Blast furnace influence factor importance analysis method and system based on big data platform, and computer readable storage medium

Abstract

The invention provides a blast furnace influence factor importance analysis method and system based on a big data platform, and a computer readable storage medium, and the blast furnace influence factor importance analysis method based on the big data platform comprises the following steps: based on a preset analysis target; a target dependent variable and a source variable or a process variable corresponding to the target dependent variable are determined from all the influence variables of the target blast furnace as independent variables, all the variables are sampled, and data values at all sampling time points form a to-be-analyzed data set; based on a to-be-analyzed data set, in combination with a screening processing method, importance analysis is carried out by utilizing GCA causal logic analysis, PCA principal component analysis and PCC correlation analysis, so that key independent variables influencing target dependent variables are determined. According to the method, the key factors influencing the furnace condition of the blast furnace can be effectively extracted, so that corresponding countermeasures can be taken according to the furnace condition of the blast furnace.

Description

Technical Field

[0001] The present invention belongs to the technical field of smelting blast furnace control, and relates to a method and system for analyzing the importance of blast furnace influencing factors based on a big data platform, and a computer-readable storage medium capable of executing the method for analyzing the importance of blast furnace influencing factors based on a big data platform. Background Art

[0002] Blast furnace smelting is a complex process, and its internal state cannot be directly observed and measured. In addition, the blast furnace condition becomes difficult to detect and analyze under the influence of multiple factors. At present, expert systems are mainly used to analyze and judge the blast furnace condition, but the application of expert systems depends on the long-term experience accumulation of staff, and with the changes in the operating environment, it takes a lot of time to update the expert system. At the same time, expert systems are more suitable for systems with fewer input, output or state variables, and the cost of using them is too high for systems with massive data.

[0003] With the improvement of process detection technology and the rapid development of data processing capabilities, data-driven fault diagnosis methods have become a current research hotspot. Therefore, how to objectively judge and analyze blast furnace conditions based on a big data platform and take advantage of data analysis has become an important part of the intelligent development of modern blast furnaces. Summary of the invention

[0004] In order to solve the above technical problems, the present invention provides a blast furnace influencing factors importance analysis method and system based on a big data platform, and a computer-readable storage medium that can execute the above method.

[0005] To achieve the above object, the present invention adopts the following technical solution:

[0006] A blast furnace influencing factor importance analysis method based on a big data platform is characterized by: analyzing the target dependent variable according to steps S1 to S5 to obtain key independent variables affecting the target dependent variable:

[0007] Step S1, based on the preset analysis target, determine the target dependent variable and the source variable or process variable corresponding to the target dependent variable from the various influencing variables of the target blast furnace as independent variables, and sample the target dependent variable and each independent variable at the same time based on the preset time interval, obtain the target dependent variable and the data values ​​of each variable corresponding to each sampling time point, construct them as the data set to be analyzed, and then execute step S2;

[0008] Step S2, screening the data set to be analyzed according to the screening processing method, and standardizing the screened data set to be analyzed to form data set A, and then executing step S3;

[0009] Step S3, for data set A, using GCA causal logic analysis, determine whether there is a causal relationship between each independent variable and the target dependent variable, and delete the independent variables in data set A that have no causal relationship with the target dependent variable and the data values ​​of each independent variable without causal relationship corresponding to each sampling time point, and then execute step S4;

[0010] Step S4, based on data set A, obtain all independent variables in data set A and the data values ​​of each variable corresponding to each sampling time point to form data set B, and use PCA principal component analysis to reduce the dimension and reconstruct data set B to form data set C, and then execute step S5;

[0011] Step S5, based on the target dependent variable and data set C, use PCC correlation analysis to calculate the correlation coefficient between each independent variable and the target dependent variable, determine the degree of correlation and correlation ranking between each independent variable and the target dependent variable, and judge the key independent variables that affect the target dependent variable through correlation ranking.

[0012] Furthermore, in step S1, when selecting source variables or process variables corresponding to the target dependent variable as independent variables, preliminary screening should be performed, that is, when any multiple variables among the source variables or process variables have a strong causal relationship, only any one of them is selected as the independent variable, and the other variables cannot be selected as independent variables.

