Digital twin model operation method based on Internet of Things log data processing, storage medium, equipment and computer program product

Through hierarchical mining of log data and ELMo model correction, combined with ridge regression algorithm, the problem of the difference between IoT log data and simulation data in the digital twin model is solved, the model operation accuracy and scenario consistency are improved, and the productivity conversion cycle is shortened.

CN119004783APending Publication Date: 2024-11-22CHINA TELECOM DIGITAL INTELLIGENCE TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411019542.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-29
Publication Date
2024-11-22

AI Technical Summary

Technical Problem

In actual application, the digital twin model has failed function due to the difference between the Internet of Things log data and simulation data, and the difference between virtual scenes and real scenes is large, so it is impossible to effectively drive the model to run.

Method used

Log data matching the IoT simulation data of the same operating conditions is extracted through log data hierarchical mining method, and the ELMo model is used for correction. The data with a difference of more than 10% is filtered out by combining the ridge regression algorithm, and the score value is calculated to drive the digital twin model run.

Benefits of technology

The gap between virtual scenes and real scenes in the operation of digital twin models has been narrowed, the accuracy of model operation has been improved, and the transformation cycle from theoretical to practical productivity has been shortened.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119004783A_ABST
    Figure CN119004783A_ABST
Patent Text Reader

Abstract

The invention discloses a digital twinborn model operation method based on Internet of Things log data processing, a storage medium, equipment and a computer program product, and the method comprises the steps: building a digital twinborn model of the Internet of Things, and adjusting the digital twinborn model through the simulation data of the Internet of Things; collecting log data in a real operation environment of the Internet of Things, and extracting the log data matched with the simulation data of the Internet of Things through a log data hierarchical mining method; the matched log data are input into an ELMo model to be corrected, and corrected log data are obtained; a fitting difference value between the corrected log data and the Internet of Things simulation data is obtained through a ridge regression algorithm, and the corrected log data with the fitting difference value larger than 10% is filtered out; and sequentially calculating score values of the remaining corrected log data and the Internet of Things simulation data, and driving the adjusted digital twinborn model by the corrected log data corresponding to the maximum score value. The digital twinborn model is driven to operate, and the gap between a virtual scene and a real scene for operation of the digital twinborn model is reduced.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of big data processing, and specifically, relates to a digital twin model operation method, storage medium, device and computer program product based on Internet of Things log data processing. Background Art

[0002] Digital twin is a technology that uses physical models, sensor data and simulation processes to create digital mappings of physical objects in virtual space. It integrates multi-disciplinary, multi-physical, multi-scale and multi-probability data to reflect the entire life cycle of physical equipment, thereby achieving comprehensive monitoring, optimization and prediction of physical objects. In particular, the application of the Internet of Things is the most in-depth, and the test of simulating the real environment is completed by combining the designed AI algorithm model with simulation data to build a digital twin model to simulate the effect in the real environment.

[0003] At the beginning of the creation of the digital twin model, IoT simulation data was used for optimization modeling. However, since the IoT simulation data are in an ideal state, the requirements for using it to drive the digital twin model are not high. Therefore, when the established digital twin model is put into actual application, the IoT log data used is quite different from the IoT simulation data. On the one hand, the IoT log data may not be able to drive the operation of the digital twin model, resulting in the failure of the digital twin model's function. On the other hand, even if it can run, it will lead to a large gap between the virtual scene and the real scene, deviating from the original design intention of the digital twin model. Summary of the invention

[0004] In response to the problems existing in the prior art, the present invention provides a digital twin model operation method, storage medium, device and computer program product based on Internet of Things log data processing. The processed Internet of Things log data can drive the operation of the digital twin model, narrowing the gap between the virtual scene and the real scene of the digital twin model operation.

