Abnormal furnace condition early warning method and device, electronic equipment and storage medium

Abstract

The invention relates to the technical field of metallurgical industry, and discloses an abnormal furnace condition early warning method and device, electronic equipment and a storage medium, and the method comprises the steps: obtaining a training data set and a to-be-detected data set, predicting the to-be-detected data set through a preset prediction model, and determining a target parameter of data abnormality, the preset prediction model is obtained through training according to a training data set, if the target parameters in a preset continuous time period are all represented as data anomaly, the fluctuation coefficient of each target parameter is calculated, whether the furnace condition of the to-be-detected blast furnace is normal or not is determined according to the fluctuation coefficient, and if the fluctuation coefficient is larger than or equal to a preset threshold value, it is determined that the furnace condition of the to-be-detected blast furnace is abnormal. Early warning is carried out; if the fluctuation coefficient is smaller than a preset threshold, determining that the furnace condition of the to-be-detected blast furnace is normal; through the method, a more practical and more accurate abnormal furnace condition early warning method is provided.

Description

Technical Field

[0001] The present application relates to the technical field of metallurgical industry, and in particular to an abnormal furnace condition early warning method, device, electronic equipment and storage medium. Background Art

[0002] Blast furnace ironmaking is a core link in the steel manufacturing process, and its importance is self-evident. However, this process is also the link with the highest energy consumption and production cost, accounting for 60% to 70% of the comprehensive energy consumption of steel, more than 90% of CO2 emissions, and 60% to 70% of the total cost of steel manufacturing. The complexity of the blast furnace lies in the fact that it is a distributed continuous production system. Any failure may trigger a series of chain reactions, such as increased fuel ratio, extended wind down and maintenance time, and substandard molten iron quality. These problems will not only lead to huge resource and equipment losses, but also shorten the service life of the blast furnace, and may even cause safety accidents and threaten personnel lives.

[0003] As a large countercurrent reactor, the blast furnace has an extremely complex smelting process. On the one hand, the smelting conditions of the blast furnace are disturbed by a variety of external factors, such as changes in the physical properties and chemical composition of raw materials and ores, climate fluctuations, etc. On the other hand, the operation of the blast furnace is also affected by objective factors such as the deviation of the foreman's operating knowledge. In addition, the blast furnace may also be disturbed by other urgent and uncertain factors, which make the occurrence of abnormal conditions in the blast furnace sudden. At the same time, the blast furnace smelting process is accompanied by complex physical and chemical reactions and drastic material and phase transformation, which makes the operating parameters and state parameters of the blast furnace have strong coupling, dynamic time lag and nonlinearity. These problems not only increase the difficulty of blast furnace smelting, but also make it difficult to capture the signs of abnormal conditions in the blast furnace, which brings great challenges to the safe production and stable operation of the blast furnace. Obviously, there is an urgent need for a new abnormal furnace condition early warning method to solve at least one of the above problems.

[0004] It should be noted that the above content only provides background technical information related to the present application and does not necessarily constitute prior art. Summary of the invention

[0005] In view of the shortcomings of the prior art described above, the present application provides a method, device, electronic equipment and storage medium for warning of abnormal furnace conditions, so as to reduce the false alarm rate of abnormal blast furnace conditions and improve the accuracy of warning of abnormal blast furnace conditions.

[0006] Other features and advantages of the present application will become apparent from the following detailed description, or may be learned in part by the practice of the present application.

[0007] According to one aspect of an embodiment of the present application, a method for warning abnormal furnace conditions is provided, comprising: obtaining a training data set and a data set to be tested, the training data set comprising furnace condition characterization parameters corresponding to different abnormal furnace condition types of a blast furnace, the furnace condition characterization parameters comprising new features that reflect the dynamic change characteristics of operating parameters before the abnormal situation of the blast furnace occurs; the data set to be tested reflects the operating parameters of the blast furnace to be tested in a characterization manner using the new features; using a preset prediction model to predict the data set to be tested, and determining target parameters of data anomalies, wherein the preset prediction model is constructed based on the training data set; if the target parameters are all characterized as data anomalies within a preset continuous time period, calculating the fluctuation coefficient of each target parameter, and determining whether the furnace condition of the blast furnace to be tested is normal based on the fluctuation coefficient; if the fluctuation coefficient is greater than or equal to a preset threshold, determining that the furnace condition of the blast furnace to be tested is abnormal, and issuing a warning; if the fluctuation coefficient is less than the preset threshold, determining that the furnace condition of the blast furnace to be tested is normal.

[0008] In one embodiment of the present application, based on the aforementioned scheme, new features of each of the operating parameters are obtained in the following manner: the operating parameters of a preset sampling time period before the abnormal situation of the blast furnace occurs are determined as effective parameters; a target characteristic parameter is determined from the operating parameters, and based on the effective parameters, the minimum value corresponding to the target characteristic parameter, the maximum value corresponding to the target characteristic parameter, the maximum rising amplitude corresponding to the target characteristic parameter, the maximum falling amplitude corresponding to the target characteristic parameter, the average rising speed corresponding to the target characteristic parameter, and the average falling speed corresponding to the target characteristic parameter are determined to construct new features of the target characteristic parameter and obtain new features of each of the operating parameters.

[0009] In one embodiment of the present application, based on the aforementioned scheme, a new feature of the target feature parameter is constructed in the following manner: a minimum value corresponding to the target feature parameter, a maximum value corresponding to the target feature parameter, and a current value corresponding to the target feature parameter at the target moment are determined from the effective parameters; a maximum rise amplitude corresponding to the target feature parameter is calculated based on the minimum value and the current value, a maximum fall amplitude corresponding to the target feature parameter is calculated based on the maximum value and the current value, an average rise speed corresponding to the target feature parameter is calculated according to the maximum rise amplitude and a first time period, and an average fall speed corresponding to the target feature parameter is calculated according to the maximum fall amplitude and a second time period, wherein the first time period is determined by the moment corresponding to the minimum value and the target moment, and the second time period is determined by the moment corresponding to the maximum value and the target moment; a new feature of the target feature parameter at the target moment is constructed according to the minimum value corresponding to the target feature parameter, the maximum value corresponding to the target feature parameter, the maximum rise amplitude corresponding to the target feature parameter, the maximum fall amplitude corresponding to the target feature parameter, the average rise speed corresponding to the target feature parameter, and the average fall speed corresponding to the target feature parameter.

[0010] In one embodiment of the present application, based on the aforementioned scheme, the preset prediction model is constructed according to the training data set, including: performing data dimensionality reduction on the new features in the training data set to obtain dimensionality reduction training features; calculating the Hotelling statistic, the prediction error sum of squares statistic, the Hotelling control limit, and the prediction error sum of squares control limit of each operating parameter in the training data set according to the dimensionality reduction training features, and constructing the preset prediction model according to the Hotelling statistic, the prediction error sum of squares statistic, the Hotelling control limit, and the prediction error sum of squares control limit.

