A method for constructing a blast furnace charge trajectory model
By constructing an initial charge trajectory model and optimizing parameters using a genetic algorithm, the problem of time-consuming charge surface modeling in blast furnace production was solved, achieving efficient and accurate charge surface distribution calculation and simplifying the operation process.
Patent Information
- Application Number
- CN202111244739.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-10-26
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2041-10-26
AI Technical Summary
In existing technologies for blast furnace production, the process of modeling the material surface is cumbersome and time-consuming, making it difficult to accurately describe the continuous distribution of the material surface. Furthermore, radar equipment has a limited lifespan and is inconvenient to maintain, which affects the normal operation of the blast furnace.
By constructing an initial furnace charge trajectory model formula, combining it with a genetic algorithm to optimize the model parameters, and continuously adjusting and optimizing the furnace charge trajectory model using radar scan data, we can obtain real material surface shape data and achieve efficient calculation of simulated material surface shape.
It simplifies the modeling process, improves the accuracy and efficiency of the model, and can directly calculate the shape of the material surface after it is laid without relying on radar, thus improving the accuracy of the material surface distribution.
Smart Images

Figure CN114139438B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of blast furnace ironmaking, and in particular to a method for constructing a blast furnace charge trajectory model. Background Art
[0002] During the blast furnace production process, the charge enters from the silo at the top of the blast furnace, passes through the flow valve, fork pipe, throat pipe, chute and finally falls to the throat. Then, as the charge is consumed, the charge in the throat gradually falls to the furnace body to complete the conversion process. The shape of the charge surface is crucial to the normal operation of the blast furnace. A good charge surface can make the gas flow distribution more reasonable and effectively fill the consumption of charge in different parts.
[0003] The traditional manual method is to use a probe to measure the depth of several parts, and then make a rough estimate of the material surface by combining the gas composition, furnace top temperature, etc. This requires the operator to have rich experience in material distribution, and it is difficult to accurately describe the continuous distribution of the material surface. With the development of measurement equipment, the material surface distribution of the furnace throat can be directly measured by radar and other technical means, but in actual applications, the service life of the radar is limited, and it is inconvenient to replace and repair it. Moreover, most radars can only measure one point at a time. It takes dozens or hundreds of measurements to complete the scanning of the entire material surface. If you need to scan before and after the material is laid, it will take a long time. Summary of the invention
[0004] In order to solve the problem that the above-mentioned existing modeling methods are cumbersome and time-consuming, the present invention provides a method for constructing a blast furnace charge trajectory model, comprising the following steps:
[0005] S1, modeling the blast furnace charge descending process to obtain an initial charge trajectory model formula, the initial charge trajectory model formula is as follows:
[0006]
[0007] H=vtcosα+0.5gt 2 ;
[0008] R = vtsinα;
[0009] Among them, v is the speed of the charge when it slides out of the chute, α is the inclination angle of the chute, r is the hydraulic radius of the material flow valve, H is the distance from the chute outlet to the material surface, t is the time the charge moves in the air, R is the lateral distance from the charge landing point to the chute outlet, and a, b, c, d, and e are model parameters;
[0010] S2, obtaining a material surface data set, and inputting the material surface data set into the initial charge trajectory model formula to calculate a set of charge landing positions in the furnace throat, and obtaining simulated material surface shape data according to the material surface data set and the set of charge landing positions in the furnace throat;
[0011] S3, obtaining real material surface shape data, optimizing the model parameters in the initial charge trajectory model formula through a genetic algorithm according to the real material surface shape data and the simulated material surface shape data, iteratively updating the model parameters in the initial charge trajectory model formula, and obtaining a charge trajectory model.
[0012] Based on the above solution, the present invention can also make the following improvements:
[0013] Furthermore, the step of acquiring the real material surface shape data in S3 is specifically to detect the real material surface shape data of the furnace top material surface by radar scanning, and the real material surface shape data is a discrete point vector S of a fixed length.
