Service platform data management system and method based on artificial intelligence

Abstract

The invention belongs to the technical field of data acquisition and management, and provides a service platform data management system and method based on artificial intelligence. The data collected by the headquarters main data platform and the regional main data platform are subjected to collection timestamp consistency comparison, when the phenomenon of non-real-time synchronization of data collection occurs, a data collection comparison period is set, non-real-time synchronization collected data are screened out and analyzed, and a data collection comparison result is obtained. A non-real-time synchronization time period is determined, data collected in the non-real-time synchronization time period is specially processed, it is guaranteed that data finally entering a main data platform is accurate and complete, in the non-real-time synchronization time period, the priority or frequency of data transmission can be adjusted, and it is avoided that a large amount of data is transmitted in the peak period; and network congestion and data transmission delay are reduced.

Description

Technical Field

[0001] This invention belongs to the field of data acquisition and management technology, specifically an artificial intelligence-based service platform data management system and method. Background Technology

[0002] During data integration, it is a common and necessary operating model for both the headquarters master data platform and regional master data platforms to simultaneously collect equipment data from different production lines. This distributed data collection architecture aims to fully utilize the computing and storage resources of each region while ensuring the comprehensiveness and timeliness of the data. However, in actual operation, due to factors such as network latency, differences in equipment performance, and inconsistent data transmission protocols, the problem of non-real-time synchronization in data collection is highly likely to occur.

[0003] Non-real-time synchronization in data acquisition leads to temporal discrepancies in data acquired by different platforms. This results in inaccurate and unreliable analysis and decision-making based on this data, severely impacting the service platform's ability to monitor and optimize the production process. Non-real-time synchronization may not be continuous and stable, but rather intermittent and irregular. Accurately identifying non-real-time synchronized data from massive data collection records and determining non-real-time synchronization periods based on scientifically sound indicators is one of the key issues that urgently needs to be addressed in the field of data management.

[0004] Therefore, the present invention provides a service platform data management system and method based on artificial intelligence. Summary of the Invention

[0005] In order to overcome the shortcomings of the prior art, at least one technical problem raised in the background art is solved.

[0006] The technical solution adopted by this invention to solve its technical problem is: A data management method for an artificial intelligence-based service platform includes: During the process of simultaneously collecting equipment data from different production lines from the headquarters master data platform and the regional master data platform, the data collected by the headquarters master data platform and the regional master data platform are compared to check the consistency of the collection timestamps to assess whether there is a phenomenon of non-real-time synchronization of data collection. When data collection is not synchronized in real time, a data collection comparison period is set, non-synchronized data is filtered out, and the non-synchronized data is analyzed to determine the non-synchronized period. Analyze network fluctuations during each non-real-time synchronization period to determine the non-real-time synchronization period of network waves; Comparative analysis is performed on non-simultaneous network waveforms within multiple data acquisition and comparison periods to determine the frequency types of non-simultaneous network waveforms, and the data acquisition frequency management is adjusted based on the frequency types of non-simultaneous network waveforms.

[0007] As a further aspect of the present invention, the evaluation process for the non-real-time synchronization phenomenon of data acquisition is as follows: Extract the data collected from the headquarters master data platform and the regional master data platform respectively, and compare the collected data with similar data. Mark the similar data collected in the headquarters master data platform and the regional master data platform as similar collected data. Obtain the collection timestamp of similar collected data in the headquarters master data platform and the regional master data platform respectively, and form a group of similar data collection timestamps. Substitute each data acquisition timestamp group into the Euclidean distance formula for calculation to obtain the data acquisition time difference value. If the data acquisition time difference value is greater than the data acquisition time difference threshold, it is displayed as a non-real-time synchronized signal.

