Time series data prediction method, device, computer equipment and medium

Abstract

The present disclosure relates to the technical field of energy data processing, and provides a time series data prediction method, device, computer equipment and medium. The method includes: obtaining a benchmark data set and at least one supplementary test set; processing metadata in the benchmark data set and the supplementary test set based on a preset data set decomposition method to obtain a benchmark trend subset, a benchmark period subset, a supplementary trend subset and a supplementary period subset; generating a target trend model; generating a target period model; generating target prediction data based on a preset calculation strategy, a benchmark trend subset, a benchmark period subset, a target trend model and a target period model. Through the above steps, the embodiment of the present disclosure can greatly improve the prediction accuracy of the data.

Description

Technical Field

[0001] The present disclosure relates to the field of energy data processing technology, and in particular to a time series data prediction method, device, computer equipment and medium. Background Art

[0002] With the rapid development of data processing technology, the energy field has generated more and more data processing needs. In some cases, due to the limitations of current data or the sensitivity of data, very little data is available, and the heterogeneity of data from different sources is large, resulting in low accuracy of data prediction. Summary of the invention

[0003] In view of this, the embodiments of the present disclosure provide a time series data prediction method, apparatus, computer equipment and medium to solve the problem in the prior art that the data obtained is very small and the data heterogeneity problem of different sources is large, resulting in low accuracy of data prediction.

[0004] According to a first aspect of an embodiment of the present disclosure, a method for predicting time series data is provided, comprising: obtaining a benchmark data set and at least one supplementary test set; processing metadata in the benchmark data set and the supplementary test set based on a preset data set decomposition method to obtain a benchmark trend subset, a benchmark period subset, a supplementary trend subset, and a supplementary period subset; training an initial trend model through the benchmark trend subset and the supplementary trend subset to generate a target trend model; training an initial period model through the benchmark period subset and the supplementary period subset to generate a target period model; generating target prediction data based on a preset calculation strategy, a benchmark trend subset, a benchmark period subset, a target trend model, and a target period model.

[0005] According to a second aspect of an embodiment of the present disclosure, a time series data prediction device is provided, comprising: an acquisition module, configured to acquire a benchmark data set and at least one supplementary test set; a decomposition module, configured to process metadata in the benchmark data set and the supplementary test set based on a preset data set decomposition method to obtain a benchmark trend subset, a benchmark period subset, a supplementary trend subset and a supplementary period subset; a trend training module, configured to train an initial trend model through a benchmark trend subset and a supplementary trend subset to generate a target trend model; a period training module, configured to train an initial period model through a benchmark period subset and a supplementary period subset to generate a target period model; a generation module, configured to generate target prediction data based on a preset calculation strategy, a benchmark trend subset, a benchmark period subset, a target trend model and a target period model.

[0006] According to a third aspect of an embodiment of the present disclosure, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above method when executing the computer program.

[0007] According to a fourth aspect of the embodiments of the present disclosure, a computer-readable storage medium is provided, which stores a computer program, and when the computer program is executed by a processor, the steps of the above method are implemented.

[0008] Compared with the prior art, the beneficial effects of the embodiments of the present disclosure include at least the following: the embodiments of the present disclosure obtain a benchmark data set and at least one supplementary test set; process the metadata in the benchmark data set and the supplementary test set based on a preset data set decomposition method to obtain a benchmark trend subset, a benchmark period subset, a supplementary trend subset and a supplementary period subset; train the initial trend model through the benchmark trend subset and the supplementary trend subset to generate a target trend model; train the initial period model through the benchmark period subset and the supplementary period subset to generate a target period model; generate target prediction data based on a preset calculation strategy, a benchmark trend subset, a benchmark period subset, a target trend model and a target period model, which can greatly improve the prediction accuracy of the data. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0010] Figure 1 is a schematic diagram of a joint learning architecture of an embodiment of the present disclosure;

[0011] Figure 2 is a flow chart of a time series data prediction method provided by an embodiment of the present disclosure;

[0012] Figure 3 is a flowchart of a specific embodiment of a time series data prediction method provided by an embodiment of the present disclosure;

[0013] Figure 4 is a block diagram of a time series data prediction device provided by an embodiment of the present disclosure;

[0014] Figure 5 is a schematic diagram of a computer device provided by an embodiment of the present disclosure. DETAILED DESCRIPTION

[0015] In the following description, specific details such as specific system structures and technologies are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present disclosure. However, it should be clear to those skilled in the art that the present disclosure may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid obstructing the description of the present disclosure with unnecessary details.