[0013] Furthermore, in step S2, the screening processing method is: based on the preset value intervals of each variable, for the data values ​​of each sampling time point of the target dependent variable and each independent variable in the data set to be analyzed, the sampling time points whose data values ​​do not meet the preset value interval or are empty are marked respectively, and it is determined whether the number of marked sampling time points of each variable exceeds 1 / 3 of the total number of sampling time points of the variable. If so, the variable is no longer analyzed. Otherwise, it is determined whether there are marked sampling time points for each variable at the same collection time, and then if so, the data values ​​of each sampling time point corresponding to each variable at the collection time are deleted, otherwise the screening processing is completed.

[0014] Furthermore, step S3 uses GCA causal logic analysis to determine whether there is a causal relationship between each independent variable and the target dependent variable according to the following steps:

[0015] Step S31, construct an autoregressive model for the independent variables and target dependent variables of data set A, that is Where X is any independent variable, Y is the target dependent variable, t is the target time point, α τ and b τ is the model coefficient, n is the model order, τ is the lag time order, u t is the model residual, and then execute step S32;

[0016] Step S32, for the autoregressive model, set the null hypothesis H0: b1 = b2 = ... = b k =0, and the significance level P is 0.05, and the F test is used to determine whether there is a potential causal relationship between X and Y. That is, when the significance level P is less than 0.05, the null hypothesis is rejected, indicating that there is a causal relationship between X and Y.

[0017] Furthermore, the lag time order of the autoregressive model ranges from 1 to 10, that is, the causal relationship between X and Y under lag time orders of 1 to 10 is judged respectively. If there is one or more significant level P less than 0.05 in 10 F tests, it is considered that there is a certain causal relationship between X and Y.

[0018] Furthermore, step S4 uses PCA principal component analysis to reduce the dimension and reconstruct the data set B according to the following steps:

[0019] Step S41, constructing an original data matrix for the m independent variables and the data values ​​at the n sampling time points of the data set B, and calculating the covariance matrix of the original data matrix, and then executing step S42;

[0020] Step S42: m eigenvalues ​​and eigenvectors corresponding to the eigenvalues ​​are calculated according to the covariance matrix, and the m eigenvalues ​​are sorted in descending order to obtain λ1, λ2, ...λ m , and the eigenvectors a1, a2…a corresponding to the eigenvalues m , and the i-th principal component is Y i =Xa i (1≤i≤m), where X is the original data matrix, a i is the i-th feature vector, and then executes step S43;

[0021] Step S43, based on the m eigenvalues ​​arranged in descending order, calculate the principal component contribution rate and the cumulative contribution rate according to the following formula. When the cumulative contribution rate reaches a preset threshold, select the first k principal components to form a new feature space, and then execute step S44:

[0022]

[0023] where λ k is the eigenvalue of the kth principal component, is the sum of all eigenvalues, is the main component contribution rate, is the cumulative contribution rate;

[0024] Step S44, constructing a load matrix based on the eigenvector of the original data matrix, combining with a preset threshold, screening out the original independent variables in the load matrix whose composition coefficient is greater than the preset threshold as the main characteristic variables of k principal components, and then executing step S45;

[0025] Step S45, constructing a data set C based on the target dependent variable and all main characteristic variables of the k principal components.

[0026] Furthermore, step S5 calculates the correlation coefficient for the target dependent variable and each independent variable in the data set C using the following formula to determine the degree of correlation between each independent variable and the target dependent variable, and determines the key independent variables that affect the target dependent variable by arranging them in descending order according to the correlation coefficients:

[0027]

[0028] Among them, X and Y are the target dependent variable and any independent variable in the data set C, Cov(X,Y) is the covariance of X and Y, Var|X| is the variance of X, Var|Y| is the variance of Y, r(X,Y) is the correlation coefficient, and the value range of r(X,Y) is -1≤r(X,Y)≤1; the larger the absolute value of the correlation coefficient r(X,Y), the stronger the linear correlation between the two variables, and when r(X,Y) is close to 0, it means there is no linear correlation.

[0029] A blast furnace influencing factor importance analysis system based on a big data platform, characterized by comprising:

[0030] Variable data acquisition module: based on the preset analysis target, the target dependent variable and the source variable or process variable corresponding to the target dependent variable are determined from the various influencing variables of the target blast furnace as independent variables, and based on the preset time interval, the target dependent variable and each independent variable are sampled at the same time to obtain the target dependent variable and the data values ​​of each variable corresponding to each sampling time point;

[0031] Data processing module: Based on the target dependent variable and the data values ​​of each variable corresponding to each sampling time point, the data set to be analyzed is constructed, and screening and standardization are performed;

[0032] Importance analysis module: Based on the data set to be analyzed after screening and standardization, the key independent variables that affect the target dependent variable are determined through importance analysis.

