CN108364095A - Molten steel quality diagnosis method in steelmaking production process based on data mining - Google Patents

Abstract

The invention provides a method for diagnosing molten steel quality in a steelmaking production process based on data mining, and belongs to the technical field of molten steel quality diagnosis. This method collects and screens the data, and then standardizes the data, determines the steel moisture and temperature control targets of each process based on clustering, extracts the classification of molten steel (hot metal) and the reasonable mode of process operation technology, and predicts the quality of molten steel in the process. Provide process operation plan, and finally realize system early warning. This method fills the vacancy that most of the existing researches only focus on the quality of casting slabs, and at the same time, it can also timely regulate the quality of molten steel in the process of steelmaking production, thereby ensuring the stability of the production process and narrow window control, and can effectively Improve product quality.

Description

Diagnosis method of molten steel quality in steelmaking production process based on data mining

technical field

The invention relates to the technical field of molten steel quality diagnosis, in particular to a method for diagnosing molten steel quality in a steelmaking production process based on data mining.

Background technique

At present, the outstanding problems of my country's steel production are the overproduction of conventional steel and the gap between the quality of high-quality steel and the international advanced level, which often needs to be realized by importing. When there are problems with some key steel products, quality positioning and diagnosis are required. Therefore, it is urgent to use the method of combining modern data mining and metallurgical mechanism to realize the process quality diagnosis of the whole process. Many experts at home and abroad have made research on it. The metallurgical full-process process quality online monitoring and offline analysis system developed by Beijing University of Science and Technology High-efficiency Rolling National Engineering Research Center proposes to comprehensively collect, monitor and online grade the process quality parameters of the whole process for the quality control of automotive plates and substrates, and to detect problems in the event of problems The need for rapid traceability, analysis, optimization and improvement of the entire process. Peng Kaixiang from Beijing University of Science and Technology and others proposed a fault diagnosis method and device for strip hot rolling quality, which solves the problem that product quality in the prior art is often controlled by skilled operators with their own experience, so that once a fault occurs, It is difficult to guarantee product quality only by feedback control strategy of delay and lag.

However, the current quality diagnosis technology is difficult to realize the diagnosis and early warning of iron and steel products in the production process. It is only a sampling inspection after the end of a certain pouring production, and the quality adjustment will be made after the next pouring, which is difficult to prevent quality problems. The emergence of batches has brought great limitations to improving production efficiency. At the same time, most of the current quality diagnosis in the metallurgical industry is for the quality diagnosis of casting slabs, and there are few quality diagnoses for the three production process steps of molten iron pretreatment, converter steelmaking and external refining. At the same time, from the perspective of diagnostic objects, most of the quality diagnosis in the iron and steel industry is proposed under the premise that the process variables are independent of each other, but the present invention adopts a comprehensive quality diagnosis technology for dependent variables. There is a certain relationship with the ingredients, and it is a method of comprehensive consideration of multiple factors.

Contents of the invention

The technical problem to be solved by the present invention is to provide a method for diagnosing molten steel quality in the steelmaking production process based on data mining.

The method includes steps as follows:

(1) Data collection and screening:

Use the secondary system of the steelmaking plant to collect the production process data of the steelmaking plant, use the furnace number to connect the production data of the four processes of molten iron pretreatment, converter steelmaking, LF furnace refining, and RH furnace refining in series, and generate a joint table to filter out valid data in it;

(2) Standardization of data:

Standardize the valid data screened out in step (1);

(3) Determination of molten steel or molten iron composition and temperature control targets in each process based on clustering:

From the data after step (2) standardized processing, the sets of heats processed by different processes are screened out, and all sets of heats whose outbound composition and temperature of each process meet the operating regulations are further screened out; the temperature and composition of molten steel or molten iron are analyzed The influence of parameters on the treatment of different processes, the temperature and composition parameters of molten steel or molten iron that have an important impact on the quality of molten steel or molten iron are selected as clustering factors, and all furnaces that have been screened out of each process and whose composition and temperature meet the operating regulations are aggregated The heat data in the furnace is clustered and processed by K-Means clustering algorithm to determine the target of molten steel or molten iron composition and temperature control in each process;

