A method for preparing battery diaphragm material for electrical energy storage

Abstract

The present invention relates to the field of data processing technology, and specifically to a method for preparing a battery diaphragm material for electrical energy storage, comprising: collecting temperature data; obtaining a characteristic value of each data point; constructing a two-dimensional data plane of a first data sequence according to the characteristic value and amplitude of each data point; clustering the data points in the two-dimensional data plane; obtaining the abnormality degree value of the cluster cluster; corresponding the data points of each cluster cluster in the two-dimensional data plane in the temperature data to obtain different labeled data points in the temperature data; hierarchical clustering of different labeled data; and then the outlier value of each data point; obtaining an optimal hyperplane according to the outlier value of each data point; obtaining abnormal data points in the two-dimensional data plane according to the optimal hyperplane, and adjusting the temperature in the production and preparation process of the battery diaphragm according to the abnormal data points. The present invention realizes precise adjustment in the preparation process of the battery diaphragm by abnormal monitoring of the temperature data.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a method for preparing a battery diaphragm material for electrical energy storage. Background Art

[0002] The battery separator is a key component in the electrical energy storage system. It is located between the positive and negative electrodes. Its main function is to prevent the positive and negative electrode materials from directly contacting each other and causing a short circuit. At the same time, it allows the ions in the electrolyte to pass freely to maintain the charge and discharge process of the battery. The performance of the separator material directly affects the safety, energy density, cycle life and cost of the battery. In the preparation process of the battery separator material, the porosity and pore size distribution are regulated by the preparation technology to achieve the optimal ion transmission performance. In the production process of the separator, the pore structure can be regulated by thermal stretching. By controlling the stretching temperature, speed and direction, separators with different pore sizes and porosities can be obtained. The appropriate temperature can ensure that the polymer material will not break during the stretching process, and will not soften excessively. This ensures that a porous structure with good mechanical strength is formed in the separator. Therefore, it is necessary to monitor the temperature during the stretching process so that the material can be stretched at a stable temperature to ensure that the prepared battery separator has good performance.

[0003] In the prior art, the process of monitoring temperature data is mainly reflected in the anomaly identification of the data, so as to adjust the temperature according to the monitoring results to keep it within a reasonable range. Support vector machine (SVM) is a supervised learning algorithm, which is usually used for classification and regression problems. In the field of anomaly detection, One-Class SVM is used to identify outliers or outliers. The purpose of One-Class SVM is to find a small area that is most similar to the normal pattern of the data set, and any point not in this area is considered abnormal. However, because the battery separator material has periodic heating and cooling stages during the preparation process, the temperature data will show periodic fluctuations. Therefore, when the temperature data is detected for anomalies through the support vector machine algorithm, the detection results will be inaccurate due to the fluctuation of the data. Summary of the invention

[0004] The invention provides a method for preparing a battery diaphragm material for electrical energy storage to solve the existing problems.

[0005] A method for preparing a battery diaphragm material for electrical energy storage of the present invention adopts the following technical scheme:

[0006] The present invention provides a method for preparing a battery separator material for electrical energy storage, the method comprising the following steps:

[0007] Collecting temperature data to obtain a first data sequence;

[0008] According to the change of the amplitude of the data points in the first data sequence, the characteristic value of each data point is obtained; according to the characteristic value and amplitude of each data point in the first data sequence, a two-dimensional data plane of the first data sequence is constructed, and the two-dimensional data plane includes a plurality of data points; the data points in the two-dimensional data plane are clustered to obtain a plurality of clusters; according to the change of the characteristic value and amplitude of the data points in the clusters, the abnormality degree of each cluster is obtained;

[0009] The data points of each cluster in the two-dimensional data plane correspond to the first data sequence, and the data points of different clusters are marked respectively to obtain several types of marked data points in the first data sequence; hierarchical clustering is performed on the marked data of the same class to obtain multiple target clusters; according to the distribution of the target clusters corresponding to all data points in the same cluster, and the abnormality degree of the cluster, the outlier value of each data point is obtained;

[0010] Obtain the optimal hyperplane based on the outlier value of each data point;

[0011] Abnormal data points in a two-dimensional data plane are obtained according to the optimal hyperplane, and the temperature in the production process of the battery separator is adjusted according to the abnormal data points.

