A perovskite thin film production data analysis method

Abstract

The present application relates to the technical field of data processing, and particularly relates to a perovskite film production data analysis method, comprising: obtaining historical data after preprocessing; obtaining prior correlation of production batches and production data difference of links in the production batches; obtaining joint influence possibility of the links in the production batches according to the prior correlation of the production batches and the production data difference of the links in the production batches; further obtaining relative values and classification weights of the links; obtaining link combinations with mutual influence effects based on the classification weights of the links; determining strong influence combinations corresponding to the links according to the relative values of the links in the link combinations with the mutual influence effects; and analyzing production data to be analyzed by using the strong influence combinations corresponding to the links to obtain analysis results. The present application analyzes production data in the perovskite film production process based on mutual influence relations of different production links on production quality, and improves accuracy of analysis results.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, and particularly relates to a perovskite film production data analysis method. BACKGROUND

[0002] Perovskite film is a new type of semiconductor material, which is widely used in the field of solar cells due to its excellent photoelectric properties (high absorption coefficient, long carrier diffusion length) and solution processing. In the production process of perovskite film, problems such as pinholes, composition segregation and grain boundary defects are prone to occur, so it is necessary to analyze the production of perovskite film to improve the yield of perovskite film.

[0003] The existing problem: the existing perovskite film production data analysis method mainly judges and analyzes the production data of different production links separately. This method only considers the influence of a single production link on the production quality, but ignores the interaction between different production links which also affects the production quality, thereby making the production data analysis result not accurate enough. SUMMARY

[0004] The present application provides a perovskite film production data analysis method to solve the existing problem.

[0005] The perovskite film production data analysis method provided by the present application adopts the following technical scheme:

[0006] An embodiment of the present application provides a perovskite film production data analysis method, which comprises the following steps:

[0007] Collect historical data and pre-process the historical data to obtain pre-processed historical data; wherein the pre-processed historical data includes production data of all production links in each production batch in the production process of perovskite film, and yield of each production batch;

[0008] Obtain prior correlation of each production batch by using the yield of each production batch;

[0009] Obtain production data difference of each production link in each production batch based on the corresponding production data of each production link in different production batches;

[0010] Obtain joint influence possibility of each production link in each production batch according to the prior correlation of each production batch and the production data difference of each production link in each production batch;

[0011] Wherein, the expression of the joint influence possibility of each production link in each production batch is as follows:

[0012]

[0013] wherein, represents the joint influence possibility of the i-th production link in the j-th production batch, represents the joint influence possibility of the i-th production link in the j-th production batch, represents the variance of the production data difference of all production links in the j-th production batch, represents the variance of the production data difference of all production links in the j-th production batch, represents the number of production links in the j-th production batch, represents the production data difference of the i-th production link in the j-th production batch, represents the prior correlation of the j-th production batch, represents the production data difference of the i-th production link in the j-th production batch; represents the prior correlation of the j-th production batch, represents the production data difference of the i-th production link in the j-th production batch; represents the prior correlation of the j-th production batch, represents the production data difference of the i-th production link in the j-th production batch; represents the prior correlation of the j-th production batch, represents the production data difference of the i-th production link in the j-th production batch; represents the prior correlation of the j-th production batch,

[0014] obtaining the relative value and the classification weight of each production link by using the joint influence possibility of each production link in each production batch;

[0015] wherein, the expression of the classification weight of each production link is as follows:

[0016]

