A method for identifying abnormality in motor operation data

By performing segmented processing and box diagram analysis on motor operation data, the problem of inaccurate detection of motor operation data in the prior art is solved, and more accurate abnormal data identification and motor performance analysis are achieved.

CN118734229BActive Publication Date: 2024-11-05ZHEJIANG HILAND TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411214801.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-02
Publication Date
2024-11-05
Estimated Expiration
2044-09-02

AI Technical Summary

Technical Problem

When detecting motor running data, the existing technology is inconsistent with local changes in the overall trend of the current data, resulting in the detection system being unable to accurately identify and troubleshoot abnormal data, affecting the accuracy of the detection results.

Method used

By segmenting the motor running data, a box line chart is constructed, the fluctuation degree and distribution characteristics of each piece of data are calculated, the upper and lower edge positions of the box line chart are determined, abnormal data is selected, and the segmented data is identified and analyzed through the detection system , optimize motor operation data.

Benefits of technology

It improves the accuracy of abnormal data detection, reduces operation errors, and facilitates the detection system to identify motor performance trends and potential fault signs, improving the reliability of detection results.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118734229B_ABST
    Figure CN118734229B_ABST
Patent Text Reader

Abstract

The present invention relates to the field of data processing, and specifically to a method for identifying abnormalities in motor operation data, the method comprising: acquiring motor operation data, segmenting the motor operation data based on preset rules to obtain a plurality of segmented data; for each segmented data, constructing a box plot according to the motor operation data in the segmented data, determining the positions of the upper and lower edges of the box plot, and obtaining abnormal data; identifying and analyzing the abnormal data of all obtained segmented data through a detection system, and optimizing the motor operation data. The present invention segments the collected motor current data according to its local variation characteristics, facilitates the identification and analysis of the detection system, and facilitates the detection system to identify the performance trend and potential fault signs of the motor, thereby solving the technical problem of inaccurate detection results in the prior art.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of data processing, and in particular to a method for identifying abnormalities in motor operation data. Background Art

[0002] A motor is a device that converts electrical energy into mechanical energy. It is widely used in various devices and applications, including industrial machinery, transportation, home appliances, etc. During the use of the motor, the working status of the motor can be understood through its operating data.

[0003] In the prior art, when performing abnormal analysis on the collected motor data, the overall data is usually directly detected through the detection system, and the result is not accurate. This is because the changing trend of the working current of the motor in different time periods is different, that is, when the current data changes locally, the overall trend will also change. As a result, the detection system cannot accurately identify and troubleshoot abnormal data, thereby optimizing the motor operation data. Summary of the invention

[0004] In order to solve the above technical problems, the purpose of the present invention is to provide a method for identifying abnormal operation data of a motor, and the technical solution adopted is as follows:

[0005] A method for identifying abnormal operation data of a motor, the method comprising:

[0006] Acquire motor operation data, and segment the motor operation data based on a preset rule to obtain a plurality of segmented data;

[0007] For each segmented data, a box plot is constructed according to the motor operation data in the segmented data, and the positions of the upper and lower edges of the box plot are determined to obtain abnormal data; wherein, for each segmented data, the position coefficients of the lower and upper edge positions of the box plot of the data segment are calculated; the upper and lower edge positions of the data segment are calculated according to the correlation coefficient of the lower and upper edge positions of the data box plot and the edge determination formula of the box plot; the data with the mark exceeding the upper and lower edge positions is filtered out as abnormal data in the data segment;

[0008] The detection system identifies and analyzes abnormal data in all segmented data to optimize the motor operation data.

[0009] Furthermore, the motor operation data at least includes: sequential motor current data during the motor operation process.

[0010] Furthermore, the step of segmenting the motor operation data based on a preset rule to obtain a plurality of segmented data further includes:

[0011] For the time series motor current data, calculate the possibility of each moment being the cutoff point of the motor current data;

[0012] When the possibility is greater than a preset threshold, determining the moment as a motor current data cutoff point;

[0013] The data is segmented according to the motor current data cutoff point to obtain a plurality of segmented data.