[0033] Furthermore, the importance analysis module includes:

[0034] Causal logic analysis module: used to determine whether there is a causal relationship between each independent variable and the target dependent variable, and to delete independent variables that have no causal relationship with the target dependent variable;

[0035] Principal component analysis module: used to reduce the dimension and reconstruct the data set consisting of all independent variables and the data values ​​of each variable corresponding to each sampling time point;

[0036] Correlation analysis module: used to calculate the correlation coefficient between each independent variable and the target dependent variable, determine the degree of correlation and correlation ranking between each independent variable and the target dependent variable, and judge the key independent variables that affect the target dependent variable through correlation ranking.

[0037] A computer-readable storage medium having a computer program stored thereon, characterized in that when the computer program is executed by a processor, the steps in the blast furnace influencing factor importance analysis method based on a big data platform are implemented.

[0038] Compared with the prior art, the beneficial effects of the present invention are: based on causal logic analysis, principal component analysis and correlation analysis, it can effectively extract the key independent variables affecting the blast furnace condition and make corresponding countermeasures; the method is simple and easy to operate, and can significantly improve the judgment and processing of the blast furnace condition. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 This is the main framework diagram of the importance analysis method of the present invention.

[0040] Figure 2 This is the principle diagram of PCA principal component analysis. DETAILED DESCRIPTION

[0041] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings.

[0042] refer to Figure 1 , a blast furnace influencing factor importance analysis method based on a big data platform, analyzes the target dependent variable according to the following steps to obtain the key independent variables that affect the target dependent variable:

[0043] Step S1, the analysis goal is to determine the main reason for the fluctuation of the heat load of the blast furnace. The heat load is determined as the target dependent variable from various influencing variables of the target blast furnace, and the source variable corresponding to the heat load is determined as the independent variable, wherein the source variables are such as pressure difference, Si content of molten iron, gas utilization rate, and coal injection amount. In addition, a preliminary screening should be performed when selecting the independent variable, that is, when there is a strong causal relationship between any multiple variables, only any one of them is selected as the independent variable, and the other variables cannot be selected as independent variables. For example, among the three variables of wind pressure, pressure difference, and top pressure, the pressure difference can be obtained by the top pressure and The wind pressure is calculated, so the wind pressure and top pressure can be discarded, and the pressure difference is selected as the independent variable for analysis; after determining the target dependent variable and the independent variable, the target dependent variable and each independent variable are sampled at the same time with a time interval of one day, and the data values ​​of the target dependent variable and each variable corresponding to each sampling time point are obtained, and the data set to be analyzed is constructed, wherein the sampling time points should be as continuous as possible. In this embodiment, 40 source variables are selected as independent variables, and 45 samples are taken at intervals of one day, that is, each variable contains 45 sampling time points, and then step S2 is executed;

[0044] Step S2, screening the data set to be analyzed according to the screening processing method, and standardizing the screened data set to be analyzed to form data set A, wherein the screening processing method is specifically based on the standard value of each variable, for the data values ​​of each sampling time point of the target dependent variable and each independent variable in the data set to be analyzed, marking the sampling time points where the data value is higher or lower than the standard value, and the data value is empty or zero due to the blast furnace shutdown, and judging whether the number of marked sampling time points of each variable exceeds 1 / 3 of the total number of sampling time points of the variable, if yes, then no longer analyze the variable, otherwise, judge whether there is a marked sampling time point for each variable at the same collection time, and then delete the data value of each sampling time point corresponding to each variable at the collection time, otherwise, complete the screening process, in this embodiment, the screening process is used to screen out 3 independent variables, delete the data value of each sampling time point corresponding to each variable at 4 sampling times, and then execute step S3;

[0045] Step S3, for data set A, using GCA causal logic analysis, determine whether there is a causal relationship between each independent variable and the target dependent variable, and delete the independent variables in data set A that have no causal relationship with the target dependent variable and the data values ​​of each variable corresponding to each sampling time point. In this embodiment, using GCA causal logic analysis, delete 25 independent variables that have no causal relationship with the target dependent variable and the data values ​​of each variable corresponding to each sampling time point, and then execute step S4;

