Data fusion method for artificial intelligence-based industrial data quality consistency test

Abstract

The application relates to the technical field of data processing, and particularly discloses a data fusion method for industrial data quality consistency testing based on artificial intelligence, which comprises the following steps: an industrial data platform is connected in communication with a target device to collect target industrial data of the target device, and the time sequence quality consistency evaluation result of the target industrial data is tested based on artificial intelligence; the data transmission process of the target device can be dynamically adjusted and the time sequence of the target industrial data can be synchronized and calibrated according to the time sequence quality consistency evaluation result of the target industrial data, so that an intelligent closed-loop system is formed, and the intelligent level is further improved; the application provides a full-aspect intelligent data transmission fusion method for industrial data fusion, ensures the reliability of a data fusion basis, and significantly improves the precision and value of a fusion result.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, specifically to a data fusion method for industrial data quality consistency testing based on artificial intelligence. Background Technology

[0002] In the field of industrial big data processing, the existing technology system has formed a basic framework for the fusion of multi-source heterogeneous data. It performs time-series alignment and feature extraction on multi-source raw data, and then standardizes the data to establish unified rules across data sources, thereby achieving structural compatibility of multi-source heterogeneous data. The pre-fusion static testing technology is used to perform offline integrity verification and range threshold detection during the data input stage to ensure the consistency of data fusion. Based on the feedback correction of the data fusion results, the accuracy and consistency of industrial data fusion are significantly improved.

[0003] For example, the invention patent with announcement number CN117992921B announces an artificial intelligence-based method for multi-source data fusion in shipping locks. This method includes: obtaining ship state coefficients at each time based on the ground speed and draft of ships reporting for port at each time; obtaining channel information for ships arriving at the lock; constructing the channel curvature of each channel area based on the ship state coefficients; obtaining channel speed influence factors for the channel area based on the length and width information of ships reporting for port and the channel curvature; constructing and marking navigation weather influence factors; performing multi-source data fusion on the channel speed influence factors, navigation weather influence factor marking values, and ship numbers to construct the arrival and passage degree of ships reporting for port in the channel area; and obtaining the optimal navigation path for ships based on the arrival and passage degree and lock opening and closing time data combined with ant colony algorithm.

[0004] For example, invention patent CN114970667B provides a method, apparatus, computer equipment, and storage medium for multi-source heterogeneous energy data fusion, belonging to the field of big data processing technology. The method includes: collecting multi-source heterogeneous energy data; assessing the data quality of the collected multi-source heterogeneous energy data; evaluating one or more indicators among the data accuracy, data completeness, data consistency, data correlation, and data uniformity of the multi-source heterogeneous energy data to obtain data quality; and fusing the multi-source heterogeneous energy data based on the results of the data quality assessment.

[0005] However, in implementing the embodiments of this application, it was discovered that the above-mentioned technology has at least the following technical problems: existing industrial data fusion methods and technical systems lack a time-series quality pass-through channel during the operation of the fusion engine. Users cannot know the real-time quality decay status of key time-series parameters such as channel speed influence factors and energy data fluctuation patterns during fusion calculations. More seriously, such time-series deviations will continue to propagate with the fusion process and cannot be interrupted or corrected, ultimately leading to untraceable time-series misalignments and logical distortions in the output results. The essence of this problem is the long-standing black box effect in the field of industrial data fusion: the decay path of time-series quality during the fusion process is invisible, its transmission is uncontrollable, and its consequences are irreversible. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a data fusion method for industrial data quality consistency testing based on artificial intelligence, which can effectively solve the problems mentioned in the background technology.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a data fusion method for industrial data quality consistency testing based on artificial intelligence, comprising: S1. An industrial data platform establishes a communication connection with a target device, the industrial data platform collects target industrial data from the target device, acquires and analyzes the attribute parameters of the target industrial data, and thereby tests the time-series quality consistency assessment result of the target industrial data based on artificial intelligence; S2. Based on the time-series quality consistency assessment result of the target industrial data, it determines whether to adjust the data transmission process of the target device, and simultaneously performs time-series synchronization calibration on the target industrial data based on the time-series quality consistency assessment result, thereby determining whether to perform secondary adjustment on the data transmission process of the target device; S3. The industrial data platform fuses the target industrial data with historical data, thereby acquiring and analyzing the data fusion process parameters of the target industrial data, thereby determining whether to issue a data fusion warning.

[0008] As a further method, it is determined whether to adjust the data transmission process of the target equipment. The specific determination process is as follows: acquire and analyze the data fusion process parameters of the target industrial data; determine whether to adjust the data transmission process of the target equipment based on the time-series quality consistency assessment results of the target industrial data, specifically by using the time-series deviation anomaly index of the target industrial data; compare the time-series deviation anomaly index of the target industrial data with a time-series deviation threshold range; if the time-series deviation anomaly index of the target industrial data falls within the time-series deviation threshold range, it is determined that no adjustment will be made to the data transmission process of the target equipment; if the time-series deviation anomaly index of the target industrial data is less than .... If the deviation threshold range is at its minimum, it is determined that the data transmission process of the target device will be adjusted according to the first adjustment strategy. If the time-series deviation anomaly index of the target industrial data is greater than the maximum value of the time-series deviation threshold range, the deviation value of the time-series deviation anomaly index of the target industrial data is obtained and compared with the preset time-series deviation anomaly index deviation value threshold in the database. If the deviation value of the time-series deviation anomaly index of the target industrial data is greater than or equal to the time-series deviation anomaly index deviation value threshold, it is determined that the data transmission process of the target device will be adjusted according to the second adjustment strategy. If the deviation value of the time-series deviation anomaly index of the target industrial data is less than the time-series deviation anomaly index deviation value threshold, it is determined that the data transmission process of the target device will not be adjusted.

[0009] As a further method, time-series synchronization calibration is performed on the target industrial data based on the time-series quality consistency assessment results. The specific calibration process is as follows: if the time-series deviation anomaly index of the target industrial data falls within the time-series deviation threshold range, the first data calibration scheme is maintained; if the time-series deviation anomaly index of the target industrial data is less than the minimum value of the time-series deviation threshold range, the second data calibration scheme is executed to update the target industrial data; if the time-series deviation anomaly index of the target industrial data is greater than the maximum value of the time-series deviation threshold range, the third data calibration scheme is executed to update the target industrial data; the target industrial data is then transmitted to the target industrial data platform, which integrates the target industrial data with historical data.

[0010] As a further method, the data fusion process of the target industrial data is adjusted. Specifically, the adjustment process is as follows: if the time-series deviation anomaly index of the target industrial data is less than the minimum value of the time-series deviation threshold range, the data fusion process of the target industrial data is adjusted through a first anomaly correction measure; if the time-series deviation anomaly index of the target industrial data is greater than the maximum value of the time-series deviation threshold range, the data fusion process of the target industrial data is adjusted through a second anomaly correction measure. The first anomaly correction measure refers to matching the signal wheel gap increase coefficient and the protocol synchronization period increase coefficient from the attribute database based on the anomaly correction factor of the target industrial data, and adjusting the signal wheel gap and the protocol synchronization period by increasing these coefficients. The second anomaly correction measure refers to matching the signal wheel gap decrease coefficient and the protocol synchronization period decrease coefficient from the attribute database based on the anomaly correction factor of the target industrial data, and adjusting the signal wheel gap and the protocol synchronization period by decreasing these coefficients.