[0014] Furthermore, the possibility of calculating each moment as a motor current data cutoff point further includes:

[0015] For each moment, determine the left neighboring data points and the right neighboring data points at that moment;

[0016] According to the left neighborhood data points and the right neighborhood data points, respectively determine the mean of the left neighborhood data points, the mean of the right neighborhood data points, the average time interval between the maximum value points in the left neighborhood data points, and the average time interval between the maximum value points in the right neighborhood data points;

[0017] The possibility of this moment being the motor current data cutoff point is calculated based on the left side neighborhood data points and the right side neighborhood data points, the mean of the left side neighborhood data points, the mean of the right side neighborhood data points, the average time interval between the maximum points in the left side neighborhood data points, and the average time interval between the maximum points in the right side neighborhood data points.

[0018] Furthermore, the calculation of the position coefficients of the lower and upper edge positions of the box plot of the data segment further includes:

[0019] For each segmented data, calculate the fluctuation degree of the data segment;

[0020] Calculate the fluctuation degree of the maximum and minimum points in the data segment relative to the median of the data segment;

[0021] The position coefficients of the lower and upper edge positions of the box plot of the data segment are calculated according to the fluctuation degree of the data segment and the fluctuation degree of the maximum point and the minimum point in the data segment relative to the median of the data segment.

[0022] Furthermore, the calculation formula of the position coefficient of the lower and upper edge positions of the data segment box plot is:

[0023] ; (3)

[0024] ; (4)

[0025] (5)

[0026] ; (6)

[0027] in; Indicates the degree of fluctuation of the minimum point of the jth data segment relative to the median of the data segment; Indicates the degree of fluctuation of the maximum point of the jth data segment relative to the median of the data segment; The position coefficient representing the lower edge position of the box plot of the jth data segment; represents the position coefficient of the upper edge position of the box plot of the jth data segment; w and v represent the number of minimum and maximum points of the jth data segment respectively; , Respectively represent the amplitudes of the rth minimum point and the r+1th minimum point in the jth data segment; , Respectively represent the amplitudes of the rth maximum point and the r+1th maximum point in the jth data segment; represents the median of the jth data segment; Indicates the degree of fluctuation of the j-th data segment.

[0028] Furthermore, the step of calculating the fluctuation degree of the data segment further includes:

[0029] Determine the extreme point, the amplitude of the extreme point, the number of extreme points, the mean value of the motor current data of the data segment, and the number of data points;

[0030] Determine the mean of the absolute values ​​of the differences between the adjacent extreme value points of the data segment, the mean of the sum of the differences between the amplitude of each data point and the amplitude mean of all extreme value points of the data segment according to the amplitudes of each two adjacent extreme value points;

[0031] Calculate the fluctuation degree of this data segment based on the above data.

[0032] Furthermore, the calculation of the fluctuation degree of the maximum value point and the minimum value point in the data segment relative to the median of the data segment further includes:

[0033] Calculate the mean of the sum of the differences between every two adjacent minimum value points and the mean of the sum of the differences between every two adjacent maximum value points in the data segment;

[0034] Calculate the sum of the differences between the minimum point in the data segment and the median of the data segment, and the sum of the differences between the maximum point and the median of the data segment;

[0035] Calculate the degree of fluctuation of the minimum point in the data segment relative to the median of the data segment according to the mean of the sum of differences between every two adjacent minimum points in the data segment and the sum of differences between the minimum point in the data segment and the median of the data segment;

[0036] The degree of fluctuation of the maximum point in the data segment relative to the median of the data segment is calculated based on the average of the sum of differences between every two adjacent maximum points in the data segment and the sum of differences between the maximum point in the data segment and the median of the data segment.

