Time series smoothing method and device based on master data, equipment and storage medium
By using a time series smoothing method based on master data, time series data is acquired and repaired, solving the problem of data inconsistency or incompleteness, improving repair accuracy and reducing noise, and is suitable for time series forecasting and anomaly detection.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- TSINGHUA UNIVERSITY
- Filing Date
- 2024-12-04
- Publication Date
- 2026-06-19
AI Technical Summary
In existing technologies, time series data cannot be accurately applied due to inconsistency or incompleteness. Traditional time series recovery techniques have low accuracy and high noise, failing to meet the requirements of consistency and smoothness.
By acquiring the time series data to be repaired and its associated master data, smoothing is performed based on the evaluation data to determine the smoothed series data that meets the consistency and smoothness constraints, and repair is carried out on this basis, selecting the method with the lowest smoothing cost for repair.
It improves the accuracy of time series data restoration, reduces noise, and ensures that the restored data meets the requirements of consistency and smoothness, making it suitable for tasks such as time series forecasting, anomaly detection, and clustering.
Smart Images
Figure CN119829554B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of data processing technology, and in particular to a time series smoothing method, apparatus, device, and storage medium based on master data. Background Technology
[0002] In recent years, the Industrial Internet of Things (IIoT) has developed rapidly, with a large number of sensors being applied in production and daily life scenarios. With the widespread adoption of various smart devices, massive amounts of real-time data are generated across everything from home life to industrial production. This data, presented in time-series format, records various information such as device status, sensor readings, and user behavior, reflecting the changes in an observed entity within a specific time frame. With the increasing prevalence of IoT devices and the continuous improvement in sensor acquisition frequency, the generation rate of time-series data is experiencing explosive growth.
[0003] However, due to a series of reasons such as sensor malfunctions, network latency, and read / write errors, real-world time-series data often contains "dirty" data, meaning it is incomplete or inconsistent. Incompleteness refers to missing values, while inconsistency refers to discrepancies between the collected data and theoretical values. For example, for an engine, its instantaneous torque and instantaneous speed determine its instantaneous fuel consumption rate, but the collected data may be affected by environmental factors, leading to deviations and thus differing from theoretical values. Using and analyzing this data can negatively impact researchers and even ultimately lead to completely erroneous conclusions. Therefore, it is necessary to process the collected time-series data before use.
[0004] In relational data scenarios, master data is commonly used for data processing. Master data refers to consistent and shared business entities within an enterprise, typically used across various business scenarios, serving as the cornerstone for describing core entities. Master data is characterized by its globality, shareability, and stability, enabling sharing and reuse across systems, processes, and departments. In recent years, with the development of Master Data Management (MDM) technology, master data has received increasing attention. For example, e-commerce platforms rely on master data to manage key information related to users and products, and utilize master data to clean and maintain relational data from business transactions. Unfortunately, there is currently a lack of research on master data management in time-series scenarios.
[0005] Traditional time series cleaning methods cannot accurately repair data and may introduce new errors and increase noise, thereby affecting the accuracy of downstream tasks such as time series forecasting, anomaly detection, and clustering. Summary of the Invention
[0006] This invention provides a time series smoothing method, apparatus, device, and storage medium based on master data to address the shortcomings of existing technologies, such as the inability to accurately further apply the collected time series data due to inconsistencies or incompleteness, and the low accuracy and high noise of traditional time series recovery techniques. The invention achieves this by generating smoothed sequence data that satisfies consistency and smoothness constraints at the lowest generation cost, thereby improving the accuracy of the repair process and reducing noise.
[0007] In a first aspect, the present invention provides a time series smoothing method based on master data, comprising the following steps:
[0008] Acquire time series data to be repaired, and acquire master data associated with the time series data to be repaired, wherein the master data includes at least evaluation data for evaluating the time series data to be repaired;
[0009] Based on the evaluation data, the time series data to be repaired is smoothed to determine the smoothed series data that meets the target conditions;
[0010] Based on the smoothed sequence data that meets the target conditions, the time series data to be repaired is repaired to obtain the repaired target time series data;
[0011] The target condition is that the smoothed sequence data satisfies preset constraints and preset cost conditions. The constraints are that the smoothed sequence data simultaneously satisfies consistency constraints and smoothness constraints. The cost condition is that when smoothing the time series data to be repaired, the minimum smoothing cost is determined from multiple smoothing costs.
[0012] Preferably, according to the time series smoothing method based on master data provided by the present invention,
[0013] The consistency constraint is that the smooth sequence data meets the requirements of the master data, and the requirements of the master data include data length requirements, data format requirements, and data time requirements.
[0014] The smoothness constraint is that the time difference between at least two time-adjacent smoothed sequence data is less than or equal to the smoothing window, where the smoothing window is a moving window set when smoothing the time series data to be repaired.
[0015] Preferably, according to the time series smoothing method based on master data provided by the present invention, the step of determining that two smoothed series data are temporally adjacent specifically includes:
[0016] Obtain the timestamp corresponding to each of the smoothed sequence data;
[0017] Calculate the timestamp difference between the timestamp of the current smoothed sequence data and the timestamps of the adjacent smoothed sequence data; the adjacent smoothed sequence data are determined according to the time series order of the smoothed sequence dataset and the timestamp of the current smoothed sequence data; the smoothed sequence dataset is a set composed of multiple smoothed sequence data arranged according to the chronological order of the timestamps of each smoothed sequence data.
[0018] The timestamp difference is compared with a preset time interval threshold.
[0019] If the timestamp difference is less than or equal to the preset time interval threshold, the current smoothed sequence data and the adjacent smoothed sequence data are determined to be time neighbors.
[0020] Preferably, according to the time series smoothing method based on master data provided by the present invention, after the step of comparing the timestamp difference with a preset time interval threshold, the method further includes:
[0021] If the timestamp difference is greater than the preset time interval threshold, it is determined that the current smoothed sequence data is not temporally adjacent to the adjacent smoothed sequence data.
