Data transmission test and optimization method based on digital twin cloud test platform

Abstract

The invention discloses a data transmission testing and optimizing method based on a digital twinning cloud testing platform, and is applied to the technical field of digital twinning. Comprising the following steps: constructing a digital twin model of a physical equipment entity on a cloud test platform, and deploying data acquisition and transmission equipment on the physical equipment entity side to form a digital twin cloud test platform; different working conditions are simulated respectively, and functions of the physical equipment entity are tested through the digital twin cloud test platform; the real-time performance and the consistency of the data transmission process are analyzed; inputting the real-time performance and consistency analysis result, the working condition data and the log data into a data transmission diagnosis model, and identifying a transmission performance reduction reason; and generating an optimization strategy according to the diagnosis result, and carrying out optimization adjustment on the digital twin cloud test platform. According to the method, the problems of insufficient data transmission real-time performance and consistency distortion of the digital twin cloud test platform under complex working conditions are solved, and the test accuracy and reliability are guaranteed.

Description

Technical Field

[0001] This invention relates to the field of digital twin technology, and more specifically to a data transmission testing and optimization method based on a digital twin cloud testing platform. Background Technology

[0002] Cloud testing is a testing model that leverages cloud platforms to allocate testing tools, equipment, and engineer resources, providing full-lifecycle testing services for software functionality, compatibility, and performance. It offers advantages such as low testing costs, high efficiency, and rapid setup. Digital twins fully utilize physical models, sensors, and operational history data, integrating multi-disciplinary, multi-physical, multi-scale, and multi-probability simulation processes to create a virtual map that reflects the entire lifecycle of the corresponding physical equipment. For applications combining software and hardware, using a digital twin cloud testing platform can combine these advantages, visually displaying the operational status of the physical equipment through a digital twin model in virtual space. However, under complex operating conditions and with variable network environments, data interaction between the digital twin cloud testing platform and the physical equipment may experience errors and delays. Data transmission delays cause the virtual digital twin model's state to lag behind the physical equipment's state, hindering real-time decision-making. Data loss, out-of-order delivery, or synchronization errors can cause deviations between the model's presented state and the actual physical state, severely impacting testing accuracy and control reliability. Existing cloud testing platforms still focus on testing the software functionality itself. For data transmission, they can only assess latency and data accuracy based on the transmission results, lacking a comprehensive testing solution that covers the entire data acquisition, transmission, and mapping process, and can diagnose and dynamically optimize data transmission performance. Therefore, providing a data transmission testing and optimization method based on a digital twin cloud testing platform is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0003] In view of this, the present invention provides a data transmission testing and optimization method based on a digital twin cloud testing platform, which evaluates the real-time performance and consistency of data transmission under different working conditions, intelligently analyzes transmission bottlenecks, generates optimization strategies, and ensures that the digital twin model in the cloud testing platform can realistically and synchronously reflect the physical device entity status.

[0004] To achieve the above objectives, the present invention provides the following technical solution: A method for data transmission testing and optimization based on a digital twin cloud testing platform includes the following steps: S1. Construct a digital twin model of the physical device entity on the cloud testing platform, and deploy data acquisition and transmission equipment on the physical device entity side to form a digital twin cloud testing platform; S2. Generate multiple sets of different working conditions, simulate different working conditions, and test the functionality of the physical equipment entity through the digital twin cloud testing platform; S3. Based on the data transmission between the digital twin cloud test platform and the physical device entity during the test, analyze the real-time performance and consistency of the data transmission process; S4. Input the real-time performance and consistency analysis results, operating condition data and log data into the data transmission diagnostic model to identify the causes of transmission performance degradation and obtain diagnostic results.

[0005] Optionally, S2 specifically involves: presetting multiple typical scenario operating condition parameters, randomly generating multiple sets of different operating conditions, and using operating condition data including network basic parameters, platform performance parameters, acquisition device parameters, and transmission fault parameters to simulate different operating conditions. The functions of the physical device entity are then tested in the digital twin cloud test platform, and the transmission data between the digital twin model and the physical device entity is collected during the test.

[0006] Optionally, S3 specifically involves: under different operating conditions, collecting the sending time of the transmitter, the arrival time of the receiver, and the state update time of the digital twin model after receiving the data during data transmission; calculating the end-to-end delay and the digital twin model delay; and calculating the average delay, delay standard deviation, data transmission ratio, and maximum delay based on the delay. During data transmission, calculating the actual state data of the physical device entity and the state data updated by the digital twin model after receiving the data, and calculating numerical consistency, distribution consistency, and temporal consistency respectively; and calculating the mean of numerical differences, the maximum of numerical differences, and the number of consistency offsets based on the consistency.