[0012] Preferably, the step of obtaining the characteristic value of each data point according to the change of the amplitude of the data point in the first data sequence comprises the following specific steps:

[0013] For any data point in the first data sequence, the number of neighborhood data points is preset , the left and right sides of the data point data points as the neighborhood data points of the data point; then the calculation formula for the eigenvalue of each data point in the first data sequence is as follows:

[0014]

[0015] In the formula, Indicates the first data sequence The characteristic value of the data point, Indicates the first data sequence The amplitude of the data point, represents the mean value of the amplitudes of all data points in the first data sequence, Indicates the first data sequence The data point The amplitude of the neighboring data points, Indicates the first data sequence The number of neighboring data points of a data point, represents the absolute value function.

[0016] Preferably, the step of constructing a two-dimensional data plane of the first data sequence according to the characteristic value and amplitude of each data point in the first data sequence, wherein the two-dimensional data plane includes a plurality of data points; and clustering the data points in the two-dimensional data plane to obtain a plurality of clusters comprises the following specific steps:

[0017] A two-dimensional coordinate system is constructed with the amplitude of each data point as the horizontal coordinate and the characteristic value as the vertical coordinate, so as to obtain a two-dimensional data plane corresponding to the first data sequence;

[0018] The DBSCAN density clustering algorithm is used to cluster the data points on the two-dimensional data plane. The distance measurement adopts the Euclidean distance between the data points to obtain multiple clusters. The preset clustering radius is U and the minimum number of clusters is E.

[0019] Preferably, the step of obtaining the abnormality degree of each cluster according to the characteristic value and amplitude change of the data points in the cluster comprises the following specific steps:

[0020]

[0021] In the formula, Indicates The abnormality of the clusters, Indicates The mean of the feature values ​​of all data points in the clusters, Indicates The mean value of the amplitude of all data points in the clusters, Indicates The first Data points and The Euclidean distance between data points is Indicates The first The amplitude of the data point in the first data sequence, Represents the maximum value of the mean amplitude of the data points in each cluster. Indicates The number of data points in a cluster, represents the absolute value function, Represents the normalization function.

[0022] Preferably, the data points of each cluster in the two-dimensional data plane are made to correspond in the first data sequence, the data points of different clusters are respectively marked to obtain several types of marked data points in the first data sequence; and the marked data of the same type are hierarchically clustered to obtain multiple target clusters, including the following specific steps:

[0023] The data points of each cluster in the two-dimensional data plane correspond to the first data sequence, and the data points of different clusters are marked respectively. The data points marked by the same cluster are recorded as the same type of marked data points, and several types of marked data points in the first data sequence are obtained; the marked data points of the same type are clustered using a hierarchical clustering algorithm, and the number of clustering layers is set to , obtain several clusters of all data points of any cluster in the two-dimensional data plane in the first data sequence, which are recorded as several target clusters.

[0024] Preferably, the method of obtaining abnormal data points in a two-dimensional data plane according to the optimal hyperplane, and adjusting the temperature in the battery separator production process according to the abnormal data points to obtain the outlier value of each data point includes the following specific steps:

[0025]

[0026] In the formula, Indicates The first outliers of data points, Indicates The abnormality of the clusters, Indicates The first The target cluster in the first data sequence where the data points are located The time value of the center point, Indicates The first The target cluster in the first data sequence where the data points are located The time value of the center point of the adjacent previous target cluster, Indicates The first The target cluster in the first data sequence where the data points are located The time point of the center point of the adjacent target cluster, Indicates The mean of the feature values ​​of all data points in the clusters, Indicates The mean value of the amplitudes of all data points in a cluster.

[0027] Preferably, the step of obtaining the optimal hyperplane according to the outlier value of each data point comprises the following specific steps:

[0028]

[0029] In the formula, represents the ordinate of the hyperplane function, represents the slope of the hyperplane function, represents the horizontal coordinate of the hyperplane function, Represents the vertical intercept of the hyperplane function.

[0030] Preferably, the calculation formula of the slope of the hyperplane function is as follows:

[0031]

[0032] In the formula, represents the maximum value of outliers among all data points. represents the minimum outlier value of all data points, Represents the maximum value of the mean of the eigenvalues ​​of all data points in each cluster in the two-dimensional data plane. Represents the minimum value of the mean of the eigenvalues ​​of all data points in each cluster in the two-dimensional data plane. Represents the maximum value of the mean amplitude of all data points in each cluster in the two-dimensional data plane. Represents the minimum value of the mean amplitude of all data points in each cluster in the two-dimensional data plane.