[0017] wherein, represents the classification weight of the i-th production link, represents the variance of the joint influence possibility of the i-th production link corresponding to the j-th production batch; represents the mean of the joint influence possibility of the i-th production link corresponding to the j-th production batch; represents the mean of the joint influence possibility of the i-th production link corresponding to the j-th production batch; represents the mean of the joint influence possibility of the i-th production link corresponding to the j-th production batch; represents the mean of the joint influence possibility of the i-th production link corresponding to the j-th production batch; represents the mean of the joint influence possibility of the i-th production link corresponding to the j-th production batch; represents the mean of the joint influence possibility of the i-th production link corresponding to the j-th production batch; represents the mean of the joint influence possibility of the i-th production link corresponding to the j-th production batch; represents the mean of the joint influence possibility of the i-th production link corresponding to the j-th production batch; represents the mean of the joint influence possibility of the i-th production link corresponding to the j-th production batch; represents the mean of the joint influence possibility of the i-th production link corresponding to the j-th production batch; represents the mean of the joint influence possibility of the i-th production link corresponding to the j-th production batch;

[0018] obtaining all production link combinations with mutual influence based on the classification weight of each production link;

[0019] determining the strong influence combination corresponding to each production link according to the relative value of each production link in each production link combination with mutual influence;

[0020] The production data of the perovskite film to be analyzed is analyzed by using the strong influence combination corresponding to each production link to obtain an analysis result.

[0021] Further, the use of the yield of each production batch to obtain the prior correlation of each production batch includes the following specific steps:

[0022] Obtain the mean of the yield of all production batches;

[0023] Obtain the prior correlation of each production batch by using the yield of each production batch and the mean of the yield of all production batches.

[0024] Further, the production data difference of each production link in each production batch is obtained based on the production data corresponding to each production link in different production batches, which includes the following specific steps:

[0025] Calculate the difference between the production data corresponding to each production link in the current production batch and the production data corresponding to each production link in each other production batch to obtain the production data difference of each production link in different production batches;

[0026] Obtain the number of production batches; based on the number of production batches and the production data difference of each production link in different production batches, obtain the production data difference of each production link in each production batch.

[0027] Further, the joint influence possibility of each production link in each production batch is obtained according to the prior correlation of each production batch and the production data difference of each production link in each production batch, which includes the following specific steps:

[0028] Obtain the variance of the production data difference of all production links in each production batch by using the production data difference of each production link in each production batch;

[0029] Obtain the number of production links;

[0030] Based on the prior correlation of each production batch, the number of production links, the variance of the production data difference of all production links in each production batch, and the production data difference of each production link in each production batch, obtain the joint influence possibility of each production link in each production batch.

[0031] Further, the relative value and classification weight of each production link are obtained by using the joint influence possibility of each production link in each production batch, which includes the following specific steps:

[0032] Calculate the mean of the joint influence possibility corresponding to each production link in different production batches;

[0033] calculate the mean value of the joint influence possibility corresponding to each other production link in different production batches, and calculate the mean value of the joint influence possibility corresponding to all other production links in different production batches to obtain the mean value of the joint influence possibility corresponding to the other production links in different production batches;

[0034] obtain the relative value of each production link using the mean value of the joint influence possibility corresponding to each production link in different production batches and the mean value of the joint influence possibility corresponding to the other production links in different production batches;

[0035] obtain the variance of the joint influence possibility corresponding to each production link in different production batches;

[0036] obtain the classification weight of each production link based on the variance of the joint influence possibility corresponding to each production link in different production batches and the relative value of each production link.

[0037] Further, the classification weight of each production link is obtained based on the classification weight of each production link, and the specific steps include:

[0038] sort all production links in descending order of classification weight, and randomly combine the sorted production links to obtain all production link combination sequences;

[0039] all production link combination sequences are used as a transaction set, the classification weight of each production link is used as support, and the transaction set and support are used as input to build a frequent pattern tree, and all production link combinations with mutual influence are output.

[0040] Further, the strong influence combination corresponding to each production link is determined according to the relative value of each production link in each production link combination with mutual influence, and the specific steps include:

[0041] count the production links in all production link combinations with mutual influence to obtain all link combinations corresponding to each production link;

[0042] obtain the classification intensity of each production link in each link combination based on the relative value of each production link in each link combination corresponding to each production link;

[0043] obtain the maximum value of the classification intensity of each production link in all link combinations, and determine the link combination corresponding to the maximum value of the classification intensity as the strong influence combination corresponding to each production link.

