A real-time statistic method for blast furnace burden distribution parameters

Abstract

The present invention belongs to the technical field of blast furnace production. The present invention provides a real-time distribution parameter statistics method for a blast furnace. When a parameter missing signal is generated, the missing number ratio and the missing discrete value are obtained, the missing number ratio and the missing discrete value are ratio-calculated to obtain a missing degree value, all missing degree values ​​within a collection period are added and averaged to obtain a collection adjustment value, and the collection adjustment value is compared with a collection adjustment threshold. If the collection adjustment value is greater than the collection adjustment threshold, a collection adjustment signal is generated. Therefore, the collection adjustment value comprehensively reflects the missing number ratio of distribution parameters in all missing data groups and the discrete degree of the missing value distribution. The larger this value is, the greater the missing number ratio of parameters in all missing data groups within the collection period is, and the distribution of the missing values ​​is more concentrated, thereby reducing the accuracy of blast furnace distribution parameter collection and resulting in incomplete blast furnace distribution parameter collection.

Description

Technical Field

[0001] The invention belongs to the technical field of blast furnace data collection, in particular to a real-time blast furnace burden parameter statistics method. Background Art

[0002] Blast furnace charging refers to the process of loading charge (including ore and coke) into the throat of a blast furnace in a certain manner and according to certain parameters through charging equipment. During this process, it is necessary to collect and count the blast furnace charging parameters in real time to provide data basis for subsequent blast furnace condition adjustments. In actual operation, blast furnace process operators are unable to collect and count charging parameters for each angle. When using the background system to collect and count parameters, it is also necessary to avoid the problem of incomplete blast furnace charging parameter collection due to missing parameters.

[0003] In the existing technology, the zoned charging method is to divide the interior of the blast furnace into different areas and carry out differentiated charging according to the characteristics of the areas. Since in the actual operation process, the blast furnace process operators are unable to collect and count the charging parameters for each angle, and parameter missing is likely to occur when the background system is used to collect and count the parameters. Therefore, it is necessary to analyze whether the charging parameters collected in real time are abnormal, and evaluate whether the abnormality type is parameter missing. When it is determined that the abnormality type is parameter missing, the degree of missing of the collected charging parameters is analyzed. If the degree of missing of the collected charging parameters is large, a collection adjustment signal is generated to avoid incomplete charging parameter information during the entire collection cycle, thereby solving the problem of incomplete collection of blast furnace charging parameters due to a large degree of missing of charging parameters.

[0004] To this end, the present invention provides a method for statistically analyzing real-time burden distribution parameters of a blast furnace. Summary of the Invention

[0005] In order to make up for the deficiencies of the prior art, the problem of incomplete collection of blast furnace burden distribution parameters due to a large degree of missing data as mentioned in the background art is solved.

[0006] The technical solution adopted by the present invention to solve its technical problem is:

[0007] A method for calculating real-time burden distribution parameters of a blast furnace, comprising:

[0008] Step 1: During the collection period, real-time collection is performed on the blast furnace charge distribution area to obtain charge distribution parameters, the charge distribution parameters are processed to obtain a charge distribution data group, and the charge distribution parameters in the charge distribution data group are pre-analyzed to obtain an abnormal data group and a non-abnormal data group;

[0009] Step 2: Based on the abnormal data group and the non-abnormal data group, the abnormal data group is processed and analyzed to obtain a type identification value, which is compared with the threshold. If the type identification value is greater than or equal to the type identification threshold, a parameter missing signal is generated, and the abnormal data group corresponding to the parameter missing signal is marked as a missing data group;

[0010] Step 3: Based on the parameter missing signal, analyze and process the missing data group to obtain an acquisition adjustment value, and compare it with the threshold. If the acquisition adjustment value is greater than the acquisition adjustment threshold, generate an acquisition adjustment signal;

[0011] Step 4: Based on the acquisition adjustment signal, obtain the acquisition adjustment coefficient and complete the dynamic adjustment of the blast furnace burden parameters by adjusting the acquisition frequency.

[0012] As a further solution of the present invention, the abnormal data group is obtained by:

[0013] In a collection cycle, the collection cycle is divided into several collection periods, and in the collection period, each collection period is divided into several collection nodes, and the cloth parameters corresponding to each collection node are obtained, and all the cloth parameters in each collection period are integrated to obtain a cloth data group;

[0014] Calculate the standard deviation of all fabric parameters in the fabric data set to obtain the discrete value of the parameters;

[0015] Compare the parameter discrete value with the parameter discrete threshold. If the parameter discrete value is greater than the parameter discrete threshold, mark the fabric data group as an abnormal data group.