[0033] Preferably, the calculation formula of the vertical intercept of the hyperplane function is as follows:

[0034]

[0035] In the formula, represents the cosine of the inverse tangent of the slope, It represents the Euclidean distance between the centroid of the cluster with the largest mean of the eigenvalues ​​of all data points in the cluster and the centroid of the cluster with the smallest mean of the eigenvalues ​​of all data points in the cluster. It represents the Euclidean distance between the centroid of the cluster with the largest mean amplitude of all data points in the cluster and the centroid of the cluster with the smallest mean amplitude of all data points in the cluster.

[0036] Preferably, the method of obtaining abnormal data points in a two-dimensional data plane according to the optimal hyperplane and adjusting the temperature in the production process of the battery separator according to the abnormal data points comprises the following specific steps:

[0037] In the two-dimensional data plane, the data points on the right side of the hyperplane are regarded as abnormal data points;

[0038] Preset the optimal temperature during the preparation of battery separator materials When the temperature value of the abnormal data point is greater than G, the temperature is lowered to the optimal temperature G; when the temperature value of the abnormal data point is less than G, the temperature is raised to the optimal temperature G.

[0039] The beneficial effect of the technical solution of the present invention is that the temperature change of the battery diaphragm material during the preparation process will have a great influence on the performance of the material. Therefore, the present invention analyzes and monitors the temperature data in the material preparation process through the support vector machine algorithm to obtain abnormal data points in the temperature data, thereby adjusting the temperature in the battery diaphragm material preparation process in real time to achieve precise temperature control in the battery diaphragm material preparation process.

[0040] When monitoring temperature data through the support vector machine algorithm, it is necessary to determine the optimal hyperplane in the support vector machine algorithm. Therefore, the original temperature data is obtained and converted into a two-dimensional data plane. Then the DBSCAN clustering algorithm is used to cluster the data points in the two-dimensional data plane, and then the abnormality of each cluster is obtained according to the amplitude and eigenvalue changes of the data points in the cluster cluster; then the outlier value of each data point is obtained according to the data distribution in the original temperature data and the data distribution in the two-dimensional data plane; and then the optimal hyperplane is obtained according to the outlier value of each data point. This hyperplane can separate normal data from abnormal data, and can intuitively and accurately detect abnormal data points in the original temperature data. When abnormal data points are detected, it is necessary to adjust the temperature of the battery diaphragm material in time. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0042] Figure 1 The present invention is a flowchart of the steps of a method for preparing a battery separator material for electrical energy storage. DETAILED DESCRIPTION

[0043] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of a method for preparing a battery diaphragm material for electrical energy storage according to the present invention, its specific implementation, structure, characteristics and effects, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0044] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0045] The specific scheme of the method for preparing a battery diaphragm material for electrical energy storage provided by the present invention is described in detail below with reference to the accompanying drawings.

[0046] See also Figure 1 , which shows a flow chart of the steps of a method for preparing a battery diaphragm material for electrical energy storage provided by an embodiment of the present invention, the method comprising the following steps:

[0047] Step S001: Collect temperature data during the preparation process of the battery separator to obtain a first data sequence.

[0048] The main purpose of this embodiment is to monitor the temperature of the battery diaphragm material during the production and preparation process, and adjust the temperature according to the monitoring results, so that the temperature during the preparation process is within a reasonable range, and the porosity of the diaphragm material is guaranteed to meet the standard. Therefore, it is necessary to first collect and obtain the temperature monitoring data during the preparation of the diaphragm material. This embodiment uses a temperature sensor to obtain the temperature data during the preparation of the diaphragm material, and the collection time interval is set to 1 second. The temperature data is time series data, which contains several data points, and this section of temperature data is normal data, and there are no abnormal data points, which are used as reference data. Then the real-time temperature monitoring data during the preparation of the battery diaphragm is collected. The real-time monitoring temperature data contains several data points obtained by real-time monitoring. The collection time interval is still 1 second. The reference data and the real-time temperature monitoring data together constitute a first data sequence, and then the first data sequence is analyzed to detect whether there is an abnormality in the real-time temperature data.

[0049] At this point, the first data sequence is obtained.

