A method, system, storage medium and program product for generating real data

Abstract

A method, system, storage medium, and program product for generating real data. In this method, trend analysis is performed on target data to obtain a trend coefficient of the change trend of the target data; a global trend function is constructed based on the trend coefficient; a main period value and a secondary period value of the target data are obtained; a periodic superposition function is constructed based on the main period value and the secondary period value; the global trend function and the periodic superposition function are synthesized to obtain a time series feature function; the time series feature function is applied to each time sampling point in the target time period to obtain a reference data sequence; the reference data sequence is randomly perturbed according to a preset fluctuation parameter to obtain a target data sequence with random fluctuation characteristics; the target data sequence is adjusted according to preset data constraint conditions to generate real data that meets the data constraint conditions. This application improves the accuracy of the real data in reflecting the time series evolution law, and further improves the ability to detect defects in time series related function tests.

Description

Technical Field

[0001] This application belongs to the field of data processing, and particularly relates to a real data generation method, system, storage medium, and program product. Background Art

[0002] With the rapid development of the software industry, there is a large demand for test data in the software R & D process. To obtain test data, the current main method is to extract data from the production environment as test data. However, this method has the risks of data leakage and privacy security, and the data in the production environment may not cover all the required test scenarios.

[0003] In the related art, test data that meets the constraint conditions can be automatically generated by setting data generation rules, including constraint conditions such as data type, value range, and dependency relationship, and using a random algorithm. This method can avoid using real data in the production environment and can flexibly customize data characteristics according to test requirements.

[0004] However, when it is necessary to simulate the trend and periodic characteristics of data changing over time, the existing methods often simply regard time-series data as independent data points for generation, ignoring the time correlation between data, resulting in the generated data being difficult to reflect the time-series evolution law in the real scenario, making it difficult to discover potential system defects during time-series related functional tests. Summary of the Invention

[0005] This application provides a real data generation method, system, storage medium, and program product, which is used to improve the accuracy of real data reflecting the time-series evolution law, and further improve the ability to discover defects in time-series related functional tests.

[0006] In a first aspect, this application provides a real data generation method, which obtains target data within a preset time period, and the target data includes multiple data points arranged in chronological order;

[0007] Performs trend analysis on the target data to obtain a trend coefficient representing the change trend of the target data;

[0008] Constructs a global trend function according to the trend coefficient, and the global trend function is used to represent the change trend of the target data over time;

[0009] Performs periodic analysis on the target data to obtain the main period value and the secondary period value of the target data;

[0010] Constructs a period superposition function according to the main period value and the secondary period value, and the period superposition function is used to represent the periodic characteristics of the target data;

[0011] Performs function synthesis on the global trend function and the period superposition function to obtain a time-series feature function;

[0012] Obtain the target time period of the data to be generated, where the target time period includes multiple time sampling points;

[0013] Apply the time series feature function to each time sampling point in the target time period to obtain a reference data sequence;

[0014] Perform random perturbation processing on the reference data sequence according to a preset fluctuation parameter to obtain a target data sequence with random fluctuation characteristics;

[0015] Adjust the target data sequence according to the preset data constraint conditions to generate real data that meets the data constraint conditions.

[0016] By adopting the above technical solution, by obtaining the target data within a preset time period and performing trend analysis to obtain a trend coefficient, the accuracy of determining the overall change direction of the data can be improved. The global trend function constructed based on the trend coefficient can quantitatively describe the long-term change law of the target data and improve the characterization accuracy of the data evolution characteristics. Performing periodic analysis on the target data to obtain the primary and secondary period values and constructing a period superposition function based on this can effectively identify and characterize the multiple periodic characteristics in the data. The time series feature function synthesized by combining the global trend function and the period superposition function realizes the unified expression of the long-term trend and periodic characteristics of the data. Applying the time series feature function to the sampling points in the target time period to generate a reference data sequence and introducing fluctuation characteristics through random perturbation processing makes the generated data closer to the real scenario. Finally, adjust according to the preset constraint conditions to ensure that the generated data meets the actual application requirements. Through multi-dimensional analysis and refined modeling of the data characteristics, the authenticity and usability of the generated data are improved, so that the generated data not only maintains the essential characteristics of the original data but also has the randomness and constraints required by the actual scenario.

[0017] Combined with some embodiments of the first aspect, in some embodiments, performing trend analysis on the target data to obtain a trend coefficient representing the change trend of the target data specifically includes:

[0018] Divide the target data into multiple data subsets according to a preset time window;

[0019] Calculate the trend feature values for each data subset respectively to obtain a trend feature sequence;

[0020] Perform piecewise fitting on the trend feature sequence to obtain multiple local trend functions;

[0021] Extract the trend coefficient based on the change characteristics of the multiple local trend functions.

[0022] By adopting the above technical solution, the target data is divided into multiple data subsets according to a preset time window, making the analysis of data trends more refined and localized. By calculating the trend characteristic values of each data subset to form a trend characteristic sequence, the local change characteristics of the data at different time periods can be captured. Segmentally fitting the trend characteristic sequence to obtain multiple local trend functions realizes the segmented quantitative expression of the data trend. Extracting trend coefficients based on the change characteristics of multiple local trend functions enables the trend coefficients to more accurately reflect the change laws of the data at different time scales. This segmented analysis method based on time windows improves the accuracy and adaptability of trend analysis, enables the extracted trend coefficients to more comprehensively characterize the change characteristics of the data, and provides a more reliable characteristic basis for constructing a global trend function in the subsequent stage.

[0023] Combined with some embodiments of the first aspect, in some embodiments, a time series feature function is obtained by performing function synthesis on the global trend function and the periodic superposition function, specifically including:

[0024] Construct a time series feature basis function set, which includes multiple trend basis functions generated based on the global trend function;

[0025] Calculate the correlation coefficients between the periodic superposition function and each trend basis function;

[0026] Decompose the periodic superposition function according to the correlation coefficients to obtain the periodic components corresponding to each trend basis function;

[0027] Perform weighted combination of each trend basis function and the corresponding periodic component to obtain the time series feature function.

[0028] By adopting the above technical solution, a time series feature basis function set containing multiple trend basis functions is constructed, providing a complete function basis for the decomposition of the periodic superposition function. By calculating the correlation coefficients between the periodic superposition function and each trend basis function, the quantitative evaluation of periodic characteristics is realized. Decomposing the periodic superposition function according to the correlation coefficients to obtain the periodic components corresponding to each trend basis function enables the effective fusion of periodic characteristics and trend characteristics. Performing weighted combination of each trend basis function and the corresponding periodic component to obtain the time series feature function realizes the unified expression of the data trend characteristics and periodic characteristics. This method based on function decomposition and weighted combination enhances the expression ability of the time series feature function for data characteristics, enabling the generated feature function to more accurately depict the overall characteristics of the data.

