Disclosure of Invention
The invention aims to provide a method and equipment for setting a range of process parameters of a blast furnace and a storage medium.
In order to achieve one of the above objects, an embodiment of the present invention provides a method for setting a range of process parameters of a blast furnace, the method including:
acquiring data of one parameter in the blast furnace process parameters and the correlation parameter of the parameter;
analyzing the parameters and the correlation parameters by using an interval analysis method to obtain a linear regression relationship between the parameters and each correlation parameter;
and obtaining a reasonable range of the parameters by combining known target indexes of one or more correlation parameters according to the linear regression relationship.
As a further improvement of an embodiment of the present invention, the method further comprises:
and obtaining the alarm range of the parameters by combining with the known lower limit indexes of one or more correlation parameters according to the linear regression relationship.
As a further improvement of an embodiment of the present invention, the "interval analysis method" includes:
acquiring sample data of a plurality of parameters at different time points, and performing interval division on the fluctuation range of the sample data of a first parameter;
according to the time corresponding relation between other parameters and the first parameter, carrying out same interval division on the sample data of all other parameters, and calculating the average value of each parameter in each interval;
and respectively taking the average value of the first parameter and other parameters in each interval as coordinate values of two coordinate axes, and respectively calculating the linear regression relationship of the first parameter and other parameters.
As a further improvement of an embodiment of the present invention, the fluctuation range of the sample data of the first parameter is divided into intervals by means of average division.
As a further improvement of an embodiment of the present invention, the "section dividing a fluctuation range of sample data of the first parameter" includes:
dividing the fluctuation range of the sample data of the first parameter into a plurality of intervals, counting the total sample size of the first parameter and the sample size in each interval, and calculating the sample size ratio of each interval;
and deleting the interval with the sample volume ratio less than the preset threshold value to obtain the finally divided interval.
As a further improvement of an embodiment of the present invention, the correlation parameter includes the yield of the blast furnace, and a reasonable range of the parameter is obtained by a linear regression relationship between the parameter and the yield in combination with a known target index of the yield;
or the correlation parameter comprises the fuel ratio of the blast furnace, and a reasonable range of the parameter is obtained by combining a known target index of the fuel ratio through a linear regression relation of the parameter and the fuel ratio.
As a further improvement of an embodiment of the present invention, the "acquiring data of one of the blast furnace process parameters and the correlation parameter thereof" includes:
classifying blast furnace process parameters, wherein the blast furnace process parameters are divided into input parameters, process parameters and output parameters;
establishing a time corresponding relation among the input parameters, the process parameters and the output parameters;
establishing a blast furnace database according to the time corresponding relation and the collected data of the blast furnace process parameters;
and acquiring data of one parameter and the related parameter from the database.
As a further improvement of an embodiment of the present invention, the input parameters include a coke quality parameter, a sintered mineral quality parameter, a pellet quality parameter, a lump ore quality parameter, a batching structure parameter and a bin level parameter, the process parameters include an operation parameter, a furnace condition characterization parameter and a furnace body management parameter, and the output parameters include a technical and economic index parameter of the blast furnace.
As a further improvement of an embodiment of the present invention, the establishing a time correspondence between the input parameter and the process parameter and the output parameter specifically includes:
the time corresponding relation of the input parameters, the process parameters and the output parameters of the blast furnace is calculated or obtained through a tracer test by detecting and testing data of raw materials, factory arrival time, arrival quantity, change of a finished product bin position, belt transfer speed and transfer quantity from the finished product bin to a blast furnace raw material bin, a blast furnace raw material bin position, transfer speed and transfer quantity after the blast furnace raw materials are loaded and dynamic monitoring of a smelting period of the blast furnace raw materials in the blast furnace.