[0050] Step S002: obtaining a characteristic value of each data point according to a change in the amplitude of the data points in the first data sequence; constructing a two-dimensional data plane of the first data sequence according to the characteristic value and amplitude of each data point in the first data sequence; clustering the data points in the two-dimensional data plane to obtain a plurality of clusters; obtaining the degree of abnormality of each cluster according to the change in the characteristic value and amplitude of the data points in the cluster.

[0051] It should be noted that in the preparation process of battery separator materials, temperature is a key control parameter, which has a significant impact on the porous structure, mechanical strength, thermal stability and final battery performance of the separator. The temperature during the preparation process can affect the pore size and pore distribution of the separator material. For example, in the process of thermally induced phase separation, higher temperatures generally lead to larger pores and lower pore density. Temperature changes can also affect the shape of the pores, thereby affecting the anisotropy and structural uniformity of the separator. The mechanical strength of the separator usually needs to be enhanced by stretching or directional crystallization at a higher temperature. The increase in temperature can enhance the fluidity of the polymer chain and help the orientation of the polymer molecular chain, thereby improving the mechanical strength of the separator. Therefore, it is necessary to monitor the temperature during the production and preparation of the battery separator to determine whether the temperature is abnormal. When an abnormality occurs, it is necessary to adjust it in time to ensure that the quality of the battery separator preparation meets the standards.

[0052] It should be further explained that when using the support vector machine algorithm to detect anomalies in temperature data, the goal of the algorithm is to find a hyperplane that can completely separate the two types of samples in the data space and maximize the distance from the boundary to the nearest sample point. This boundary is the so-called "maximum boundary hyperplane". Any point that is not in the maximum boundary area is considered abnormal. However, because the battery separator material has periodic heating and cooling stages during the preparation process, the temperature data will show periodic fluctuations. Therefore, when using the support vector machine algorithm to detect anomalies in temperature data, some data will have outlier features, which will lead to inaccurate detection results due to data fluctuations. When using the support vector machine algorithm to detect anomalies in data, the position of the hyperplane needs to be obtained. Therefore, it is necessary to first visualize the first data sequence, calculate the features of each data point in the first data sequence, and then place it in a two-dimensional coordinate system. The position parameters of the hyperplane are obtained based on the change relationship of the data points in the two-dimensional coordinate system.

[0053] Specifically, first, for any data point in the first data sequence, the number of neighborhood data points is preset , this embodiment adopts Describe the data point on both sides. data points are used as neighborhood data points of the data point. It is particularly noted that if the data point is close to the boundary of the first data sequence and it is not possible to obtain enough neighborhood data points, no additional data points are made and the actual existing data points are used as neighborhood data points. The characteristic value of each data point is obtained according to the change in the amplitude of the data point in the first data sequence. The calculation formula is as follows:

[0054]

[0055] In the formula, Indicates the first data sequence The characteristic value of the data point, Indicates the first data sequence The amplitude of the data point, represents the mean value of the amplitudes of all data points in the first data sequence, Indicates the first data sequence The data point The amplitude of the neighboring data points, Indicates the first data sequence The number of neighboring data points of a data point, represents the absolute value function.

[0056] It should be noted that Indicates the first data sequence The larger the difference between the amplitude of the data point and the mean amplitude of all data points in the first data sequence, the greater the difference. The more the amplitude of a data point deviates from normal data, the greater its abnormality may be. Indicates The larger the difference is, the greater the difference between the current data point and its neighboring data points, and therefore the greater the degree of abnormality. Indicates The mean value of the difference between the amplitude of a data point and all its neighboring data points.

[0057] Furthermore, according to the characteristic value of each data point in the first data sequence obtained by the above calculation, a two-dimensional coordinate system is constructed with the amplitude of each data point as the horizontal coordinate and the characteristic value as the vertical coordinate to obtain a two-dimensional data plane corresponding to the first data sequence.

[0058] It should be noted that because the battery diaphragm material will have periodic heating and cooling stages during the preparation process, the temperature data will show periodic fluctuations, and the periodic fluctuations show that the temperature data fluctuates up and down. Therefore, when the data is detected for abnormality through the support vector machine algorithm, the data outlier characteristics are more obvious, that is, the farther away from the hyperplane, the more local outliers will appear in the data. Therefore, in order to accurately detect the abnormal data points in the first data sequence, this embodiment analyzes the data in the two-dimensional data plane to obtain the degree of outliers of the data points, and then detects the abnormal data points; the support vector machine algorithm is a binary classification algorithm. If the data point is on the wrong side of the hyperplane, this point may be misclassified or an outlier. Therefore, in order to accurately classify the data, it is necessary to characterize the normal characteristics of the data.