[0029] Combined with some embodiments of the first aspect, in some embodiments, after adjusting the target data sequence according to preset data constraint conditions to generate real data that meets the data constraint conditions, the method further includes:

[0030] Construct a feature marker sequence, which is used to characterize the temporal variation law of real data;

[0031] Analyze the importance degree of real data based on the feature marker sequence;

[0032] Classify the data elements of real data according to the importance degree to obtain a classification result;

[0033] Establish a multi-level association mapping relationship according to the classification result;

[0034] Adjust the sequence structure of real data through the association mapping relationship.

[0035] By adopting the above technical solution, the temporal variation law of real data is characterized by constructing a feature marker sequence, realizing the quantitative description of the dynamic characteristics of data. Analyzing the importance degree of real data based on the feature marker sequence makes the value evaluation of data more objective and accurate. Classifying the data elements according to the importance degree to obtain a classification result realizes the hierarchical management of data. Establishing a multi-level association mapping relationship according to the classification result enables the effective expression of the association characteristics between data of different importance degrees. Adjusting the sequence structure of real data through the association mapping relationship optimizes the organization form of data. This data structure optimization method based on importance classification and association mapping improves the organization efficiency and utilization value of data, making the data structure more suitable for actual application requirements.

[0036] Combined with some embodiments of the first aspect, in some embodiments, establishing a multi-level association mapping relationship according to the classification result specifically includes:

[0037] Extract the temporal characteristics of data elements to generate a temporal characteristic sequence;

[0038] Calculate the association degree between data elements at different levels based on the temporal characteristic sequence;

[0039] Construct an initial mapping rule according to the association degree;

[0040] Perform correlation analysis on data elements based on the initial mapping rule to obtain an optimized mapping rule;

[0041] Generate a multi-level association mapping relationship according to the optimized mapping rule.

[0042] By adopting the above technical solution, by extracting the temporal features of data elements to generate a temporal feature sequence, the changing rules and characteristics of data elements in the time dimension can be comprehensively captured. Based on the temporal feature sequence, calculating the degree of association between data elements at different levels can accurately quantify the mutual influence and dependence relationship between data elements. The initial mapping rule constructed according to the degree of association provides a basic framework for subsequent optimization. Conducting a correlation analysis on data elements to obtain an optimized mapping rule can more accurately reflect the actual correlation relationship between data elements. Finally, the multi-level association mapping relationship generated according to the optimized mapping rule not only reflects the hierarchy and correlation between data elements, but also reflects the internal law of the data structure, making the generated data more in line with the data characteristics in the real scenario, and improving the authenticity and usability of the data. The establishment process of this multi-level association mapping relationship takes into account the temporal features and correlation features of data elements, making the generated data not only maintain the continuity of the time series, but also maintain a reasonable correlation between data at different levels, so that the generated data is closer to the requirements of the actual application scenario.

[0043] Combined with some embodiments of the first aspect, in some embodiments, after adjusting the sequence structure of the real data through the association mapping relationship, the method further includes:

[0044] Obtain the adjustment effect index of the real data;

[0045] Construct a data flow evaluation model based on the adjustment effect index;

[0046] Generate a data flow optimization strategy according to the data flow evaluation model;

[0047] Adopt the data flow optimization strategy to dynamically optimize the real data.

[0048] By adopting the above technical solution, by obtaining the adjustment effect index of the real data, the actual effect of data adjustment can be quantitatively evaluated. The data flow evaluation model constructed based on the adjustment effect index can systematically evaluate the flow performance of data. The data flow optimization strategy generated according to the data flow evaluation model can specifically improve the organizational structure of data. Adopting the data flow optimization strategy to dynamically optimize the real data enables the data to be continuously adjusted and improved according to actual needs. This dynamic optimization mechanism based on evaluation indicators and optimization strategies makes the generated data have better adaptability and usability, and can adjust data characteristics in a timely manner with the change of the application environment to maintain the real-time effectiveness of the data. By establishing a closed-loop mechanism for data flow evaluation and optimization, the quality and efficiency of data generation are improved, the dynamic adaptation ability of the generated data is enhanced, and the data better meets the dynamic requirements in actual applications.

[0049] In some embodiments in combination with some embodiments of the first aspect, a data flow optimization strategy is adopted to dynamically optimize real data, which specifically includes:

[0050] Collect real-time flow parameters of real data;

[0051] Calculate the data flow efficiency based on the real-time flow parameters;

[0052] Adjust the optimization parameters based on the data flow efficiency;

[0053] Update the data flow optimization strategy according to the optimization parameters;

[0054] Apply the updated data flow optimization strategy to adjust the organization mode of real data.

[0055] By adopting the above technical solutions, by collecting the real-time flow parameters of real data, the flow state of the data can be grasped in real time. The data flow efficiency calculated based on the real-time flow parameters can objectively reflect the actual performance of the data flow. The optimization parameters adjusted based on the data flow efficiency can accurately guide the direction of data optimization. The data flow optimization strategy updated according to the optimization parameters can adapt to the dynamic changes of the data flow characteristics. Applying the updated data flow optimization strategy to adjust the organization mode of real data enables the data organization structure to dynamically adapt to the actual needs. This dynamic optimization mechanism based on real-time parameters improves the utilization efficiency and applicability of the data by continuously adjusting and optimizing the data organization mode, making the generated data have stronger practical value and adaptability in actual applications and being able to better meet the requirements for data real-time and effectiveness in actual applications.

[0056] In a second aspect, an embodiment of the present application provides a real data generation system, which includes: one or more processors and a memory; the memory is coupled to the one or more processors, and the memory is used to store computer program code, and the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the system to execute the methods described in the first aspect and any possible implementation manner in the first aspect.

[0057] In a third aspect, an embodiment of the present application provides a computer-readable storage medium, including instructions, when the above instructions run on the system, enabling the system to execute the methods described in the first aspect and any possible implementation manner in the first aspect.

[0058] In a fourth aspect, an embodiment of the present application provides a computer program product, characterized in that when the computer program product runs on the system, enabling the system to execute the methods described in any possible implementation manner in the first aspect.