[0016] If the parameter discrete value is less than or equal to the parameter discrete threshold, the fabric data set is marked as a non-abnormal data set.

[0017] As a further solution of the present invention, the type identification value is obtained by:

[0018] Substitute the parameter number difference coefficient and the parameter acquisition difference coefficient into the formula: The type identification value SB is calculated, where GS represents the abnormal number difference coefficient, CJ represents the abnormal collection difference coefficient, and α and β represent the preset weight coefficients.

[0019] As a further solution of the present invention, the abnormal number difference coefficient is obtained as follows:

[0020] Get the number of fabric parameters in the abnormal data group and mark it as the number of abnormal parameters;

[0021] Subtract the number of abnormal parameters from the threshold of the number of abnormal parameters, take the absolute value, and get the abnormal number difference;

[0022] The anomaly number difference is calculated by comparing the anomaly number threshold with the anomaly number difference coefficient.

[0023] As a further solution of the present invention, the abnormal acquisition difference coefficient is obtained as follows:

[0024] In the abnormal data group, obtain the abnormal collection times, subtract the abnormal collection times from the abnormal collection times threshold, take the absolute value, and obtain the abnormal collection difference;

[0025] The abnormal acquisition difference value is calculated by ratioing the abnormal acquisition number threshold to obtain the abnormal acquisition difference coefficient.

[0026] As a further solution of the present invention, the acquisition method of the acquisition adjustment value is:

[0027] All missing degree values ​​within the collection period are added together and averaged to obtain the collection adjustment value.

[0028] As a further solution of the present invention, the method for obtaining the missing degree value is:

[0029] The missing degree value is obtained by calculating the ratio of the missing number proportion to the missing discrete value.

[0030] As a further solution of the present invention, the method for obtaining the missing number ratio is as follows:

[0031] Obtain the number of missing parameters in each missing data group, and sum the number of missing parameters in each missing data group to obtain the number of missing parameters;

[0032] The ratio of the number of missing parameters to the total number of parameters collected during the collection period is calculated to obtain the missing parameter ratio.

[0033] As a further solution of the present invention, the missing discrete coefficient is obtained as follows:

[0034] Get the collection nodes corresponding to the missing parameters in each missing data group and mark them as missing nodes;

[0035] Get the interval time between all missing nodes, calculate the standard deviation of the interval time between all missing nodes, and get the missing discrete value;

[0036] The missing discrete value is calculated by ratioing the time length value corresponding to the acquisition period to obtain the missing discrete coefficient.

[0037] As a further solution of the present invention, the dynamic adjustment process of blast furnace burden parameters is as follows:

[0038] Substitute the missingness value into the formula: The acquisition adjustment coefficient K is calculated, where QS is the missing degree value, QSy It is expressed as the missingness threshold;

[0039] Based on the collection adjustment coefficient, the collection adjustment coefficient and the missing degree value are substituted into the formula: F t =F d +F d ×K, calculate the adjusted acquisition frequency F t , where F d It represents the current acquisition frequency, and K represents the acquisition adjustment coefficient.

[0040] The beneficial effects of the present invention are as follows:

[0041] (1) The present invention performs real-time data collection on the blast furnace charging area within the collection period, obtains charging parameters, processes the charging parameters to obtain a charging data group, performs pre-analysis processing on the charging parameters in the charging data group to obtain an abnormal data group and a non-abnormal data group, and processes and analyzes the charging parameters in the abnormal data group and the non-abnormal data group respectively to obtain a type identification value, and compares it with a threshold value. If the type identification value is greater than or equal to the type identification threshold value, a parameter missing signal is generated, thereby analyzing the parameters in the abnormal data group and identifying whether the charging parameters in the abnormal data group are missing;

