Heating furnace gas consumption prediction method, device, equipment, medium and product

Abstract

The invention discloses a heating furnace gas consumption prediction method and device, equipment, a medium and a product, and relates to the technical field of metallurgy thermal energy conservation, and the method comprises the steps: dividing a first sample set and a second sample set into a first training set, a first verification set, a second training set and a second verification set according to a set proportion; training and verifying the ARIMA model according to the second training set and the second verification set to obtain a gas calorific value prediction model; clustering the first training set to obtain a cluster corresponding to the first training set; clustering the first verification set by taking a cluster corresponding to the first training set as a reference to obtain a cluster corresponding to the first sample set; and training and verifying the burnup prediction mathematical model corresponding to each cluster according to the second training set, the second verification set, the first training set, the clusters corresponding to the first sample set and the predicted gas calorific value average value to obtain a burnup prediction model set. According to the method and the device, the prediction accuracy of the gas consumption of the heating furnace can be improved.

Description

Technical Field

[0001] The present application relates to the field of metallurgical thermal energy-saving technology, and in particular to a method, device, equipment, medium and product for predicting coal gas consumption of a heating furnace. Background Art

[0002] As steel companies continue to grow and develop, more and more steel companies are not only concerned about steel production, but also energy consumption. With the increase in energy consumption, it will inevitably lead to an increase in corporate energy costs and output costs, and also bring more pollution emissions. Therefore, energy consumption is of great significance to the development of steel companies.

[0003] In the related research on heating furnace energy saving and consumption reduction, the relevant data is captured, and the law of heating furnace gas consumption is tried to be discovered from the data, so as to predict the heating furnace gas consumption according to the production plan, and then production is carried out according to the predicted heating furnace gas consumption to avoid heating furnace gas waste. However, the heating furnace gas consumption predicted by this method is not accurate. Summary of the invention

[0004] The purpose of this application is to provide a method, device, equipment, medium and product for predicting the gas consumption of a heating furnace, which can improve the prediction accuracy of the gas consumption of a heating furnace.

[0005] To achieve the above objectives, this application provides the following solutions:

[0006] In a first aspect, the present application provides a method for predicting gas consumption of a heating furnace, comprising:

[0007] Divide the preset time period into multiple time periods, and obtain a first sample set and a second sample set; the first sample set includes a production data set when the heating furnace heats the steel billet in each time period; the production data set includes an average value of each furnace-entry production data of the heating furnace; the second sample set includes a calorific value fuel consumption set when the heating furnace heats the steel billet in each time period; the calorific value fuel consumption set includes an average gas calorific value and an average gas fuel consumption;

[0008] Divide the first sample set into a first training set and a first validation set according to a set ratio;

[0009] The second sample set is divided into a second training set and a second validation set according to a set ratio; the second training set includes a set of calorific value combustion when the steel billet is heated from the 1st time period to the mth time period, and the second validation set includes a set of calorific value combustion when the steel billet is heated from the m+1th time period to the nth time period; n represents the number of all time periods included in the preset time period;

[0010] The ARIMA model is trained and verified according to the second training set and the second verification set to obtain a gas calorific value prediction model;

[0011] The first training set is clustered to obtain a cluster corresponding to the first training set; one cluster corresponds to all steel billets with the same characteristics entering the furnace;

[0012] Taking the cluster corresponding to the first training set as a benchmark, clustering the first validation set is performed to obtain the cluster corresponding to the first sample set;

[0013] According to the second training set, the second validation set, the cluster corresponding to the first training set, the cluster corresponding to the first sample set and the predicted average gas calorific value when the heating furnace heats the steel billet in each time period, the fuel consumption prediction mathematical model corresponding to each cluster in the target cluster set is trained and verified to obtain a fuel consumption prediction model set, wherein the fuel consumption prediction model set is used to predict the gas consumption of the heating furnace; the target cluster set includes the cluster corresponding to the first training set; the predicted average gas calorific value is obtained by the gas calorific value prediction model.

[0014] Optionally, the furnace production data specifically includes:

[0015] The plan includes the average furnace time, the average length of the billet, the average width of the billet, the average thickness of the billet, the average specific heat capacity of the billet, the average density of the billet, the average temperature of the billet entering the furnace and the target average temperature of the billet out of the furnace.

[0016] Optionally, the step of training and validating the ARIMA model according to the second training set and the second validation set to obtain a gas calorific value prediction model specifically includes:

[0017] At the current iteration number, the number of autoregressive items and the number of moving average items in the ARIMA model at the current iteration number are calculated based on the ARIMA model at the previous iteration number, and the ARIMA model at the current iteration number is constructed; the number of autoregressive items and the number of moving average items in the ARIMA model at the initial iteration number are obtained based on the average gas calorific value in the second training set;

[0018] The ARIMA model at the current iteration number is trained according to the average gas calorific value in the second training set to obtain the trained ARIMA model at the current iteration number;

[0019] If i=m+1, the average gas calorific value of the ipth time period to the mth time period in the second training set is input into the ARIMA model trained at the current number of iterations to obtain the predicted average gas calorific value of the ith time period; p is the number of autoregressive terms in the ARIMA model at the current number of iterations;

[0020] When m+1+p is less than or equal to n, if m+1<i<m+1+p, the average gas calorific value of the ipth time period to the average gas calorific value of the mth time period and the predicted gas calorific value of the m+1th time period to the predicted gas calorific value of the i-1th time period in the second training set are input into the ARIMA model trained at the current number of iterations to obtain the predicted gas calorific value of the ith time period; p is the number of autoregressive terms in the ARIMA model at the current number of iterations;