[0022] Preferably, in the time series smoothing method based on master data provided by the present invention, the formula for the smoothing cost is:
[0023]
[0024] In the formula, For time series data to be repaired, For smooth sequence data, This is a distance metric between the time series data to be repaired and the smoothed series data.
[0025] Preferably, in the time series smoothing method based on master data provided by the present invention, the smoothing window includes at least a first time window;
[0026] The step of smoothing the time series data to be repaired based on the evaluation data to determine the smoothed series data that meets the target conditions includes:
[0027] Based on the timestamp of each time series data to be repaired and the smoothing window, it is determined whether each time series data to be repaired is within the first time window;
[0028] If so, each of the time series data to be repaired within the first time window is used as the initial node data for smoothing, and the first smoothed sequence data that meets the target condition is determined.
[0029] If not, based on the mapping relationship between the time series data to be repaired that is not in the first time window and the master data, a candidate sequence set is determined from the evaluation data, and the candidate sequence set is used to smooth the time series data to be repaired that is not in the first time window to generate a second smoothed sequence data that meets the target conditions.
[0030] Based on the first smoothed sequence data and the second smoothed sequence data, smoothed sequence data that meets the target conditions is determined.
[0031] Preferably, according to a time series smoothing method based on master data provided by the present invention, the step of smoothing the time series data to be repaired that is not within the first time window using the candidate sequence set to generate second smoothed sequence data that meets the target condition includes:
[0032] Determine the neighboring windows for each of the time series data to be repaired that is not within the first time window;
[0033] Using the candidate sequence data in the candidate sequence set, a smoothness verification process is performed on each of the time series data to be repaired within the neighboring window. The time series data to be repaired that passes the smoothness verification is then smoothed to generate the second smoothed sequence data.
[0034] Preferably, according to the time series smoothing method based on master data provided by the present invention, the step of smoothing each of the time series data to be repaired within the first time window as initial node data to determine the first smoothed sequence data that meets the target condition includes:
[0035] Never randomly select simulated time series data from the multiple time series data to be repaired within the first time window, wherein the timestamp of the simulated time series data is greater than the timestamp of any of the time series data to be repaired within the first time window;
[0036] The smoothing cost generated when smoothing each of the time series data to be repaired within the first time window is determined as the initial node data and the simulated time series data is determined as the termination node data;
[0037] Determine the initial node data corresponding to the smoothing process that minimizes the smoothing cost;
[0038] The time series data to be repaired corresponding to the initial node data with the lowest smoothing cost is used as the initial node data. Multiple time series data to be repaired within the first time window are smoothed to generate the first smoothed sequence data.
[0039] Secondly, the present invention also provides a time series smoothing device based on master data, comprising the following modules:
[0040] An acquisition module is used to acquire time series data to be repaired and to acquire master data associated with the time series data to be repaired, wherein the master data includes at least evaluation data for evaluating the time series data to be repaired;
[0041] The determination module is used to perform smoothing processing on the time series data to be repaired based on the evaluation data, and determine the smoothed series data that meets the target conditions;
[0042] The repair module is used to repair the time series data to be repaired based on the smoothed sequence data that meets the target conditions, to obtain the repaired target time series data; wherein, the target conditions are that the smoothed sequence data meets preset constraints and preset cost conditions, the constraints are that the smoothed sequence data simultaneously meets consistency constraints and smoothness constraints, and the cost conditions are that the minimum smoothing cost is determined from multiple smoothing costs generated when smoothing the time series data to be repaired.
[0043] Thirdly, the present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the time series smoothing method based on master data as described above.
[0044] Fourthly, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the time series smoothing method based on master data as described above.
[0045] Fifthly, the present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the time series smoothing method based on master data as described above.
[0046] This invention provides a time series smoothing method, apparatus, device, and storage medium based on master data. The method involves acquiring time series data to be repaired and acquiring master data associated with the time series data, wherein the master data includes at least evaluation data for evaluating the time series data to be repaired. Based on the evaluation data, the time series data to be repaired is smoothed to determine smoothed sequence data that meets target conditions. Based on the smoothed sequence data that meets the target conditions, the time series data to be repaired is repaired to obtain repaired target time series data. The target conditions are that the smoothed sequence data meets preset constraints and preset cost conditions. The constraints are that the smoothed sequence data simultaneously meets consistency constraints and smoothness constraints. The cost conditions are that the minimum smoothing cost is determined from multiple smoothing costs generated during the smoothing process of the time series data to be repaired. This invention addresses the shortcomings of existing technologies, such as inconsistent or incomplete time series data, which prevents accurate further application of the acquired time series data, and the low accuracy and high noise of traditional time series recovery techniques. It improves the accuracy of repairing defective time series data and reduces noise. Attached Figure Description
[0047] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.
[0048] Figure 1 This is a flowchart illustrating a time series smoothing method based on master data provided by the present invention.
[0049] Figure 2 This is a schematic diagram of the structure of a time series smoothing device based on master data provided by the present invention.
[0050] Figure 3 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation
[0051] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.
[0052] First, let's analyze some of the terms used in this invention:
[0053] Master data: Master data is core business entity data shared across departments and systems within an enterprise, and it has high requirements for stability and accuracy.
[0054] Time series data: Time series data is data collected at different points in time, used to describe how a phenomenon changes over time. It refers to data collected at different points in time to reflect the state or extent of change of a thing or phenomenon over time. This type of data can be quarterly data, monthly data, etc., and can be further subdivided into various types such as stationary processes and detrended stationary processes.
[0055] Smoothing: Smoothing is a statistical method that uses mathematical algorithms to reduce noise or random fluctuations in raw data, making the data more stable and easier to analyze.
[0056] The main types of smoothing include:
[0057] Simple Moving Average (SMA): Smooths data by calculating the average of consecutive subsets. This method is suitable for time series without obvious trends or seasonality.
[0058] Weighted Moving Average (WMA): Similar to SMA, but gives higher weight to the most recent observation to reflect its greater influence on the current value.
[0059] Exponential smoothing: gives greater weight to recent observations, which decays exponentially over time.
