Digital twin platform based on multi-source data access

Abstract

The invention discloses a digital twinborn platform based on multi-source data access, relates to the technical field of digital twinborn platforms, and solves the problem that the data accuracy difference of equipment models in the digital twinborn platform is large. The digital twin platform can receive related data with uniform data formats, the optimal selected format is determined when the formats are uniformly processed, the conversion rate is higher, the efficiency is higher and a better data conversion effect is achieved when format conversion is carried out on multi-source data, and in order to guarantee the related accuracy of the data output by the digital twin platform, the conversion efficiency is improved. And the data output in real time is checked with the collected data, whether time delay or related errors exist or not is identified, and the data output by the digital twin model in real time is calibrated in real time based on the determined time delay and related errors, so that the accuracy of data output of the digital twin model is guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of digital twin platforms, and in particular to a digital twin platform based on multi-source data access. Background Art

[0002] Digital twins make full use of physical models, sensors, operation history and other data, integrate multi-disciplinary, multi-physical quantity, multi-scale, and multi-probability simulation processes, complete mapping in virtual space, and thus reflect the entire life cycle of the corresponding physical equipment; digital twins are a concept that transcends reality and can be regarded as a digital mapping system of one or more important and interdependent equipment systems.

[0003] The application with publication number CN115834629A relates to a physical equipment interconnection platform and method for a digital twin manufacturing workshop. The interconnection platform is characterized in that it includes an interface conversion platform, a protocol integration platform and a digital twin management and control system; the interface conversion platform connects the multi-source heterogeneous physical equipment in the manufacturing workshop, and is used to convert the physical interfaces of the multi-source heterogeneous physical equipment in the manufacturing workshop into a unified interface type; the protocol integration platform is used to integrate multiple communication protocols using multi-threading, obtain the equipment data of each multi-source heterogeneous physical equipment and process it; the digital twin management and control system is used to determine the working status of each multi-source heterogeneous physical equipment according to the equipment data processed by the protocol integration platform. The present invention can be widely used in the field of equipment communication.

[0004] When its digital twin platform is used for related processing of multi-source data, it generally converts several groups of different data formats based on the different data formats of the multi-source data, and then uses the digital twin platform to display the multi-source data in real time. However, the original conversion method is relatively simple, and its data format conversion is slow. In addition, the subsequent displayed device model has related errors in data output, resulting in inconsistency between the actual output data and the data output by the virtual device model, which causes related problems with data accuracy. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present invention provides a digital twin platform based on multi-source data access, which solves the problem of large differences in the accuracy of device model data within the digital twin platform.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: A digital twin platform based on multi-source data access, including:

[0007] Access the data collection end, collect the data connected to the multiple data source interfaces connected to the digital twin platform, and transmit the collected multi-source data to the data format processing end, where each collection is a collection cycle, and the collection cycle is a preset cycle;

[0008] The data format processing end performs format analysis and processing on the multiple sets of multi-source data collected during the current collection cycle, unifies the formats of the multiple sets of multi-source data, and transmits the multiple sets of multi-source data after the unified formats to the virtual space generation end. The specific method is as follows:

[0009] Confirm the data formats that can be processed by this digital twin platform and mark such data formats as the formats to be converted;

[0010] Identify whether there is multi-source data in the same format as the one to be converted among several different sets of multi-source data:

[0011] If it exists, such multi-source data is calibrated as standard data, and the format corresponding to the standard data with the largest data capacity is calibrated as the selected format;

[0012] If there is no standard data, select the selected format from the formats to be converted: confirm the average conversion speed V generated during the conversion of different data formats to the selected format from the historical completed data i-k , where i represents different data formats, k represents different formats to be converted, and based on the label k, several groups belonging to the same format to be converted are converted at an average speed V i-k Perform mean processing to determine the undetermined mean J k , and then convert several groups of the same format to be converted at an average speed V i-k Perform variance processing to determine the unknown variance F k , using PD k =(J k ×C1)÷(F k ×C2) determine its evaluation value PD k , where C1 and C2 are preset fixed coefficient factors, from several groups of evaluation values ​​PD k Select the format to be converted corresponding to the maximum value, and mark the format to be converted as the selected format;