[0021] If m+1+p≤i≤n, the average predicted gas calorific value in the ipth time period to the average predicted gas calorific value in the i-1th time period are input into the ARIMA model trained at the current number of iterations to obtain the average predicted gas calorific value in the i-th time period; p is the number of autoregressive terms in the ARIMA model at the current number of iterations;

[0022] When m+1+p is greater than n, if m+1<i<m+1+p, the average gas calorific value of the ipth time period to the average gas calorific value of the mth time period and the predicted gas calorific value of the m+1th time period to the predicted gas calorific value of the i-1th time period in the second training set are input into the ARIMA model trained at the current number of iterations to obtain the predicted gas calorific value of the ith time period; p is the number of autoregressive terms in the ARIMA model at the current number of iterations;

[0023] Perform error analysis on the average value of the predicted gas calorific value in the i-th time period and the average value of the gas calorific value in the i-th time period in the second verification set to obtain an error value;

[0024] If the error value does not meet the iteration stop condition, the iteration number is updated and the next iteration is entered;

[0025] If the error value meets the iteration stop condition, the ARIMA model trained at the current iteration number is determined to be the gas calorific value prediction model.

[0026] Optionally, the error value is a mean square error and a mean percentage error.

[0027] Optionally, the fuel consumption prediction mathematical model corresponding to each cluster in the target cluster set is trained and verified according to the second training set, the second validation set, the cluster corresponding to the first training set, the cluster corresponding to the first sample set, and the average value of the predicted gas calorific value when the heating furnace heats the steel billet in each time period to obtain a fuel consumption prediction model set, specifically including:

[0028] For any cluster corresponding to the first training set, the calorific value burnup set when the heating furnace heats the steel billet in each time period within the first time period set in the second training set and the second verification set is added to the cluster corresponding to the first training set to obtain a third training set of the cluster corresponding to the first training set; the first time period set includes all time periods corresponding to the cluster corresponding to the first training set;

[0029] For any cluster corresponding to the first sample set, the average predicted gas calorific value when the heating furnace heats the steel billet in each time period in the second time period set and the gas consumption when the heating furnace heats the steel billet in each time period in the second time period set in the second training set and the second validation set are added to the cluster corresponding to the first sample set to obtain a third validation set of the cluster corresponding to the first sample set; the second time period set includes all time periods corresponding to the cluster corresponding to the first sample set;

[0030] According to the third training set of all clusters corresponding to the first training set and the third verification set of all clusters corresponding to the first sample set, the fuel consumption prediction mathematical model corresponding to each cluster in the target cluster set is trained and verified to obtain a second verification result. If the second verification result does not meet the second set threshold, the preset number is changed, and the step of clustering the first training set to obtain the clusters corresponding to the first training set is returned. If the second verification result meets the second set threshold, the fuel consumption prediction mathematical model corresponding to each cluster in the target cluster set is determined to be the fuel consumption prediction model set.

[0031] Optionally, the fuel consumption prediction mathematical model corresponding to each cluster in the target cluster set is trained and verified according to a third training set of all clusters corresponding to the first training set and a third verification set of all clusters corresponding to the first sample set to obtain a second verification result, specifically including:

[0032] For any cluster corresponding to the first training set, calculating the average energy loss of the cluster corresponding to the first training set according to the third training set of the cluster corresponding to the first training set;

[0033] Based on the third verification set of the target cluster corresponding to the first sample set and the average energy loss of the cluster corresponding to the first training set, the fuel consumption prediction mathematical model corresponding to the cluster corresponding to the first training set is verified to obtain a verification result; the target cluster is the cluster corresponding to the cluster corresponding to the first training set among all the clusters corresponding to the first sample set.

[0034] In a second aspect, the present application provides a heating furnace gas consumption prediction device, comprising:

[0035] an acquisition module, for dividing a preset time period into a plurality of time periods, and acquiring a first sample set and a second sample set; the first sample set includes a set of production data when a heating furnace heats a steel billet in each time period; the production data set includes an average value of each furnace-entry production data of the heating furnace; the second sample set includes a set of calorific value and fuel consumption when the heating furnace heats the steel billet in each time period; the calorific value and fuel consumption set includes an average value of coal gas calorific value and an average value of coal gas fuel consumption;

[0036] A first division module, used for dividing the first sample set into a first training set and a first validation set according to a set ratio;

[0037] The second division module is used to divide the second sample set into a second training set and a second verification set according to a set ratio; the second training set includes a calorific value combustion set when the steel billet is heated from the 1st time period to the mth time period, and the second verification set includes a calorific value combustion set when the steel billet is heated from the m+1th time period to the nth time period; n represents the number of all time periods included in the preset time period;

[0038] A gas calorific value prediction model determination module is used to train and verify the ARIMA model according to the second training set and the second verification set to obtain a gas calorific value prediction model;

[0039] A first clustering module is used to cluster the first training set to obtain clusters corresponding to the first training set; one cluster corresponds to all steel billets with the same characteristics entering the furnace;

[0040] A second clustering module is used to cluster the first validation set based on the cluster corresponding to the first training set to obtain the cluster corresponding to the first sample set;

[0041] A fuel consumption prediction model set determination module is used to train and verify the fuel consumption prediction mathematical models corresponding to each cluster in the target cluster set according to the second training set, the second verification set, the cluster corresponding to the first training set, the cluster corresponding to the first sample set and the predicted average gas calorific value when the heating furnace heats the steel billet in each time period to obtain a fuel consumption prediction model set, wherein the fuel consumption prediction model set is used to predict the gas consumption of the heating furnace; the target cluster set includes the cluster corresponding to the first training set; the predicted average gas calorific value is obtained by the gas calorific value prediction model.

[0042] In a third aspect, the present application provides a computer device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement any of the above-described methods for predicting heating furnace gas consumption.

