Mechanical part machining data dynamic safety protection method

Abstract

The invention relates to the technical field of data processing, in particular to a mechanical part machining data dynamic safety protection method which comprises the steps that a machining current data sequence of a mechanical part is collected; obtaining an initial standard current factor according to the processing current data sequence; obtaining an abnormal current data segment according to the initial standard current factor; obtaining a current trend difference factor according to the abnormal current data segment; obtaining abnormal current data according to the current trend difference factor; obtaining data similarity according to fluctuation trend differences between different abnormal current data in the abnormal current data segment; obtaining an associated abnormal data set according to the data similarity; obtaining a vibration abnormal factor according to the associated abnormal data set; and compressing the processing current data sequence according to the vibration abnormal factor. According to the method, the segmentation of the current data is more reasonable, the compression efficiency of a revolving door compression algorithm is improved, and the dynamic safety of processing data is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, in particular to a mechanical part machining data dynamic security protection method. BACKGROUND

[0002] The machining process of mechanical parts usually uses a turning tool on a numerical control machine tool to rub the mechanical parts, thereby completing the machining. The friction generated during the machining process changes constantly, causing the current data of the numerical control machine tool to also change constantly. In order to ensure the normal operation of the machining process of the mechanical parts, the current data of the numerical control machine tool needs to be saved and analyzed. However, due to the large amount of collected current data, in order to improve the efficiency of data security protection, the current data needs to be compressed. The traditional method compresses the current data by using a rotating door compression algorithm. However, the traditional rotating door compression algorithm cannot effectively distinguish abnormal data generated by machine tool vibration and other interference factors, resulting in uniform segmentation compression of all abnormal data and normal data, reducing the compression efficiency and the dynamic security of the machining data. SUMMARY

[0003] The present application provides a mechanical part machining data dynamic security protection method to solve the existing problems: the traditional rotating door compression algorithm cannot effectively distinguish abnormal data generated by machine tool vibration and other interference factors, resulting in uniform segmentation compression of all abnormal data and normal data, reducing the compression efficiency.

[0004] The mechanical part machining data dynamic security protection method of the present application adopts the following technical scheme: The method comprises the following steps: Collecting machining current data sequences of a plurality of mechanical parts, the machining current data sequences comprising a plurality of current data, each current data corresponding to a collection time; Obtaining an initial standard current factor for each current data according to the change difference of the current data between different mechanical parts, the initial standard current factor being used to describe the difference between the current data and the normal current data; selecting a plurality of initial abnormal current data and a plurality of normal current data from the current data according to the maximum change difference of the initial standard current factors between the current data of different mechanical parts at the same collection time; The processing current data sequence is divided into several abnormal current data segments based on the initial abnormal current data and normal current data. Based on the changing trends of abnormal current data before and after different initial abnormal current data within each abnormal current data segment, a current trend difference factor is obtained for each initial abnormal current data segment. Several abnormal current data are selected from the initial abnormal current data based on the current trend difference factor. The data similarity between any two abnormal current data segments is obtained based on the fluctuation trend differences between different abnormal current data segments. The abnormal current data are then categorized based on data similarity to obtain several associated abnormal datasets. Finally, the vibration anomaly factor for each abnormal current data segment is obtained based on the distribution of different abnormal current data within the associated abnormal datasets. The machining current data sequence is compressed based on the vibration anomaly factor.

[0005] Preferably, the method for obtaining the initial standard current factor for each current data based on the variation differences in current data between different mechanical parts includes: Take any one mechanical part as the target mechanical part, and denote each mechanical part other than the target mechanical part as the reference mechanical part of the target mechanical part; In the formula, The first data in the sequence of machining current data for the target mechanical part Initial standard current factor for each current data; This indicates the sequence of machining current data for all reference mechanical parts of the target mechanical part. In the current data, the first one is related to the machining current data sequence of the target mechanical part. The number of equal current data; This indicates the number of all reference mechanical parts for the target mechanical part.

[0006] Preferably, the method for selecting several initial abnormal current data and several normal current data from the current data based on the maximum difference in the initial standard current factor among the current data of different mechanical parts at the same acquisition time includes: The first in the sequence of machining current data for all mechanical parts From the initial standard current factors of the current data, the current data corresponding to the largest initial standard current factor is taken as the first current data in the entire processing current data sequence. Standard current data for each current data point; obtain standard current data for all current data in the processing current data sequence for each mechanical part. For any one of the machining current data sequence of any one of the mechanical parts, the absolute value of the difference between the standard current data of the current data and the current data is recorded as the standard current difference of the current data, the standard current difference of all current data in the machining current data sequence of the mechanical parts is obtained, all standard current differences are linearly normalized, and each normalized standard current difference is recorded as a standard current factor; A standard current factor threshold T1 is preset, for any one of the machining current data sequence of any one of the mechanical parts, if the standard current factor of the current data is greater than or equal to T1, the current data is recorded as initial abnormal current data; if the standard current factor of the current data is less than T1, the current data is recorded as normal current data.