[0008] As a further aspect of the present invention, the filtering process for non-real-time synchronously collected data is as follows: The set data collection and comparison period is divided into several data collection and comparison periods. Within the data collection and comparison period, the time difference between the collection timestamps of the same type of collected data in the main master data platform and in the regional master data platform is obtained, and the ratio is calculated with the duration corresponding to the data collection and comparison period to obtain the data collection timestamp difference. If the data acquisition timestamp difference is greater than the data acquisition timestamp difference threshold, the analyzed data of the same type will be marked as non-real-time synchronous acquisition data.

[0009] As a further aspect of the present invention, the process for determining the non-real-time synchronization period is as follows: The proportion of non-real-time synchronously collected data during the data collection and comparison period to the total synchronously collected data during the same period is used as the ratio of non-real-time synchronously collected data. Extract the data collection timestamp difference corresponding to each non-real-time synchronous data collection within the data collection and comparison period, and calculate the summation and average to obtain the non-real-time synchronization degree value; The non-real-time synchronization analysis value is obtained by summing the ratio of non-real-time data collection quantity to the non-real-time synchronization degree value. If the non-real-time synchronization analysis value is greater than the non-real-time synchronization analysis threshold, it is marked as a non-real-time synchronization period.

[0010] As a further aspect of the present invention, the process of analyzing network fluctuations during each non-real-time synchronization period is as follows: The non-real-time synchronization period is divided into several bandwidth traffic monitoring points. During the non-real-time synchronization period, the outgoing bandwidth traffic of each bandwidth traffic monitoring point is obtained and the ratio with the total bandwidth of the node is calculated to obtain the actual bandwidth utilization value. The values ​​are then substituted into a two-dimensional coordinate system to construct a bandwidth utilization change curve. Extract the peak and trough coordinates on the bandwidth utilization change curve, and take the actual bandwidth utilization value corresponding to the peak coordinate as the actual bandwidth utilization peak value and the actual bandwidth utilization value corresponding to the trough coordinate as the actual bandwidth utilization trough value. The average bandwidth utilization for a given time period is calculated by summing the peak and trough values ​​of all actual bandwidth utilization.

[0011] As a further aspect of the present invention, the process for determining the non-real simultaneous period of the network waves is as follows: The difference between adjacent peak values ​​and valley values ​​of actual bandwidth utilization is obtained, and the absolute value is then summed and averaged to calculate the actual bandwidth utilization fluctuation value. The average bandwidth utilization during a given time period is summed with the actual bandwidth utilization fluctuation value to obtain the bandwidth analysis value for that time period. If the bandwidth analysis value for that time period is greater than the bandwidth analysis threshold, then the non-real-time synchronous time period analyzed is marked as a non-real-time synchronous time period of the network wave.

[0012] As a further aspect of the present invention, the process of comparing and analyzing non-simultaneous periods of network waves within multiple data acquisition and comparison cycles is as follows: Within the data acquisition and comparison period, the interval duration between adjacent network waves that are not real simultaneous is obtained, and the standard deviation is calculated to obtain the standard deviation of the interval between adjacent network waves that are not real simultaneous. Obtain the number of time intervals between non-simultaneous periods of adjacent network waves, and calculate the standard deviation to obtain the standard deviation of the number of non-simultaneous periods of adjacent network waves; The non-real analysis value of the network waves is obtained by summing the standard deviation of the non-real time interval between adjacent network waves and the standard deviation of the non-real number of adjacent network waves.

[0013] As a further aspect of the present invention, the process for determining the frequency type of non-real simultaneous occurrence of network waves is as follows: If the non-real analysis value of the network wave is greater than the non-real analysis threshold of the network wave, it is displayed as a non-real isostatic signal of the network wave; If the non-real analysis value of the network wave is less than or equal to the non-real analysis threshold of the network wave, it will be displayed as a non-real wave signal.