[0059] One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:

[0060] 1. The present application provides a real data generation method. By obtaining target data within a preset time period and performing trend analysis to obtain a trend coefficient, the accuracy of determining the overall change direction of the data can be improved. The global trend function constructed based on the trend coefficient can quantitatively describe the long-term change law of the target data, improving the accuracy of characterizing the data evolution characteristics. By performing periodic analysis on the target data to obtain primary and secondary period values and constructing a periodic superposition function accordingly, the multiple periodic characteristics in the data can be effectively identified and characterized. The time series feature function synthesized by combining the global trend function and the periodic superposition function realizes the unified expression of the long-term trend and periodic characteristics of the data. Applying the time series feature function to the sampling points in the target time period to generate a reference data sequence and introducing fluctuation characteristics through random perturbation processing makes the generated data closer to the real scenario. Finally, adjusting according to the preset constraint conditions ensures that the generated data meets the actual application requirements. Through multi-dimensional analysis and refined modeling of data characteristics, the authenticity and usability of the generated data are improved, enabling the generated data to not only maintain the essential characteristics of the original data but also possess the randomness and constraints required by the actual scenario.

[0061] 2. The present application provides a real data generation method. By constructing a feature marker sequence to represent the time series change law of real data, the quantitative description of the dynamic characteristics of the data is realized. Analyzing the importance degree of real data based on the feature marker sequence makes the value evaluation of the data more objective and accurate. Grading data elements according to the importance degree to obtain a grading result realizes the hierarchical management of the data. Establishing a multi-level association mapping relationship according to the grading result enables the effective expression of the association characteristics between data of different importance degrees. Adjusting the sequence structure of real data through the association mapping relationship optimizes the organization form of the data. This data structure optimization method based on importance grading and association mapping improves the organization efficiency and utilization value of the data, making the data structure more suitable for actual application requirements.

[0062] 3. The present application provides a real data generation method. By obtaining the adjustment effect indicators of real data, the actual effect of data adjustment can be quantitatively evaluated. The data flow evaluation model constructed based on the adjustment effect indicators can systematically evaluate the data flow performance. The data flow optimization strategy generated according to the data flow evaluation model can specifically improve the data organization structure. Using the data flow optimization strategy to dynamically optimize the real data enables the data to be continuously adjusted and improved according to actual needs. This dynamic optimization mechanism based on evaluation indicators and optimization strategies makes the generated data have better adaptability and practicability, can adjust data characteristics in a timely manner with the change of the application environment, and maintain the real-time effectiveness of the data. By establishing a closed-loop mechanism for data flow evaluation and optimization, the quality and efficiency of data generation are improved, the dynamic adaptation ability of the generated data is enhanced, and the data can better meet the dynamic requirements in actual applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 is a schematic flowchart of a real data generation method in an embodiment of the present application.

[0064] Figure 2 is a schematic flowchart of a real data generation method based on data flow optimization in an embodiment of the present application.

[0065] Figure 3 is a schematic structural diagram of an entity device of a real data generation system provided in an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0066] The terms used in the following embodiments of the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in the specification and appended claims of the present application, the singular forms "a", "an", "the", "above-mentioned", "said", and "this" are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term "and / or" used in the present application refers to any and all possible combinations including one or more of the listed items.

[0067] Hereinafter, the terms "first" and "second" are only used for descriptive purposes and cannot be construed as implying or suggesting relative importance or implicitly indicating the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include one or more of such features. In the description of the embodiments of the present application, unless otherwise stated, the meaning of "a plurality" is two or more.

[0068] The following uses an embodiment and in combination with Figure 1 to describe a real data generation method in an embodiment of the present application:

[0069] Please refer to Figure 1 , which is a schematic flowchart of a real data generation method in an embodiment of this application.

[0070] S101. Obtain target data within a preset time period, perform trend analysis on the target data, and obtain a trend coefficient representing the change trend of the target data;

[0071] The system obtains target data within a preset time period, performs trend analysis on the target data, and obtains a trend coefficient representing the change trend of the target data. The target data includes multiple data points arranged in chronological order. Specifically: divide the target data into multiple data subsets according to a preset time window; calculate trend feature values for each data subset respectively to obtain a trend feature sequence; perform piecewise fitting on the trend feature sequence to obtain multiple local trend functions; extract the trend coefficient based on the change characteristics of the multiple local trend functions.

[0072] In this step, the system first obtains the target data within a preset time period. The target data can be various time series data, such as sales data, environmental monitoring data collected by sensors, etc. After obtaining the target data, the system performs trend analysis on it to obtain a trend coefficient representing the change trend of the target data. Trend analysis can reveal the overall change direction and law of the data.

[0073] In specific implementation, the system can use a variety of methods for trend analysis. A feasible way is to divide the target data into multiple data subsets according to a preset time window, and then calculate trend feature values for each data subset respectively to obtain a trend feature sequence. Then, the system can perform piecewise fitting on the trend feature sequence to obtain multiple local trend functions. Finally, the system can extract the trend coefficient based on the change characteristics of the multiple local trend functions as the overall trend feature of the entire target data.

[0074] In practical applications, due to the complexity and diversity of the target data, accurate trend analysis of it may face some challenges. To address this issue, the system can also introduce some optimization measures. For example, the system can adopt a certain overlap when dividing the data subsets to smooth the local fluctuations of the data; for data with large noise, the system can first perform smoothing filtering and then perform trend analysis; in addition, the system can also select appropriate trend feature values (such as mean, slope, etc.) and fitting methods (such as linear regression, polynomial fitting, etc.) according to the characteristics of the data to improve the accuracy of trend analysis. Through the above measures, the system can more accurately capture the change trend of the target data and provide a reliable trend coefficient for the subsequent steps.

[0075] S102. Construct a global trend function according to the trend coefficient;

[0076] The system constructs a global trend function based on the trend coefficient, and the global trend function is used to characterize the change trend of the target data over time. In this step, the system constructs a global trend function according to the trend coefficient obtained in the previous step. This function is used to characterize the overall change trend of the target data over the entire time span. By constructing the global trend function, the system can better depict and predict the long-term trend of the data.

[0077] In specific implementation, the system can adopt various function forms to construct the global trend function, such as linear function, exponential function, logarithmic function, etc. The choice of function form depends on the characteristics of the target data and the change pattern revealed by the trend coefficient. For example, if the trend coefficient indicates that the data shows a stable linear growth, then the system can choose a linear function as the global trend function; if the trend coefficient shows that the data has a trend of accelerating growth, then an exponential function may be more suitable. The system can use the method of parameter fitting to find the optimal parameters of the global trend function according to the trend coefficient.

