industrial equipment, and the virtual space data is operation data of each component in the digital twin model corresponding to the industrial equipment;
inputting the target virtual space data into a pre-trained reconstruction model, and reconstructing the target virtual space data by using the reconstruction model to obtain reconstruction data in a physical space corresponding to the target virtual space data;
and determining an error index according to the reconstruction data and the target physical space data, and correcting the digital twin model under the condition that the error index meets a correction condition.
2. The method of claim 1, wherein the reconstructed model is obtained by training:
acquiring a plurality of training samples with preset time sequence lengths; each training sample comprises a physical space data sample and a virtual space data sample corresponding to each moment in the preset time sequence length;
inputting all virtual space data samples into a pre-constructed reconstruction model, and reconstructing each virtual space data sample by using the reconstruction model to obtain reconstruction data in a physical space corresponding to each virtual space data sample;
calculating an error index corresponding to each training sample according to reconstruction data in a physical space corresponding to each virtual space data sample and the physical space data sample corresponding to each virtual space data sample;
and under the condition that the error index does not meet the model convergence condition, performing iterative optimization on the parameters of the reconstruction model to obtain the trained reconstruction model.
3. The method according to claim 1, wherein the intercepting the physical space data and the virtual space data obtained at the current time and before the current time according to the preset time sequence length to obtain the target physical space data and the target virtual space data comprises:
denoising physical space data and virtual space data obtained at the current moment and before the current moment to obtain denoised physical space data and virtual space data;
and intercepting the denoised physical space data and virtual space data according to the preset time sequence length to obtain target physical space data and target virtual space data.
4. The method according to claim 1, wherein the inputting the target virtual space data into a previously trained reconstruction model, and reconstructing the target virtual space data by using the reconstruction model to obtain the reconstruction data in the physical space corresponding to the target virtual space data comprises:
inputting the target virtual space data into a coding layer of the pre-trained reconstruction model to obtain a context vector corresponding to the target virtual space data;
inputting the context vector and the target virtual space data into a decoding layer of the reconstruction model to obtain initial reconstruction data under a physical space corresponding to the target virtual space data;
and inputting the initial reconstruction data into a nonlinear projection layer of the reconstruction model to obtain the reconstruction data under the physical space corresponding to the target virtual space data.
5. The method of claim 1, wherein determining an error index from the reconstructed data and the target physical space data comprises:
calculating a shape difference and a distortion difference between the target virtual space data and the target physical space data according to the reconstruction data and the target physical space data;
calculating the error index based on the shape difference and the distortion difference.
6. The method of claim 5, wherein said calculating shape differences and distortion differences between said target virtual space data and said target physical space data from said reconstruction data and said target physical space data comprises:
calculating a shape difference and a distortion difference between the target virtual space data and the target physical space data by:
wherein,
representing the reconstruction data under the physical space corresponding to the target virtual space data,
the target physical space data is characterized,
the difference in the shape is characterized by,
the difference in distortion is characterized by the difference in distortion,
characterizing a smoothness index of 0 or greater,
representing the reconstruction data under the physical space corresponding to the target virtual space data and the regular path between the reconstruction data and the target physical space data,
and characterizing the reconstruction data under the physical space corresponding to the target virtual space data and a regular overhead matrix of the target physical space data.
7. The method of claim 5, wherein said calculating the error index based on the shape variance and the distortion variance comprises:
calculating the error index by the following formula:
wherein,
the error index is characterized in that it is,
the hyper-parameters are characterized in that,
(ii) characterizing the difference in shape,
characterizing the distortion difference.
8. An apparatus for modifying a digital twin model, the apparatus comprising:
the acquisition module is used for intercepting the current time and the physical space data and the virtual space data which are acquired before the current time according to the preset time sequence length to acquire target physical space data and target virtual space data; the physical space data is operation data of each component on the industrial equipment, and the virtual space data is operation data of each component in the digital twin model corresponding to the industrial equipment;
the processing module is used for inputting the target virtual space data into a pre-trained reconstruction model, reconstructing the target virtual space data by using the reconstruction model and obtaining reconstruction data under a physical space corresponding to the target virtual space data;
and the correction module is used for determining an error index according to the reconstruction data and the target physical space data and correcting the digital twin model under the condition that the error index meets a correction condition.
9. An electronic device comprising a processor and a memory, the memory storing a computer program executable by the processor, the processor being operable to execute the computer program to implement the method of any one of claims 1 to 7.