[0046] Specifically, step S3 includes:

[0047] Step S31, construct an autoregressive model for the independent variables and target dependent variables of data set A, that is Where X is any independent variable, Y is the target dependent variable, t is the target time point, α τ and b τ is the model coefficient, n is the model order, τ is the lag time order, u t is the model residual, and then execute step S32;

[0048] Step S32, for the autoregressive model, set the null hypothesis H0: b1 = b2 = ... = b k =0, and the significance level P is 0.05, the lag time order is 10, and the F test is used to determine whether there is a potential causal relationship between X and Y. That is, when the significance level P is less than 0.05, the null hypothesis is rejected, indicating that there is a causal relationship between X and Y;

[0049] Step S4, based on data set A, obtain all independent variables in data set A and the data values ​​of each variable corresponding to each sampling time point to form data set B, and refer to Figure 2, using PCA principal component analysis, reduce the dimension and reconstruct the data set B to form the data set C, and then execute step S5;

[0050] Specifically, step S4 performs dimension reduction and reconstruction on the data set B according to the following steps:

[0051] Step S41, construct an original data matrix for the 12 independent variables of data set B and the data values ​​at the sampling time points corresponding to the respective variables, and calculate the covariance matrix corresponding to the original data matrix, and then execute step S42

[0052] Step S42, 12 eigenvalues ​​and eigenvectors corresponding to the eigenvalues ​​are calculated according to the covariance matrix, and the 12 eigenvalues ​​are sorted in descending order to obtain λ1, λ2, ...λ 12 , and the eigenvectors a1, a2…a corresponding to the eigenvalues 12 , and the i-th principal component is Y i =Xa i (1≤i≤12), where X is the original data matrix, a i is the i-th feature vector, and then executes step S43;

[0053] Step S43, based on the 12 eigenvalues ​​arranged in descending order, calculate the principal component contribution rate and the cumulative contribution rate according to the following formula. When the cumulative contribution rate reaches 80%, select the first k principal components to form a new feature space. In this embodiment, select the first 4 principal components to form a new feature space, and then execute step S44:

[0054]

[0055] where λ k is the eigenvalue of the kth principal component, is the sum of all eigenvalues, is the main component contribution rate, is the cumulative contribution rate;

[0056] Step S44, based on the eigenvector of the original data matrix, construct a load matrix, combine the preset threshold, filter out the original independent variables in the load matrix whose composition coefficient is greater than the preset threshold as the main characteristic variables of k principal components. In this embodiment, the threshold is set to 0.3, and 6 independent variables are filtered out as the main characteristic variables of 4 principal components, and then execute step S45;

[0057] Step S45, constructing a data set C based on the target dependent variable and all main characteristic variables of the k principal components.

[0058] Step S5, based on the target dependent variable and data set C, PCC correlation analysis is used to calculate the correlation coefficient between each independent variable and the target dependent variable, determine the correlation degree and correlation ranking between each independent variable and the target dependent variable, and judge the key independent variables affecting the target dependent variable through the correlation ranking. In this embodiment, the following formula is used to calculate the correlation coefficient between each independent variable and the target dependent variable, and determine the coal injection amount as the key independent variable affecting the target dependent variable:

[0059]

[0060] Among them, X and Y are the target dependent variable and any independent variable in the data set C, Cov(X,Y) is the covariance of X and Y, Var|X| is the variance of X, Var|Y| is the variance of Y, r(X,Y) is the correlation coefficient, and the value range of r(X,Y) is -1≤r(X,Y)≤1; the larger the absolute value of the correlation coefficient r(X,Y), the stronger the linear correlation between the two variables, and when r(X,Y) is close to 0, it means there is no linear correlation.

[0061] The present invention also provides a blast furnace influencing factor importance analysis system based on a big data platform, comprising:

[0062] Variable data acquisition module: based on the preset analysis target, the target dependent variable and the source variable or process variable corresponding to the target dependent variable are determined from the various influencing variables of the target blast furnace as independent variables, and based on the preset time interval, the target dependent variable and each independent variable are sampled at the same time to obtain the target dependent variable and the data values ​​of each variable corresponding to each sampling time point;

[0063] Data processing module: Based on the target dependent variable and the data values ​​of each variable corresponding to each sampling time point, the data set to be analyzed is constructed, and screening and standardization are performed;

[0064] Importance analysis module: Based on the data set to be analyzed after screening and standardization, the key independent variables that affect the target dependent variable are determined through importance analysis.