[0011] In one embodiment of the present application, based on the above-mentioned scheme, the data set to be tested is predicted using a preset prediction model to determine the target parameter of data anomaly, including: predicting the data set to be tested using the preset prediction model to obtain the Hotelling statistic and the prediction error sum of squares statistic of each operating parameter in the data set to be tested at the target time; comparing the Hotelling statistic with the corresponding Hotelling control limit, and the prediction error sum of squares statistic with the corresponding prediction error sum of squares control limit, respectively; if the Hotelling statistic is greater than the corresponding Hotelling control limit, and / or the prediction error sum of squares statistic is greater than the corresponding prediction error sum of squares control limit, then it is determined that the operating parameter at the target time is characterized as data anomaly; if the Hotelling statistic is less than or equal to the corresponding Hotelling control limit, and the prediction error sum of squares statistic is less than or equal to the corresponding prediction error sum of squares control limit, then it is determined that the operating parameter at the target time is characterized as data normal; and the operating parameter characterized as data anomaly is determined as the target parameter.

[0012] In one embodiment of the present application, based on the aforementioned scheme, if the target parameters within a preset continuous time period are all characterized as data anomalies, the fluctuation coefficient of each target parameter is calculated, including: if the target parameters within a preset continuous time period are all characterized as data anomalies, the parameter mean of each target parameter within the preset time period is calculated, wherein the preset continuous time period is composed of multiple consecutive target moments; based on the parameter mean and the parameter values ​​of each target parameter at the target moment, the variation range of each target parameter within the preset time period is calculated; based on the data standard deviation of each target parameter within the preset time period and the variation range, the fluctuation coefficient of each target parameter within the preset time period is calculated.

[0013] In one embodiment of the present application, based on the aforementioned scheme, if the fluctuation coefficient is greater than or equal to a preset threshold, after determining that the condition of the blast furnace to be tested is abnormal, the method further includes: counting the number of target parameters whose fluctuation coefficient is greater than or equal to the preset threshold; if the number of target parameters is greater than a first quantity threshold, a first-level warning is issued; if the number of target parameters is greater than a second quantity threshold, a second-level warning is issued; if the number of target parameters is greater than a third quantity threshold, a third-level warning is issued, wherein the first quantity threshold is greater than the second quantity threshold, and the second quantity threshold is greater than the third quantity threshold.

[0014] According to one aspect of an embodiment of the present application, an abnormal furnace condition early warning device is provided, comprising: an acquisition module, for acquiring a training data set and a data set to be tested, wherein the training data set comprises furnace condition characterization parameters corresponding to different abnormal furnace condition types of a blast furnace, and the furnace condition characterization parameters comprise new features reflecting the dynamic change characteristics of operating parameters before the abnormal situation of the blast furnace occurs; the data set to be tested reflects the operating parameters of the blast furnace to be tested in the manner of characterization of the new features; a prediction module, for predicting the data set to be tested by using a preset prediction model to determine target parameters of data abnormality, wherein the preset prediction model is obtained by training the training data set; a calculation module, for calculating the fluctuation coefficient of each target parameter if the target parameters are all characterized as data abnormality within a preset continuous time period, and determining whether the furnace condition of the blast furnace to be tested is normal according to the fluctuation coefficient; a determination module, for determining that the furnace condition of the blast furnace to be tested is abnormal and issuing an early warning if the fluctuation coefficient is greater than or equal to a preset threshold; and for determining that the furnace condition of the blast furnace to be tested is normal if the fluctuation coefficient is less than the preset threshold.

[0015] According to one aspect of an embodiment of the present application, an electronic device is provided, comprising: one or more processors; a storage device for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the electronic device implements the abnormal furnace condition warning method as described in any one of the above embodiments.

[0016] The present application also provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor of a computer, the computer is enabled to execute the abnormal furnace condition warning method as described in any one of the above embodiments.

[0017] The beneficial effects of the present application are as follows: the present application obtains a training data set and a data set to be tested, wherein the training data set includes furnace condition characterization parameters corresponding to different abnormal furnace condition types of the blast furnace, and the furnace condition characterization parameters include new features that reflect the dynamic change characteristics of operating parameters before the abnormal situation of the blast furnace occurs. The data set to be tested reflects the operating parameters of the blast furnace to be tested in a new feature characterization manner, analyzes the changes of the blast furnace operating parameters under different types of abnormal blast furnace conditions before the occurrence of the blast furnace, constructs and expands new features for the operating parameters, and makes full use of existing feature information to reduce the false alarm rate of abnormal blast furnace conditions and improve the accuracy; uses a preset prediction model to predict the data set to be tested, and determines the target parameters of the data anomaly. number, wherein the preset prediction model is constructed according to the training data set, if the target parameters in the preset continuous time period are characterized as data anomalies, the fluctuation coefficient of each target parameter is calculated, and whether the furnace condition of the blast furnace to be tested is normal is determined according to the fluctuation coefficient, if the fluctuation coefficient is greater than or equal to the preset threshold value, it is determined that the furnace condition of the blast furnace to be tested is abnormal, and an early warning is issued; if the fluctuation coefficient is less than the preset threshold value, it is determined that the furnace condition of the blast furnace to be tested is normal, and the influence of data mutation on early warning is effectively avoided by determining the state of the data in the continuous time period, and the above method is used to timely issue an early warning of the blast furnace condition, which provides the blast furnace operator with sufficient time to adjust the blast furnace, and relieves the pressure of the blast furnace operator to a certain extent.

[0018] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The drawings herein are incorporated into the specification and constitute a part of the specification, showing embodiments consistent with the present application, and together with the specification, are used to explain the principles of the present application. Obviously, the drawings described below are only some embodiments of the present application, and for those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative work. In the drawings:

[0020] Figure 1 is a schematic diagram of an exemplary system architecture shown in an exemplary embodiment of the present application;

[0021] Figure 2 is a flow chart of an abnormal furnace condition early warning method shown in an exemplary embodiment of the present application;

[0022] Figure 3 It is a schematic diagram of an abnormal data time segment warning process of an abnormal furnace condition warning method shown in an exemplary embodiment of the present application;

[0023] Figure 4 It is a schematic diagram of the abnormal furnace condition classification warning knowledge mechanism judgment process of the abnormal furnace condition warning method shown in an exemplary embodiment of the present application;

[0024] Figure 5 It is a schematic diagram of the overall process of an abnormal furnace condition early warning method shown in an exemplary embodiment of the present application;

[0025] Figure 6 is a block diagram of an abnormal furnace condition early warning device shown in an exemplary embodiment of the present application;

[0026] Figure 7 A schematic diagram of the structure of a computer system suitable for implementing an electronic device of an embodiment of the present application is shown. DETAILED DESCRIPTION

[0027] The following will describe the implementation methods of the present application with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can easily understand other advantages and effects of the present application from the contents disclosed in this specification. The present application can also be implemented or applied through other different specific implementation methods, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present application. It should be understood that the preferred embodiments are only for illustrating the present application, not for limiting the scope of protection of the present application.