[0042] (2) When generating a parameter missing signal, the present invention obtains the missing number ratio and the missing discrete value, calculates the ratio of the missing number ratio to the missing discrete value, obtains the missing degree value, adds and averages all the missing degree values ​​in the acquisition period, obtains the acquisition adjustment value, and compares the acquisition adjustment value with the acquisition adjustment threshold. If the acquisition adjustment value is greater than the acquisition adjustment threshold, an acquisition adjustment signal is generated. Therefore, the acquisition adjustment value comprehensively reflects the missing number ratio of the distribution parameters in all missing data groups and the discrete degree of the missing value distribution. The larger this value is, the greater the missing number ratio of the parameters in all missing data groups in the acquisition period is, and the distribution of the missing values ​​is more concentrated, thereby reducing the accuracy of the blast furnace distribution parameter acquisition, resulting in incomplete blast furnace distribution parameter acquisition;

[0043] (3) When generating an acquisition adjustment signal, the present invention substitutes the missing degree value into the formula to obtain the acquisition adjustment coefficient, and substitutes the acquisition adjustment coefficient and the current acquisition frequency into the formula to obtain the adjusted acquisition frequency. Thus, when the blast furnace burden distribution parameters are collected and counted in real time, if parameter collection is missing, the acquisition frequency can be adjusted in time to avoid incomplete burden distribution parameter information in the entire collection cycle, thereby solving the problem of incomplete blast furnace burden distribution parameter collection due to a large missing degree. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] The present invention will be further described below with reference to the accompanying drawings.

[0045] Figure 1 is a flowchart of the steps of Example 1 of the present invention;

[0046] Figure 2 is a flowchart of the steps of Example 2 of the present invention;

[0047] Figure 3 This is a flowchart of the steps of Example 3 of the present invention. DETAILED DESCRIPTION

[0048] In order to make the technical means, creative features, objectives and effects achieved by the present invention easier to understand, the present invention is further described below in conjunction with specific implementation methods.

[0049] Example 1

[0050] like Figure 1 As shown, a method for calculating real-time burden distribution parameters of a blast furnace according to an embodiment of the present invention includes:

[0051] Step 1: During the collection period, real-time collection is performed on the blast furnace charge distribution area to obtain charge distribution parameters, the charge distribution parameters are processed to obtain a charge distribution data group, and the charge distribution parameters in the charge distribution data group are pre-analyzed to obtain an abnormal data group and a non-abnormal data group;

[0052] In some embodiments, within a collection cycle, the collection cycle is divided into a number of collection periods, within the collection period, each collection period is divided into a number of collection nodes, the cloth parameters corresponding to each collection node are obtained, and all cloth parameters within each collection period are integrated to obtain a cloth data group;

[0053] It can be understood that: the collection period is divided into a number of collection periods at equal time intervals;

[0054] It should be noted that: the fabric parameters collected in each collection period are integrated and arranged according to the time sequence of the collection nodes to obtain the fabric data group;

[0055] The pre-analysis process of the cloth parameters in the cloth data group is as follows:

[0056] Calculate the standard deviation of all fabric parameters in the fabric data set to obtain the discrete value of the parameters;

[0057] It can be understood that the parameter dispersion value means that in the fabric data set, by calculating the standard deviation of all fabric parameters, it reflects the degree of dispersion of the fabric parameters in the fabric data set. Specifically, the larger the value, the more dispersed the fabric parameters in the fabric data set, the greater the difference between each fabric parameter, and the greater the possibility of data anomaly. Conversely, the smaller the value, the more concentrated the fabric parameters in the fabric data set, the smaller the difference between each fabric parameter, and the less likely the possibility of data anomaly.

[0058] The parameter discrete value is compared with the parameter discrete threshold. The comparison process is as follows:

[0059] If the parameter discrete value is greater than the parameter discrete threshold, it means that the fabric parameters in the fabric data set are more dispersed and the differences between the fabric parameters are larger, and the fabric data set is marked as an abnormal data set;

[0060] If the parameter discrete value is less than or equal to the parameter discrete threshold, it means that the fabric parameters in the fabric data group are more concentrated and the differences between the fabric parameters are smaller, and the fabric data group is marked as a non-abnormal data group;

[0061] Step 2: Based on the abnormal data group, process and analyze the fabric parameters in the abnormal data group to obtain the type identification value, and compare it with the threshold to generate a parameter missing signal;

[0062] The parameter missing signal includes a parameter missing signal and a parameter non-missing signal;

[0063] In some embodiments, the number of fabric parameters in the abnormal data group is obtained and marked as the number of abnormal parameters;

[0064] Subtract the number of abnormal parameters from the threshold of the number of abnormal parameters, take the absolute value, and get the abnormal number difference;