[0007] Preferably, the method of dividing the machining current data sequence into several abnormal current data segments according to the initial abnormal current data and the normal current data comprises: For any one of the machining current data sequence of any one of the mechanical parts, an initial abnormal current data quantity threshold T2 is preset; in the machining current data sequence, a data segment composed of every T2 initial abnormal current data is recorded as an abnormal current data segment.

[0008] Preferably, the method of obtaining the current trend difference factor of each initial abnormal current data according to the change trend of the abnormal current data before and after different initial abnormal current data in the abnormal current data segment comprises: For any one of the machining current data sequence of any one of the mechanical parts, an initial abnormal current data quantity threshold T2 is preset; in the machining current data sequence, a data segment composed of every T2 initial abnormal current data is recorded as an abnormal current data segment. In the formula, represents the initial current trend difference factor of the initial abnormal current data; represents the number of all pre-abnormal current data of the initial abnormal current data; represents the first pre-abnormal current data; represents the first pre-abnormal current data; represents the first pre-abnormal current data; represents the first pre-abnormal current data; represents the number of all post-abnormal current data of the initial abnormal current data; represents the first post-abnormal current data; represents the first post-abnormal current data fitting current value; represents the number of all current data in the abnormal current data segment; an initial current trend difference factor of all initial abnormal current data in the abnormal current data segment is obtained, and all initial current trend difference factors are linearly normalized, and each initial current trend difference factor after normalization is recorded as a current trend difference factor.

[0009] Preferably, the method of screening a plurality of abnormal current data from the initial abnormal current data according to the current trend difference factor comprises the following specific method: A current trend difference factor threshold T3 is preset, and the initial abnormal current data with a current trend difference factor greater than T3 is recorded as abnormal current data.

[0010] Preferably, the data similarity of any two abnormal current data is obtained according to the fluctuation trend difference between different abnormal current data in the abnormal current data segment, and the specific method comprises the following: For any two adjacent abnormal current data in any abnormal current data segment, in the two abnormal current data, the difference value between the second abnormal current data and the first abnormal current data is recorded as a first difference value, the difference value between the sequence numbers of the second abnormal current data and the first abnormal current data in the abnormal current data segment is recorded as a second difference value, and the ratio of the first difference value to the second difference value is recorded as the slope of the first abnormal current data. For any three adjacent abnormal current data in the abnormal current data segment, in the three abnormal current data, the absolute value of the difference between the slopes of the first abnormal current data and the second abnormal current data is recorded as the pre-trend difference value of the second abnormal current data, and the absolute value of the difference between the slopes of the third abnormal current data and the second abnormal current data is recorded as the post-trend difference value of the second abnormal current data. In the formula, represents the data similarity between the first abnormal current data and the first abnormal current data, ; represents a preset hyperparameter; represents the pre-trend difference value of the first abnormal current data; represents the pre-trend difference value of the first abnormal current data; represents a preset hyperparameter; represents the data similarity between the first an abnormal current data; an abnormal current data; an abnormal current data; an abnormal current data; a post-trend difference value of the abnormal current data; a post-trend difference value of the abnormal current data; a post-trend difference value of the abnormal current data; a sign function.

[0011] Preferably, the abnormal current data is classified according to data similarity to obtain a plurality of associated abnormal data sets, and the specific method comprises: a similarity threshold KL is preset, any abnormal current data in any abnormal current data segment is recorded as target abnormal current data, each abnormal current data except the target abnormal current data is recorded as reference abnormal current data of the target abnormal current data, the reference abnormal current data with data similarity equal to KL of the target abnormal current data is recorded as associated abnormal current data of the target abnormal current data, and a data set composed of all the associated abnormal current data is taken as the associated abnormal data set of the target abnormal current data.

