Defect identification and positioning method, device and equipment for time series data and medium
Patent Information
- Application Number
- CN202411741239.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-29
- Publication Date
- 2025-03-25
AI Technical Summary
The prior art is difficult to subdividing the types of defects in timing data, accurately locate the location of defects and evaluate the scope of their impact, making it difficult to conduct detailed quality analysis and traceability analysis of subsequent process flows in steel rolling production.
By obtaining the time sequence data to be identified and the time sequence data of defect characteristics, sliding intercepting the time sequence data to be identified using the preset span interval to obtain multiple sub-sequence data, and calculate the distance matrix to determine the minimum cumulative distance and the corresponding optimal alignment path according to the distance between the time sequence data of defect characteristics and the sub-sequence data, so as to update the result dictionary and perform defect interval filtering to obtain the target defect interval and type.
The efficiency and accuracy of time-series data defect identification and positioning are improved, the defect interval can be positioned more flexibly, and the overall computing efficiency of the distance matrix is improved through the preset span interval.
Smart Images
Figure CN119691500A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of defect detection technology, and in particular to a defect identification and positioning method, device, equipment and medium for time series data. Background Art
[0002] In the steel rolling production process, time series data plays a vital role. Time series data is not only generated in large quantities, but also closely monitored to ensure the smoothness of the production process and the stability of product quality. Timely discovery and accurate location of defects in these time series data are of immeasurable value in improving product quality, reducing scrap rates, and maintaining the stable operation of the production line.
[0003] Current time series data defect discovery mainly relies on threshold rules, that is, defect detection is achieved by comparing the quantitative relationship between specific statistical values in time series data and preset thresholds. For example, by comparing the maximum / minimum value with the set upper / lower limit value, or evaluating the difference between the overall mean and the standard value, these methods have clear logic and show good robustness. However, the degree of defect recognition is relatively limited. In most cases, this method can only determine whether a defect has occurred, but it is difficult to further subdivide the type of defect, accurately locate the location of the defect, and evaluate its impact range. This limitation poses a considerable challenge to the analysis of quality problems such as sickle bend recognition and plate shape recognition in steel rolling production, as well as traceability analysis of subsequent process flows such as roller rolling force analysis and looper tension analysis. Summary of the invention
[0004] The present invention provides a defect identification and positioning method, device, equipment and medium for time series data to solve the above-mentioned technical problems of being unable to subdivide the types of defects, accurately locate the occurrence locations of defects and evaluate their impact range.
[0005] In one embodiment of the present application, the present application provides a defect identification and positioning method for time series data, including: obtaining time series data to be identified and defect feature time series data; slidingly intercepting the time series data to be identified according to a preset span interval to obtain multiple sub-sequence data, and obtaining a distance matrix corresponding to each sub-sequence data according to the distance between the defect feature time series data and each sub-sequence data; if the minimum cumulative distance in a distance matrix is less than a preset distance threshold, and the distance column index of the minimum cumulative distance is greater than the lower limit of the preset span interval, then updating the result dictionary according to the sliding interception position corresponding to the sub-sequence data, the minimum cumulative distance and the distance column index, and the minimum cumulative distance is located in the last row of the distance matrix; performing defect interval screening on the sub-sequence data corresponding to the result dictionary to obtain a target defect interval and a target defect type.
[0006] In one embodiment of the present application, the result dictionary is updated according to the sliding interception position corresponding to the subsequence data, the minimum cumulative distance and the distance column index, including: using the interception starting point in the sliding interception position as the result row index of the result dictionary, and the interception starting point is used to represent the interception start position corresponding to the subsequence data in the time series data to be identified; determining the sum of the distance column index and the result row index as the result column index of the result dictionary; storing the minimum cumulative distance in the result dictionary according to the result row index and the result column index.