[0044] Further, the relative value of each production link in each link combination corresponding to each production link is obtained, and the classification strength of each production link in each link combination is obtained, including the following specific steps:

[0045] The average value of the relative values of all production links in each link combination corresponding to each production link is calculated, and the average value is determined as the classification strength of each production link in each link combination.

[0046] Further, the strong influence combination corresponding to each production link is used to analyze the perovskite thin film production data to be analyzed, and an analysis result is obtained, including the following specific steps:

[0047] The data to be analyzed is determined, wherein the data to be analyzed includes a production batch to be analyzed and production data of each production link in the production batch to be analyzed in a perovskite thin film production process;

[0048] According to the strong influence combination corresponding to each production link in the production batch to be analyzed, the production data of other production links having mutual influence on each production link is obtained;

[0049] The production data of each production link in the production batch to be analyzed and the production data of other production links having mutual influence on each production link are analyzed by using a joint analysis method, and an analysis result is obtained.

[0050] Further, the historical data is preprocessed to obtain preprocessed historical data, including the following specific steps:

[0051] The historical data is cleaned to obtain preprocessed historical data.

[0052] The technical scheme of the present application has the following advantages: the perovskite thin film production data analysis method proposed in the embodiment of the present application analyzes the production data of different production links collected, obtains the mutual influence relationship of different production links on production quality, analyzes the production data in the perovskite thin film production process based on the mutual influence relationship of different production links on production quality, and improves the accuracy of the analysis result. BRIEF DESCRIPTION OF DRAWINGS

[0053] In order to more clearly illustrate the technical scheme in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiment or the prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and those skilled in the art can obtain other drawings according to these drawings without creative labor.

[0054] Figure 1A step flow chart of a perovskite film production data analysis method. DETAILED DESCRIPTION

[0055] In order to further clarify the technical means and effects taken by the present application to achieve the predetermined object of the application, the specific implementation, structure, characteristics and effects of a perovskite film production data analysis method according to the present application are described in detail as follows in combination with the drawings and preferred embodiments. Different "one embodiment" or "another embodiment" in the following description 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.

[0056] 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.

[0057] The specific scheme of the perovskite film production data analysis method provided by the present application is specifically described below in combination with the drawings.

[0058] Please refer to Figure 1 which shows a step flow chart of a perovskite film production data analysis method provided by one embodiment of the present application, which includes the following steps:

[0059] Step S001: Collect historical data and pre-process the historical data to obtain pre-processed historical data; wherein the pre-processed historical data includes production data of all production links of each production batch in the perovskite film production process, and the yield rate of each production batch.

[0060] The perovskite film production data analysis method proposed in this embodiment analyzes the production data collected from different production links to obtain the mutual influence relationship of different production links on production quality, and analyzes the production data in the perovskite film production process based on the mutual influence relationship of different production links on production quality, thereby improving the accuracy of the analysis results.

[0061] It should be noted that: the main purpose of this embodiment is to analyze the influence of the interaction between different production data in the perovskite film production process on the production quality of the perovskite film by establishing an analysis model, so it is necessary to collect different production data in the perovskite film production process, and it is necessary to verify the analysis model using the actual production results of the perovskite film under the corresponding production data, so it is necessary to collect historical data, which includes production data and actual production results.

[0062] Specifically, first, collect production data: take the production batch of the perovskite film as a node, and obtain the production data of each production batch in the perovskite film production process, including the production data of each production link in each production batch. We will collect production data from each production stage and create a dataset.

[0063] With the first Taking a production batch as an example, the corresponding production data is as follows: ,in Indicates the first The first in the production batch Production data for each stage of the production process.

[0064] Next, actual production results were collected: taking the production batch of the perovskite film as a node, the yield rate of the perovskite film in each production batch was statistically analyzed.