[0014] Further, in S3, the optimization operation of the model parameters in the initial charge trajectory model formula is performed by a genetic algorithm according to the real charge surface shape data and the simulated charge surface shape data, and the iterative update of the model parameters in the initial charge trajectory model formula is specifically as follows:
[0015] S301, setting the lower limit and the upper limit of the solution space of the model parameters, constructing an initial population by a chromosome encoding method, inputting the population individuals in the initial population into the initial charge trajectory model formula to obtain simulated material surface shape data of the population individuals, wherein the simulated material surface shape data is a discrete point vector C of a fixed length;
[0016] S302, interpolate or eliminate the discrete point vector S to make the length of the discrete point vector S of the real material surface shape data the same as the length of the discrete point vector C of the simulated material surface shape data, and input the discrete point vectors S and C into the fitness function to obtain the fitness of the individuals in the population. The fitness function is the inverse of the two norms of the discrete point vectors. The fitness function f is specifically:
[0017]
[0018] S303, sorting the fitness of the individuals in the population in descending order, and retaining the first N individuals in the population;
[0019] S304, performing a crossover mutation operation on the retained population individuals to obtain the next generation population, inputting the population individuals in the current population into the initial charge trajectory model formula to obtain the simulated material surface shape data of the population individuals, and repeating the above S303 and S304 operations until the number of iterations is equal to the preset number of iterations to enter step S305;
[0020] S305 , selecting the population individual with the greatest fitness to iteratively update the model parameters in the initial charge trajectory model formula to obtain a charge trajectory model.
[0021] Furthermore, the construction of the initial population by the chromosome encoding method in S301 is specifically as follows: the chromosome encoding method adopts floating point encoding, and the population individuals of the initial population include model parameters a, b, c, d, and e.
[0022] Furthermore, in S304, the crossover mutation operation is performed on the retained individuals of the population to obtain the next generation population operation. Specifically, the retained individuals of the population are subjected to gene crossover processing through a crossover operator, and genes of two paired individuals of the population are exchanged with each other through a single-point crossover method to obtain individuals of the next generation population; the retained individuals of the population are subjected to gene mutation processing through a mutation operator, and after the gene mutation position in the individuals of the population is determined, a mutation operation is performed through a basic bit mutation method, and the original gene at the gene mutation position is inverted according to a preset probability.
[0023] Another object of the present invention is to provide a computer system, which is configured to perform the following steps: first, model the blast furnace charge descending process to obtain an initial charge trajectory model formula; then, obtain a charge surface data set, and input the charge surface data set into the initial charge trajectory model formula to calculate a set of charge landing positions in the throat part, and obtain simulated charge surface shape data based on the charge surface data set and the set of charge landing positions in the throat part; finally, obtain real charge surface shape data, and optimize the model parameters in the initial charge trajectory model formula through a genetic algorithm based on the real charge surface shape data and the simulated charge surface shape data, and iteratively update the model parameters in the initial charge trajectory model formula to obtain a charge trajectory model.
[0024] Another object of the present invention is to provide a computer-readable storage medium, including a memory, wherein the memory stores a computer program, and when the computer program is executed by a processor, the above-mentioned blast furnace charge trajectory model construction method is implemented.
[0025] The beneficial effects of the present invention are as follows: based on the material surface data scanned by radar, multiple parameters of the charge trajectory model are continuously adjusted and optimized, and finally the most suitable charge trajectory model for the current blast furnace is obtained, so that the shape of the material surface after laying is directly calculated through the material laying matrix without the need for radar, and by simulating the genetic evolution in nature, the optimal model parameters of the charge trajectory model can be efficiently found in the solution space, which not only simplifies the modeling process but also improves the accuracy of the model. BRIEF DESCRIPTION OF THE DRAWINGS
[0026] Figure 1A flowchart of a method for constructing a blast furnace charge trajectory model according to the present invention;
[0027] Figure 2 It is a structural schematic diagram of the throat of a blast furnace of the present invention;
[0028] Figure 3 The present invention is a flow chart of a method for optimizing model parameters of a blast furnace charge trajectory model. DETAILED DESCRIPTION
[0029] The principles and features of the present invention are described below in conjunction with the accompanying drawings. The examples given are only used to explain the present invention and are not used to limit the scope of the present invention.