[0028] It should be noted that the illustrations provided in the following embodiments are only used to illustrate the basic concept of the present application in a schematic manner, and therefore the illustrations only show components related to the present application rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.

[0029] In the following description, a large number of details are discussed to provide a more thorough explanation of the embodiments of the present application. However, it is obvious to those skilled in the art that the embodiments of the present application can be implemented without these specific details. In other embodiments, well-known structures and devices are shown in the form of block diagrams rather than in detail to avoid making the embodiments of the present application difficult to understand.

[0030] First of all, it should be noted that dimensional difference refers to the situation where there are obvious differences in the measurement scales or units of different features in the same set of data.

[0031] Covariance is used in probability theory and statistics to measure the overall error of two variables. Variance is a special case of covariance, that is, when the two variables are the same. Covariance represents the overall error of the two variables, which is different from the variance that only represents the error of one variable. If the two variables have the same trend, that is, if one is greater than its expected value and the other is also greater than its expected value, then the covariance between the two variables is positive. If the two variables have opposite trends, that is, one is greater than its expected value and the other is less than its expected value, then the covariance between the two variables is negative.

[0032] Cumulative variance contribution rate is an indicator used in statistics, especially in principal component analysis (PCA), to measure the proportion of data variation explained by the first n principal components. It represents the sum of the variance contribution rates of the first n principal components, reflecting the ability of these principal components to explain the overall variation of the data.

[0033] Figure 1 It is a schematic diagram of an exemplary system architecture shown in an exemplary embodiment of the present application.

[0034] Reference Figure 1 As shown, the system architecture may include a data acquisition device 101 and a computer device 102. Among them, the computer device 102 may be at least one of a desktop graphics processing unit (Graphic Processing Unit, GPU) computer, a GPU computing cluster, a neural network computer, etc. The data acquisition device 101 is used to collect training data sets and test data sets. In this embodiment, the data acquisition device 101 obtains the above data and provides it to the computer device 102 for processing. Relevant technicians can use the computer device 102 to predict the test data set using a preset prediction model to determine the target parameters of data anomalies, wherein the preset prediction model is constructed according to the training data set. If the target parameters in the preset continuous time period are characterized as data anomalies, the fluctuation coefficient of each target parameter is calculated, and the furnace condition of the blast furnace to be tested is determined to be normal according to the fluctuation coefficient. If the fluctuation coefficient is greater than or equal to the preset threshold, it is determined that the furnace condition of the blast furnace to be tested is abnormal, and an early warning is performed; if the fluctuation coefficient is less than the preset threshold, it is determined that the furnace condition of the blast furnace to be tested is normal. It should be noted that the data acquisition device 101 and the computer device 102 provided in this embodiment are only an example, and should not bring any restrictions to the functions and scope of use of the embodiments of the present application.

[0035] It should be noted that the abnormal furnace condition warning method provided in the embodiment of the present application is generally executed by the computer device 102 , and accordingly, the abnormal furnace condition warning device is generally arranged in the computer device 102 .

[0036] Figure 2 is a flow chart of an abnormal furnace condition early warning method shown in an exemplary embodiment of the present application. The abnormal furnace condition early warning method can be executed by a computing and processing device. The computing and processing device can be Figure 1 The computer device 102 shown in FIG. Figure 2 As shown, the abnormal furnace condition early warning method at least includes steps S210 to S240, which are described in detail as follows:

[0037] In step S210, a training data set and a test data set are obtained.

[0038] Among them, the training data set includes the furnace condition characterization parameters corresponding to different types of abnormal furnace conditions of the blast furnace, and the furnace condition characterization parameters include new features that reflect the dynamic change characteristics of the operating parameters before the abnormal situation of the blast furnace occurs; the test data set reflects the operating parameters of the blast furnace to be tested in a new feature characterization method.

[0039] In the present embodiment, according to the actual needs of the blast furnace smelting site, the abnormal furnace condition type of the blast furnace that needs early warning and the main parameters characterizing the blast furnace condition, i.e., the furnace condition characterizing parameters, are determined. The essence of the upper adjustment and lower adjustment of blast furnace smelting is to change the material layer structure and the gas distribution, so as to ensure the stable and smooth operation of the blast furnace. For this reason, the abnormal furnace condition of the blast furnace caused by the blast furnace unloading situation and the abnormal gas flow distribution can be divided into pipelines, collapsed materials and suspended materials, that is, the abnormal furnace condition types include but are not limited to pipelines, collapsed materials and suspended materials, and their main characterizing parameters include furnace top temperature, furnace top pressure, gas utilization, total pressure difference, upper pressure difference, lower pressure difference, furnace body static pressure, air permeability index, cold air pressure, hot air pressure, hot air temperature, wind pressure, Bitoba air volume, etc.

[0040] In this embodiment, different abnormal furnace conditions of the blast furnace have different definitions, as follows: The blast furnace pipeline stroke refers to the particularly strong air permeability of a certain area on the cross section of the blast furnace, which makes the coal gas flow in this area develop abnormally like in the pipeline. Before the pipeline occurs, the furnace condition is often poor, and then there are phenomena such as unstable wind pressure, furnace top pressure peak, pressure difference peak, furnace top temperature increase, wind pressure decrease, and increased air volume; when the pipeline occurs, there are often phenomena such as sudden rise in furnace top temperature, sudden rise in furnace top pressure, sudden drop in gas utilization rate, and sudden drop in air volume. Suspended material refers to the slope of the height change of the mechanical probe after the probe reaches the material surface becomes smaller, and the furnace charge descends very slowly or stagnates, or even goes up. Before the suspended material occurs, there are often phenomena such as sudden rise or peak of wind pressure, increase in wind pressure, increase in upper pressure difference (upper suspended material), increase in lower pressure difference (lower suspended material), decrease in air volume, decrease in permeability index, decrease in furnace top pressure, and increase in furnace top temperature. Collapse of material refers to a descent in the form of collapse, and the height of the collapse exceeds a certain range. Before material collapse occurs, there will be phenomena such as increased wind pressure, reduced air volume, drastic fluctuations in furnace top temperature, large fluctuations in furnace top pressure, large fluctuations in furnace body static pressure, and large fluctuations in pressure difference.

[0041] In one embodiment of the present application, the process after obtaining the training data set and the test data set also includes the following steps: calculating the mean and standard deviation of each operating parameter according to the blast furnace operation data; using the standard score formula to calculate the standardized standard value of each operating parameter according to the mean and standard deviation of each operating parameter, so as to standardize each operating parameter.