[0014] As a further aspect of the present invention, the process of adjusting the data acquisition frequency management operation according to the frequency type of non-simultaneous occurrence of network waves is as follows: Based on the non-real stable signal of the network wave, the number of data avoidance periods is calculated by summing and averaging the number of time intervals between adjacent non-real simultaneous network waves. Based on the non-real synchronous wave signal of the network wave, the number of time intervals between adjacent non-real synchronous waves is compared, the number of the longest and shortest time intervals is extracted, and the summation and average are calculated to obtain the number of data avoidance time periods. In subsequent data collection, after a period of data collection is completed, the next data collection period is determined based on the number of data collection periods.

[0015] An artificial intelligence-based service platform data management system includes: Non-real-time synchronization assessment module: During the process of the headquarters master data platform and the regional master data platform simultaneously collecting equipment data from different production lines, the module compares the data collection timestamps collected by the headquarters master data platform and the regional master data platform to assess whether there is a phenomenon of non-real-time synchronization in data collection. Non-real-time synchronization period determination module: When the phenomenon of non-real-time synchronization of data collection occurs, the data collection comparison period is set, the non-real-time synchronized data is filtered out, and the non-real-time synchronized data is analyzed to determine the non-real-time synchronization period. Network Waveform Non-Real-Time Synchronization Analysis Module: Analyzes network fluctuations within each non-real-time synchronization period to determine the non-real-time synchronization period of network waves; Data acquisition and management module: Performs comparative analysis on non-simultaneous network waveforms within multiple data acquisition comparison periods, determines the frequency type of non-simultaneous network waveforms, and adjusts the data acquisition frequency management based on the frequency type of non-simultaneous network waveforms.

[0016] The beneficial effects of this invention are as follows: 1. In the process of simultaneously collecting equipment data from different production lines on both the headquarters master data platform and the regional master data platform, this invention compares the consistency of the collection timestamps of the data collected by the headquarters master data platform and the regional master data platform. When a non-real-time synchronization of data collection occurs, a data collection comparison period is set to filter out the non-real-time synchronized data. The non-real-time synchronized data is then analyzed to determine the non-real-time synchronization period. The data collected during the non-real-time synchronization period is then specially processed, such as marked, reviewed, or re-collected, or synchronized and corrected, to ensure that the data finally entering the master data platform is accurate and complete. This provides a reliable foundation for data analysis and decision-making based on artificial intelligence. During the non-real-time synchronization period, the priority or frequency of data transmission can be adjusted to avoid large-scale data transmission during peak hours, reducing network congestion and data transmission delays.

[0017] 2. This invention analyzes network fluctuations during each non-real-time synchronization period to determine non-real-time synchronization periods. It identifies data acquisition delays occurring when the headquarters master data platform and regional master data platforms simultaneously collect data from equipment in different production lines, attributing these delays to network fluctuations. Furthermore, by comparing and analyzing non-real-time synchronization periods across multiple data acquisition comparison cycles, it determines the frequency types of these periods and adjusts data acquisition frequency management based on these frequency types. Data acquisition periods are determined according to the number of data avoidance periods, thus preventing data acquisition delays and reducing data acquisition interruptions or loss caused by network fluctuations. Attached Figure Description

[0018] The invention will now be further described with reference to the accompanying drawings.

[0019] Figure 1 This is a flowchart illustrating the steps of a data management method for an artificial intelligence-based service platform according to the present invention. Figure 2 This is a flowchart illustrating the judgment process in a data management method for an artificial intelligence-based service platform according to the present invention. Figure 3 This is a flowchart of modules within a data management system for an artificial intelligence-based service platform according to the present invention. Detailed Implementation