(4) Classification of molten steel or molten iron and extraction of reasonable mode of process operation process:

The outbound composition and temperature of each process screened in step (3) meet all the heats set as hit heats, and the rest of the heats are non-hit heats. Analyze and compare the different processes of hit heats and non-hit heats Treatment process, summarizing the treatment process characteristics of hit heats and non-hit heats, the treatment process of hit heats is the reasonable corresponding process process mode of this type of molten steel, analyzing and collecting process modes of different processes, And enter the database management;

(5) Process molten steel quality prediction:

The incoming molten steel or molten iron temperature and composition of a certain heat are collected through the secondary system, if it exceeds the incoming control target of the corresponding process, but within the controllable range (that is, the composition temperature of all hit heats traced back to the incoming time The collection of information is within the controllable range of the process), then according to the temperature and composition values of molten steel or molten iron, it is classified into the corresponding category after processing in step (3), and then according to the hit of this category The rate predicts the probability that the temperature and composition of the molten steel of this furnace or molten iron after corresponding process treatment will reach the end point control target, and the value is equal to the hit rate;

(6) The process operation plan provides:

According to the prediction of the molten steel quality in the step (5), the process operation mode scheme is provided for the corresponding heat of the corresponding process, so as to ensure that the control target requirements are finally met;

(7) Warning:

If the incoming molten steel temperature and composition of a certain furnace exceed the controllable range, the system will alarm.

Among them, the effective data in step (1) is the data of a heat in the whole steelmaking process, from KR pretreatment to the end of refining, the temperature, composition, various process operations and parameter information are complete and within a reasonable range.

The standardization process in step (2) specifically uses the Min-Max standardization method to turn all clustering factors into scalars and map them in the interval [0,1].

The beneficial effects of above-mentioned technical scheme of the present invention are as follows:

The method of the present invention fills up the vacancy that most of the existing researches only focus on the research on the quality of the slab, and at the same time, it can also timely regulate the quality of molten steel during the steelmaking production process, thereby ensuring the stability of the production process and narrow window control, and can Effectively improve product quality.

Description of drawings

Fig. 1 is the flow chart of the method for diagnosing molten steel quality in the steelmaking production process based on data mining of the present invention;

Fig. 2 is a flowchart of the actual application of the system diagnosis method for molten steel temperature and composition in the steelmaking production process provided by the embodiment of the present invention;

Fig. 3 is a flow chart of the K-Means clustering algorithm used when clustering production data provided by an embodiment of the present invention;

FIG. 4 is a flowchart of a decision tree algorithm provided by an embodiment of the present invention.

Detailed ways

In order to make the technical problems, technical solutions and advantages to be solved by the present invention clearer, the following will describe in detail with reference to the drawings and specific embodiments.

The present invention provides a method for diagnosing molten steel quality in the steelmaking production process based on data mining, as shown in Figure 1, the method includes the following steps:

(1) Data collection and screening:

Use the secondary system of the steelmaking plant to collect the production process data of the steelmaking plant, use the furnace number to connect the production data of the four processes of molten iron pretreatment, converter steelmaking, LF furnace refining, and RH furnace refining in series, and generate a joint table to filter out valid data in it;

(2) Standardization of data:

Standardize the valid data screened out in step (1);

(3) Determination of molten steel or molten iron composition and temperature control targets in each process based on clustering:

From the data after step (2) standardized processing, the sets of heats processed by different processes are screened out, and all sets of heats whose outbound composition and temperature of each process meet the operating regulations are further screened out; the temperature and composition of molten steel or molten iron are analyzed The influence of parameters on the treatment of different processes, the temperature and composition parameters of molten steel or molten iron that have an important impact on the quality of molten steel or molten iron are selected as clustering factors, and all furnaces that have been screened out of each process and whose composition and temperature meet the operating regulations are aggregated The heat data in the furnace is clustered and processed by K-Means clustering algorithm to determine the target of molten steel or molten iron composition and temperature control in each process;

(4) Classification of molten steel or molten iron and extraction of reasonable mode of process operation process:

The outbound composition and temperature of each process screened in step (3) meet all the heats set as hit heats, and the rest of the heats are non-hit heats. Analyze and compare the different processes of hit heats and non-hit heats Treatment process, summarizing the treatment process characteristics of hit heats and non-hit heats, the treatment process of hit heats is the reasonable corresponding process process mode of this type of molten steel, analyzing and collecting process modes of different processes, And enter the database management;

(5) Process molten steel quality prediction:

The incoming molten steel or molten iron temperature and composition of a furnace is collected through the secondary system. If the incoming control target of the corresponding process is exceeded, but within the controllable range (the composition temperature information of all hit furnaces can be traced back to the incoming time) within the collection), then according to the temperature and composition values of molten steel or molten iron, classify it into the corresponding category after treatment in step (3), and then predict the heat of molten steel or molten iron according to the hit rate of this category The probability that the temperature and composition of the molten steel after the corresponding process reaches the end point control target, and the value is equal to the hit rate;

(6) The process operation plan provides:

According to the prediction of the molten steel quality in the step (5), the process operation mode scheme is provided for the corresponding heat of the corresponding process, so as to ensure that the control target requirements are finally met;

(7) Warning:

If the incoming molten steel temperature and composition of a certain furnace exceed the controllable range, the system will alarm.

As shown in Figure 2, the specific implementation process is as follows:

(1) Data collection and screening

Collect historical sample data of a certain type of molten steel in the steelmaking plant, the data includes KR molten iron pretreatment data, converter steelmaking data, LF refining process data, RH refining process data, slab continuous casting production control Records, ladle process query records, and use the database to link these production data based on the furnace number to form a joint data table for the entire process of the steelmaking plant and filter out valid data. The main research is on the composition and temperature data that have an important impact on the quality of molten steel when the molten steel enters and exits the station. For different steel types, all the heats processed by selecting the steel type from the historical sample data are recorded as a set S, and in all the heat set S, all the heats whose outbound composition and temperature meet the operating regulations are hit heats , denoted as set P.

(2) Standardization of data

Since these temperature and composition data are different in distribution intervals and units, it is unreasonable to cluster them directly, so they must be standardized first. The processing method used here is the Min-Max standardized processing method. It is a linear transformation of the original data. Let minA and maxA be the maximum value and minimum value of attribute A respectively, and an original value x of A is mapped to the value x` in the interval [0,1] through Min-Max standardization. The formula is:

(3) Determination of molten steel (hot metal) components and temperature control targets in each process based on clustering

First trace the information of the set P back to the time of entering the station, and then use the K-Means clustering algorithm to perform cluster analysis on the historical sample data of its composition and temperature information. As shown in Figure 3, the composition and temperature of a furnace of molten steel are Multiple variables such as temperature information are stored in the form of a matrix, and the rows of the matrix are used to represent the molten steel of each furnace, and the columns are used to represent the information of the molten steel of the furnace, so the set of p pieces of information on n furnaces of molten steel can be represented by an n× It is represented by a p-dimensional matrix, and the jth information of the i-th heat molten steel is expressed as x _ij in the matrix, and the data matrix is as follows:

Then use the dissimilarity matrix to store the differences between different heats of the same steel type. The dissimilarity matrix of n heats of molten steel is expressed as an n×n-dimensional matrix, and d(A,B) is used to represent A The dissimilarity with B, the dissimilarity matrix of the set X={x ₁ ,x ₂ ,…,x _n } containing n heats is expressed as follows:

Here, d( _xi , x _j ) is a similarity measurement function. When _xi is similar or similar, the value of d( _xi , x _j ) is close to 0, and when d( _xi , x _j ) is When it is larger, it means that there is a big difference between heats x _i and x _j . At present, the most commonly used similarity measure function is Euclidean distance, which is defined as:

x _i and x _j represent any two furnaces of molten steel, p is the variable number of molten steel, and n is the number of furnaces.