[0005] To achieve the above technical objectives, the present invention adopts the following technical solution: a digital twin model operation method based on IoT log data processing, specifically comprising the following steps:

[0006] Step 1: Establish a digital twin model of the Internet of Things and adjust the digital twin model according to the simulation data of the Internet of Things under abnormal conditions;

[0007] Step 2: Collect log data in the real operating environment of the Internet of Things, and extract log data that matches the Internet of Things simulation data under the same working conditions through the log data hierarchical mining method;

[0008] Step 3: Input the matched log data into the ELMo model for correction to obtain the corrected log data;

[0009] Step 4: Obtain the fitting difference between the corrected log data and the corresponding IoT simulation data through the ridge regression algorithm, and filter out the corrected log data with a fitting difference greater than 10%;

[0010] Step 5: Calculate the score values ​​of the remaining corrected log data belonging to the same working condition, and use the corrected log data corresponding to the maximum score value under the same working condition to drive the adjusted digital twin model.

[0011] Furthermore, the specific process of step 2 is as follows: under the same working conditions, the log data is filtered according to the business corresponding to the IoT simulation data, and the filtered log data is sorted according to the business weight, the similarity between the sorted log data and the corresponding IoT simulation data is calculated, and the similarities are arranged in order from large to small, and the log data with the top 60% similarity is extracted.

[0012] Furthermore, step 3 includes the following sub-steps:

[0013] Step 3.1, divide each IoT simulation data into keywords;

[0014] Step 3.2: Input keywords into the ELMo model to obtain Embedding and corresponding weights, perform weighted fusion on the obtained Embedding, and obtain corrected simulation data;

[0015] Step 3.3, repeat step 3.2 until the mean square error between the corrected simulation data and the simulation data converges, completing the training of the ELMo model;

[0016] Step 3.4: Input the keywords of the matched log data into the trained ELMo model to obtain the corrected log data. Furthermore, the calculation process of the fitting difference in step 4 is:

[0017] s=||Xθ-y|| 2 +||Γθ|| 2

[0018] Among them, s represents the fitting difference between the corrected log data and the IoT simulation data, X represents the corrected log data, y represents the corresponding IoT simulation data, θ represents the fitting hyperparameter, and Γ represents the weight constant.

[0019] Furthermore, step 5 includes the following sub-steps:

[0020] Step 5.1, classify the remaining modified log data belonging to the same working condition into one category;

[0021] Step 5.2: For each type of correction log data, the correction log data is standardized based on the corresponding IoT simulation data, and the score of each correction log data under the same working condition is calculated according to the weight of each business in the correction log data;

[0022] Step 5.3: Use the corrected log data corresponding to the maximum score value under the same working conditions to drive the adjusted digital twin model.

[0023] Furthermore, the calculation process of the score value of each modified log data under the same working condition in step 5.2 is:

[0024]

[0025] Among them, P i represents the score value of the i-th modified log data under the same working condition, n represents the number of all businesses in the modified log data, j represents the index of n, and w ij represents the weight ratio of the jth business in the i-th correction log data, x ij represents the quantized value of the jth business item in the i-th modified log data, y j It represents the quantitative value of the jth business in the IoT simulation data under the corresponding working condition.

[0026] Furthermore, the present invention also provides a computer-readable storage medium storing a computer program, wherein the computer program enables a computer to execute the digital twin model operation method based on IoT log data processing.

[0027] Furthermore, the present invention also provides an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the digital twin model operation method based on IoT log data processing is implemented.

[0028] Furthermore, the present invention also provides a computer program product, including a computer program, which, when executed by a processor, implements the digital twin model operation method based on Internet of Things log data processing.