[0020] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments. Example 1

[0021] Under a group-type service platform, equipment data from different production lines needs to be synchronized to the headquarters master data platform. However, when the headquarters master data platform and regional master data platforms simultaneously collect equipment data from different production lines, data collection delays may occur due to cross-regional network fluctuations or insufficient dedicated line bandwidth. Therefore, please refer to [the relevant documentation / reference needed]. Figure 1 - Figure 2 As shown in the embodiment of the present invention, a data management method for an artificial intelligence-based service platform includes the following steps: Step 1: During the process of collecting equipment data from different production lines simultaneously on the headquarters master data platform and the regional master data platform, the data collected by the headquarters master data platform and the regional master data platform are compared to check the consistency of the collection timestamps to assess whether there is a phenomenon of non-real-time synchronization of data collection. In some embodiments, data collected from the headquarters master data platform and the regional master data platform are extracted respectively, and the collected data are compared with similar data. Similar data collected from the headquarters master data platform and the regional master data platform are marked as similar collected data. It should be noted that similar data collection refers to the same data collected in both the headquarters master data platform and the regional master data platform. Collect the collection timestamps of the same type of data in the main master data platform and in the regional master data platform respectively, and form a group of the same data collection timestamps; Substitute each data collection timestamp group into the Euclidean distance formula for calculation to obtain the time difference value of data collection. If the time difference of the same data collection is greater than the threshold of the same data collection time difference, it indicates that the time difference of the same data collected in the headquarters master data platform and the regional master data platform is large, that is, the degree of data collection asynchrony is high, which is displayed as a non-real-time synchronous signal of collection. If the time difference of the same data collection is less than or equal to the threshold of the same data collection time difference, it means that the time difference of the same data collected in the headquarters master data platform and the regional master data platform is small, that is, the degree of data collection asynchrony is low, and it is displayed as a real-time synchronization signal. Step 2: When data collection is not synchronized in real time, set a data collection comparison period, filter out the non-synchronized data, analyze the non-synchronized data, and determine the non-synchronized period. In some embodiments, the set data collection and comparison period is equally divided into several data collection and comparison time periods, wherein the duration of each data collection and comparison time period is equal. During the data collection and comparison period, the time difference between the collection timestamps of similar collected data on the main master data platform and the regional master data platform is obtained, and the ratio is calculated with the duration corresponding to the data collection and comparison period to obtain the data collection timestamp difference. If the data collection timestamp difference is greater than the data collection timestamp difference threshold, it indicates that there is a large time difference between the main master data platform and the regional master data platform when collecting the same type of data during the data collection and comparison period. The collected data of the same type will be marked as non-real-time synchronous collected data. If the data collection timestamp difference is less than or equal to the data collection timestamp difference threshold, it means that there is a small time difference between the main master data platform and the regional master data platform when collecting the same type of data during the data collection and comparison period, and no processing operation is required. The proportion of non-real-time synchronously collected data during the data collection and comparison period to the total synchronously collected data during the same period is used as the ratio of non-real-time synchronously collected data. Extract the data collection timestamp difference corresponding to each non-real-time synchronous data collection within the data collection and comparison period, and calculate the summation and average to obtain the non-real-time synchronization degree value; The non-real-time synchronization analysis value is obtained by summing the ratio of non-real-time data collection quantity to the non-real-time synchronization degree value. It is understandable that the meaning of the non-real-time synchronization analysis value is: an indicator that comprehensively measures the non-real-time synchronization status of data collection, integrating the proportion of non-real-time synchronized data in quantity and the degree of non-real-time synchronization. On the one hand, the proportion of non-real-time synchronized data reflects the scale of non-real-time synchronized data in the overall data collection during the data collection and comparison period. On the other hand, the degree of non-real-time synchronization indicates the degree of time deviation of non-real-time synchronized data in the data collection and comparison period. Specifically, the larger the non-real-time synchronization analysis value, the larger the scale of non-real-time synchronized data in the overall data collection during the data collection and comparison period, and the larger the degree of time deviation of non-real-time synchronized data in the data collection and comparison period. Conversely, the smaller the non-real-time synchronization analysis value, the smaller the scale of non-real-time synchronized data in the overall data collection during the data collection and comparison period, and the smaller the degree of time deviation of non-real-time synchronized data in the data collection and comparison period. If the non-real-time synchronization analysis value is greater than the non-real-time synchronization analysis threshold, it indicates that the time deviation of the non-real-time synchronized data during the data collection and comparison period is large, and the data collection and comparison period analyzed is marked as a non-real-time synchronization period. If the non-real-time synchronization analysis value is less than or equal to the non-real-time synchronization analysis threshold, it indicates that the scale of non-real-time synchronization data in the overall data collection during the data collection and comparison period is small, and the time deviation of the non-real-time synchronization data during the data collection and comparison period is small. Therefore, the analyzed data collection and comparison period is marked as a real-time synchronization period.