[0059] Specifically, the DBSCAN density clustering algorithm is used to cluster the data points on the two-dimensional data plane, and the clustering radius is set to , the minimum number of clusters is , and obtain multiple clusters. The above thresholds are empirical values, and implementers can set them according to different implementation environments; the DBSCAN density clustering algorithm is a well-known technology and will not be described here. Then analyze the obtained clusters to obtain the abnormality degree of each cluster, and the calculation formula is as follows:

[0060]

[0061] In the formula, Indicates The abnormality of the clusters, Indicates The mean of the feature values ​​of all data points in the clusters, Indicates The mean value of the amplitude of all data points in the clusters, Indicates The first Data points and The Euclidean distance between data points is Indicates The first The amplitude of the data point in the first data sequence, Represents the maximum value of the mean amplitude of all data points in the clusters. Indicates The number of data points in a cluster, represents the absolute value function, Represents a linear normalization function; the normalized object is all clusters .

[0062] It should be noted that Indicates The density of data points in a cluster is represented by the distance between data points. The smaller the average distance, the greater the density. Therefore, the more clustered the data points in the cluster are, the smaller the abnormality is. It should be noted that the calculation formula of the Euclidean distance between two data points is as follows: ,in, Indicates Data points and The Euclidean distance between data points is Indicates The horizontal coordinate of the data point, Indicates The horizontal coordinate of the data point, Indicates The vertical coordinate of the data point, Indicates The vertical coordinate of the data point. Indicates The difference between the amplitude of the data point in the cluster and the maximum value. The larger the difference, the smaller the abnormality of the current cluster. Multiply by the number of cluster data points This is to indicate that when the number of data points in a cluster is greater, it means that in the first data sequence, there are more data points with similar temperatures, so the data points in the cluster are more likely to be normal in the region; on the contrary, the fewer the number of data points in the cluster, the greater the difference between the changes in its temperature data points and other data in the first data sequence, so it is more likely to be an abnormal cluster.

[0063] At this point, the abnormality levels of different clusters are obtained.

[0064] Step S003: correspond the data points of each cluster in the two-dimensional data plane in the first data sequence, mark the data points of different clusters respectively, and obtain several types of marked data points in the first data sequence; perform hierarchical clustering on the marked data of the same class to obtain multiple target clusters; obtain the outlier value of each data point according to the distribution of the target clusters corresponding to all data points in the same cluster and the abnormality degree of the cluster.

[0065] According to the abnormal degree of different clusters obtained by the above calculation, the original data is then analyzed to obtain the outlier values ​​of different data points.

[0066] It should be noted that, when converting the first data sequence to a two-dimensional data plane, although the data in the first data sequence presents certain periodic changes and the amplitude changes of the data will vary, the conversion is performed based on the changes of each data point and its neighborhood data points. Therefore, abnormal data points in the two-dimensional data plane will have outlier characteristics. The larger the abnormality value, the greater the degree of outlier. Therefore, the outlier value of each data point can be obtained based on the distribution of the data in the two-dimensional data plane.

[0067] Specifically, firstly, the data points of each cluster in the two-dimensional data plane are corresponding to the first data sequence, and the data points of different clusters are marked respectively, and the data points marked by the same cluster are recorded as the same type of marked data points, and several types of marked data points in the first data sequence are obtained. Then, the hierarchical clustering algorithm is used to cluster the different price data points, and the number of clustering layers is set to , obtain several clusters of data points of any cluster in the two-dimensional data plane in the first data sequence, which are recorded as several target clusters. It should be noted that when multiple data points contained in a cluster in the two-dimensional data plane are hierarchically clustered in the first data sequence, multiple target clusters will be obtained, and each target cluster is distributed at a different position in the first data sequence.

[0068] It should be noted that, because the temperature of the battery separator material changes periodically during processing, the purpose of this step is to describe the periodic characteristics of the data points in the two-dimensional data plane in the first data sequence. Therefore, the outlier value of each data point is obtained based on the clustering result of the first data sequence.