[0065] Finally, data preprocessing is performed. In this embodiment, the data preprocessing method used is data cleaning, which includes deleting duplicate data and filling in missing values.

[0066] This completes the acquisition of production data for all production stages of each production batch, as well as the yield rate for each production batch.

[0067] Step S002: Use the yield rate of each production batch to obtain the prior correlation of each production batch.

[0068] In this embodiment, the average yield rate of all production batches is obtained; using the yield rate of each production batch and the average yield rate of all production batches, the prior correlation of each production batch is obtained.

[0069] It should be noted that the production of perovskite thin films requires the coordinated operation of multiple processes and procedures. Therefore, quality issues in perovskite thin film production are not necessarily caused by a single production stage, but rather are the result of the combined effects of multiple stages. Existing technologies often analyze single production data, failing to analyze the mutual influence of different production data on perovskite thin films. Therefore, this embodiment utilizes different production data from the perovskite thin film production process to establish an analytical model, and uses correlation verification between different production data to obtain the combined effects between them.

[0070] The yield rate of each production batch reflects the combined impact of production data from different production stages within that batch. Therefore, it is necessary to obtain the prior correlation of the combined impact of production data for each production batch.

[0071] Specifically, with the first Taking a production batch as an example, the expression for its corresponding prior correlation is as follows:

[0072]

[0073] in, Indicates the first prior correlation of each production batch, indicates the yield rate of the first production batch; indicates the average of the yield rates of the collected production batches; indicates an inverse proportional normalization function.

[0074] It should be noted that: in different production batches, the lower the yield rate of the perovskite film produced by the first production batch, the greater the possibility that the production data corresponding to different production links in the first production batch is abnormal, and vice versa.

[0075] Step S003: Based on the production data corresponding to each production link in different production batches, obtain the production data difference of each production link in each production batch.

[0076] In this embodiment, the difference between the production data corresponding to each production link in the current production batch and the production data corresponding to each production link in each other production batch is calculated to obtain the production data difference corresponding to each production link in different production batches; the number of production batches is obtained; and based on the number of production batches and the production data difference corresponding to each production link in different production batches, the production data difference of each production link in each production batch is obtained.

[0077] It should be noted that: the joint influence possibility of the production data corresponding to all production links in each production batch is obtained by using the prior correlation of each production batch and combining the differences of different production links in different production batches.

[0078] Specifically, first, the difference between the production data of each production link in the first production batch and the production data of the corresponding production link in other different production batches is obtained, and the difference between the production data of the first production batch and the production data of the corresponding production link in other different production batches is obtained. For example, the difference between the production data of the first production link in the first

[0079] production batch and the production data of the corresponding production link in other different production batches, i.e., the expression of the production data difference of each production link in each production batch, is as follows:

[0080] wherein, indicates the production data difference of the first production link in the first production batch, indicates the number of production batches, and respectively indicate the first the first production batch and the first production link in the second production batch correspond to the same production data, the first production batch and the first production link in the second production batch correspond to the same production data, the production data of the first production link in the first production batch and the production data of the first production link in the second production batch, ; , and .

[0081] It should be noted that: the greater, the greater the difference between the production data of the first production link in the first production batch and the production data of the first production link in the second production batch, and vice versa. the greater, the greater the difference between the production data of the first production link in the first production batch and the production data of the first production link in the second production batch, and vice versa. the greater, the greater the difference between the production data of the first production link in the first production batch and the production data of the first production link in the second production batch, and vice versa. the greater, the greater the difference between the production data of the first production link in the first production batch and the production data of the first production link in the second production batch, and vice versa. the greater, the greater the difference between the production data of the first production link in the first production batch and the production data of the first production link in the second production batch, and vice versa.

[0082] Step S004: According to the prior correlation of each production batch and the production data difference of each production link in each production batch, the joint influence possibility of each production link in each production batch is obtained.