[0022] In this specific implementation, the solution is as follows: During the simultaneous collection of equipment data from different production lines by the headquarters master data platform and the regional master data platform, the timestamp consistency of the collected data is compared between the headquarters master data platform and the regional master data platform. If data collection is not synchronized in real time, a data collection comparison period is set to filter out the non-synchronized data. The non-synchronized data is then analyzed to determine the non-synchronized period. The data collected during the non-synchronized period is then specially processed, such as marked, reviewed, or re-collected, or synchronized and corrected, to ensure that the data finally entering the master data platform is accurate and complete, providing a reliable foundation for data analysis and decision-making based on artificial intelligence. During the non-synchronized period, the priority or frequency of data transmission can be adjusted to avoid large amounts of data transmission during peak hours, reducing network congestion and data transmission delays. Example 2

[0023] Please see Figure 1 - Figure 2 As shown in the embodiment of the present invention, a data management method for an artificial intelligence-based service platform further includes the following steps: Step 3: Analyze the network fluctuations during each non-real-time synchronization period to determine the non-real-time synchronization period of the network waves; It should be noted that network fluctuations can be caused by fluctuations in bandwidth and traffic. In some embodiments, the non-real-time synchronization period is equally divided into several bandwidth traffic monitoring points, wherein the interval between adjacent bandwidth traffic monitoring points is equal. During non-real-time synchronization periods, the outgoing bandwidth traffic of each bandwidth traffic monitoring point is acquired and the ratio with the total bandwidth of the node is calculated to obtain the actual bandwidth utilization value. The value is then substituted into a two-dimensional coordinate system to construct a bandwidth utilization change curve with the X-axis representing time and the Y-axis representing bandwidth traffic. Extract the peak and trough coordinates on the bandwidth utilization change curve, and take the actual bandwidth utilization value corresponding to the peak coordinate as the actual bandwidth utilization peak value and the actual bandwidth utilization value corresponding to the trough coordinate as the actual bandwidth utilization trough value. The average bandwidth utilization for a given period is calculated by summing all the peak and valley values ​​of actual bandwidth utilization. The difference between adjacent peak values ​​and valley values ​​of actual bandwidth utilization is obtained, and the absolute value is then summed and averaged to calculate the actual bandwidth utilization fluctuation value. The average bandwidth of the time period is summed with the fluctuation value of the actual bandwidth to obtain the bandwidth analysis value of the time period. It is understandable that the meaning of the time period bandwidth analysis value is: an indicator that comprehensively considers the average utilization and fluctuation of network bandwidth during a specific non-real-time synchronization period, reflecting the usage characteristics and status of the non-real-time synchronization period as a whole. On the one hand, the average bandwidth utilization of the time period reflects the overall degree of network bandwidth resource utilization during the non-real-time synchronization period. On the other hand, the actual bandwidth utilization fluctuation value reflects the fluctuation range of network bandwidth during the non-real-time synchronization period, that is, the stability of bandwidth usage. Specifically, if the time period bandwidth analysis value is larger, it indicates that network bandwidth is used more frequently during the non-real-time synchronization period, resources are occupied in large quantities, and network congestion and idleness alternate more frequently. If the time period bandwidth analysis value is smaller, it indicates that network bandwidth is used less frequently during the non-real-time synchronization period, and network congestion and idleness alternate less frequently. If the bandwidth analysis value of a time period is greater than the bandwidth analysis threshold of a time period, it indicates that network bandwidth is used frequently during the non-real-time synchronization period, resources are occupied in large quantities, and network congestion and idle periods alternate frequently. The non-real-time synchronization period analyzed is marked as a non-real-time synchronization period. If the bandwidth analysis value for a given time period is less than or equal to the bandwidth analysis threshold for that time period, it indicates that network bandwidth usage is less frequent during the non-real-time synchronization period, and that network congestion and idle periods alternate less frequently. The non-real-time synchronization period analyzed is then marked as a non-network-wave non-real-time synchronization period.