[0060] Kalman filtering: a recursive algorithm used to estimate the invisible portion of observed data in dynamic systems. It is particularly well-suited for handling data containing missing values.
[0061] In the relevant technologies, at least the following technical problems exist:
[0062] Table 1 below shows the vehicle data recorded by the on-board sensors of a certain mixer truck at a sampling frequency of 60 Hz, including latitude and longitude coordinates, engine speed, engine torque, and fuel consumption rate.
[0063] Table 1
[0064]
[0065] However, due to a series of reasons such as sensor failure, network latency, and read / write errors, real-world time-series data often contains dirty data, meaning that the data is incomplete or inconsistent. Incompleteness refers to the presence of missing values in the data, such as some attribute values being empty for data 4 and data 5 in Table 1.
[0066] Inconsistency refers to the difference between the collected data and the theoretical values. For an engine, its instantaneous torque and instantaneous speed determine the instantaneous fuel consumption rate. However, the collected data may be affected by environmental factors, leading to deviations and discrepancies from the theoretical values. For example, in Table 1, data 3 (third row) and data 7 (seventh row) have the same torque and speed, but different fuel consumption rates. Using and analyzing this data may have negative impacts on researchers, potentially leading to completely erroneous conclusions. Therefore, it is necessary to clean the collected time-series data before use.
[0067] In relational data scenarios, master data is commonly used for data cleansing. Master data refers to consistent and shared business entities within an enterprise, typically used extensively in various business scenarios, serving as the cornerstone for describing core entities. Master data is characterized by its globality, shareability, and stability, enabling sharing and reuse across systems, processes, and departments. In recent years, with the development of Master Data Management (MDM) technology, master data has received increasing attention. For example, e-commerce platforms rely on master data to manage key information related to users and products, utilizing master data to clean and maintain relational data related to business transactions.
[0068] However, there is currently no research on master data management in time series scenarios. Traditional average smoothing algorithms ignore important information from the original time series, while exponential smoothing algorithms increase noise and have low accuracy in processing the time series data to be repaired.
[0069] The following is combined Figures 1-3 This invention describes a time series smoothing method, apparatus, device, and storage medium based on master data, which addresses the shortcomings of existing technologies where inconsistent or incomplete time series data prevents accurate further application of the collected time series data, and where traditional time series recovery techniques suffer from low accuracy and high noise. The invention achieves this by generating smoothed sequence data that satisfies consistency and smoothness constraints at the lowest generation cost, thereby improving the accuracy of the repair process and reducing noise.
[0070] Figure 1 This is a flowchart illustrating a time series smoothing method based on master data provided by the present invention, as shown below. Figure 1 As shown, the method may include, but is not limited to, steps S100 to S300:
[0071] S100, acquire the time series data to be repaired, and acquire the master data associated with the time series data to be repaired, wherein the master data includes at least evaluation data for evaluating the time series data to be repaired;
[0072] S200, Based on the evaluation data, the time series data to be repaired is smoothed to determine the smoothed series data that meets the target conditions;
[0073] S300, based on the smoothed sequence data that meets the target conditions, the time series data to be repaired is repaired to obtain the repaired target time series data.
[0074] In step S100 of some embodiments, the time series data to be repaired is obtained, and the master data associated with the time series data to be repaired is obtained.
[0075] It should be noted that the master data includes at least evaluation data for evaluating the time series data to be repaired.
[0076] To more clearly and completely explain the time series smoothing method based on master data provided by this invention, we take vehicle data recorded by the on-board sensors of a mixer truck at a sampling frequency of 60 Hz, including latitude and longitude coordinates, engine speed, engine torque, and fuel consumption rate, as an example.
[0077] For example, the time series data to be repaired includes the 7 rows of data in Table 1 above. Each row of data represents a time series data to be repaired, and each time series data to be repaired corresponds to a time value.
[0078] Acquire the master data associated with the time series data to be repaired. Master data refers to critical business data that is reused in multiple systems and needs to maintain a high degree of consistency. In an embodiment of the present invention, for example, if the time series data to be repaired is collected vehicle time series data, then the master data associated with the vehicle time series data can be internal management data of the car company, as well as some data that can be shared. This shared data is used to guide and evaluate the survival of business data such as vehicle time series data.
[0079] In embodiments of the present invention, the master data may include evaluation data, such as vehicle information, vehicle usage time information, and some standard parameter information of the vehicle, such as vehicle engine information, vehicle spark plug information, etc. Based on these evaluation data information with standard significance, data supplementation or data description can be performed on the time series data to be repaired.
[0080] Master data is core business entity data shared across departments and systems within an enterprise, and it has high requirements for stability and accuracy.
[0081] Master data management is a crucial part of enterprise information management. It involves the standardization, integration, and sharing of data to ensure consistency and accuracy of data across different systems.
[0082] In step S200 of some embodiments, the time series data to be repaired is smoothed based on the evaluation data to determine smoothed series data that meets the target conditions.
[0083] First, it should be noted that the target condition is that the smoothed sequence data satisfies preset constraints and preset cost conditions.
[0084] Furthermore, the constraint condition is that the smooth sequence data simultaneously satisfies the consistency constraint and the smoothness constraint.
[0085] In some embodiments of the present invention, the consistency constraint is that the smooth sequence data meets the requirements of the master data, and the requirements of the master data include data length requirements, data format requirements, and data time requirements.
[0086] The smoothness constraint is that the time difference between at least two time-adjacent smoothed sequence data is less than or equal to the smoothing window, where the smoothing window is a moving window set when smoothing the time series data to be repaired.
[0087] In some embodiments of the present invention, considering that the time series data to be repaired is typically read directly from a database, it is desirable to perform data cleaning efficiently in a streaming manner in these embodiments. The smoothness constraint S considers tuples that are temporally close, with timestamp differences not exceeding [a certain value]. .
[0088] Therefore, for each tuple in the time series data to be repaired, i.e., each piece of time series data to be repaired, we consider a tuple of size . The window.
[0089]
[0090] In the formula, For a length of The time series window, also known as the smoothing window, includes time series from arrive All data points, For window length, Used to describe the definition.
[0091] Furthermore, consistency constraints To ensure that every tuple in any time series data meets the requirements of the master data, the smoothness constraint requires that tuples within a time window be close to each other. Therefore, the tuples of the smooth series data in the expected window are also neighbors in the master data.
[0092] Furthermore, the cost condition is the condition for determining the minimum smoothing cost from among multiple smoothing costs generated when smoothing the time series data to be repaired.
[0093] Define smoothing cost,
[0094] Suppose the time series to be repaired is The formula for smoothing costs is:
[0095]
[0096] In the formula, For time series data to be repaired, For smooth sequence data, This is a distance metric between the time series data to be repaired and the smoothed series data.
[0097] In some embodiments of the present invention, the time series data to be repaired is defined as follows: and master data as By balancing, a certain Smoothed sequence data The determined smoothed sequence data must meet the smoothing cost. Minimum, while simultaneously satisfying consistency constraints and smoothness constraints ,Right now
[0098]
[0099]
[0100]
[0101] In the formula, To smooth out costs, Y is the primary data. For window length, and Used to describe the definition, it can be understood as any two data points in the window. for timestamp, for timestamp, This is the smoothness threshold.
[0102] Furthermore, the smoothness threshold is used to determine whether the distance between two data points satisfies the smoothness constraint. That is, when the distance between two data points is less than or equal to this smoothness threshold, it is determined that the distance between the two data points satisfies the smoothness constraint; when the distance between two data points is greater than this smoothness threshold, it is determined that the distance between the two data points does not satisfy the smoothness constraint.
[0103] Furthermore, it can be understood that the smoothness constraint is that the time difference between at least two time-adjacent smoothed sequence data is less than or equal to the smoothing window, which is a moving window set when smoothing the time series data to be repaired.
[0104] For example, if Table 1 above contains time series data to be repaired, and the moving window value of the smoothing window is 3, then when smoothing the first row of time series data to be repaired in Table 1, the first, second, and third rows of time series data to be repaired are selected as the time series data to be repaired within the first smoothing window.
[0105] When smoothing the time series data to be repaired in the second row of Table 1, the time series data to be repaired in the second, third and fourth rows are selected as the time series data to be repaired in the second smoothing window.
[0106] When smoothing the time series data to be repaired in the third row of Table 1, the time series data to be repaired in the third, fourth and fifth rows are selected as the time series data to be repaired in the third smoothing window.
[0107] Of course, the value of the moving window of the smoothing window can be set arbitrarily according to the needs, or it can be 4. So when smoothing the time series data to be repaired in the third row of Table 1, the time series data to be repaired in the third, fourth, fifth and sixth rows are selected as the time series data to be repaired in the third smoothing window.
[0108] Furthermore, the smoothness constraint requires that the time difference between at least two time-adjacent smoothed sequence data is less than or equal to the smoothing window. The specific steps for determining whether two smoothed sequence data are time-adjacent can be as follows: first, obtain the timestamp corresponding to each smoothed sequence data; then, calculate the timestamp difference between the timestamp of the current smoothed sequence data and the timestamps of adjacent smoothed sequence data.
[0109] The timestamp difference is compared with a preset time interval threshold.
[0110] If the timestamp difference is less than or equal to the preset time interval threshold, the current smoothed sequence data and the adjacent smoothed sequence data are determined to be time neighbors.
[0111] If the timestamp difference is greater than the preset time interval threshold, it is determined that the current smoothed sequence data is not temporally adjacent to the adjacent smoothed sequence data.
[0112] Furthermore, adjacent smoothed sequence data are determined based on the time series order of the smoothed sequence dataset and the timestamp of the current smoothed sequence data.
[0113] A smoothed sequence dataset is a collection of multiple smoothed sequence data arranged in chronological order according to the timestamps of each smoothed sequence data.
[0114] It is understandable that, for example, smoothed sequence data 1 corresponds to timestamp 1 of 10:16:15. The fuel consumption information and engine speed information of smoothed sequence data 1 will not be elaborated here. This example only illustrates the timestamp information of smoothed sequence data to explain the time proximity.
[0115] Smoothed sequence data 2 has a corresponding timestamp of 10:18:15. Smoothed sequence data 3 has a corresponding timestamp of 10:19:15. Smoothed sequence data 4 has a corresponding timestamp of 10:40:15.
[0116] The smoothed sequence dataset consists of smoothed sequence data 1, smoothed sequence data 2, smoothed sequence data 3, and smoothed sequence data 4.
[0117] Assuming the current smoothed sequence data is smoothed sequence data 3, then based on the time series order of the smoothed sequence dataset and the timestamp of the current smoothed sequence data 3, it can be determined that smoothed sequence data 2 and smoothed sequence data 4 are adjacent smoothed sequence data.
[0118] Calculate the timestamp difference of 1 between the timestamp of the current smoothed sequence data 3 and the timestamp of the adjacent smoothed sequence data 2.
[0119] At the same time, the timestamp difference 2 between the timestamp of the current smoothed sequence data 3 and the timestamp of the adjacent smoothed sequence data 4 is calculated.
[0120] Timestamp difference 1 is 1 minute, and timestamp difference 2 is 21 minutes.
[0121] The timestamp difference is compared with a preset time interval threshold.
[0122] Assume the preset time interval threshold is 2 minutes.
[0123] The timestamp difference 1 and timestamp difference 2 are compared with a preset time interval threshold, respectively.
[0124] If the timestamp difference 1 is less than or equal to the preset time interval threshold, the current smoothed sequence data 3 and the adjacent smoothed sequence data 2 are determined to be time neighbors.
[0125] If the timestamp difference 2 is greater than the preset time interval threshold, it is determined that the current smoothed sequence data 3 and the adjacent smoothed sequence data 4 are not temporally adjacent.
[0126] Furthermore, after determining that the current smoothed sequence data 3 and the adjacent smoothed sequence data 2 are time-proximities, it is then determined whether the smoothed sequence data 1 adjacent to the smoothed sequence data 2 is also time-proximities to the current smoothed sequence data 3, until adjacent smoothed sequence data that are not time-proximities are determined.
[0127] In some embodiments of the present invention, the smoothing window includes at least a first time window.
[0128] In some embodiments of the present invention, a first time window
[0129] In the formula, Let X be the first time window, and X be the time series data to be repaired. For window length, and For time intervals less than Two data points.
[0130] Using Table 1 above as an example, the value of the moving window of the smoothing window is 3. Therefore, when smoothing the time series data to be repaired in the first row of Table 1, the time series data to be repaired in the first row, the second row, and the third row are selected as the time series data to be repaired in the first smoothing window.
[0131] The first time window includes the first, second, and third rows of time series data to be repaired.
[0132] The step of smoothing the time series data to be repaired based on the evaluation data to determine the smoothed series data that meets the target conditions includes:
[0133] Based on the timestamp of each time series data to be repaired and the smoothing window, it is determined whether each time series data to be repaired is within the first time window;
[0134] If so, each of the time series data to be repaired within the first time window is used as the initial node data for smoothing, and the first smoothed sequence data that meets the target condition is determined.
[0135] If not, based on the mapping relationship between the time series data to be repaired that is not in the first time window and the master data, a candidate sequence set is determined from the evaluation data, and the candidate sequence set is used to smooth the time series data to be repaired that is not in the first time window to generate a second smoothed sequence data that meets the target conditions.
[0136] Based on the first smoothed sequence data and the second smoothed sequence data, smoothed sequence data that meets the target conditions is determined.
[0137] It is understood that, based on the timestamp of each time series data to be repaired and the smoothing window, it is determined whether each time series data to be repaired is within the first time window. The timestamp of the time series data to be repaired within the first time window is related to the moving window value of the smoothing window. For example, if the moving window value is 4, then there are 4 time series data to be repaired within the first time window. By sorting the multiple time series data to be repaired according to the timestamp of each time series data to be repaired in the order of the timestamps, the first 4 time series data to be repaired within the first time window can be determined.
[0138] For each time series data to be repaired, it is determined whether it is within the first time window.
[0139] If each time series data to be repaired is determined to be within the first time window, each time series data to be repaired within the first time window is used as the initial node data for smoothing processing to determine the first smoothed sequence data that meets the target condition.
[0140] Simulated time series data is randomly selected from multiple time series data to be repaired within the first time window.
[0141] Furthermore, the timestamp of the simulated time series data is greater than the timestamp of any of the time series data to be repaired within the first time window.
[0142] For example, the simulated time series data is the data in the seventh row of Table 1, and the time series data to be repaired within the first time window is the data in the first, second, and third rows of Table 1. Therefore, the timestamp corresponding to the simulated time series data is greater than the timestamps in the first, second, and third rows.
[0143] The smoothing cost is generated when each of the time series data to be repaired within the first time window is determined as the initial node data and the simulated time series data is determined as the termination node data, respectively, and smoothing processing is performed.
[0144] For example, the time series data to be repaired in the first row is the initial node, and the time series data to be repaired in the seventh row is the terminal node data. The smoothing cost is determined to be F1.
[0145] The second row contains the time series data to be repaired, which is the initial node data. The seventh row contains the time series data to be repaired, which is the termination node data. The smoothing cost is determined to be F2.
[0146] The time series data to be repaired in the third row is the initial node, and the time series data to be repaired in the seventh row is the terminal node data. The smoothing cost is determined to be F3.
[0147] By comparing F1, F2, and F3, for example, if F2 is less than F1 and F1 is less than F3, then we can determine that F2 has the lowest smoothing cost.
[0148] Then, the initial node data corresponding to the smoothing process with the minimum smoothing cost F2 is determined, and the time series data to be repaired corresponding to the initial node data with the minimum smoothing cost is used as the initial node data. That is, the time series data to be repaired in the second row is used as the initial node data. The multiple time series data to be repaired within the first time window are smoothed to generate the first smoothed sequence data. This can solve the cold start problem and avoid the first smoothed sequence data within the first time window from generating high noise.
[0149] In some embodiments of the present invention, the step of smoothing the time series data to be repaired that is not within the first time window using the candidate sequence set to generate second smoothed sequence data that meets the target condition includes:
[0150] The process involves identifying neighboring windows for each time series data to be repaired that is not within the first time window, and then acquiring the time series data to be repaired within those nearby windows. Finally, using candidate sequence data from the candidate sequence set, a smoothness verification process is performed on each time series data to be repaired within the neighboring windows. The time series data to be repaired that passes the smoothness verification is then smoothed to generate the second smoothed sequence data. This avoids generating the second smoothed sequence data based on flawed time series data to be repaired, thereby improving the accuracy of the second smoothed sequence data.
[0151] In step S300 of some embodiments, the time series data to be repaired is repaired based on the smoothed sequence data that meets the target conditions to obtain the repaired target time series data.
[0152] It is understandable that after step S200 is executed, the specific execution steps can be as follows: use the smoothed sequence data that meets the target conditions to repair the time series data to be repaired, and obtain the repaired target time series data.
[0153] For example, if the fuel consumption rate information in row 7 and row 3 of Table 1 is inconsistent, based on the smoothed sequence data that meets the target conditions, it can be predicted that the fuel consumption information in row 7 is incorrect. Therefore, the fuel consumption information in row 3 is used to repair and update the information in row 7, and the repaired target time series data is obtained.
[0154] For example, if the fuel consumption rate information in row 4 of Table 1 is missing, the fuel consumption rate information in row 4 can be predicted to be 35.1 based on the smoothed sequence data that meets the target conditions. Then, the missing fuel consumption rate information in row 4 is updated to obtain the corrected target time series data.
[0155] Furthermore, based on the repaired target time series data, time series prediction, anomaly detection, clustering, etc., can be performed to verify the accuracy of the repair. If the repair accuracy is greater than or equal to the preset accuracy threshold, the repair can be considered successful; otherwise, the repair should be performed again.
[0156] This invention provides a time series smoothing method, apparatus, device, and storage medium based on master data. The method involves acquiring time series data to be repaired and acquiring master data associated with the time series data, wherein the master data includes at least evaluation data for evaluating the time series data to be repaired. Based on the evaluation data, the time series data to be repaired is smoothed to determine smoothed sequence data that meets target conditions. Based on the smoothed sequence data that meets the target conditions, the time series data to be repaired is repaired to obtain repaired target time series data. The target conditions are that the smoothed sequence data meets preset constraints and preset cost conditions. The constraints are that the smoothed sequence data simultaneously meets consistency constraints and smoothness constraints. The cost conditions are that the minimum smoothing cost is determined from multiple smoothing costs generated during the smoothing process of the time series data to be repaired. This invention addresses the shortcomings of existing technologies, such as inconsistent or incomplete time series data, which prevents accurate further application of the acquired time series data, and the low accuracy and high noise of traditional time series recovery techniques. It improves the accuracy of repairing defective time series data and reduces noise.
[0157] The time series smoothing device based on master data provided by the present invention is described below. The time series smoothing device based on master data described below and the time series smoothing method based on master data described above can be referred to in correspondence.
[0158] like Figure 2 The diagram shown is a structural schematic of a time series smoothing device based on master data provided by the present invention. The time series smoothing device based on master data includes the following modules:
[0159] The acquisition module 210 is used to acquire time series data to be repaired and to acquire master data associated with the time series data to be repaired, wherein the master data includes at least evaluation data for evaluating the time series data to be repaired.
[0160] The determination module 220 is used to perform smoothing processing on the time series data to be repaired based on the evaluation data, and determine the smoothed series data that meets the target conditions;
[0161] Repair module 230 is used to repair the time series data to be repaired based on the smoothed sequence data that meets the target conditions, to obtain the repaired target time series data; wherein, the target conditions are that the smoothed sequence data meets preset constraints and preset cost conditions, the constraints are that the smoothed sequence data simultaneously meets consistency constraints and smoothness constraints, and the cost conditions are that the minimum smoothing cost is determined from multiple smoothing costs generated when smoothing the time series data to be repaired.
[0162] Preferably, the time series smoothing device based on master data provided by the present invention is further configured such that the consistency constraint is that the smoothed sequence data meets the requirements of the master data, wherein the requirements of the master data include data length requirements, data format requirements, and data time requirements;
[0163] The smoothness constraint is that the time difference between at least two time-adjacent smoothed sequence data is less than or equal to the smoothing window, where the smoothing window is a moving window set when smoothing the time series data to be repaired.
[0164] Preferably, according to the time series smoothing device based on master data provided by the present invention, determining that two smoothed sequence data are time-adjacent is further used to obtain the timestamp corresponding to each of the smoothed sequence data;
[0165] Calculate the timestamp difference between the timestamp of the current smoothed sequence data and the timestamps of the adjacent smoothed sequence data; the adjacent smoothed sequence data are determined according to the time series order of the smoothed sequence dataset and the timestamp of the current smoothed sequence data; the smoothed sequence dataset is a set composed of multiple smoothed sequence data arranged according to the chronological order of the timestamps of each smoothed sequence data.
[0166] The timestamp difference is compared with a preset time interval threshold.
[0167] If the timestamp difference is less than or equal to the preset time interval threshold, the current smoothed sequence data and the adjacent smoothed sequence data are determined to be time neighbors.
[0168] Preferably, the time series smoothing device based on master data provided by the present invention is further configured to determine that the current smoothed sequence data is not temporally adjacent to the adjacent smoothed sequence data when the timestamp difference is greater than the preset time interval threshold.
[0169] Preferably, in the time series smoothing device based on master data provided by the present invention, the formula for the smoothing cost is further as follows:
[0170]
[0171] In the formula, For time series data to be repaired, For smooth sequence data, This is a distance metric between the time series data to be repaired and the smoothed series data.
[0172] Preferably, according to the time series smoothing device based on master data provided by the present invention,
[0173] The smoothing window includes at least a first time window;
[0174] The determining module 220 is further configured to determine whether each of the time series data to be repaired is within the first time window based on the timestamp of each of the time series data to be repaired and the smoothing window;
[0175] If so, each of the time series data to be repaired within the first time window is used as the initial node data for smoothing, and the first smoothed sequence data that meets the target condition is determined.
[0176] If not, based on the mapping relationship between the time series data to be repaired that is not in the first time window and the master data, a candidate sequence set is determined from the evaluation data, and the candidate sequence set is used to smooth the time series data to be repaired that is not in the first time window to generate a second smoothed sequence data that meets the target conditions.
[0177] Based on the first smoothed sequence data and the second smoothed sequence data, smoothed sequence data that meets the target conditions is determined.
[0178] Specifically, the determining module 220 is further configured to determine the neighboring window of each of the time series data to be repaired that is not in the first time window;
[0179] Using the candidate sequence data in the candidate sequence set, a smoothness verification process is performed on each of the time series data to be repaired within the neighboring window. The time series data to be repaired that passes the smoothness verification is then smoothed to generate the second smoothed sequence data.
[0180] The determining module 220 is further configured to randomly determine simulated time series data from among the multiple time series data to be repaired that are not in the first time window, wherein the timestamp of the simulated time series data is greater than the timestamp of any of the time series data to be repaired in the first time window;
[0181] The smoothing cost generated when smoothing each of the time series data to be repaired within the first time window is determined as the initial node data and the simulated time series data is determined as the termination node data;
[0182] Determine the initial node data corresponding to the smoothing process that minimizes the smoothing cost;
[0183] The time series data to be repaired corresponding to the initial node data with the lowest smoothing cost is used as the initial node data. Multiple time series data to be repaired within the first time window are smoothed to generate the first smoothed sequence data.
[0184] This invention provides a time series smoothing method, apparatus, device, and storage medium based on master data. The method involves acquiring time series data to be repaired and acquiring master data associated with the time series data, wherein the master data includes at least evaluation data for evaluating the time series data to be repaired. Based on the evaluation data, the time series data to be repaired is smoothed to determine smoothed sequence data that meets target conditions. Based on the smoothed sequence data that meets the target conditions, the time series data to be repaired is repaired to obtain repaired target time series data. The target conditions are that the smoothed sequence data meets preset constraints and preset cost conditions. The constraints are that the smoothed sequence data simultaneously meets consistency constraints and smoothness constraints. The cost conditions are that the minimum smoothing cost is determined from multiple smoothing costs generated during the smoothing process of the time series data to be repaired. This invention addresses the shortcomings of existing technologies, such as inconsistent or incomplete time series data, which prevents accurate further application of the acquired time series data, and the low accuracy and high noise of traditional time series recovery techniques. It improves the accuracy of repairing defective time series data and reduces noise.
[0185] Figure 3 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 3As shown, the electronic device may include: a processor 310, a communications interface 320, a memory 330, and a communications bus 340, wherein the processor 310, the communications interface 320, and the memory 330 communicate with each other through the communications bus 340. Processor 310 can invoke logic instructions in memory 330 to execute a time series smoothing method based on master data. The method includes: acquiring time series data to be repaired, and acquiring master data associated with the time series data to be repaired, the master data including at least evaluation data for evaluating the time series data to be repaired; smoothing the time series data to be repaired based on the evaluation data to determine smoothed series data that meets target conditions; and repairing the time series data to be repaired based on the smoothed series data that meets the target conditions to obtain repaired target time series data. The target conditions are that the smoothed series data meets preset constraints and preset cost conditions, the constraints are that the smoothed series data simultaneously meets consistency constraints and smoothness constraints, and the cost conditions are that, when smoothing the time series data to be repaired, the minimum smoothing cost is determined from multiple smoothing costs generated.
[0186] Furthermore, the logical instructions in the aforementioned memory 330 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.
[0187] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the time series smoothing method based on master data provided by the above methods. The method includes: acquiring time series data to be repaired, and acquiring master data associated with the time series data to be repaired, wherein the master data includes at least evaluation data for evaluating the time series data to be repaired; smoothing the time series data to be repaired based on the evaluation data to determine smoothed series data that meets the target conditions; and repairing the time series data to be repaired based on the smoothed series data that meets the target conditions to obtain repaired target time series data. The target conditions are that the smoothed series data meets preset constraints and preset cost conditions. The constraints are that the smoothed series data simultaneously meets consistency constraints and smoothness constraints. The cost conditions are that, when smoothing the time series data to be repaired, the minimum smoothing cost is determined from multiple smoothing costs generated.
[0188] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements a time series smoothing method based on master data provided by the methods described above. The method includes: acquiring time series data to be repaired, and acquiring master data associated with the time series data to be repaired, the master data including at least evaluation data for evaluating the time series data to be repaired; smoothing the time series data to be repaired based on the evaluation data to determine smoothed sequence data that meets target conditions; and repairing the time series data to be repaired based on the smoothed sequence data that meets the target conditions to obtain repaired target time series data; wherein the target conditions are that the smoothed sequence data meets preset constraint conditions and preset cost conditions, the constraint conditions are conditions that the smoothed sequence data simultaneously meets consistency constraints and smoothness constraints, and the cost conditions are conditions that determine the minimum smoothing cost from multiple smoothing costs generated during the smoothing process of the time series data to be repaired.
[0189] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.
[0190] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.
[0191] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A time series smoothing method based on principal data, characterized by, Applied to vehicles, this method is used to repair time-series data collected by vehicle sensors that suffers from incomplete or inconsistent data. The method includes: Acquire time series data to be repaired, and acquire master data associated with the time series data to be repaired, wherein the master data includes at least evaluation data for evaluating the time series data to be repaired; the time series data to be repaired is time series data collected by vehicle sensors, including at least one of instantaneous engine torque, instantaneous speed and instantaneous fuel consumption rate; The evaluation data is selected from at least one of the following: vehicle information, vehicle usage time information, vehicle standard parameter information, and theoretical fuel consumption rate defined in the master data corresponding to engine operating conditions. The vehicle standard parameter information includes at least the vehicle's engine information and the vehicle's spark plug information. Based on the evaluation data, the time series data to be repaired is smoothed to determine smoothed series data that meet the target conditions; wherein, the smoothing process of the time series data to be repaired based on the evaluation data to determine smoothed series data that meet the target conditions includes: Based on the timestamp of each time series data to be repaired and a preset smoothing window, it is determined whether each time series data to be repaired is within the first time window; the smoothing window includes at least the first time window; If so, each of the time series data to be repaired within the first time window is used as the initial node data for smoothing, and the first smoothed sequence data that meets the target condition is determined. If not, based on the mapping relationship between the time series data to be repaired that is not in the first time window and the master data, a candidate sequence set is determined from the evaluation data, and the candidate sequence set is used to smooth the time series data to be repaired that is not in the first time window to generate a second smoothed sequence data that meets the target conditions. Based on the first smoothed sequence data and the second smoothed sequence data, determine the smoothed sequence data that satisfies the target condition; Based on the smoothed sequence data that meets the target conditions, the time series data to be repaired is repaired to obtain the repaired target time series data; The target conditions are that the smoothed sequence data meets preset constraints and preset cost conditions. The constraints are that the smoothed sequence data simultaneously meets consistency constraints and smoothness constraints. The consistency constraints are that the smoothed sequence data meets the requirements of the master data, which include data length requirements, data format requirements, and data time requirements. Inconsistency refers to the difference between the engine's collected data and theoretical values, specifically the difference between the collected data and the theoretical values of the instantaneous fuel consumption rate determined by the engine's instantaneous torque and instantaneous speed. The smoothness constraints are used to ensure that the variation in data values collected by the same sensor in smoothed sequence data with adjacent time periods does not exceed the variation range defined by the smoothing window. The cost conditions are the conditions for determining the minimum smoothing cost from multiple smoothing costs generated when smoothing the time series data to be repaired.
2. The time series smoothing method based on master data according to claim 1, characterized in that, The smoothness constraint is that the time difference between at least two time-adjacent smoothed sequence data is less than or equal to the smoothing window, where the smoothing window is a moving window set when smoothing the time series data to be repaired.
3. The time series smoothing method based on master data according to claim 2, characterized in that, The step of determining that two smoothed data sequences are temporally adjacent specifically includes: Obtain the timestamp corresponding to each of the smoothed sequence data; Calculate the timestamp difference between the timestamp of the current smoothed sequence data and the timestamps of the adjacent smoothed sequence data; the adjacent smoothed sequence data are determined according to the time series order of the smoothed sequence dataset and the timestamp of the current smoothed sequence data; the smoothed sequence dataset is a set composed of multiple smoothed sequence data arranged according to the chronological order of the timestamps of each smoothed sequence data. The timestamp difference is compared with a preset time interval threshold. If the timestamp difference is less than or equal to the preset time interval threshold, the current smoothed sequence data and the adjacent smoothed sequence data are determined to be time neighbors.
4. The time series smoothing method based on master data according to claim 3, characterized in that, After the step of comparing the timestamp difference with a preset time interval threshold, the method further includes: If the timestamp difference is greater than the preset time interval threshold, it is determined that the current smoothed sequence data is not temporally adjacent to the adjacent smoothed sequence data.
5. The time series smoothing method based on master data according to claim 1, characterized in that, The formula for the smoothing cost is: ； In the formula, For time series data to be repaired, For smooth sequence data, This is a distance metric between the time series data to be repaired and the smoothed series data.
6. The time series smoothing method based on master data according to claim 3, characterized in that, The step of smoothing the time series data to be repaired that is not within the first time window using the candidate sequence set to generate a second smoothed sequence data that meets the target condition includes: Determine the neighboring windows for each of the time series data to be repaired that is not within the first time window; Using the candidate sequence data in the candidate sequence set, a smoothness verification process is performed on each of the time series data to be repaired within the neighboring window. The time series data to be repaired that passes the smoothness verification is then smoothed to generate the second smoothed sequence data.
7. The time series smoothing method based on master data according to claim 6, characterized in that, The step of smoothing each of the time series data to be repaired within the first time window as initial node data to determine the first smoothed sequence data that meets the target condition includes: Never randomly select simulated time series data from the multiple time series data to be repaired within the first time window, wherein the timestamp of the simulated time series data is greater than the timestamp of any of the time series data to be repaired within the first time window; The smoothing cost generated when smoothing each of the time series data to be repaired within the first time window is determined as the initial node data and the simulated time series data is determined as the termination node data; Determine the initial node data corresponding to the smoothing process that minimizes the smoothing cost; The time series data to be repaired corresponding to the initial node data with the lowest smoothing cost is used as the initial node data. Multiple time series data to be repaired within the first time window are smoothed to generate the first smoothed sequence data.
8. A master data based time series smoothing device applied to the master data based time series smoothing method according to any one of claims 1 to 7, characterized by, Applied to vehicles, this device is used to repair time-series data collected by vehicle sensors that has incomplete or inconsistent data. The device includes: The acquisition module is used to acquire time series data to be repaired and master data associated with the time series data to be repaired. The master data includes at least evaluation data for evaluating the time series data to be repaired. The time series data to be repaired is time series data collected by vehicle sensors, including at least one of instantaneous torque, instantaneous speed, and instantaneous fuel consumption rate of the engine. The evaluation data is selected from at least one of vehicle information, vehicle usage time information, vehicle standard parameter information, and theoretical fuel consumption rate corresponding to engine operating conditions defined in the master data. The vehicle standard parameter information includes at least vehicle engine information and vehicle spark plug information. A determination module is used to smooth the time series data to be repaired based on the evaluation data to determine smoothed sequence data that meets the target conditions. The smoothing process based on the evaluation data to determine smoothed sequence data that meets the target conditions includes: determining whether each time series data to be repaired is within a first time window based on the timestamp of each data point and a preset smoothing window; the smoothing window includes at least the first time window; if so, smoothing each time series data to be repaired within the first time window as initial node data to determine first smoothed sequence data that meets the target conditions; if not, determining a candidate sequence set from the evaluation data based on the mapping relationship between the time series data not within the first time window and the master data, and using the candidate sequence set to smooth the time series data not within the first time window to generate second smoothed sequence data that meets the target conditions; and determining smoothed sequence data that meets the target conditions based on the first smoothed sequence data and the second smoothed sequence data. The repair module is used to repair the time series data to be repaired based on the smoothed sequence data that meets the target conditions, to obtain the repaired target time series data. The target conditions are that the smoothed sequence data meets preset constraints and preset cost conditions. The constraints are that the smoothed sequence data simultaneously meets consistency constraints and smoothness constraints. The consistency constraints are that the smoothed sequence data meets the requirements of the master data, including data length requirements, data format requirements, and data time requirements. Inconsistency refers to a difference between the engine's collected data and theoretical values, specifically a difference between the collected data and the theoretical values of the instantaneous fuel consumption rate determined by the engine's instantaneous torque and instantaneous speed. The smoothness constraints ensure that the variation in data values collected by the same sensor in time-adjacent smoothed sequence data does not exceed the variation range defined by the smoothing window. The cost conditions are the conditions for determining the minimum smoothing cost from multiple smoothing costs generated during the smoothing process of the time series data to be repaired.
9. An electronic device comprising a memory, a processor, and a computer program stored on the memory and running on the processor, characterized in that, When the processor executes the program, it implements the time series smoothing method based on master data as described in any one of claims 1 to 7.
10. A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that, When the computer program is executed by a processor, it implements the time series smoothing method based on master data as described in any one of claims 1 to 7.
11. A computer program product comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the time series smoothing method based on master data as described in any one of claims 1 to 7.