[0013] Based on the determined selected format, converting the multi-source data in different formats into the selected format, and transmitting the converted multi-source data in the same format to the virtual space generation end;

[0014] The virtual space generation end constructs the virtual space based on the processed multi-source data in the same format, performs data mapping in the virtual space, and generates digital twin models corresponding to several groups of multi-source data;

[0015] The data verification processing end verifies the operating data generated by the digital twin model with the corresponding multi-source data collected, identifies abnormal data, and makes parameter adjustments;

[0016] And the data verification processing end includes a data checking unit, a periodic verification unit and a parameter adjustment unit at the same time;

[0017] Among them, the data verification unit verifies the parameter data at the same time, determines the operation data generated by its digital twin model at the current moment, and then determines the multi-source data of the data source associated with the operation data at the current moment, and verifies the operation data with the multi-source data to determine the abnormal moment. The specific method is:

[0018] The running data generated at the current moment is marked as YX q , where q represents different moments, and the corresponding multi-source data collected at the current moment is calibrated as DY q ;

[0019] Identify whether its operating data and multi-source data meet: |YX q -DY q |>Y1, where Y1 is the preset value. If yes, the current moment is marked as an abnormal moment. If no, monitoring continues;

[0020] The periodic verification unit performs periodic data verification based on the determined abnormal time, and identifies whether there is data delay based on the specific data verification result, specifically in the following manner:

[0021] The determined abnormal moment T1 is periodically confirmed, the first group of generation cycles is locked, the generation cycle is a preset value, and the initial moment of the generation cycle is T1. The operation data generated by the digital twin model within the generation cycle are confirmed one by one and sorted according to the timeline to generate an operation data sequence;

[0022] Based on the determined initial time and end time of the generation cycle, the initial time is extended forward by half of the generation cycle, and the end time is extended backward by half of the generation cycle to lock a group of verification cycles, and the corresponding multi-source data monitored in the verification cycle are sorted to determine the multi-source data sequence;

[0023] The running data sequence is checked with the multi-source data sequence, and each checking process is determined: the initial data of the running data sequence is first aligned with the initial data of the multi-source data sequence, the first checking process is determined, and it is identified whether the running data and the multi-source data at each same position in the checking process are consistent. If they are inconsistent, the second checking process is performed. If they are consistent, this checking process is marked as the standard process; the initial data of the running data sequence is aligned with the second set of data of the multi-source data sequence, the second checking process is determined, and whether they are consistent in this checking process, and so on, until the last running data of the running data sequence is aligned with the last running data of the multi-source data sequence, and the process stops;

[0024] Identify whether there is a standard process among several verification processes:

[0025] If a standard process exists, the determined standard process is transmitted to the parameter adjustment unit;

[0026] If there is no standard process, the operation data at the same position in each verification process is subjected to difference processing with the multi-source data to determine several groups of difference values ​​in the corresponding verification process, and then the several groups of difference values ​​are subjected to variance processing to determine the standard value in the corresponding verification process. Based on the different standard values ​​corresponding to each verification process, a group of minimum values ​​is determined, and the verification process corresponding to the minimum value is calibrated as a sub-standard process;

[0027] The parameter adjustment unit, based on the determined standard process or substandard process, makes parameter corrections to the generated digital twin model to ensure the data output accuracy of the digital twin model; the specific method is:

[0028] If the confirmed verification process is a standard process, the time difference correction is performed: based on the time difference CZ between the multi-source data and the operating data at the same position in the corresponding standard process, the digital twin model is time-corrected. If CZ>0, the operating data corresponding to the time point with a time difference of CZ before the current moment is used as the output data. If CZ<0, the operating data corresponding to the time point with a time difference of CZ after the current moment is used as the output data.