[0043] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the above-described methods for predicting heating furnace gas consumption.

[0044] In a fifth aspect, the present application provides a computer program product, including a computer program, which, when executed by a processor, implements any of the above-mentioned methods for predicting heating furnace gas consumption.

[0045] According to the specific embodiments provided in this application, this application discloses the following technical effects:

[0046] The present application provides a method, device, equipment, medium and product for predicting the gas consumption of a heating furnace. The K-means method is used to cluster the heating furnace entry production data, and the steel billets with the same entry characteristics are classified into one category, so as to study the influence of the entry characteristics on the fuel consumption of the heating furnace. The ARIMA model is used to predict the change of the average calorific value of the gas, and on this basis, the influence of the average calorific value of the gas on the gas consumption is analyzed, which can improve the prediction accuracy of the gas consumption of the heating furnace and make the prediction of the gas consumption more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0048] Figure 1 This is an application environment diagram of a method for predicting gas consumption of a heating furnace in one embodiment of the present application;

[0049] Figure 2 This is a flow chart of a method for predicting gas consumption of a heating furnace in one embodiment of the present application;

[0050] Figure 3 This is a specific flow chart of a method for predicting gas consumption of a heating furnace in one embodiment of the present application;

[0051] Figure 4 This is a steel billet data division diagram in a time period of Δt in an embodiment of the present application;

[0052] Figure 5 This is a flow chart of the K-means algorithm in one embodiment of the present application;

[0053] Figure 6 ARIMA model parameter determination flow chart in one embodiment of the present application;

[0054] Figure 7 This is a flowchart of selecting the K value in the K-means algorithm in one embodiment of the present application;

[0055] Figure 8 A schematic diagram of the structure of a computer device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0056] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0057] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0058] The heating furnace gas consumption prediction method provided in the embodiment of the present application can be applied to Figure 1In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be set up separately, integrated on the server 104, or placed on the cloud or other servers. The terminal 102 can send the first sample set and the second sample set to the server 104. After the server 104 receives the first sample set and the second sample set, for the first sample set and the second sample set, the server 104 divides the first sample set into a first training set and a first verification set according to a set ratio. The second sample set is divided into a second training set and a second validation set according to a set ratio; the ARIMA model is trained and validated according to the second training set and the second validation set to obtain a gas calorific value prediction model; the first training set is clustered to obtain a cluster corresponding to the first training set; one cluster corresponds to all steel billets with the same furnace characteristics (production rhythm and hot and cold loading mode); the first validation set is clustered based on the cluster corresponding to the first training set to obtain a cluster corresponding to the first sample set; the fuel consumption prediction mathematical model corresponding to each cluster in the target cluster set is trained and validated according to the second training set, the second validation set, the cluster corresponding to the first training set, the cluster corresponding to the first sample set and the predicted gas calorific value average value when the heating furnace heats the steel billet in each time period to obtain a fuel consumption prediction model set, and the fuel consumption prediction model set is used to predict the gas consumption of the heating furnace; the target cluster set includes the cluster corresponding to the first training set; the predicted gas calorific value average value is obtained by the gas calorific value prediction model. The server 104 can feed back the obtained fuel consumption prediction model set to the terminal 102. In addition, in some embodiments, the heating furnace gas consumption prediction method can also be implemented independently by the server 104 or the terminal 102. For example, the terminal 102 can directly predict the heating furnace gas consumption for the first sample set and the second sample set, or the server 104 can obtain the first sample set and the second sample set from the data storage system and predict the heating furnace gas consumption for the first sample set and the second sample set.

[0059] The terminal 102 may be, but is not limited to, various desktop computers, laptop computers, smart phones, tablet computers, IoT devices, and portable wearable devices. The IoT devices may be smart speakers, smart TVs, smart air conditioners, smart vehicle-mounted devices, etc. The portable wearable devices may be smart watches, smart bracelets, head-mounted devices, etc. The server 104 may be implemented as an independent server or a server cluster consisting of multiple servers, or may be a cloud server.

[0060] In an exemplary embodiment, Figure 2 and Figure 3As shown, a method for predicting the gas consumption of a heating furnace is provided. The method is executed by a computer device, and can be executed by a computer device such as a terminal or a server alone, or by a terminal and a server together. In the embodiment of the present application, the method is applied to Figure 1 The server 104 in the example is used as an example to illustrate the method, which includes the following steps. Among them:

[0061] Step 201: Divide the preset time period into multiple time periods (one time period is Δt minutes), and obtain the first sample set and the second sample set. The first sample set includes the production data set when the heating furnace heats the steel billet in each time period; the production data set includes the average value of each furnace entry production data; the second sample set includes the calorific value consumption set when the heating furnace heats the steel billet in each time period; the calorific value consumption set includes the average value of the calorific value of the gas and the average value of the gas consumption; in the i-th time period Δt i The production data set for heating the billet is based on all the data in the i-th time period Δt i The production data of the billets entering, leaving or being in the heating furnace are averaged. The specific calculation process is as follows: Figure 4 Specifically, the furnace production data includes: the average time in the furnace Average length of billet Average width of billet Average thickness of billet Average specific heat capacity of steel billet Average density of billet Average temperature of billet entering furnace and the target average temperature of the billet out of the furnace

[0062] Step 202: Divide the first sample set into a first training set according to a set ratio (8:2) and the first validation set Among them, Δt n represents the nth time period, It represents the average time that the heating furnace plans to spend in the furnace to heat the steel billet in the nth time period. It represents the average length of the steel billet when the heating furnace heats the steel billet in the nth time period. It indicates the target average temperature of the steel billet when the heating furnace heats the steel billet in the nth time period.