[0030] like Figure 1 The present invention provides a method for constructing a blast furnace charge trajectory model, comprising the following steps:
[0031] S1, model the blast furnace charge descending process and obtain the initial charge trajectory model formula,
[0032] S2, obtaining a material surface data set, and inputting the material surface data set into the initial charge trajectory model formula to calculate a set of charge landing positions in the furnace throat, and obtaining simulated material surface shape data according to the material surface data set and the set of charge landing positions in the furnace throat;
[0033] S3, obtaining real material surface shape data, optimizing the model parameters in the initial charge trajectory model formula through a genetic algorithm according to the real material surface shape data and the simulated material surface shape data, iteratively updating the model parameters in the initial charge trajectory model formula, and obtaining a charge trajectory model.
[0034] The present invention continuously adjusts and optimizes multiple parameters of the charge trajectory model according to the charge surface data scanned by radar, and finally obtains the charge trajectory model that is most suitable for the current blast furnace. Therefore, the charge surface shape after charge distribution can be directly calculated through the charge distribution matrix without the need for radar. By simulating the genetic evolution in nature, the optimal model parameters of the charge trajectory model can be efficiently found in the solution space, which not only simplifies the modeling process but also improves the accuracy of the model.
[0035] Specifically, in this embodiment, the charge descending process is modeled using the classical mechanics formula to obtain the charge trajectory model formula. Figure 2 As shown, the blast furnace charge descending process is modeled to obtain the initial charge trajectory model formula, which is as follows:
[0036]
[0037] H=vtcosα+0.5gt2 ;
[0038] R = vtsinα;
[0039] Among them, v is the speed of the charge when it slides out of the chute, α is the inclination angle of the chute, r is the hydraulic radius of the material flow valve, H is the distance from the chute outlet to the material surface, t is the time the charge moves in the air, R is the lateral distance from the charge landing point to the chute outlet, and a, b, c, d, and e are model parameters.
[0040] In this embodiment, the initial charge trajectory model formula is used, such as Figure 3 As shown in the figure, after obtaining the material surface data set, the location of the charge at the furnace throat can be calculated. The material surface data set includes the hydraulic radius of the material flow valve, the chute angle and drop height, the total weight, density, and distribution matrix of the material batch. The distribution volume of each gear of the blast furnace can be calculated through the total weight, density, and distribution matrix of the material batch. The shape of the material surface of the gear can be simulated based on the drop point position and distribution volume calculated based on the trajectory model. Repeating the above method can calculate the shape of each gear of the current distribution, and obtain the simulated material surface shape data of the material surface after the current batch of material is distributed.
[0041] Specifically in this embodiment, the real surface shape data of the furnace top material surface is detected by radar scanning, and the real surface shape data is a discrete point vector S of fixed length. Set the evolutionary algebra counter (preset number of iterations) t=0, set the maximum evolutionary algebra T, and randomly generate M individuals as the initial population. If the current number of completed iterations t=T, the individual with the maximum fitness obtained in the entire evolutionary process (all iterations) is selected as the optimal solution output, and the genetic calculation is terminated. When the fitness of the optimal individual reaches a given threshold, or the fitness of the optimal individual and the fitness of the population no longer increase, or the number of iterations reaches a preset algebra, the algorithm terminates. The preset algebra is generally set to 100-500 generations.