[0011] Compared with the prior art, the embodiments of the present invention have at least the following advantages or beneficial effects:

[0012] (1) This invention provides a data fusion method for industrial data quality consistency testing based on artificial intelligence. By connecting the industrial data platform with the target equipment, the target industrial data of the target equipment is collected, the attribute parameters of the target industrial data are collected and analyzed, and the time series quality consistency assessment result of the target industrial data is tested based on artificial intelligence. Compared with the traditional manual monitoring and testing methods, this monitoring and testing method based on artificial intelligence technology can greatly improve the accuracy, depth, efficiency and foresight of industrial time series data quality consistency assessment, and provide a strong intelligent guarantee for the efficient and stable operation of industrial equipment and industrial data platform. This method can dynamically adjust the data transmission process of the target equipment and perform time series synchronization calibration of the target industrial data according to the time series quality consistency assessment result of the target industrial data, forming an intelligent closed-loop system, further improving the level of intelligence. This invention provides a comprehensive intelligent data transmission fusion method for industrial data fusion, ensuring the basic reliability of data fusion and significantly improving the accuracy and value of fusion results.

[0013] (2) Based on the time series quality consistency assessment results of the target industrial data, the present invention constructs a dynamic time series deviation anomaly index, and the system can adaptively adjust the data transmission process of the target equipment to realize autonomous feedback secondary optimization adjustment of the data transmission process (such as after adjusting the data transmission process, autonomously obtaining the time series deviation anomaly index of the target monitored industrial data and comparing it with the time series deviation threshold range). At the same time, based on the time series consistency assessment results of the target industrial data, the system performs time series synchronization calibration of the target industrial data, realizing closed-loop adjustment that integrates intelligent assessment, dynamic optimization and feedback adjustment, and solving the defects of the prior art in dynamic response lag and lack of adaptiveness.

[0014] (3) This invention acquires and analyzes the data fusion process parameters of the target industrial data in real time, obtains the adjustment anomaly coefficient representing the fusion state, and dynamically adjusts the threshold range in combination with the time-series asynchronous correction coefficient to avoid misjudgment caused by static threshold. It adopts a three-level judgment logic (when the adjustment anomaly coefficient is judged to exceed the threshold range, the second-level time-series threshold correction is triggered, the secondary adjustment operation is performed and the anomaly coefficient is recalculated for verification; if the secondary verification result meets the requirements, the fusion process parameters are initialized to eliminate the accumulated error; if the anomaly persists, the early warning mechanism is immediately triggered and the deep anomaly correction measures matched in the attribute database are called for targeted intervention), to ensure the real-time and accuracy of data fusion quality assessment and significantly improve the long-term operational stability and reliability of the target data platform. Attached Figure Description

[0015] The present invention will be further described with reference to the accompanying drawings, but the embodiments in the drawings do not constitute any limitation on the present invention. For those skilled in the art, other drawings can be obtained based on the following drawings without creative effort.

[0016] Figure 1 This is a schematic diagram of the method flow of the present invention;

[0017] Figure 2 This is a schematic diagram of the attribute parameter analysis and aggregation method of the present invention;

[0018] Figure 3 This is a schematic diagram of the adjustment data transmission determination method of the present invention;

[0019] Figure 4 This is a schematic diagram of the data calibration and transmission method of the present invention;

[0020] Figure 5 This is a schematic diagram of the data transmission secondary adjustment determination method of the present invention;

[0021] Figure 6 This is a schematic diagram of the data fusion early warning determination method of the present invention;

[0022] Figure 7 This is a schematic diagram of the data fusion and early warning method of the present invention. Detailed Implementation

[0023] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0024] Reference Figure 1 As shown, this invention provides a data fusion method for industrial data quality consistency testing based on artificial intelligence, comprising: S1. An industrial data platform establishes a communication connection with a target device, the industrial data platform collects target industrial data from the target device, acquires and analyzes the attribute parameters of the target industrial data, and thereby tests the time-series quality consistency assessment result of the target industrial data based on artificial intelligence; S2. Based on the time-series quality consistency assessment result of the target industrial data, it determines whether to adjust the data transmission process of the target device, and simultaneously performs time-series synchronization calibration on the target industrial data based on the time-series quality consistency assessment result, thereby determining whether to perform secondary adjustment on the data transmission process of the target device; S3. The industrial data platform fuses the target industrial data with historical data, thereby acquiring and analyzing the data fusion process parameters of the target industrial data, thereby determining whether to issue a data fusion warning.

[0025] Figure 2 The flowchart of the attribute parameter analysis and aggregation method of the present invention is as follows: First, attribute parameters of the target industrial data are collected, including clock drift rate, signal phase difference, and load fluctuation factor; then, the collected attribute parameters are analyzed in depth to calculate the weight of clock drift rate, signal phase difference, and load fluctuation factor respectively; finally, the weight of each factor is aggregated and integrated to provide the core parameter basis for subsequent weight-based calculation of industrial data time series deviation anomaly index and system adjustment.

[0026] Specifically, the attribute parameters of the target industrial data are collected and analyzed. The specific analysis process is as follows: The attribute parameters of the target industrial data include the clock drift rate, the signal phase difference, and the load fluctuation factor. The clock drift rate refers to the clock frequency deviation between the target industrial data and the target industrial data platform. The signal phase difference refers to the timing offset of the target industrial data in adjacent sampling periods. The load fluctuation factor refers to the load fluctuation change value of the target industrial data in the sampling period. The clock drift rate can be monitored by a precision clock synchronization device, the signal phase difference can be obtained by directly reading the timestamp from the PLC controller, and the load fluctuation factor can be monitored by an edge computing gateway.

[0027] The aforementioned target industrial data refers to interactive control data, process flow data, and output result data collected by the target equipment, including communication protocol data, process start and stop timestamps, process parameter sequences, quality inspection data, and output and energy consumption data.

[0028] The aforementioned clock frequency deviation refers to the absolute value of the difference between the clock frequency of the target device and the clock frequency of the target industrial data platform; the aforementioned load fluctuation change value refers to the standard deviation of the fluctuation change in the data utilization rate of the target industrial data.

[0029] It should be explained that the load fluctuation factor is obtained by monitoring the dynamic load rate of the target industrial data in real time through the edge computing gateway and responding immediately to the fluctuation of the load rate. The aforementioned dynamic load rate refers to the data utilization rate of the target industrial data at a certain moment.

[0030] It needs to be explained that the influence of the time series deviation anomaly index on the proportional relationship between the clock drift rate of the target industrial data and the defined clock drift rate, the proportional relationship between the signal phase difference of the target industrial data and the defined signal phase difference, and the proportional relationship between the load fluctuation factor of the target industrial data and the defined load fluctuation factor is quantified by introducing weighting coefficients. These influences are then aggregated to derive the time series deviation anomaly index of the target industrial data. The specific expression for the time series deviation anomaly index of the target industrial data is as follows:

[0031] ;

[0032] In the formula, TAI is the timing deviation anomaly index of the target industrial data, CDR is the clock drift rate of the target industrial data, SPD is the signal phase difference of the target industrial data, LFF is the load fluctuation factor of the target industrial data, J_CDR is the preset defined clock drift rate in the attribute database, J_SPD is the preset defined signal phase difference in the attribute database, J_LFF is the preset defined load fluctuation factor in the attribute database, kx1 is the weight coefficient corresponding to the preset clock drift rate factor in the attribute database, kx2 is the weight coefficient corresponding to the preset signal phase difference factor in the attribute database, and kx3 is the weight coefficient of the preset load fluctuation factor in the attribute database.