[0007] In one embodiment of the present application, the time series data to be identified is slidingly intercepted according to a preset span interval to obtain multiple sub-sequence data, including: determining the minimum index value of the time series data to be identified as the interception starting point; determining the smallest value between the index value to be compared and the maximum index value of the time series data to be identified as the interception end point, and the index value to be compared is obtained based on the sum of the interception starting point and the upper limit of the preset span interval; intercepting the time series data to be identified according to the interception starting point and the interception end point to obtain sub-sequence data, and increasing the value of the interception starting point to obtain a new interception starting point; if the new interception starting point is greater than the threshold to be intercepted , then stop intercepting data of the time series data to be identified, and the threshold to be intercepted is obtained based on the difference between the maximum index value of the time series data to be identified and the lower limit of the preset span interval; if the new interception starting point is less than or equal to the threshold to be intercepted, then repeat the steps of determining the minimum value between the new index value to be compared and the maximum index value of the time series data to be identified as the new interception end point, intercepting the time series data to be identified according to the new interception starting point and the new interception end point to obtain new subsequence data, and increasing the value of the new interception starting point to obtain a new interception starting point, until the new interception starting point is greater than the length to be intercepted.
[0008] In one embodiment of the present application, a distance matrix corresponding to each subsequence data is obtained according to the distance between the defect feature time series data and each subsequence data, including: normalizing the defect feature time series data and the subsequence data respectively; according to the distance between the defect feature time series data and each subsequence data, and according to the distance between the normalized defect feature time series data and the normalized subsequence data, an alignment path is calculated to obtain the distance matrix corresponding to the subsequence data, so as to determine the optimal alignment path corresponding to the minimum cumulative distance.
[0009] In one embodiment of the present application, before obtaining the time series data to be identified, it also includes: obtaining initial time series data; performing mathematical statistics on the initial time series data to obtain an initial statistical value; if the initial statistical value meets the defect existence condition corresponding to a preset statistical threshold, the initial time series data is determined as the time series data to be identified.
[0010] In one embodiment of the present application, defect interval screening is performed on the subsequence data corresponding to the result dictionary to obtain a target defect interval and a target defect type, including: determining the subsequence data corresponding to the result dictionary as at least one candidate defect interval; performing mathematical statistics on the at least one candidate defect interval to obtain at least one candidate statistical value; if a candidate statistical value satisfies a defect existence condition corresponding to a preset statistical threshold, determining the candidate defect interval as a target defect interval, and determining the defect feature type corresponding to the defect feature time series data as a target defect type.
[0011] In one embodiment of the present application, the mathematical statistics include extreme value statistics, mean and standard deviation statistics; extreme value statistics include maximum value and minimum value.
[0012] In one embodiment of the present application, the present application provides a defect identification and positioning device for time series data, including: a data acquisition module, used to acquire time series data to be identified and defect feature time series data; a multi-sequence regularization module, used to slide and intercept the time series data to be identified according to a preset span interval to obtain multiple sub-sequence data, and obtain a distance matrix corresponding to each sub-sequence data according to the distance between the defect feature time series data and each sub-sequence data; a defect positioning module, used to update a result dictionary according to the sliding interception position corresponding to the sub-sequence data, the minimum cumulative distance and the distance column index if the minimum cumulative distance in a distance matrix is less than a preset distance threshold and the distance column index of the minimum cumulative distance is greater than the lower limit of the preset span interval, and the minimum cumulative distance is located in the last row of the distance matrix; a defect screening module, used to screen the defect interval of the sub-sequence data corresponding to the result dictionary to obtain a target defect interval and a target defect type.
[0013] In one embodiment of the present application, the present application provides an electronic device, which includes: one or more processors; a storage device for storing one or more programs, and when the one or more programs are executed by the one or more processors, the electronic device implements the defect identification and positioning method of timing data as described in any of the above embodiments.
[0014] In one embodiment of the present application, the present application provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor of a computer, the computer executes the defect identification and positioning method of the timing data described in any of the above embodiments.
[0015] Beneficial effects of embodiments of the present invention: The present invention provides a defect identification and positioning method, device, equipment and medium for time series data. The embodiments of the present invention obtain multiple sub-sequence data by slidingly intercepting the time series data to be identified through a preset span interval, and obtains a distance matrix corresponding to each sub-sequence data according to the distance between the defect feature time series data and each sub-sequence data. The flexibility of defect interval positioning can be improved by slidingly intercepting the time series data to be identified, and the overall computational efficiency of the distance matrix can be improved by presetting the span interval, thereby improving the defect identification and positioning efficiency and accuracy of the time series data to be identified.