Determine the data set X={x ₁ ,x ₂ ,…,x _n } and the number of clusters according to the historical sample data;

①Initialization: Randomly specify k cluster centers (m ₁ ,m ₂ ,…,m _k );

②Assign _xi : For each sample _xi , find the nearest cluster center and assign it to this class;

③Recalculate the centers of each cluster:

④ Calculate the deviation:

⑤ Judgment of convergence: If the J value converges, the algorithm terminates; otherwise, return to the second step.

⑥ Obtaining modes: through repeated operations until convergence, k modes can be obtained

These k modalities correspond to k clustering centers, and then perform clustering again with these k centers as the clustering centers of all data sets S, so that all production data are divided into k categories, and for each type of data Excavate the qualified rate after the process ends, and consider that the set of high hit rate class is the operating procedure when the process enters the station, and the molten steel temperature value m and each component value _ni of the cluster center of this type of molten steel are the optimal control values , according to the requirements of smelting steel, determine the target range of the end point ΔT`, combined with the temperature drop of molten steel between processes ΔT, the target temperature of the end point of the previous process is m+ΔT±ΔT`. According to the requirements of smelting steel, the target range of the end point Δn is determined, and the composition of the process does not change much, so the target composition of the end point of the previous process is _ni + Δn.

(4) Classification of molten steel (hot metal) and extraction of reasonable mode of process operation process

In the previous step, the k-type molten steel was obtained. In each type of molten steel, some heats belong to hit heats. The set of all hit heats is the controllable range of the molten steel in this process. Calculate each type The proportion of hit heats to all heats of this type is recorded as the hit rate j of this type of molten steel entering the station for this process.

Analyze and compare the differences in the treatment process of the hit furnace and the non-hit furnace in this process, and summarize and extract the treatment process characteristics of the hit furnace. This process is done using the decision tree method:

As shown in Figure 4, the continuous attribute is processed using the decision tree method. Assuming that the continuous attribute a has n different values in the sample set D, the proportion of qualified and unqualified samples is p _k (k=1, 2,...,|y|)

①Sort the n different values of a continuous attribute appearing in the sample set from small to large, and record it as {a ¹ ,a ² ,...,a ⁿ }

②The midpoint between two adjacent values a ⁱ and a ⁱ⁺¹ of an attribute As a possible split point t, the data set is divided into two parts, and the information gain of each possible split point is calculated:

where information entropy

③ Select the split point with the largest information gain after correction as the best split point for this feature

④ Calculate the information gain rate of the best split point and correct it as the information gain of this attribute

⑤Comparing the information gain rate of each attribute to construct a decision tree, the one with a large gain rate is used as the root node, and the decision tree is extended in turn and the decision tree is pruned

The treatment process of the hit heat is the reasonable treatment process mode of this type of molten steel in this process. Analyze and collect the treatment process modes of all k-type molten steels in a certain process, and enter them into the database for management.

(5) Process molten steel quality prediction

The incoming molten steel temperature and composition of a furnace in a certain process are collected through the secondary system. If the incoming control target of this process is exceeded but within the controllable range, it will be taken according to the value of its temperature and composition. Classify it into one of the k categories, and then predict the probability that the molten steel temperature and composition of the heat of molten steel after RH treatment will reach the end control target according to the hit rate of this category, and the value is equal to the hit rate.

(6) Provide process operation plan

According to the process mode corresponding to the k-type incoming molten steel in a certain process, a reasonable process operation mode plan is provided for the heat, and the process is guided to carry out reasonable processing to ensure that the control target requirements are finally met.