[0029] Compared with the prior art, the present invention has the following beneficial effects: the digital twin model operation method based on the processing of the Internet of Things log data of the present invention extracts the log data matching the Internet of Things simulation data of the same working condition through the log data hierarchical mining method, and the massive log data can be hierarchically optimized through the log hierarchical mining method to reduce the reading load and retrieval positioning, and the extracted log data matching the Internet of Things simulation data of the same working condition can preliminarily reduce the difference between the real data and the Internet of Things simulation data, and provide a basis for using real data to drive the operation of the digital twin model; the matched log data is further corrected through the ELMo model, so that the log data is corrected according to the logic of the Internet of Things simulation data, and is as close as possible to the data that the digital twin model can run, thereby improving the operation accuracy of the digital twin model; in addition, the corrected log data is filtered through the ridge regression algorithm to filter out the data with a large deviation from the Internet of Things simulation data; and the remaining corrected log data is scored to find the corrected log data closest to the Internet of Things simulation data under the same working condition, which can drive the digital twin model, reduce the gap between the virtual scene and the real scene of the digital twin model operation, and shorten the cycle of the digital twin model from theory to actual productivity transformation and have certain social value. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a flow chart of the digital twin model operation method based on IoT log data processing of the present invention;

[0031] Figure 2 This is a flowchart of correcting log data acquisition in the present invention. DETAILED DESCRIPTION

[0032] The technical solution of the present invention is further explained below in conjunction with the accompanying drawings.

[0033] like Figure 1 The present invention is a flowchart of a digital twin model operation method based on IoT log data processing, which specifically includes the following steps:

[0034] Step 1: Establish a digital twin model of the Internet of Things and adjust the digital twin model according to the Internet of Things simulation data under different working conditions.

[0035] Step 2, collect log data under the real operating environment of the Internet of Things, and extract log data matching the Internet of Things simulation data of the same working condition through the log data hierarchical mining method. During the operation of the Internet of Things, the log data is very large. For example, in some large systems that provide global services, the amount of daily log data can reach tens of TB, and Microsoft's online service system even generates more than 1pbe of logs every day. If the massive log data of the Internet of Things is directly collected by the server to the log for analysis, not only the data processing volume is large, but also the network operation pressure is huge. The present invention can hierarchically optimize the massive log data through the log hierarchical mining method, reduce the reading load and retrieval positioning. At the same time, due to the large difference between the simulation data used for virtual and reality modeling and the real data in the digital twin modeling process, the application of the digital twin model in the real scene and the simulation scene will be large, which cannot well reflect the problem of the actual application of the digital twin model in the production scene. The present invention can preliminarily reduce the difference between the real data and the Internet of Things simulation data by extracting the log data matching the Internet of Things simulation data of the same working condition, and provide a basis for using real data to drive the operation of the digital twin model. Specifically, under the same working conditions, the log data is filtered according to the business corresponding to the IoT simulation data, and the filtered log data is sorted according to the business weight. The similarity between the sorted log data and the corresponding IoT simulation data is calculated, and the similarities are arranged in order from large to small, and the log data with the top 60% similarity is extracted.

[0036] Step 3. The IoT simulation data is the data closest to the satisfaction of the digital twin model simulated in order to verify the feasibility of the digital twin model. Therefore, the IoT simulation data does not take into account the complex business conditions in the actual environment; and the log data is the real data of the IoT operation. Since the real data is not ideal data, the real data may not be able to drive the digital twin model. The real data needs to be corrected according to the IoT simulation data, and the matching log data is input into the ELMo model for correction to obtain the corrected log data. Since the ELMo model uses bidirectional LSTM, the IoT simulation data is used in the training process, and the forward and reverse sequence information of the IoT simulation data is considered at the same time, which can obtain richer feature information, so that the log data is corrected according to the IoT simulation data logic, as close as possible to the data that the digital twin model can run, thereby improving the operation accuracy of the digital twin model; Figure 2 , specifically including the following sub-steps:

[0037] Step 3.1, divide each IoT simulation data into keywords;

[0038] Step 3.2: Input keywords into the ELMo model to obtain Embedding and corresponding weights, perform weighted fusion on the obtained Embedding, and obtain corrected simulation data;

[0039] Step 3.3, repeat step 3.2 until the mean square error between the corrected simulation data and the simulation data converges, completing the training of the ELMo model;

[0040] Step 3.4: Input the keywords of the matched log data into the trained ELMo model to obtain the corrected log data.