[0029] If the verification process is determined to be a substandard process, determine the difference between its operating data and multi-source data at the same position, then average several groups of differences to determine the correlation value, and then use the same method of standard process time difference correction to determine the output data of the digital twin model at the current moment, and combine the output data with the correlation value to determine the accurate data output at the current moment, and the accurate data = output data + correlation value.

[0030] The present invention provides a digital twin platform based on multi-source data access. Compared with the prior art, it has the following beneficial effects:

[0031] The present invention unifies the formats of different multi-source data received by the digital twin platform, so that the digital twin platform can receive relevant data with unified data formats. When the formats are unified, the best selected format is determined, so that when the multi-source data is converted into a format, the conversion rate is faster and the efficiency is higher, thereby achieving a better data conversion effect.

[0032] When its digital twin platform performs data verification and output, in order to ensure the relevant accuracy of the data output by its digital twin platform, the real-time output data is compared with the collected data to identify whether there is a time delay or related error. Based on the determined time extension and related errors, the real-time output data of the digital twin model is calibrated in real time to ensure the accuracy of the data output of its digital twin model. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 It is a schematic diagram of the principle framework of the present invention;

[0034] Figure 2 It is a schematic diagram of moving and checking two sets of data sequences of the present invention. DETAILED DESCRIPTION

[0035] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0036] Example 1

[0037] See also Figure 1 , the present application provides a digital twin platform based on multi-source data access, including an access data acquisition end, a data format processing end, a virtual space generation end, and a data verification processing end, wherein the access data acquisition end is electrically connected to an input node of the data format processing end, wherein the data format processing end is electrically connected to an input node of the virtual space generation end or the data verification processing end, respectively, and the virtual space generation end and the data verification processing end are bidirectionally connected;

[0038] The data verification processing end includes a data checking unit, a period checking unit and a parameter adjustment unit, wherein the data checking unit, the period checking unit and the parameter adjustment unit are electrically connected from the output node to the input node in sequence;

[0039] Among them, the access data acquisition end collects data connected to multiple data source interfaces connected to the digital twin platform, and transmits the collected multi-source data to the data format processing end, where each collection is a collection cycle, and the collection cycle is a preset cycle, which is formulated by the operator based on experience, and is generally 10 minutes;

[0040] The data format processing end performs format analysis on the multiple sets of multi-source data collected during the current collection cycle, unifies the formats of the multiple sets of multi-source data, and transmits the multiple sets of multi-source data after the unified formats to the virtual space generation end. The specific method of format unification is as follows:

[0041] Confirm the data formats that can be processed by this digital twin platform and mark such data formats as the formats to be converted;

[0042] Identify whether there is multi-source data in the same format as the one to be converted among several different sets of multi-source data:

[0043] If it exists, such multi-source data is calibrated as standard data, and the format corresponding to the standard data with the largest data capacity is calibrated as the selected format;

[0044] If there is no standard data, select the selected format from the formats to be converted: confirm the average conversion speed V generated during the conversion of different data formats to the selected format from the historical completed data i-k , where i represents different data formats, k represents different formats to be converted, and based on the label k, several groups belonging to the same format to be converted are converted at an average speed V i-k Perform mean processing to determine the undetermined mean J k , and then convert several groups of the same format to be converted at an average speed V i-k Perform variance processing to determine the unknown variance F k , using PD k =(J k ×C1)÷(F k ×C2) determine its evaluation value PD k , where C1 and C2 are preset fixed coefficient factors, and their specific values ​​are determined by the operator based on experience, from several groups of evaluation values ​​PD k Select the format to be converted corresponding to the maximum value, and mark the format to be converted as the selected format;

[0045] Based on the determined selected format, converting the multi-source data in different formats into the selected format, and transmitting the converted multi-source data in the same format to the virtual space generation end;

[0046] For example: The digital twin platform can process three data formats: CSV, JSON, and XML, which are marked as formats to be converted.

[0047] Now there are 5 different sets of multi-source data, namely:

[0048] Data set 1: The format is CSV and the data capacity is 100MB.

[0049] Data group 2: The format is JSON and the data capacity is 80MB.