(7) early warning

If the temperature and composition of the incoming molten steel in a certain process of a heat exceed the controllable range, the system will alarm and warn the process that the temperature and composition range of the molten steel in this heat exceeds the temperature and composition range of all hit heats in the historical data. The hit rate of this process is very low, so we need to focus on it and choose an unconventional treatment process.

The above description is a preferred embodiment of the present invention, it should be pointed out that for those of ordinary skill in the art, without departing from the principle of the present invention, some improvements and modifications can also be made, these improvements and modifications It should also be regarded as the protection scope of the present invention.

Claims (4)

1. A method for diagnosing molten steel quality in a steelmaking production process based on data mining, characterized in that: the steps are as follows: (1) Data collection and screening: Use the secondary system of the steelmaking plant to collect the production process data of the steelmaking plant, use the furnace number to connect the production data of the four processes of molten iron pretreatment, converter steelmaking, LF furnace refining, and RH furnace refining in series, and generate a joint table to filter out valid data in it; (2) Standardization of data: Standardize the valid data screened out in step (1); (3) Determination of molten steel or molten iron composition and temperature control targets in each process based on clustering: From the data after step (2) standardized processing, the sets of heats processed by different processes are screened out, and all sets of heats whose outbound composition and temperature of each process meet the operating regulations are further screened out; the temperature and composition of molten steel or molten iron are analyzed The influence of parameters on the treatment of different processes, the temperature and composition parameters of molten steel or molten iron that have an important impact on the quality of molten steel or molten iron are selected as clustering factors, and all furnaces that have been screened out of each process and whose composition and temperature meet the operating regulations are aggregated The heat data in the furnace is clustered and processed by K-Means clustering algorithm to determine the target of molten steel or molten iron composition and temperature control in each process; (4) Classification of molten steel or molten iron and extraction of reasonable mode of process operation process: The outbound composition and temperature of each process screened in step (3) meet all the heats set as hit heats, and the rest of the heats are non-hit heats. Analyze and compare the different processes of hit heats and non-hit heats Treatment process, summarizing the treatment process characteristics of hit heats and non-hit heats, the treatment process of hit heats is the reasonable corresponding process process mode of this type of molten steel, analyzing and collecting process modes of different processes, And enter the database management; (5) Process molten steel quality prediction: The incoming molten steel or molten iron temperature and composition of a furnace are collected through the secondary system. If the entry control target of the corresponding process is exceeded but within the controllable range, the temperature and composition of the molten steel or molten iron will be selected. Classify it into the corresponding category after treatment in step (3), and then predict the probability that the furnace molten steel or the molten steel temperature and composition of the molten steel after corresponding process treatment will reach the terminal control target according to the hit rate of this category, and the value is equal to the hit rate. Rate; (6) The process operation plan provides: According to the prediction of the molten steel quality in the step (5), the process operation mode scheme is provided for the corresponding heat of the corresponding process, so as to ensure that the control target requirements are finally met; (7) Warning: If the incoming molten steel temperature and composition of a certain furnace exceed the controllable range, the system will alarm. 2. The method for diagnosing molten steel quality in the steelmaking production process based on data mining according to claim 1 is characterized in that: in the described step (1), the effective data is the data of a heat in the whole steelmaking process from the KR preset Complete temperature, composition, various process operations and parameter information from processing to the end of refining. 3. The method for diagnosing molten steel quality in the steelmaking production process based on data mining according to claim 1, characterized in that: the standardization process in the step (2) is specifically to utilize the Min-Max standardization method to convert all clustering factors into scalar and all map in the interval [0,1]. 4. The method for diagnosing molten steel quality in the steelmaking production process based on data mining according to claim 1, characterized in that: the controllable range in the step (5) is the composition of all hit heats traced back to the moment of entering the station A collection of temperature information.