[0033] It should be explained that the attribute database is used to store parameters of the data fusion method for AI-based industrial data quality consistency testing.

[0034] It should be explained that the Time Series Deviation Anomaly Index of the Target Industrial Data is a quantitative indicator that measures the degree to which the time series quality of the target industrial data deviates from the expected value, thus constituting an anomaly. It is used to characterize the degree of time series deviation anomaly of the target industrial data.

[0035] The above definition of clock drift rate represents the maximum critical reference value of clock drift rate; the above definition of signal phase difference represents the maximum critical reference value of signal phase difference; the above definition of load ripple factor represents the maximum critical reference value of load ripple factor.

[0036] The weighting coefficients corresponding to the aforementioned clock drift rate factor represent the weighting strength of the ratio between the clock drift rate of the target industrial data and the defined clock drift rate on the timing deviation anomaly index. These coefficients are preset in the attribute database and range from (0, 1). The weighting coefficients corresponding to the aforementioned signal phase difference factor represent the weighting strength of the ratio between the signal phase difference of the target industrial data and the defined signal phase difference on the timing deviation anomaly index. These coefficients are preset in the attribute database and range from (0, 1). The weighting coefficients corresponding to the aforementioned load fluctuation factor represent the weighting strength of the ratio between the load fluctuation factor of the target industrial data and the defined load fluctuation factor on the timing deviation anomaly index. These coefficients are preset in the attribute database and range from (0, 1).

[0037] It needs to be explained that an increase in clock drift rate means a widening of the timing misalignment between the target device and the target data platform, which in turn leads to timing inaccuracies in signal transmission. The signal phase difference increases accordingly. In other words, an increase in clock drift rate leads to an increase in the signal phase difference, which in turn increases the proportional relationship between the clock drift rate of the target industrial data and the defined clock drift rate, as well as the proportional relationship between the signal phase difference of the target industrial data and the defined signal phase difference. This increases the timing deviation anomaly index. On the other hand, an increase in clock drift rate and signal phase difference also increases the dynamic complexity rate, causing the load fluctuation factor to deviate from the preset load fluctuation factor. The proportional relationship between the load fluctuation factor and the defined load fluctuation factor increases, further increasing the timing deviation anomaly index. In summary, clock drift rate indirectly affects the timing deviation anomaly index by influencing the signal phase difference, which in turn affects the load fluctuation factor.

[0038] Figure 3 The flowchart of the data transmission adjustment and determination method of the present invention is as follows: First, the time-series deviation anomaly index of the target industrial data is calculated; then, the calculated time-series deviation anomaly index is compared with a preset threshold range for determination: if the index is within the time-series deviation threshold range, it is determined that the data transmission process will not be adjusted; if the index is less than the minimum value of the range, the data transmission process is immediately optimized through the first adjustment strategy; if the index is greater than the maximum value of the range, the index deviation value is further calculated and compared with the preset threshold. When the deviation value is greater than or equal to the preset threshold, the second adjustment strategy is activated for deep adjustment; otherwise, the data transmission process remains unchanged. Through this hierarchical and progressive determination and adjustment mechanism, dynamic optimization and precise control of the industrial data transmission process are achieved.

[0039] Specifically, the determination of whether to adjust the data transmission process of the target equipment is made by using the time series quality consistency assessment results of the target industrial data to determine whether to adjust the data transmission process of the target equipment, which means using the time series deviation anomaly index of the target industrial data to determine whether to adjust the data transmission process of the target equipment.

[0040] In one specific embodiment, the present invention, based on the time-series quality consistency assessment results of target industrial data, constructs a dynamic time-series deviation anomaly index. The system can adaptively adjust the data transmission process of the target equipment, realize autonomous feedback secondary optimization adjustment of the data transmission process, and simultaneously perform time-series synchronous calibration of the target industrial data based on the time-series consistency assessment results of the target industrial data. This achieves a closed-loop adjustment that integrates intelligent assessment, dynamic optimization, and feedback adjustment, solving the defects of existing technologies such as lag in dynamic response and lack of adaptability.

[0041] The time series deviation anomaly index of the target industrial data is compared with the time series deviation threshold range. If the time series deviation anomaly index of the target industrial data falls within the time series deviation threshold range, it is determined that the data transmission process of the target equipment will not be adjusted.

[0042] If the time series deviation anomaly index of the target industrial data is less than the minimum value of the time series deviation threshold range, it is determined that the data transmission process of the target equipment will be adjusted through the first adjustment strategy.

[0043] If the time-series deviation anomaly index of the target industrial data is greater than the maximum value of the time-series deviation threshold range, the deviation value of the time-series deviation anomaly index of the target industrial data is obtained and compared with the preset time-series deviation anomaly index deviation value threshold in the database. If the deviation value of the time-series deviation anomaly index of the target industrial data is greater than or equal to the time-series deviation anomaly index deviation value threshold, it is determined that the data transmission process of the target equipment shall be adjusted through the second adjustment strategy.

[0044] If the deviation value of the time series deviation anomaly index of the target industrial data is less than the threshold value of the time series deviation anomaly index, it is determined that no adjustment will be made to the data transmission process of the target equipment.

[0045] It should be explained that the time series deviation threshold range refers to the time series deviation anomaly index threshold range, which is the set between the maximum value and the minimum value of the time series deviation anomaly index that are preset in the attribute database.

[0046] It should be explained that the deviation value of the time series deviation anomaly index is the difference between the time series deviation anomaly index and the maximum value of the time series deviation threshold range.

[0047] It should be explained that the time-series quality consistency assessment result of the target industrial data is derived from the attribute parameters of the target industrial data collected and analyzed by artificial intelligence testing, and refers to the quantitative assessment of the time-series quality consistency of the target industrial data.

[0048] Specifically, the determination is to adjust the data transmission process of the target device through the first adjustment strategy, which refers to matching the clock calibration frequency reduction coefficient, the database buffer capacity reduction coefficient, and the data acquisition frequency reduction coefficient from the attribute database based on the first deviation degree of the time deviation anomaly of the target industrial data.

[0049] By reducing the clock calibration frequency, database buffer capacity, and data acquisition frequency during data transmission of the target device, the clock calibration frequency, database buffer capacity, and data acquisition frequency are adjusted.

[0050] The first deviation degree of the time series deviation anomaly of the target industrial data mentioned above refers to the minimum value of the time series deviation threshold interval minus the time series deviation anomaly index of the target industrial data, divided by the minimum value of the time series deviation threshold interval.

[0051] It needs to be explained that the clock calibration frequency reduction coefficient, database buffer capacity reduction coefficient, and data acquisition frequency reduction coefficient are matched from the attribute database. When the keywords "clock calibration frequency," "database buffer capacity," and "data acquisition frequency" are input, the database maps the keywords to preset fields. Based on the coefficient configuration table of the attribute database, the database performs precise field matching and returns the preset coefficient values ​​to complete the matching.