[0037] Furthermore, the step of calculating the upper and lower edge positions of the data segment according to the correlation coefficient of the lower and upper edge positions of the data box plot and the edge determination formula of the box plot further includes:

[0038] Determine the upper quartile and the lower quartile of the data segment;

[0039] Calculate the absolute value of the difference between the upper quartile data and the lower quartile;

[0040] The upper and lower edge positions of the data segment are calculated based on the upper quartile and the lower quartile, the absolute value of the difference between the upper quartile and the lower quartile, and the correlation coefficient of the lower and upper edge positions of the box plot of the data segment, and the edge determination formula of the box plot.

[0041] Furthermore, the preset threshold is an empirical value.

[0042] The present invention has the following beneficial effects: the present invention segments the collected motor current data according to its local variation characteristics, thereby avoiding the problem of inaccurate abnormal data detection results caused by large changes in local motor current data while the overall motor current data tends to be stable and the changes are not obvious, and then calculates the fluctuation degree and distribution characteristics of each segmented data to obtain the upper and lower edge positions of the box plot corresponding to the corresponding data segment, and the abnormal data in each segment of data is screened and marked according to the upper and lower edge positions. The box plot calculation method reduces the error in the calculation process of each segment of data, further improves the accuracy of the abnormal data detection results, facilitates the identification and analysis of the detection system, and facilitates the detection system to identify the performance trend and potential fault signs of the motor. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0044] Figure 1 A flow chart of a method for identifying abnormal motor operation data provided by an embodiment of the present invention;

[0045] Figure 2A flow chart of obtaining abnormal data from segmented data provided by an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0046] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the specific implementation method, structure, features and effects of a motor operation data abnormality identification method proposed by the present invention in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0047] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0048] A specific scheme of a method for identifying abnormal operation data of a motor provided by the present invention is described in detail below with reference to the accompanying drawings.

[0049] See also Figure 1 , which shows a flow chart of a method for identifying abnormal motor operation data provided by an embodiment of the present invention, the method comprising:

[0050] Step S110: Obtain motor operation data.

[0051] In an optional embodiment, the motor operation data at least includes: sequential motor current data during the motor operation process. The main purpose of this embodiment is to screen the collected motor current data for abnormal values, so the sequential motor current data during the motor operation process can be collected by a current sensor; the sensor can be contact or inductive (the contact type is used in this embodiment), and the current data collected when the motor is running is collected using a sensor with high sensitivity and rapid response to reduce the error in the data collection process. The frequency of collection is set to 0.1 seconds (empirical value, specifically set according to the equipment environment and the required accuracy), and the collection time can be collected throughout the motor operation process or collect data for a period of time (the collection time in this embodiment is 1 hour, which is specifically determined according to the equipment detection requirements), and then the collected data is sent to the computer system by wired or wireless means to facilitate the analysis of the subsequent detection system.

[0052] Step S120: segmenting the motor operation data based on preset rules to obtain a plurality of segmented data.

[0053] In this embodiment, by analyzing the overall current data collected, it is found that while the current data fluctuates in the local time period, its overall change trend will also change. Compared with the characteristics of the overall current data change, the local current data has a certain regularity, while the current data does not have a certain regularity in its overall nature and is difficult to analyze. This is because the overall change trend of the motor's current data is caused by uncertain environmental factors such as the load of the motor when it is working, and the local change of the motor is due to a short time, and the possibility of sudden changes in uncertain environmental factors such as the load at this time is small, and the current data change is more regular. Based on this analysis, if a box plot is constructed overall through the collected current data, the degree of change and the change law of the current data in different time periods may be quite different. For example, the collected current data is normal in a certain section of the data, but may become abnormal data in another section, which leads to inaccurate detection results. Therefore, it is necessary to divide the collected current data into multiple sections according to its local change characteristics.

