Battery production line operation status monitoring method and system based on big data
By introducing the calculation of the degree and density difference between segmentation characterization and density in the operating status monitoring of the battery production line, selecting the target dimension and building an isolated forest, the problem of the isolated forest algorithm being difficult to screen abnormal data points in the monitoring of the battery production line is improved, and the detection accuracy and efficiency.
Patent Information
- Application Number
- CN202411413871.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-11
- Publication Date
- 2025-01-03
- Estimated Expiration
- 2044-10-11
AI Technical Summary
In the monitoring of battery production line operation status, the isolated forest algorithm is difficult to effectively screen out the dimensions of abnormal data points, resulting in a decrease in detection accuracy and the inability to accurately segment abnormal data points and normal data points.
The degree of segmentation characterization is introduced to select the target dimension, and the degree of density difference calculation is calculated by principal component analysis and the degree of density difference is constructed, solitary forests are constructed to achieve anomaly detection, and the tree depth threshold is controlled to avoid excessive or undersegment.
The segmentation accuracy and efficiency of abnormal data points and normal data points are improved, and the efficiency and accuracy problems caused by excessive or undersegment are prevented, thus achieving more efficient abnormal detection.
Smart Images
Figure CN118916797B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of data processing, and in particular to a method and system for monitoring the operating status of a battery production line based on big data. Background Art
[0002] With the development of science and technology, the degree of automation of battery production lines is higher. In order to ensure the normal operation of battery production lines, the operating status of battery production lines needs to be automatically monitored.
[0003] When the battery production line runs abnormally, the operating status data of the battery production line will be abnormal. For example, when the inventory of the buffer area at a station of the battery production line is small, it is very likely that the station is forced to stop working due to lack of parts; when the inventory of the buffer area at a station of the battery production line is large, it means that the processing efficiency of the station is low, thereby increasing unnecessary inventory costs. Therefore, it is necessary to detect abnormalities in the operating status data of the battery production line.
[0004] The isolation forest algorithm has the advantages of low time complexity and high applicability, so the algorithm is often used for anomaly detection of big data. The process of anomaly detection by the isolation forest algorithm is: first, a value is randomly selected from the value interval of a dimension data as the partition value, and then the data points greater than the partition value are segmented together, and the data points less than or equal to the partition value are segmented together, and anomaly detection is achieved through continuous segmentation iteration; wherein the abnormal data points will deviate from the normal data, and the abnormal data points can be segmented in an independent interval through a small number of segmentations, so the algorithm is based on this principle to achieve abnormal data detection. The operating status data points of the production line collected at each moment will have data of multiple dimensions, and since the abnormal data points only have deviation characteristics in certain dimensions, the partition value obtained on the dimension with small deviation cannot better segment the abnormal data, and the segmentation processing using the dimension with small deviation will only increase the abnormal detection time, will not improve the detection accuracy, but will reduce the detection accuracy to a certain extent. Therefore, how to select a dimension that is easier to filter out abnormal data points from many dimensions has become the research focus of the present invention.
[0005] The patent application document with publication number CN116192531A discloses a log anomaly detection system based on isolation forest. The method in the document improves the accuracy of the detection system by combining the isolation forest algorithm with pattern matching technology. The method in the document does not consider the dimension screening problem, so the method in the document cannot be used to monitor the operation status of the battery production line. Summary of the invention
[0006] In order to solve the problem of how to select a dimension that is easier to filter out abnormal data points from among many dimensions, the present invention provides a battery production line operation status monitoring method and system based on big data.