[0069] Specifically, according to the horizontal coordinate difference of the data points in the same target cluster and the abnormality of the cluster corresponding to the marked data point of the corresponding class of the target cluster, the calculation formula for obtaining the outlier value of each data point is as follows:

[0070]

[0071] In the formula, Indicates The first outliers of data points, Indicates The abnormality of the clusters, Indicates The first The target cluster in the first data sequence where the data points are located The time value of the center point, Indicates The first The target cluster in the first data sequence where the data points are located The time value of the center point of the adjacent previous target cluster, Indicates The first The target cluster in the first data sequence where the data points are located The time point of the center point of the adjacent target cluster, Indicates The mean of the feature values ​​of all data points in the clusters, Indicates The mean value of the amplitudes of all data points in a cluster.

[0072] It should be noted that Represents the first The data distribution of the data points of the clusters in the first data sequence is the data distribution of the first data sequence. Because the data in the same data segment will be clustered into one category when performing hierarchical clustering, the period of the data points is represented by describing the temporal change relationship of the target cluster. The stronger the periodicity, the stronger the periodicity of the first data sequence in the two-dimensional data plane. The smaller the outlier degree of the data points in a cluster, that is, The closer the ratio is to 1, the more The stronger the periodic characteristics of the data points in a cluster, the smaller their outlier degree. Represents the first The distribution position of the clusters in the plane, the larger or smaller the ratio is, the more it indicates that the data points in the cluster are outliers, so the outlier value of the data point is larger.

[0073] At this point, the outlier value of each data point is obtained.

[0074] Step S004: Obtain an optimal hyperplane according to the outlier value of each data point.

[0075] It should be noted that the role of the optimal hyperplane is to find a decision boundary that can maximize the interval between data points of different categories and can separate the samples in the data set according to category. The goal of the optimal hyperplane is to maximize the interval between two categories. The interval is twice the minimum distance from the nearest data point (support vector) of the opposite category to the hyperplane. This maximum interval feature helps to improve the generalization ability of the model. Therefore, the outlier value of each data point obtained by the above calculation is then used to obtain the optimal hyperplane. In two-dimensional data, the optimal hyperplane is a straight line.

[0076] Specifically, the calculation formula of the optimal hyperplane is as follows:

[0077]

[0078] In formula (1), represents the hyperplane function, where represents the ordinate of the hyperplane function, represents the slope of the hyperplane function, represents the horizontal coordinate of the hyperplane function, Represents the vertical intercept of the hyperplane function.

[0079] In formula (2), represents the maximum value of all outliers of data points, represents the minimum outlier value of all data points, Represents the maximum value of the mean of the eigenvalues ​​of all data points in each cluster in the two-dimensional data plane. Represents the minimum value of the mean of the eigenvalues ​​of all data points in each cluster in the two-dimensional data plane. Represents the maximum value of the mean amplitude of all data points in each cluster in the two-dimensional data plane. Represents the minimum value of the mean amplitude of all data points in each cluster in the two-dimensional data plane.

[0080] It should be noted that The ratio of the maximum and minimum outliers of all data points indicates the degree of divergence of the data points. The larger the ratio, the more divergent the data points are in the two-dimensional data plane. Therefore, the larger the ratio, the greater the slope. represents the distribution of clusters in the two-dimensional data plane, represents the distribution of the ordinate, It represents the distribution of the horizontal coordinate, and its ratio determines the slope of the hyperplane function. The larger the numerator is, the greater the change in the vertical coordinate of the data point in the two-dimensional data plane is. Therefore, the larger the slope should be, so that the data points with large eigenvalues ​​can be separated. Similarly, the larger the denominator is, the greater the change in the horizontal coordinate of the data point in the two-dimensional data plane is. Therefore, the smaller the slope should be, so that the data points with large amplitudes can be separated.

[0081] In formula (3), represents the cosine of the inverse tangent of the slope, It represents the Euclidean distance between the centroid of the cluster with the largest mean of the eigenvalues ​​of all data points in the cluster and the centroid of the cluster with the smallest mean of the eigenvalues ​​of all data points in the cluster. It represents the Euclidean distance between the centroid of the cluster with the largest mean amplitude of all data points in the cluster and the centroid of the cluster with the smallest mean amplitude of all data points in the cluster.