[0065] Furthermore, the importance analysis module includes:

[0066] Causal logic analysis module: used to determine whether there is a causal relationship between each independent variable and the target dependent variable, and to delete independent variables that have no causal relationship with the target dependent variable;

[0067] Principal component analysis module: used to reduce the dimension and reconstruct the data set consisting of all independent variables and the data values ​​of each variable corresponding to each sampling time point;

[0068] Correlation analysis module: used to calculate the correlation coefficient between each independent variable and the target dependent variable, determine the degree of correlation and correlation ranking between each independent variable and the target dependent variable, and judge the key independent variables that affect the target dependent variable through correlation ranking.

[0069] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps in a method for analyzing the importance of blast furnace influencing factors based on a big data platform.

[0070] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any way. Any simple modification, change and equivalent structural change made to the above embodiment based on the technical essence of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A method for analyzing the importance of blast furnace influencing factors based on a big data platform, characterized in that: Analyze the target dependent variable according to steps S1 to S5 to obtain the key independent variables that affect the target dependent variable: Step S1, based on the preset analysis target, determine the target dependent variable and the source variable or process variable corresponding to the target dependent variable from the various influencing variables of the target blast furnace as independent variables, and sample the target dependent variable and each independent variable at the same time based on the preset time interval, obtain the target dependent variable and the data values ​​of each variable corresponding to each sampling time point, construct them as the data set to be analyzed, and then execute step S2; Step S2, screening the data set to be analyzed according to the screening processing method, and standardizing the screened data set to be analyzed to form data set A, and then executing step S3; Step S3, for data set A, using GCA causal logic analysis, determine whether there is a causal relationship between each independent variable and the target dependent variable, and delete the independent variables in data set A that have no causal relationship with the target dependent variable and the data values ​​of each independent variable without causal relationship corresponding to each sampling time point, and then execute step S4; Step S4, based on data set A, obtain all independent variables in data set A and the data values ​​of each variable corresponding to each sampling time point to form data set B, and use PCA principal component analysis to reduce the dimension and reconstruct data set B to form data set C, and then execute step S5; Step S5, based on the target dependent variable and data set C, use PCC correlation analysis to calculate the correlation coefficient between each independent variable and the target dependent variable, determine the degree of correlation and correlation ranking between each independent variable and the target dependent variable, and judge the key independent variables that affect the target dependent variable through correlation ranking.

2. The blast furnace influencing factor importance analysis method based on a big data platform according to claim 1 is characterized in that: In step S1, when selecting the source variable or process variable corresponding to the target dependent variable as the independent variable, a preliminary screening should be performed, that is, when any multiple variables among the source variables or process variables have a strong causal relationship, only any one of them is selected as the independent variable, and the other variables cannot be selected as independent variables.

3. The method for analyzing the importance of blast furnace influencing factors based on a big data platform according to claim 1 is characterized in that: In step S2, the screening processing method is: based on the preset value intervals of each variable, for the data values ​​of each sampling time point of the target dependent variable and each independent variable in the data set to be analyzed, the sampling time points whose data values ​​do not meet the preset value interval or are empty are marked respectively, and it is determined whether the number of marked sampling time points of each variable exceeds 1 / 3 of the total number of sampling time points of the variable. If so, the variable is no longer analyzed. Otherwise, it is determined whether there are marked sampling time points for each variable at the same collection time, and then if so, the data values ​​of each sampling time point corresponding to each variable at the collection time are deleted, otherwise the screening processing is completed.

4. The method for analyzing the importance of blast furnace influencing factors based on a big data platform according to claim 1 is characterized in that: The step S3 uses the following steps to determine whether there is a causal relationship between each independent variable and the target dependent variable using GCA causal logic analysis: Step S31, construct an autoregressive model for the independent variables and target dependent variables of data set A, that is Where X is any independent variable, Y is the target dependent variable, t is the target time point, α τ and b τ is the model coefficient, n is the model order, τ is the lag time order, u t is the model residual, and then execute step S32; Step S32, for the autoregressive model, set the null hypothesis H0: b1 = b2 = ... = b k =0, and the significance level P is 0.05, and the F test is used to determine whether there is a potential causal relationship between X and Y. That is, when the significance level P is less than 0.05, the null hypothesis is rejected, indicating that there is a causal relationship between X and Y.