[0016] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention. BRIEF DESCRIPTION OF THE DRAWINGS
[0017] The drawings herein are incorporated into the specification and constitute a part of the specification, showing embodiments consistent with the present application, and together with the specification, are used to explain the principles of the present application. Obviously, the drawings described below are only some embodiments of the present application, and for those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative work. In the drawings:
[0018] Figure 1 A schematic diagram showing a flow chart of a defect identification and location method of time series data according to an embodiment of the present application is shown;
[0019] Figure 2 A schematic diagram of a DTW algorithm according to an embodiment of the present application is shown;
[0020] Figure 3 A schematic diagram of an implementation flow of a defect identification and location method of time series data according to an embodiment of the present application is shown;
[0021] Figure 4 A block diagram of a device for defect identification and location of time series data according to an embodiment of the present application is shown;
[0022] Figure 5 A schematic diagram of the structure of a computer system suitable for implementing an electronic device of an embodiment of the present application is shown. DETAILED DESCRIPTION
[0023] The following describes the embodiments of the present application through specific examples, and those skilled in the art can easily understand other advantages and effects of the present application from the contents disclosed in this specification. The present application can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present application. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict.
[0024] It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present application, and thus the drawings only show components related to the present application rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed at will, and the component layout may also be more complicated.
[0025] In the following description, a large number of details are discussed to provide a more thorough explanation of the embodiments of the present application. However, it is obvious to those skilled in the art that the embodiments of the present application can be implemented without these specific details. In other embodiments, well-known structures and devices are shown in the form of block diagrams rather than in detail to avoid making the embodiments of the present application difficult to understand.
[0026] In the related technology, there is a technical problem that it is impossible to subdivide the types of defects, accurately locate the locations of defects, and evaluate their impact range.
[0027] In order to solve the above technical problems, the present application provides a defect identification and positioning method, device, equipment and medium for time series data. The implementation details of the technical solution of the embodiment of the present application are elaborated in detail below.
[0028] See also Figure 1 , Figure 1 FIG. 1 is a flow chart showing a defect identification and location method of time series data according to an embodiment of the present application. Figure 1 As shown, in an exemplary embodiment, the defect identification and positioning method of time series data includes at least steps S110 to S140, which are described in detail as follows:
[0029] Step S110, obtaining the time series data to be identified and the defect feature time series data.
[0030] In one embodiment of the present application, before obtaining the time series data to be identified, it also includes: obtaining initial time series data; performing mathematical statistics on the initial time series data to obtain an initial statistical value; if the initial statistical value meets the defect existence condition corresponding to a preset statistical threshold, the initial time series data is determined as the time series data to be identified.
[0031] In one embodiment of the present application, the time series data to be identified is used to represent an initial time series data with defects.
[0032] In one embodiment of the present application, the time series data to be identified includes a steel rolling production process, and time series data of the steel rolling production that has been determined to have defects.
[0033] In one embodiment of the present application, defect feature time series data is used to characterize a segment with typical features selected from a certain historical time series data as a basis for identifying defects.
[0034] In one embodiment of the present application, mathematical statistics include extreme value statistics, mean value and standard deviation statistics; extreme value statistics include maximum value and minimum value.
[0035] Step S120 , slidingly intercepting the time series data to be identified according to a preset span interval to obtain a plurality of subsequence data, and obtaining a distance matrix corresponding to each subsequence data according to the distance between the defect feature time series data and each subsequence data.