[0016] The present disclosure will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.

[0017] Federated learning refers to the comprehensive use of various AI (Artificial Intelligence) technologies to jointly explore the value of data and give rise to new intelligent formats and models based on joint modeling, while ensuring data security and user privacy. Federated learning has at least the following characteristics:

[0018] (1) A weakly centralized joint training model in which participating nodes control their own data to ensure data privacy and security in the process of co-creating intelligence.

[0019] (2) In different application scenarios, we use screening and / or combination of AI algorithms and privacy-preserving computing to establish multiple model aggregation optimization strategies to obtain high-level, high-quality models.

[0020] (3) Under the premise of ensuring data security and user privacy, methods to improve the performance of the federated learning engine are obtained based on multiple model aggregation optimization strategies. The performance method can be achieved by solving problems including computing architecture parallelism, information interaction in large-scale cross-domain networks, intelligent perception, and exception handling mechanisms to improve the overall performance of the federated learning engine.

[0021] (4) Obtain the needs of multiple users in each scenario, determine a reasonable assessment of the true contribution of each joint participant through a mutual trust mechanism, and distribute incentives.

[0022] Based on the above methods, we can establish an AI technology ecosystem based on federated learning, give full play to the value of industry data, and promote the implementation of scenarios in vertical fields.

[0023] The present disclosure will be described in detail below with reference to the accompanying drawings.

[0024] Figure 1 Schematic diagram of a joint learning architecture of an embodiment of the present disclosure. Figure 1 As shown, the architecture of the federated learning may include a server (central node) 101 and participants 102 , 103 , and 104 .

[0025] In the joint learning process, the basic model can be established by the server 101, and the server 101 sends the model to the participants 102, 103, and 104 with which the communication connection is established. The basic model can also be uploaded to the server 101 by any participant after establishment, and the server 101 sends the model to other participants with which the communication connection is established. The participants 102, 103, and 104 build the model according to the downloaded basic structure and model parameters, use local data to train the model, obtain updated model parameters, and encrypt and upload the updated model parameters to the server 101. The server 101 aggregates the model parameters sent by the participants 102, 103, and 104 to obtain the global model parameters, and transmits the global model parameters back to the participants 102, 103, and 104. The participants 102, 103, and 104 iterate their respective models according to the received global model parameters until the model finally converges, thereby realizing the training of the model. In the joint learning process, the data uploaded by participants 102, 103 and 104 are model parameters, local data will not be uploaded to server 101, and all participants can share the final model parameters, so joint modeling can be achieved on the basis of ensuring data privacy. It should be noted that the number of participants is not limited to the three as described above, but can be set as needed, and the embodiments of the present disclosure do not limit this.

[0026] Figure 2 It is a flowchart of a time series data prediction method provided by an embodiment of the present disclosure. Figure 2 The time series data prediction method can be obtained by Figure 1 The terminal device or server 2 executes. Figure 2 As shown, the time series data prediction method includes:

[0027] S201, obtaining a benchmark data set and at least one supplementary test set.

[0028] The benchmark data set and the supplementary test set include data sets composed of multiple metadata, where metadata refers to a data structure composed of one or more feature data. The feature data can be basic numerical units, such as "average temperature: 35.4", "daily gas consumption: 55.65", etc.

[0029] S202, processing metadata in a benchmark dataset and at least one supplementary test set based on a preset dataset decomposition method to obtain a benchmark trend subset, a benchmark period subset, at least one supplementary trend subset and at least one supplementary period subset.

[0030] The data set decomposition method may be a method of decomposing a data set into multiple sub-data sets with the same structure as the data set, wherein the same structure may refer to the number of feature data, categories and other aspects of the metadata in two or more data sets are the same, the trend subset refers to the data subset decomposed based on trend changes, and the period subset refers to the data subset decomposed based on period changes. Trend changes refer to a tendency or state of continuous change over a period of time. For example, when the predicted value has a tendency to change upward or downward over time, it is considered to have a trend.