[0082] It should be noted that It represents the sum of the vertical distances from the cluster with the largest eigenvalue mean and the cluster with the smallest eigenvalue mean to the straight line with a slope of k. It represents the sum of the vertical distances from the cluster with the largest mean amplitude and the cluster with the smallest mean amplitude to the straight line with slope k.

[0083] At this point, the optimal hyperplane is obtained.

[0084] Step S005: obtaining abnormal data points in a two-dimensional data plane according to the optimal hyperplane, and adjusting the temperature in the production process of the battery separator according to the abnormal data points.

[0085] According to the optimal hyperplane obtained above, the abnormal data points in the two-dimensional data plane are then obtained. In the support vector machine algorithm, the hyperplane is the decision boundary that separates different categories in the feature space, and the abnormal data is located on one side of the hyperplane. In this embodiment, because the data with large eigenvalues ​​or large amplitudes are abnormal, the abnormal data is located on the right side of the hyperplane. Then, in the two-dimensional data plane, the data on the right side of the hyperplane is an abnormal data point, thereby obtaining the abnormal data point. Then, the temperature in the production process of the battery diaphragm is adjusted according to the abnormal data points.

[0086] Specifically, in a two-dimensional data plane, data points located on the right side of the hyperplane are regarded as abnormal data points.

[0087] Furthermore, when an abnormal data point is detected in the temperature data, it means that the temperature data is abnormal at this time, so it is necessary to adjust the temperature in time. During the adjustment, the optimal temperature in the preparation process of the battery diaphragm material is obtained based on prior knowledge. When the temperature of the abnormal data point is detected to be too high, it needs to be adjusted downward; when the temperature of the abnormal data point is detected to be too low, it needs to be adjusted upward.

[0088] At this point, a method for preparing battery separator materials for electrical energy storage has been completed.

[0089] 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 in the protection scope of the present invention.

Claims

1. A method for preparing a battery diaphragm material for electrical energy storage, characterized in that: The method comprises the following steps: Collecting temperature data to obtain a first data sequence; According to the change of the amplitude of the data points in the first data sequence, the characteristic value of each data point is obtained; according to the characteristic value and amplitude of each data point in the first data sequence, a two-dimensional data plane of the first data sequence is constructed, and the two-dimensional data plane includes a plurality of data points; the data points in the two-dimensional data plane are clustered to obtain a plurality of clusters; according to the change of the characteristic value and amplitude of the data points in the clusters, the abnormality degree of each cluster is obtained; The data points of each cluster in the two-dimensional data plane correspond to the first data sequence, and the data points of different clusters are marked respectively to obtain several types of marked data points in the first data sequence; hierarchical clustering is performed on the marked data of the same class to obtain multiple target clusters; according to the distribution of the target clusters corresponding to all data points in the same cluster, and the abnormality degree of the cluster, the outlier value of each data point is obtained; Obtain the optimal hyperplane based on the outlier value of each data point; Obtaining abnormal data points in a two-dimensional data plane according to the optimal hyperplane, and adjusting the temperature in the production process of the battery separator according to the abnormal data points; The specific steps of obtaining the characteristic value of each data point according to the change of the amplitude of the data point in the first data sequence include the following: For any data point in the first data sequence, the number of neighborhood data points is preset , the left and right sides of the data point data points as the neighborhood data points of the data point; then the calculation formula for the eigenvalue of each data point in the first data sequence is as follows: In the formula, Indicates the first data sequence The characteristic value of the data point, Indicates the first data sequence The amplitude of the data point, represents the mean value of the amplitudes of all data points in the first data sequence, Indicates the first data sequence The data point The amplitude of the neighboring data points, Indicates the first data sequence The number of neighboring data points of a data point, represents the absolute value function; The step of constructing a two-dimensional data plane of the first data sequence according to the characteristic value and amplitude of each data point in the first data sequence, wherein the two-dimensional data plane includes a plurality of data points; and clustering the data points in the two-dimensional data plane to obtain a plurality of clusters includes the following specific steps: A two-dimensional coordinate system is constructed with the amplitude of each data point as the horizontal coordinate and the characteristic value as the vertical coordinate, so as to obtain a two-dimensional data plane corresponding to the first data sequence; The DBSCAN density clustering algorithm is used to cluster the data points on the two-dimensional data plane. The distance measurement adopts the Euclidean distance between the data points to obtain multiple clusters. The preset cluster radius is U and the minimum number of clusters is E. The specific steps of obtaining the abnormality degree of each cluster according to the characteristic value and amplitude change of the data points in the cluster are as follows: In the formula, Indicates The abnormality of the clusters, Indicates The mean of the feature values ​​of all data points in the clusters, Indicates The mean value of the amplitude of all data points in the clusters, Indicates The first Data points and The Euclidean distance between data points is Indicates The first The amplitude of the data point in the first data sequence, Represents the maximum value of the mean amplitude of the data points in each cluster. Indicates The number of data points in a cluster, represents the absolute value function, represents the normalization function; The method of making the data points of each cluster in the two-dimensional data plane correspond to each other in the first data sequence, marking the data points of different clusters respectively, and obtaining several types of marked data points in the first data sequence; and performing hierarchical clustering on the marked data of the same type to obtain multiple target clusters includes the following specific steps: The data points of each cluster in the two-dimensional data plane correspond to the first data sequence, and the data points of different clusters are marked respectively. The data points marked by the same cluster are recorded as the same type of marked data points, and several types of marked data points in the first data sequence are obtained; the marked data points of the same type are clustered using a hierarchical clustering algorithm, and the number of clustering layers is set to , obtain several clusters of all data points of any cluster in the two-dimensional data plane in the first data sequence, recorded as several target clusters; The method of obtaining abnormal data points in a two-dimensional data plane according to the optimal hyperplane and adjusting the temperature in the production process of the battery separator according to the abnormal data points to obtain the outlier value of each data point includes the following specific steps: In the formula, Indicates The first outliers of data points, Indicates The abnormality of the clusters, Indicates The first The target cluster in the first data sequence where the data points are located The time value of the center point, Indicates The first The target cluster in the first data sequence where the data points are located The time value of the center point of the adjacent previous target cluster, Indicates The first The target cluster in the first data sequence where the data points are located The time point of the center point of the adjacent target cluster, Indicates The mean of the feature values ​​of all data points in the clusters, Indicates The mean value of the amplitude of all data points in the cluster; The specific steps of obtaining the optimal hyperplane according to the outlier value of each data point are as follows: In the formula, represents the ordinate of the hyperplane function, represents the slope of the hyperplane function, represents the horizontal coordinate of the hyperplane function, Represents the vertical intercept of the hyperplane function.