[0007] In the first aspect, the present invention provides a battery production line operation status monitoring method based on big data, which adopts the following technical solutions:
[0008] A battery production line operation status monitoring method based on big data includes the following steps:
[0009] Acquire several historical operation status data points of a battery production line, wherein the historical operation status data points include data of multiple dimensions;
[0010] All historical operating status data points are randomly sampled to obtain an analysis data set, and principal component analysis is performed on the data points in the analysis data set to obtain several eigenvalues and the principal component direction corresponding to each eigenvalue; the segmentation representation degree of the dimension is calculated: , Indicates eigenvalues, Indicates The angle between the perpendicular direction of the principal component direction corresponding to the eigenvalue and the dimension, Indicates that all data in the analysis dataset are in The density difference of the projection values in the direction perpendicular to the principal component direction corresponding to the eigenvalues, represents the number of eigenvalues;
[0011] All dimensions are clustered into two categories according to the degree of segmentation representation, and the dimension in the category with a large mean value of the segmentation representation degree is used as the target dimension; the target dimension is used as the dimension used when the isolation forest algorithm performs segmentation analysis, and the analysis data set is analyzed and processed to obtain an isolated tree, and the isolated trees obtained from all the analysis data sets are used to form an isolation forest;
[0012] Isolation forest is used to perform anomaly detection on the current operating status data to realize production line operating status monitoring.
[0013] The present invention reflects the difficulty of separating abnormal data points from normal data points in the data of each dimension by introducing the segmentation representation degree, and then controls the dimension selection by the segmentation representation degree, so that the dimension selected based on this method can separate abnormal data points from normal data points more efficiently and accurately; further, when calculating the segmentation representation degree, by introducing the eigenvalues of each principal component direction and the angle between each dimension and the principal component direction, the abnormal features and normal features of each dimension at a unit distance are accurately reflected, so as to accurately reflect the sensitivity of each dimension to distinguishing abnormal data points from normal data points; at the same time, when calculating the segmentation representation degree, the density difference degree is also introduced to reflect the deviation difference of the data of each dimension from the principal component direction, so as to accurately reflect the difficulty of separating data points by distance; by combining the eigenvalues of each principal component direction, the angle between each dimension and the principal component direction and the density difference degree, the difficulty of separating normal data points from abnormal data points based on the data of each dimension is accurately reflected.
[0014] Preferably, the method for obtaining the density difference degree includes:
[0015] Obtaining the density of the projection value;
[0016] The density of all projection values is clustered into two categories;
[0017] The absolute value of the difference between the mean of all densities in one category and the mean of all densities in another category is taken as the degree of density difference.
[0018] The present invention accurately reflects the density difference between data points by analyzing the density difference between different categories, and provides a basis for accurately analyzing the difficulty of segmenting abnormal data points and normal data points.
[0019] Preferably, obtaining the density of the projection value includes:
[0020] The absolute values of the differences between the projection value and the left adjacent projection value and the right adjacent projection value are respectively obtained, and the reciprocal of the average of the absolute values of the differences between the left adjacent projection value and the right adjacent projection value is taken as the density of the projection value.
[0021] The present invention reflects the density of the projection value by the distance between the projection value and the adjacent projection values on both sides, and the density calculation method is simpler and faster.
[0022] Preferably, the step of using the target dimension as the dimension used in the segmentation analysis by the isolation forest algorithm, and analyzing and processing the analysis data set to obtain an isolation tree comprises:
[0023] The isolation forest algorithm is used to construct an isolation tree for the analysis data set. In each iteration of the isolation tree construction, the bottom node is obtained, a dimension is randomly selected from all target dimensions as a reference dimension, and data randomly selected from the value range of all data points in the bottom node on the reference dimension is used as segmentation data. The data points in the bottom node are divided into two parts using the segmentation data, which are respectively used as the left subtree and the right subtree of the bottom node.
[0024] In response to the tree depth being greater than or equal to the tree depth threshold, the iteration ends to obtain an isolated tree.
[0025] When using the isolation forest algorithm to perform anomaly segmentation, the present invention improves the accuracy and efficiency of anomaly detection by selecting a dimension that can segment abnormal data points more quickly.
[0026] Preferably, the tree depth threshold acquisition method includes:
[0027] The inverse of the mean of the segmentation representation degree of all target dimensions is multiplied by the preset initial tree depth threshold to obtain the rounded-up value of the data as the tree depth threshold.
[0028] The present invention controls the tree depth threshold by the segmentation representation degree, thereby preventing the problem of low segmentation efficiency caused by over-segmentation, and also preventing the problem of segmentation accuracy angle caused by under-segmentation.