[0012] Preferably, the vibration abnormality factor of each abnormal current data segment is obtained according to the distribution of different abnormal current data in the associated abnormal data set, and the specific method comprises: In the formula, a vibration abnormality degree of any abnormal current data segment; a mean value of all abnormal current data in the abnormal current data segment; a mean value of all current data in the abnormal current data segment; a number of all abnormal current data in the abnormal current data segment; a number of the abnormal current data in the associated abnormal data set of all abnormal current data in the abnormal current data segment; a number of the abnormal current data in the associated abnormal data set of all abnormal current data in the abnormal current data segment; a number of all abnormal current data in the associated abnormal data set of the abnormal current data; a number of all abnormal current data in the associated abnormal data set of the abnormal current data;

[0013] Preferably, the processing current data sequence is compressed according to the vibration abnormality factor, and the specific method comprises: a vibration abnormality factor threshold For any abnormal current data segment, the first and second current data points within the segment are both designated as a dividing point. For any other abnormal current data point within the segment besides the first and second current data points... Current data; In the formula, Indicates the first The degree of boundary between individual current data; Indicates the preset hyperparameters; Indicates the first The slope of the first dividing point before each current data point; Indicates the first The slope of the second dividing point before the current data; The vibration anomaly factor represents the abnormal current data segment; This represents the preset threshold for the vibration anomaly factor; Indicates the first The first dividing point before the current data and the first... The maximum upslope of all current data in a data segment consisting of current data; Indicates the first The first dividing point before the current data and the first... The minimum downslope of all current data in a data segment consisting of current data; A threshold T5 is preset for the degree of separation; current data with a degree of separation greater than T5 in the abnormal current data segment are recorded as the separation point; all separation points in the abnormal current data segment are obtained; the data segment formed between any two separation points is recorded as the data segment to be compressed; each data segment to be compressed is used as a segment, and rotating door compression is performed on all segments to obtain several compressed data segments; all compressed data segments are stored in the database.

[0014] The beneficial effects of the technical solution of this invention are as follows: An initial standard current factor is obtained based on the processing current data sequence; abnormal current data segments are obtained based on the initial standard current factor; a current trend difference factor is obtained based on the abnormal current data segments; data similarity is obtained based on the current trend difference factor; the degree of vibration anomaly is obtained based on the data similarity; and the processing current data sequence is compressed based on the degree of vibration anomaly. The initial standard current factor of this invention reflects the difference between current data and normal current data; the current trend difference factor reflects the degree to which the initial abnormal current data belongs to current data with a significant change in actual trend; and the data similarity reflects the similarity of trend changes and numerical change patterns between current data. This makes the segmentation of current data more reasonable, improves the compression efficiency of the rotating door compression algorithm, and enhances the dynamic security of the processing data. BRIEF DESCRIPTION OF DRAWINGS

[0015] In order to more clearly illustrate the technical solutions of the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or prior art description will be briefly introduced. Obviously, the drawings in the following description only constitute some embodiments of the present application, and other drawings can be obtained by those skilled in the art without any creative effort on the basis of these drawings.

[0016] Figure 1 A flow chart of the steps of the mechanical part machining data dynamic security protection method of the present application; Figure 2 A lathe machining part schematic diagram of the present application. DETAILED DESCRIPTION

[0017] In order to further illustrate the technical means and effects adopted by the present application to achieve the predetermined purpose, the specific implementation, structure, features and effects of the mechanical part machining data dynamic security protection method according to the present application are described in detail as follows in combination with the drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs.

[0019] The specific scheme of the mechanical part machining data dynamic security protection method provided by the present application is specifically described below in combination with the drawings.

[0020] Please refer to Figure 1 , which shows a flow chart of the steps of the mechanical part machining data dynamic security protection method provided by one embodiment of the present application. The method comprises the following steps: Step S001: Collecting the machining current data sequence of a plurality of mechanical parts.

[0021] It should be noted that the traditional method compresses the current data by using the rotation door compression algorithm. However, the traditional rotation door compression algorithm cannot effectively distinguish abnormal data generated by machine tool vibration and other interference factors, resulting in uniform segmented compression of all abnormal data and normal data, reducing the compression efficiency and reducing the dynamic security of the machining data. Therefore, the present embodiment proposes a mechanical part machining data dynamic security protection method.

[0022] Specifically, in order to realize the mechanical part processing data dynamic security protection method, first need to collect processing current data sequence, the specific process is: using numerical control machine tool to 50 same type of mechanical parts are respectively processed, using current sensor every 1 second to collect a numerical control machine tool in processing each mechanical part when the current data, a total of 100 seconds; With any one mechanical part as an example, arrange all the collected current data in order from small to large according to the collection time, and the arranged sequence is recorded as the processing current data sequence of the mechanical part; Get all the mechanical parts processing current data sequence. Wherein the processing mode of each same type of mechanical part is consistent. In addition, it should be pointed out that the number of mechanical parts, collection interval and total collection time are not specifically limited in the embodiment, and the number of mechanical parts, collection interval and total collection time can be determined according to the specific implementation. Please refer to Figure 2 It shows the schematic diagram of machine tool processing parts, which shows chuck, direction, mechanical parts, rotation direction and turning tool.

[0023] At this point, the processing current data sequence is obtained by the above method.