[0042] In this embodiment, a large amount of historical data is stored in the blast furnace history database, but the data collected on site have large dimensional differences. In order to eliminate the impact of dimensional differences on the model, it is necessary to standardize the parameters. Assume that the parameters selected for a certain abnormal furnace condition are recorded as X = [X1, X2, ..., X n ], where n is the number of parameters selected; X i =[x i,1 ,x i,2 ,...,x i,m ], where x i,j represents the value of the jth sample of the i-th parameter (where 0≤j≤m), and each operating parameter is normalized by the following equations (1) to (3):

[0043]

[0044]

[0045]

[0046] In formula (1) to formula (3), represents the mean of the i-th operating parameter, m is the number of samples, xi,j represents the value of the jth sample of the i-th parameter (where 0≤j≤m), σ i represents the standard deviation of the ith operating parameter, represents the normalized value of the jth sample of the ith operating parameter. All parameters of the ith parameter after normalization are expressed as The parameters selected for this abnormal furnace condition are recorded as

[0047] In this embodiment, the historical data is divided into a training data set and a test data set. For example, the data of the previous D days is used as training data to train the model to obtain a preset prediction model, and the data of the current day is used as test data to test the model, and D is determined according to the recent operation of the blast furnace.

[0048] In one embodiment of the present application, new features of each operating parameter are obtained in the following manner: the operating parameters of a preset sampling time period before the occurrence of an abnormal situation in the blast furnace are determined as effective parameters; the target characteristic parameters are determined from the operating parameters, and the minimum value corresponding to the target characteristic parameter, the maximum value corresponding to the target characteristic parameter, the maximum rising amplitude corresponding to the target characteristic parameter, the maximum falling amplitude corresponding to the target characteristic parameter, the average rising speed corresponding to the target characteristic parameter, and the average falling speed corresponding to the target characteristic parameter are determined based on the effective parameters to construct new features of the target characteristic parameters and obtain new features of each operating parameter.

[0049] In one embodiment of the present application, a new feature of the target feature parameter is constructed in the following manner: the minimum value corresponding to the target feature parameter, the maximum value corresponding to the target feature parameter, and the current value corresponding to the target feature parameter at the target time are determined from the effective parameters; the maximum rise amplitude corresponding to the target feature parameter is calculated based on the minimum value and the current value, the maximum fall amplitude corresponding to the target feature parameter is calculated based on the maximum value and the current value, the average rise speed corresponding to the target feature parameter is calculated according to the maximum rise amplitude and the first time period, and the average fall speed corresponding to the target feature parameter is calculated according to the maximum fall amplitude and the second time period, wherein the first time period is determined by the moment corresponding to the minimum value and the target time, and the second time period is determined by the moment corresponding to the maximum value and the target time; the new feature of the target feature parameter at the target time is constructed according to the minimum value corresponding to the target feature parameter, the maximum value corresponding to the target feature parameter, the maximum rise amplitude corresponding to the target feature parameter, the maximum fall amplitude corresponding to the target feature parameter, the average rise speed corresponding to the target feature parameter, and the average fall speed corresponding to the target feature parameter. The target time is any time in the preset sampling time period before the abnormal situation of the blast furnace occurs, and can be a time selected in advance from the preset sampling time period before the abnormal situation of the blast furnace occurs.

[0050] In this embodiment, since the occurrence of abnormal furnace conditions is the result of long-term accumulation, the corresponding parameters will show obvious changes when abnormal furnace conditions occur. Before the abnormal furnace conditions occur, the parameters change, but the changes are very subtle. Therefore, it is difficult to capture the signs of abnormal furnace conditions by using only the original data set for data mining. For this reason, it is necessary to construct a new feature data set to mine the changes of various parameters before the abnormality occurs. The steps for constructing new features are as follows:

[0051] Step 1: Original characteristics of abnormal furnace conditions before they occur Signs have appeared in the previous kT time, where T is the parameter sampling period and k is a multiple determined by historical operation data or expert experience.

[0052] Step 2: Determine the construction of original features based on the potential action time length kT The valid historical data of the new feature is the valid parameters

[0053] Step 3: Calculate the new feature of the parameter (target feature parameter) at the current moment.

[0054] Step 3.1: New feature 1 is valid historical data The maximum value of

[0055]

[0056] Step 3.2: New feature 2 is valid historical data The minimum value of

[0057]

[0058] Step 3.3: New feature 3 is valid historical data The maximum rise of Among them, x o,0 The current value of the parameter at the current moment.

[0059] Step 3.4: New feature 4 is valid historical data The maximum decrease in

[0060]

[0061] Step 3.5: New feature 5 is valid historical data The average rising speed from the minimum value to the current value is expressed as Where Δd1 is The time step from the minimum value to the current moment is the first time period.

[0062] Step 3.6: New feature 4 is valid historical data The average rate of decrease from the maximum value to the current value is expressed as Where Δd2 is The step length from the maximum value to the current moment is the second time period.

[0063] In summary, the i-th feature The new features constructed at the current moment can be expressed as Then the representation of the new feature set constructed by all the features of the current abnormal furnace condition is referred to formula (4):

[0064]

[0065] According to the above steps, the new feature set of the selected features of the abnormal furnace condition at any time, that is, the new feature representation of the target feature parameters at the target time, can be calculated by referring to formula (5):

[0066]

[0067] in, It is a new feature set, i.e., a new feature, with 6n features, and the number of samples is updated in real time; is the i-th new feature.

[0068] In step S220, a preset prediction model is used to predict the data set to be tested, and target parameters of data anomalies are determined.

[0069] In one embodiment of the present application, the process of constructing a preset prediction model based on a training data set includes the following steps: performing data dimensionality reduction on new features in the training data set to obtain reduced-dimensionality training features; calculating the Hotelling statistic, the prediction error sum of squares statistic, the Hotelling control limit, and the prediction error sum of squares control limit of each operating parameter in the training data set based on the reduced-dimensionality training features, and constructing a preset prediction model based on the Hotelling statistic, the prediction error sum of squares statistic, the Hotelling control limit, and the prediction error sum of squares control limit.

[0070] In this embodiment, a feature data set, i.e., a data set mean of a new feature, is calculated, and a covariance matrix of the feature data set is obtained based on the feature data set and the data set mean; the covariance matrix is ​​decomposed by eigenvalues ​​to obtain eigenvalues ​​and eigenvectors of each principal component in the covariance matrix; the variance contribution rate of each principal component is calculated based on the eigenvalues, the variance contribution rates are accumulated to obtain a cumulative variance contribution rate, and a target dimensionality reduction principal component is determined based on the cumulative variance contribution rate and a preset contribution rate threshold; a reduced feature data set is obtained based on the feature data set and the eigenvector matrix, and a data state determination is performed on the reduced feature data set through a data state determination model, wherein the eigenvector matrix is ​​a matrix composed of eigenvectors corresponding to the target dimensionality reduction principal component.

[0071] In this embodiment, model training includes the following steps:

[0072] 1. Data dimensionality reduction. After feature reconstruction, the feature reconstruction dimension is expanded, which increases the complexity of model training. Therefore, it is necessary to reduce the dimensionality of the feature data set. The specific steps are as follows:

[0073] Step 1: Find the mean of the new feature data set. New feature data set The representation of the mean value of refers to formula (6):

[0074]

[0075] in, is the mean of the i-th feature of the new data set, i.e. the new feature data set. is the jth sample of the i-th feature in the new data set, and KT+1 is the number of samples.