As a further improvement of an embodiment of the present invention, the "establishing a blast furnace database of the collected data of the blast furnace process parameters" specifically includes:
the method comprises the steps of carrying out data cleaning, data mining and data fusion on data in a blast furnace database, and carrying out data analysis, monitoring and alarming by using the fused data in the blast furnace database, wherein the data cleaning refers to removing abnormal points in the acquired data, the data mining refers to calculating the data of indirect parameters through an existing formula on the basis of the data, and the data fusion refers to unifying the data frequency or data cycle of all parameters to obtain cycle data.
In order to achieve one of the above objects, an embodiment of the present invention provides an electronic device, which includes a memory and a processor, wherein the memory stores a computer program operable on the processor, and the processor implements the steps of any one of the above methods for setting the range of the blast furnace process parameter when executing the program.
In order to achieve one of the above objects, an embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored, wherein the computer program, when executed by a processor, implements the steps of any one of the above methods for setting the range of the blast furnace process parameter.
Compared with the prior art, the method for setting the range of the process parameters of the blast furnace analyzes the data of the process parameters of the blast furnace by using an interval analysis method, and scientifically appoints a reasonable range and an alarm range for the process parameters of the blast furnace. Meanwhile, according to the reasonable range and the alarm range, real-time data of the process parameters of the blast furnace can be monitored and alarmed, the production of the blast furnace is effectively guided, the stability of the furnace condition of the blast furnace is facilitated, and the economic benefit of the blast furnace is improved.
Detailed Description
The present invention will be described in detail below with reference to specific embodiments shown in the drawings. These embodiments are not intended to limit the present invention, and structural, methodological, or functional changes made by those skilled in the art according to these embodiments are included in the scope of the present invention.
The blast furnace process parameters comprise blast furnace operation process operation parameters, blast furnace cooling system monitoring parameters, blast furnace raw material parameters, blast furnace distribution matrix parameters, blast furnace blanking parameters, furnace top gas temperature parameters, blast furnace gas composition parameters, molten iron weight, quality and temperature parameters, and slag weight and quality parameters. The blast furnace operation process operation parameters comprise theoretical combustion temperature of a tuyere zone, blast kinetic energy, a furnace belly coal gas index, a ventilation resistance coefficient, wind speed of the tuyere zone, wind quantity of the tuyere zone, wind temperature of the tuyere zone, wind pressure of the tuyere zone, humidification quantity, oxygen-rich quantity, coal injection quantity and the like. The monitoring parameters of the blast furnace cooling system comprise cooling wall temperature, cooling system flow, cooling water pressure, cooling water temperature and the like. The blast furnace raw material parameters comprise the quality, bin position, batching structure and the like of coke, sinter, lump ore and pellets used by the blast furnace. The furnace top gas temperature parameters comprise furnace top gas temperature, furnace top gas pressure, cross temperature measurement temperature, furnace top Z/W and the like.
From historical data, it can be seen that for so many blast furnace process parameters, there is little linear relationship, almost non-linear relationship, or even disorder, and analysis of these data using various statistical methods cannot simplify the relationship between these blast furnace process parameters. Therefore, the traditional blast furnace process parameter range can be established only by the blast furnace operation experience, and scientific basis and data support are lacked.
After long-time research, the inventor invents an interval analysis method, which can linearize the data of the nonlinear relations of the blast furnace process parameters, even disordered data, thereby simplifying the relation among the blast furnace process parameters and providing a scientific means for setting the range of the blast furnace process parameters.
As shown in fig. 1, the interval analysis method includes the steps of:
step S110: and acquiring sample data of a plurality of parameters at different time points, and performing interval division on the fluctuation range of the sample data of the first parameter.
For the convenience of division, it is preferable to divide the fluctuation range of the sample data of the first parameter into intervals by means of average division.
The number of intervals may be many or few, but since the average value of each interval is subjected to linear regression subsequently, the number of the divided intervals is preferably 6 to 8, if the sample data size is large, the number of the divided intervals may be 8, if the sample data size is small, the number of the divided intervals may be 6, and so on.