[0042] Specifically in this embodiment, the chromosome encoding method adopts floating point encoding, and the population individuals of the initial population include model parameters a, b, c, d, and e. The lower limit and upper limit of the solution space of the model parameters are set, and the lower limit of the solution space of the five parameters is set to {-50,-50,-50,-50,-50}, and the upper limit of the solution space of the parameters is set to {50,50,50,50,50}. The primary population {{a1,b1,c1,d1,e1}, {a2,b2,c2,d2,e2},...} is generated according to the chromosome encoding method, and the population size is 100. The initial population is constructed by the chromosome encoding method, and the population individuals in the initial population are input into the initial charge trajectory model formula to obtain the simulated material surface shape data of the population individuals, wherein the simulated material surface shape data is a discrete point vector C of fixed length. The discrete point vector S is interpolated or eliminated, and the fitness of the simulated material surface shape data is obtained according to the fitness function, so that the length of the discrete point vector S of the real material surface shape data is the same as the length of the discrete point vector C of the simulated material surface shape data. The fitness function is the inverse of the two norms of the discrete point vector, and the fitness function f is specifically:
[0043]
[0044] Specifically in this embodiment, the first N individuals of the population are retained according to the applicability of the simulated material surface shape data of the population individuals in descending order; a crossover mutation operation is performed on the retained population individuals to obtain the next generation population, and after the population individuals in the current population are input into the initial charge trajectory model formula to obtain the simulated material surface shape data of the population individuals, the retained population individuals are subjected to gene crossover processing through a crossover operator, and two paired population individuals exchange genes with each other through a single-point crossover method to form the next generation population individuals; the retained population individuals are subjected to gene mutation processing through a mutation operator, and after the gene mutation position in the population individual is determined, a mutation operation is performed through a basic bit mutation method, and the original gene at the gene mutation position is inverted according to a preset probability.
[0045] The population selects individuals with higher fitness by the fitness ratio method of the selection operation to be retained, and the optimized individuals in the population are directly inherited to the next generation or generated by pairing crossover through the selection operator and then inherited to the next generation. Then, the crossover operator of the single-point crossover is used to select two chromosomes, split them at a randomly selected position point and exchange the right part, so as to obtain two different sub-chromosomes. Under the condition that the single-point crossover can cause less damage, the next generation population is generated by performing mutation and crossover operations on the selected individuals, and the genes of the two paired chromosomes are exchanged in a certain way through the crossover operation, so as to form a new individual. The crossover operation plays a key role in the genetic algorithm, and changing the gene values on certain loci of the individual strings in the population is the main method to generate new individuals. The mutation operation is to change the gene value on a certain locus according to a small probability, which is also an operation method to generate new individuals. The present invention adopts the method of basic bit mutation to perform mutation operation, and its specific operation process is: first determine the gene mutation position of each individual, and then reverse the original gene of the mutation point according to a certain probability. After the population P(t) is selected, crossover and mutated, the next generation population P(t+1) is obtained. The genetic algorithm has a local random search capability. When the genetic algorithm has approached the parameter neighborhood in the charge trajectory model formula through the crossover operator, the local random search capability of the mutation operator can accelerate the convergence to the parameters in the optimal charge trajectory model formula, and can also enable the genetic algorithm to maintain population diversity to prevent premature convergence. The population individual with the largest fitness is selected to iteratively update the model parameters in the initial charge trajectory model formula to obtain the charge trajectory model.
[0046] Another object of the present invention is to provide a computer system, which is configured to perform the following steps: first, model the blast furnace charge descending process to obtain an initial charge trajectory model formula; then, obtain a charge surface data set, and input the charge surface data set into the initial charge trajectory model formula to calculate a set of charge landing positions in the throat part, and obtain simulated charge surface shape data based on the charge surface data set and the set of charge landing positions in the throat part; finally, obtain real charge surface shape data, and optimize the model parameters in the initial charge trajectory model formula through a genetic algorithm based on the real charge surface shape data and the simulated charge surface shape data, and iteratively update the model parameters in the initial charge trajectory model formula to obtain a charge trajectory model.
[0047] Another object of the present invention is to provide a computer-readable storage medium, including a memory, wherein a computer program is stored in the memory, and when the computer program is executed by a processor, the above-mentioned method for constructing a blast furnace charge trajectory model is implemented. The computer device mainly includes a processor, a memory 302 and a bus, wherein the memory stores at least one program, and the program is executed by the processor to implement the configuration of the computer as described above. The processor includes one or more processing cores, the processor is connected to the memory through a bus, the memory is used to store program instructions, and when the processor executes the program instructions in the memory, the method for constructing a blast furnace charge trajectory model provided by the above-mentioned method embodiment is implemented. The memory can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, disk or optical disk.
[0048] The present invention proposes a computer-readable storage medium, in which at least one program is stored. When the at least one program is run, it is executed to realize the configuration of the computer system as described above and / or to execute the method for constructing a blast furnace charge trajectory model as described above.