[0052] The aforementioned clock frequency reduction factor refers to the reduction ratio corresponding to the preset clock frequency in the attribute database; the aforementioned database buffer capacity reduction factor refers to the reduction ratio corresponding to the preset database buffer capacity in the attribute database; the aforementioned data acquisition frequency reduction factor refers to the reduction ratio corresponding to the preset data acquisition frequency in the attribute database.

[0053] It should be explained that the reduction adjustment is achieved by multiplying the clock calibration frequency reduction factor by the clock calibration frequency during the data transmission process of the target device; the database buffer capacity reduction factor is multiplied by the database buffer capacity during the data transmission process of the target device; and the data acquisition frequency reduction factor is multiplied by the data acquisition frequency during the data transmission process of the target device.

[0054] Specifically, the determination is to adjust the data transmission process of the target device through the second adjustment strategy, which refers to matching the clock calibration frequency increase coefficient, the database buffer capacity increase coefficient, and the data acquisition frequency increase coefficient from the attribute database based on the second deviation degree of the time deviation anomaly of the target industrial data.

[0055] By increasing the clock calibration frequency, database buffer capacity, and data acquisition frequency, the clock calibration frequency, database buffer capacity, and data acquisition frequency of the target device during data transmission are adjusted.

[0056] The second deviation degree of the time series deviation anomaly of the target industrial data mentioned above refers to the time series deviation anomaly index of the target industrial data minus the maximum value of the time series deviation threshold range, and then divided by the maximum value of the time series deviation threshold range.

[0057] It needs to be explained that the clock calibration frequency increase factor, database buffer capacity increase factor, and data acquisition frequency increase factor are matched from the attribute database. When the keywords "clock calibration frequency," "database buffer capacity," and "data acquisition frequency" are input, the database maps the keywords to preset fields. Based on the coefficient configuration table of the attribute database, the database performs precise field matching and returns the preset coefficient values ​​to complete the matching.

[0058] The system configuration table of the aforementioned attribute database refers to the core data table used by the attribute database to define and manage database parameters, modules, and association rules.

[0059] The aforementioned clock frequency increase factor refers to the increase ratio corresponding to the preset clock frequency in the attribute database; the aforementioned database buffer capacity increase factor refers to the increase ratio corresponding to the preset database buffer capacity in the attribute database; the aforementioned data acquisition frequency increase factor refers to the increase ratio corresponding to the preset data acquisition frequency in the attribute database.

[0060] It should be explained that the increase adjustment is achieved by multiplying the clock calibration frequency increase factor by the clock calibration frequency during the data transmission process of the target device; the database buffer capacity increase factor is multiplied by the database buffer capacity during the data transmission process of the target device; and the data acquisition frequency increase factor is multiplied by the data acquisition frequency during the data transmission process of the target device.

[0061] Figure 4 The flowchart of the data calibration and transmission method of the present invention is as follows: First, the calculated time-series deviation anomaly index is compared with a preset threshold range. If the index is within the time-series deviation threshold range, the first data calibration scheme is maintained; if it is not within the threshold range, it is further determined whether the index is less than the minimum value of the threshold range—if it is less than the minimum value, the second data calibration scheme is executed; if it is greater than the maximum value, the third data calibration scheme is executed. Regardless of the calibration scheme used, the target industrial data is ultimately updated. The updated target industrial data will be transmitted to the industrial data platform to provide a basis for subsequent data fusion and analysis, ensuring that the time-series quality of the data meets system requirements.

[0062] Furthermore, based on the time series quality consistency assessment results of the target industrial data, time series synchronization calibration is performed on the target industrial data. The specific calibration process is as follows: if the time series deviation anomaly index of the target industrial data belongs to the time series deviation threshold range, then the first data calibration scheme is maintained.

[0063] If the time series deviation anomaly index of the target industrial data is less than the minimum value of the time series deviation threshold range, then the second data calibration scheme is executed to update the target industrial data.

[0064] If the time series deviation anomaly index of the target industrial data is greater than the maximum value of the time series deviation threshold range, then the third data calibration scheme is executed to update the target industrial data.

[0065] The target industrial data is transmitted to the target industrial data platform, which then integrates the target industrial data with historical data.

[0066] It should be explained that the comparison results of the time series deviation anomaly index and threshold range of different target industrial data correspond to corresponding data calibration schemes. Specifically, the first data calibration scheme obtains the baseline PIP synchronization protocol periodic execution interval, baseline anomaly judgment boundary range, and baseline signal fusion weight allocation from the attribute database to calibrate the target industrial data. The second data calibration scheme matches the PIP synchronization protocol periodic execution interval increase coefficient, anomaly judgment boundary range increase coefficient, and main signal fusion weight decrease coefficient from the attribute database. The PIP synchronization protocol periodic interval and anomaly judgment boundary range of the target industrial data are increased by the PIP synchronization protocol periodic execution interval increase coefficient and anomaly judgment boundary range increase coefficient, and the main signal fusion weight is decreased by the main signal fusion weight decrease coefficient. The third data calibration scheme matches the PIP synchronization protocol periodic execution interval decrease coefficient, anomaly judgment boundary range decrease coefficient, and main signal fusion weight increase coefficient from the attribute database. The PIP synchronization protocol periodic interval and anomaly judgment boundary range of the target industrial data are decreased by the PIP synchronization protocol periodic execution interval decrease coefficient and anomaly judgment boundary range decrease coefficient, and the main signal fusion weight is increased by the main signal fusion weight increase coefficient.

[0067] It needs to be explained that: PIP synchronization protocol period interval is the interval at which the target industrial data is clocked and calibrated using the PIP protocol; anomaly detection boundary range is the range of the target industrial data for anomaly detection, and expanding the boundary range can reduce the proportion of edge data and lower the probability of false judgment; signal fusion weight is the weight of each signal data in the target industrial data during the data fusion process.

[0068] Figure 5 The flowchart of the data transmission secondary adjustment judgment method of the present invention is as follows: First, the updated target industrial data is transmitted to the industrial data platform, and the platform performs data fusion operation. After the data fusion is completed, the data transmission adjustment effect is judged a second time, that is, whether the adjusted time deviation anomaly index is within a preset threshold range: if the adjusted index is still not within the time deviation threshold range, the secondary adjustment process is initiated; if it is within the threshold range, it is determined that no secondary adjustment is needed. On this basis, the adjustment anomaly coefficient is further compared with the preset threshold: if the adjustment anomaly coefficient is greater than or equal to the threshold, the subsequent data fusion early warning and correction mechanism will be triggered; otherwise, the data status will continue to be monitored to ensure the time stability and accuracy of industrial data during transmission and fusion.

[0069] Furthermore, it is determined whether to perform secondary adjustment on the data transmission process of the target device. The specific determination process is as follows: after timing synchronization calibration, the timing deviation anomaly index of the target monitored industrial data is obtained and compared with the timing deviation threshold range. If the timing deviation anomaly index of the target monitored industrial data still does not belong to the timing deviation threshold range, it is determined that the data transmission process of the target device should be adjusted secondary.

[0070] If the time-series deviation anomaly index of the target monitored industrial data falls within the time-series deviation threshold range, it is determined that no secondary adjustment will be made to the data transmission process of the target device.

[0071] The aforementioned target industrial data monitoring refers to the target industrial data of the target equipment monitored and acquired by the target industrial data platform.