5. The method for analyzing the importance of blast furnace influencing factors based on a big data platform according to claim 4 is characterized in that: The lag time order of the autoregressive model ranges from 1 to 10, that is, the causal relationship between X and Y under lag time orders of 1 to 10 is determined respectively. If there is one or more significant level P less than 0.05 in 10 F tests, it is considered that there is a certain causal relationship between X and Y.

6. The blast furnace influencing factor importance analysis method based on a big data platform according to claim 1 is characterized in that: The step S4 is based on the following steps, using PCA principal component analysis to reduce the dimension and reconstruct the data set B: Step S41, constructing an original data matrix for the m independent variables and the data values ​​at the n sampling time points of the data set B, and calculating the covariance matrix of the original data matrix, and then executing step S42; Step S42: m eigenvalues ​​and eigenvectors corresponding to the eigenvalues ​​are calculated according to the covariance matrix, and the m eigenvalues ​​are sorted in descending order to obtain λ1, λ2, ...λ m , and the eigenvectors a1, a2…a corresponding to the eigenvalues m , and the i-th principal component is Y i =Xa i (1≤i≤m), where X is the original data matrix, a i is the i-th feature vector, and then executes step S43; Step S43, based on the m eigenvalues ​​arranged in descending order, calculate the principal component contribution rate and the cumulative contribution rate according to the following formula. When the cumulative contribution rate reaches a preset threshold, select the first k principal components to form a new feature space, and then execute step S44: where λ k is the eigenvalue of the kth principal component, is the sum of all eigenvalues, is the main component contribution rate, is the cumulative contribution rate; Step S44, constructing a load matrix based on the eigenvector of the original data matrix, combining with a preset threshold, screening out the original independent variables in the load matrix whose composition coefficient is greater than the preset threshold as the main characteristic variables of k principal components, and then executing step S45; Step S45, constructing a data set C based on the target dependent variable and all main characteristic variables of the k principal components.

7. The method for analyzing the importance of blast furnace influencing factors based on a big data platform according to claim 1 is characterized in that: In step S5, the correlation coefficient is calculated for the target dependent variable and each independent variable in the data set C using the following formula to determine the degree of correlation between each independent variable and the target dependent variable, and the key independent variables affecting the target dependent variable are determined by arranging them in descending order according to the correlation coefficients: Among them, X and Y are the target dependent variable and any independent variable in the data set C, Cov(X,Y) is the covariance of X and Y, Var|X| is the variance of X, Var|Y| is the variance of Y, r(X,Y) is the correlation coefficient, and the value range of r(X,Y) is -1≤r(X,Y)≤1; the larger the absolute value of the correlation coefficient r(X,Y), the stronger the linear correlation between the two variables, and when r(X,Y) is close to 0, it means there is no linear correlation.

8. A blast furnace influencing factor importance analysis system based on a big data platform, characterized in that: include: Variable data acquisition module: based on the preset analysis target, the target dependent variable and the source variable or process variable corresponding to the target dependent variable are determined from the various influencing variables of the target blast furnace as independent variables, and based on the preset time interval, the target dependent variable and each independent variable are sampled at the same time to obtain the target dependent variable and the data values ​​of each variable corresponding to each sampling time point; Data processing module: Based on the target dependent variable and the data values ​​of each variable corresponding to each sampling time point, the data set to be analyzed is constructed, and screening and standardization are performed; Importance analysis module: Based on the data set to be analyzed after screening and standardization, the key independent variables that affect the target dependent variable are determined through importance analysis.

9. The blast furnace influencing factor importance analysis system based on a big data platform according to claim 8 is characterized in that: The importance analysis module includes: Causal logic analysis module: used to determine whether there is a causal relationship between each independent variable and the target dependent variable, and to delete independent variables that have no causal relationship with the target dependent variable; Principal component analysis module: used to reduce the dimension and reconstruct the data set consisting of all independent variables and the data values ​​of each variable corresponding to each sampling time point; Correlation analysis module: used to calculate the correlation coefficient between each independent variable and the target dependent variable, determine the degree of correlation and correlation ranking between each independent variable and the target dependent variable, and judge the key independent variables that affect the target dependent variable through correlation ranking.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps in a method for analyzing the importance of blast furnace influencing factors based on a big data platform as described in any one of claims 1 to 7 are implemented.

Images