[0054] When analyzing the collected motor current data, the overall construction of the box plot cannot accurately filter out the abnormal data in the segment of data. Therefore, the similarity of the left and right neighboring data segments of the segment of data at each moment is analyzed through the local characteristics of the current data, and the possibility of the selected moment being the data truncation point is judged based on the similarity, and then the motor current data is divided into multiple segments through a preset threshold method; by calculating the overall fluctuation degree of the current data segment, and then calculating the distribution change degree between the maximum point and the minimum point of the median sum of the segment of data, the position coefficients of the upper and lower edge positions are obtained according to the fluctuation degree of the two compared with the overall data segment, and then the upper and lower edge positions of the segment of data are determined, and the abnormal data of the segment of data is determined by the upper and lower edge positions of the segment of data. Similarly, abnormal data of other data segments are obtained by the same method, and then these abnormal data are screened and marked.

[0055] In an optional embodiment, step S120 further includes: for the time-series motor current data, calculating the possibility that each moment is a motor current data cutoff point; when the possibility is greater than a preset threshold, determining the moment as the motor current data cutoff point; and segmenting the data according to the motor current data cutoff point to obtain multiple segmented data.

[0056] In an optional embodiment, step S120 further includes: for each moment, determining the left neighborhood data points and the right neighborhood data points at that moment; determining the mean of the left neighborhood data points, the mean of the right neighborhood data points, the average time interval between the maximum points in the left neighborhood data points, and the average time interval between the maximum points in the right neighborhood data points according to the left neighborhood data points and the right neighborhood data points; calculating the possibility that the moment is a truncation point of the motor current data according to the left neighborhood data points and the right neighborhood data points, the mean of the left neighborhood data points, the mean of the right neighborhood data points, the average time interval between the maximum points in the left neighborhood data points, and the average time interval between the maximum points in the right neighborhood data points.

[0057] In this step, the collected current data is divided into multiple segments according to its local change characteristics and change trends, which is more conducive to obtaining abnormal values ​​of current data in different segmented data; the segmented calculation formula for the collected motor current data is as follows (1):

[0058] ; (1)

[0059] In the formula, Indicates the possibility that the i-th moment is the motor current data cutoff point, that is, the data between the i-th moment (including the data at the i-th moment) and the previous data cutoff point is a segment (for the first segment of data, it is the data between the initial moment and the first data cutoff point); Represents the mean of n data points at the i-th moment and its left neighborhood; It represents the mean of n data points at the ith moment and its right neighborhood; n is 10 (an empirical value); it should be noted that the reason why the data at the ith moment is added to its left and right neighborhoods for calculation is that the value may be an abnormal value, and the calculation formulas on both sides of the neighborhood include this value to reduce the error in the similarity calculation of the left and right neighborhood data points at the ith moment caused by the abnormality of the value, making the judgment and selection of the data segment cutoff point more accurate.

[0060] In the formula, Represents the average time interval between the maximum points in the i-th moment and its left neighborhood n data points; Represents the average time interval between the maximum points in the i-th moment and its right neighboring n data points; , Represents the sequence of left and right neighboring data segments at the i-th moment; , represents the absolute value of the Pearson correlation coefficient between two sequences, which is a well-known technique and will not be described in detail here; represents an exponential function with a natural constant as base; represents the linear normalization function.

[0061] in, It represents the absolute value of the difference between the mean of the n data points at the ith moment and its left neighboring data points and the mean of the n data points at the ith moment and its left neighboring data points. Because when the ith moment is the cutoff point, the difference in the change characteristics of the data segments on its left and right neighbors is relatively large, so it is necessary to cut off from this point, and the average amplitude of the left and right neighboring data at the ith moment reflects the overall amplitude level of the two data segments. Therefore, the larger the value, the greater the difference in the change characteristics of the left and right neighboring data segments at the ith moment, and the greater the possibility that this point is the cutoff point. It represents the absolute value of the difference between the average time interval between the maximum points in the i-th moment and the n data points in its left neighborhood and the average time interval between the maximum points in the i-th moment and the n data points in its right neighborhood. The time interval between the maximum points can reflect the periodicity of the change of the data segment to a certain extent. When the periodicity difference of the change of the neighborhood data segment around the i-th moment is relatively large, it means that the change characteristics of the two data are relatively large, and the possibility that the point is a truncation point is greater. It represents the absolute value of the Pearson correlation coefficient between the segment of data consisting of n data points at the i-th moment and its left neighborhood and the segment of data consisting of n data points at the i-th moment and its right neighborhood. The larger the value, the greater the similarity difference between the two pieces of data, and the greater the possibility that the i-th moment is the cutoff point.