[0076] Step 2: Solve the covariance of the new feature data set. The covariance matrix of the new feature data set is represented by reference formula (7):

[0077]

[0078] Among them, C is the covariance matrix of the new feature, is the i-th feature in the new data set, is the mean of the i-th feature of the new data set, i.e. the new feature data set, and T is the parameter sampling period.

[0079] Step 3: Solve the eigenvalue λ and eigenvector Λ of the covariance matrix of the new feature data set. The calculation formula refers to formula (8):

[0080] CΛ=λΛ Formula (8)

[0081] Among them, C is the covariance matrix of the new feature, λ is the eigenvalue of the covariance matrix of the new feature dataset, Λ is the eigenvector of the covariance matrix of the new feature dataset, and the number of eigenvalues ​​λ is 6n, each λ i (i=1,2,...,6n) corresponds to a eigenvector Λ i (i=1,2,...,6n), requiring λ1≥λ2≥…≥λ 6n .

[0082] Step 4: Determine the principal component to be reduced according to the cumulative variance contribution rate. The calculation formula of the cumulative variance contribution rate is referenced to formula (9):

[0083]

[0084] Where P is the cumulative variance contribution rate of the principal element (pivot) of the first q rows, and p ≥ plimit , p limit Artificially determined, requiring 0≤p limit ≤1;λ i is the eigenvalue of the principal element in the i-th row, λ j is the eigenvalue of the jth principal element, and the number of eigenvalues ​​λ is 6n; when the appropriate p is determined limit After that, the q-row principal elements (q≤6n) after dimension reduction can be determined. The dimension-reduced data Y is represented by Λ q The solution formula is shown in formula (10):

[0085]

[0086] Among them, Y is the feature data set after dimensionality reduction, Λ q Represents the matrix composed of eigenvectors corresponding to the principal element to be reduced in dimension, is a new feature dataset.

[0087] 2. Determine the control limits. Calculate T 2 The statistics are the Hotelling statistic and the SPE statistic (also called the Q statistic), which is the sum of squared prediction errors and their control limits, to determine the normal range of the data, T 2 The statistical calculation formula is shown in formula (11):

[0088]

[0089] Among them, T 2 Indicates T 2 Statistics, S -1 / 2 It means taking the -1 / 2 exponent for each non-zero element in the matrix; represents the Euclidean norm; P represents the new principal component, i.e., the principal element to be reduced in dimension; Y T It means to transpose the feature data set after dimensionality reduction, that is, the feature data set after dimensionality reduction after transposition, P T It means transposing the new principal component, that is, the new principal component after transposition, and Y represents the feature data set after dimensionality reduction.

[0090] T 2 The calculation formula of the control limit of the statistic, i.e. Hotelling's control limit, is expressed as follows:

[0091]

[0092] Among them, T α Indicates T 2 The control limit of the statistic, 1-α is the confidence level, F α (k1,n-k1) is the distribution of furnace top F with the first degree of freedom being k1 and the second degree of freedom being n-k1.

[0093] The calculation formula of Q statistic, i.e., the prediction error sum of squares statistic, is expressed as follows:

[0094] Q=Y T (I-PP T )Y Formula (13)

[0095] Where Q represents the Q statistic, I represents the identity matrix, which has the same number of rows as P, P represents the new principal component, and Y T Indicates the transposition of the feature data set after dimensionality reduction, P T It indicates the transposition of the new principal component, and Y indicates the feature data set after dimensionality reduction.

[0096] The calculation formula of the control limit of Q statistic is as follows:

[0097]

[0098] Among them, Q α represents the control limit of the Q statistic, λ j is the eigenvalue of the jth principal element, (λ j ) r Represents λ j to the rth power (r can be 1, 2 or 3), θ r Including θ1, θ2 and θ3, θ r is j The sum of the rth power, c α is the confidence limit of the standard normal distribution, c α Satisfies the following formula (15):

[0099]

[0100] In one embodiment of the present application, a preset prediction model is used to predict the test data set to determine the target parameter of data anomaly, including: using the preset prediction model to predict the test data set to obtain the Hotelling statistic and the prediction error sum of squares statistic of each operating parameter in the test data set at the target moment; comparing the Hotelling statistic with the corresponding Hotelling control limit, and the prediction error sum of squares statistic with the corresponding prediction error sum of squares control limit, respectively; if the Hotelling statistic is greater than the corresponding Hotelling control limit, and / or the prediction error sum of squares statistic is greater than the corresponding prediction error sum of squares control limit, then it is determined that the operating parameter at the target moment is characterized by data anomaly; if the Hotelling statistic is less than or equal to the corresponding Hotelling control limit, and the prediction error sum of squares statistic is less than or equal to the corresponding prediction error sum of squares control limit, then it is determined that the operating parameter at the target moment is characterized by data normality; and the operating parameter characterized by data anomaly is determined as the target parameter.

[0101] In this embodiment, the model prediction includes the judgment of data anomalies at the current moment and the judgment of data anomalies within a period of time. The specific steps are as follows:

[0102] Step 1: Input the test data set into the model and calculate T at the current moment according to the above formulas (11) and (13). 2 Statistics and Q statistics, if T 2 If the statistic or Q statistic exceeds its control limit, the data at the current moment is considered abnormal.

[0103] Step 2: If T occurs continuously for aT (where a is a positive integer and T is the parameter sampling period) 2 >T α or Q>Q α , then the current data is considered to be obviously abnormal, and further mechanism is needed to determine whether this type of furnace condition abnormality occurs; otherwise, the furnace condition at this moment is considered to be normal. Figure 3 As shown, Figure 3 It is a schematic diagram of an abnormal data time segment warning process of an abnormal furnace condition warning method shown as an exemplary embodiment of the present application.

[0104] In step S230, if the target parameters in the preset continuous time period are all characterized by data anomalies, the fluctuation coefficient of each target parameter is calculated, and whether the furnace condition of the blast furnace to be tested is normal is determined based on the fluctuation coefficient.

[0105] In one embodiment of the present application, if the target parameters in a preset continuous time period are all characterized as data anomalies, the parameter mean of each target parameter in the preset time period is calculated, wherein the preset continuous time period is composed of multiple consecutive target moments; the variation range of each target parameter in the preset time period is calculated based on the parameter mean and the parameter value of each target parameter at the target moment; the fluctuation coefficient of each target parameter in the preset time period is calculated based on the data standard deviation and variation range of each target parameter in the preset time period.

[0106] In this embodiment, when the continuous time segment data is abnormal, whether the furnace condition will be abnormal in the future is further determined by the mechanism, and the specific steps are as follows:

[0107] Step 1: Determine the main influencing parameters. According to the blast furnace conditions at a specific site and the types of abnormal blast furnace conditions to be warned, select a preset number of influencing parameters, for example, select three main influencing parameters, set as α1, α2, α3.