[0029] Preferably, the random sampling of all historical operating status data points to obtain an analysis data set includes:
[0030] A random sampling process is performed on all historical operation status data points, and a data set consisting of historical operation status data points obtained by random sampling is recorded as an analysis data set, and several analysis data sets are obtained by sampling multiple times.
[0031] Preferably, the method of using the isolation forest to perform anomaly detection processing on the operating status data at the current moment to realize production line operating status monitoring includes:
[0032] The current operating status data is input into the isolation forest to obtain the abnormality judgment result. When the abnormality judgment result is abnormal, it is determined that the production line is operating abnormally and an early warning is issued; when the abnormality judgment result is normal, it is determined that the production line is operating normally and no early warning is issued.
[0033] In the second aspect, the present invention provides a battery production line operation status monitoring system based on big data, which adopts the following technical solutions:
[0034] A battery production line operation status monitoring system based on big data includes: a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned battery production line operation status monitoring method based on big data is implemented.
[0035] By adopting the above technical solution, the above-mentioned battery production line operation status monitoring method based on big data is generated into a computer program and stored in a memory to be loaded and executed by a processor, so that a terminal device is made according to the memory and the processor for easy use.
[0036] The present invention has the following technical effects:
[0037] The present invention introduces the segmentation representation degree to reflect the difficulty of segmenting abnormal data points from normal data points in each dimension, and then controls the dimension selection by the segmentation representation degree, so that the dimension selected based on this method can more efficiently and accurately segment abnormal data points from normal data points;
[0038] Furthermore, the tree depth threshold is controlled by the degree of segmentation representation, thereby preventing the problem of low segmentation efficiency caused by over-segmentation and the problem of segmentation accuracy angle caused by under-segmentation.
[0039] Furthermore, when calculating the degree of segmentation representation, the eigenvalues of each principal component direction and the angle between each dimension and the principal component direction are introduced to accurately reflect the distinction between abnormal features and normal features of each dimension at a unit distance, thereby accurately reflecting the sensitivity of each dimension to distinguishing abnormal data points from normal data points; at the same time, when calculating the degree of segmentation representation, the density difference degree is also introduced to reflect the deviation difference of the data of each dimension from the principal component direction, thereby accurately reflecting the difficulty of separating the data points by distance; the eigenvalues of each principal component direction, the angle between each dimension and the principal component direction and the density difference degree are combined to accurately reflect the difficulty of separating normal data points from abnormal data points based on the data of each dimension. BRIEF DESCRIPTION OF THE DRAWINGS
[0040] By reading the detailed description below with reference to the accompanying drawings, the above and other purposes, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-restrictive manner, and the same or corresponding reference numerals represent the same or corresponding parts.
[0041] Figure 1 It is a flow chart of a method in a battery production line operation status monitoring method based on big data in an embodiment of the present invention. DETAILED DESCRIPTION
[0042] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.
[0043] It should be understood that when the terms "first", "second", etc. are used in the claims, descriptions, and drawings of the present invention, they are only used to distinguish different objects, rather than to describe a specific order. The terms "include" and "comprise" used in the description and claims of the present invention indicate the presence of the described features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their collections.
[0044] The embodiment of the present invention discloses a battery production line operation status monitoring method based on big data, referring to Figure 1 , comprising steps S1 to S4:
[0045] S1: Acquire several historical operating status data points of a battery production line, wherein the historical operating status data points include data of multiple dimensions.
[0046] Specifically, each type of operating status data of the battery production line collected historically at each moment is obtained, and the data points consisting of all types of operating status data of the battery production line collected historically at the same moment are recorded as historical operating status data points. The types of operating status data include but are not limited to the following aspects: the amount of goods buffered at each workstation, the processing time of each process at each workstation, the yield rate, the equipment utilization rate, and the productivity.
[0047] S2: Randomly sample all historical operating status data points to obtain an analysis data set, perform principal component analysis on the data points in the analysis data set to obtain a number of eigenvalues and the principal component direction corresponding to each eigenvalue; and calculate the segmentation representation degree of the dimension.
[0048] S20: Randomly sample all historical operating status data points to obtain an analysis data set.