[0065] Calculate the ratio of the anomaly number difference to the anomaly number threshold to obtain the anomaly number difference coefficient;

[0066] The non-abnormal interval values ​​corresponding to all non-abnormal data groups are summed up to obtain the non-abnormal acquisition value;

[0067] In the abnormal data group, obtain the abnormal collection times, subtract the abnormal collection times from the abnormal collection times threshold, take the absolute value, and obtain the abnormal collection difference;

[0068] Calculate the ratio of the abnormal acquisition difference to the abnormal acquisition number threshold to obtain the abnormal acquisition difference coefficient;

[0069] Substitute the parameter number difference coefficient and the parameter acquisition difference coefficient into the formula: The type identification value SB is calculated, where GS represents the abnormal number difference coefficient, CJ represents the abnormal collection difference coefficient, and α and β represent the preset weight coefficients;

[0070] It can be understood that the meaning of the type identification value is: by calculating the abnormal number difference coefficient and the abnormal collection difference coefficient, it reflects whether there are missing fabric parameters in the abnormal data group. Specifically, the larger the value, the greater the difference in the number of parameters in the abnormal data group, the greater the deviation in the number of collections, and the greater the missing degree of fabric parameters in the abnormal data group. Conversely, the smaller the value, the smaller the difference in the number of parameters in the abnormal data group, the smaller the deviation in the number of collections, and the smaller the missing degree of fabric parameters in the abnormal data group.

[0071] The type identification value is compared with the type identification threshold. The comparison process is as follows:

[0072] If the type identification value is greater than or equal to the type identification threshold, it means that the number of parameters in the abnormal data group is relatively different and the deviation of the number of acquisitions is relatively large, and a parameter missing signal is generated. The abnormal data group corresponding to the parameter missing signal is marked as a missing data group;

[0073] If the type identification value is less than the type identification threshold, it means that the difference in the number of parameters in the abnormal data group is small and the deviation in the number of acquisitions is small, and a parameter non-missing signal is generated. The abnormal data group corresponding to the parameter missing signal is marked as a non-missing data group;

[0074] The specific implementation scheme of the embodiment of the present invention is as follows: within the collection period, real-time collection is performed on the blast furnace charging area to obtain charging parameters, the charging parameters are processed to obtain a charging data group, the charging parameters in the charging data group are pre-analyzed and processed to obtain an abnormal data group and a non-abnormal data group, and the charging parameters in the abnormal data group and the non-abnormal data group are processed and analyzed respectively to obtain a type identification value, and compared with a threshold value. If the type identification value is greater than or equal to the type identification threshold value, a parameter missing signal is generated, thereby analyzing the parameters in the abnormal data group and identifying whether the charging parameters in the abnormal data group are missing.

[0075] Example 2

[0076] like Figure 2 As shown, based on Example 1, a real-time blast furnace burden distribution parameter statistics method according to an embodiment of the present invention includes:

[0077] Step 3: Based on the parameter missing signal, analyze and process the missing data group to obtain the acquisition adjustment value, and compare it with the threshold to generate the acquisition adjustment signal;

[0078] The collection of whether to adjust the signal package includes collecting the adjustment signal and collecting the non-adjustment signal;

[0079] In some embodiments, when a parameter missing signal is generated, the missing number ratio and the missing discrete value are obtained to obtain the acquisition adjustment value. The specific process is as follows:

[0080] Obtain the number of missing parameters in each missing data group, and sum the number of missing parameters in each missing data group to obtain the number of missing parameters;

[0081] Calculate the ratio of the number of missing parameters to the total number of parameters collected during the collection period to obtain the missing number ratio;

[0082] Get the collection nodes corresponding to the missing parameters in each missing data group and mark them as missing nodes;

[0083] Get the interval time between all missing nodes, calculate the standard deviation of the interval time between all missing nodes, and get the missing discrete value;

[0084] Calculate the ratio of the missing discrete value to the time length value corresponding to the acquisition period to obtain the missing discrete coefficient;

[0085] Calculate the ratio of missing number proportion to missing discrete value to obtain missing degree value;

[0086] It can be understood that the meaning of the missing degree value is: by calculating the proportion of missing parameters in the missing data group and the proportion of the dispersion of missing values, it comprehensively reflects the missing degree of parameters in the missing data group. Specifically, the larger the value, the larger the proportion of missing parameters in the missing data group, and the more concentrated the distribution of missing values. Conversely, the smaller the value, the smaller the proportion of missing parameters in the missing data group, and the more dispersed the distribution of missing values.