[0108] Step 2: Calculate the change of each parameter in the last b sampling periods (where b is a positive integer determined by on-site expert knowledge). The calculation formula is as follows:

[0109]

[0110] in, It represents the mean of the b consecutive data before the i-th parameter, excluding the current moment; is the i-th parameter, value at time j; is the current value of the i-th parameter; Δα i It can represent the variation of the i-th parameter within a period of time; σ i represents the standard deviation of the previous b consecutive data of the ith parameter, which can be used to represent the fluctuation of the ith parameter in the previous period of time; k i represents the fluctuation coefficient of the i-th parameter at the current moment, k i The larger it is, the greater the fluctuation is and the more likely anomalies are to occur.

[0111] In step S240, if the fluctuation coefficient is greater than or equal to the preset threshold, it is determined that the condition of the blast furnace to be tested is abnormal, and an early warning is issued; if the fluctuation coefficient is less than the preset threshold, it is determined that the condition of the blast furnace to be tested is normal.

[0112] In this embodiment, the fluctuation coefficient k of the i-th parameter is determined i The control limit is the preset threshold k i,lim , when k i >k i,lim This parameter is considered abnormal.

[0113] In one embodiment of the present application, if the fluctuation coefficient is greater than or equal to a preset threshold, the process after determining that the condition of the blast furnace to be tested is abnormal also includes the following steps: counting the number of target parameters whose fluctuation coefficient is greater than or equal to the preset threshold; if the number of target parameters is greater than a first quantity threshold, a first-level warning is issued; if the number of target parameters is greater than a second quantity threshold, a second-level warning is issued; if the number of target parameters is greater than a third quantity threshold, a third-level warning is issued, wherein the first quantity threshold is greater than the second quantity threshold, and the second quantity threshold is greater than the third quantity threshold.

[0114] In this embodiment, based on the example in the above embodiment, if all three parameters satisfy k i >k i,lim , then a first-level warning is issued; if any two parameters satisfy k i >k i,lim , then a second-level warning is issued; if only one parameter or no parameter satisfies k i >k i,lim , then the third level warning, that is, the third level warning is a data abnormality warning, and the probability of abnormal furnace conditions is small. Figure 4 As shown, Figure 4 It is a schematic diagram of the abnormal furnace condition classification warning knowledge mechanism judgment process of the abnormal furnace condition warning method shown in an exemplary embodiment of the present application.

[0115] In one embodiment of the present application, Figure 5 is a schematic diagram of the overall process of an abnormal furnace condition early warning method shown in an exemplary embodiment of the present application, with reference to Figure 5 As shown, the abnormal furnace condition early warning method includes the following steps: according to the actual needs of the blast furnace smelting site, determine the blast furnace abnormal condition category and the main parameters characterizing the blast furnace condition that need to be warned; analyze the changes in the blast furnace operating parameters before the occurrence of different types of abnormal blast furnace conditions, propose a new feature construction method, and construct a new data set; in view of the problem that there are few samples of blast furnace abnormal conditions, the blast furnace abnormal condition is warned based on the principal component analysis method, and the blast furnace condition data is monitored in real time; based on expert knowledge, abnormal furnace condition warning rules are formulated, and the mechanism judgment is made on the time segment of data abnormality, so as to realize the graded warning of blast furnace abnormal conditions. The specific implementation steps have been described in detail in the aforementioned embodiments, and will not be repeated here.

[0116] In one embodiment of the present application, the No. 1 blast furnace of a steel plant is used as the scenario, and Python3.9 is used for product development to verify the effectiveness and feasibility of early warning of blast furnace pipeline travel in real time. First, combined with the on-site operation experience of blast furnace smelting, the changes in the operating parameters and operation parameters of the blast furnace before the pipeline travel occurs are analyzed, and the blast furnace operation parameters are selected, including furnace top temperature, cross temperature measurement, furnace top pressure, furnace body static pressure, total pressure difference, upper pressure difference, lower pressure difference, permeability index, gas utilization rate, air volume, wind pressure, wind temperature and other parameters; secondly, according to the proposed feature construction method, the selected parameters are feature constructed, a new feature data set is established, and it is stored in the database. Next, an algorithm model is built and the corresponding parameters are set. The data of the previous D days in the feature data set are reduced in dimension and model trained to obtain the Q statistic and T 2 The control limits of the statistic are calculated and the current Q statistic and T are predicted in real time. 2 Whether the statistic exceeds the control limit, if it exceeds, the data at the current moment is judged to be abnormal, if not, the data at the current moment is judged to be normal. If data anomalies occur within aT time, the graded warning mechanism is triggered to further judge the mechanism. Then, based on expert knowledge, parameters such as top temperature, full top pressure, and gas utilization rate are selected as mechanism rule parameters, and the current data anomaly is further judged according to the mechanism rule to give the anomaly level. Through the above steps, data mining technology is used to detect whether the data at the current moment is abnormal; at the same time, the alarm mechanism for data anomalies within aT time is used to avoid false detection of data mutations; in addition, expert knowledge is added for abnormal graded warning, which further determines whether a pipeline trip occurs and the severity of the current pipeline trip.

[0117] In one embodiment of the present application, the No. 1 blast furnace of a steel plant is used as the scenario, and Python3.9 is used for product development to verify the effectiveness and feasibility of the blast furnace suspended material warning in real time. First, combined with the on-site operation experience of blast furnace smelting, the changes in the operating parameters and operation parameters of the blast furnace before the pipeline trip occur are analyzed, and the blast furnace operation parameters are selected, including wind pressure, upper pressure difference, lower pressure difference, permeability index, furnace top pressure, furnace top temperature, air volume, wind pressure, wind temperature and other parameters; secondly, according to the proposed feature construction method, the selected parameters are feature constructed, a new feature data set is established, and it is stored in the database. Next, an algorithm model is built and the corresponding parameters are set. The data of the previous D days in the feature data set are reduced in dimensionality and model training are performed to calculate the Q statistic and T 2 The control limit of the statistic and the real-time judgment of the current moment Q statistic and T 2 Whether the statistic exceeds the control limit, if it exceeds, the current data is judged to be abnormal, if not, the current data is judged to be normal. If data abnormalities occur within aT time, the graded warning mechanism is triggered to further judge the mechanism. Then, based on expert knowledge, parameters such as wind pressure, air volume and permeability index are selected as mechanism rule parameters, and the current data abnormality is further judged according to the mechanism rule, and the abnormality level is given. Through the above steps, data mining technology is used to detect whether the current data is abnormal; at the same time, the alarm mechanism for data abnormalities within aT time is used to avoid false detection of data mutations; in addition, expert knowledge is added for abnormal graded warning, which further determines whether blast furnace overhang occurs and the severity of the current overhang.

[0118] This application takes blast furnace conditions as the research object, analyzes the changes in blast furnace operating parameters before the occurrence of abnormal conditions of different blast furnaces, constructs new features and expands features for the selected parameters based on the proposed feature construction method, and makes full use of the existing feature information; then, the new feature set is reduced in dimension based on the principal component analysis method, and the Q statistic and T are calculated. 2 The control limit of the statistic can judge in real time whether the current data is abnormal and whether the continuous time segment data is abnormal, effectively avoiding the impact of data mutation on early warning; then, based on the abnormal continuous time segment data, the abnormal condition of the blast furnace is further judged according to the process logic criteria, and further graded early warning is performed. This application makes full use of data mining technology and expert experience to provide real-time early warning of the blast furnace condition, providing the blast furnace operator with sufficient time to adjust the blast furnace, effectively reducing the false alarm rate of abnormal blast furnace conditions, improving accuracy, and relieving the pressure of the blast furnace on-site operator to a certain extent.