[0072] Specifically, if the time series deviation anomaly index of the target industrial data is less than the minimum value of the time series deviation threshold range, the data transmission process of the target equipment will be adjusted a second time through the first adjustment strategy.

[0073] If the time-series deviation anomaly index of the target industrial data is greater than the maximum value of the time-series deviation threshold range, the deviation value of the time-series deviation anomaly index of the target industrial data is obtained and compared with the preset time-series deviation anomaly index deviation value threshold in the database. If the deviation value of the time-series deviation anomaly index of the target industrial data is greater than or equal to the time-series deviation anomaly index deviation value threshold, it is determined that the data transmission process of the target equipment will be adjusted a second time through the second adjustment strategy.

[0074] If the deviation value of the time series deviation anomaly index of the target industrial data is less than the threshold value of the time series deviation anomaly index, then no secondary adjustment will be made to the data transmission process of the target equipment.

[0075] After the second adjustment, the time series deviation anomaly index of the target monitored industrial data is obtained and compared with the time series deviation threshold range. If the time series deviation anomaly index of the target monitored industrial data still does not fall within the time series deviation threshold range, a time series warning is issued for the data transmission process of the target equipment.

[0076] The aforementioned target monitoring industrial data refers to the industrial data obtained by monitoring the target equipment again after secondary adjustment.

[0077] It should be explained that the first adjustment strategy refers to matching the clock calibration frequency reduction coefficient, database buffer capacity reduction coefficient, and data acquisition frequency reduction coefficient from the attribute database based on the first deviation degree of the time series deviation of the target monitored industrial data. Then, the clock calibration frequency reduction coefficient, database buffer capacity reduction coefficient, and data acquisition frequency reduction coefficient are used to perform a second reduction adjustment on the clock calibration frequency, database buffer capacity, and data acquisition frequency of the target monitored industrial data.

[0078] It should be explained that the second adjustment strategy refers to matching the clock calibration frequency reduction coefficient, database buffer capacity reduction coefficient, and data acquisition frequency reduction coefficient from the attribute database based on the second deviation degree of the time series deviation of the target monitoring industrial data. The clock calibration frequency reduction coefficient, database buffer capacity reduction coefficient, and data acquisition frequency reduction coefficient are then used to further increase and adjust the clock calibration frequency, database buffer capacity, and data acquisition frequency of the target monitoring industrial data.

[0079] The first deviation degree of the time series deviation anomaly of the aforementioned target monitored industrial data refers to the difference between the minimum value of the time series deviation threshold range and the time series deviation anomaly index of the target monitored industrial data.

[0080] The second deviation degree of the time series deviation anomaly of the aforementioned target monitoring industrial data refers to the difference between the time series deviation anomaly index of the target monitoring industrial data and the maximum value of the time series deviation threshold range.

[0081] It should be explained that the secondary reduction adjustment involves multiplying the clock calibration frequency reduction factor by the reduced clock calibration frequency; multiplying the database buffer capacity reduction factor by the reduced database buffer capacity; and multiplying the data acquisition frequency reduction factor by the reduced data acquisition frequency.

[0082] It should be explained that the secondary increase adjustment is performed by multiplying the clock calibration frequency increase factor by the increased clock calibration frequency; the database buffer capacity increase factor by multiplying the database buffer capacity by the increased database buffer capacity; and the data acquisition frequency increase factor by multiplying the data acquisition frequency by the increased data acquisition frequency.

[0083] The aforementioned time-series warning refers to a visual pop-up window displayed by the target industrial data platform indicating that there is an anomaly in the current time series.

[0084] The data fusion process parameters of the target industrial data and the adjustment anomaly coefficient of the target industrial data are acquired and analyzed. The specific analysis process is as follows: The data fusion process parameters of the target industrial data include the data fusion channel bit error rate, the data fusion decision fluctuation frequency, and the time series deviation anomaly index of the target industrial data. The aforementioned data fusion channel bit error rate refers to the proportion of data transmission and fusion errors caused by signal distortion in the data fusion channel during the data fusion process. The aforementioned data fusion decision fluctuation frequency refers to the frequency of abnormal fluctuations in the target industrial data due to improper weight allocation during the data fusion process.

[0085] In one specific embodiment, the present invention acquires and analyzes the data fusion process parameters of the target industrial data in real time, obtains the adjustment anomaly coefficient characterizing the fusion state, and dynamically adjusts the threshold range in combination with the time-series asynchronous correction coefficient to avoid misjudgment caused by static thresholds. A three-level judgment logic is adopted (when the adjustment anomaly coefficient is determined to exceed the threshold range, the second-level time-series threshold correction is triggered, a secondary adjustment operation is performed, and the anomaly coefficient is recalculated for verification; if the secondary verification result meets the requirements, the fusion process parameters are initialized to eliminate accumulated errors; if the anomaly persists, an early warning mechanism is immediately triggered, and the deep anomaly correction measures matched in the attribute database are called for targeted intervention).

[0086] The median value of the time series deviation threshold range is marked as the reference value of the time series deviation anomaly index.

[0087] By introducing weighting coefficients, the influence of the following on the adjustment anomaly coefficient of the target industrial data is quantified: the ratio of the target industrial data's data fusion channel bit error rate to the defined data fusion channel bit error rate; the ratio of the target industrial data's data fusion decision fluctuation frequency to the defined data fusion decision fluctuation frequency; and the deviation relationship between the absolute value of the difference between the target industrial data's time series deviation anomaly index and its reference value and the reference value. These influences are then coupled to derive the adjustment anomaly coefficient of the target industrial data. Specifically:

[0088] ;

[0089] In the formula, RAI is the adjustment anomaly coefficient of the target industrial data, BER is the bit error rate of the data fusion channel of the target industrial data, DFR is the data fusion decision fluctuation frequency of the target industrial data, TAI is the time series anomaly deviation index of the target industrial data, J_BER is the pre-defined data fusion channel bit error rate in the attribute database, J_DFR is the pre-defined data fusion decision fluctuation frequency in the attribute database, C_TAI is the reference value of the time series deviation anomaly index, kw1 is the weight coefficient corresponding to the pre-defined data fusion channel bit error rate factor in the attribute database, kw2 is the weight coefficient corresponding to the pre-defined data fusion decision fluctuation frequency factor in the attribute database, and kw3 is the weight coefficient corresponding to the pre-defined time series deviation anomaly index factor in the attribute database.

[0090] The aforementioned adjustment anomaly coefficient of the target industrial data is a quantitative indicator composed of the degree of anomaly in the transmission and fusion of the target industrial data, used to quantify the degree of adjustment anomaly in the target industrial data.

[0091] The above definition of the data fusion channel bit error rate represents the maximum allowable value of the data fusion channel bit error rate; the above definition of the data fusion decision fluctuation frequency represents the maximum allowable value of the data fusion decision fluctuation frequency.

[0092] The above reference value for the time series deviation anomaly index refers to the median value of the time series deviation anomaly index.

[0093] The weighting coefficient of the aforementioned data fusion channel bit error rate factor represents the degree of influence of the ratio between the target industrial data's data fusion channel bit error rate and the defined data fusion channel bit error rate on the adjustment anomaly coefficient. It is preset in the attribute database and has a value range of (0, 1]. The weighting coefficient of the aforementioned data fusion decision fluctuation frequency factor represents the degree of influence of the ratio between the target industrial data's data fusion decision fluctuation frequency and the defined data fusion decision fluctuation frequency on the adjustment anomaly coefficient. It is preset in the attribute database and has a value range of (0, 1). The influence coefficient of the aforementioned time series deviation anomaly index factor represents the degree of influence of the ratio between the absolute value of the difference between the target industrial data's time series deviation anomaly index and the time series deviation anomaly index reference value on the adjustment anomaly coefficient. It is preset in the attribute database and has a value range of (0, 1).