[0036] In one embodiment of the present application, the time series data to be identified is subjected to sliding interception according to a preset span interval to obtain a plurality of subsequence data, including: determining the minimum index value of the time series data to be identified as the interception starting point; determining the minimum value between the index value to be compared and the maximum index value of the time series data to be identified as the interception end point, and the index value to be compared is obtained based on the sum of the interception starting point and the upper limit of the preset span interval; intercepting the time series data to be identified according to the interception starting point and the interception end point to obtain subsequence data, and increasing the value of the interception starting point to obtain a new interception starting point; if the new interception starting point is greater than the interception threshold value, then stop intercepting data of the time series data to be identified, and the interception threshold is obtained based on the difference between the maximum index value of the time series data to be identified and the lower limit of the preset span interval; if the new interception starting point is less than or equal to the interception threshold, then repeat the steps of determining the minimum value between the new index value to be compared and the maximum index value of the time series data to be identified as the new interception end point, intercepting the time series data to be identified according to the new interception starting point and the new interception end point to obtain new subsequence data, and increasing the value of the new interception starting point to obtain a new interception starting point, until the new interception starting point is greater than the length to be intercepted.
[0037] In one embodiment of the present application, the preset span interval is used to characterize the upper and lower limits of the span of the identified defect interval to ensure the rationality of the identification result.
[0038] In one embodiment of the present application, a distance matrix corresponding to each subsequence data is obtained based on the distance between the defect feature time series data and each subsequence data, including: normalizing the defect feature time series data and the subsequence data respectively; calculating an alignment path based on the distance between the defect feature time series data and each subsequence data and the distance between the normalized defect feature time series data and the normalized subsequence data, and obtaining a distance matrix corresponding to the subsequence data to determine the optimal alignment path corresponding to the minimum cumulative distance.
[0039] In one embodiment of the present application, the alignment path is calculated by a dynamic time warping (DTW) algorithm.
[0040] In one embodiment of the present application, see Figure 2 , Figure 2 FIG. 4 shows a schematic diagram of a DTW algorithm according to an embodiment of the present application. Figure 2 As shown in the figure, if the defect feature time series data and the time series data to be identified are directly dynamically time-warped, the defect feature time series data T = {T1, T2, ..., T m} and the time series data to be identified S={S1,S2,…,S n}, the Dynamic Time Warping (DTW) algorithm creates an m×n DTW matrix D, where the first element D[1,1] = dist(S1,T1), dist() is a distance function, including Euclidean distance, Manhattan distance, etc., and then the calculation formula for the i-th row and j-th column element D[i,j] is as follows:
[0041] D[i,j]=dist(S i ,T j )+min(D[i-1,j],D[i,j-1],D[i-1,j-1]) Formula (1)
[0042] Where D[i,j] is the i-th row and j-th column element of the DTW matrix, dist(S i ,T j ) is the i-th data S in the time series data to be identified i and the jth data T in the defect feature time series data j The distance between them, D[i-1,j] is the element of the i-1th row and j-1th column of the DTW matrix, D[i,j-1] is the element of the i-th row and j-1th column of the DTW matrix, and D[i-1,j-1] is the element of the i-1th row and j-1th column of the DTW matrix.
[0043] In one embodiment of the present application, there is a minimum element D[m,k] in the last row of elements of D, such that S'={S1, S2, ..., Sk} has the smallest DTW distance with T among all subsequences of S that start with S1, that is, it is more similar in morphology than other subsequences. Since the DTW algorithm is a typical dynamic programming algorithm with limited computational efficiency, the defect localization based on subsequence data proposed in this application can effectively reduce the number of subsequent cyclic calculations of the DTW of the subsequence starting with S1. In addition, the introduction of the preset span interval can effectively reduce the required DTW computation scale and further improve the overall computational efficiency.
[0044] In one embodiment of the present application, the upper and lower limits of the preset span interval are δ max and δ min , j represents the interception starting point, S' represents the subsequence data, n represents the maximum index value of the time series data to be identified, and a represents the minimum index value of the time series data to be identified. The subsequence data can be expressed as: S' = {S j ,S j+1 ,S j+2 ,…,S min(n,j+δmax-1)}, the value of j is in the interval [a,n-δ min ]Inside.
[0045] In one embodiment of the present application, the defect feature time series data is represented by T, and after normalizing T and S', D' corresponding to the DTW matrix is calculated.