[0119] Figure 6 is a block diagram of an abnormal furnace condition warning device shown in an exemplary embodiment of the present application. The device can be applied to Figure 1The implementation environment shown in the figure is specifically configured in the computer device 102. The device can also be applied to other exemplary implementation environments and specifically configured in other devices. This embodiment does not limit the implementation environment to which the device is applied.

[0120] like Figure 6 As shown, the exemplary abnormal furnace condition warning device includes: an acquisition module 610 , a prediction module 620 , a calculation module 630 and a determination module 640 .

[0121] Among them, the acquisition module 610 is used to obtain a training data set and a data set to be tested, the training data set includes furnace condition characterization parameters corresponding to different abnormal furnace condition types of the blast furnace, and the furnace condition characterization parameters include new features that reflect the dynamic change characteristics of the operating parameters before the abnormal situation of the blast furnace occurs; the data set to be tested reflects the operating parameters of the blast furnace to be tested in a new feature characterization manner; the prediction module 620 is used to use a preset prediction model to predict the data set to be tested and determine the target parameters of the data anomaly, wherein the preset prediction model is obtained by training the training data set; the calculation module 630 is used to calculate the fluctuation coefficient of each target parameter if the target parameters in a preset continuous time period are all characterized as data anomalies, and determine whether the furnace condition of the blast furnace to be tested is normal according to the fluctuation coefficient; the determination module 640 is used to determine that the furnace condition of the blast furnace to be tested is abnormal and issue an early warning if the fluctuation coefficient is greater than or equal to the preset threshold; if the fluctuation coefficient is less than the preset threshold, it is determined that the furnace condition of the blast furnace to be tested is normal.

[0122] It should be noted that the abnormal furnace condition warning device provided in the above embodiment and the abnormal furnace condition warning method provided in the above embodiment belong to the same concept, wherein the specific manner in which each module and unit performs the operation has been described in detail in the method embodiment, and will not be repeated here. In practical applications, the abnormal furnace condition warning device provided in the above embodiment can allocate the above functions to different functional modules as needed, that is, divide the internal structure of the device into different functional modules to complete all or part of the functions described above, and this is not limited here.

[0123] An embodiment of the present application also provides an electronic device, comprising: one or more processors; a storage device for storing one or more programs, and when the one or more programs are executed by one or more processors, the electronic device implements the abnormal furnace condition warning method provided in the above-mentioned embodiments.

[0124] Figure 7 The structure diagram of the computer system suitable for implementing the electronic device of the embodiment of the present application is shown. It should be noted that: Figure 7 The computer system 700 of the electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0125] like Figure 7 As shown, the computer system 700 includes a central processing unit (CPU) 701, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 702 or the program loaded from the storage part 708 to the random access memory (RAM) 703, such as executing the methods provided in the above-mentioned various embodiments. In the RAM 703, various programs and data required for system operation are also stored. The CPU 701, the ROM 702 and the RAM 703 are connected to each other through the bus 704. The input / output (I / O) interface 705 is also connected to the bus 704.

[0126] The following components are connected to the I / O interface 705: an input section 706 including a keyboard, a mouse, etc.; an output section 707 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 708 including a hard disk, etc.; and a communication section 709 including a network interface card such as a LAN (Local Area Network) card, a modem, etc. The communication section 709 performs communication processing via a network such as the Internet. A drive 710 is also connected to the I / O interface 705 as needed. A removable medium 711, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 710 as needed so that a computer program read therefrom is installed into the storage section 708 as needed.

[0127] In particular, according to an embodiment of the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present application includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a computer program for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication section 709, and / or installed from a removable medium 711. When the computer program is executed by a central processing unit (CPU) 701, various functions defined in the system of the present application are executed.

[0128] It should be noted that the computer-readable medium shown in the embodiment of the present application can be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium can be, for example, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media can include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, wherein a computer-readable computer program is carried. This propagated data signal can take a variety of forms, including but not limited to an electromagnetic signal, an optical signal, or any suitable combination of the above. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. A computer program contained on a computer-readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wired, etc., or any suitable combination of the above.

[0129] The flowchart and block diagram in the accompanying drawings illustrate the possible architecture, functions and operations of the system, method and computer program product according to various embodiments of the present application. Wherein, each box in the flowchart or block diagram can represent a module, a program segment, or a part of the code, and the above-mentioned module, program segment, or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flowchart, and the combination of boxes in the block diagram or flowchart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0130] The units involved in the embodiments described in this application may be implemented by software or hardware, and the units described may also be set in a processor. The names of these units do not, in some cases, constitute limitations on the units themselves.

[0131] Another aspect of the present application further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor of a computer, causes the computer to execute the abnormal furnace condition early warning method provided in the above-mentioned embodiments. The computer-readable storage medium may be included in the electronic device described in the above-mentioned embodiments, or may exist independently without being assembled into the electronic device.

[0132] It should be noted that, although several modules or units of the equipment for action execution are mentioned in the above detailed description, this division is not mandatory. In fact, according to the embodiments of the present application, the features and functions of two or more modules or units described above can be embodied in one module or unit. On the contrary, the features and functions of one module or unit described above can be further divided into being embodied by multiple modules or units.

[0133] Another aspect of the present application also provides a computer program product or a computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the abnormal furnace condition early warning method provided in each of the above embodiments.

[0134] Through the description of the above implementation methods, it is easy for those skilled in the art to understand that the example implementation methods described here can be implemented by software or by combining software with necessary hardware. Therefore, the technical solution according to the implementation methods of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, and includes several instructions to enable a computing device (which can be a personal computer, a server, a touch terminal, or a network device, etc.) to execute the method according to the implementation methods of the present application.

[0135] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the embodiments disclosed herein. The present application is intended to cover any variations, uses or adaptations of the present application, which follow the general principles of the present application and include common knowledge or customary technical means in the art that are not disclosed in the present application.

[0136] The above embodiments are merely illustrative of the principles and effects of the present application, and are not intended to limit the present application. Anyone familiar with the technology may modify or change the above embodiments without violating the spirit and scope of the present application. Therefore, all equivalent modifications or changes made by a person of ordinary skill in the art without departing from the spirit and technical ideas disclosed in the present application shall still be covered by the claims of the present application.