[0007] Optionally, S4 specifically involves: taking the real-time and consistency analysis results, operating condition data, and log data as input data, refining and filtering the input data, inputting the refined and filtered data into the data transmission diagnostic model, identifying the key operating condition feature combinations that lead to the degradation of consistency and real-time performance, locating the defect location of the digital twin cloud test platform, and outputting diagnostic results.

[0008] Optionally, the input data can be purified and filtered by classifying the input data and standardizing the features of numerical data. ; In the formula , for the standardized first i Each feature data, For the first i One set of original feature data, for The minimum value, for The maximum value; the pre-trained TF-IDF model is used to process word vectors for text data to obtain feature word vectors; the DBSCAN clustering algorithm is used to cluster the data, and the cluster centers are used as the purification and screening results of the data clusters to obtain the purified and screened dataset.

[0009] Optionally, the data transmission diagnostic model can be combined with data transmission expertise graphs to diagnose the causes of performance degradation and generate optimization and adjustment strategies.

[0010] Optionally, after S4, it also includes: S5. Generate optimization strategies based on the diagnostic results obtained in S4, and optimize and adjust the digital twin cloud test platform.

[0011] Optionally, S5 specifically involves: based on the diagnostic results of the data transmission diagnostic model, outputting targeted optimization strategies, including: data transmission strategy optimization, computational processing strategy optimization, and digital twin model mapping step size optimization; sending the targeted optimization strategies to the digital twin cloud test platform and physical device entities; and executing the optimization strategies.

[0012] Optionally, after S5, it also includes: S6. After optimizing and adjusting the digital twin cloud test platform, set the working condition data to be consistent with the data before optimization and adjustment. Test the function of the physical device entity in the digital twin cloud test platform. Analyze the changes in the real-time performance and consistency of the data transmission process based on the data transmission between the digital twin cloud test platform and the physical device entity during the test, and evaluate the effect of optimization and adjustment.

[0013] As can be seen from the above technical solution, compared with the prior art, the present invention provides a data transmission testing and optimization method based on a digital twin cloud testing platform, which has the following beneficial effects: The present invention tests the real-time performance and consistency of digital twin cloud testing scenarios. By utilizing the cloud testing platform to flexibly simulate complex and ever-changing real network environments, it can expose transmission problems that may occur under specific harsh working conditions in advance, thereby improving system stability; The present invention introduces a diagnostic model to analyze test data, determine the causes of data transmission performance problems, and generate optimization strategies, thereby achieving proactive optimization of network transmission performance; The present invention tests and verifies the optimization scheme, forming a complete process scheme of testing, analysis, optimization, and verification, ensuring the long-term reliable operation of the digital twin cloud testing platform. Attached Figure Description

[0014] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0015] Figure 1 This is a flowchart of the data transmission testing and optimization method based on the digital twin cloud testing platform of the present invention. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] This invention discloses a data transmission testing and optimization method based on a digital twin cloud testing platform, such as... Figure 1 As shown, it includes the following steps: S1. Construct a digital twin model of the physical device entity on the cloud testing platform, and deploy data acquisition and transmission equipment on the physical device entity side to form a digital twin cloud testing platform; S2. Generate multiple sets of different working conditions, simulate different working conditions, and test the functionality of the physical equipment entity through the digital twin cloud testing platform; S3. Based on the data transmission between the digital twin cloud test platform and the physical device entity during the test, analyze the real-time performance and consistency of the data transmission process; S4. Input the real-time performance and consistency analysis results, operating condition data and log data into the data transmission diagnostic model to identify the causes of transmission performance degradation and obtain diagnostic results.

[0018] Furthermore, S2 specifically involves: pre-setting multiple typical scenario operating condition parameters, randomly generating multiple sets of different operating conditions, and using operating condition data including network basic parameters, platform performance parameters, acquisition device parameters, and transmission fault parameters to simulate different operating conditions. The functions of the physical device entity are then tested in the digital twin cloud test platform, and the transmission data between the digital twin model and the physical device entity is collected during the test.