[0050] Data group 3: The format is CSV and the data capacity is 120MB.

[0051] Data group 4: The format is XML and the data capacity is 90MB.

[0052] Data group 5: The format is JSON and the data capacity is 70MB.

[0053] First, since there are multi-source data in CSV format (data group 1 and data group 3), they are calibrated as standard data; among them, the standard data with the largest data capacity is CSV (220MB), so the CSV format is calibrated as the selected format.

[0054] Next, let's assume that there is no multi-source data in CSV format, but only in JSON and XML. From the historical completion data, we get the average conversion speed of different data formats to CSV, JSON and XML:

[0055] Average speed of converting CSV to JSON: V 1-2 =20MB / s

[0056] Average speed of CSV to XML conversion: V 1-3 =15MB / s

[0057] Average speed of converting JSON to CSV: V 2-1 =18MB / s

[0058] Average speed of converting JSON to XML: V 2-3 =12MB / s

[0059] Average speed of XML to CSV conversion: V 3-1 =16MB / s

[0060] Average speed of XML to JSON conversion: V 3-2 =10MB / s

[0061] Assume C1=2, C2=1.

[0062] For CSV format as the format to be converted (k=1):

[0063] Average J1 = (20 + 15 + 18 + 16) ÷ 4 = 17.25 MB / s

[0064] Variance F1 = [(20-17.25)2 + (15-17.25)2 + (18-17.25)2 + (16-17.25)2] ÷ 4 = 5.1875

[0065] Rating value PD1 = (17.25 × 2) ÷ (5.1875 × 1) = 6.8

[0066] For JSON format as the format to be converted (k=2):

[0067] Average J2 = (18 + 12 + 10) ÷ 3 = 13.33 MB / s

[0068] Variance F2 = [(18-13.33)2 + (12-13.33)2 + (10-13.33)2] ÷ 3 = 11.11

[0069] Rating value PD2 = (13.33 × 2) ÷ (11.11 × 1) = 2.4

[0070] For XML format as the format to be converted (k=3):

[0071] Average J3 = (16 + 10) ÷ 2 = 13 MB / s

[0072] Variance F3 = [(16-13)2 + (10-13)2] ÷ 2 = 9

[0073] Rating value PD3 = (13 × 2) ÷ (9 × 1) = 2.89

[0074] Since PD1 (6.8) is the largest, the CSV format is selected.

[0075] Finally, the multi-source data in different formats (such as JSON and XML) are converted to the selected CSV format, and the converted multi-source data in the same format are transmitted to the virtual space generation end.

[0076] Among them, the virtual space generation end constructs the virtual space based on the processed multi-source data of the same format, performs data mapping in the virtual space, and generates several groups of digital twin models corresponding to the multi-source data. Because the digital twin makes full use of physical models, sensors, operation history and other data, integrates multi-disciplinary, multi-physical quantity, multi-scale, and multi-probability simulation processes, and completes mapping in the virtual space, it reflects the entire life cycle of the corresponding physical equipment. Based on the collected multi-source data, spatial mapping can be performed to determine the virtual model. This type of digital twin processing technology is relatively common in the existing technology, so it will not be elaborated here.

[0077] Example 2

[0078] The data verification processing end verifies the operating data generated by the digital twin model with the corresponding multi-source data collected, identifies abnormal data, and makes the digital twin model more realistic by adjusting parameters.

[0079] The parameter data at the same moment is checked based on the data checking unit at the same moment first, the operation data generated by the digital twin model at the current moment is determined, and then the multi-source data of the data source associated with the operation data at the current moment is determined, the operation data is checked with the multi-source data to determine the abnormal moment, and the determined abnormal moment is transmitted to the periodic verification unit. The operation data is generally the result data, such as the specific speed value generated by a certain device under a certain power parameter state, wherein the specific method of checking is:

[0080] The running data generated at the current moment is marked as YX q , where q represents different moments, and the corresponding multi-source data collected at the current moment is calibrated as DY q ;

[0081] Identify whether its operating data and multi-source data meet: |YX q -DY q |>Y1, where Y1 is a preset value, and its specific value is determined by the operator based on experience. If yes, the current moment is marked as an abnormal moment, if not, continue monitoring.