[0024] Step S002: obtaining the initial standard current factor of each current data according to the change difference of current data between different mechanical parts; according to the maximum change difference of initial standard current factor between current data of different mechanical parts at the same collection time, screening out a number of initial abnormal current data and a number of normal current data from the current data.

[0025] It should be pointed out that when the numerical control machine tool processes the mechanical parts, the material of the mechanical parts, the material of the blade and the vibration of the machine tool will interfere with the friction between the blade and the mechanical parts, so as to change the corresponding current data and produce part of the abnormal current data; Among them, the machine tool vibration is usually produced by the machine tool in the processing process, and it will not have a great impact on the quality of the mechanical parts, which is a common interference factor within the range of normal interference, and the actual reference significance is not great; In order to improve the security protection efficiency of current data, the abnormal data generated by machine tool vibration and the rest of the current data are divided into data segments, and each data segment is compressed respectively to improve the compression efficiency.

[0026] It should be further pointed out that for the same type of mechanical parts, the shapes of these mechanical parts are basically consistent, and the processing mode is also basically consistent, which leads to high similarity of the corresponding current data change. The embodiment can obtain the standard range of different processing time according to the difference between the corresponding current data between different mechanical parts, and determine the initial abnormal current data according to the standard range, so as to facilitate subsequent analysis and processing.

[0027] Specifically, any one mechanical part is designated as the target mechanical part, and every other mechanical part is designated as a reference mechanical part for the target mechanical part; the machining current data sequence of the target mechanical part is used as the reference mechanical part. Taking the current data as an example, based on the machining current data sequence of all reference mechanical parts, the first... The current data is used to obtain the first current data in the machining current data sequence of the target mechanical part. The initial standard current factor of the current data. Wherein, in the machining current data sequence of the target mechanical part... The method for calculating the initial standard current factor for each current data point is as follows: In the formula, This represents the first data point in the sequence of machining current data for the target mechanical part. Initial standard current factor for each current data; This indicates the number of machining current data sequences of all reference mechanical parts for the target mechanical part. In the current data, the first current data in the sequence of machining current data for the target mechanical part is... The number of equal current data; This indicates the number of all reference mechanical parts for the target mechanical part. Where the processing current data sequence of the target mechanical part is... The larger the initial standard current factor of a current data point, the closer the current data point is to the normal range of current data fluctuations.

[0028] Furthermore, the processing current data sequence of each mechanical part is obtained. The initial standard current factor of the current data is used, and the current data corresponding to the largest initial standard current factor is taken as the first current data in the entire processing current data sequence. Standard current data is obtained for each current data point. Standard current data for all current data in the processing current data sequence of each mechanical part is obtained. Taking any current data in the processing current data sequence of any mechanical part as an example, the absolute value of the difference between the standard current data and the current data is recorded as the standard current difference of the current data. The standard current difference of all current data in the processing current data sequence of the mechanical part is obtained. All standard current differences are linearly normalized, and each normalized standard current difference is recorded as the standard current factor.

[0029] Further, a standard current factor threshold T1 is preset, wherein the embodiment takes T1 = 0.5 as an example for description, and the embodiment is not specifically limited, wherein T1 can be determined according to specific implementation; taking any current data in the machining current data sequence of any mechanical part as an example, if the standard current factor of the current data is greater than or equal to T1, the current data is recorded as initial abnormal current data; if the standard current factor of the current data is less than T1, the current data is recorded as normal current data; all initial abnormal current data and all normal current data in the machining current data sequence of the mechanical part are obtained, and all initial abnormal current data and all normal current data in the machining current data sequence of all mechanical parts are obtained.

[0030] At this point, all initial abnormal current data and all normal current data in the machining current data sequence of all mechanical parts are obtained by the above method.

[0031] Step S003: dividing the machining current data sequence into a plurality of abnormal current data segments according to the initial abnormal current data and the normal current data; obtaining a current trend difference factor of each initial abnormal current data according to the change trend of the abnormal current data before and after different initial abnormal current data in the abnormal current data segment; screening a plurality of abnormal current data from the initial abnormal current data according to the current trend difference factor; obtaining data similarity of any two abnormal current data according to the fluctuation trend difference between different abnormal current data in the abnormal current data segment; classifying the abnormal current data according to the data similarity to obtain a plurality of associated abnormal data sets; and obtaining a vibration abnormality factor of each abnormal current data segment according to the distribution of different abnormal current data in the associated abnormal data set.