[0063] Step 203: Divide the second sample set into a second training set according to a set ratio (8:2) and the second validation set in, It represents the average calorific value of the gas when the heating furnace heats the steel billet in the nth time period. represents the average value of gas consumption when the heating furnace heats the steel billet in the nth time period. The second training set includes the calorific value consumption set when the steel billet is heated from the 1st time period to the mth time period, and the second verification set includes the calorific value consumption set when the steel billet is heated from the m+1th time period to the nth time period; n represents the number of all time periods included in the preset time period.

[0064] Step 204: Train and verify the ARIMA model based on the second training set and the second verification set to obtain a gas calorific value prediction model.

[0065] Step 205: clustering the first training set to obtain clusters corresponding to the first training set; one cluster corresponds to all steel billets with the same furnace characteristics.

[0066] Step 206: clustering the first validation set based on the cluster corresponding to the first training set to obtain the cluster corresponding to the first sample set. The production data set in the first validation set is divided into the cluster corresponding to the first training set.

[0067] Step 207: According to the second training set, the second validation set, the cluster corresponding to the first training set, the cluster corresponding to the first sample set and the predicted average gas calorific value when the heating furnace heats the steel billet in each time period, the fuel consumption prediction mathematical model corresponding to each cluster in the target cluster set is trained and verified to obtain a fuel consumption prediction model set, wherein the fuel consumption prediction model set is used to predict the gas consumption of the heating furnace; the target cluster set includes the cluster corresponding to the first training set; the predicted average gas calorific value is obtained by the gas calorific value prediction model.

[0068] In practical applications, the ARIMA model is trained and verified according to the second training set and the second verification set to obtain a gas calorific value prediction model, which specifically includes:

[0069] At the current number of iterations, the number of autoregressive items and the number of moving average items in the ARIMA model at the current number of iterations are calculated based on the ARIMA model at the previous number of iterations, and the ARIMA model at the current number of iterations is constructed; the number of autoregressive items and the number of moving average items in the ARIMA model at the initial number of iterations are obtained based on the average gas calorific value in the second training set.

[0070] The ARIMA model at the current iteration number is trained according to the average gas calorific value in the second training set to obtain the trained ARIMA model at the current iteration number.

[0071] If i=m+1, the average gas calorific value of the ipth time period to the mth time period in the second training set is input into the ARIMA model trained at the current number of iterations to obtain the predicted average gas calorific value of the ith time period; p is the number of autoregressive terms in the ARIMA model at the current number of iterations.

[0072] When m+1+p is less than or equal to n, if m+1<i<m+1+p, then the average gas calorific value of the ipth time period to the average gas calorific value of the mth time period and the predicted average gas calorific value of the m+1th time period to the predicted average gas calorific value of the i-1th time period in the second training set are input into the ARIMA model trained under the current number of iterations to obtain the predicted average gas calorific value of the i-th time period; p is the number of autoregressive terms in the ARIMA model under the current number of iterations.

[0073] If m+1+p≤i≤n, the predicted average gas calorific value of the ipth time period to the predicted average gas calorific value of the i-1th time period are input into the ARIMA model trained under the current number of iterations to obtain the predicted average gas calorific value of the i-th time period; p is the number of autoregressive terms in the ARIMA model under the current number of iterations.

[0074] When m+1+p is greater than n, if m+1<i<m+1+p, the average gas calorific value of the ipth time period in the second training set to the average gas calorific value of the mth time period and the predicted average gas calorific value of the m+1th time period to the predicted average gas calorific value of the i-1th time period are input into the ARIMA model trained under the current number of iterations to obtain the predicted average gas calorific value of the i-th time period; p is the number of autoregressive terms in the ARIMA model under the current number of iterations.

[0075] The error value is obtained by performing error analysis on the predicted average gas calorific value in the i-th time period and the average gas calorific value in the i-th time period in the second verification set; the error value is the mean square error and the average percentage error.

[0076] If the error value does not meet the iteration stop condition (the error is within the set range), the number of iterations is updated and the next iteration is entered.

[0077] If the error value meets the iteration stop condition, the ARIMA model trained at the current iteration number is determined to be the gas calorific value prediction model.

[0078] In practical applications, constructing the ARIMA model at the current number of iterations under the initial number of iterations specifically includes:

[0079] Step A: Constructing a mathematical model of gas calorific value prediction model based on ARIMA model:

[0080]

[0081] in, They represent the average calorific value of the gas when the heating furnace heats the steel billet in the i-th, i-1-th, i-2-th, and ip-th time periods respectively; c is a constant term, arrive represents the autoregressive coefficients, θ1 to θ q represents the moving average coefficient; ∈ i ,∈ i-1 ,∈ i-2 ,∈ i-q They respectively represent the errors of the average calorific value of coal gas when the heating furnace heats the steel billet in the i-th, i-1-th, i-2-th, and ip-th time periods.

[0082] Step B: Determine the values ​​of the number of autoregressive terms p, the number of moving average terms q and the number of differentials d in the mathematical model based on the gas calorific value prediction model according to the data in the time period column and the gas calorific value average value column in the second training set (X2), and train the mathematical model of the gas calorific value prediction model with the determined p and q values ​​according to the second training set, adjust the values ​​of the autoregressive coefficient and the moving average coefficient in the model, and obtain the ARIMA model under the current number of iterations.

[0083] The specific process of determining p, q, and d can be found in Figure 5 Specifically, it includes: firstly, the ADF test should be performed on the average gas calorific value data in the second training set to check whether the data is stable. When the data is not stable, the data should be differentially processed, and the differential number is the value of d. When the data is always stable, d is 0. After determining the value of d, the values ​​of p and q are determined according to the following principles:

[0084]

[0085] In practical applications, the number of autoregressive terms and the number of moving average terms in the ARIMA model at the current iteration number are calculated based on the ARIMA model at the previous iteration number. Specifically, the number of autoregressive terms and the number of moving average terms in the ARIMA model at the previous iteration number are selected around the number of autoregressive terms and the number of moving average terms in the ARIMA model at the current iteration number to determine the number of autoregressive terms and the number of moving average terms in the ARIMA model at the previous iteration number.