[0094] Figure 6 The flowchart of the data fusion early warning determination method of the present invention is as follows: First, the adjustment anomaly coefficient is compared with a preset adjustment anomaly coefficient threshold. If the adjustment anomaly coefficient is less than the threshold, it is determined that no data fusion early warning will be issued. If the adjustment anomaly coefficient is greater than or equal to the threshold, the correction and adjustment process is initiated, that is, the timing deviation threshold range is corrected, and the data fusion process is optimized and adjusted based on the corrected threshold. After the adjustment is completed, a secondary adjustment anomaly coefficient is obtained and compared with the adjustment anomaly coefficient threshold again. If the secondary adjustment anomaly coefficient is still greater than or equal to the threshold, it is determined that a data fusion early warning will be issued. If the secondary adjustment anomaly coefficient is less than the threshold, the data fusion process is initialized. Through this cyclic verification mechanism, the stability and reliability of industrial data during the fusion process are ensured, and potential risks are detected and dealt with in a timely manner.

[0095] Furthermore, the determination of whether to issue a data fusion warning is made by extracting a preset adjustment anomaly coefficient threshold from the attribute database and comparing it with the adjustment anomaly coefficient of the target industrial data. If the adjustment anomaly coefficient of the target industrial data is less than the adjustment anomaly coefficient threshold, it is determined that no data fusion warning will be issued.

[0096] The aforementioned threshold for the adjustment anomaly coefficient refers to the maximum value of the adjustment anomaly coefficient, which is the maximum value allowed by the preset definition of the adjustment anomaly coefficient in the attribute database.

[0097] If the adjustment anomaly coefficient of the target industrial data is greater than or equal to the adjustment anomaly coefficient threshold, then the time-series asynchronous increase correction coefficient is matched from the attribute database to increase the maximum value of the time-series deviation threshold range. The time-series asynchronous decrease correction coefficient is matched from the database to decrease the minimum value of the time-series deviation threshold range. The data fusion process of the target industrial data is also adjusted. After the adjustment is completed, the secondary adjustment anomaly coefficient of the target industrial data is obtained and compared with the adjustment anomaly coefficient threshold. If the secondary adjustment anomaly coefficient of the target industrial data is still greater than or equal to the adjustment anomaly coefficient threshold, then a data fusion warning is issued.

[0098] The aforementioned asynchronous time-series increase correction coefficient refers to the proportion used to increase the maximum value of the time-series deviation threshold range of the target industrial data; the aforementioned asynchronous time-series decrease correction coefficient refers to the proportion used to decrease the minimum value of the time-series deviation threshold range of the target industrial data.

[0099] It should be explained that after increasing the maximum value of the time series deviation threshold range and implementing the second adjustment strategy to adjust the target industrial data a second time, the time series deviation anomaly index is obtained again. If it still does not belong to the threshold range, then the maximum value of the corrected time series deviation threshold range is the obtained time series deviation anomaly index.

[0100] It should be explained that after reducing the minimum value of the time series deviation threshold range and implementing the first adjustment strategy to adjust the target industrial data a second time, the time series deviation anomaly index is obtained again. If it still does not belong to the threshold range, then the minimum value of the corrected time series deviation threshold range is the obtained time series deviation anomaly index.

[0101] The aforementioned secondary adjustment anomaly coefficient refers to a quantitative indicator composed of the degree of anomaly in the data transmission and fusion of the target industrial data after adjustment, used to quantify the degree of adjustment anomaly in the target industrial data after adjustment.

[0102] It needs to be explained that the asynchronous increase correction coefficient is matched from the attribute database. When the keyword "asynchronous increase correction" is input, the database maps the keyword to a preset field. Based on the coefficient configuration table of the attribute database, the field is matched precisely, and the preset coefficient value is returned to complete the matching.

[0103] It needs to be explained that the asynchronous reduction correction coefficient is matched from the attribute database. When the keyword "asynchronous reduction correction" is input, the database maps the keyword to a preset field. Based on the coefficient configuration table of the attribute database, the field is matched precisely, and the preset coefficient value is returned to complete the matching.

[0104] If the secondary adjustment anomaly coefficient of the target industrial data is less than the adjustment anomaly coefficient threshold, then the data fusion process is initialized.

[0105] It needs to be explained that during the initialization of the data fusion process, when the secondary adjustment anomaly coefficient is less than the adjustment anomaly coefficient threshold, the data fusion process of the target industrial data is reset, temporary variables generated by historical adjustment behaviors (including time-series asynchronous increase and time-series asynchronous decrease of the correction coefficient, etc.) are cleared, and the original operating state without correction intervention is restored.

[0106] The system then re-evaluates whether the time-series deviation anomaly index of the target industrial data falls within the corrected time-series deviation threshold range. If the time-series deviation anomaly index falls within the corrected time-series deviation threshold range, no data fusion warning is issued. If the time-series deviation anomaly index of the target industrial data is less than the minimum value of the corrected time-series deviation threshold range, a data fusion warning is issued, and the data fusion process of the target industrial data platform is adjusted based on the first anomaly correction measure. If the time-series deviation anomaly index of the target industrial data is greater than the maximum value of the corrected time-series deviation threshold range, a data fusion warning is issued, and the data fusion process of the target industrial data platform is adjusted based on the second anomaly correction measure.

[0107] The aforementioned corrected timing deviation threshold range refers to the timing deviation threshold range after dynamic correction based on secondary adjustment.

[0108] The aforementioned data fusion warning indicates that the target industrial data platform will display a visual pop-up window showing that there is an anomaly in the current data fusion.

[0109] Figure 7 The diagram illustrates the data fusion and early warning method of this invention. First, the time-series deviation anomaly index of the target industrial data is compared with a corrected threshold range. If the time-series deviation anomaly index is within the corrected time-series deviation threshold range, no data fusion early warning is issued. If it exceeds this range, it is further determined whether the time-series deviation anomaly index is less than the minimum value of the corrected time-series deviation threshold range. If it is less than the minimum value, a data fusion early warning is immediately triggered, and a first anomaly correction measure is executed. If it is greater than the maximum value, an early warning is triggered, and a second anomaly correction measure is executed. Regardless of which anomaly correction measure is executed, the ultimate goal is to optimize the time-series quality of the industrial data by adjusting the data fusion process, ensuring the stability and reliability of the data fusion process, and reducing anomaly risks.

[0110] Furthermore, the data fusion process of the target industrial data is adjusted. Specifically, if the time-series deviation anomaly index of the target industrial data is less than the minimum value of the time-series deviation threshold range, the data fusion process of the target industrial data is adjusted through a first anomaly correction measure. If the time-series deviation anomaly index of the target industrial data is greater than the maximum value of the time-series deviation threshold range, the data fusion process of the target industrial data is adjusted through a second anomaly correction measure. The first anomaly correction measure refers to matching the signal wheel gap increase coefficient and the protocol synchronization period increase coefficient from the attribute database based on the anomaly correction factor of the target industrial data, and adjusting the signal wheel gap and the protocol synchronization period by increasing these coefficients. The second anomaly correction measure refers to matching the signal wheel gap decrease coefficient and the protocol synchronization period decrease coefficient from the attribute database based on the anomaly correction factor of the target industrial data, and adjusting the signal wheel gap and the protocol synchronization period by decreasing these coefficients.