[0083] In this embodiment, the variance of the production data difference of all production links in each production batch is obtained by using the production data difference of each production link in each production batch; the number of production links is obtained; based on the prior correlation of each production batch, the number of production links, the variance of the production data difference of all production links in each production batch, and the production data difference of each production link in each production batch, the joint influence possibility of each production link in each production batch is obtained.

[0084] Specifically: the initial possibility of the production data of each production link in the first production batch and the production data of other production links in the same production batch jointly influencing the first production batch is calculated, and the initial possibility of the production data of each production link in the second production batch and the production data of other production links in the same production batch jointly influencing the second production batch is calculated. the first production batch and the first production link in the second production batch correspond to the same production data, the first production batch and the first production link in the second production batch correspond to the same production data, the first production batch and the first production link in the second production batch correspond to the same production data, the first production batch and the first production link in the second production batch correspond to the same production data,

[0085]

[0086] wherein, the joint influence possibility of the first production link in the first production batch, the joint influence possibility of the first production link in the first production batch, the joint influence possibility of the first production link in the first production batch, the variance of the production data difference of all production links in the first production batch, the variance of the production data difference of all production links in the first production batch, the number of production links in the first production batch, the number of production links in the first production batch, the number of production links in the first production batch, the number of production links in the first production batch, the production data of the production links in the first production batch.

[0087] It should be noted that: The smaller the value is, the more likely it is that the production data of all the production links in the first production batch have the same trend of difference. The larger the value is, the more likely it is that the production data of some production links in the first production batch have the same trend of difference. The larger the value is, the more likely it is that the production data of different production links in the first production batch affect each other. The smaller the value is, the more likely it is that the production data of different production links in the first production batch affect each other. The larger the value is, the more likely it is that the production data of some production links in the first production batch have independent influence on the first production batch. The smaller the value is, the more likely it is that the production data of some production links in the first production batch have independent influence on the first production batch. The smaller the value is, the more likely it is that the production data of the production links in the first production batch. The larger the value is, the more likely it is that the production data of the production links in the first production batch. The smaller the value is, the more likely it is that the production data of the production links in the first production batch.

[0088] Step S005: obtaining the relative value and the classification weight of each production link by using the joint influence possibility of each production link in each production batch.

[0089] In this embodiment, the mean value of the joint influence possibility of each production link in different production batches is calculated; the mean value of the joint influence possibility of each other production link in different production batches is calculated, and the mean value of the joint influence possibility of all other production links in different production batches is calculated to obtain the mean value of the joint influence possibility of other production links in different production batches; the relative value of each production link is obtained by using the mean value of the joint influence possibility of each production link in different production batches and the mean value of the joint influence possibility of other production links in different production batches; the variance of the joint influence possibility of each production link in different production batches is obtained; and the classification weight of each production link is obtained based on the variance of the joint influence possibility of each production link in different production batches and the relative value of each production link.

[0090] It should be noted that the production data of each production stage in each production batch, and the probability of joint influence within the current production batch, can only determine whether the corresponding production stage has a joint effect with other production stages, which will affect product production. However, it cannot clearly identify which production stages have a joint effect with each other. Therefore, further analysis is needed. Since it is necessary to obtain the mutual influence rules between different production stages, this embodiment builds an analysis model based on a frequent pattern tree. This embodiment first needs to obtain the classification weight of each production stage, and then use the classification weight of each production stage to construct a frequent pattern tree.

[0091] Specifically: By utilizing the joint influence probability of production data from each production stage within each production batch, the classification weights of different production stages across all production batches are obtained, with the first... Taking a single production stage as an example, the expression for its corresponding classification weight is as follows:

[0092]

[0093] in, Indicates the first The classification weight of each production stage, Indicates the first Each production link Variance of the probability of combined effects for each production batch; Indicates the first Each production link The mean of the combined effects probability corresponding to each production batch; Indicates except the first Other production stages in each production stage The mean of the combined impact probability corresponding to each production batch.