[0086] In practical applications, K-means clustering is performed on the first training set (X1), and the data therein is divided according to K cluster centers, so that the steel billets with similar furnace characteristics are grouped into one cluster. In practical applications, the K value in the K-means clustering algorithm is determined to be a preset number. The specific steps are:

[0087] The steps to determine the K value in the K-means algorithm are as follows Figure 6 As shown, we should first calculate the sum of squared sample errors (SSE(K)) under different K values, draw a line graph of SSE(K) and K values, and select the K value corresponding to the point with the largest slope change (i.e., the elbow) on the line graph as the K value of the K-means method. The formula for the sum of squared sample errors (SSE(K)) is:

[0088]

[0089] Where K is the number of clusters; C j is the cluster corresponding to the jth initialized cluster center; C ji It is C j The production data set of the steel billet heating in the i-th time period in the cluster; C j0 is the production data set corresponding to the jth cluster center; ||C ji -C j0 || 2 Represents C j The Euclidean distance between the production data set when the steel billet is heated in the i-th time period in the cluster and the production data set corresponding to the j-th cluster center.

[0090] For the specific process of K-means algorithm, please refer to Figure 7 , its mathematical model can be expressed as:

[0091]

[0092] Among them C ji Represents C j The production data set of the steel billet heating in the i-th time period in the cluster; C jis Represents C j The sth furnace production data in the production data set when the steel billet is heated in the i-th time period in the cluster, such as the average time in the furnace Average length of billet Width thick Average specific heat capacity of steel billet Average density of billet Average temperature of billet entering furnace Or average target billet outlet temperature C j0 represents the production data set corresponding to the jth cluster center; C j0t represents the t-th furnace production data under the j-th cluster center; dis(C ji ,C j0 ) means C jThe Euclidean distance between the production data set when the steel billet is heated in the i-th time period in the cluster and the production data set corresponding to the j-th cluster center.

[0093] In practical applications, the fuel consumption prediction mathematical model corresponding to each cluster in the target cluster set is trained and verified according to the second training set, the second validation set, the cluster corresponding to the first training set, the cluster corresponding to the first sample set, and the average predicted gas calorific value when the heating furnace heats the steel billet in each time period to obtain a fuel consumption prediction model set, specifically including:

[0094] For any cluster corresponding to the first training set, add the calorific value consumption set of the heating furnace in the second training set and the second validation set when heating the steel billet in each time period in the first time period set to the cluster corresponding to the first training set, and obtain the third training set of the cluster corresponding to the first training set. The first time period set includes all time periods corresponding to the cluster corresponding to the first training set; the number of data groups in different clusters is different, that is, v is different in different clusters.

[0095] For any cluster corresponding to the first sample set, the average predicted gas calorific value when the heating furnace heats the steel billet in each time period in the second time period set and the gas consumption when the heating furnace heats the steel billet in each time period in the second training set and the second validation set are added to the cluster corresponding to the first sample set to obtain a third validation set of the cluster corresponding to the first sample set. The second time period set includes all time periods corresponding to the cluster corresponding to the first sample set, and the number of data groups in different clusters is different, that is, w is different in different clusters.

[0096] According to the third training set of all clusters corresponding to the first training set and the third verification set of all clusters corresponding to the first sample set, the fuel consumption prediction mathematical model corresponding to each cluster in the target cluster set is trained and verified to obtain a second verification result. If the second verification result does not meet the second set threshold, the preset number is changed, and the step of clustering the first training set to obtain the clusters corresponding to the first training set is returned. If the second verification result meets the second set threshold, the fuel consumption prediction mathematical model corresponding to each cluster in the target cluster set is determined to be the fuel consumption prediction model set.

[0097] In practical applications, the fuel consumption prediction mathematical model corresponding to each cluster in the target cluster set is trained and verified according to the third training set of all clusters corresponding to the first training set and the third verification set of all clusters corresponding to the first sample set to obtain the second verification result, which specifically includes:

[0098] For any cluster corresponding to the first training set, the average energy loss of the cluster corresponding to the first training set is calculated according to the third training set of the cluster corresponding to the first training set.

[0099] Based on the third verification set of the target cluster corresponding to the first sample set and the average energy loss of the cluster corresponding to the first training set, the fuel consumption prediction mathematical model corresponding to the cluster corresponding to the first training set is verified to obtain a verification result; the target cluster is the cluster corresponding to the cluster corresponding to the first training set among all the clusters corresponding to the first sample set.

[0100] In practical applications, calculating the average energy loss of the cluster corresponding to the first training set according to the third training set of the cluster corresponding to the first training set specifically includes:

[0101] Substitute the data in the third training set (X3) of the jth cluster corresponding to the first training set into the following expression to find the energy loss Q in the i-th time period in the jth cluster: (loss)ji And the average energy loss Q under the jth cluster (loss)j* , where the calculation formula is as follows:

[0102]

[0103] Among them, Q (loss)ji is the energy loss in the jth cluster at the i-th time period, J; is the average gas consumption when heating the steel billet in the i-th time period in the j-th cluster, m 3 ; is the average specific heat capacity of the steel billet when it is heated in the i-th time period in the j-th cluster, J / kg·℃; is the average length of the billet when it is heated in the i-th time period in the j-th cluster, m; is the average width of the steel billet when it is heated in the i-th time period in the j-th cluster, m; is the average thickness of the steel billet when it is heated in the i-th time period in the j-th cluster, m; T (out)ji is the average target temperature of the billet when the billet is heated in the i-th time period in the j-th cluster, °C; T (in)ji is the average temperature of the billet entering the furnace when the billet is heated in the i-th time period in the j-th cluster, °C; is the average density of the steel billet when it is heated in the i-th time period in the j-th cluster, kg / m 3 ; is the average calorific value of the gas used to heat the steel billet in the i-th time period in the j-th cluster, J / m 3 ;Q (loss)j*is the average energy loss under the jth cluster, J; v is the number of data groups in the fuel consumption prediction training set (X3), that is, the number of production data sets.