[0078] S103. Conduct a periodic analysis on the target data to obtain the main period value and the secondary period value of the target data;

[0079] In this step, the system conducts a periodic analysis on the target data to obtain the main period value and the secondary period value of the data. The periodic analysis can reveal the implicit periodic patterns in the data, such as seasonality, daily periodicity, etc. By finding these periodic components, the system can more accurately depict the fluctuation characteristics of the data.

[0080] In specific implementation, the system can adopt various methods for periodic analysis, such as Fourier transform, wavelet analysis, etc. Taking Fourier transform as an example, the system can perform Fourier transform on the target data to obtain the representation of the data in the frequency domain. By analyzing the power spectral density in the frequency domain, the system can find the main frequency components in the data, and then determine the corresponding period values. Generally speaking, the frequency with the maximum power corresponds to the main period of the data, and the frequency with the second highest power corresponds to the secondary period of the data. The system can calculate the corresponding period lengths according to these frequency values.

[0081] In practical applications, since the periodic components of the data may be interfered by the trend term and noise, the effect of directly conducting periodic analysis may not be ideal enough. To improve the analysis accuracy, the system can, before the periodic analysis, first perform detrending processing and smoothing filtering on the data. The detrending processing can eliminate the long-term trend components in the data, making the periodic components more prominent; the smoothing filtering can suppress the high-frequency noise in the data, making the periodic signal clearer. In addition, the system can also combine the characteristics of the data, select appropriate periodic analysis methods and parameter settings, such as using period estimation based on the autocorrelation function, scale selection in wavelet transform, etc., to further optimize the effect of periodic analysis.

[0082] S104. Construct a period superposition function based on the main period value and the secondary period value;

[0083] The system constructs a period superposition function based on the main period value and the secondary period value, and the period superposition function is used to characterize the periodic characteristics of the target data.

[0084] In this step, the system constructs a period superposition function according to the main period value and the secondary period value obtained in the previous step. This function is used to characterize the periodic fluctuations of the target data and depict the oscillation modes of the data at different period scales. By constructing the period superposition function, the system can better restore the periodic components of the data.

[0085] In specific implementation, the system can adopt various function forms to construct the period superposition function, such as trigonometric functions, sine functions, etc. The period superposition function is usually composed of multiple periodic functions with different periods, and each periodic function corresponds to a period component of the data. The main period value and the secondary period value determine the frequency parameters of each periodic function. The system can use the method of parameter fitting to solve the amplitude and phase parameters of the period superposition function according to the actual period characteristics of the data, so that it can fit the periodic fluctuations of the data to the greatest extent.

[0086] It should be noted that in some cases, the periodic components of the data may be relatively complex, including multiple mutually coupled period components. In this regard, the system can consider using some high-order periodic functions to construct the period superposition function, such as Fourier series, nonlinear periodic functions, etc. These high-order functions can describe more complex periodic patterns and provide greater expressive power. At the same time, the system can also introduce some optimization measures, such as adding a regularization term in the fitting process to control the complexity of the period superposition function and avoid the problem of overfitting.

[0087] By reasonably constructing the period superposition function, the system can accurately depict the periodic characteristics of the target data, laying a foundation for subsequent function synthesis and data generation. The period superposition function and the global trend function complement each other and jointly constitute a complete description of the data characteristics.

[0088] S105. Synthesize the global trend function and the period superposition function to obtain a time series feature function;

[0089] The system synthesizes the global trend function and the period superposition function to obtain a time series feature function. Specifically: construct a time series feature basis function set, and the time series feature basis function set includes multiple trend basis functions generated based on the global trend function; calculate the correlation coefficients between the period superposition function and each trend basis function; decompose the period superposition function according to the correlation coefficients to obtain the period components corresponding to each trend basis function; and perform weighted combination of each trend basis function and the corresponding period component to obtain the time series feature function.

[0090] In this step, the system synthesizes the previously obtained global trend function and periodic superposition function to generate a time series feature function. This function combines the trend feature and periodic feature of the data, and can comprehensively describe the evolution law of the data in the time dimension. Through function synthesis, the system can obtain a more accurate and complete data feature model.

[0091] In specific implementation, the core of function synthesis is to organically combine the global trend function and periodic superposition function. A commonly used method is to use the periodic superposition function as the modulation term of the global trend function, and combine the two functions through multiplication or addition. For example, the system can multiply the periodic superposition function by the global trend function, so that the periodic fluctuations of the data show different amplitude changes along with the overall trend; or, the system can superimpose the periodic superposition function on the global trend function, so that the data shows periodic fluctuations while maintaining the overall trend.

[0092] To further optimize the effect of function synthesis, the system can adopt some more refined synthesis strategies. For example, the system can construct a time series feature basis function set, which includes multiple trend basis functions generated based on the global trend function. Then, the system calculates the correlation coefficients between the periodic superposition function and each trend basis function, decomposes the periodic superposition function according to the correlation coefficients, and obtains the periodic components corresponding to each trend basis function. Finally, the system weights and combines each trend basis function with the corresponding periodic component to generate the final time series feature function. This synthesis strategy based on correlation decomposition can better consider the mutual influence between the trend term and the periodic term, and improve the expression ability of the synthesized function.

[0093] Another possible problem is that in some cases, there may be certain conflicts or incompatibilities between the global trend function and the periodic superposition function. For example, the value ranges of the two functions may deviate greatly in some time periods, resulting in the synthesized function losing its physical meaning. In response, the system can introduce some constraint conditions during the function synthesis process to limit and optimize the properties such as the value range, continuity, and smoothness of the synthesized function, so as to ensure that the generated time series feature function meets the characteristics and requirements of the actual data. At the same time, the system can also evaluate the effects of different synthesis strategies through methods such as cross-validation and select the optimal solution.

[0094] S106. Obtain the target time period of the data to be generated, and apply the time series feature function to each time sampling point in the target time period to obtain a reference data sequence;

[0095] The system obtains the target time period of the data to be generated, applies the time series feature function to each time sampling point in the target time period, and obtains a reference data sequence. The target time period includes multiple time sampling points.

[0096] In this step, the system first obtains the target time period of the data to be generated. The target time period refers to the time interval for which data needs to be generated, which can be a past historical period or a future prediction interval. After determining the target time period, the system applies the time series feature function to each time sampling point within this time period, and generates a reference data sequence through function mapping.