[0031] S203: Train the initial trend model using the reference trend subset and at least one supplementary trend subset to generate a target trend model.

[0032] The initial trend model refers to an existing or self-set mathematical formula related to trend changes, and the parameters of the mathematical formula can be constants, arrays, vectors, etc. Training the initial trend model can refer to the process of determining the model parameters through a series of steps or methods based on the obtained data of the baseline trend subset and the supplementary trend subset, and finally obtaining a target trend model that meets the requirements.

[0033] S204: Train the initial cycle model using the reference cycle subset and at least one supplementary cycle subset to generate a target cycle model.

[0034] The initial cycle model may refer to a pre-set model whose parameters are initial default values. The model is trained based on the data of the reference cycle subset and the supplementary cycle subset, and its parameters may be optimized to obtain a trained target cycle model.

[0035] S205, generating target prediction data based on a preset calculation strategy, a benchmark trend subset, a benchmark period subset, a target trend model and a target period model.

[0036] The disclosed embodiment acquires a benchmark data set and at least one supplementary test set; processes metadata in the benchmark data set and at least one supplementary test set based on a preset data set decomposition method to obtain a benchmark trend subset, a benchmark period subset, at least one supplementary trend subset and at least one supplementary period subset; trains an initial trend model through the benchmark trend subset and at least one supplementary trend subset to generate a target trend model; trains an initial period model through the benchmark period subset and at least one supplementary period subset to generate a target period model; generates target prediction data based on a preset calculation strategy, a benchmark trend subset, a benchmark period subset, a target trend model and a target period model, which can greatly improve the prediction accuracy of data.

[0037] In some embodiments, obtaining a benchmark data set and at least one supplementary test set includes: obtaining an original data set, wherein the original data set includes at least one metadata; processing at least one metadata in the original data set based on a basic data processing strategy to generate at least one basic processed metadata to obtain a benchmark data set; obtaining at least one original supplementary set; processing each original supplementary set in at least one original supplementary set based on a supplementary processing strategy to generate at least one supplemented processed original supplementary set to obtain at least one supplementary test set.

[0038] The original data set may refer to a data structure consisting of the original metadata obtained. The original metadata may refer to a data structure consisting of data in the original format obtained. The timestamp data may refer to the data representing the time in the metadata. The format of the timestamp data can be set as needed and is not specifically limited here. The basic data processing strategy may refer to the steps or methods for processing the data in the original format so that the data can be used to process the model. The supplementary processing strategy may refer to the steps or methods for processing the data in at least one supplementary test set in a certain way.

[0039] In some embodiments, the basic data processing strategy includes: performing exception processing on at least one acquired metadata to obtain metadata after exception processing, and the number of metadata can be one or more; performing smoothing processing on the metadata after exception processing to obtain the smoothed metadata to construct a smoothed data set; processing the smoothed data set based on the splitting processing strategy to obtain processed metadata.

[0040] Specifically, abnormal data can refer to one or some data in the metadata that do not meet the preset requirements. Exception processing can include checking data consistency or processing invalid values ​​and missing values, and deleting or replacing abnormal data in the metadata. Invalid values ​​can refer to null values, values ​​that do not meet the data type requirements, or other abnormal values. Missing values ​​can refer to the incomplete values ​​of one or some attributes in the existing data set.

[0041] Smoothing can refer to reducing the amplitude of changes in metadata, and smoothing can make the trend of data more obvious. As an example, smoothing can use the following mathematical formula:

[0042] F(t+n) / 2=(F(t+1)+F(t+1)+F(t+1)+...+F(t+n)) / n

[0043] Here, t represents the data number, F(t+n) represents the data with the t+nth number, and t and n are integers.

[0044] The splitting processing strategy refers to the steps or methods of splitting out one or more feature data based on the timestamp data. As an example, a timestamp data may be "September 4, 2017", and the timestamp data may be split out into any of the following data: "week: 6", "day of the year: 245", "week of the year: 35", "year information: 2017" or "monthly information: 09". By splitting the timestamp into other time feature data, the data dimension can be increased, making the training results more accurate.