[0094] It should be noted that: The calculation method is as follows: First, calculate the remaining... Each of the production stages (other production stages) in the production process The average of the combined impact probabilities corresponding to each production batch is used, and then the probability of all other production stages is calculated. The average probability of combined effects for each production batch. For example: remaining The first stage in the production process is... The mean of the combined effects probability for each production batch is The two production stages are in The mean of the combined effects probability for each production batch is ,Will and The mean is denoted as .

[0095] The above formula is the ratio of weight value x relative value as the relative value of each production link. and The purpose of the weight value is to obtain whether the joint influence possibility of the first production link in each production batch remains stable in all production batches, the more stable, the greater the weight value; the purpose of the relative value is to obtain the relative size of the joint influence possibility corresponding to the first production link compared with other production links. The greater the relative value, the greater the possibility that the first production link and the remaining other production links interact with each other to cause changes in the production quality of the perovskite film. The greater the relative value, the greater the possibility that the first production link and the remaining other production links interact with each other to cause changes in the production quality of the perovskite film.

[0096] Step S006: Based on the classification weight of each production link, obtain all production link combinations with mutual influence.

[0097] In this embodiment, all production links are sorted in descending order of classification weight, and all production link sorting combinations are obtained by arbitrarily combining the sorted production links; all production link sorting combinations are used as a transaction set, the classification weight of each production link is used as support, and the transaction set and support are used as input to build a frequent pattern tree, and all production link combinations with mutual influence are output.

[0098] It should be noted that: in this embodiment, the classification weight of each production link is used to obtain frequent items by building a tree. The construction of the Frequent Pattern Tree frequent pattern tree and the acquisition of frequent items are prior art and will not be described here.

[0099] This embodiment improves the tree and describes the specific input and output as follows:

[0100] First, the improvement of the tree: the data processing process of the existing tree is to calculate the support of each transaction, but all items in this embodiment, i.e., production links, have the same number of occurrences in all production batches, so support processing cannot be performed, therefore, the classification weight of each production link is used instead of its support.

[0101] Secondly, the FP ​​The input to the tree is to sort all production processes in descending order of their classification weights, and then combine them arbitrarily to obtain a sorted set of all processes. This set is then used as the input for the transaction set.

[0102] Finally, improvements. FP The tree's output: its output is a combination of different production stages, for example [ , +1]、[ -2, -3, -5]. All production stages within each different combination of production stages in the output will interact with each other, leading to changes in the production quality of perovskite thin films.

[0103] Step S007: Based on the relative value of each production link in each production link combination with mutual influence, determine the strong influence combination corresponding to each production link.

[0104] In this embodiment, statistics are performed on all production links in the production link combinations that have mutual influence, and all link combinations corresponding to each production link are obtained; based on the relative value of each production link in each link combination corresponding to each production link, the classification strength of each production link in each link combination is obtained; the maximum classification strength of each production link in all link combinations is obtained, and the link combination corresponding to the maximum classification strength is determined as the strong influence combination corresponding to each production link.

[0105] In this embodiment, the mean of the relative values ​​of all production links in each link combination corresponding to each production link is calculated, and the mean is determined as the classification strength of each production link in each link combination.

[0106] It should be noted that different production stages may appear simultaneously in the above combinations of production stages. Therefore, it is necessary to verify the relative strengths of the same production stages in different combinations, as shown below:

[0107] With the first Taking a single production stage as an example, it exists in In a combination of production stages, each combination of production stages contains Each production stage.

[0108] It should be further noted that the total number of elements in each production stage combination can be equal or unequal. For ease of representation, this embodiment uses the term "elementary number". express.

[0109] With the first Taking a combination of production stages as an example, calculate the components it includes. The average of the relative values corresponding to the production links is denoted as the average of the relative values corresponding to the production links in the first production link combination. The classification intensity of the first production link in the first production link combination.

[0110] The classification intensity of the first production link in the first production link combination. The classification intensity of the first production link in the first production link combination. The classification intensity of the first production link in the first production link combination. The classification intensity of the first production link in the first production link combination. The classification intensity of the first production link in the first production link combination.