[0019] In this embodiment of the invention, the network basic parameters specifically include: bandwidth, which, by setting upper and lower limits on bandwidth, can simulate narrowband IoT and other situations in complex working conditions; basic latency, which simulates data transmission distance by adding a fixed basic latency; latency fluctuation, which simulates network fluctuation by introducing random latency fluctuations; and packet loss, which simulates data loss by adding random packet loss. Platform performance parameters specifically include: the performance of computing resources such as CPU, memory, and I / O, and the CPU load. Data acquisition device parameters specifically include: sensor sampling frequency, reported data volume, and reported data priority. Transmission fault parameters specifically include: communication network interruption, which simulates network interruption by randomly setting the interruption time node and duration; device restart; and erroneous data, which randomly injects abnormal data into the data. Based on the randomly generated working conditions, multiple typical scenarios can be preset, such as 4G network scenarios and channel interference scenarios.

[0020] Furthermore, S3 specifically involves: under different operating conditions, collecting the sending time of the transmitter, the arrival time of the receiver, and the state update time of the digital twin model after receiving the data during data transmission; calculating the end-to-end delay and the digital twin model delay; and calculating the average delay, delay standard deviation, data transmission ratio, and maximum delay based on the delay. During data transmission, calculating the actual state data of the physical device entity and the updated state data of the digital twin model after receiving the data, and calculating the following respectively: numerical consistency (for the values ​​corresponding to sensors, directly calculating the numerical differences, such as temperature and pressure data); distribution consistency (used to evaluate the spatial distribution difference of a data point, calculating the cosine similarity or Euclidean distance between two state vectors in the feature space to characterize distribution consistency, the closer the distance, the higher the distribution consistency); and temporal consistency (used to determine whether the curves, phases, and frequency characteristics changing over time are consistent, calculating the mean of numerical differences, the maximum of numerical differences, and the number of consistency offsets based on consistency).

[0021] In this embodiment of the invention, the real-time performance index is calculated as follows: Calculate end-to-end delay: ; In the formula, For the first k End-to-end latency of a data packet For the first k The cloud test platform receives the timestamp of each data packet. For the first k The timestamp of the data packet's transmission; Calculate the latency of the digital twin model: ; In the formula, For the first k The digital twin model of the data packet delay For the first k The digital twin model of each data packet's state update timestamp; Calculate the average delay: ; In the formula, For average delay, N For the number of data packets, Indicates the first k The delay of a data packet can be or ; Calculate the standard deviation of delay: ; In the formula, The standard deviation of the delay; Calculate the data transfer ratio: ; In the formula, This represents the data transmission ratio. The number of data packets sent. The number of data packets received; Calculate the maximum latency: ; In the formula, Maximum delay; The consistency index is calculated as follows: Numerical consistency calculation: Calculate the numerical difference: ; In the formula, For the first k The numerical difference between individual data packets, For the first k The actual physical device status value of each data packet. For the first k The digital twin model of each data packet updates the state value; Calculate the mean of numerical differences: ; In the formula, The mean of the numerical differences; Calculate the maximum numerical difference: ; In the formula, This represents the maximum numerical difference. Distribution consistency calculation: Calculate cosine similarity: ; ; ; In the formula, For cosine similarity, This represents the actual state vector of the physical device. Update the state vector for the digital twin model. m The number of state dimensions; The closer the value is to 1, the higher the consistency of the distribution; Calculate Euclidean distance: ; In the formula, For Euclidean distance, For the first j The actual value of the physical device in each state. For the first j The digital twin model updates the values ​​for each state; The smaller the value, the higher the distribution consistency; Timing consistency calculation: For time series data and Phase shift is calculated through cross-correlation analysis: ; Take The largest As a timing offset; if If the value is less than the preset threshold, it is recorded as a consistency offset, and the total number of offsets during the test is counted.

[0022] In addition to the raw data, the transmitted data during the test also includes an identifier and a high-precision timestamp from the sender. The timestamp is added just before the data is transmitted to separate the sensing delay from the transmission delay.

[0023] Furthermore, S4 specifically involves: taking real-time and consistency analysis results, operating condition data, and log data as input data, refining and filtering the input data, inputting the refined and filtered data into the data transmission diagnostic model, identifying key operating condition feature combinations that lead to consistency and real-time degradation, locating the defect locations of the digital twin cloud test platform, and outputting diagnostic results.

[0024] Furthermore, the purification and filtering of the input data specifically involves classifying the input data and performing feature standardization on numerical data. ; In the formula , for the standardized first i Each feature data, For the first iOne set of original feature data, for The minimum value, for The maximum value; the pre-trained TF-IDF model is used to process word vectors for text data to obtain feature word vectors; the DBSCAN clustering algorithm is used to cluster the data, and the cluster centers are used as the purification and screening results of the data clusters to obtain the purified and screened dataset.