[0097] Specifically, when implementing, the system needs to perform appropriate discretization processing on the target time period, dividing it into multiple time sampling points. The density of the time sampling points can be set according to the time granularity of the data and the generation requirements, such as by day, by hour, etc. Then, the system applies the time series feature function to each time sampling point, calculates the corresponding function value, and obtains a reference data sequence corresponding to the time sampling points. This reference data sequence reflects the basic trend of the data in terms of trend and periodicity within the target time period.

[0098] S107. Perform random perturbation processing on the reference data sequence according to the preset fluctuation parameters to obtain a target data sequence containing random fluctuation characteristics;

[0099] The system performs random perturbation processing on the reference data sequence according to the preset fluctuation parameters to obtain a target data sequence containing random fluctuation characteristics. The preset fluctuation parameters are used to control the fluctuation amplitude and frequency of the target data sequence, so that the generated target data sequence has more realistic random fluctuation characteristics. By introducing random perturbation into the reference data sequence, the common uncertainties and randomness in real data can be simulated, improving the authenticity and diversity of the generated data.

[0100] Specifically, when implementing, the system can use a variety of random perturbation methods to process the reference data sequence. For example, the system can generate random perturbation amounts based on Gaussian noise, and superimpose the random perturbation amounts on each data point of the reference data sequence to introduce random fluctuations. The amplitude of the random perturbation amount can be adjusted according to the preset fluctuation parameters to control the fluctuation degree of the target data sequence. In addition, the system can also use other types of random noise, such as uniform noise, Poisson noise, etc., to generate random perturbations with different characteristics. The system can select a suitable random noise model according to the specific application scenario and data characteristics to generate random perturbations that conform to the real data distribution.

[0101] S108. Adjust the target data sequence according to the preset data constraint conditions to generate real data that meets the data constraint conditions.

[0102] The system adjusts the target data sequence according to the preset data constraint conditions to generate real data that meets the data constraint conditions. The preset data constraint conditions include requirements and restrictions on various indicators or attributes of the generated data, such as the value range, statistical characteristics, correlation relationships, etc. By adjusting the target data sequence to conform to the preset data constraint conditions, it can be ensured that the generated real data matches the actual scenario in all aspects and has higher quality and usability.

[0103] In the specific implementation process, the system can adopt a variety of data adjustment techniques to meet the preset data constraint conditions. For example, the system can perform scaling or translation transformations on the target data sequence to make its numerical distribution meet specified statistical characteristics, such as mean, variance, quantiles, etc. The system can also fine-tune the target data sequence through optimization algorithms or iterative methods to better conform to the data constraint conditions while maintaining the time series characteristics and random fluctuation characteristics. In addition, for the case where there are correlation relationships between multiple target data sequences, the system can perform joint adjustment on each sequence through techniques such as correlation analysis and causal inference to maintain their logical consistency and correlation.

[0104] In the process of adjusting the target data sequence, there may be a problem that the adjusted data sequence deviates from the original time series characteristics and random fluctuation characteristics. To solve this problem, the system can minimize the impact on the original characteristics while meeting the data constraint conditions. A feasible method is to introduce a regularization term or penalty term and consider both the retention of the original characteristics and the satisfaction of the constraint conditions in the objective function of data adjustment. By balancing the weights between different objectives, the system can generate real data while maximizing the retention of the time series characteristics and random fluctuation characteristics of the target data sequence. In addition, the system can also explore the optimal data adjustment scheme through techniques such as iterative optimization or multi-objective optimization to generate high-quality real data that meets the constraint conditions.

[0105] In the above embodiments, by obtaining the target data within a preset time period and performing trend analysis to obtain a trend coefficient, the accuracy of determining the overall change direction of the data can be improved. The global trend function constructed based on the trend coefficient can quantitatively describe the long-term change law of the target data and improve the characterization accuracy of the data evolution characteristics. By performing periodic analysis on the target data to obtain the primary and secondary period values and constructing a periodic superposition function based on them, the multiple periodic characteristics in the data can be effectively identified and characterized. The time series feature function synthesized by combining the global trend function and the periodic superposition function realizes the unified expression of the long-term trend and periodic characteristics of the data. Applying the time series feature function to the sampling points in the target time period generates a reference data sequence, and introducing fluctuation characteristics through random perturbation processing makes the generated data closer to the real scenario. Finally, adjustment is made according to the preset constraint conditions to ensure that the generated data meets the actual application requirements. Through multi-dimensional analysis and refined modeling of data characteristics, the authenticity and usability of the generated data are improved, so that the generated data not only retains the essential characteristics of the original data but also has the randomness and constraints required by the actual scenario.

[0106] Based on the real data generation method provided in the above embodiments, through trend analysis, periodic analysis, and function synthesis of the target data, accurate modeling of data characteristics and high-quality generation of real data are achieved. However, in practical applications, in addition to paying attention to data feature modeling, it is also necessary to consider the correlation relationship between data elements and the dynamic optimization problem of data. The following combines Figure 2 to describe a real data generation method based on data flow optimization in the embodiments of the present application:

[0107] Please refer to Figure 2 , which is a schematic flowchart of a real data generation method based on data flow optimization in the embodiments of the present application.

[0108] S201. Construct a feature marker sequence and analyze the importance degree of real data based on the feature marker sequence;

[0109] The system constructs a feature marker sequence and analyzes the importance degree of real data based on the feature marker sequence. The feature marker sequence is used to characterize the time series change law of real data.

[0110] The system constructs a feature marker sequence and analyzes the importance degree of real data. The feature marker sequence is a time series obtained by feature extraction and encoding of real data, which can effectively characterize the change law and feature distribution of real data at different time points. By analyzing the feature marker sequence, the importance of each data element in the real data can be evaluated, providing a basis for subsequent data classification and optimization.

[0111] In specific implementation, the system can adopt various feature extraction and encoding techniques to construct a feature marker sequence. For example, the system can use time series analysis methods such as moving average, trend decomposition, and seasonal adjustment to extract the trend features, periodic features, and random features of real data, and map these features into discrete marker symbols to form a feature marker sequence. In addition, the system can also apply machine learning algorithms such as clustering, classification, and anomaly detection to perform feature learning and encoding on real data, generating a feature marker sequence with high generalization and recognition. The system can select a suitable feature extraction and encoding method according to the specific application scenario and data characteristics to obtain the optimal feature marker sequence.