[0062] When the probability is greater than a preset threshold, the moment is determined as the motor current data cutoff point. In this embodiment, the preset threshold is an empirical value. (empirical value), the data point at the i-th moment is the cutoff point; through this method, the motor current data is segmented according to the motor current data cutoff point to obtain multiple segmented data.

[0063] Step S130: For each segmented data, a box plot is constructed according to the motor operation data in the segmented data, and the positions of the upper and lower edges of the box plot are determined to obtain abnormal data.

[0064] In this embodiment, the positions of the upper and lower edges will affect the detection accuracy of abnormal data, so it is necessary to determine the positions of the upper and lower edges of the box plot for each section of current data, which is conducive to accurate screening of abnormal data.

[0065] See also Figure 2 , which shows a flowchart of obtaining abnormal data for segmented data provided by an embodiment of the present invention. In an optional implementation, step S130 further includes the following steps S1301-S1305:

[0066] Step S1301: For each segmented data, calculate the fluctuation degree of the data segment.

[0067] Step S1302: Calculate the degree of fluctuation of the maximum and minimum points in the data segment relative to the median of the data segment.

[0068] Step S1303: Calculate the position coefficients of the lower and upper edge positions of the box plot of the data segment according to the fluctuation degree of the data segment and the fluctuation degree of the maximum point and the minimum point in the data segment relative to the median of the data segment.

[0069] In an optional embodiment, step S1301 further includes: determining the extreme points, extreme point amplitudes, the number of extreme points, the mean of the motor current data of the data segment, and the number of data points of the data segment; determining the mean of the absolute values ​​of the differences between adjacent extreme points of the data segment, the mean of the sum of the differences between the amplitudes of each data point and the amplitude mean of all extreme points of the data segment based on the amplitudes of every two adjacent extreme points; and calculating the degree of fluctuation of the data segment based on the above data.

[0070] Through the above steps, multiple data segments are obtained. When determining the upper and lower edges of the box plot of each segmented data, the position of the upper and lower edges depends on the fluctuation degree of the current data segment. If the current data segment has a large change degree, then when determining the upper and lower edges, the position is relatively far, that is, the distance between the upper and lower edges is relatively far. Therefore, it is necessary to first calculate the fluctuation degree of each data segment, which is as follows (2):

[0071] ; (2)

[0072] In the formula, Indicates the degree of fluctuation of the jth data segment; , They represent the amplitudes of the sth extreme point and the s+1th extreme point of the jth data segment respectively; u represents the number of extreme points in the jth data segment; represents the amplitude of the i-th data point in the j-th data segment; represents the mean of the j-th data segment; m represents the number of data points in the j-th data segment.

[0073] in, It represents the mean of the absolute values ​​of the differences between the adjacent extreme value points of the j-th data segment. The larger the value, the larger the fluctuation range of the data segment, that is, the greater the fluctuation degree of the data segment, and the greater the position distance between the upper and lower edges of the data segment; It represents the mean of the sum of the differences between the amplitude of each data point in the jth data segment and the mean amplitude of the data points in this data segment. The larger the value, the greater the fluctuation degree of the data end, and the greater the position distance between the upper and lower edges of the data segment.