[0082] The periodic verification unit thereof performs periodic data verification based on the determined abnormal time, and identifies whether data delay exists based on the specific data verification result, wherein the specific method of identifying is:

[0083] The determined abnormal moment T1 is periodically confirmed, and the first group of generation cycles is locked. The generation cycle is a preset value, and its specific value is determined by the operator based on experience. Generally, it is 3 minutes. The initial moment of the generation cycle is T1. The operation data generated by the digital twin model within the generation cycle are confirmed one by one and sorted according to the timeline to generate an operation data sequence.

[0084] Combination Figure 2 , based on the determined initial time and end time of the generation cycle, the initial time is extended forward by half a generation cycle, and the end time is extended backward by half a generation cycle to lock a group of verification cycles (that is, the length of the verification cycle is twice the length of the generation cycle), and the corresponding multi-source data monitored in the verification cycle are sorted to determine the multi-source data sequence;

[0085] The running data sequence is checked with the multi-source data sequence, and each checking process is determined: the initial data of the running data sequence is first aligned with the initial data of the multi-source data sequence, the first checking process is determined, and it is identified whether the running data and the multi-source data at each same position in the checking process are consistent. If they are inconsistent, the second checking process is performed. If they are consistent, this checking process is marked as the standard process; the initial data of the running data sequence is aligned with the second set of data of the multi-source data sequence, the second checking process is determined, and whether they are consistent in this checking process, and so on, until the last running data of the running data sequence is aligned with the last running data of the multi-source data sequence, and the process stops;

[0086] Identify whether there is a standard process among several verification processes:

[0087] If a standard process exists, the determined standard process is transmitted to the parameter adjustment unit;

[0088] If there is no standard process, the operation data at the same position in each verification process is subjected to difference processing with the multi-source data to determine several groups of difference values ​​in the corresponding verification process, and then the several groups of difference values ​​are subjected to variance processing to determine the standard value in the corresponding verification process. Based on the different standard values ​​corresponding to each verification process, a group of minimum values ​​is determined, and the verification process corresponding to the minimum value is calibrated as a sub-standard process;

[0089] For example, the generation cycle of a certain multi-source data is 1 hour, the initial time is 9:00, and the end time is 10:00.

[0090] Then the verification period is from 8:00 to 11:00, and the multi-source data monitored within this verification period are sorted to obtain a multi-source data sequence.

[0091] Assume the running data sequence is: [10,15,20,25,30]

[0092] The multi-source data sequence is: [8,12,18,22,28]

[0093] First, the initial data 10 of the running data sequence is aligned with the initial data 8 of the multi-source data sequence, and a first verification process is performed. If inconsistency is found, a second verification process is performed.

[0094] In the second checking process, the initial data 10 of the running data sequence is aligned with the second set of data 12 of the multi-source data sequence. If inconsistency is found, the next checking process is continued.

[0095] In the third checking process, the initial data 10 of the running data sequence is aligned with the third set of data 18 of the multi-source data sequence. If inconsistency is found, the process continues.

[0096] In the fourth checking process, the initial data 10 of the running data sequence is aligned with the fourth set of data 22 of the multi-source data sequence. If inconsistency is found, the process continues.

[0097] In the fifth checking process, the initial data 10 of the running data sequence is aligned with the fifth set of data 28 of the multi-source data sequence, and inconsistencies are found.

[0098] Since there is no standard process, the difference between each verification process is calculated next:

[0099] The first check process difference: [2,3,2,3,2]

[0100] The second verification process difference: [-2,3,-2,3,-2]

[0101] The third verification process difference: [-8,-3,2,3,-2]

[0102] The fourth check process difference: [-12,-7,-3,3,2]

[0103] The fifth verification process difference: [-18,-13,-8,-3,2]

[0104] The variance of each check process difference is then calculated:

[0105] First check process variance: 2.8

[0106] Second check process variance: 2.8

[0107] The third check process variance: 18.8

[0108] Fourth check process variance: 42.8

[0109] Fifth check process variance: 92.8

[0110] Since the variances of the first and second verification processes are the same and the smallest, either one of them is calibrated as the substandard process.