[0112] In the process of analyzing the importance degree of real data, there may be a problem that some feature markers have a greater impact on the importance assessment, while other feature markers contribute less. To solve this problem, the system can introduce a feature weight mechanism to assign different weight coefficients to different feature markers to balance their impact on the importance assessment. The system can adaptively adjust the feature weights through methods such as statistical analysis, expert knowledge, or machine learning, making the importance assessment result more accurate and reasonable. At the same time, the system can also adopt feature selection or dimensionality reduction techniques to screen out a group of feature markers that contribute the most to the importance assessment, reduce the computational complexity and noise interference, and improve the analysis efficiency and reliability. By optimizing the feature marker sequence and the importance assessment method, the system can better depict the importance distribution of real data and lay a foundation for subsequent data grading and optimization.

[0113] S202. Classify the data elements of the real data according to the importance degree to obtain a classification result;

[0114] The system classifies the data elements of the real data according to the importance degree to obtain a classification result. A data element is the basic unit that constitutes real data and can be a single data point, a data segment, or a data interval, etc. By classifying the importance of data elements, the key information and secondary information in the real data can be identified, providing guidance for subsequent data association and optimization.

[0115] In the specific implementation process, the system can adopt various classification methods to divide the importance of data elements. For example, the system can sort and group the data elements based on the importance assessment score, dividing the data elements with high importance scores into high levels and the data elements with low importance scores into low levels. The system can also use clustering algorithms such as K-means and hierarchical clustering to automatically group according to the importance characteristics of data elements to form different importance levels. In addition, the system can combine expert knowledge and business rules to manually label and adjust the importance of data elements to meet the classification requirements of specific fields.

[0116] S203. Establish a multi-level association mapping relationship according to the classification result;

[0117] The system establishes a multi-level association mapping relationship according to the classification result, specifically including: extracting the temporal characteristics of data elements to generate a temporal characteristic sequence; calculating the association degree between data elements at different levels based on the temporal characteristic sequence; constructing an initial mapping rule according to the association degree; performing correlation analysis on data elements based on the initial mapping rule to obtain an optimized mapping rule; generating a multi-level association mapping relationship according to the optimized mapping rule.

[0118] The system establishes a multi-level association mapping relationship according to the classification result. The multi-level association mapping relationship refers to the association and dependence relationship between data elements established between different importance levels, which is used to characterize the evolution law and mutual influence of data elements in time series. By constructing a multi-level association mapping relationship, the complex connections and conduction mechanisms inside the real data can be revealed, providing an important basis for data optimization.

[0119] In the specific implementation process, the system first extracts the temporal characteristics of data elements, such as trendiness, periodicity, mutability, etc., to generate a temporal characteristic sequence. Then, the system calculates the association degree between data elements at different levels based on the temporal characteristic sequence, and can use indicators such as correlation coefficient, mutual information, and causal strength to quantify the correlation. Next, the system constructs an initial mapping rule according to the association degree to map and associate data elements with strong correlation. On this basis, the system can also optimize the initial mapping rule and generate a more perfect and consistent mapping rule through strategies such as association transfer and conflict resolution. Finally, the system generates a multi-level association mapping relationship according to the optimized mapping rule to form an association network between data elements.

[0120] S204. Adjust the sequence structure of real data through the association mapping relationship;

[0121] The system adjusts the sequence structure of real data through the association mapping relationship. The sequence structure adjustment refers to optimizing the temporal arrangement and organization method of real data according to the association and dependence relationship between data elements to improve the correlation and consistency of data. Through the sequence structure adjustment, the real data can be made to more conform to the internal association pattern and evolution law, reduce data redundancy and noise, and improve data quality and usability.

[0122] In the specific implementation process, the system can adopt various sequence adjustment strategies to optimize the structure of real data. For example, the system can reorder and aggregate data elements based on the association mapping relationship, placing strongly associated data elements in adjacent positions to form association-compact data blocks. The system can also use techniques such as time series compression and curve fitting to smooth the real data and extract trends, removing random fluctuations and outliers to make the sequence structure more stable and continuous. In addition, the system can hierarchically organize and index the real data according to the association dependency relationship, constructing a multi-level data view and a fast retrieval mechanism to improve the efficiency of data access and analysis.

[0123] S205. Obtain the adjustment effect indicators of the real data, and construct a data flow evaluation model based on the adjustment effect indicators;

[0124] The system obtains the adjustment effect indicators of the real data and constructs a data flow evaluation model based on the adjustment effect indicators. The adjustment effect indicators are quantitative metrics for measuring the quality of sequence structure adjustment, such as the improvement of association strength, data compression rate, information fidelity, etc. By collecting and analyzing the adjustment effect indicators, the performance and advantages of different adjustment strategies can be objectively evaluated, providing a decision-making basis for optimizing data flow.

[0125] In the specific implementation process, the system can adopt various methods to obtain the adjustment effect indicators. For example, the system can calculate the association strength improvement indicator by comparing the data association before and after adjustment, reflecting the contribution of the adjustment to mining the internal relationship of the data. The system can also use information theory methods to measure the data compression rate and information loss before and after adjustment, evaluating the impact of the adjustment on data reduction and fidelity. In addition, the system can introduce domain expert knowledge and user feedback to subjectively evaluate and score the adjustment effect, comprehensively considering the applicability and interpretability of the adjustment.

[0126] Based on the obtained adjustment effect indicators, the system constructs a data flow evaluation model for quantifying and predicting the impact of different adjustment strategies on data flow. The evaluation model can adopt methods such as statistical learning and machine learning, and fit the mapping relationship between the adjustment effect indicators and the data flow performance through training and optimization. The input of the model can include adjustment strategy parameters, data characteristics, environmental factors, etc., and the output is performance indicators such as corresponding data flow efficiency, response time, and resource consumption. Through the data flow evaluation model, the system can quickly evaluate and compare the advantages and disadvantages of different adjustment strategies, predict the impact of the adjustment on data flow, and support the automatic selection and optimization of adjustment strategies.

[0127] In the process of constructing a data flow evaluation model, problems such as high model complexity and poor generalization ability may occur. To solve these problems, the system can adopt techniques such as model integration and regularization, and improve the stability and generalization performance of the model by combining multiple simple models and introducing prior knowledge. At the same time, the system can adopt mechanisms such as incremental learning and online update, enabling the evaluation model to adapt to the dynamic changes in the data flow environment and continuously optimize and improve. In addition, the system can introduce interpretability analysis techniques, such as feature importance ranking and decision visualization, to improve the transparency and credibility of the evaluation model, facilitating user understanding and application. By establishing an efficient, robust, and interpretable data flow evaluation model, the system can provide intelligent decision-making support for optimizing data flow and enhance the efficiency and quality of data processing and analysis.