[0041] Step 4: For the corrected log data that has a large gap with the IoT simulation data, it cannot drive the digital twin model well, and this part of the data will occupy the data space of the digital twin model and consume computing resources. The ridge regression algorithm is used to obtain the fitting difference between the corrected log data and the corresponding IoT simulation data, and the corrected log data with a fitting difference greater than 10% is filtered out. The ridge regression algorithm can intuitively obtain the difference between the IoT simulation data and the corrected log data, thereby performing data filtering, thereby improving computing efficiency and reducing resource consumption.

[0042] The calculation process of the fitting difference in the present invention is:

[0043] s=||Xθ-y|| 2 +||Γθ|| 2

[0044] Among them, s represents the fitting difference between the corrected log data and the IoT simulation data, X represents the corrected log data, y represents the corresponding IoT simulation data, θ represents the fitting hyperparameter, and Γ represents the weight constant.

[0045] Step 5: Calculate the score values ​​of the remaining corrected log data belonging to the same working condition, and use the corrected log data corresponding to the maximum score value under the same working condition to drive the adjusted digital twin model. For the corrected log data under the same working condition, the higher the score, the closer the corrected log data is to the simulation data. The more accurate and effective the calculation is, the smaller the gap between the virtual scene and the real scene of the digital twin model. The specific steps include the following:

[0046] Step 5.1, classify the remaining modified log data belonging to the same working condition into one category;

[0047] Step 5.2: For each type of correction log data, standardize the correction log data based on the corresponding IoT simulation data, and calculate the score of each correction log data under the same working conditions according to the weight of each business in the correction log data:

[0048]

[0049] Among them, P i represents the score value of the i-th modified log data under the same working condition, n represents the number of all businesses in the modified log data, j represents the index of n, and w ij represents the weight ratio of the jth business in the i-th correction log data, x ij represents the quantized value of the jth business item in the i-th modified log data, y j It represents the quantitative value of the jth business in the IoT simulation data under the corresponding working condition.

[0050] Step 5.3: Use the corrected log data corresponding to the maximum score value under the same working conditions to drive the adjusted digital twin model.

[0051] The digital twin model operation method based on the processing of the Internet of Things log data of the present invention can perform hierarchical optimization on massive log data through a log hierarchical mining method, reduce the reading load and retrieval positioning, and the extracted log data matching the Internet of Things simulation data under the same working condition can preliminarily reduce the difference between the real data and the Internet of Things simulation data, and provide a basis for using real data to drive the operation of the digital twin model; the matched log data is further corrected by the ELMo model, so that the log data is corrected according to the logic of the Internet of Things simulation data, and is as close as possible to the data that the digital twin model can run, thereby improving the operation accuracy of the digital twin model; in addition, the data with a large deviation from the Internet of Things simulation data is filtered out, and the remaining corrected log data is scored to find the corrected log data that is closest to the Internet of Things simulation data under the same working condition, which can drive the digital twin model, reduce the gap between the virtual scene and the real scene of the digital twin model operation, and shorten the cycle of the digital twin model from theory to actual productivity transformation and have certain social value.

[0052] In one technical solution of the present invention, a computer-readable storage medium is also provided, storing a computer program, wherein the computer program enables a computer to execute the digital twin model operation method based on IoT log data processing.

[0053] In one technical solution of the present invention, an electronic device is also provided, including: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the digital twin model operation method based on IoT log data processing is implemented.

[0054] In one technical solution of the present invention, a computer program product is also provided, including a computer program, which, when executed by a processor, implements the digital twin model operation method based on Internet of Things log data processing.

[0055] In the embodiments disclosed in the present application, the computer storage medium can be a tangible medium that can contain or store a program for use by or in conjunction with an instruction execution system, device or equipment. The computer storage medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the above. More specific examples of computer storage media can include electrical connections based on one or more lines, portable computer disks, hard disks, random access memories (RAM), read-only memories (ROM), erasable programmable read-only memories (EPROM or flash memory), optical fibers, portable compact disk read-only memories (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the above.