[0024] Step 4: Compare and analyze the non-simultaneous network waves within multiple data acquisition comparison periods to determine the frequency type of non-simultaneous network waves, and adjust the data acquisition frequency management operation according to the frequency type of non-simultaneous network waves. In some embodiments, within the data acquisition and comparison period, the interval duration between adjacent non-real simultaneous periods of network waves is obtained, and the standard deviation is calculated to obtain the standard deviation of the non-real simultaneous interval between adjacent network waves. Obtain the number of time intervals between non-simultaneous periods of adjacent network waves, and calculate the standard deviation to obtain the standard deviation of the number of non-simultaneous periods of adjacent network waves; The non-real analysis value of the network waves is obtained by summing the standard deviation of the non-real time interval between adjacent network waves and the standard deviation of the non-real number of adjacent network waves. If the non-real analysis value of the network wave is greater than the non-real analysis threshold of the network wave, it indicates that the frequency of the non-real network wave occurring in the same period is relatively stable within the data acquisition and comparison period, which is displayed as a stable non-real network wave signal. If the non-real analysis value of the network wave is less than or equal to the non-real analysis threshold of the network wave, it indicates that the frequency of the non-real network wave occurring in the same period is relatively unstable within the data acquisition and comparison period, and it is displayed as a non-real network wave synchronous wave signal. Based on the non-real synchronous signal of the network wave, the number of time intervals between adjacent non-real synchronous time intervals of the network wave is summed and averaged to obtain the number of data avoidance time intervals. In the subsequent data acquisition process, after the data acquisition operation of one time interval is completed, the next data acquisition time interval is determined according to the number of data avoidance time intervals for data acquisition. Based on the non-real synchronous wave signal of the network wave, the number of time intervals between adjacent non-real synchronous waves is compared, the number of the longest and shortest time intervals is extracted, and the summation and average are calculated to obtain the number of data avoidance time intervals. In the subsequent data acquisition process, after the data acquisition operation of one time interval is completed, the next data acquisition time interval is determined based on the number of data avoidance time intervals for data acquisition. The specific solution in this embodiment is as follows: Analyze network fluctuations within each non-real-time synchronization period to determine the non-real-time synchronization periods. This allows us to identify data acquisition delays that occur when the headquarters master data platform and the regional master data platform simultaneously collect data from equipment in different production lines. These delays are caused by network fluctuations. Furthermore, by comparing and analyzing non-real-time synchronization periods across multiple data acquisition comparison cycles, we can determine the frequency types of these periods. Based on these frequency types, we can adjust the data acquisition frequency management, determining the data acquisition period according to the number of data avoidance periods. This helps avoid data acquisition delays during the acquisition process and reduces data acquisition interruptions or missing data due to network fluctuations. Example 3

[0025] like Figure 3 As shown in the figure, this embodiment of the invention also provides a service platform data management system based on artificial intelligence, including the following modules: Non-real-time synchronization assessment module: During the process of the headquarters master data platform and the regional master data platform simultaneously collecting equipment data from different production lines, the module compares the data collection timestamps collected by the headquarters master data platform and the regional master data platform to assess whether there is a phenomenon of non-real-time synchronization in data collection. Non-real-time synchronization period determination module: When the phenomenon of non-real-time synchronization of data collection occurs, the data collection comparison period is set, the non-real-time synchronized data is filtered out, and the non-real-time synchronized data is analyzed to determine the non-real-time synchronization period. Network Waveform Non-Real-Time Synchronization Analysis Module: Analyzes network fluctuations within each non-real-time synchronization period to determine the non-real-time synchronization period of network waves; Data acquisition and management module: Performs comparative analysis on non-simultaneous network waveforms within multiple data acquisition comparison periods, determines the frequency type of non-simultaneous network waveforms, and adjusts the data acquisition frequency management based on the frequency type of non-simultaneous network waveforms.