[0128] S206. Generate a data flow optimization strategy according to the data flow evaluation model;

[0129] The system generates a data flow optimization strategy according to the data flow evaluation model. The data flow optimization strategy refers to a series of adjustment and optimization measures formulated for specific data flow scenarios and requirements, aiming to improve the efficiency, reliability, and security of data flow. Through the data flow evaluation model, the system can quantitatively analyze the impact of different strategies on data flow, automatically generate the optimal combination of optimization strategies, and guide the dynamic adjustment and continuous optimization of data flow.

[0130] In the specific implementation process, the system can adopt various optimization algorithms and strategy generation techniques, and automatically search for and combine the optimal adjustment measures according to the data flow evaluation model. For example, the system can use methods such as heuristic search and evolutionary computation to explore and optimize in the candidate strategy space and find the strategy combination with the highest comprehensive evaluation index. The system can also utilize techniques such as reinforcement learning and multi-objective optimization, and continuously improve and perfect the optimization strategy through a trial-and-error and feedback mechanism to adapt to the complex and changing data flow environment.

[0131] The generated data flow optimization strategy can include adjustment and optimization measures in multiple aspects, such as data compression, redundancy elimination, cache management, load balancing, security encryption, etc. These measures can provide customized optimization solutions for different stages and components of data flow, such as data collection, transmission, processing, storage, etc. The system can dynamically adjust the parameters and weights of the optimization strategy according to specific application requirements and resource limitations, and balance multiple objectives such as the performance, cost, and security of data flow.

[0132] S207. Dynamically optimize the real data by adopting the data flow optimization strategy.

[0133] The system adopts a data flow optimization strategy to dynamically optimize real data, specifically including: collecting real-time flow parameters of real data; calculating data flow efficiency based on the real-time flow parameters; adjusting optimization parameters based on the data flow efficiency; updating the data flow optimization strategy according to the optimization parameters; and applying the updated data flow optimization strategy to adjust the organization mode of real data.

[0134] The system adopts a data flow optimization strategy to dynamically optimize real data. Dynamic optimization refers to dynamically adjusting and updating the optimization strategy according to the real-time state and feedback of data flow, enabling it to adapt to changes in the data environment and continuously improve the performance of data flow. Through dynamic optimization, the system can achieve adaptive control and real-time optimization of data flow, and improve the intelligent level of data processing and analysis.

[0135] In the specific implementation process, the system can achieve dynamic optimization in various ways. First, the system collects various indicators and parameters of data flow in real time, such as flow rate, latency, error rate, etc., to form real-time monitoring data of data flow. Then, the system inputs the real-time monitoring data into the data flow evaluation model to calculate the execution effect and performance of the current optimization strategy. The system can evaluate the effectiveness and applicability of the optimization strategy by comparing the actual effect with the expected goal.

[0136] If the effect of the current optimization strategy fails to meet the expectation, or new data flow patterns and requirements emerge, the system can promptly trigger the adjustment and update of the optimization strategy. The system can adapt to the changing trend of data flow by adjusting parameters, weights, thresholds, etc. of the optimization strategy. At the same time, the system can also expand and enhance the capabilities of the optimization strategy by introducing new optimization measures, algorithms, models, etc. The adjustment of the optimization strategy can be progressive, stabilizing the optimization effect through small-scale and multi-step corrections; or it can be leapfrog, coping with drastic environmental changes through large-scale and subversive reforms.

[0137] During the process of dynamic optimization, the system needs to balance the stability and adaptability of the optimization strategy. Too frequent or drastic strategy adjustments may introduce uncertainties and fluctuations, affecting the continuity and consistency of data flow. To solve this problem, the system can adopt mechanisms such as smooth transition and progressive update to reduce the impact and interference of strategy adjustments on data flow. At the same time, the system can also set reasonable adjustment thresholds and cooling times to avoid overly sensitive or short-sighted optimization behaviors.

[0138] In addition, the system can also introduce intelligent mechanisms such as self-learning and self-adaptation, enabling the optimization strategy to continuously evolve and improve based on historical experience and feedback. The system can adopt technologies such as incremental learning and online learning to continuously update and expand the knowledge base and rule base of the optimization strategy, improving the accuracy and generalization ability of optimization. At the same time, the system can also establish a user feedback and interaction mechanism, introduce artificial intelligence and expert experience, and guide and calibrate the optimization strategy.

[0139] Through dynamic optimization and continuous improvement, the system can keep the optimization strategy in an optimal state, adapt to the dynamic changes of the data flow environment, and provide users with efficient, reliable, and intelligent data processing and analysis services. The dynamic optimization mechanism enables the data system to transform from a static and passive state to a dynamic and active state, thus truly realizing the self-optimization and self-driving of data flow.

[0140] In the above embodiments, by constructing a feature marking sequence to represent the temporal variation law of real data, a quantitative description of the dynamic features of the data is realized. Based on the feature marking sequence, the importance degree of real data is analyzed, making the value evaluation of the data more objective and accurate. The data elements are graded according to the importance degree to obtain a grading result, realizing the hierarchical management of the data. According to the grading result, a multi-level association mapping relationship is established, enabling the effective expression of the association features between data of different importance degrees. By adjusting the sequence structure of the real data through the association mapping relationship, the organization form of the data is optimized. This data structure optimization method based on importance grading and association mapping improves the organization efficiency and utilization value of the data, making the data structure more suitable for actual application requirements.

[0141] The system in the embodiments of the present invention application will be described from the perspective of hardware processing. Please refer to Figure 3 which is a schematic structural diagram of an entity device of a real data generation system provided by an embodiment of the present application.

[0142] It should be noted that Figure 3 the structure of the system shown is only an example and should not bring any restrictions to the functions and usage scope of the embodiments of the present invention.

[0143] Such as Figure 3As shown, the system includes a Central Processing Unit (CPU) 301, which can perform various appropriate actions and processes according to the program stored in the Read-Only Memory (ROM) 302 or the program loaded from the storage section 308 into the Random Access Memory (RAM) 303, such as executing the method in the above embodiments. In the RAM 303, various programs and data required for system operation are also stored. The CPU 301, ROM 302, and RAM 303 are connected to each other via a bus 304. An Input / Output (I / O) interface 305 is also connected to the bus 304.

[0144] The following components are connected to the I / O interface 305: an input section 306 including a camera, an infrared sensor, etc.; an output section 307 including a Liquid Crystal Display (LCD), a speaker, etc.; a storage section 308 including a hard disk, etc.; and a communication section 309 including a network interface card such as a LAN (Local Area Network) card, a modem, etc. The communication section 309 performs communication processing via a network such as the Internet. A drive 310 is also connected to the I / O interface 305 as needed. A removable medium 311, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 310 as needed so that a computer program read from it can be installed into the storage section 308 as needed.