In addition, after the interval division is performed, the sample size of some intervals may be small, and the subsequent processing is not helpful, so in a preferred embodiment, after the fluctuation range of the sample data of the first parameter is divided into a plurality of intervals, the total sample size of the first parameter and the sample size in each interval are counted, and the sample size ratio of each interval is calculated. And deleting the interval with the sample volume ratio less than the preset threshold value to obtain the finally divided interval. The predetermined threshold may be 5%, that is, when the sample size of a certain interval is less than 5% of the total sample size, the interval is deleted or removed, and the data in the interval does not enter the subsequent processing.
Step S120: and according to the time corresponding relation between other parameters and the first parameter, carrying out same interval division on the sample data of all other parameters, and calculating the average value of each parameter in each interval.
For example, sample data of a first parameter is divided into M intervals, the first interval includes four sample data of the first parameter at time points A, B, C and D, and according to the time correspondence relationship between other parameters and the first parameter, sample data of other parameters at corresponding time points A, B, C and D are also divided into the first interval, and so on. In this way, the sample data of the other parameters is also divided into M sections having the same correspondence relationship as the first parameter.
After the interval division is finished, calculating the average value of each parameter in each interval, including the average value of the first parameter in M intervals, and the average value of each other parameter in M intervals.
Step S130: and respectively taking the average value of the first parameter and other parameters in each interval as coordinate values of two coordinate axes, and respectively calculating the linear regression relationship of the first parameter and other parameters.
The two coordinate axes may be a horizontal axis and a vertical axis, and a linear regression relationship between the first parameter and one of the other parameters is calculated by taking an average value of the first parameter in each interval as a coordinate value of the horizontal axis/the vertical axis and taking an average value of the one of the other parameters in each interval as a coordinate value of the vertical axis/the horizontal axis, respectively.
All other parameters are processed in the same manner to obtain a plurality of linear regression relationships of the first parameter to all other parameters.
It should be noted that, when the linear regression relationship between the parameters is analyzed by using the interval analysis method, the data of all the parameters involved in the analysis are acquired, and the time correspondence relationship exists, and for the blast furnace process parameters, many times we cannot accurately know that the parameter data of the raw materials reacted in the blast furnace, that is, the data of the raw materials and the acquired data of the blast furnace condition have no time correspondence relationship, so that the blast furnace process parameters need to be sorted, the time correspondence relationship is established for the sorted parameters, and then the blast furnace database is established for the acquired data according to the time correspondence relationship.
Specifically, the blast furnace process parameters are sorted, and all the blast furnace process parameters are divided into input parameters, process parameters and output parameters. Wherein:
the input parameters refer to raw material parameters, including quality parameters, bin position parameters, burden structure parameters and the like of coke, sinter, lump ore and pellets used by a blast furnace, and are shown in the following table 1.
The process parameters include operating parameters, furnace condition characterization parameters, and furnace management parameters, as shown in table 2 below.
The output parameters refer to the technical and economic index parameters of the blast furnace and the like, including yield, fuel ratio and the like, and are shown in the following table 3.
TABLE 1
TABLE 2
TABLE 3
As can be seen from tables 1 to 3, the process parameters and the output parameters are collected at the same time, or can be calculated according to data collected at the same time, and only if the input parameters are not collected at the same time, the corresponding relationship between the input parameters and the process parameters and the time of the output parameters needs to be established.
The time corresponding relation of the input parameters, the process parameters and the output parameters of the blast furnace is calculated or obtained through a tracer test by dynamically monitoring the inspection and test data of the raw materials, the time to the factory, the arrival quantity, the change of the position of the finished product bin, the belt transfer speed and the belt transfer quantity from the finished product bin to the blast furnace raw material bin, the position of the blast furnace raw material bin, the transfer speed and the transfer quantity after the blast furnace raw materials are loaded, the smelting period of the blast furnace raw materials in the blast furnace and the like.