[0049] It should be noted that when using the isolation forest algorithm to construct an isolation tree, it is necessary to first randomly select some data points from all historical operating status data points to construct the isolation tree.
[0050] Optionally, as an example, all historical operating status data points are randomly sampled to obtain an analysis data set, including:
[0051] A random sampling process is performed on all historical operation status data points, and a data set consisting of historical operation status data points obtained by random sampling is recorded as an analysis data set, and several analysis data sets are obtained by sampling multiple times.
[0052] S21: performing principal component analysis on the data points in the analysis data set to obtain a number of eigenvalues and a principal component direction corresponding to each eigenvalue.
[0053] It should be noted that the degree to which abnormal data points deviate from normal data in each dimension is different; in order to better segment the abnormal data points, it is necessary to select the dimension with a greater degree of deviation.
[0054] It should be further explained that since it is impossible to distinguish normal data points from abnormal data points, it is impossible to directly determine the deviation of abnormal data points from normal data points in each dimension. Since the data of different dimensions of the production line are correlated, for example, when the productivity is high, the processing time of each station and each process is shorter, so the productivity and processing time have a certain correlation. Normal data points have certain distribution characteristics in two dimensions, and abnormal data points will deviate from this distribution in two dimensions, so the deviation of abnormal data points from normal data points in each dimension can be analyzed based on this information. The principal component analysis algorithm can parse out the correlation distribution characteristics of data in different dimensions, so the principal component analysis algorithm can be used to assist in analyzing the deviation of abnormal data points from normal data points.
[0055] Optionally, as an example, principal component analysis is performed on the data in the analysis data set to obtain a number of eigenvalues and a principal component direction corresponding to each eigenvalue, including:
[0056] The principal component analysis algorithm is used to perform principal component analysis on all data points in the analysis data set to obtain several eigenvalues and the principal component direction corresponding to each eigenvalue.
[0057] It should be noted that using a principal component analysis algorithm to perform principal component analysis on all data points in an analysis data set to obtain a number of eigenvalues and a principal component direction corresponding to each eigenvalue is a prior art and will not be described in detail here.
[0058] S22: Calculate the segmentation representation degree of the dimension.
[0059] Preferably, as an example, calculating the segmentation representation degree of the dimension includes:
[0060]
[0061] in, Indicates An eigenvalue reflects the distribution of all data points in the analysis data set along the principal component direction corresponding to the eigenvalue. The principal component direction corresponding to the eigenvalue reflects the association between data of different dimensions. The larger the value, the more data points in the analysis data set are distributed along the principal component direction corresponding to the eigenvalue. If a data point deviates from the principal component direction, it means that the data point is more likely to be an abnormal data point. Indicates The angle between the vertical direction of the principal component direction corresponding to the eigenvalue and the dimension. This value reflects the correlation between the dimension and the principal component direction. The larger the value, the smaller the correlation between the dimension and the principal component direction. equal , indicating that the dimension has no correlation with the principal component direction, so The closer This indicates that the dimension has less correlation with the direction of the principal component; This value reflects the unit distance representation of abnormal features and normal features of the data on the dimension. The larger the value, the smaller the distance feature of the data on the dimension is, the more abnormal data points and normal data points can be distinguished. Indicates that all data points in the analysis data set are in The density difference of the projection values in the direction perpendicular to the principal component direction corresponding to the eigenvalues reflects the difference in the deviation of all data points in the analysis data set from the principal component direction. The larger the value, the easier it is to separate data points with different degrees of deviation. It reflects the difficulty of separating abnormal data points from normal data points in the data of the dimension. The larger the value, the easier it is to separate abnormal data points from normal data points. Indicates the degree of segmentation representation of the dimension.
[0062] The above embodiments involve the degree of density difference. The method for determining the degree of density difference is described below:
[0063] Obtaining the absolute values of the differences between the projection value and the left adjacent projection value and the right adjacent projection value respectively, and taking the reciprocal of the average of the absolute values of the differences between the projection value and the left adjacent projection value and the right adjacent projection value as the density of the projection value;
[0064] The density of all projection values is clustered into two categories;
[0065] The absolute value of the difference between the mean of all densities in one category and the mean of all densities in another category is taken as the degree of density difference.