[0111] The classification intensity of the first production link in the first production link combination. The classification intensity of the first production link in the first production link combination.

[0112] Step S008: using the strong influence combination corresponding to each production link to analyze the perovskite thin film production data to be analyzed, and obtaining an analysis result.

[0113] In this embodiment, the data to be analyzed is determined; wherein the data to be analyzed includes a production batch to be analyzed and production data of each production link in the production batch to be analyzed in the perovskite thin film production process; according to the strong influence combination corresponding to each production link in the production batch to be analyzed, the production data of other production links having mutual influence on each production link is obtained; and a joint analysis method is used to analyze the production data of each production link in the production batch to be analyzed and the production data of other production links having mutual influence on each production link, and an analysis result is obtained.

[0114] It should be noted that the strong influence combination corresponding to different production links of the perovskite thin film has been obtained, and now the production data of the perovskite thin film is analyzed using the same, and the specific analysis process is as follows:

[0115] First, the perovskite thin film production batch that needs to be analyzed is determined, and the production data corresponding to different production links in the production batch is obtained. Then, taking the production data of any one production link as an example, the production data corresponding to all production links in the strong influence combination corresponding to the production link is extracted first; and the production data of the production link and the extracted production data are jointly analyzed by simultaneous analysis.

[0116] Thus, the present application is completed.

[0117] ​To sum up, the method for analyzing production data of perovskite thin film is proposed in the embodiment of the application, the mutual influence relationship of different production links on production quality is obtained by analyzing the collected production data of different production links, and the production data in the production process of perovskite thin film is analyzed based on the mutual influence relationship of different production links on production quality, so that the accuracy of the analysis result is improved.

[0118] The above merely describes preferred embodiments of the present application and is not intended to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method of analyzing production data of a perovskite thin film, characterized by, The method comprises the following steps: Collect historical data and pre-process the historical data to obtain pre-processed historical data; wherein the pre-processed historical data comprises production data of all production links of each production batch in the production process of the perovskite film, and a yield rate of each production batch; Obtain a priori correlation of each production batch by using the yield rate of each production batch; Obtain production data difference of each production link in each production batch based on corresponding production data of each production link in different production batches; Obtain joint influence possibility of each production link in each production batch according to the a priori correlation of each production batch and the production data difference of each production link in each production batch; Wherein, the expression of the joint influence possibility of each production link in each production batch is as follows: in, Indicates the first The first in the production batch The potential for combined effects across various production stages. Indicates the first The variance of production data differences across all production stages in a production batch. Indicates the first The number of production stages in a production batch Indicates the first The first production batch Differences in production data at each stage of the production process Indicates the first Prior correlation of each production batch Indicates the first The first production batch Differences in production data across different production stages; Obtain relative value and classification weight of each production link by using the joint influence possibility of each production link in each production batch; wherein, the expression of the classification weight of each production link is as follows: wherein, represents the classification weight of the th production link, represents the variance of the joint influence possibility of the th production link corresponding to the production batches; represents the mean of the joint influence possibility of the th production link corresponding to the production batches; represents the mean of the joint influence possibility of the other production links except the th production link corresponding to the production batches; the relative value of each production link is the ratio of and . Obtain all production link combinations with mutual influence based on the classification weight of each production link; determine a strong influence combination corresponding to each production link according to the relative value of each production link in each production link combination with mutual influence. Analyze the perovskite film production data to be analyzed by using the strong influence combination corresponding to each production link to obtain an analysis result.

2. The method of claim 1, wherein the method comprises: The specific steps of obtaining the a priori correlation of each production batch by using the yield rate of each production batch include the following: Obtain the mean value of the yield rates of all production batches; Obtain the a priori correlation of each production batch by using the yield rate of each production batch and the mean value of the yield rates of all production batches.