[0045] In some embodiments, a smoothed data set is processed based on a split processing strategy to obtain processed metadata, including: obtaining at least one split index; generating an intermediate timestamp data set based on the split index and the timestamp data of each metadata in the smoothed data set; and updating each timestamp data in the smoothed data set to each metadata in the smoothed data set to obtain processed metadata.

[0046] In some embodiments, an intermediate timestamp data set is generated based on the splitting indicator and the timestamp data of each metadata in the smoothed data set, including: obtaining one of the splitting indicators that is not marked as split to obtain an intermediate indicator; processing the timestamp data of each metadata in the smoothed data set based on the intermediate indicator to generate intermediate timestamp data to obtain an intermediate timestamp data set; marking the intermediate indicator as split; repeating the above steps until each splitting indicator is marked as split to obtain an intermediate timestamp data set.

[0047] In some embodiments, the supplementary processing strategy includes: obtaining an original supplementary set, wherein the original supplementary set includes at least one metadata; obtaining a conversion coefficient corresponding to the original supplementary set; generating at least one converted metadata based on at least one metadata in the original supplementary set and the conversion coefficient; processing the at least one converted metadata based on the basic data processing strategy to generate at least one basic processed metadata, and obtaining the original supplementary set after the supplementary processing.

[0048] The conversion coefficient may refer to the coefficient for converting the metadata in the supplementary data. Since there are differences between the data in the supplementary set and the benchmark data set, a certain conversion coefficient may be set to reduce the weight of the data in the supplementary set for training the model. The conversion coefficient may be specified by human experience or calculated in a certain way. As an example, the statistical target is daily gas consumption, and the benchmark data set contains 100 daily gas consumption data. The 100 gas consumption data are summed to obtain the baseline data sum. A supplementary data set contains 200 data, and the 200 data are summed to obtain the supplementary data sum. Then the coefficient may be the supplementary data sum / (benchmark data sum + supplementary data sum). It should be pointed out that the above calculation method is only one method of use, and other methods of use may be set as needed, without specific restrictions here.

[0049] In some embodiments, the preset data set decomposition method is the STL additive data set decomposition method.

[0050] STL (Seasonal and Trend decomposition using Loess) is an implementation method based on local weighted regression. Among them, Loess (locally weighted scatterplot smoothing, LOWESS or LOESS) is a local polynomial regression fitting, which is a common method for smoothing two-dimensional scatter plots. It combines the simplicity of traditional linear regression and the flexibility of nonlinear regression. The STL additive data set decomposition method can refer to dividing the data set into three data subsets: trend subset, period subset and residual subset. Among them, the trend subset and period subset refer to the above description and will not be repeated here. The residual subset can refer to the data subset decomposed based on the residual phenomenon. The residual phenomenon can refer to the impact of many accidental factors on the time series. The corresponding metadata in the trend subset, period subset and residual subset are added together to obtain the data before the split. It should be pointed out that since the proportion of residuals is very small, generally not more than 1%, the residual subset is generally ignored. As an example, the initial data is 100. After being processed by STL additive data and decomposition method, the trend value can be 60 and the period value can be 40 (the residual value is ignored). Then, the trend value and the period value are added to get the initial data 100 before the split.

[0051] In some embodiments, the calculation strategy includes: importing the baseline trend subset into the target trend model to obtain trend target data; importing the baseline period subset into the target period model to obtain period target data; and obtaining target prediction data based on the trend target data and the period target data. The trend target data may refer to the predicted trend value, and the period target data may refer to the predicted period value. If the data set decomposition method is the STL additive data set decomposition method, the target prediction data can be obtained by adding the trend value to the target value.

[0052] Figure 3 This is a flow chart of the method for predicting daily gas consumption of Company A in 2021 provided in an embodiment of the present disclosure. Figure 3 The daily gas consumption forecast method of Company A in 2021 can be obtained by Figure 1 Server execution. Figure 3 As shown, the daily gas consumption prediction method includes:

[0053] S301, obtain the original daily gas consumption dataset of Company A in 2020, wherein the original daily gas consumption includes at least one metadata.