[0066] S3: All dimensions are clustered into two categories according to the degree of segmentation representation, and the dimension in the category with a large mean value of the segmentation representation degree is used as the target dimension; the target dimension is used as the dimension used when the isolation forest algorithm performs segmentation analysis, and the analysis data set is analyzed and processed to obtain an isolated tree, and the isolated trees obtained from all analysis data sets are used to form an isolation forest.
[0067] S30: All dimensions are clustered into two categories according to the segmentation representation degree, and the dimension in the category with a larger mean value of the segmentation representation degree is taken as the target dimension.
[0068] It should be noted that in order to better separate abnormal data points from normal data points, it is necessary to select a dimension that can easily distinguish normal data points from abnormal data points. From step S2, it can be seen that the degree of segmentation representation can reflect the ability of data in each dimension to distinguish normal data points from abnormal data points, so the dimension selection can be based on the degree of segmentation representation.
[0069] Optionally, as an example, all dimensions are clustered into two categories according to the segmentation representation degree, and the dimension in the category with a large mean value of the segmentation representation degree is used as the target dimension, including:
[0070] The segmentation representation degree is used as the clustering index, and the K-mean clustering algorithm is used to cluster all dimensions into two categories. The mean of the segmentation representation degree of all dimensions in each category is calculated, and the dimension in the category with the largest mean of the segmentation representation degree is taken as the target dimension.
[0071] S31: Using the target dimension as the dimension used for segmentation analysis using the isolation forest algorithm, and analyzing and processing the analysis data set to obtain an isolation tree.
[0072] Preferably, as an example, the target dimension is used as the dimension used when the isolation forest algorithm performs segmentation analysis, and the analysis data set is analyzed and processed to obtain an isolation tree, including:
[0073] The isolation forest algorithm is used to construct an isolation tree for the analysis data set. In each iteration of the isolation tree construction, the bottom node is obtained, a dimension is randomly selected from all target dimensions as a reference dimension, and data randomly selected from the value range of all data points in the bottom node on the reference dimension is used as segmentation data. The data points in the bottom node are divided into two parts using the segmentation data, which are respectively used as the left subtree and the right subtree of the bottom node.
[0074] In response to the tree depth being greater than or equal to the tree depth threshold, the iteration ends to obtain an isolated tree.
[0075] The above embodiments involve a tree depth threshold. The method for determining the tree depth threshold is described below:
[0076] The inverse of the mean of the segmentation representation degree of all target dimensions is multiplied by the preset initial tree depth threshold to obtain the rounded-up value of the data as the tree depth threshold. , It represents the number of analyzed data sets, and other embodiments may take other values, which are not specifically limited in this embodiment. The preset initial tree depth threshold is generally the tree depth threshold commonly used in the isolation forest algorithm.
[0077] It is understandable that the degree of segmentation representation can reflect the difficulty of segmenting abnormal data points and normal data points to a certain extent. The tree depth threshold in the isolation forest algorithm is used to control the segmentation refinement. If the segmentation difficulty of abnormal data points and normal data points is relatively small, the abnormal data points and normal data points can be separated by a few rough segmentations. If the segmentation difficulty of abnormal data points and normal data points is relatively large, multiple fine segmentations are required to separate the abnormal data points and normal data points. Therefore, setting the tree depth threshold by the degree of segmentation representation can not only save segmentation time, but also more accurately separate abnormal data points and normal data points.
[0078] S32: The isolated trees obtained from all analyzed data sets form an isolation forest.
[0079] Optionally, as an example, an isolation forest is constructed using isolated trees from all analyzed datasets, including:
[0080] According to the method in steps S2 to S31, an isolation tree is constructed for each analysis data set, and a forest consisting of isolated trees constructed from all analysis data sets is used as an isolation forest.
[0081] S4: Use the isolation forest to perform anomaly detection on the current operating status data to achieve production line operating status monitoring.
[0082] It should be noted that in order to determine the operating status of the battery production line at the current moment, the operating status of the battery production line at the current moment needs to be monitored.