Specifically, the raw material quality parameters (including the quality parameters of coke, sintered ore, pellet and ore lump) of the input parameters and the process parameters or the output parameters have a time difference, wherein the time difference is the reaction time in the furnace-the sampling time of the blast furnace raw material, the belt transfer time from a finished product bin to a blast furnace raw material bin after the blast furnace raw material is sampled, the storage time of the blast furnace raw material in the blast furnace raw material bin, the transfer time after the blast furnace raw material is loaded, and the smelting period of the blast furnace raw material in the blast furnace.
In a specific embodiment, a time correspondence of the coke quality parameter of the input parameters and the process parameters is established. Sampling time T for collecting cokeGetBelt transit time delta t of sampling point to blast furnace coke binCoke (coke)Collecting the coke bin of the blast furnace at TGet+ΔtCoke (coke)The storage capacity H at the moment, the charging speed V of blast furnace coke and the charging transit time delta t of the blast furnaceFurnace with a heat exchangerCollecting the smelting period delta t of furnace charge in the blast furnaceSmelting. Acquisition time T of process parametersFurnace with a heat exchangerThereby determining the time correspondence of the coke quality parameter and the process parameter as follows:
Tfurnace with a heat exchanger=TGet+ΔtCoke (coke)+H/V+ΔtFurnace with a heat exchanger+TSmelting。
After the time corresponding relation among the input parameters, the process parameters and the output parameters is established, a blast furnace database is established according to the time corresponding relation of the input parameters, the process parameters and the output parameters, and the collected blast furnace process parameter data. And then analyzing the data of each parameter in the blast furnace database by using an interval analysis method to obtain a linear regression relationship among the blast furnace process parameters.
It should be noted that the collected data of the blast furnace process parameters may be all data collected in a certain period of time, such as in the last two years. For the collected data of the blast furnace process parameters, after the blast furnace database is established according to the time correspondence, the data in the blast furnace database needs to be cleaned, mined and fused, and then the fused data is used for data analysis, monitoring and alarming, for example, the data in the blast furnace database is used for analysis by using an interval analysis method or a normalized interval analysis method, and the data in the blast furnace database is used in the whole text and refers to the fused data in the blast furnace database.
The data cleaning is to remove abnormal bad point data and supplement missing data. Such as data cleaning of the thermocouple temperature of the cooling wall and elimination of bad data. And data which are not in a reasonable fluctuation range are rejected according to different heights and different materials of each layer of cooling wall in the furnace body and different temperature fluctuation ranges during normal production. For example, 13 sections of cast iron cooling walls on the upper part of a furnace body are protected by cooling water, the temperature of the cast iron cooling walls is generally 70-300 ℃, the thermoelectric even data outside 70-300 ℃ are firstly removed, and finally, for the data within 70-300 ℃, if a certain point does not fluctuate or change in one day, the thermocouple of the monitoring point is considered to be damaged, the temperature data of the monitoring point is removed, and the bad data of the blast furnace thermocouple is left after being removed, so that the fault of furnace condition judgment caused by data distortion is avoided. And for the detection test data, carrying out abnormal data point elimination according to whether the detection test data is in a normal detection range. And judging whether missing data exists according to the test frequency, automatically filling the missing data, and filling the average test data of nearly three times.
Data mining refers to statistical analysis of parameter data on the basis of data acquisition, such as statistical average, maximum, minimum, data distribution, standard deviation and the like. Meanwhile, data mining also comprises mining data of indirect parameters, wherein the indirect parameters are parameter data which cannot be directly obtained by collecting data and are obtained by calculating through an existing formula. For example, blast kinetic energy of a blast furnace, activity index of a hearth, ore-coke ratio radial distribution of distribution, heat balance, theoretical combustion temperature and the like are indirect parameters.