[0032] It should be noted that in actual situations, if the blade and the mechanical part itself have material problems, the internal impurity material of the blade and the mechanical part is not uniformly distributed, which will not change the friction force generated by the contact between the blade and the mechanical part uniformly, and further, the corresponding generated abnormal current data will have a large fluctuation and a relatively irregular change fluctuation trend; and the vibration of the data machine tool will vibrate at a certain regular frequency, so that the corresponding generated abnormal current data will have a large fluctuation and a relatively regular change fluctuation trend. In order to better protect the current data, the embodiment analyzes the change trend between the initial abnormal current data to obtain the vibration abnormality degree of the abnormal current data, so as to facilitate subsequent analysis and processing.

[0033] Specifically, taking the processing current data sequence of any mechanical part as an example, a preset threshold T2 for the number of initial abnormal current data is set. In this embodiment, T2=20 is used as an example, but this embodiment is not specifically limited, and T2 can be determined according to the specific implementation. In this processing current data sequence, the data segment consisting of every T2 initial abnormal current data is denoted as an abnormal current data segment. Taking any abnormal current data segment as an example, the least squares method is used to fit all the current data in the abnormal current data segment to obtain a fitting curve, and the value of each current data on the fitting curve is denoted as the fitted current value of each current data. Taking any initial abnormal current data in the abnormal current data segment as an example, each initial abnormal current data before the initial abnormal current data is denoted as the preceding abnormal current data, and each initial abnormal current data after the initial abnormal current data is denoted as the following abnormal current data. There may be multiple normal current data in each abnormal current data segment; the least squares method is a well-known technique and will not be described in detail in this embodiment. It should be noted that if the number of remaining initial abnormal current data does not meet the preset T2, the data segment consisting of the remaining initial abnormal current data will be recorded as one abnormal current data.

[0034] Furthermore, based on the fitted current values ​​of the preceding and subsequent abnormal current data of the initial abnormal current data, the initial current trend difference factor of the initial abnormal current data is obtained. The calculation method for the initial current trend difference factor of the initial abnormal current data is as follows: In the formula, This represents the initial current trend difference factor of the initial abnormal current data; This indicates the number of all preceding abnormal current data for this initial abnormal current data; Indicates the first One set of abnormal current data; Indicates the first Fitted current values ​​from the preceding abnormal current data; This indicates the number of all subsequent abnormal current data points for this initial abnormal current data. Indicates the first One post-abnormal current data; Indicates the first Fitted current values ​​of each post-abnormal current data; represents the number of all current data in the abnormal current data segment. Wherein, the greater the initial current trend difference factor of the initial abnormal current data is, the greater the difference between the initial abnormal current data before and after the initial abnormal current data and the value under the ideal trend is, reflecting that the initial abnormal current data is more likely to belong to the current data with large changes in actual trend. The initial current trend difference factor of all initial abnormal current data in the abnormal current data segment is obtained, and all initial current trend difference factors are linearly normalized. Each normalized initial current trend difference factor is recorded as a current trend difference factor.

[0035] Further, a current trend difference factor threshold T3 is preset, wherein T3=0.6 is taken as an example in the embodiment, and the embodiment is not specifically limited, wherein T3 can be determined according to specific implementation; the initial abnormal current data with a current trend difference factor greater than T3 is recorded as abnormal current data, and all abnormal current data in the abnormal current data segment are obtained; taking any two adjacent abnormal current data in the abnormal current data segment as an example, in the two abnormal current data, the difference between the second abnormal current data and the first abnormal current data is recorded as a first difference, the difference between the sequence numbers of the second abnormal current data and the first abnormal current data in the abnormal current data segment is recorded as a second difference, and the ratio of the first difference to the second difference is recorded as the slope of the first abnormal current data. The slopes of all abnormal current data in the abnormal current data segment are obtained.

[0036] Further, taking any three adjacent abnormal current data in the abnormal current data segment as an example, in the three abnormal current data, the absolute value of the difference between the slopes of the first abnormal current data and the second abnormal current data is recorded as the pre-trend difference value of the second abnormal current data, and the absolute value of the difference between the slopes of the third abnormal current data and the second abnormal current data is recorded as the post-trend difference value of the second abnormal current data; the pre-trend difference values and the post-trend difference values of all abnormal current data are obtained. It should be noted that the pre-trend difference value of the first abnormal current data in the abnormal current data segment is 0 by default in the embodiment, and the post-trend difference value of the last abnormal current data in the abnormal current data segment is 0 by default in the embodiment.