[0111] The aforementioned anomaly correction factor is a dynamic quantitative indicator used to characterize the degree of time-series deviation anomalies in target industrial data and the need to adjust for anomalies.

[0112] It needs to be explained that the signal wheel gap increase coefficient and the protocol synchronization period increase coefficient are matched from the attribute database. An anomaly correction factor is input, and the database maps the anomaly correction factor to a preset field. Based on the coefficient configuration table of the attribute database, the field is precisely matched, and the preset coefficient value is returned to complete the increase coefficient matching.

[0113] It needs to be explained that the signal wheel gap reduction coefficient and the protocol synchronization period reduction coefficient are matched from the attribute database. An anomaly correction factor is input, and the database maps the anomaly correction factor to a preset field. Based on the coefficient configuration table of the attribute database, the field is precisely matched, and the preset coefficient value is returned to complete the reduction coefficient matching.

[0114] The aforementioned signal wheel gap increase factor refers to the percentage value preset in the attribute database to increase the signal wheel gap; the aforementioned protocol synchronization period increase factor refers to the percentage value preset in the attribute database to increase the protocol synchronization period.

[0115] The aforementioned signal wheel gap reduction coefficient refers to the percentage value preset in the attribute database to reduce the signal wheel gap; the aforementioned protocol synchronization period reduction coefficient refers to the percentage value preset in the attribute database to reduce the protocol synchronization period.

[0116] It should be explained that the signal wheel gap is the gap between the position sensor and the signal wheel. Adjusting the signal wheel gap can adjust the data fusion signal transmission rate and fusion accuracy.

[0117] It should be explained that increasing the signal wheel gap means increasing the gap between the position sensor and the signal wheel, reducing the sensor's sensitivity to capture signals, thereby reducing misjudgments or over-response caused by high-frequency micro-fluctuations. At the same time, increasing the matching protocol period coefficient extends the protocol synchronization period, further reducing misjudgments or over-response. Conversely, decreasing the signal wheel gap means decreasing the gap between the position sensor and the signal wheel, increasing the sensor's sensitivity to capture signals, thereby more accurately capturing weak signal fluctuations, improving the real-time performance and accuracy of data transmission fusion, ensuring signal integrity. At the same time, decreasing the matching protocol period coefficient shortens the protocol synchronization period, reducing accumulated errors, and promptly correcting signal delays or misjudgments, further improving the real-time performance and accuracy of data transmission fusion.

[0118] The above description is merely an example and illustration of the structure of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the structure of the invention or exceed the scope defined by the present invention, and all such modifications and additions should fall within the protection scope of the present invention.

Claims

1. A data fusion method for industrial data quality consistency testing based on artificial intelligence, characterized in that, include: S1. The industrial data platform establishes a communication connection with the target equipment. The industrial data platform collects the target industrial data of the target equipment, collects and analyzes the attribute parameters of the target industrial data, and then uses artificial intelligence to test the time-series quality consistency evaluation results of the target industrial data. S2. Determine whether to adjust the data transmission process of the target equipment based on the time series quality consistency assessment results of the target industrial data. At the same time, perform time series synchronization calibration on the target industrial data based on the time series quality consistency assessment results of the target industrial data. After time series synchronization calibration, obtain the time series deviation anomaly index of the target monitored industrial data. Determine whether to perform secondary adjustment on the data transmission process of the target equipment based on the time series deviation anomaly index. S3. The industrial data platform fuses target industrial data with historical data to obtain and analyze the data fusion process parameters of the target industrial data, thereby determining whether to issue a data fusion warning; The specific analysis process for collecting and analyzing the attribute parameters of the target industrial data is as follows: The attribute parameters of the target industrial data include the clock drift rate of the target industrial data, the signal phase difference of the target industrial data, and the load fluctuation factor of the target industrial data. By introducing weighting coefficients to quantify the proportional relationship between the clock drift rate of the target industrial data and the defined clock drift rate, the proportional relationship between the signal phase difference of the target industrial data and the defined signal phase difference, and the proportional relationship between the load fluctuation factor of the target industrial data and the defined load fluctuation factor, the influence of each influence degree on the timing deviation anomaly index is obtained by pooling the influence degrees. The time-series deviation anomaly index of the target industrial data is used to characterize the degree of time-series deviation anomaly of the target industrial data; The determination of whether to adjust the data transmission process of the target device is specifically as follows: Determining whether to adjust the data transmission process of the target equipment based on the time series quality consistency assessment results of the target industrial data refers to determining whether to adjust the data transmission process of the target equipment based on the time series deviation anomaly index of the target industrial data. The time series deviation anomaly index of the target industrial data is compared with the time series deviation threshold range. If the time series deviation anomaly index of the target industrial data falls within the time series deviation threshold range, it is determined that the data transmission process of the target equipment will not be adjusted. If the time series deviation anomaly index of the target industrial data is less than the minimum value of the time series deviation threshold range, it is determined that the data transmission process of the target equipment shall be adjusted through the first adjustment strategy. If the time series deviation anomaly index of the target industrial data is greater than the maximum value of the time series deviation threshold range, the deviation value of the time series deviation anomaly index of the target industrial data is obtained and compared with the preset time series deviation anomaly index deviation value threshold in the database. If the deviation value of the time series deviation anomaly index of the target industrial data is greater than or equal to the time series deviation anomaly index deviation value threshold, it is determined that the data transmission process of the target equipment is adjusted through the second adjustment strategy. If the deviation value of the time series deviation anomaly index of the target industrial data is less than the threshold value of the time series deviation anomaly index, it is determined that no adjustment will be made to the data transmission process of the target equipment.

2. The data fusion method for industrial data quality consistency testing based on artificial intelligence according to claim 1, characterized in that: The adjustment of the data transmission process of the target device through the first adjustment strategy is as follows: Based on the time series deviation anomaly index of the target industrial data and the minimum value of the time series deviation threshold range, the first deviation degree of the time series deviation anomaly of the target industrial data is obtained; The first adjustment strategy refers to matching the clock calibration frequency reduction coefficient, the database buffer capacity reduction coefficient, and the data acquisition frequency reduction coefficient from the attribute database based on the first deviation degree of the time-series deviation anomaly of the target industrial data. By reducing the clock calibration frequency, database buffer capacity, and data acquisition frequency during data transmission of the target device, the clock calibration frequency, database buffer capacity, and data acquisition frequency are adjusted.

3. The data fusion method for industrial data quality consistency testing based on artificial intelligence according to claim 1, characterized in that: The adjustment of the data transmission process of the target device through the second adjustment strategy is as follows: Based on the time series deviation anomaly index of the target industrial data and the maximum value of the time series deviation threshold range, the second deviation degree of the time series deviation anomaly of the target industrial data is obtained; The second adjustment strategy refers to matching the clock calibration frequency increase factor, the database buffer capacity increase factor, and the data acquisition frequency increase factor from the attribute database based on the second deviation degree of the time series deviation anomaly of the target industrial data. By increasing the clock calibration frequency, database buffer capacity, and data acquisition frequency, the clock calibration frequency, database buffer capacity, and data acquisition frequency of the target device during data transmission are adjusted.