[0104] In practical applications, based on the third verification set of the target cluster corresponding to the first sample set and the average energy loss of the cluster corresponding to the first training set, the fuel consumption prediction mathematical model corresponding to the cluster corresponding to the first training set is verified to obtain verification results, specifically including:

[0105] Since the K-means method classifies billets with similar initial conditions into one category, all Q (loss)ji Theoretically it is approximate.

[0106] The mathematical model for fuel consumption prediction is The data in the third validation set (Y3) of the jth cluster corresponding to the first sample set is input into the above model to predict the fuel consumption to verify whether the fuel consumption prediction is accurate. If the error is too large, the corrected K value is returned and the subsequent steps are repeated. If the prediction is accurate, the fuel consumption prediction modeling and verification are completed, where: is the predicted gas consumption when heating the steel billet in the i-th time period in the j-th cluster, i.e., the gas consumption, m 3 .

[0107] In practical applications, when a fuel consumption prediction model set is used to predict the gas consumption of a heating furnace, when the acquired production data set is input into the fuel consumption prediction model set, the fuel consumption prediction model set will automatically cluster it, match the average energy loss according to the clusters, and match the corresponding calorific value according to time, and finally complete the fuel consumption prediction according to the fuel consumption prediction model.

[0108] Heating furnaces are large gas consumers. Accurately predicting the gas consumption of heating furnaces is important for optimizing the scheduling and energy conservation of steel enterprises' gas systems. In fact, the first study on energy conservation and consumption reduction of heating furnaces is based on the parameter dome temperature T. 热 , hot air temperature T 风 , exhaust temperature T 烟, determine the furnace condition of the hot blast furnace after the furnace change, and calculate the residual heat of the hot blast furnace after the heat exchange according to the determined furnace condition. According to the calculated residual heat of the hot blast furnace, the amount of gas for burning the furnace is predicted based on the furnace condition; but gas consumption is not only affected by the furnace condition, so more factors need to be considered to affect gas consumption. The second method uses the production plan of the steel rolling system, equipment maintenance plan, production data of the heating furnace in recent years, gas consumption indicators and heat balance test data as basic parameters, and the daily heating steel type plan, heating furnace hourly production plan, average temperature of steel billets entering and leaving the furnace, and average preheating temperature of air and gas as dynamic prediction correction parameters. Through heat balance test, a mathematical model combining statistical analysis and mechanism is established to make medium-, long- and short-term predictions of gas consumption of steel rolling heating furnaces; but there are too many factors affecting the production of heating furnaces, and the use of dynamic prediction correction parameters and heat balance test cannot meet the prediction accuracy requirements. This application uses the K-means method to divide the furnace entry plan data into K categories, and builds a fuel consumption prediction model based on each category. Through the time arrangement of the production plan, and based on the predicted value of the change in the average gas calorific value at the corresponding time obtained by the gas calorific value prediction model, on this basis, the billet data in the production plan is imported into the fuel consumption prediction model, so as to calculate the amount of gas required for the batch of billets to be sent to the heating furnace. This application has the following technical effects:

[0109] (1) This application effectively solves the problem of the inability to accurately predict the gas consumption of heating furnaces in the process of balancing and optimizing the scheduling of the gas system of large-scale steel joint ventures by dynamically and accurately predicting the gas consumption of heating furnaces. It plays an important role in improving the gas utilization efficiency of enterprises and saving energy and reducing emissions.

[0110] (2) This application proposes that the average calorific value of gas will also have an impact on the final gas consumption, and uses the ARIMA model to predict the change of the average calorific value of gas. On this basis, the impact of the average calorific value of gas on gas consumption is analyzed, making the prediction of gas consumption more accurate.

[0111] (3) This application takes into account the impact of the production rhythm and hot and cold charging methods of steel billets on the fuel consumption of the heating furnace, and clusters the production data of the heating furnace by the K-means method, classifying steel billets with the same production rhythm and hot and cold charging methods into one category, thereby studying the impact of the above factors on the fuel consumption of the heating furnace, making the prediction more accurate.

[0112] (4) The gas demand of the heating furnace can be predicted according to the production plan, thereby avoiding the waste of gas on the heating furnace by the steel enterprise. At the same time, it can also provide a basis for the gas scheduling of the steel enterprise, thereby achieving energy conservation and emission reduction and promoting it to industrial applications.

[0113] Based on the same inventive concept, the embodiment of the present application also provides a heating furnace gas consumption prediction device for implementing the heating furnace gas consumption prediction method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more heating furnace gas consumption prediction device embodiments provided below can refer to the limitations of the heating furnace gas consumption prediction method above, and will not be repeated here.

[0114] In an exemplary embodiment, a heating furnace gas consumption prediction device is provided, comprising:

[0115] An acquisition module is used to divide a preset time period into multiple time periods, and acquire a first sample set and a second sample set; the first sample set includes a set of production data when a heating furnace heats steel billets in each time period; the production data set includes an average value of each production data entering the heating furnace; the second sample set includes a set of calorific value and fuel consumption when the heating furnace heats steel billets in each time period; the calorific value and fuel consumption set includes an average gas calorific value and an average gas fuel consumption.