[0083] Isolation forest is used to perform anomaly detection on the current operating status data to achieve production line operating status monitoring, including:
[0084] The current operating status data is input into the isolation forest to obtain the abnormality judgment result. When the abnormality judgment result is abnormal, it is determined that the production line is operating abnormally and an early warning is issued; when the abnormality judgment result is normal, it is determined that the production line is operating normally and no early warning is issued.
[0085] An embodiment of the present invention also discloses a battery production line operation status monitoring system based on big data, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a battery production line operation status monitoring method based on big data according to the present invention is implemented.
[0086] The above system also includes other components well known to those skilled in the art, such as a communication bus and a communication interface, and their configuration and functions are known in the art, so they will not be described in detail here.
[0087] In the present invention, the aforementioned memory may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium may be any suitable magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory, a dynamic random access memory, a static random access memory, an enhanced dynamic random access memory, a high bandwidth memory, a hybrid storage cube, etc., or any other medium that can be used to store the required information and can be accessed by an application, a module, or both. Any such computer storage medium may be part of the device or accessible or connectable to the device.
[0088] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.
[0089] The above are all preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the protection scope of the present invention.
Claims
1. A battery production line operation status monitoring method based on big data, characterized in that: Includes steps: Acquire several historical operation status data points of a battery production line, wherein the historical operation status data points include data of multiple dimensions; All historical operating status data points are randomly sampled to obtain an analysis data set, and principal component analysis is performed on the data points in the analysis data set to obtain several eigenvalues and the principal component direction corresponding to each eigenvalue; the segmentation representation degree of the dimension is calculated: , Indicates eigenvalues, Indicates The angle between the perpendicular direction of the principal component direction corresponding to the eigenvalue and the dimension, Indicates that all data in the analysis dataset are in The density difference of the projection values in the direction perpendicular to the principal component direction corresponding to the eigenvalues, represents the number of eigenvalues; The method for obtaining the density difference degree includes: The inverse of the mean of the absolute value of the difference between the projection value and the adjacent projection value on the left and the absolute value of the difference between the projection value and the adjacent projection value on the right is used as the density of the projection value; the densities of all projection values are clustered into two categories; the absolute value of the difference between the density means of the two categories is used as the degree of density difference; All dimensions are clustered into two categories according to the degree of segmentation representation, and the dimensions in the category with the larger mean value of the segmentation representation degree are taken as the target dimensions; the inverse of the mean value of the segmentation representation degree of all target dimensions multiplied by the preset initial tree depth threshold is used as the rounded-up value of the data as the tree depth threshold; the isolation forest algorithm is used to construct an isolated tree for the analysis data set, and in each iteration of the construction of the isolated tree, the bottom node is obtained, a dimension is randomly selected from all target dimensions as the reference dimension, and any data within the value range of the data on the reference dimension of all data points in the bottom node is used as the segmentation data, and the data points in the bottom node are divided into two parts using the segmentation data, which are respectively used as the left subtree and the right subtree of the bottom node; in response to the tree depth being not less than the tree depth threshold, the iteration ends to obtain an isolated tree; the isolated trees obtained from all analysis data sets are used to form an isolation forest; Isolation forest is used to perform anomaly detection on the current operating status data to realize production line operating status monitoring.
2. The battery production line operation status monitoring method based on big data according to claim 1 is characterized in that: The random sampling of all historical operating status data points to obtain an analysis data set includes: A random sampling process is performed on all historical operation status data points, and a data set consisting of historical operation status data points obtained by random sampling is recorded as an analysis data set, and several analysis data sets are obtained by sampling multiple times.
3. The battery production line operation status monitoring method based on big data according to claim 1 is characterized in that: The method of using the isolation forest to perform anomaly detection on the current operation status data to realize production line operation status monitoring includes: The current operating status data is input into the isolation forest to obtain the abnormality judgment result. When the abnormality judgment result is abnormal, it is determined that the production line is operating abnormally and an early warning is issued; when the abnormality judgment result is normal, it is determined that the production line is operating normally and no early warning is issued.
4. The battery production line operation status monitoring system based on big data is characterized by: include: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a battery production line operation status monitoring method based on big data according to any one of claims 1 to 3 is implemented.