[0054] S302: Process at least one metadata in the original daily gas consumption data set based on the basic data processing strategy to generate at least one basic processed metadata to obtain a reference daily gas consumption data set.

[0055] S303, obtain the daily supplementary gas consumption dataset of Company B in 2020.

[0056] S304: Obtain a conversion coefficient corresponding to the supplementary daily gas consumption data set and the original daily gas consumption data set.

[0057] S305 , generating at least one converted metadata based on each metadata in the supplementary daily gas consumption dataset and the conversion coefficient, and obtaining a target daily gas consumption supplementary dataset.

[0058] S306 , based on the STL additive data set decomposition method, the metadata in the benchmark daily gas consumption data set and the target daily gas consumption supplementary data set are processed respectively to obtain a benchmark trend subset, a benchmark period subset, a supplementary trend subset and a supplementary period subset.

[0059] S307, training the initial trend model through the reference trend subset and the supplementary trend subset to generate a target trend model.

[0060] S308, training the initial cycle model through the reference cycle subset and the supplementary cycle subset to generate a target cycle model.

[0061] S309, importing the benchmark trend subset into the target trend model to obtain the trend target data of company A.

[0062] S310, importing the benchmark cycle subset into the target cycle model to obtain the cycle target data of Company A.

[0063] S311, sum the trend target data and the cycle target data to obtain the forecast data for Company A in 2021.

[0064] All the above optional technical solutions can be arbitrarily combined to form optional embodiments of the present application, which will not be described one by one here.

[0065] The following are embodiments of the device disclosed herein, which can be used to execute the method embodiments disclosed herein. For details not disclosed in the device embodiments disclosed herein, please refer to the method embodiments disclosed herein.

[0066] Figure 4 Schematic diagram of a time series data prediction device provided by an embodiment of the present disclosure. Figure 4 As shown, the time series data prediction device includes:

[0067] The acquisition module 401 is configured to acquire a benchmark data set and at least one supplementary test set.

[0068] The decomposition module 402 is configured to process metadata in the benchmark dataset and at least one supplementary test set based on a preset dataset decomposition method to obtain a benchmark trend subset, a benchmark period subset, at least one supplementary trend subset and at least one supplementary period subset.

[0069] The trend training module 403 is configured to train the initial trend model through the reference trend subset and at least one supplementary trend subset to generate a target trend model.

[0070] The cycle training module 404 is configured to train the initial cycle model through the reference cycle subset and at least one supplementary cycle subset to generate a target cycle model.

[0071] The generation module 405 is configured to generate target prediction data based on a preset calculation strategy, a reference trend subset, a reference period subset, a target trend model and a target period model.

[0072] According to the technical solution provided by the embodiments of the present disclosure, a benchmark data set and at least one supplementary test set are obtained; metadata in the benchmark data set and at least one supplementary test set are processed based on a preset data set decomposition method to obtain a benchmark trend subset, a benchmark period subset, at least one supplementary trend subset and at least one supplementary period subset; an initial trend model is trained through the benchmark trend subset and at least one supplementary trend subset to generate a target trend model; an initial period model is trained through the benchmark period subset and at least one supplementary period subset to generate a target period model; target prediction data is generated based on a preset calculation strategy, a benchmark trend subset, a benchmark period subset, a target trend model and a target period model, which can greatly improve the prediction accuracy of the data.

[0073] In some embodiments, the acquisition module 401 of the time series data prediction device is further configured to: acquire an original data set, wherein the original data set includes at least one metadata; process at least one metadata in the original data set based on a basic data processing strategy, generate at least one basic processed metadata, and obtain a benchmark data set; acquire at least one original supplementary set; process each original supplementary set in at least one original supplementary set based on a supplementary processing strategy, generate at least one supplemented original supplementary set, and obtain at least one supplementary test set.

[0074] In some embodiments, the basic data processing strategy includes: obtaining timestamp data of each metadata in at least one metadata to obtain at least one timestamp data; performing exception processing on the at least one metadata obtained to obtain at least one metadata after exception processing; performing smoothing processing on the at least one metadata after exception processing to obtain at least one metadata after smoothing processing to obtain a smoothed data set; processing the smoothed data set based on the split processing strategy to obtain at least one basic processed metadata.