[0025] In this embodiment of the invention, the DBSCAN clustering algorithm is used to cluster the data as follows: For a sample point If a point is defined as a core point, it is considered a core point if its radius eps neighborhood contains at least the minimum number of samples. If the number of samples in its radius eps neighborhood is less than the minimum number of samples, but it is within the radius eps neighborhood of other core points, it is considered a boundary point. If it is neither a core point nor within the radius eps neighborhood of other core points, it is considered a noise point. In the sample set Choose any point ,judge Can it be used as the core point? If so, then... Data clusters are formed around the central point; from Select other points within the radius eps neighborhood ,judge Can it be used as the new core point of this data cluster? If the core point definition is not met, then... As a boundary point, if the condition is met, then... Points within the radius eps of the given data cluster are added to the resulting data cluster. Simultaneously, the process of identifying core points continues within each data cluster until all core points in the cluster have been traversed. All points in the data cluster are then removed from the sample set. After removing the data points, the process of building data clusters is repeated for the remaining sample points until the sample set is complete. All points are subjected to core point judgment and clustering. The remaining points are noise points, which are then removed. The multiple data clusters obtained from the clustering are the purified and filtered datasets.

[0026] Furthermore, the data transmission diagnostic model combines data transmission expertise graphs to diagnose the causes of performance degradation and generate optimization and adjustment strategies.

[0027] In this embodiment of the invention, the knowledge graph is built based on professional data from fields such as data transmission, digital twin models, and cloud platform testing. Specifically: After acquiring professional materials, the data is divided into structured and unstructured data. Structured data includes equipment lists, historical fault records, configuration files, etc., while unstructured data includes operation and maintenance manuals, fault analysis reports, technical documents, expert experience records, and other texts. Knowledge entities and entity management are defined for the knowledge graph, such as equipment categories, software, events, and policies. Relationships between different entities are manually labeled. Then, knowledge entities are extracted from the data, converting text data into triples in the form (entity, relation, entity). Knowledge entity extraction uses the Word2Vec-BiLSTM-CRF model. Word2Vec converts sentences in the data into feature vectors, BiLSTM extracts hidden states based on the feature vectors, and CRF calculates the probability that data features belong to different labels. The label sequence with the highest score is obtained as the knowledge entity extraction result. CRF computation, with input feature H, outputs the label. The probability of: ; In the formula, The potential function of the CRF module. i ∈ n , , They represent , The true label.

[0028] After identifying all knowledge entities, relation extraction of triples is performed based on matching rules, and the extracted triples are stored in a graph database. In this embodiment, the Neo4j database is selected, the data transmission diagnostic model is Nemotron 3Nano, and the model input is a Prompt structure, for example: You are a data transmission diagnostic expert for a digital twin cloud testing platform. Please analyze the following pre-processed data and, combining your expertise in network transmission and digital twins, complete the diagnostic task.

[0029] The data includes: real-time metrics, consistency metrics, operating parameters, and log summaries.

[0030] Your task is to correlate performance metrics with specific operating parameters and log events, distinguishing whether the problem stems from the network environment, platform computing resources, or the data acquisition device itself, and to invoke professional knowledge during the inference process. It should be noted that the LLMs used in this embodiment employ existing technologies disclosed prior to the application date. The improvement to the testing optimization method for the cloud testing platform in this embodiment does not involve the model architecture, training methods, or parameter optimization of the LLMs themselves; those skilled in the art can implement this invention without making any modifications to the internal structure of the LLMs.

[0031] Furthermore, after S4, it also includes: S5. Generate optimization strategies based on the diagnostic results obtained in S4, and optimize and adjust the digital twin cloud test platform.

[0032] Furthermore, S5 specifically involves: based on the diagnostic results of the data transmission diagnostic model, outputting targeted optimization strategies, including: data transmission strategy optimization, such as adjusting the data transmission frequency, compression algorithm, and switching transmission protocols; computational processing strategy optimization, such as adjusting the data processing algorithm; digital twin model mapping step size optimization, using an interpolation algorithm to compensate for latency in the digital twin model; and sending the targeted optimization strategies to the digital twin cloud testing platform and physical device entities for execution.

[0033] Furthermore, after S5, it also includes: S6. After optimizing and adjusting the digital twin cloud test platform, set the working condition data to be consistent with the data before optimization and adjustment. Test the function of the physical device entity in the digital twin cloud test platform. Analyze the changes in the real-time performance and consistency of the data transmission process based on the data transmission between the digital twin cloud test platform and the physical device entity during the test, and evaluate the effect of optimization and adjustment.