[0046] Step S130, if the minimum cumulative distance in a distance matrix is less than a preset distance threshold, and the distance column index of the minimum cumulative distance is greater than the lower limit of the preset span interval, the result dictionary is updated according to the sliding interception position, the minimum cumulative distance and the distance column index corresponding to the subsequence data.
[0047] The minimum cumulative distance is located in the last row of the distance matrix.
[0048] In one embodiment of the present application, the result dictionary is updated according to the sliding interception position, the minimum cumulative distance and the distance column index corresponding to the subsequence data, including: using the interception starting point in the sliding interception position as the result row index of the result dictionary, and the interception starting point is used to characterize the interception start position corresponding to the subsequence data in the time series data to be identified; determining the sum of the distance column index and the result row index as the result column index of the result dictionary; storing the minimum cumulative distance in the result dictionary according to the result row index and the result column index.
[0049] In one embodiment of the present application, a preset distance threshold is used to characterize the upper limit of the DTW distance between the defect feature time series data and the subsequence data. Exceeding the upper limit indicates that there is a large difference in morphology between the subsequence data and the defect feature time series data, thereby determining that the subsequence data does not belong to the target defect type.
[0050] In one embodiment of the present application, the minimum value D'[m,k] is found in the last row of D', that is, the minimum cumulative distance. If k≥δ max and D'[m,k]<d, then R[(j,j + k)] = D'[m,k], where m is the distance row index of the minimum cumulative distance, k is the distance column index of the minimum cumulative distance, d is the DTW threshold, that is, the preset distance threshold, and j is the intercept start point.
[0051] Step S140, perform defect interval screening on the subsequence data corresponding to the result dictionary to obtain a target defect interval and a target defect type.
[0052] In one embodiment of the present application, performing defect interval screening on the subsequence data corresponding to the result dictionary to obtain a target defect interval and a target defect type includes: determining the subsequence data corresponding to the result dictionary as at least one candidate defect interval; performing mathematical statistics on the at least one candidate defect interval to obtain at least one candidate statistical value; if a candidate statistical value satisfies the defect existence condition corresponding to the preset statistical threshold, then determining the candidate defect interval as the target defect interval and determining the defect feature type corresponding to the defect feature time series data as the target defect type.
[0053] In one embodiment of the present application, the result dictionary R is a set of all subsequences in the time series data S to be recognized that are similar to the defect feature time series data T. Since the normalization operation is performed on S and T in advance, the range information is masked. Therefore, it is also necessary to map the intervals in R back to S to form at least one candidate defect interval, and then perform threshold rule judgment on each of the at least one candidate defect interval to screen out the true target defect interval. If no target defect interval is finally formed, it means that there is no defect feature in S similar to T.
[0054] In one embodiment of the present application, there is a corresponding relationship between the result row index of the result dictionary and the intercept start position in the subsequence data corresponding to the subsequence data, and there is a corresponding relationship between the result column index of the result dictionary and the intercept end position in the subsequence data corresponding to the subsequence data. Thus, the subsequence data before normalization is determined through the values of the result row index and the result column index to obtain a candidate defect interval.
[0055] In one embodiment of the present application, please refer to Figure 3 , Figure 3 shows a schematic flowchart of the implementation process of the defect recognition and positioning method for time series data according to an embodiment of the present application. As Figure 3As shown, time series data: obtain initial time series data; threshold rule: the initial time series data first passes through the threshold rule to determine whether there are defects; whether there are defects: if there are defects, the initial time series data is determined as the time series data to be identified, if there are no defects, then end; defect identification and location: use the pre-determined defect classification rules to identify and locate the defects of the initial time series data, and obtain a result dictionary. The defect classification rules include sliding interception of the time series data to be identified according to the preset span interval to obtain multiple sub-sequence data, and dynamically time warping the defect feature time series data and each sub-sequence data to obtain the distance matrix corresponding to each sub-sequence data, and store the elements in the distance matrix that meet the preset distance threshold into the result dictionary; candidate defect interval: determine a batch of candidate defect intervals through the result dictionary; threshold rule: perform threshold rule judgment on these candidate defect intervals; whether there are defects: if the candidate defect intervals are all judged by the threshold rule, it means that the defect identification has failed, otherwise, the identified defect category and location interval are returned. This application innovatively introduces a dynamic time warping algorithm based on the threshold rule to preliminarily judge whether a defect has occurred. The DTW algorithm can deeply analyze and determine whether the classified defect features are hidden in the time series data. Once similar defect features appear in the time series data to be identified, it can not only accurately identify them, but also return the specific interval location of the defect feature, thereby providing a more accurate and comprehensive defect detection and analysis solution for steel rolling production.