[0037] Further, the data similarity between the first abnormal current data and the second abnormal current data in the abnormal current data segment is obtained according to the difference between the pre-trend difference value and the post-trend difference value between the first abnormal current data and the second abnormal current data. ​​​​​The method for calculating the data similarity of abnormal current data is as follows: In the formula, Indicates the first The abnormal current data and the first Data similarity of abnormal current data, ; This represents the preset hyperparameters; in this embodiment, the preset hyperparameters are... This is used to limit the range of variation of the preceding trend difference value and the following trend difference value; Indicates the first The preceding trend difference value of each abnormal current data; Indicates the first The preceding trend difference value of each abnormal current data; This represents the preset hyperparameters; in this embodiment, the preset hyperparameters are... This is used to limit the range of variation in abnormal current data; Indicates the first One abnormal current data point; Indicates the first One abnormal current data point; Indicates the first Post-trend difference value of abnormal current data; Indicates the first Post-trend difference value of abnormal current data; Indicates restriction The condition for taking the value 1; This represents a sign function. Where if... Indicates the first The abnormal current data and the first The greater the data similarity of the abnormal current data, the more it indicates that the first abnormal current data... The abnormal current data and the first The more similar the trend and numerical variation patterns among the abnormal current data, the better.

[0038] Furthermore, any one abnormal current data point in the abnormal current data segment is designated as the target abnormal current data point. Each abnormal current data point other than the target abnormal current data point is designated as a reference abnormal current data point for the target abnormal current data point. Reference abnormal current data points with a similarity of 1 to the target abnormal current data point are designated as associated abnormal current data points for the target abnormal current data point. The dataset comprised of all associated abnormal current data points is then used as the associated abnormal dataset for the target abnormal current data point. The associated abnormal datasets for all abnormal current data points in the abnormal current data segment are then obtained. Each abnormal current data segment corresponds to multiple associated abnormal datasets.

[0039] Furthermore, the vibration anomaly factor of this abnormal current data segment is obtained based on the associated anomaly dataset of all abnormal current data in this abnormal current data segment. The calculation method for the vibration anomaly factor of this abnormal current data segment is as follows: In the formula, This indicates the degree of vibration abnormality in the abnormal current data segment; This represents the average of all abnormal current data in this abnormal current data segment; This represents the average value of all current data in this abnormal current data segment; This indicates the number of all abnormal current data in this abnormal current data segment; This indicates the first abnormal data set in the associated abnormal data set of all abnormal current data in this abnormal current data segment. The number of times each abnormal current data point occurs; Indicates the first The number of all abnormal current data points in the associated abnormal dataset for each abnormal current data point is determined. A higher vibration anomaly level in a given abnormal current data segment indicates that the segment is less likely to be caused by machine tool vibration. The vibration anomaly level of all abnormal current data segments is obtained, and all vibration anomaly levels are linearly normalized. Each normalized vibration anomaly level is then denoted as a vibration anomaly factor.

[0040] Thus, the vibration anomaly factors for all abnormal current data segments are obtained using the above method.

[0041] Step S004: Compress the processing current data sequence according to the vibration anomaly factor.

[0042] Specifically, a threshold value for a vibration anomaly factor is preset. In this embodiment, This example is used for illustration; no specific limitations are set in this embodiment. This can be determined based on the specific implementation; taking any abnormal current data segment as an example, the first and second current data points in the abnormal current data segment are both recorded as a dividing point, and the third current data point in the abnormal current data segment other than the first and second current data points is used as the dividing point. Taking the current data as an example, according to the first current data... The difference between the current data and other boundary points is obtained from the first... The degree of boundary of each current data point. Among them, the first... The method for calculating the boundary degree of individual current data is as follows: In the formula, Indicates the first The degree of boundary between current data; This represents the preset hyperparameters; in this embodiment, the preset hyperparameters are... , used to limit the range of slope variation; Indicates the first The slope of the first dividing point before each current data point; Indicates the first The slope of the second dividing point before the current data; This indicates the vibration anomaly factor for the abnormal current data segment; This represents the preset threshold for the vibration anomaly factor; Indicates the first The first dividing point before the current data and the first... The maximum upslope of all current data in a data segment consisting of current data; Indicates the first The first dividing point before the current data and the first... The minimum downslope of all current data within a data segment composed of current data points is used to determine the boundary degree of all current data within that abnormal current data segment. The determination of the upslope and downslope is well-known in the rotating door compression algorithm and will not be elaborated upon in this embodiment.

[0043] Furthermore, a threshold value T5 is preset, where this embodiment uses T5=0 as an example. This embodiment does not impose specific limitations, and T5 can be determined according to the specific implementation. Current data in the abnormal current data segment with a threshold value greater than T5 are recorded as threshold points. All threshold points in the abnormal current data segment are obtained. The data segment formed between any two threshold points is recorded as a data segment to be compressed. Each data segment to be compressed is treated as a sub-segment, and rotating door compression is performed on all sub-segments to obtain several compressed data segments. All compressed data segments are stored in the database. The process of compressing sub-segments is a well-known part of the rotating door compression algorithm and will not be elaborated upon in this embodiment.

[0044] This concludes the embodiment.