[0074] In an optional embodiment, step S1302 further includes: calculating the mean of the sum of differences between every two adjacent minimum points in the data segment and the mean of the sum of differences between every two adjacent maximum points; calculating the sum of differences between the minimum point in the data segment and the median of the data segment and the sum of differences between the maximum point and the median of the data segment; calculating the degree of fluctuation of the minimum point in the data segment relative to the median of the data segment based on the mean of the sum of differences between every two adjacent minimum points in the data segment and the sum of differences between the minimum point in the data segment and the median of the data segment; calculating the degree of fluctuation of the maximum point in the data segment relative to the median of the data segment based on the mean of the sum of differences between every two adjacent maximum points in the data segment and the sum of differences between the maximum point in the data segment and the median of the data segment.

[0075] Furthermore, the fluctuation degree of each data segment can only roughly determine the relative distance between the upper and lower edges of the box plot of the data segment, but cannot accurately determine the position of the upper and lower edges. Therefore, it is necessary to combine the fluctuation degree of the data points on the left and right sides of the median of the data segment to determine the position coefficient of the upper and lower edges. The calculation formula is as follows (3)-(6):

[0076] ; (3)

[0077] ; (4)

[0078] (5)

[0079] ; (6)

[0080] in; Indicates the degree of fluctuation of the minimum point of the jth data segment relative to the median of the data segment; Indicates the degree of fluctuation of the maximum point of the jth data segment relative to the median of the data segment; The position coefficient representing the lower edge position of the box plot of the jth data segment; represents the position coefficient of the upper edge position of the box plot of the jth data segment; w and v represent the number of minimum and maximum points of the jth data segment respectively; , Respectively represent the amplitudes of the rth minimum point and the r+1th minimum point in the jth data segment; , Respectively represent the amplitudes of the rth maximum point and the r+1th maximum point in the jth data segment; Represents the median of the jth data segment.

[0081] In the formula, It represents the mean of the sum of the differences between the adjacent minimum points of the j-th data segment. The larger the value, the greater the fluctuation of the part of the data segment that is smaller than the median value of the data segment, and the position determined by the lower edge is farther away; similarly, It represents the mean of the sum of differences between adjacent maximum points in the jth data segment. The larger the value, the greater the fluctuation of the part of the data segment that is greater than the median of the data segment, and the position determined by the upper edge is farther away. It represents the sum of the differences between the minimum point of the jth data segment and the median of the data segment. The larger the value, the larger the distribution range of the data segment that is less than the median, and the farther the position determined by the lower edge is. Similarly, It represents the sum of the differences between the maximum point of the jth data segment and the median of the data segment. The larger the value is, the farther the upper edge is determined.

[0082] According to the ratio of the variation characteristics of each segment of local current data relative to the median and the overall fluctuation degree of the segment of data, the coefficients of the upper and lower edges are determined, reflecting the fluctuation degree of the two parts of data relative to the median for the overall data. The larger the value, the larger the corresponding coefficients of the upper and lower edges, and vice versa. In formula (5), It represents the ratio of the fluctuation degree of the minimum point of the jth data segment relative to the median of the data segment and the overall fluctuation degree of the data segment. Formula (6) It represents the ratio of the fluctuation degree of the maximum point of the j-th data segment relative to the median of the data segment and the overall fluctuation degree of the data segment. The larger the corresponding ratio is, the larger the coefficient of the corresponding edge is.

[0083] In an optional embodiment, step S130 further includes: for each segmented data, calculating the position coefficients of the lower and upper edge positions of the box plot of the data segment; calculating the upper and lower edge positions of the data segment based on the correlation coefficient of the lower and upper edge positions of the data box plot and the edge determination formula of the box plot; filtering and marking data that exceeds the upper and lower edge positions as abnormal data in the data segment.

[0084] In an optional implementation, step S130 further includes steps S1304-S1305:

[0085] Step S1304: determine the upper quartile and the lower quartile of the data segment; and calculate the absolute value of the difference between the upper quartile data and the lower quartile.