[0056] In one embodiment of the present application, during a steel rolling production process, when a technician discovers that a special segment exists in a steel rolling production time series data, and the special segment has not been identified before, the technician can search for similar time series data segments in the historical data to verify whether the special segment belongs to a production process abnormality or a production quality defect. After determining that it is a defect feature, the various parameters of the defect feature can be set and stored in the defect database in the form of defect classification rules.
[0057] In one embodiment of the present application, during the steel rolling production process, when new time series data to be identified is generated, the corresponding defect classification rules in the defect database can be automatically matched with the new time series data to be identified. If a defect occurs, the system will automatically mark the defect category and defect occurrence interval for the time series data.
[0058] See also Figure 4 , Figure 4 A block diagram of a device for identifying and locating defects in time series data according to an embodiment of the present application is shown. The device can be applied to Figure 1 The implementation environment shown in the figure is specifically configured in the computer device 101. The device can also be applied to other exemplary implementation environments and specifically configured in other devices. This embodiment does not limit the implementation environment to which the device is applied.
[0059] like Figure 4 As shown, a defect identification and location device 400 for time series data according to an embodiment of the present application includes: a data acquisition module 401, a multi-sequence regularization module 402, a defect location module 403 and a defect screening module 404.
[0060] The data acquisition module 401 is used to acquire the time series data to be identified and the defect feature time series data;
[0061] The multi-sequence regularization module 402 is used to perform sliding interception of the time series data to be identified according to a preset span interval to obtain multiple sub-sequence data, and obtain a distance matrix corresponding to each sub-sequence data according to the distance between the defect feature time series data and each sub-sequence data;
[0062] The defect location module 403 is used to update the result dictionary according to the sliding interception position, the minimum cumulative distance and the distance column index corresponding to the subsequence data if the minimum cumulative distance in a distance matrix is less than a preset distance threshold and the distance column index of the minimum cumulative distance is greater than the lower limit of the preset span interval, and the minimum cumulative distance is located in the last row of the distance matrix;
[0063] The defect screening module 404 is used to screen the defect intervals of the subsequence data corresponding to the result dictionary to obtain a target defect interval and a target defect type.
[0064] It should be noted that the defect identification and positioning device for time series data provided in the above embodiment and the defect identification and positioning method for time series data provided in the above embodiment belong to the same concept, wherein the specific manner in which each module and unit performs the operation has been described in detail in the method embodiment and will not be repeated here. In practical applications, the defect identification and positioning device for time series data provided in the above embodiment can distribute the above functions to different functional modules as needed, that is, divide the internal structure of the device into different functional modules to complete all or part of the functions described above, and this is not limited here.
[0065] An embodiment of the present application also provides an electronic device, comprising: one or more processors; a storage device for storing one or more programs, and when the one or more programs are executed by one or more processors, the electronic device implements the defect identification and positioning method of the timing data provided in the above-mentioned embodiments.
[0066] See also Figure 5 , Figure 5 The structure diagram of the computer system suitable for implementing the electronic device of the embodiment of the present application is shown. It should be noted that: Figure 5The computer system 500 of the electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.
[0067] like Figure 5 As shown, the computer system 500 includes a central processing unit (CPU) 501, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 502 or the program loaded from the storage part 508 to the random access memory (RAM) 503, such as executing the method in the above embodiment. In the RAM 503, various programs and data required for system operation are also stored. The CPU 501, the ROM 502 and the RAM 503 are connected to each other through the bus 504. The input / output (I / O) interface 505 is also connected to the bus 504.