[0075] In some embodiments, a smoothed data set is processed based on a split processing strategy to obtain at least one basic processed metadata, including: obtaining at least one split index; generating at least one intermediate timestamp data set based on the at least one split index and the timestamp data of each metadata in the smoothed data set; updating each timestamp data in the smoothed data set to each metadata in the smoothed data set to obtain at least one basic processed metadata.

[0076] In some embodiments, a smoothed data set is processed based on a split processing strategy to obtain at least one basic processed metadata, including: obtaining at least one split index; generating at least one intermediate timestamp data set based on the at least one split index and the timestamp data of each metadata in the smoothed data set; updating each timestamp data in the smoothed data set to each metadata in the smoothed data set to obtain at least one basic processed metadata.

[0077] In some embodiments, the supplementary processing strategy includes: obtaining an original supplementary set, wherein the original supplementary set includes at least one metadata; obtaining a conversion coefficient corresponding to the original supplementary set; generating at least one converted metadata based on at least one metadata and the conversion coefficient in the original supplementary set; processing the at least one converted metadata based on the basic data processing strategy to generate at least one basic processed metadata, and obtaining the original supplementary set after supplementary processing.

[0078] In some embodiments, the preset data set decomposition method is the STL additive data set decomposition method.

[0079] In some embodiments, the calculation strategy includes: importing the baseline trend subset into the target trend model to obtain trend target data; importing the baseline cycle subset into the target cycle model to obtain cycle target data; and obtaining target prediction data based on the trend target data and the cycle target data.

[0080] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present disclosure.

[0081] Figure 5 is a schematic diagram of a computer device 500 provided in an embodiment of the present disclosure. Figure 5 As shown, the computer device 500 of this embodiment includes: a processor 501, a memory 502, and a computer program 503 stored in the memory 502 and executable on the processor 501. When the processor 501 executes the computer program 503, the steps in the above-mentioned method embodiments are implemented. Alternatively, when the processor 501 executes the computer program 503, the functions of the modules / units in the above-mentioned device embodiments are implemented.

[0082] Exemplarily, the computer program 503 may be divided into one or more modules / units, which are stored in the memory 502 and executed by the processor 501 to complete the present disclosure. The one or more modules / units may be a series of computer program instruction segments capable of completing specific functions, which are used to describe the execution process of the computer program 503 in the computer device 500.

[0083] The computer device 500 may be a desktop computer, a notebook computer, a PDA, a cloud server, or other computer devices. The computer device 500 may include, but is not limited to, a processor 501 and a memory 502. Those skilled in the art will appreciate that Figure 5 This is only an example of the computer device 500 and does not constitute a limitation of the computer device 500. The computer device 500 may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the computer device may also include input and output devices, network access devices, buses, etc.

[0084] The processor 501 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or the processor may be any conventional processor, etc.

[0085] The memory 502 may be an internal storage unit of the computer device 500, for example, a hard disk or memory of the computer device 500. The memory 502 may also be an external storage device of the computer device 500, for example, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the computer device 500. Further, the memory 502 may also include both an internal storage unit of the computer device 500 and an external storage device. The memory 502 is used to store computer programs and other programs and data required by the computer device. The memory 502 may also be used to temporarily store data that has been output or is to be output.

[0086] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0087] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0088] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this disclosure.

[0089] In the embodiments provided in the present disclosure, it should be understood that the disclosed apparatus / computer equipment and methods can be implemented in other ways. For example, the apparatus / computer equipment embodiments described above are merely schematic. For example, the division of modules or units is only a logical function division. There may be other division methods in actual implementation. Multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces, indirect coupling or communication connection of the apparatus or unit, which may be electrical, mechanical or other forms.

[0090] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0091] In addition, each functional unit in each embodiment of the present disclosure may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0092] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present disclosure implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. The computer program may include computer program code, and the computer program code may be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electric carrier signals and telecommunication signals.