[0087] All missing degree values ​​within the collection period are summed and averaged to obtain the collection adjustment value;

[0088] It should be noted that the meaning of the acquisition adjustment value is: a comprehensive evaluation index after quantifying the missing degree of all missing data groups during the acquisition period. This value comprehensively reflects the proportion of missing parameters in all missing data groups and the discrete degree of the missing value distribution. Specifically, the larger this value is, the larger the proportion of missing parameters in all missing data groups during the acquisition period, the more concentrated the distribution of missing values, and the less complete the parameter information in all missing data groups during the acquisition period. Conversely, the smaller this value is, the smaller the proportion of missing parameters in all missing data groups during the acquisition period, the more dispersed the distribution of missing values, and the more complete the parameter information in all missing data groups during the acquisition period.

[0089] Compare the acquisition adjustment value with the acquisition adjustment threshold. The comparison process is as follows:

[0090] If the acquisition adjustment value is greater than the acquisition adjustment threshold, it means that the number of missing parameters in all missing data groups during the acquisition period accounts for a large proportion, the distribution of missing values ​​is relatively concentrated, and the parameter information in all missing data groups during the acquisition period is relatively incomplete, and an acquisition adjustment signal is generated;

[0091] If the acquisition adjustment value is less than or equal to the acquisition adjustment threshold, it means that the number of missing parameters in all missing data groups during the acquisition period is small, the distribution of missing values ​​is relatively dispersed, and the parameter information in all missing data groups during the acquisition period is relatively complete, and a non-acquisition adjustment signal is generated;

[0092] The specific implementation plan of the embodiment of the present invention is: when a parameter missing signal is generated, the missing number ratio and the missing discrete value are obtained, the missing number ratio and the missing discrete value are calculated to obtain a missing degree value, all the missing degree values ​​in the acquisition period are added and averaged to obtain an acquisition adjustment value, and the acquisition adjustment value is compared with the acquisition adjustment threshold. If the acquisition adjustment value is greater than the acquisition adjustment threshold, an acquisition adjustment signal is generated. Therefore, the acquisition adjustment value comprehensively reflects the missing number ratio of the distribution parameters in all missing data groups and the discrete degree of the missing value distribution. The larger this value is, the larger the missing number ratio of the parameters in all missing data groups in the acquisition period is, and the distribution of the missing values ​​is more concentrated, thereby reducing the accuracy of the blast furnace distribution parameter acquisition, resulting in incomplete blast furnace distribution parameter acquisition.

[0093] Example 3

[0094] like Figure 3 As shown, based on the above-mentioned embodiment 1 and embodiment 2, a method for statistically analyzing real-time burden distribution parameters of a blast furnace according to an embodiment of the present invention includes:

[0095] Step 4: Based on the acquisition adjustment signal, the acquisition adjustment coefficient is obtained, and the dynamic adjustment of the blast furnace burden parameters is completed by adjusting the acquisition frequency;

[0096] In some embodiments, when an acquisition adjustment signal is generated, an acquisition adjustment coefficient is obtained, and the acquisition process is as follows:

[0097] Substitute the missingness value into the formula: The acquisition adjustment coefficient K is calculated, where QS is the missing degree value, QS y It is expressed as the missingness threshold;

[0098] Based on the collection adjustment coefficient, the collection adjustment coefficient and the missing degree value are substituted into the formula: F t =F d +F d×K, calculate the adjusted acquisition frequency F t , where F d It represents the current acquisition frequency, and K represents the acquisition adjustment coefficient;

[0099] The specific implementation scheme of the embodiment of the present invention is as follows: when generating an acquisition adjustment signal, the missing degree value is substituted into the formula to obtain the acquisition adjustment coefficient, and the acquisition adjustment coefficient and the current acquisition frequency are substituted into the formula to obtain the adjusted acquisition frequency. Therefore, when performing real-time acquisition and statistics of blast furnace burden distribution parameters, if parameter collection is missing, the acquisition frequency can be adjusted in time to avoid incomplete burden distribution parameter information throughout the entire acquisition cycle, thereby solving the problem of incomplete blast furnace burden distribution parameter collection due to a large missing degree.