[0045] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for dynamically securing machining data of a mechanical part, characterized in that, The method comprises the following steps: Collecting a machining current data sequence of a plurality of mechanical parts, the machining current data sequence comprising a plurality of current data, each current data corresponding to a collection time; Obtaining an initial standard current factor of each current data according to the variation difference of the current data between different mechanical parts, the initial standard current factor being used to describe the difference between the current data and normal current data; screening a plurality of initial abnormal current data and a plurality of normal current data from the current data according to the maximum variation difference of the initial standard current factors between the current data of different mechanical parts at the same collection time; Dividing the machining current data sequence into a plurality of abnormal current data segments according to the initial abnormal current data and the normal current data; obtaining a current trend difference factor of each initial abnormal current data according to the variation trend of the abnormal current data before and after the different initial abnormal current data in the abnormal current data segment; screening a plurality of abnormal current data from the initial abnormal current data according to the current trend difference factor; obtaining data similarity of any two abnormal current data according to the fluctuation trend difference between the different abnormal current data in the abnormal current data segment; classifying the abnormal current data according to the data similarity to obtain a plurality of associated abnormal data sets; obtaining a vibration abnormality factor of each abnormal current data segment according to the distribution of the different abnormal current data in the associated abnormal data set; Compressing the machining current data sequence according to the vibration abnormality factor.

2. The method of claim 1, wherein the method further comprises: The specific method for obtaining the initial standard current factor of each current data according to the variation difference of the current data between different mechanical parts comprises: Taking any one mechanical part as a target mechanical part, and taking each mechanical part other than the target mechanical part as a reference mechanical part of the target mechanical part; In the formula, represents an initial standard current factor of the initial current data in the machining current data sequence of the target mechanical part; represents the number of the current data equal to the initial current data in the machining current data sequence of the target mechanical part in the machining current data sequence of all the reference mechanical parts of the target mechanical part; represents the number of all the reference mechanical parts of the target mechanical part.​​​ 3. The method of claim 1, wherein the method further comprises: The specific method for screening a plurality of initial abnormal current data and a plurality of normal current data from the current data according to the maximum variation difference of the initial standard current factors between the current data of different mechanical parts at the same collection time comprises: In the initial standard current factors of the first current data on the machining current data sequence of all mechanical parts, the current data corresponding to the maximum initial standard current factor is taken as the standard current data of the first current data on the machining current data sequence of all mechanical parts; the standard current data of all current data on the machining current data sequence of each mechanical part is obtained. For any one current data in the machining current data sequence of any one mechanical part, taking the absolute value of the difference between the standard current data of the current data and the current data as the standard current difference amount of the current data, obtaining the standard current difference amount of all current data in the machining current data sequence of the mechanical part, linearly normalizing all standard current difference amounts, and taking each normalized standard current difference amount as a standard current factor; Pre-setting a standard current factor threshold T1, for any one current data in the machining current data sequence of any one mechanical part, if the standard current factor of the current data is greater than or equal to T1, then the current data is recorded as an initial abnormal current data; If the standard current factor of the current data is less than T1, then the current data is recorded as a normal current data.

4. The method of claim 1, wherein the method further comprises: The specific method for dividing the machining current data sequence into a plurality of abnormal current data segments according to the initial abnormal current data and the normal current data comprises: For any one of the machining current data sequence of mechanical parts, an initial abnormal current data quantity threshold T2 is preset; in the machining current data sequence, a data segment composed of every T2 initial abnormal current data is recorded as an abnormal current data segment.

5. The method of claim 1, wherein the method further comprises: The specific method for obtaining the current trend difference factor of each initial abnormal current data according to the change trend of the abnormal current data before and after the different initial abnormal current data in the abnormal current data segment comprises: For any one of the abnormal current data segment, the least square method is used to fit all the current data in the abnormal current data segment to obtain a fitting curve, and the value of each current data on the fitting curve is recorded as the fitting current value of each current data; for any one of the initial abnormal current data in the abnormal current data segment, each initial abnormal current data before the initial abnormal current data in the abnormal current data segment is recorded as the pre-abnormal current data of the initial abnormal current data, and each initial abnormal current data after the initial abnormal current data is recorded as the post-abnormal current data of the initial abnormal current data; wherein represents an initial current trend difference factor of the initial abnormal current data; represents a number of all preceding abnormal current data of the initial abnormal current data; represents the preceding abnormal current data; represents the preceding abnormal current data; represents a number of all succeeding abnormal current data of the initial abnormal current data; represents the succeeding abnormal current data; represents the succeeding abnormal current data; represents a number of all current data in the abnormal current data segment; The initial current trend difference factors of all the initial abnormal current data in the abnormal current data segment are obtained, and all the initial current trend difference factors are linearly normalized, and each normalized initial current trend difference factor is recorded as a current trend difference factor.