[0026] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A data management method for an artificial intelligence-based service platform, characterized in that: include: During the process of simultaneously collecting equipment data from different production lines from the headquarters master data platform and the regional master data platform, the data collected by the headquarters master data platform and the regional master data platform are compared to check the consistency of the collection timestamps to assess whether there is a phenomenon of non-real-time synchronization of data collection. When data collection is not synchronized in real time, a data collection comparison period is set, non-synchronized data is filtered out, and the non-synchronized data is analyzed to determine the non-synchronized period. Analyze network fluctuations during each non-real-time synchronization period to determine the non-real-time synchronization period of network waves; Comparative analysis is performed on non-simultaneous network waveforms within multiple data acquisition and comparison periods to determine the frequency types of non-simultaneous network waveforms, and the data acquisition frequency management is adjusted based on the frequency types of non-simultaneous network waveforms.

2. The data management method for an artificial intelligence-based service platform according to claim 1, characterized in that: The evaluation process for the phenomenon of non-real-time synchronization in data acquisition is as follows: Extract the data collected from the headquarters master data platform and the regional master data platform respectively, and compare the collected data with similar data. Mark the similar data collected in the headquarters master data platform and the regional master data platform as similar collected data. Obtain the collection timestamp of similar collected data in the headquarters master data platform and the regional master data platform respectively, and form a group of similar data collection timestamps. Substitute each data acquisition timestamp group into the Euclidean distance formula for calculation to obtain the data acquisition time difference value. If the data acquisition time difference value is greater than the data acquisition time difference threshold, it is displayed as a non-real-time synchronized signal.

3. The data management method for an artificial intelligence-based service platform according to claim 1, characterized in that: The filtering process for non-real-time synchronously collected data is as follows: The set data collection and comparison period is divided into several data collection and comparison periods. Within the data collection and comparison period, the time difference between the collection timestamps of the same type of collected data in the main master data platform and in the regional master data platform is obtained, and the ratio is calculated with the duration corresponding to the data collection and comparison period to obtain the data collection timestamp difference. If the data acquisition timestamp difference is greater than the data acquisition timestamp difference threshold, the analyzed data of the same type will be marked as non-real-time synchronous acquisition data.

4. The data management method for an artificial intelligence-based service platform according to claim 1, characterized in that: The process for determining the non-real-time synchronization period is as follows: The proportion of non-real-time synchronously collected data during the data collection and comparison period to the total synchronously collected data during the same period is used as the ratio of non-real-time synchronously collected data. Extract the data collection timestamp difference corresponding to each non-real-time synchronous data collection within the data collection and comparison period, and calculate the summation and average to obtain the non-real-time synchronization degree value; The non-real-time synchronization analysis value is obtained by summing the ratio of non-real-time data collection quantity to the non-real-time synchronization degree value. If the non-real-time synchronization analysis value is greater than the non-real-time synchronization analysis threshold, it is marked as a non-real-time synchronization period.

5. The data management method for an artificial intelligence-based service platform according to claim 1, characterized in that: The process of analyzing network fluctuations during each non-real-time synchronization period is as follows: The non-real-time synchronization period is divided into several bandwidth traffic monitoring points. During the non-real-time synchronization period, the outgoing bandwidth traffic of each bandwidth traffic monitoring point is obtained and the ratio with the total bandwidth of the node is calculated to obtain the actual bandwidth utilization value. The values ​​are then substituted into a two-dimensional coordinate system to construct a bandwidth utilization change curve. Extract the peak and trough coordinates on the bandwidth utilization change curve, and take the actual bandwidth utilization value corresponding to the peak coordinate as the actual bandwidth utilization peak value and the actual bandwidth utilization value corresponding to the trough coordinate as the actual bandwidth utilization trough value. The average bandwidth utilization for a given time period is calculated by summing the peak and trough values ​​of all actual bandwidth utilization.