[0068] The following components are connected to the I / O interface 505: an input section 506 including a keyboard, a mouse, etc.; an output section 507 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker; a storage section 508 including a hard disk, etc.; and a communication section 509 including a network interface card such as a LAN (Local Area Network) card, a modem, etc. The communication section 509 performs communication processing via a network such as the Internet. A drive 510 is also connected to the I / O interface 505 as needed. A removable medium 511, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 510 as needed so that a computer program read therefrom is installed into the storage section 508 as needed.
[0069] In particular, according to an embodiment of the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present application includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a computer program for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication section 509, and / or installed from a removable medium 511. When the computer program is executed by a central processing unit (CPU) 501, various functions defined in the system of the present application are executed.
[0070] It should be noted that the computer-readable medium shown in the embodiment of the present application can be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium can be, for example, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media can include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, wherein a computer-readable computer program is carried. This propagated data signal can take a variety of forms, including but not limited to an electromagnetic signal, an optical signal, or any suitable combination of the above. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. A computer program contained on a computer-readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wired, etc., or any suitable combination of the above.
[0071] The flowchart and block diagram in the accompanying drawings illustrate the possible architecture, functions and operations of the system, method and computer program product according to various embodiments of the present application. Wherein, each box in the flowchart or block diagram can represent a module, a program segment, or a part of the code, and the above-mentioned module, program segment, or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flowchart, and the combination of boxes in the block diagram or flowchart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.
[0072] The units involved in the embodiments described in the present application can be implemented by software or by hardware, and the units described can also be set in a processor. Among them, the names of these units do not constitute a limitation on the units themselves under certain circumstances. Therefore, the technical solution according to the implementation mode of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, including a number of instructions to enable a computing device (which can be a personal computer, a server, a touch terminal, or a network device, etc.) to execute the method according to the implementation mode of the present application.
[0073] Another aspect of the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor of a computer, causes the computer to execute the defect identification and location method of time series data provided in the above-mentioned embodiments. The computer-readable storage medium may be included in the electronic device described in the above-mentioned embodiments, or may exist independently without being assembled into the electronic device.
[0074] In the above embodiments, unless otherwise specified, by using serial numbers such as "first" and "second" to describe common objects, it only means that they refer to different instances of the same object, rather than indicating that the objects being described must adopt a given order, whether in time, space, sorting or any other way.
[0075] The above embodiments are merely illustrative of the principles and effects of the present application, and are not intended to limit the present application. Anyone familiar with the technology may modify or change the above embodiments without violating the spirit and scope of the present application. Therefore, all equivalent modifications or changes made by a person of ordinary skill in the art without departing from the spirit and technical ideas disclosed in the present application shall still be covered by the claims of the present application.
Claims
1. A defect identification and location method for time series data, characterized in that: The method comprises: Acquire time series data to be identified and defect feature time series data; Slidingly intercepting the time series data to be identified according to a preset span interval to obtain a plurality of subsequence data, and obtaining a distance matrix corresponding to each subsequence data according to the distance between the defect feature time series data and each subsequence data; If the minimum cumulative distance in a distance matrix is less than a preset distance threshold, and the distance column index of the minimum cumulative distance is greater than the lower limit of the preset span interval, then the result dictionary is updated according to the sliding interception position corresponding to the subsequence data, the minimum cumulative distance and the distance column index, and the minimum cumulative distance is located in the last row in the distance matrix; The subsequence data corresponding to the result dictionary is screened for defect intervals to obtain target defect intervals and target defect types.
2. The defect identification and positioning method of time series data according to claim 1, characterized in that: The result dictionary is updated according to the sliding interception position corresponding to the subsequence data, the minimum cumulative distance and the distance column index, including: Using the interception starting point in the sliding interception position as the result row index of the result dictionary, the interception starting point is used to represent the interception start position in the time series data to be identified corresponding to the subsequence data; Determine the sum of the distance column index and the result row index as the result column index of the result dictionary; The minimum cumulative distance is stored in the result dictionary according to the result row index and the result column index.