[0100] The basic principles, main features, and advantages of the present invention are shown and described above. Those skilled in the art should understand that the present invention is not limited to the foregoing embodiments. The foregoing embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for calculating real-time blast furnace burden parameters, characterized by: include: Step 1: During the collection period, real-time collection is performed on the blast furnace charge distribution area to obtain charge distribution parameters, the charge distribution parameters are processed to obtain a charge distribution data group, and the charge distribution parameters in the charge distribution data group are pre-analyzed to obtain an abnormal data group and a non-abnormal data group; Step 2: Based on the abnormal data group and the non-abnormal data group, the abnormal data group is processed and analyzed to obtain a type identification value, which is compared with the threshold. If the type identification value is greater than or equal to the type identification threshold, a parameter missing signal is generated, and the abnormal data group corresponding to the parameter missing signal is marked as a missing data group; Step 3: Based on the parameter missing signal, analyze and process the missing data group to obtain an acquisition adjustment value, and compare it with the threshold. If the acquisition adjustment value is greater than the acquisition adjustment threshold, generate an acquisition adjustment signal; Get the number of fabric parameters in the abnormal data group and mark it as the number of abnormal parameters; Subtract the number of abnormal parameters from the threshold of the number of abnormal parameters, take the absolute value, and get the abnormal number difference; Calculate the ratio of the anomaly number difference to the anomaly number threshold to obtain the anomaly number difference coefficient; In the abnormal data group, obtain the abnormal collection times, subtract the abnormal collection times from the abnormal collection times threshold, take the absolute value, and obtain the abnormal collection difference; Calculate the ratio of the abnormal acquisition difference to the abnormal acquisition number threshold to obtain the abnormal acquisition difference coefficient; Substitute the anomaly number difference coefficient and the anomaly collection difference coefficient into the formula: , calculate the type identification value SB, where GS represents the abnormal number difference coefficient, CJ represents the abnormal collection difference coefficient, 、 Expressed as a preset weight coefficient; The ratio of the missing number ratio to the missing dispersion coefficient is calculated to obtain the missing degree value; All missing degree values ​​within the collection period are summed and averaged to obtain the collection adjustment value; Step 4: Based on the acquisition adjustment signal, the acquisition adjustment coefficient is obtained, and the dynamic adjustment of the blast furnace burden parameters is completed by adjusting the acquisition frequency; Substitute the missingness value into the formula: , calculate the acquisition adjustment coefficient K, where, Expressed as the missing degree value, It is expressed as the missingness threshold; Based on the acquisition adjustment coefficient, substitute the acquisition adjustment coefficient and the current acquisition frequency into the formula: , calculate the adjusted acquisition frequency ,in, It represents the current acquisition frequency, and K represents the acquisition adjustment coefficient.

2. The method for calculating real-time blast furnace burden distribution parameters according to claim 1, wherein: The abnormal data group is obtained as follows: In a collection cycle, the collection cycle is divided into several collection periods, and in the collection period, each collection period is divided into several collection nodes, and the cloth parameters corresponding to each collection node are obtained, and all the cloth parameters in each collection period are integrated to obtain a cloth data group; Calculate the standard deviation of all fabric parameters in the fabric data set to obtain the discrete value of the parameters; Compare the parameter discrete value with the parameter discrete threshold. If the parameter discrete value is greater than the parameter discrete threshold, mark the fabric data group as an abnormal data group. If the parameter discrete value is less than or equal to the parameter discrete threshold, the fabric data set is marked as a non-abnormal data set.

3. The method for calculating real-time blast furnace burden distribution parameters according to claim 1, wherein: The missing number ratio is obtained as follows: Obtain the number of missing parameters in each missing data group, and sum the number of missing parameters in each missing data group to obtain the number of missing parameters; The ratio of the number of missing parameters to the total number of parameters collected during the collection period is calculated to obtain the missing parameter ratio.

4. The method for calculating real-time blast furnace burden distribution parameters according to claim 1, wherein: The missing dispersion coefficient is obtained as follows: Get the collection nodes corresponding to the missing parameters in each missing data group and mark them as missing nodes; Get the interval time between all missing nodes, calculate the standard deviation of the interval time between all missing nodes, and get the missing discrete value; The missing discrete value is calculated by ratioing the time length value corresponding to the acquisition period to obtain the missing discrete coefficient.

Images