Claims

1. A method for early warning of abnormal furnace conditions, characterized in that: include: Acquire a training data set and a test data set, wherein the training data set includes furnace condition characterization parameters corresponding to different abnormal furnace condition types of the blast furnace, and the furnace condition characterization parameters include new features reflecting the dynamic change characteristics of the operating parameters before the abnormal situation of the blast furnace occurs; the test data set reflects the operating parameters of the blast furnace to be tested in the manner of characterization of the new features; Using a preset prediction model to predict the data set to be tested, and determining target parameters of data anomalies, wherein the preset prediction model is constructed based on the training data set; If the target parameters are all characterized as data anomalies within a preset continuous time period, the fluctuation coefficient of each target parameter is calculated, and whether the furnace condition of the blast furnace to be tested is normal is determined according to the fluctuation coefficient; If the fluctuation coefficient is greater than or equal to the preset threshold, it is determined that the condition of the blast furnace to be tested is abnormal, and an early warning is issued; if the fluctuation coefficient is less than the preset threshold, it is determined that the condition of the blast furnace to be tested is normal.

2. The abnormal furnace condition early warning method according to claim 1, characterized in that: The new characteristics of each of the operating parameters are obtained in the following manner: Determine the operating parameters of a preset sampling time period before the abnormal situation of the blast furnace occurs as valid parameters; Determine a target characteristic parameter from the operating parameters, and determine a minimum value corresponding to the target characteristic parameter, a maximum value corresponding to the target characteristic parameter, a maximum rising amplitude corresponding to the target characteristic parameter, a maximum falling amplitude corresponding to the target characteristic parameter, an average rising speed corresponding to the target characteristic parameter, and an average falling speed corresponding to the target characteristic parameter based on the effective parameters, so as to construct a new feature of the target characteristic parameter and obtain a new feature of each of the operating parameters.

3. The abnormal furnace condition early warning method according to claim 2, characterized in that: A new feature of the target feature parameter is constructed in the following way: Determine, from the valid parameters, a minimum value corresponding to the target characteristic parameter, a maximum value corresponding to the target characteristic parameter, and a current value corresponding to the target characteristic parameter at a target time; The maximum rising amplitude corresponding to the target characteristic parameter is calculated based on the minimum value and the current value, the maximum falling amplitude corresponding to the target characteristic parameter is calculated based on the maximum value and the current value, the average rising speed corresponding to the target characteristic parameter is calculated according to the maximum rising amplitude and a first time period, and the average falling speed corresponding to the target characteristic parameter is calculated according to the maximum falling amplitude and a second time period, wherein the first time period is determined by the moment corresponding to the minimum value and the target time, and the second time period is determined by the moment corresponding to the maximum value and the target time; A new feature of the target feature parameter at the target moment is constructed according to the minimum value corresponding to the target feature parameter, the maximum value corresponding to the target feature parameter, the maximum rising amplitude corresponding to the target feature parameter, the maximum falling amplitude corresponding to the target feature parameter, the average rising speed corresponding to the target feature parameter and the average falling speed corresponding to the target feature parameter.

4. The abnormal furnace condition early warning method according to claim 1, characterized in that: The preset prediction model is constructed according to the training data set, including: Performing data dimensionality reduction on the new features in the training data set to obtain dimensionality-reduced training features; The Hotelling statistic, the prediction error sum of squares statistic, the Hotelling control limit, and the prediction error sum of squares control limit of each operating parameter in the training data set are calculated respectively according to the dimensionality reduction training features, and the preset prediction model is constructed according to the Hotelling statistic, the prediction error sum of squares statistic, the Hotelling control limit, and the prediction error sum of squares control limit.

5. The abnormal furnace condition early warning method according to claim 4, characterized in that: The preset prediction model is used to predict the data set to be tested, and the target parameters of the data anomaly are determined, including: Using the preset prediction model to predict the data set to be tested, and obtaining the Hotelling statistic and the prediction error square sum statistic of each operating parameter in the data set to be tested at the target time; Comparing the Hotelling statistic with the corresponding Hotelling control limit, and the prediction error sum of squares statistic with the corresponding prediction error sum of squares control limit respectively; If the Hotelling statistic is greater than the corresponding Hotelling control limit, and / or the forecast error sum of squares statistic is greater than the corresponding forecast error sum of squares control limit, it is determined that the operating parameter at the target time is characterized by data anomaly; If the Hotelling statistic is less than or equal to the corresponding Hotelling control limit, and the forecast error sum of squares statistic is less than or equal to the corresponding forecast error sum of squares control limit, it is determined that the operating parameter at the target time is characterized as normal data; An operating parameter characterized by data anomaly is determined as the target parameter.

6. The abnormal furnace condition early warning method according to any one of claims 1 to 5, characterized in that: If the target parameters are all characterized as data anomalies within a preset continuous time period, the fluctuation coefficient of each target parameter is calculated, including: If the target parameters in a preset continuous time period are all characterized as data anomalies, then the parameter mean of each target parameter in the preset time period is calculated, wherein the preset continuous time period is composed of a plurality of consecutive target moments; Calculating the variation range of each target parameter within the preset time period based on the parameter mean and the parameter value of each target parameter at the target time; The fluctuation coefficient of each target parameter within the preset time period is calculated according to the data standard deviation and the variation range of each target parameter within the preset time period.

7. The abnormal furnace condition early warning method according to any one of claims 1 to 5, characterized in that: If the fluctuation coefficient is greater than or equal to the preset threshold, after determining that the condition of the blast furnace to be tested is abnormal, the method further includes: Counting the number of target parameters whose fluctuation coefficient is greater than or equal to the preset threshold; If the number of target parameters is greater than the first quantity threshold, a first-level warning is issued; if the number of target parameters is greater than the second quantity threshold, a second-level warning is issued; if the number of target parameters is greater than the third quantity threshold, a third-level warning is issued, wherein the first quantity threshold is greater than the second quantity threshold, and the second quantity threshold is greater than the third quantity threshold.

8. An abnormal furnace condition early warning device, characterized in that: include: An acquisition module is used to acquire a training data set and a test data set, wherein the training data set includes furnace condition characterization parameters corresponding to different abnormal furnace condition types of the blast furnace, and the furnace condition characterization parameters include new features that reflect the dynamic change characteristics of operating parameters before the abnormal situation of the blast furnace occurs; the test data set reflects the operating parameters of the blast furnace to be tested in a manner characterized by the new features; A prediction module, used to predict the data set to be tested by using a preset prediction model to determine target parameters of data anomalies, wherein the preset prediction model is trained based on the training data set; A calculation module, for calculating the fluctuation coefficient of each target parameter if the target parameters are all characterized as data abnormalities within a preset continuous time period, and determining whether the furnace condition of the blast furnace to be tested is normal according to the fluctuation coefficient; The determination module is used to determine that the condition of the blast furnace to be tested is abnormal and issue an early warning if the fluctuation coefficient is greater than or equal to a preset threshold; if the fluctuation coefficient is less than the preset threshold, determine that the condition of the blast furnace to be tested is normal.

9. An electronic device, characterized in that: The electronic device comprises: one or more processors; A storage device for storing one or more programs, when the one or more programs are executed by the one or more processors, enables the electronic device to implement the abnormal furnace condition warning method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed by a processor of a computer, the computer is caused to execute the abnormal furnace condition early warning method according to any one of claims 1 to 7.

Images