[0086] Step S1305: The upper and lower edge positions of the data segment are calculated according to the upper quartile and the lower quartile, the absolute value of the difference between the upper quartile data and the lower quartile, and the correlation coefficient of the lower and upper edge positions of the box plot of the data segment, and the edge determination formula of the box plot.

[0087] The position coefficients of the upper and lower edges are obtained through the above steps. According to the existing edge determination formula of the box plot, the positions of the upper and lower edges are obtained. The calculation formula is as follows (7)-(8):

[0088] ; (7)

[0089] ; (8)

[0090] In the formula Indicates the upper edge position of the jth data segment; Indicates the lower edge position of the jth data segment; represents the upper quartile of the j-th data segment; represents the lower quartile of the jth data segment, , represent the position coefficient of the lower edge position and the position coefficient of the upper edge position of the j-th data segment obtained in the above steps respectively; It represents the value of the upper quartile minus the lower quartile of the j-th data segment.

[0091] The upper and lower edge positions of the box plot of each data segment are obtained through the above steps. When any data in the data segment exceeds the upper and lower edge positions in the box plot corresponding to the data segment, it can be screened and marked as abnormal data to facilitate further analysis by the subsequent detection system.

[0092] Step S140: The detection system is used to identify and analyze abnormal data of all segmented data obtained, and optimize the motor operation data.

[0093] Through the above steps, the abnormal data of each segment of the motor current data are obtained, and these abnormal data are further analyzed by the detection system, which is conducive to the troubleshooting and analysis of the motor's performance and potential faults.

[0094] The method of this embodiment is adopted to segment the collected motor current data according to its local change characteristics, and then calculate the fluctuation degree and distribution characteristics of each segmented data to obtain the upper and lower edge positions of the box plot corresponding to the corresponding data segment. The abnormal data in each segment of data is screened and marked according to the upper and lower edge positions, which is convenient for the detection system to identify and analyze, and to facilitate the detection system to identify the performance trends and potential fault signs of the motor.

[0095] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0096] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A method for identifying abnormal motor operation data, characterized in that: The method comprises: Acquire motor operation data, and segment the motor operation data based on a preset rule to obtain a plurality of segmented data; For each segmented data, a box plot is constructed according to the motor operation data in the segmented data, and the positions of the upper and lower edges of the box plot are determined to obtain abnormal data; wherein, for each segmented data, the position coefficients of the lower and upper edge positions of the box plot of the data segment are calculated; the upper and lower edge positions of the data segment are calculated according to the correlation coefficient of the lower and upper edge positions of the data box plot and the edge determination formula of the box plot; the data with the mark exceeding the upper and lower edge positions is filtered out as abnormal data in the data segment; The detection system identifies and analyzes the abnormal data of all segmented data obtained, and optimizes the motor operation data; The method for obtaining the position coefficients of the lower and upper edge positions of the data segment box plot is as follows: for each segmented data, the fluctuation degree of the data segment is calculated; Calculate the fluctuation degree of the maximum and minimum points in the data segment relative to the median of the data segment; The position coefficients of the lower and upper edge positions of the box plot of the data segment are calculated according to the fluctuation degree of the data segment and the fluctuation degree of the maximum value point and the minimum value point in the data segment relative to the median of the data segment; Among them, the calculation formula for the position coefficient of the lower and upper edge positions of the data segment box plot is: ;(3) ;(4) (5) ;(6) in; Indicates the degree of fluctuation of the minimum point of the jth data segment relative to the median of the data segment; Indicates the degree of fluctuation of the maximum point of the jth data segment relative to the median of the data segment; The position coefficient representing the lower edge position of the box plot of the jth data segment; represents the position coefficient of the upper edge position of the box plot of the jth data segment; w and v represent the number of minimum and maximum points of the jth data segment respectively; , Respectively represent the amplitudes of the rth minimum point and the r+1th minimum point in the jth data segment; , Respectively represent the amplitudes of the rth maximum point and the r+1th maximum point in the jth data segment; represents the median of the jth data segment; Indicates the degree of fluctuation of the j-th data segment.