3. The method for analyzing perovskite thin film production data according to claim 2, characterized in that, The specific steps of obtaining the production data difference of each production link in each production batch based on the corresponding production data of each production link in different production batches include the following: Calculate the difference between the corresponding production data of each production link in the current production batch and the corresponding production data of the production link in each other production batch to obtain the production data difference of each production link in different production batches; Obtain the number of production batches; obtain the production data difference of each production link in each production batch based on the number of production batches and the production data difference of each production link in different production batches.

4. The method of claim 3, wherein the method is characterized by: The specific steps of obtaining the joint influence possibility of each production link in each production batch according to the a priori correlation of each production batch and the production data difference of each production link in each production batch include the following: Obtain the variance of the production data difference of all production links in each production batch by using the production data difference of each production link in each production batch; Obtain the number of production links; Obtain the joint influence possibility of each production link in each production batch based on the a priori correlation of each production batch, the number of production links, the variance of the production data difference of all production links in each production batch, and the production data difference of each production link in each production batch.

5. The method of claim 4, wherein the method is characterized by: The specific steps of obtaining the relative value and classification weight of each production link by using the joint influence possibility of each production link in each production batch include the following: Calculate the mean value of the joint influence possibility of each production link in different production batches; Calculate the mean value of the joint influence possibility of each other production link in different production batches, and calculate the mean value of the mean value of the joint influence possibility of all other production links in different production batches to obtain the mean value of the joint influence possibility of other production links in different production batches; Obtain the relative value of each production link by using the mean value of the joint influence possibility of each production link in different production batches and the mean value of the joint influence possibility of other production links in different production batches; Obtain the variance of the joint influence possibility of each production link in different production batches; Obtain the classification weight of each production link based on the variance of the joint influence possibility of each production link in different production batches and the relative value of each production link.

6. The method of claim 5, wherein the method is characterized by: The specific steps of obtaining all production link combinations with mutual influence based on the classification weight of each production link include the following: Sort all production links in descending order of classification weight, and randomly combine the sorted production links to obtain all production link sorting combinations; Take all production link sorting combinations as a transaction set, take the classification weight of each production link as support, and take the transaction set and support as input to build a frequent pattern tree, and output all production link combinations with mutual influence.

7. The method of claim 6, wherein the method comprises: The specific steps of determining the strong influence combination corresponding to each production link based on the relative value of each production link in each production link combination with mutual influence include the following: Count the production links in all production link combinations with mutual influence to obtain all link combinations corresponding to each production link; Obtain the classification intensity of each production link in each link combination based on the relative value of each production link in each link combination corresponding to each production link; Obtain the maximum classification intensity of each production link in all link combinations, and determine the link combination corresponding to the maximum classification intensity as the strong influence combination corresponding to each production link.

8. The method for analyzing perovskite thin film production data according to claim 7, characterized in that, The specific steps of obtaining the classification intensity of each production link in each link combination based on the relative value of each production link in each link combination corresponding to each production link include the following: Calculate the mean value of the relative values of all production links in each link combination corresponding to each production link, and determine the mean value as the classification intensity of each production link in each link combination. 9.The method of claim 7, wherein the method further comprises: The specific steps of analyzing the perovskite thin film production data to be analyzed by using the strong influence combination corresponding to each production link to obtain an analysis result include the following: ​ Determine the data to be analyzed; wherein the data to be analyzed includes the production data of each production link in the production batch to be analyzed in the perovskite thin film production process. According to the strong influence combination corresponding to each production link in the production batch to be analyzed, production data of other production links having mutual influence on each production link are obtained; By using the joint analysis method, the production data of each production link in the production batch to be analyzed and the production data of other production links having mutual influence on each production link are analyzed to obtain an analysis result. 10.The method of claim 1, wherein the method further comprises: The specific steps of preprocessing the historical data to obtain the preprocessed historical data include the following: ​ The historical data is subjected to data cleaning to obtain the preprocessed historical data.

Images