[0116] The first division module is used to divide the first sample set into a first training set and a first verification set according to a set ratio.

[0117] The second division module is used to divide the second sample set into a second training set and a second verification set according to a set ratio.

[0118] The gas calorific value prediction model determination module is used to train and verify the ARIMA model according to the second training set and the second verification set to obtain the gas calorific value prediction model.

[0119] The first clustering module is used to cluster the first training set to obtain clusters corresponding to the first training set; one cluster corresponds to all steel billets with the same furnace characteristics.

[0120] The second clustering module is used to cluster the first validation set based on the cluster corresponding to the first training set to obtain the cluster corresponding to the first sample set.

[0121] A fuel consumption prediction model set determination module is used to train and verify the fuel consumption prediction mathematical models corresponding to each cluster in the target cluster set according to the second training set, the second verification set, the cluster corresponding to the first training set, the cluster corresponding to the first sample set and the predicted average gas calorific value when the heating furnace heats the steel billet in each time period to obtain a fuel consumption prediction model set, wherein the fuel consumption prediction model set is used to predict the gas consumption of the heating furnace; the target cluster set includes the cluster corresponding to the first training set; the predicted average gas calorific value is obtained by the gas calorific value prediction model.

[0122] In an exemplary embodiment, a computer device is provided. The computer device may be a server or a terminal. The internal structure diagram thereof may be as follows: Figure 8 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store heating furnace gas consumption prediction data. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for predicting heating furnace gas consumption is implemented.

[0123] Those skilled in the art will understand that Figure 8 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components. In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the above-mentioned method embodiments when executing the computer program.

[0124] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by a processor, the above-mentioned method embodiments are implemented.

[0125] In an exemplary embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the above-mentioned method embodiments are implemented.

[0126] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0127] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0128] The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. The non-relational database may include a distributed database based on blockchain, etc., but is not limited thereto. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but is not limited thereto.

[0129] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0130] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A method for predicting heating furnace gas consumption, characterized in that: The heating furnace gas consumption prediction method comprises: Divide the preset time period into multiple time periods, and obtain a first sample set and a second sample set; the first sample set includes a production data set when the heating furnace heats the steel billet in each time period; the production data set includes an average value of each furnace-entry production data of the heating furnace; the second sample set includes a calorific value fuel consumption set when the heating furnace heats the steel billet in each time period; the calorific value fuel consumption set includes an average gas calorific value and an average gas fuel consumption; Divide the first sample set into a first training set and a first validation set according to a set ratio; The second sample set is divided into a second training set and a second validation set according to a set ratio; the second training set includes a set of calorific value combustion when the steel billet is heated from the 1st time period to the mth time period, and the second validation set includes a set of calorific value combustion when the steel billet is heated from the m+1th time period to the nth time period; n represents the number of all time periods included in the preset time period; The ARIMA model is trained and verified according to the second training set and the second verification set to obtain a gas calorific value prediction model; The first training set is clustered to obtain a cluster corresponding to the first training set; one cluster corresponds to all steel billets with the same characteristics entering the furnace; Taking the cluster corresponding to the first training set as a benchmark, clustering the first validation set is performed to obtain the cluster corresponding to the first sample set; According to the second training set, the second validation set, the cluster corresponding to the first training set, the cluster corresponding to the first sample set and the predicted average gas calorific value when the heating furnace heats the steel billet in each time period, the fuel consumption prediction mathematical model corresponding to each cluster in the target cluster set is trained and verified to obtain a fuel consumption prediction model set, wherein the fuel consumption prediction model set is used to predict the gas consumption of the heating furnace; the target cluster set includes the cluster corresponding to the first training set; the predicted average gas calorific value is obtained by the gas calorific value prediction model.

2. The method for predicting heating furnace gas consumption according to claim 1, characterized in that: The furnace production data specifically includes: The plan includes the average furnace time, the average length of the billet, the average width of the billet, the average thickness of the billet, the average specific heat capacity of the billet, the average density of the billet, the average temperature of the billet entering the furnace and the target average temperature of the billet out of the furnace.

3. The method for predicting heating furnace gas consumption according to claim 1, characterized in that: The gas calorific value prediction model is obtained by training and verifying the ARIMA model according to the second training set and the second verification set, specifically including: At the current iteration number, the number of autoregressive items and the number of moving average items in the ARIMA model at the current iteration number are calculated based on the ARIMA model at the previous iteration number, and the ARIMA model at the current iteration number is constructed; the number of autoregressive items and the number of moving average items in the ARIMA model at the initial iteration number are obtained based on the average gas calorific value in the second training set; The ARIMA model at the current iteration number is trained according to the average gas calorific value in the second training set to obtain the trained ARIMA model at the current iteration number; If i=m+1, the average gas calorific value of the ipth time period to the mth time period in the second training set is input into the ARIMA model trained at the current number of iterations to obtain the predicted average gas calorific value of the ith time period; p is the number of autoregressive terms in the ARIMA model at the current number of iterations; When m+1+p is less than or equal to n, if m+1<i<m+1+p, the average gas calorific value of the ipth time period to the average gas calorific value of the mth time period and the predicted gas calorific value of the m+1th time period to the predicted gas calorific value of the i-1th time period in the second training set are input into the ARIMA model trained at the current number of iterations to obtain the predicted gas calorific value of the ith time period; p is the number of autoregressive terms in the ARIMA model at the current number of iterations; If m+1+p≤i≤n, the average predicted gas calorific value in the ipth time period to the average predicted gas calorific value in the i-1th time period are input into the ARIMA model trained at the current number of iterations to obtain the average predicted gas calorific value in the i-th time period; p is the number of autoregressive terms in the ARIMA model at the current number of iterations; When m+1+p is greater than n, if m+1<i<m+1+p, the average gas calorific value of the ipth time period to the average gas calorific value of the mth time period and the predicted gas calorific value of the m+1th time period to the predicted gas calorific value of the i-1th time period in the second training set are input into the ARIMA model trained at the current number of iterations to obtain the predicted gas calorific value of the ith time period; p is the number of autoregressive terms in the ARIMA model at the current number of iterations; Perform error analysis on the average value of the predicted gas calorific value in the i-th time period and the average value of the gas calorific value in the i-th time period in the second verification set to obtain an error value; If the error value does not meet the iteration stop condition, the iteration number is updated and the next iteration is entered; If the error value meets the iteration stop condition, the ARIMA model trained at the current iteration number is determined to be the gas calorific value prediction model.