6. The method of claim 1, wherein the method further comprises: The specific method for screening a plurality of abnormal current data from the initial abnormal current data according to the current trend difference factor comprises: A current trend difference factor threshold T3 is preset, and the initial abnormal current data with a current trend difference factor greater than T3 is recorded as abnormal current data.

7. The method of claim 1, wherein the method further comprises: The specific method for obtaining the data similarity of any two abnormal current data according to the fluctuation trend difference between different abnormal current data in the abnormal current data segment comprises: For any two adjacent abnormal current data in any one of the abnormal current data segments, the difference between the second abnormal current data and the first abnormal current data is recorded as a first difference, the difference between the sequence numbers of the second abnormal current data and the first abnormal current data in the abnormal current data segment is recorded as a second difference, and the ratio of the first difference to the second difference is recorded as the slope of the first abnormal current data; For any three adjacent abnormal current data in the abnormal current data segment, the absolute value of the difference between the slopes of the first abnormal current data and the second abnormal current data is recorded as the pre-trend difference value of the second abnormal current data, and the absolute value of the difference between the slopes of the third abnormal current data and the second abnormal current data is recorded as the post-trend difference value of the second abnormal current data. In the formula, represents the data similarity of the first abnormal current data and the second abnormal current data, ; represents a preset hyperparameter; represents the preceding trend difference value of the first abnormal current data; represents the preceding trend difference value of the first abnormal current data; represents a preset hyperparameter; represents the first abnormal current data; represents the first abnormal current data; represents the following trend difference value of the first abnormal current data; represents the following trend difference value of the first abnormal current data; represents a sign function.

8. The method of claim 1, wherein the method further comprises: The specific method for obtaining a plurality of associated abnormal data sets by classifying the abnormal current data according to the data similarity comprises: A similarity threshold KL is preset, any one of the abnormal current data in any one of the abnormal current data segments is recorded as target abnormal current data, each abnormal current data except the target abnormal current data is recorded as reference abnormal current data of the target abnormal current data, the reference abnormal current data with data similarity equal to KL with the target abnormal current data is recorded as associated abnormal current data of the target abnormal current data, and a data set composed of all the associated abnormal current data is taken as associated abnormal data set of the target abnormal current data.

9. The method of claim 1, wherein the method further comprises: The specific method for obtaining the vibration abnormality factor of each abnormal current data segment according to the distribution of different abnormal current data in the associated abnormal data set comprises the following steps: In the formula, It indicates the degree of vibration abnormality in any abnormal current data segment; This represents the mean of all abnormal current data in the abnormal current data segment. This represents the mean of all current data in the abnormal current data segment. This indicates the number of all abnormal current data in the abnormal current data segment; This represents the associated abnormal dataset of all abnormal current data in the abnormal current data segment, specifically the first abnormal data. The number of times each abnormal current data point occurs; Indicates the first The number of all abnormal current data in the associated abnormal dataset of each abnormal current data is obtained; the vibration abnormality degree of all abnormal current data segments is obtained, and all vibration abnormality degrees are linearly normalized. Each normalized vibration abnormality degree is recorded as a vibration abnormality factor.

10. The method of claim 6, wherein the method further comprises: The specific method for compressing the machining current data sequence according to the vibration abnormality factor comprises the following steps: A preset vibration anomaly factor threshold value is set ; for any one abnormal current data segment, the first current data and the second current data in the abnormal current data segment are recorded as a demarcation point, and the first current data and the second current data in the abnormal current data segment are recorded as a demarcation point. ​ In the formula, Indicates the first The degree of boundary between individual current data; Indicates the preset hyperparameters; Indicates the first The slope of the first dividing point before each current data point; Indicates the first The slope of the second dividing point before the current data; The vibration anomaly factor represents the abnormal current data segment; This represents the preset threshold for the vibration anomaly factor; Indicates the first The first dividing point before the current data and the first... The maximum upslope of all current data in a data segment consisting of current data; Indicates the first The first dividing point before the current data and the first... The minimum downslope of all current data in a data segment consisting of current data; A demarcation degree threshold T5 is preset; the current data with demarcation degree greater than T5 in the abnormal current data segment is recorded as a demarcation point; All the demarcation points in the abnormal current data segment are obtained, all the demarcation points in all the abnormal current data segments are obtained; the data segment formed between any two demarcation points is recorded as a to-be-compressed data segment; each to-be-compressed data segment is taken as a segment, and all the segments are subjected to a rotating door compression to obtain a plurality of compressed data segments, and all the compressed data segments are stored in a database.

Images