[0034] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0035] Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for data transmission testing and optimization based on a digital twin cloud testing platform, characterized in that, Includes the following steps: S1. Construct a digital twin model of the physical device entity on the cloud testing platform, and deploy data acquisition and transmission equipment on the physical device entity side to form a digital twin cloud testing platform; S2. Generate multiple sets of different working conditions, simulate different working conditions, and test the functionality of the physical equipment entity through the digital twin cloud testing platform; S3. Based on the data transmission between the digital twin cloud test platform and the physical device entity during the test, analyze the real-time performance and consistency of the data transmission process; S4. Input the real-time performance and consistency analysis results, operating condition data and log data into the data transmission diagnostic model to identify the causes of transmission performance degradation and obtain diagnostic results.

2. The data transmission testing and optimization method based on a digital twin cloud testing platform according to claim 1, characterized in that, S2 specifically involves: pre-setting operating parameters for multiple typical scenarios, randomly generating multiple sets of different operating conditions, and using operating condition data including network basic parameters, platform performance parameters, acquisition device parameters, and transmission fault parameters to simulate different operating conditions. The functions of the physical device entity are then tested in the digital twin cloud testing platform, and the transmission data between the digital twin model and the physical device entity is collected during the testing process.

3. The data transmission testing and optimization method based on a digital twin cloud testing platform according to claim 1, characterized in that, S3 specifically involves: under different operating conditions, collecting the sending time of the transmitter, the arrival time of the receiver, and the state update time of the digital twin model after receiving the data during data transmission; calculating the end-to-end delay and the digital twin model delay; and calculating the average delay, delay standard deviation, data transmission ratio, and maximum delay based on the delay. During data transmission, calculating the actual state data of the physical device entity and the updated state data of the digital twin model after receiving the data, and calculating numerical consistency, distribution consistency, and temporal consistency respectively. Based on the consistency, calculating the mean of numerical differences, the maximum of numerical differences, and the number of consistency offsets.

4. The data transmission testing and optimization method based on a digital twin cloud testing platform according to claim 1, characterized in that, S4 specifically involves taking real-time and consistency analysis results, operating condition data, and log data as input data, refining and filtering the input data, inputting the refined and filtered data into the data transmission diagnostic model, identifying key operating condition feature combinations that lead to consistency and real-time degradation, locating the defects in the digital twin cloud test platform, and outputting diagnostic results.

5. The data transmission testing and optimization method based on a digital twin cloud testing platform according to claim 4, characterized in that, The specific steps for refining and filtering the input data are: classifying the input data and performing feature standardization on numerical data. ； In the formula , for the standardized first i Each feature data, For the first i One set of original feature data, for The minimum value, for The maximum value; the pre-trained TF-IDF model is used to process word vectors for text data to obtain feature word vectors; the DBSCAN clustering algorithm is used to cluster the data, and the cluster centers are used as the purification and screening results of the data clusters to obtain the purified and screened dataset.

6. The data transmission testing and optimization method based on a digital twin cloud testing platform according to claim 1, characterized in that, The data transmission diagnostic model combines data transmission expertise graphs to diagnose the causes of performance degradation and generate optimization strategies.

7. The data transmission testing and optimization method based on a digital twin cloud testing platform according to claim 1, characterized in that, S4 is followed by: S5. Generate optimization strategies based on the diagnostic results obtained in S4, and optimize and adjust the digital twin cloud test platform.

8. The data transmission testing and optimization method based on a digital twin cloud testing platform according to claim 5, characterized in that, S5 specifically involves: based on the diagnostic results of the data transmission diagnostic model, outputting targeted optimization strategies, including: data transmission strategy optimization, computational processing strategy optimization, and digital twin model mapping step size optimization, and sending the targeted optimization strategies to the digital twin cloud test platform and physical device entities to execute the optimization strategies.

9. The data transmission testing and optimization method based on a digital twin cloud testing platform according to claim 5, characterized in that, Following S5 are: S6. After optimizing and adjusting the digital twin cloud test platform, set the working condition data to be consistent with the data before optimization and adjustment. Test the function of the physical device entity in the digital twin cloud test platform. Analyze the changes in the real-time performance and consistency of the data transmission process based on the data transmission between the digital twin cloud test platform and the physical device entity during the test, and evaluate the effect of optimization and adjustment.

Images