4. The method for predicting heating furnace gas consumption according to claim 3, characterized in that: The error values ​​are mean square error and mean percentage error.

5. The method for predicting heating furnace gas consumption according to claim 1, characterized in that: According to the second training set, the second validation set, the clusters corresponding to the first training set, the clusters corresponding to the first sample set and the average predicted gas calorific value when the heating furnace heats the steel billet in each time period, the fuel consumption prediction mathematical model corresponding to each cluster in the target cluster set is trained and validated to obtain a fuel consumption prediction model set, specifically including: For any cluster corresponding to the first training set, the calorific value burnup set when the heating furnace heats the steel billet in each time period within the first time period set in the second training set and the second verification set is added to the cluster corresponding to the first training set to obtain a third training set of the cluster corresponding to the first training set; the first time period set includes all time periods corresponding to the cluster corresponding to the first training set; For any cluster corresponding to the first sample set, the average predicted gas calorific value when the heating furnace heats the steel billet in each time period in the second time period set and the gas consumption when the heating furnace heats the steel billet in each time period in the second time period set in the second training set and the second validation set are added to the cluster corresponding to the first sample set to obtain a third validation set of the cluster corresponding to the first sample set; the second time period set includes all time periods corresponding to the cluster corresponding to the first sample set; According to the third training set of all clusters corresponding to the first training set and the third verification set of all clusters corresponding to the first sample set, the fuel consumption prediction mathematical model corresponding to each cluster in the target cluster set is trained and verified to obtain a second verification result. If the second verification result does not meet the second set threshold, the preset number is changed, and the step of clustering the first training set to obtain the clusters corresponding to the first training set is returned. If the second verification result meets the second set threshold, the fuel consumption prediction mathematical model corresponding to each cluster in the target cluster set is determined to be the fuel consumption prediction model set.

6. The method for predicting heating furnace gas consumption according to claim 5, characterized in that: The fuel consumption prediction mathematical model corresponding to each cluster in the target cluster set is trained and verified according to the third training set of all clusters corresponding to the first training set and the third verification set of all clusters corresponding to the first sample set to obtain a second verification result, specifically including: For any cluster corresponding to the first training set, calculating the average energy loss of the cluster corresponding to the first training set according to the third training set of the cluster corresponding to the first training set; Based on the third verification set of the target cluster corresponding to the first sample set and the average energy loss of the cluster corresponding to the first training set, the fuel consumption prediction mathematical model corresponding to the cluster corresponding to the first training set is verified to obtain a verification result; the target cluster is the cluster corresponding to the cluster corresponding to the first training set among all the clusters corresponding to the first sample set.

7. A heating furnace gas consumption prediction device, characterized in that: The heating furnace gas consumption prediction device comprises: an acquisition module, for dividing a preset time period into a plurality of time periods, and acquiring a first sample set and a second sample set; the first sample set includes a set of production data when a heating furnace heats a steel billet in each time period; the production data set includes an average value of each furnace-entry production data of the heating furnace; the second sample set includes a set of calorific value and fuel consumption when the heating furnace heats the steel billet in each time period; the calorific value and fuel consumption set includes an average value of coal gas calorific value and an average value of coal gas fuel consumption; A first division module, used for dividing the first sample set into a first training set and a first validation set according to a set ratio; The second division module is used to divide the second sample set into a second training set and a second verification set according to a set ratio; the second training set includes a calorific value combustion set when the steel billet is heated from the 1st time period to the mth time period, and the second verification set includes a calorific value combustion set when the steel billet is heated from the m+1th time period to the nth time period; n represents the number of all time periods included in the preset time period; A gas calorific value prediction model determination module is used to train and verify the ARIMA model according to the second training set and the second verification set to obtain a gas calorific value prediction model; A first clustering module is used to cluster the first training set to obtain clusters corresponding to the first training set; one cluster corresponds to all steel billets with the same characteristics entering the furnace; A second clustering module is used to cluster the first validation set based on the cluster corresponding to the first training set to obtain the cluster corresponding to the first sample set; A fuel consumption prediction model set determination module is used to train and verify the fuel consumption prediction mathematical models corresponding to each cluster in the target cluster set according to the second training set, the second verification set, the cluster corresponding to the first training set, the cluster corresponding to the first sample set and the predicted average gas calorific value when the heating furnace heats the steel billet in each time period to obtain a fuel consumption prediction model set, wherein the fuel consumption prediction model set is used to predict the gas consumption of the heating furnace; the target cluster set includes the cluster corresponding to the first training set; the predicted average gas calorific value is obtained by the gas calorific value prediction model.

8. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method for predicting heating furnace gas consumption according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for predicting heating furnace gas consumption according to any one of claims 1 to 6 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method for predicting heating furnace gas consumption according to any one of claims 1 to 6 is implemented.

Images