[0111] Assume that the first verification process is selected as the substandard process.

[0112] Example 3

[0113] The parameter adjustment unit, based on the determined standard process or substandard process, makes parameter corrections to the generated digital twin model to ensure the data output accuracy of the digital twin model. The specific method of performing parameter correction is as follows:

[0114] If the determined verification process is a standard process, a time difference correction is performed: based on the time difference CZ between the multi-source data and the operating data at the same position in the corresponding standard process, the digital twin model is time-corrected. If CZ>0 (indicating that the operating data is output in advance), the operating data corresponding to the time point with a time difference of CZ before the current moment is used as the output data. If CZ<0 (indicating that the operating data is output with a delay), the operating data corresponding to the time point with a time difference of CZ after the current moment is used as the output data. There is no need to count CZ=0 here, because when CZ=0, it means that the operating data corresponding to the current moment and the multi-source data are of the same type and there is no time delay, so there is no need for relevant statistical processing.

[0115] If the verification process is determined to be a substandard process, determine the difference between its operating data and multi-source data at the same position, then average several groups of differences to determine the correlation value, and then use the same method of standard process time difference correction to determine the output data of the digital twin model at the current moment, and combine the output data with the correlation value (its accurate data = output data + correlation value, regardless of whether the correlation value is positive or negative) to determine the accurate data output at the current moment.

[0116] Example 4

[0117] The specific implementation process of this embodiment includes all the implementation processes of the above three groups of embodiments.

[0118] Some of the data in the above formulas are dimensionless and numerically calculated. Meanwhile, the contents not described in detail in this specification belong to the prior art known to those skilled in the art.

[0119] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A digital twin platform based on multi-source data access, characterized by: include: Access the data collection end, collect the data connected to the multiple data source interfaces connected to the digital twin platform, and transmit the collected multi-source data to the data format processing end, where each collection is a collection cycle, and the collection cycle is a preset cycle; The data format processing end performs format analysis on the multiple sets of multi-source data collected in this collection cycle, unifies the formats of the multiple sets of multi-source data, and transmits the multiple sets of multi-source data with unified formats to the virtual space generation end; The virtual space generation end constructs the virtual space based on the processed multi-source data in the same format, performs data mapping in the virtual space, and generates digital twin models corresponding to several groups of multi-source data; At the data verification processing end, the operating data generated by the digital twin model is compared with the corresponding multi-source data collected, abnormal data is identified, and parameter adjustments are made.

2. The digital twin platform based on multi-source data access according to claim 1, characterized in that: The specific method of the data format processing end to unify the formats of several groups of multi-source data is as follows: Confirm the data formats that can be processed by this digital twin platform and mark such data formats as the formats to be converted; Identify whether there is multi-source data in the same format as the one to be converted among several different sets of multi-source data: If it exists, such multi-source data is calibrated as standard data, and the format corresponding to the standard data with the largest data capacity is calibrated as the selected format; If there is no standard data, select the selected format from the formats to be converted: confirm the average conversion speed V generated during the conversion of different data formats to the selected format from the historical completed data i-k , where i represents different data formats, k represents different formats to be converted, and based on the label k, several groups belonging to the same format to be converted are converted at an average speed V i-k Perform mean processing to determine the undetermined mean J k , and then convert several groups of the same format to be converted at an average speed V i-k Perform variance processing to determine the unknown variance F k , using PD k =(J k ×C1)÷(F k ×C2) determine its evaluation value PD k , where C1 and C2 are preset fixed coefficient factors, from several groups of evaluation values ​​PD k Select the format to be converted corresponding to the maximum value, and mark the format to be converted as the selected format; Based on the determined selected format, the multi-source data in different formats are converted into the selected format, and the converted multi-source data in the same format are transmitted to the virtual space generation end.