[0093] The above embodiments are only used to illustrate the technical solutions of the present disclosure, rather than to limit them. Although the present disclosure has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present disclosure, and should all be included in the protection scope of the present disclosure.

Claims

1. A time series data prediction method, characterized in that: include: Obtaining a benchmark dataset and at least one supplemental test set; Processing the metadata in the benchmark dataset and the supplementary test dataset based on a preset dataset decomposition method to obtain a benchmark trend subset, a benchmark period subset, a supplementary trend subset, and a supplementary period subset; Training the initial trend model through the reference trend subset and the supplementary trend subset to generate a target trend model; Training the initial cycle model through the reference cycle subset and the supplementary cycle subset to generate a target cycle model; Generate target prediction data based on a preset calculation strategy, the benchmark trend subset, the benchmark period subset, the target trend model and the target period model; The obtaining of at least one supplementary test set comprises: obtaining at least one original complement set; Processing each original supplementary set in the original supplementary set based on the supplementary processing strategy to generate at least one original supplementary set after supplementary processing to obtain the supplementary test set; The supplementary treatment strategies include: Acquire an original supplemental set, wherein the original supplemental set includes at least one metadata; Obtaining conversion coefficients corresponding to the original supplementary set; generating at least one converted metadata based on at least one metadata in the original supplemental set and the conversion coefficient; The converted metadata is processed based on a basic data processing strategy to generate basic processed metadata to construct an original supplementary set after supplementary processing.

2. The method according to claim 1, characterized in that The obtaining of the benchmark data set comprises: Acquire an original data set, wherein the original data set includes at least one metadata; At least one metadata in the original data set is processed based on a basic data processing strategy to generate at least one basic processed metadata to obtain the benchmark data set.

3. The method according to claim 2, characterized in that The basic data processing strategy includes: Performing exception processing on at least one of the obtained metadata to obtain metadata after the exception processing; Smoothing the metadata after the abnormal processing to obtain the smoothed metadata to construct a smoothed data set; The smoothed data set is processed based on a split processing strategy to obtain processed metadata.

4. The method according to claim 3, characterized in that The step of processing the smoothed data set based on the split processing strategy to obtain basic processed metadata includes: Get at least one split indicator; Generate an intermediate timestamp data set based on the splitting indicator and the timestamp data of each metadata in the smoothed data set; Each time stamp data in the smoothed data set is updated to each metadata in the smoothed data set to obtain the basic processed metadata.

5. The method according to claim 1, characterized in that: The preset data set decomposition method is a time series decomposition method STL additive data set decomposition method using robust local weighted regression as a smoothing method.

6. The method according to any one of claims 1 to 5, characterized in that: The calculation strategy includes: Importing the benchmark trend subset into the target trend model to obtain trend target data; Importing the reference period subset into the target period model to obtain period target data; Based on the trend target data and the period target data, target prediction data is obtained.

7. A time series data prediction device, characterized in that: include: An acquisition module, configured to acquire a benchmark data set and at least one supplementary test set; a decomposition module configured to process the metadata in the benchmark data set and the at least one supplementary test set based on a preset data set decomposition method to obtain a benchmark trend subset, a benchmark period subset, at least one supplementary trend subset and at least one supplementary period subset; A trend training module, configured to train an initial trend model through the reference trend subset and the at least one supplementary trend subset to generate a target trend model; A cycle training module, configured to train the initial cycle model through the reference cycle subset and the at least one supplementary cycle subset to generate a target cycle model; A generating module configured to generate target prediction data based on a preset calculation strategy, the benchmark trend subset, the benchmark period subset, the target trend model and the target period model; The acquisition module is further configured to: acquire at least one original supplementary set; process each original supplementary set in the original supplementary set based on a supplementary processing strategy to generate at least one original supplementary set after supplementary processing to obtain the supplementary test set; the supplementary processing strategy includes: acquiring an original supplementary set, wherein the original supplementary set includes at least one metadata; acquiring a conversion coefficient corresponding to the original supplementary set; generating at least one converted metadata based on at least one metadata in the original supplementary set and the conversion coefficient; processing the converted metadata based on a basic data processing strategy to generate basic processed metadata to construct the original supplementary set after supplementary processing.

8. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images