2. A method for preparing a battery diaphragm material for electrical energy storage according to claim 1, characterized in that: The calculation formula of the slope of the hyperplane function is as follows: In the formula, represents the maximum value of all outliers of data points, represents the minimum outlier value of all data points, Represents the maximum value of the mean of the eigenvalues ​​of all data points in each cluster in the two-dimensional data plane. Represents the minimum value of the mean of the eigenvalues ​​of all data points in each cluster in the two-dimensional data plane. Represents the maximum value of the mean amplitude of all data points in each cluster in the two-dimensional data plane. Represents the minimum value of the mean amplitude of all data points in each cluster in the two-dimensional data plane.

3. A method for preparing a battery diaphragm material for electrical energy storage according to claim 2, characterized in that: The calculation formula of the vertical intercept of the hyperplane function is as follows: In the formula, represents the cosine of the inverse tangent of the slope, It represents the Euclidean distance between the centroid of the cluster with the largest mean of the eigenvalues ​​of all data points in the cluster and the centroid of the cluster with the smallest mean of the eigenvalues ​​of all data points in the cluster. It represents the Euclidean distance between the centroid of the cluster with the largest mean amplitude of all data points in the cluster and the centroid of the cluster with the smallest mean amplitude of all data points in the cluster.

4. The method for preparing a battery diaphragm material for electrical energy storage according to claim 1, characterized in that: The specific steps of obtaining abnormal data points in a two-dimensional data plane according to the optimal hyperplane and adjusting the temperature in the production process of the battery separator according to the abnormal data points are as follows: In the two-dimensional data plane, the data points on the right side of the hyperplane are regarded as abnormal data points; Preset the optimal temperature during the preparation of battery separator materials When the temperature value of the abnormal data point is greater than G, the temperature is lowered to the optimal temperature G; when the temperature value of the abnormal data point is less than G, the temperature is raised to the optimal temperature G.

Images