[0145] Specifically, according to an embodiment of the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present invention includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains a computer program for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network via the communication section 309, and / or installed from the removable medium 311. When the computer program is executed by the Central Processing Unit (CPU) 301, various functions defined in the present invention are executed.

[0146] It should be noted that the computer-readable medium shown in the embodiments of the present invention may be a computer-readable signal medium, a computer-readable storage medium, or any combination of the two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples of the computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present invention, the computer-readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In the present invention, the computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable computer program. Such a propagated data signal may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above.

[0147] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present invention. Among them, each block in the flowchart or block diagram may represent a module, a program segment, or a part of code, and the above module, program segment, or part of code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than marked in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram or flowchart, and the combination of blocks in the block diagram or flowchart, may be implemented by a dedicated hardware-based system for performing the specified functions or operations, or may be implemented by a combination of dedicated hardware and computer instructions.

[0148] As another aspect, the present invention also provides a computer-readable storage medium, which may be included in the system described in the above embodiments; or may exist alone without being assembled into the system. The above storage medium carries one or more computer programs, and when the above one or more computer programs are executed by a processor of a system, the system implements the method provided in the above embodiments.

[0149] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, and are not intended to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the various embodiments of the present application.

[0150] As used in the above embodiments, depending on the context, the term "when..." may be interpreted to mean "if...", or "after...", or "in response to determining...", or "in response to detecting...". Similarly, depending on the context, the phrase "when determining..." or "if detecting (the stated condition or event)" may be interpreted to mean "if determining...", or "in response to determining...", or "when detecting (the stated condition or event)", or "in response to detecting (the stated condition or event)".

[0151] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer may be a general-purpose computer, a dedicated computer, a computer network, or other programmable devices. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from a website, computer, server, or data center to another website, computer, server, or data center by wire (such as coaxial cable, optical fiber, digital subscriber line) or wirelessly (such as infrared, wireless, microwave, etc.). The computer-readable storage medium may be any available medium that can be accessed by a computer or a data storage device such as a server or data center that includes one or more available media integrated. The available medium may be a magnetic medium (for example, a floppy disk, a hard disk, a magnetic tape), an optical medium (for example, a DVD), or a semiconductor medium (for example, a solid-state drive), etc.

[0152] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by relevant hardware instructed by a computer program. This program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the above method embodiments. The aforementioned storage medium includes: various media that can store program codes, such as ROM, random access memory (RAM), magnetic disks, or optical discs.

Claims

1. A method for generating real data, characterized in that, Including: Obtain target data within a preset time period, where the target data includes multiple data points arranged in chronological order; Conduct trend analysis on the target data to obtain a trend coefficient characterizing the change trend of the target data; Construct a global trend function based on the trend coefficient, where the global trend function is used to characterize the change trend of the target data over time; Conduct periodic analysis on the target data to obtain the main period value and the secondary period value of the target data; Construct a periodic superposition function based on the main period value and the secondary period value, where the periodic superposition function is used to characterize the periodic characteristics of the target data; Perform function synthesis on the global trend function and the periodic superposition function to obtain a time series feature function; Obtain the target time period of the data to be generated, where the target time period includes multiple time sampling points; Apply the time series feature function to each time sampling point in the target time period to obtain a reference data sequence; Perform random perturbation processing on the reference data sequence according to a preset fluctuation parameter to obtain a target data sequence with random fluctuation characteristics; Adjust the target data sequence according to preset data constraint conditions to generate real data that meets the data constraint conditions; Construct a feature marking sequence, where the feature marking sequence is used to characterize the time series change law of the real data; Analyze the importance degree of the real data based on the feature marking sequence; Classify the data elements of the real data according to the importance degree to obtain a classification result; Establish a multi-level association mapping relationship according to the classification result, specifically including: Extract the time series features of the data elements to generate a time series feature sequence; Calculate the association degree between the data elements at different levels based on the time series feature sequence; Construct an initial mapping rule according to the association degree; Conduct association analysis on the data elements based on the initial mapping rule to obtain an optimized mapping rule; Generate a multi-level association mapping relationship according to the optimized mapping rule; Adjust the sequence structure of the real data through the association mapping relationship.

2. The method according to claim 1, characterized in that, The conducting trend analysis on the target data to obtain a trend coefficient characterizing the change trend of the target data specifically includes: Divide the target data into multiple data subsets according to a preset time window; Calculate trend feature values for each of the data subsets respectively to obtain a trend feature sequence; Perform piecewise fitting on the trend feature sequence to obtain multiple local trend functions; Extract the trend coefficient based on the change characteristics of the multiple local trend functions.

3. The method according to claim 1, wherein The performing function synthesis on the global trend function and the periodic superposition function to obtain a time series feature function specifically includes: Construct a time series feature basis function set, where the time series feature basis function set includes multiple trend basis functions generated based on the global trend function; Calculate the correlation coefficients between the periodic superposition function and each of the trend basis functions; Decompose the periodic superposition function according to the correlation coefficients to obtain periodic components corresponding to each of the trend basis functions; Perform weighted combination on each of the trend basis functions and the corresponding periodic components to obtain a time series feature function.

4. The method according to claim 1, wherein After adjusting the sequence structure of the real data through the associated mapping relationship, the method further includes: Obtaining an adjustment effect index of the real data; Constructing a data flow evaluation model based on the adjustment effect index; Generating a data flow optimization strategy according to the data flow evaluation model; Dynamically optimizing the real data by using the data flow optimization strategy.

5. The method according to claim 4, wherein The dynamically optimizing the real data by using the data flow optimization strategy specifically includes: Collecting real-time flow parameters of the real data; Calculating the data flow efficiency according to the real-time flow parameters; Adjusting optimization parameters based on the data flow efficiency; Updating the data flow optimization strategy according to the optimization parameters; Adjusting the organization mode of the real data by applying the updated data flow optimization strategy.

6. A real data generation system, characterized in that, The system includes: One or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the system to execute the method according to any one of claims 1-5.

7. A computer-readable storage medium, comprising instructions, characterized in that, When the instruction runs on the system, enabling the system to execute the method according to any one of claims 1-5.

8. A computer program product, characterized in that, When the computer program product runs on the system, enabling the system to execute the method according to any one of claims 1-5.

Images