3. The defect identification and positioning method of time series data according to claim 1, characterized in that: The time series data to be identified is subjected to sliding interception according to a preset span interval to obtain a plurality of subsequence data, including: Determine the minimum index value of the time series data to be identified as the interception starting point; Determine the smallest value between the index value to be compared and the maximum index value of the time series data to be identified as the interception end point, wherein the index value to be compared is obtained based on the sum of the interception start point and the upper limit of the preset span interval; According to the interception starting point and the interception end point, data is intercepted on the time series data to be identified to obtain subsequence data, and the value of the interception starting point is increased to obtain a new interception starting point; If the new interception starting point is greater than the threshold to be intercepted, then the data interception of the time series data to be identified is stopped, and the threshold to be intercepted is obtained based on the difference between the maximum index value of the time series data to be identified and the lower limit of the preset span interval; If the new interception starting point is less than or equal to the threshold to be intercepted, the steps of determining the smallest value between the new index value to be compared and the maximum index value of the time series data to be identified as the new interception end point are repeated, and the time series data to be identified is intercepted according to the new interception starting point and the new interception end point to obtain new subsequence data, and the value of the new interception starting point is increased to obtain a new interception starting point, until the new interception starting point is greater than the length to be intercepted.
4. The defect identification and positioning method of time series data according to any one of claims 1 to 3, characterized in that: According to the distance between the defect feature time series data and each of the subsequence data, a distance matrix corresponding to each of the subsequence data is obtained, including: Normalizing the defect feature time series data and the subsequence data respectively; According to the distance between the defect characteristic time series data and each of the subsequence data According to the distance between the normalized defect feature time series data and the normalized subsequence data, the alignment path is calculated to obtain a distance matrix corresponding to the subsequence data to determine the optimal alignment path corresponding to the minimum cumulative distance.
5. The defect identification and positioning method of time series data according to any one of claims 1 to 3, characterized in that: Before obtaining the time series data to be identified, it also includes: Get initial time series data; Performing mathematical statistics on the initial time series data to obtain initial statistical values; If the initial statistical value satisfies the defect existence condition corresponding to the preset statistical threshold, the initial time series data is determined as the time series data to be identified.
6. The defect identification and positioning method of time series data according to claim 5, characterized in that: The subsequence data corresponding to the result dictionary is screened for defect intervals to obtain target defect intervals and target defect types, including: Determine the subsequence data corresponding to the result dictionary as at least one candidate defect interval; Performing mathematical statistics on the at least one candidate defect interval to obtain at least one candidate statistical value; If a candidate statistical value satisfies a defect existence condition corresponding to a preset statistical threshold, the candidate defect interval is determined as a target defect interval, and the defect feature type corresponding to the defect feature time series data is determined as a target defect type.
7. The defect identification and positioning method of time series data according to claim 6, characterized in that: The mathematical statistics include extreme value statistics, mean and standard deviation statistics; extreme value statistics include maximum value and minimum value.
8. A defect identification and location device for time series data, characterized in that: The device comprises: A data acquisition module, used to acquire time series data to be identified and defect feature time series data; A multi-sequence regularization module is used to perform sliding interception on the time series data to be identified according to a preset span interval to obtain multiple sub-sequence data, and obtain a distance matrix corresponding to each sub-sequence data according to the distance between the defect feature time series data and each sub-sequence data; A defect location module, configured to update a result dictionary according to the sliding interception position corresponding to the subsequence data, the minimum cumulative distance and the distance column index if the minimum cumulative distance in a distance matrix is less than a preset distance threshold and the distance column index of the minimum cumulative distance is greater than the lower limit of the preset span interval, wherein the minimum cumulative distance is located in the last row of the distance matrix; The defect screening module is used to screen the defect interval of the subsequence data corresponding to the result dictionary to obtain the target defect interval and the target defect type.
9. An electronic device, characterized in that: The electronic device comprises: one or more processors; A storage device for storing one or more programs, which, when executed by the one or more processors, enables the electronic device to implement the defect identification and positioning method for time series data as described in any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed by a processor of a computer, the computer is caused to execute the defect identification and positioning method for time series data according to any one of claims 1 to 7.