The data fusion refers to unifying the data frequency or data period of all parameters to obtain periodic data. Because the data acquisition frequencies of the blast furnace process parameters are different, for example, some parameters are acquired once per second, some parameters are acquired once per minute, some parameters are acquired once per hour, and some parameters are acquired once per even per day, the parameter data of the different data acquisition frequencies need to be subjected to data fusion, and the data frequencies or data periods of all parameters are unified to obtain periodic data. For example, the data frequency of all the parameters is unified to be one hour and the data period is one hour. Because the data volume of the blast furnace is relatively large and the whole period is relatively long, the preferred data frequency is one data per day, namely the data period is day. The method for obtaining the periodic data of one parameter comprises the following steps: the average value or the latest value of all data of the parameter in the data period is obtained as one period data of the parameter. The data of a certain parameter in the blast furnace database is used later, and the data refers to the period data of the parameter.
As shown in fig. 2, the present invention provides a method for setting a range of a blast furnace process parameter, wherein the method uses an interval analysis method to analyze data of the blast furnace process parameter, and a reasonable range is established for the blast furnace process parameter. The method comprises the following steps:
step S210: data of one of the blast furnace process parameters and its associated parameter is obtained.
And acquiring a parameter to be analyzed and data of a parameter having correlation with the parameter from the blast furnace database established with the time correspondence.
The correlation parameter is a parameter having a correlation with the parameter, and may be considered as all other blast furnace process parameters except the parameter, which are correlated because they are blast furnace process parameters. Of course, the correlation parameter may be considered as some other blast furnace process parameters having strong correlation with the parameter, and in this case, the method of obtaining the parameter having correlation with the parameter may be determined by historical experience or obtained by correlation analysis.
Correlation analysis is prior art, i.e. the correlation between two parameters can be analyzed by data. The method for obtaining the correlation parameter of one parameter by using the correlation analysis comprises the following steps:
and respectively analyzing the correlation between the parameter and other blast furnace process parameters through data, and selecting other blast furnace process parameters which are ranked N before the correlation of the parameter according to the strength of the correlation as the parameters having the correlation with the parameter.
Since the output parameter includes a technical economic indicator of the blast furnace, it is preferable that the correlation parameter includes the output parameter.
In a preferred embodiment, the correlation parameter comprises the yield of the blast furnace, and a reasonable range of the parameter is obtained by combining a linear regression relation between the parameter and the yield and the target index of the known yield.
In another preferred embodiment, the correlation parameter comprises the fuel ratio of the blast furnace, and a reasonable range of the parameter is obtained by a linear regression relationship of the parameter and the fuel ratio in combination with a known target index of the fuel ratio.
Step S220: and analyzing the parameters and the correlation parameters by using an interval analysis method to obtain a linear regression relationship between the parameters and each correlation parameter.
And carrying out interval division on the data of the parameters according to the fluctuation range. And then, carrying out the same interval division on the data of all the correlation parameters according to the time corresponding relation between the correlation parameters and the parameters. And after the interval division is finished, calculating the average value of the parameters in each interval, and simultaneously calculating the average value of each correlation parameter in each interval.
Taking the average value of the parameter in each interval as a coordinate value of a coordinate axis (such as a horizontal axis), taking the average value of a correlation parameter in each interval as a coordinate value of another coordinate value (such as a vertical axis), and calculating the linear regression relationship between the parameter and the correlation parameter. Using the same method, a linear regression relationship of the parameter with other correlation parameters was calculated.
Step S230: and obtaining a reasonable range of the parameters by combining known target indexes of one or more correlation parameters according to the linear regression relationship.
The known target index refers to the existing target range or target attribute of the parameter, for example, the target range of the output for a certain blast furnace is between 13500-. For another example, in the target range of the yield, it is considered that the higher the yield is, the better the yield is, the target property is, that is, the known target index is.