6. The data management method for an artificial intelligence-based service platform according to claim 5, characterized in that: The process for determining the non-real simultaneous time interval of the network wave is as follows: The difference between adjacent peak values ​​and valley values ​​of actual bandwidth utilization is obtained, and the absolute value is then summed and averaged to calculate the actual bandwidth utilization fluctuation value. The average bandwidth utilization during a given time period is summed with the actual bandwidth utilization fluctuation value to obtain the bandwidth analysis value for that time period. If the bandwidth analysis value for that time period is greater than the bandwidth analysis threshold, then the non-real-time synchronous time period analyzed is marked as a non-real-time synchronous time period of the network wave.

7. The data management method for an artificial intelligence-based service platform according to claim 1, characterized in that: The process of comparing and analyzing non-real simultaneous periods of network waves within multiple data acquisition and comparison periods is as follows: Within the data acquisition and comparison period, the interval duration between adjacent network waves that are not real simultaneous is obtained, and the standard deviation is calculated to obtain the standard deviation of the interval between adjacent network waves that are not real simultaneous. Obtain the number of time intervals between non-simultaneous periods of adjacent network waves, and calculate the standard deviation to obtain the standard deviation of the number of non-simultaneous periods of adjacent network waves; The non-real analysis value of the network waves is obtained by summing the standard deviation of the non-real time interval between adjacent network waves and the standard deviation of the non-real number of adjacent network waves.

8. The data management method for an artificial intelligence-based service platform according to claim 7, characterized in that: The process for determining the frequency type of non-real simultaneous occurrence of network waves is as follows: If the non-real analysis value of the network wave is greater than the non-real analysis threshold of the network wave, it is displayed as a non-real isostatic signal of the network wave; If the non-real analysis value of the network wave is less than or equal to the non-real analysis threshold of the network wave, it will be displayed as a non-real wave signal.

9. The data management method for an artificial intelligence-based service platform according to claim 8, characterized in that: Based on the frequency types of non-simultaneous occurrence of network waves, the process for adjusting the data acquisition frequency management is as follows: Based on the non-real stable signal of the network wave, the number of data avoidance periods is calculated by summing and averaging the number of time intervals between adjacent non-real simultaneous network waves. Based on the non-real synchronous wave signal of the network wave, the number of time intervals between adjacent non-real synchronous waves is compared, the number of the longest and shortest time intervals is extracted, and the summation and average are calculated to obtain the number of data avoidance time periods. In subsequent data collection, after a period of data collection is completed, the next data collection period is determined based on the number of data collection periods.

10. A data management system for an artificial intelligence-based service platform, characterized in that: Includes the following modules: Non-real-time synchronization assessment module: During the process of the headquarters master data platform and the regional master data platform simultaneously collecting equipment data from different production lines, the module compares the data collection timestamps collected by the headquarters master data platform and the regional master data platform to assess whether there is a phenomenon of non-real-time synchronization in data collection. Non-real-time synchronization period determination module: When the phenomenon of non-real-time synchronization of data collection occurs, the data collection comparison period is set, the non-real-time synchronized data is filtered out, and the non-real-time synchronized data is analyzed to determine the non-real-time synchronization period. Network Waveform Non-Real-Time Synchronization Analysis Module: Analyzes network fluctuations within each non-real-time synchronization period to determine the non-real-time synchronization period of network waves; Data acquisition and management module: Performs comparative analysis on non-simultaneous network waveforms within multiple data acquisition comparison periods, determines the frequency type of non-simultaneous network waveforms, and adjusts the data acquisition frequency management based on the frequency type of non-simultaneous network waveforms.

Images