10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the method according to any one of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210776233.5A CN114841021B (en) | 2022-07-04 | 2022-07-04 | Method and device for correcting digital twin model, electronic device and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210776233.5A CN114841021B (en) | 2022-07-04 | 2022-07-04 | Method and device for correcting digital twin model, electronic device and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114841021A true CN114841021A (en) | 2022-08-02 |
| CN114841021B CN114841021B (en) | 2022-10-11 |
Family
ID=82575031
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210776233.5A Active CN114841021B (en) | 2022-07-04 | 2022-07-04 | Method and device for correcting digital twin model, electronic device and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114841021B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115408890A (en) * | 2022-11-02 | 2022-11-29 | 南京体育学院 | Data processing method based on digital twin intelligent phase space multi-dimensional reconstruction |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110851966A (en) * | 2019-10-30 | 2020-02-28 | 同济大学 | Digital twin model correction method based on deep neural network |
| CN112070279A (en) * | 2020-08-19 | 2020-12-11 | 浙江工业大学 | Product processing control method based on digital twinning technology |
| US20210262439A1 (en) * | 2018-11-18 | 2021-08-26 | Zhejiang University | Method of identification and compensation of inherent deviation of yaw error of wind turbine based on true power curve |
| US20210319156A1 (en) * | 2020-04-09 | 2021-10-14 | Wuhan University | Method and system for fault diagnosis with small samples of power equipment based on virtual and real twin spaces |
| CN113868898A (en) * | 2021-11-29 | 2021-12-31 | 联想新视界（北京）科技有限公司 | Data processing method and device based on digital twin model |
| CN114199156A (en) * | 2021-12-15 | 2022-03-18 | 桂林航天工业学院 | Wing profile curved surface shape reconstruction method based on double FBG shape sensors |
| CN114298443A (en) * | 2022-03-03 | 2022-04-08 | 北京航空航天大学杭州创新研究院 | Industrial equipment predictive maintenance method and device based on health state index and electronic equipment |
-
2022
- 2022-07-04 CN CN202210776233.5A patent/CN114841021B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20210262439A1 (en) * | 2018-11-18 | 2021-08-26 | Zhejiang University | Method of identification and compensation of inherent deviation of yaw error of wind turbine based on true power curve |
| CN110851966A (en) * | 2019-10-30 | 2020-02-28 | 同济大学 | Digital twin model correction method based on deep neural network |
| US20210319156A1 (en) * | 2020-04-09 | 2021-10-14 | Wuhan University | Method and system for fault diagnosis with small samples of power equipment based on virtual and real twin spaces |
| CN112070279A (en) * | 2020-08-19 | 2020-12-11 | 浙江工业大学 | Product processing control method based on digital twinning technology |
| CN113868898A (en) * | 2021-11-29 | 2021-12-31 | 联想新视界（北京）科技有限公司 | Data processing method and device based on digital twin model |
| CN114199156A (en) * | 2021-12-15 | 2022-03-18 | 桂林航天工业学院 | Wing profile curved surface shape reconstruction method based on double FBG shape sensors |
| CN114298443A (en) * | 2022-03-03 | 2022-04-08 | 北京航空航天大学杭州创新研究院 | Industrial equipment predictive maintenance method and device based on health state index and electronic equipment |
Non-Patent Citations (3)
| Title |
|---|
| XIN LI等: "Multisource Model-Driven Digital Twin System of Robotic Assembly", 《IEEE SYSTEMS JOURNAL》 * |
| 林润泽等: "基于数字孪生的智能装配机械臂实验系统", 《实验室研究与探索》 * |
| 谢红兰等: "X射线荧光全息术中消除光源偏振效应和孪生像的重构新算法", 《物理学报》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115408890A (en) * | 2022-11-02 | 2022-11-29 | 南京体育学院 | Data processing method based on digital twin intelligent phase space multi-dimensional reconstruction |
| CN115408890B (en) * | 2022-11-02 | 2023-03-24 | 南京体育学院 | Data processing method based on digital twin intelligent phase space multi-dimensional reconstruction |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114841021B (en) | 2022-10-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111047482B (en) | Knowledge tracking system and method based on hierarchical memory network | |
| CN114841021B (en) | Method and device for correcting digital twin model, electronic device and storage medium | |
| CN113011085A (en) | Equipment digital twin modeling method and system | |
| CN109065176B (en) | Blood glucose prediction method, device, terminal and storage medium | |
| JP7152938B2 (en) | Machine learning model building device and machine learning model building method | |
| CN109514553A (en) | A kind of method, system and the equipment of the mobile control of robot | |
| WO2022009010A1 (en) | Model fidelity monitoring and regeneration for manufacturing process decision support | |
| CN117076171A (en) | Abnormality detection and positioning method and device for multi-element time sequence data | |
| CN115359321A (en) | Model training method and device, electronic equipment and storage medium | |
| CN116738617A (en) | Power system modeling method and device, electronic equipment and storage medium | |
| CN113484882B (en) | GNSS sequence prediction method and system of multi-scale sliding window LSTM | |
| JP7079294B2 (en) | Video blur detection method and equipment | |
| CN115972216B (en) | Parallel robot forward motion solving method, control method, equipment and storage medium | |
| CN114298299A (en) | Model training method, device, equipment and storage medium based on course learning | |
| CN111242379A (en) | Nuclear recursive maximum correlation entropy time sequence online prediction method based on random Fourier features | |
| CN112882934A (en) | Test analysis method and system based on defect growth | |
| CN113361380B (en) | Human body key point detection model training method, detection method and device | |
| CN114399901B (en) | Method and equipment for controlling traffic system | |
| CN116522070A (en) | Non-supervision intelligent fault diagnosis method and system for mechanical parts | |
| CN115660981A (en) | Image denoising method and device, electronic equipment and computer storage medium | |
| CN113569479B (en) | Long-term multi-step control method, device and storage medium for rock mass crack development of stone cave temple | |
| CN112884129B (en) | Multi-step rule extraction method, device and storage medium based on teaching data | |
| CN116151092A (en) | Loss metering method and system for extra-high voltage direct current transmission system | |
| CN108763010B (en) | Performance prediction method and device and data analysis equipment | |
| CN117109593B (en) | Submersible robot positioning method and system based on robust Kalman filtering |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |