Similar day search method, apparatus, device and computer readable storage medium

Abstract

A similar day search method, device, equipment and computer readable storage medium, comprising: scene label screening is carried out on historical data pool according to target period to be matched, and scene screening data is obtained, target period attribute characteristics of target period to be matched are used for statistical feature matching on scene screening data, and initial matching data is obtained, and dynamic time warping feature matching is carried out on initial matching data according to target period attribute characteristics, and similar day matching data is obtained;Redundancy verification and score disintegration output are carried out on similar day matching data, and similar day search result is obtained.The elastic time alignment capability of dynamic time warping can effectively capture the trend similar but local time dislocation period, improve the accuracy of similarity judgment, and solve the technical problems that the prior art cannot identify the similar day of local time point fluctuation dislocation, resulting in misjudgment.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of electric energy transaction, in particular to a similar day searching method, device and equipment and computer readable storage medium. BACKGROUND

[0002] In the current electric energy transaction auxiliary decision system, similar day is a common function. This function is to mine historical data with similar characteristics to the target period, which can be used to provide reference for transaction strategy making.

[0003] In related technologies, the scheme for searching similar days mostly adopts a fixed period matching mode. That is, when a user sets a target period, the system searches data of the same time period in historical data, and calculates the similarity of each feature through an algorithm such as Euclidean distance, and finally outputs a list of similar periods.

[0004] However, due to the dynamic shift of the operation mode of the power system with factors such as seasonal changes, user habit adjustments and policy updates, for example, the power consumption peak period may shift with the change of seasons. The existing point-to-point similarity calculation such as Euclidean distance has a high requirement for the rigidity of the time axis. Even if the overall trend of two periods is completely consistent, but there is a slight fluctuation misalignment at the local time point, it will also be judged as "not similar", which limits the accuracy of similarity judgment and leads to misjudgment. SUMMARY

[0005] The present application provides a similar day searching method, device, equipment and computer readable storage medium, which can solve the technical problems of existing technologies that are difficult to identify similar days with local time point misalignment and prone to misjudgment.

[0006] In a first aspect, the present application provides a similar day searching method, comprising:

[0007] According to the target period attribute characteristics of the to-be-matched target period, the scene screening data is subjected to statistical feature matching to obtain initial matching data, and according to the target period attribute characteristics, the initial matching data is subjected to dynamic time warping feature matching to obtain similar day matching data;

[0008] The similar day matching data is subjected to redundancy verification and score disassembly output to obtain a similar day searching result.

[0009] In combination with the first aspect, in an implementation mode, according to the to-be-matched target period, the scene label screening data of the historical data pool is obtained, comprising:

[0010] The scene labels of the to-be-matched target period and the historical data pool are constructed;

[0011] According to the scene label, the historical data pool is screened by the scene label to obtain scene screening data;

[0012] The scene label includes a date label, a season label, a day-night type label, and a special event label.

[0013] In combination with the first aspect, in an implementation, according to a target period attribute feature of a target period to be matched, statistical feature matching is performed on the scene screening data to obtain initial matching data, including:

[0014] Mathematical statistics is performed on the target period attribute feature of the target period to be matched and the historical attribute feature of the scene screening data to obtain a target data statistical feature and a historical data statistical feature;

[0015] Euclidean distance similarity calculation is performed on the target data statistical feature and the historical data statistical feature to obtain an initial matching similarity score;

[0016] According to the initial matching similarity score, the scene screening data is screened to obtain the initial matching data.

[0017] In combination with the first aspect, in an implementation, according to the target period attribute feature, dynamic time warping feature matching is performed on the initial matching data to obtain similar day matching data, including:

[0018] Trend consistency filtering is performed on the target period attribute feature and the historical attribute feature of the initial matching data to obtain filtered data;

[0019] Dynamic time warping similarity calculation is performed on the filtered data and the target period attribute feature to obtain a dynamic time warping similarity score;

[0020] According to the dynamic time warping similarity score, the initial matching data is screened to obtain the similar day matching data.

[0021] In combination with the first aspect, in an implementation, trend consistency filtering is performed on the target period attribute feature and the historical attribute feature of the initial matching data to obtain filtered data, including:

[0022] Fast Fourier transform is performed on the target period attribute feature and the historical attribute feature of the initial matching data to obtain a target curve frequency feature and a historical curve frequency feature;

[0023] According to the target curve frequency feature and the historical curve frequency feature, feature screening filtering is performed on the initial matching data to obtain the filtered data.

[0024] In combination with the first aspect, in an implementation, dynamic time warping similarity calculation is performed on the filtered data and the target period attribute feature to obtain a dynamic time warping similarity score, including:

[0025] Calculate the minimum dynamic time warping distance of each filtered data and target period attribute feature under a preset offset granularity;

[0026] Determine the dynamic time warping similarity score of each filtered data according to the minimum dynamic time warping distance.

[0027] In combination with the first aspect, in an implementation, the similar day matching data is subjected to redundancy checking and score disassembly output to obtain similar day search results, including:

[0028] The similar day matching data is subjected to same day redundancy checking to obtain checking data;

[0029] The checking data and the target period attribute feature are subjected to feature-by-feature dynamic time warping similarity calculation to obtain feature matching details and time offset information;

[0030] The similar day search results are obtained according to the feature matching details and the time offset information.

[0031] The second aspect, the embodiments of the present application provide a similar day search device, including:

[0032] The similar day matching unit is configured to perform scene label screening on a historical data pool according to a target period to be matched to obtain scene screening data, perform statistical feature matching on the scene screening data according to a target period attribute feature of the target period to be matched to obtain initial matching data, and perform dynamic time warping feature matching on the initial matching data according to the target period attribute feature to obtain similar day matching data.

[0033] The checking output unit is configured to perform redundancy checking and score disassembly output on the similar day matching data to obtain similar day search results.

[0034] The third aspect, the embodiments of the present application provide a similar day search device, the similar day search device includes a processor, a memory, and a similar day search program stored in the memory and executable by the processor, wherein when the similar day search program is executed by the processor, the steps in the similar day search method of any of the above are implemented.

[0035] The fourth aspect, the embodiments of the present application provide a computer readable storage medium, the computer readable storage medium stores a similar day search program, wherein when the similar day search program is executed by the processor, the steps in the similar day search method of any of the above are implemented.

[0036] The technical scheme provided by the embodiments of the present application has at least the following beneficial effects: through the elastic time alignment capability of dynamic time warping, the trend similar but locally time misaligned periods can be effectively captured, the accuracy of similarity judgment is improved, and the technical problem that the prior art cannot identify similar days with local time point fluctuation misalignment and leads to misjudgment is solved. BRIEF DESCRIPTION OF DRAWINGS

[0037] Figure 1 Flowchart of an embodiment of the similar date search method of the present application;

[0038] Figure 2 Flowchart of the refinement of the scene label screening step of the present application;

[0039] Figure 3 Flowchart of the refinement of the statistical feature matching step of the present application;

[0040] Figure 4 Flowchart of the refinement of the dynamic time warping feature matching of the present application;

[0041] Figure 5 Flowchart of the refinement of step S401 in the similar date search method of the present application; Figure 4

[0042] Figure 6 Flowchart of the refinement of step S402 in the similar date search method of the present application; Figure 4

[0043] Figure 7 Flowchart of the refinement of step S102 in the similar date search method of the present application; Figure 1

[0044] Flowchart of the functional modules of an embodiment of the similar date search device of the present application; Figure 8

[0045] Hardware structure of the similar date search device involved in the embodiment of the present application. Figure 9 DETAILED DESCRIPTION

[0046] In order to enable persons skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by persons skilled in the art without creative labor fall within the scope of protection of the present application.

[0047] ​​​The terms "include," "includes" or "comprise," "comprises" or "comprising" and any variations thereof in the Specification and in the Claims shall not be construed as being restrictive, but shall be given the widest possible interpretation. For example, a process, method, system, product, or apparatus that comprises a list of steps or units shall not be limited to the listed steps or units, but can also include additional steps or units not expressly listed or inherent to such process, method, system, product, or apparatus. The terms "first," "second," and "third," etc. are used to distinguish different objects, etc., and do not represent a chronological order or a limitation of different types.

[0048] In the description of the embodiments of the present application, "exemplary", "for example", "e.g." or "for instance" are used on the basis that a proper meaning of the related concept is presented in a specific manner.

[0049] In the description of the embodiments of the present application, unless otherwise specified, " / " means or, for example, A / B can mean A or B; "and / or" in the text only describes the relationship of the associated objects, which means that there can be three relationships, for example, A and / or B, which means that there are three cases of A alone, A and B together, and B alone. In addition, in the description of the embodiments of the present application, "multiple" means two or more than two.

[0050] In some of the processes described in the embodiments of the present application, a plurality of operations or steps appear in a specific order, but it should be understood that these operations or steps can be executed or performed in parallel or in a different order from that in which they appear in the embodiments of the present application. The serial number of the operation is only used to distinguish different operations, and the serial number itself does not represent any execution order. In addition, these processes can include more or fewer operations, and these operations or steps can be executed in sequence or in parallel, and these operations or steps can be combined.

[0051] In order to make the purpose, technical scheme and advantages of the present application clearer, the embodiments of the present application will be further described in detail below with reference to the accompanying drawings.

[0052] In a first aspect, the embodiments of the present application provide a similar day search method.

[0053] In an embodiment, the similar day search method comprises the following steps. Figure 1 , Figure 1 The flowchart of an embodiment of the similar day search method of the present application is shown in FIG. 1. As shown in FIG. 1, the similar day search method comprises the following steps. Figure 1 In an embodiment, the similar day search method comprises the following steps.

[0054] S101, scene label screening is performed on the historical data pool according to a target period to be matched to obtain scene screening data, statistical feature matching is performed on the scene screening data according to a target period attribute feature of the target period to be matched to obtain initial matching data, and dynamic time warping feature matching is performed on the initial matching data according to the target period attribute feature to obtain similar day matching data.

[0055] In order to improve search accuracy while ensuring search efficiency, an embodiment sets a three-layer progressive filtering mode. The first layer is pre-filtering through scene labels, such as filtering and screening historical data through scene labels such as season, date type, and day and night attribute. The second layer uses statistical values of part of the attribute features for screening, which can effectively simplify the calculation amount, quickly filter the candidate data, and narrow the candidate range. The third layer is dynamic time warping feature matching, which calculates the similarity score under the premise of allowing the feature points to have a certain shift on the time axis, and sorts and screens the similar day matching data according to the similarity score.

[0056] S102, redundancy verification and score decomposition are performed on the similar day matching data to obtain similar day search results.

[0057] After the similar day matching is completed, redundancy verification and similarity score decomposition need to be performed. The purpose of redundancy verification is to remove multiple results from the same date, and the similarity score decomposition is to analyze the matching degree of each feature dimension, generate a matching detail of each feature, and obtain similar day search results that can be visualized and displayed specific similarity analysis content.

[0058] In this embodiment, the elastic time alignment capability of dynamic time warping can effectively capture similar periods with local time misalignment, improve the accuracy of similarity judgment, and solve the technical problem of misjudgment in related technologies due to the dynamic shift of the operation mode of the power system. The existing technology cannot identify similar days with local time point fluctuation misalignment, and is prone to misjudgment.

[0059] Further, in an embodiment, with reference to Figure 2 , Figure 2 is a detailed process diagram of the scene label screening step of the present application, which includes:

[0060] S201, scene labels of a target period to be matched and a historical data pool are constructed.

[0061] S202, scene label screening is performed on the historical data pool according to the scene labels to obtain scene screening data.

[0062] The scene labels include date labels, season labels, day and night type labels, and special event labels.

[0063] Specifically, according to the target date period and the corresponding relevant attribute features of the similar day search required by the user input, the embodiment first generates the scene label of the target period. The scene label includes but is not limited to, for example, date type, including weekdays, weekends and holidays; seasonal type, including spring, summer, autumn and winter; day and night type, including early morning, morning, afternoon and night; special event, including maintenance and extreme weather. In addition, for the convenience of calculation, for the relevant attribute features input by the user, the normalization processing is first carried out, and all the input attribute features to be considered are normalized to the interval [0, 1].

[0064] In the scene label screening process, the embodiment can screen the historical database according to the set label screening type input by the user. Among them, part of the important scene label is used as the default label, and part is used as the optional label, which is selected by the user. For example, the embodiment first screens and retains the historical data consistent with the "date type" and "seasonal type" of the target period input by the user. Then, for the remaining scene labels, the user can configure whether to screen and retain the historical data consistent with the other scene labels of the target period input by the user.

[0065] In addition, in some embodiments, it can also be configured to select the historical data of how many time ranges to retain in order to avoid large changes in power system characteristics and market rule changes over a long period of time.

[0066] In this embodiment, by replacing the fixed period screening with the scene label screening, the similar day missing detection caused by the dynamic offset of the power system in the screening process can be effectively avoided.

[0067] Further, in an embodiment, with reference to Figure 3 , Figure 3 The detailed process diagram of the statistical feature matching step of the present application includes:

[0068] S301, the mathematical statistics of the target period attribute features of the target period to be matched and the historical attribute features of the scene screening data are carried out to obtain the target data statistical features and the historical data statistical features;

[0069] S302, the Euclidean distance similarity calculation of the target data statistical features and the historical data statistical features is carried out to obtain the initial matching similarity score;

[0070] S303, the scene screening data is screened according to the initial matching similarity score to obtain the initial matching data.

[0071] Specifically, the statistical features used for statistical feature matching in this embodiment include the mean, peak value, and peak position of each attribute feature. Instead of using the entire sequence, this embodiment uses statistical measures calculated from some attribute features as simplified features to perform preliminary matching between the target time period and historical data, effectively improving computational efficiency and achieving coarse matching of historical data.

[0072] In the example, the statistical feature matching process is based on the Euclidean distance algorithm to calculate the similarity between the target time period and each historical time period in terms of simplified features, and sorts them according to the similarity score, retaining the first few historical time periods with the highest scores to obtain the initial matching data.

[0073] In this embodiment, statistical feature matching by simplifying features can effectively improve computational efficiency and increase the response speed of similar day search.

[0074] Furthermore, in one embodiment, reference is made to Figure 4 , Figure 4 This is a detailed flowchart illustrating the dynamic time warping feature matching process of this application, including:

[0075] S401. Perform trend consistency filtering on the target time period attribute features and the historical attribute features of the initial matching data to obtain filtered data;

[0076] S402. Calculate the dynamic time warping similarity between the filtered data and the target time period attribute features to obtain the dynamic time warping similarity score.

[0077] S403. Based on the dynamic time warping similarity score, the initial matching data is filtered to obtain similar daily matching data.

[0078] Specifically, after completing label screening and simplified feature coarse screening, the embodiment further filters through dynamic time warping feature matching to achieve accurate matching of the full feature vector.

[0079] The dynamic time warping feature matching process includes three steps: trend consistency filtering, dynamic time warping similarity calculation, and similarity matching filtering. Trend consistency filtering quickly filters out dissimilar time periods by performing curve feature analysis on each feature. Dynamic time warping similarity calculation calculates feature similarity while allowing for time offsets, effectively identifying time periods in historical data where feature trends are similar but local time points are misaligned, thus avoiding missed detections in similar day matching.

[0080] Furthermore, in one embodiment, reference is made to Figure 5 , Figure 5 For this application Figure 4 A detailed flowchart of step S401 is shown below, including:

[0081] S501, performing fast Fourier transform on the target period attribute feature and the historical attribute feature of the initial matching data to obtain target curve frequency features and historical curve frequency features;

[0082] S502, performing feature screening and filtering on the initial matching data according to the target curve frequency features and the historical curve frequency features to obtain filtered data.

[0083] For the initial matching data, the embodiment first filters according to the trend consistency of each data with the attribute features of the target period. For this purpose, the embodiment uses fast Fourier transform to extract the curve frequency features of each feature in the target period and the historical period, and filters according to the obtained curve frequency features. For example, the overall rising and falling trend of the load in the target period and the historical period is calculated, and if the trend is the same, it is retained, and if the trend is opposite, the corresponding data is removed.

[0084] In this embodiment, the trend consistency filtering supplements the previous rough matching process, which can effectively remove periods that are similar in statistics but not similar in feature change trend, narrow the matching range, and improve the matching accuracy.

[0085] Further, in an embodiment, with reference to Figure 6 , Figure 6 is a detailed flowchart of step S402 in the present application Figure 4 , including:

[0086] S601, calculating the minimum dynamic time warping distance of each filtered data and the attribute features of the target period under a preset offset granularity;

[0087] S602, determining the dynamic time warping similarity score of each filtered data according to the minimum dynamic time warping distance.

[0088] Specifically, for the periods remaining after multiple screenings, in order to screen out similar days with similar feature trends but local time point misalignment, the embodiment uses a dynamic time warping similarity calculation method, that is, allows the feature points to have a ±1 granularity offset on the time axis. For example, if the granularity is set to 15 minutes, the peak value can be advanced or delayed by 15 minutes. And calculate the dynamic time warping distance of the feature attribute data of the historical period and the feature attribute data of the target period under the preset offset granularity as the similarity score, that is, the minimum distance between the two sets of data within the allowed offset range, and determine the similarity score based on this. According to the score, retain a certain number of periods with high ranking.

[0089] In this embodiment, through the elastic time alignment capability of dynamic time warping, periods with similar trends but local time misalignment can be effectively captured, improving the accuracy of similarity judgment and avoiding misjudgment.

[0090] Further, in an embodiment, referring to Figure 7 , Figure 7 is a detailed flowchart of step S102 in the present application Figure 1 , including:

[0091] S701, same-day redundancy verification is performed on the similar day matching data to obtain verification data;

[0092] S702, feature-by-feature dynamic time warping similarity calculation is performed on the verification data and the target period attribute features to obtain feature matching details and time offset information;

[0093] S703, the similar day search result is obtained according to the feature matching details and the time offset information.

[0094] Specifically, in the final output result stage, the embodiment first performs same-day redundancy verification on the obtained similar day matching data, that is, if multiple candidate periods come from the same date, such as different two periods of the same historical day, only the one with the highest score is retained to avoid information redundancy.

[0095] Then similarity score decomposition also needs to be performed, that is, the matching degree of each specific feature, such as load matching degree and electricity price matching degree, is calculated one by one, and according to the obtained result, feature matching details and corresponding time offset information are generated. After sorting, visual display is performed.

[0096] In the present embodiment, same-day redundancy verification can effectively remove redundancy, and feature-by-feature dynamic time warping similarity calculation can split and display the matching degree of a single feature, to provide more detailed and visual results for users.

[0097] In a second aspect, the present embodiment also provides a similar day search device.

[0098] In an embodiment, referring to Figure 8 , Figure 8 is a functional module schematic diagram of an embodiment of the similar day search device of the present application. As shown in Figure 8 , the similar day search device includes:

[0099] A similar day matching unit 801 is configured to perform scene label screening on a historical data pool according to a target period to be matched to obtain scene screening data, perform statistical feature matching on the scene screening data according to a target period attribute feature of the target period to be matched to obtain initial matching data, and perform dynamic time warping feature matching on the initial matching data according to the target period attribute feature to obtain similar day matching data.

[0100] A verification and output unit 802 is configured to perform redundancy verification and score decomposition output on the similar day matching data to obtain a similar day search result.

[0101] The functions of each module in the similar day search device correspond to the steps in the similar day search method, and the functions and implementation processes are not repeated here.

[0102] In a third aspect, the embodiments of the present application provide a similar day search device. The similar day search device can be a personal computer (PC), a notebook computer, a server, or the like.

[0103] Referring to Figure 9 , Figure 9 FIG. 1 is a schematic diagram of a hardware structure of a similar day search device according to an embodiment of the present application. In the embodiments of the present application, the similar day search device can include a processor, a memory, a communication interface, and a communication bus.

[0104] The communication bus can be of any type, and is used to interconnect the processor, the memory, and the communication interface.

[0105] The communication interface includes an input / output (I / O) interface, a physical interface, and a logical interface, and the like, which are used to interconnect devices inside the similar day search device, and are used to interconnect the similar day search device with other devices (for example, other computing devices or user devices). The physical interface can be an Ethernet interface, a fiber interface, an ATM interface, or the like; and the user device can be a display (Display), a keyboard (Keyboard), or the like.

[0106] The memory can be various types of storage media, such as random access memory (RAM), read-only memory (ROM), non-volatile RAM (NVRAM), flash memory, optical storage, hard disk, programmable ROM (PROM), erasable PROM (EPROM), electrically erasable PROM (EEPROM), or the like.

[0107] The processor can be a general-purpose processor, which can invoke a similar day search program stored in the memory and execute the similar day search method provided by the embodiments of the present application. For example, the general-purpose processor can be a central processing unit (CPU). The method executed by the similar day search program when invoked can refer to the embodiments of the similar day search method of the present application, and is not repeated here.

[0108] Those skilled in the art can understand that Figure 9 The hardware structure shown in the foregoing embodiments is not a limitation on the present application, and can include more or less components than shown, or combine certain components, or arrange different components.

[0109] In a fourth aspect, the embodiments of the present application further provide a readable storage medium.

[0110] The readable storage medium of the present application stores a similar day search program, and when the similar day search program is executed by a processor, the steps of the similar day search method are implemented.

[0111] The method implemented when the similar day search program is executed can refer to each embodiment of the similar day search method of the present application, and will not be described here.

[0112] It should be noted that the serial numbers of the embodiments of the present application are only for description, and do not represent the advantages and disadvantages of the embodiments.

[0113] From the above description of the embodiments, those skilled in the art can clearly understand that the above-mentioned embodiment method can be realized by means of software and the necessary general hardware platform, of course, it can also be realized by hardware, but in many cases the former is a better embodiment. Based on such understanding, the technical solutions of the present application can be embodied in the form of a software product, which is stored in a storage medium (such as a ROM / RAM, a magnetic disk, an optical disk) and includes a plurality of instructions for causing a terminal device to execute the methods of the embodiments of the present application.

[0114] The above is only the preferred embodiment of the present application, and does not limit the patent scope of the present application, and any equivalent structure or equivalent flow transformation using the content of the specification and drawings, or direct or indirect application in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A similar day search method characterized by, The method comprises the following steps: According to the target period attribute characteristics of the target period to be matched, the statistical characteristics of the scene screening data are matched to obtain initial matching data, and according to the target period attribute characteristics, dynamic time warping feature matching is performed on the initial matching data to obtain similar day matching data; The similar day matching data is subjected to redundancy verification and score disassembly output to obtain similar day search results; According to the target period attribute characteristics, dynamic time warping feature matching is performed on the initial matching data to obtain similar day matching data, which comprises the following steps: The target curve frequency characteristics and the historical curve frequency characteristics are obtained by performing fast Fourier transform on the target period attribute characteristics and the historical attribute characteristics of the initial matching data, the feature screening filtering of the initial matching data is performed according to the target curve frequency characteristics and the historical curve frequency characteristics to obtain filtered data, the minimum dynamic time warping distance of each filtered data and the target period attribute characteristics is calculated under a preset offset granularity, the dynamic time warping similarity score of each filtered data is determined according to the minimum dynamic time warping distance, and the initial matching data is screened according to the dynamic time warping similarity score to obtain similar day matching data.

2. The similar day search method of claim 1, wherein, The scene label screening data is obtained by screening the historical data pool according to the target period to be matched, which comprises the following steps: The scene labels of the target period to be matched and the historical data pool are constructed; The scene label screening data is obtained by screening the historical data pool according to the scene labels; The scene labels comprise date labels, season labels, day and night type labels and special event labels.

3. The similar day search method of claim 1, wherein, The scene label screening data is obtained by screening the historical data pool according to the target period attribute characteristics of the target period to be matched, which comprises the following steps: The target data statistical characteristics and the historical data statistical characteristics are obtained by performing mathematical statistics on the target period attribute characteristics of the target period to be matched and the historical attribute characteristics of the scene screening data; The initial matching similarity score is obtained by performing Euclidean distance similarity calculation on the target data statistical characteristics and the historical data statistical characteristics; The initial matching data is obtained by screening the scene screening data according to the initial matching similarity score.

4. The similar day search method of claim 1, wherein, The similar day search results are obtained by performing redundancy verification and score disassembly output on the similar day matching data, which comprises the following steps: The verification data is obtained by performing same-day redundancy verification on the similar day matching data; The feature matching details and the time offset information are obtained by performing feature-by-feature dynamic time warping similarity calculation on the verification data and the target period attribute characteristics; The similar day search results are obtained according to the feature matching details and the time offset information.

5. A similar day search apparatus characterized by comprising: The method comprises the following steps: The similar day matching unit is configured to perform scene tag screening on the historical data pool according to a target time period to be matched to obtain scene screening data, perform statistical feature matching on the scene screening data according to a target time period attribute feature of the target time period to be matched to obtain initial matching data, and perform dynamic time warping feature matching on the initial matching data according to the target time period attribute feature to obtain similar day matching data. The verification output unit is configured to perform redundancy verification and score disassembly output on the similar day matching data to obtain similar day search results. The dynamic time warping feature matching on the initial matching data according to the target time period attribute feature to obtain similar day matching data includes: performing fast Fourier transform on the target time period attribute feature and historical attribute features of the initial matching data to obtain target curve frequency features and historical curve frequency features, performing feature screening filtering on the initial matching data according to the target curve frequency features and the historical curve frequency features to obtain filtered data, calculating minimum dynamic time warping distances of each of the filtered data and the target time period attribute feature under a preset offset granularity, determining dynamic time warping similarity scores of each of the filtered data according to the minimum dynamic time warping distances, and performing screening on the initial matching data according to the dynamic time warping similarity scores to obtain similar day matching data.

6. A similar day search apparatus characterized by comprising: The similar day search device includes a processor, a memory, and a similar day search program stored on the memory and executable by the processor, and when the similar day search program is executed by the processor, the steps in the similar day search method according to any one of claims 1 to 4 are implemented.

7. A computer readable storage medium characterized in that, The computer readable storage medium stores a similar day search program, and when the similar day search program is executed by the processor, the steps in the similar day search method according to any one of claims 1 to 4 are implemented.

Images