4. The data fusion method for industrial data quality consistency testing based on artificial intelligence according to claim 1, characterized in that: The time-series quality consistency assessment results based on the target industrial data are used to perform time-series synchronization calibration on the target industrial data. The specific calibration process is as follows: If the time series deviation anomaly index of the target industrial data falls within the time series deviation threshold range, then the first data calibration scheme is maintained. If the time series deviation anomaly index of the target industrial data is less than the minimum value of the time series deviation threshold range, then the second data calibration scheme is executed to update the target industrial data. If the time series deviation anomaly index of the target industrial data is greater than the maximum value of the time series deviation threshold range, then the third data calibration scheme is executed to update the target industrial data; The target industrial data is transmitted to the target industrial data platform, which then integrates the target industrial data with historical data.

5. The data fusion method for industrial data quality consistency testing based on artificial intelligence according to claim 1, characterized in that: The determination process for whether to perform secondary adjustments on the data transmission process of the target device is as follows: After timing synchronization calibration, the timing deviation anomaly index of the target monitored industrial data is obtained and compared with the timing deviation threshold range. If the timing deviation anomaly index of the target monitored industrial data still does not fall within the timing deviation threshold range, it is determined that the data transmission process of the target equipment should be adjusted a second time. If the time-series deviation anomaly index of the target monitored industrial data falls within the time-series deviation threshold range, it is determined that no secondary adjustment will be made to the data transmission process of the target device. The secondary adjustment specifically refers to: if the time series deviation anomaly index of the target industrial data is less than the minimum value of the time series deviation threshold range, then the data transmission process of the target equipment is adjusted secondaryly through the first adjustment strategy; If the time-series deviation anomaly index of the target industrial data is greater than the maximum value of the time-series deviation threshold range, the deviation value of the time-series deviation anomaly index of the target industrial data is obtained and compared with the preset time-series deviation anomaly index deviation value threshold in the database. If the deviation value of the time-series deviation anomaly index of the target industrial data is greater than or equal to the time-series deviation anomaly index deviation value threshold, it is determined that the data transmission process of the target equipment will be adjusted a second time through the second adjustment strategy. If the deviation value of the time series deviation anomaly index of the target industrial data is less than the threshold value of the time series deviation anomaly index, then no secondary adjustment will be made to the data transmission process of the target equipment. After the second adjustment, the time series deviation anomaly index of the target monitored industrial data is obtained and compared with the time series deviation threshold range. If the time series deviation anomaly index of the target monitored industrial data still does not fall within the time series deviation threshold range, a time series warning is issued for the data transmission process of the target equipment.

6. The data fusion method for industrial data quality consistency testing based on artificial intelligence according to claim 1, characterized in that: The data fusion process parameters for acquiring and analyzing target industrial data are specifically analyzed as follows: The data fusion process parameters of the target industrial data include the data fusion channel bit error rate of the target industrial data, the data fusion decision fluctuation frequency of the target industrial data, and the time series deviation anomaly index of the target industrial data. The aforementioned data fusion channel bit error rate refers to the proportion of data transmission and fusion errors caused by signal distortion in the data fusion channel during the data fusion process of the target industrial data; The aforementioned data fusion decision fluctuation frequency refers to the frequency of abnormal fluctuations in target industrial data during the data fusion process due to improper weight allocation; The median value of the time series deviation threshold range is marked as the reference value of the time series deviation anomaly index; By introducing weighting coefficients, the influence of the target industrial data on the adjustment anomaly coefficient of the target industrial data is quantified by the ratio of the data fusion channel bit error rate to the defined data fusion channel bit error rate, the ratio of the data fusion decision fluctuation frequency of the target industrial data to the defined data fusion decision fluctuation frequency, and the deviation relationship between the absolute value of the difference between the time series deviation anomaly index and the reference value of the time series deviation anomaly index of the target industrial data and the reference value of the time series deviation anomaly index. The influence of each degree of influence is coupled to obtain the adjustment anomaly coefficient of the target industrial data. The adjustment anomaly coefficient of the target industrial data is used to quantify the degree of adjustment anomaly in the target industrial data.

7. The data fusion method for industrial data quality consistency testing based on artificial intelligence according to claim 6, characterized in that: The specific process for determining whether to issue a data fusion warning is as follows: The preset adjustment anomaly coefficient threshold is extracted from the attribute database and compared with the adjustment anomaly coefficient of the target industrial data. If the adjustment anomaly coefficient of the target industrial data is less than the adjustment anomaly coefficient threshold, it is determined that no data fusion warning will be issued. If the adjustment anomaly coefficient of the target industrial data is greater than or equal to the adjustment anomaly coefficient threshold, then the time-series asynchronous increase correction coefficient is matched from the attribute database to increase the maximum value of the time-series deviation threshold range, and the time-series asynchronous decrease correction coefficient is matched from the database to decrease the minimum value of the time-series deviation threshold range. The data fusion process of the target industrial data is also adjusted. After the adjustment is completed, the secondary adjustment anomaly coefficient of the target industrial data is obtained and compared with the adjustment anomaly coefficient threshold. If the secondary adjustment anomaly coefficient of the target industrial data is still greater than or equal to the adjustment anomaly coefficient threshold, then a data fusion warning is issued. If the secondary adjustment anomaly coefficient of the target industrial data is less than the adjustment anomaly coefficient threshold, then the data fusion process is initialized; The time series deviation anomaly index of the target industrial data is again determined to be within the corrected time series deviation threshold range. If the time series deviation anomaly index of the target industrial data is within the corrected time series deviation threshold range, it is determined that no data fusion warning will be issued. If the time series deviation anomaly index of the target industrial data is less than the minimum value of the corrected time series deviation threshold range, then a data fusion early warning is issued and the data fusion process of the target industrial data platform is adjusted based on the first anomaly correction measure. If the time series deviation anomaly index of the target industrial data is greater than the maximum value of the corrected time series deviation threshold range, a data fusion early warning will be issued and the data fusion process of the target industrial data platform will be adjusted based on the second anomaly correction measures.

8. The data fusion method for industrial data quality consistency testing based on artificial intelligence according to claim 7, characterized in that: The adjustment process for the data fusion of the target industrial data is as follows: If the time series deviation anomaly index of the target industrial data is less than the minimum value of the corrected time series deviation threshold range, the data fusion process of the target industrial data will be adjusted through the first anomaly correction measure. If the time series deviation anomaly index of the target industrial data is greater than the maximum value of the corrected time series deviation threshold range, the data fusion process of the target industrial data will be adjusted through the second anomaly correction measure. The first anomaly correction measure refers to matching the signal wheel gap increase coefficient and the protocol synchronization period increase coefficient from the attribute database based on the anomaly correction factor of the target industrial data, and adjusting the signal wheel gap and the protocol synchronization period by increasing the signal wheel gap and the protocol synchronization period. The second anomaly correction measure refers to matching the signal wheel gap reduction coefficient and the protocol synchronization cycle reduction coefficient from the attribute database based on the anomaly correction factor of the target industrial data, and adjusting the signal wheel gap and the protocol synchronization cycle by reducing the signal wheel gap and the protocol synchronization cycle.

Images