In a specific embodiment, the data in the blast furnace database is divided into M intervals according to the fluctuation range of the coke M40, then the average value of all the parameters in each class is calculated, the average value of the coke M40 is used as the abscissa, the average value of other blast furnace process parameters is used as the ordinate, the rule (linear regression relationship) of the coke M40 and other blast furnace process parameters is obtained, and then the reasonable range of the coke M40 is determined according to the known target indexes of one or more other blast furnace process parameters. Taking the linear regression relationship between coke M40 and yield as an example, as shown in fig. 3, when M40 exceeds 89.5, yield does not increase but decreases, and a reasonable range of coke M40 was determined to be [87.5,89.5] based on a known target index of yield (higher yield is better).
In another specific embodiment, the data in the blast furnace database is divided into 8 intervals according to the fluctuation range of the yield, for selecting important blast furnace process parameters (hereinafter referred to as important parameters) such as coke quality, sinter quality, blast kinetic energy and the like, the data of the important parameters are divided into the same 8 intervals according to the time corresponding relation with the yield, the yield and the average value of the important parameters in each interval are calculated, then the average value of the important parameters is taken as an abscissa, and the average value of the yield is taken as an ordinate, so as to obtain the rule (linear regression relation) between the yield and the important parameters. Then, according to the known target index of the yield, a reasonable range of the important parameters is determined. As shown in fig. 4, the linear regression relationship between the blowing kinetic energy PI and the output Ke satisfies the following relationship:
Ke=1.522×PI-10335。
as can be seen from FIG. 4, when the yield is between 13500-14500t/d (known target index of yield), the reasonable range of the blast kinetic energy is between 15600-16300J/s.
In a preferred embodiment, the method further comprises: and obtaining the alarm range of the parameters by combining with the known lower limit indexes of one or more correlation parameters according to the linear regression relationship.
The known lower limit index is a known lower limit range of a parameter in which data of the parameter is determined to be abnormal or the blast furnace is determined to be abnormal. For example, if the lower limit range is less than 11000t/d for the blast furnace output, it is determined that the blast furnace is abnormal when the output is within the lower limit range.
With continued reference to fig. 3 and 4, the alarm range for coke M40 may be obtained by a linear regression relationship between coke M40 and production, in combination with a known lower limit indicator for production. Similarly, the alarm range of the blowing kinetic energy can be obtained by combining the known lower limit index of the yield through the linear regression relationship between the blowing kinetic energy and the yield.
According to the setting method of the blast furnace process parameter range, the reasonable ranges and the alarm ranges of all blast furnace process parameters can be set, and the following table 4 shows the reasonable ranges of part of important blast furnace operation parameters and raw material quality parameters obtained by the method.
TABLE 4
By setting the reasonable range and the alarm range of the blast furnace process parameters, the real-time data of the blast furnace process parameters can be monitored and alarmed, the production of the blast furnace is effectively guided, the stability of the blast furnace condition is facilitated, and the economic benefit of the blast furnace is improved.
The invention also provides an electronic device, which comprises a memory and a processor, wherein the memory stores a computer program capable of running on the processor, and the processor realizes any one step of the setting method of the blast furnace process parameter range when executing the program, namely realizes the step of any one technical scheme of the setting method of the blast furnace process parameter range.
The present invention also provides a computer-readable storage medium, on which a computer program is stored, which, when executed by a processor, implements any one of the steps in the method for setting a range of a process parameter of a blast furnace described above, that is, implements any one of the steps in the method for setting a range of a process parameter of a blast furnace described above.
It should be understood that although the present description refers to embodiments, not every embodiment contains only a single technical solution, and such description is for clarity only, and those skilled in the art should make the description as a whole, and the technical solutions in the embodiments can also be combined appropriately to form other embodiments understood by those skilled in the art.
The above-listed detailed description is only a specific description of a possible embodiment of the present invention, and they are not intended to limit the scope of the present invention, and equivalent embodiments or modifications made without departing from the technical spirit of the present invention should be included in the scope of the present invention.