2. The method for identifying abnormal motor operation data according to claim 1, characterized in that: The motor operation data at least includes: sequential motor current data during the motor operation process.

3. The method for identifying abnormal motor operation data according to claim 2, characterized in that: The step of segmenting the motor operation data based on a preset rule to obtain a plurality of segmented data further comprises: For the time series motor current data, calculate the possibility of each moment being the cutoff point of the motor current data; When the possibility is greater than a preset threshold, determining the moment as a motor current data cutoff point; The data is segmented according to the motor current data cutoff point to obtain a plurality of segmented data.

4. The method for identifying abnormal motor operation data according to claim 3, characterized in that: The possibility of calculating each moment as the motor current data cutoff point further includes: For each moment, determine the left neighboring data points and the right neighboring data points at that moment; According to the left neighborhood data points and the right neighborhood data points, respectively determine the mean of the left neighborhood data points, the mean of the right neighborhood data points, the average time interval between the maximum value points in the left neighborhood data points, and the average time interval between the maximum value points in the right neighborhood data points; The possibility of this moment being the motor current data cutoff point is calculated based on the left side neighborhood data points and the right side neighborhood data points, the mean of the left side neighborhood data points, the mean of the right side neighborhood data points, the average time interval between the maximum points in the left side neighborhood data points, and the average time interval between the maximum points in the right side neighborhood data points.

5. The method for identifying abnormal motor operation data according to claim 1, characterized in that: The calculating the fluctuation degree of the data segment further comprises: Determine the extreme point, the amplitude of the extreme point, the number of extreme points, the mean value of the motor current data of the data segment, and the number of data points; Determine the mean of the absolute values ​​of the differences between the adjacent extreme value points of the data segment, the mean of the sum of the differences between the amplitude of each data point and the amplitude mean of all extreme value points of the data segment according to the amplitudes of each two adjacent extreme value points; Calculate the fluctuation degree of this data segment based on the above data.

6. The method for identifying abnormal motor operation data according to claim 1, characterized in that: The calculation of the fluctuation degree of the maximum value point and the minimum value point in the data segment relative to the median of the data segment further includes: Calculate the mean of the sum of the differences between every two adjacent minimum value points and the mean of the sum of the differences between every two adjacent maximum value points in the data segment; Calculate the sum of the differences between the minimum point in the data segment and the median of the data segment, and the sum of the differences between the maximum point and the median of the data segment; Calculate the degree of fluctuation of the minimum point in the data segment relative to the median of the data segment according to the mean of the sum of differences between every two adjacent minimum points in the data segment and the sum of differences between the minimum point in the data segment and the median of the data segment; The degree of fluctuation of the maximum point in the data segment relative to the median of the data segment is calculated based on the average of the sum of differences between every two adjacent maximum points in the data segment and the sum of differences between the maximum point in the data segment and the median of the data segment.

7. The method for identifying abnormal motor operation data according to claim 5 or 6, characterized in that: The step of calculating the upper and lower edge positions of the data segment according to the correlation coefficient of the lower and upper edge positions of the data box plot and the edge determination formula of the box plot further includes: Determine the upper quartile and the lower quartile of the data segment; Calculate the absolute value of the difference between the upper quartile data and the lower quartile; The upper and lower edge positions of the data segment are calculated based on the upper quartile and the lower quartile, the absolute value of the difference between the upper quartile and the lower quartile, and the correlation coefficient of the lower and upper edge positions of the box plot of the data segment, and the edge determination formula of the box plot.

8. The method for identifying abnormal motor operation data according to claim 3, characterized in that: The preset threshold is an empirical value.