3. The digital twin platform based on multi-source data access according to claim 1, characterized in that: The data verification processing end includes a moment data checking unit, a period verification unit and a parameter adjustment unit; Among them, the data verification unit at the same time verifies the parameter data at the same time, determines the operation data generated by its digital twin model at the current moment, and then determines the multi-source data of the data source associated with the operation data at the current moment, and verifies the operation data with the multi-source data to determine the abnormal moment; A periodic verification unit performs periodic data verification based on the determined abnormal time, and identifies whether there is data delay based on a specific data verification result; The parameter adjustment unit makes parameter corrections to the generated digital twin model based on the determined standard process or sub-standard process to ensure the data output accuracy of the digital twin model.

4. The digital twin platform based on multi-source data access according to claim 3 is characterized in that: The specific method of the data checking unit at the same time to check the operation data with the multi-source data is: The running data generated at the current moment is marked as YX q , where q represents different moments, and the corresponding multi-source data collected at the current moment is calibrated as DY q ; Identify whether its operating data and multi-source data meet: |YX q -DY q |>Y1, where Y1 is the preset value. If yes, the current moment is marked as an abnormal moment. If no, monitoring continues.

5. The digital twin platform based on multi-source data access according to claim 4 is characterized in that: The specific method of the period checking unit to identify whether there is data delay is: The determined abnormal moment T1 is periodically confirmed, the first group of generation cycles is locked, the generation cycle is a preset value, and the initial moment of the generation cycle is T1. The operation data generated by the digital twin model within the generation cycle are confirmed one by one and sorted according to the timeline to generate an operation data sequence; Based on the determined initial time and end time of the generation cycle, the initial time is extended forward by half of the generation cycle, and the end time is extended backward by half of the generation cycle to lock a group of verification cycles, and the corresponding multi-source data monitored in the verification cycle are sorted to determine the multi-source data sequence; The running data sequence is checked with the multi-source data sequence, and each checking process is determined: the initial data of the running data sequence is first aligned with the initial data of the multi-source data sequence, the first checking process is determined, and it is identified whether the running data and the multi-source data at each same position in the checking process are consistent. If they are inconsistent, the second checking process is performed. If they are consistent, this checking process is marked as the standard process; the initial data of the running data sequence is aligned with the second set of data of the multi-source data sequence, the second checking process is determined, and whether they are consistent in this checking process, and so on, until the last running data of the running data sequence is aligned with the last running data of the multi-source data sequence, and the process stops; Identify whether there is a standard process among several verification processes: If a standard process exists, the determined standard process is transmitted to the parameter adjustment unit; If there is no standard process, the operating data at the same position in each verification process is subjected to difference processing with the multi-source data to determine several groups of differences in the corresponding verification process. The several groups of differences are then subjected to variance processing to determine the standard value of the corresponding verification process. Based on the different standard values ​​corresponding to each verification process, a set of minimum values ​​is determined, and the verification process corresponding to this minimum value is calibrated as a sub-standard process.

6. The digital twin platform based on multi-source data access according to claim 5 is characterized in that: The specific method of the parameter adjustment unit to correct the parameters of the digital twin model is: If the confirmed verification process is a standard process, the time difference correction is performed: based on the time difference CZ between the multi-source data and the operating data at the same position in the corresponding standard process, the digital twin model is time-corrected. If CZ>0, the operating data corresponding to the time point with a time difference of CZ before the current moment is used as the output data. If CZ<0, the operating data corresponding to the time point with a time difference of CZ after the current moment is used as the output data. If the verification process is determined to be a substandard process, determine the difference between its operating data and multi-source data at the same location, then average several groups of differences to determine the correlation value, and then use the same method of standard process time difference correction to determine the output data of the digital twin model at the current moment, and combine the output data with the correlation value to determine the accurate data output at the current moment.

7. The digital twin platform based on multi-source data access according to claim 6, characterized in that: The accurate data=output data+correlation value.

Images