[0056] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in this application can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0057] The above are only preferred embodiments of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should be regarded as the protection scope of the present invention.

Claims

1. A digital twin model operation method based on IoT log data processing, characterized in that: The specific steps include: Step 1: Establish a digital twin model of the Internet of Things and adjust the digital twin model according to the simulation data of the Internet of Things under abnormal conditions; Step 2: Collect log data in the real operating environment of the Internet of Things, and extract log data that matches the Internet of Things simulation data under the same working conditions through the log data hierarchical mining method; Step 3: Input the matched log data into the ELMo model for correction to obtain the corrected log data; Step 4: Obtain the fitting difference between the corrected log data and the corresponding IoT simulation data through the ridge regression algorithm, and filter out the corrected log data with a fitting difference greater than 10%; Step 5: Calculate the score values ​​of the remaining corrected log data belonging to the same working condition, and use the corrected log data corresponding to the maximum score value under the same working condition to drive the adjusted digital twin model.

2. According to claim 1, a digital twin model operation method based on IoT log data processing is characterized in that: The specific process of step 2 is as follows: under the same working conditions, the log data is filtered according to the business corresponding to the IoT simulation data, and the filtered log data is sorted according to the business weight, the similarity between the sorted log data and the corresponding IoT simulation data is calculated, and the similarities are arranged in order from large to small, and the log data with the top 60% similarity is extracted.

3. According to claim 1, a digital twin model operation method based on IoT log data processing is characterized in that: Step 3 includes the following sub-steps: Step 3.1, divide each IoT simulation data into keywords; Step 3.2: Input keywords into the ELMo model to obtain Embedding and corresponding weights, perform weighted fusion on the obtained Embedding, and obtain corrected simulation data; Step 3.3, repeat step 3.2 until the mean square error between the corrected simulation data and the simulation data converges, completing the training of the ELMo model; Step 3.4: Input the keywords of the matched log data into the trained ELMo model to obtain the corrected log data.

4. According to claim 3, a digital twin model operation method based on IoT log data processing is characterized in that: The calculation process of the fitting difference in step 4 is: s=||Xθ-y|| 2 +||Γθ|| 2 Among them, s represents the fitting difference between the corrected log data and the IoT simulation data, X represents the corrected log data, y represents the corresponding IoT simulation data, θ represents the fitting hyperparameter, and Γ represents the weight constant.

5. According to claim 4, a digital twin model operation method based on IoT log data processing is characterized in that: Step 5 includes the following sub-steps: Step 5.1, classify the remaining modified log data belonging to the same working condition into one category; Step 5.2: For each type of correction log data, the correction log data is standardized based on the corresponding IoT simulation data, and the score of each correction log data under the same working condition is calculated according to the weight of each business in the correction log data; Step 5.3: Use the corrected log data corresponding to the maximum score value under the same working conditions to drive the adjusted digital twin model.

6. According to claim 5, a digital twin model operation method based on IoT log data processing is characterized in that: The calculation process of the score value of each corrected log data under the same working conditions in step 5.2 is: Among them, P i represents the score value of the i-th modified log data under the same working condition, n represents the number of all businesses in the modified log data, j represents the index of n, and w ij represents the weight ratio of the jth business in the i-th correction log data, x ij represents the quantized value of the jth business item in the i-th modified log data, y j It represents the quantitative value of the jth business in the IoT simulation data under the corresponding working condition.

7. A computer-readable storage medium storing a computer program, characterized in that: The computer program enables the computer to execute the digital twin model operation method based on Internet of Things log data processing as described in any one of claims 1 to 6.

8. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the digital twin model operation method based on IoT log data processing is implemented as described in any one of claims 1 to 6.

9. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the digital twin model operation method based on Internet of Things log data processing described in any one of claims 1 to 6 is implemented.