[0049] The reader should understand that in the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine different embodiments or examples described in this specification and the features of different embodiments or examples without contradiction.
[0050] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the method embodiments described above are only illustrative, for example, the division of steps is only a logical function division, and there may be other division methods in actual implementation, such as multiple steps can be combined or integrated into another step, or some features can be ignored or not executed.
[0051] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.
Claims
1. A method for constructing a blast furnace charge trajectory model, characterized in that: The following steps are included: S1, modeling the blast furnace charge descending process to obtain an initial charge trajectory model formula, the initial charge trajectory model formula is as follows: H=vtcosα+0.5gt 2 ; R = vtsinα; Among them, v is the speed of the charge when it slides out of the chute, α is the inclination angle of the chute, r is the hydraulic radius of the material flow valve, H is the distance from the chute outlet to the material surface, t is the time the charge moves in the air, R is the lateral distance from the charge landing point to the chute outlet, and a, b, c, d, and e are model parameters; S2, obtaining a material surface data set, and inputting the material surface data set into the initial charge trajectory model formula to calculate a set of charge landing positions in the furnace throat, and obtaining simulated material surface shape data according to the material surface data set and the set of charge landing positions in the furnace throat; S3, obtaining real material surface shape data, optimizing the model parameters in the initial charge trajectory model formula through a genetic algorithm according to the real material surface shape data and the simulated material surface shape data, iteratively updating the model parameters in the initial charge trajectory model formula, and obtaining a charge trajectory model.
2. The method for constructing a blast furnace charge trajectory model according to claim 1, characterized in that: The step of acquiring the real material surface shape data in S3 is specifically to detect the real material surface shape data of the furnace top material surface by radar scanning, wherein the real material surface shape data is a discrete point vector S of a fixed length.
3. The method for constructing a blast furnace charge trajectory model according to claim 2, characterized in that: In S3, the model parameters in the initial charge trajectory model formula are optimized by a genetic algorithm according to the real charge surface shape data and the simulated charge surface shape data, and the model parameters in the initial charge trajectory model formula are iteratively updated as follows: S301, setting the lower limit and the upper limit of the solution space of the model parameters, constructing an initial population by a chromosome encoding method, inputting all population individuals in the initial population into the initial charge trajectory model formula to obtain simulated material surface shape data of the population individuals, wherein the simulated material surface shape data is a discrete point vector C of a fixed length; S302, interpolate or eliminate the discrete point vector S to make the length of the discrete point vector S of the real material surface shape data the same as the length of the discrete point vector C of the simulated material surface shape data, and input the discrete point vectors S and C into the fitness function to obtain the fitness of the individuals in the population. The fitness function is the inverse of the two norms of the discrete point vectors. The fitness function f is specifically: S303, sorting the fitness of the individuals in the population in descending order, and retaining the first N individuals in the population; S304, performing a crossover mutation operation on the retained population individuals to obtain the next generation population, inputting the population individuals in the current population into the initial charge trajectory model formula to obtain the simulated material surface shape data of the population individuals, and repeating the above S303 and S304 operations until the number of iterations is equal to the preset number of iterations to enter step S305; S305 , selecting a population individual with the greatest fitness and iteratively updating the model parameters in the initial charge trajectory model formula according to the selected population individual to obtain a charge trajectory model.
4. The method for constructing a blast furnace charge trajectory model according to claim 3, characterized in that: The specific method of constructing the initial population by the chromosome encoding method in S301 is as follows: the chromosome encoding method adopts floating point encoding, and the population individuals of the initial population include model parameters a, b, c, d, and e.
5. The method for constructing a blast furnace charge trajectory model according to claim 4, characterized in that: In S304, the crossover mutation operation is performed on the retained individuals of the population to obtain the next generation population operation. Specifically, the retained individuals of the population are subjected to gene crossover processing through a crossover operator, and two paired individuals of the population are subjected to gene exchange through a single-point crossover method to obtain the next generation population individuals; the retained individuals of the population are subjected to gene mutation processing through a mutation operator, and after the gene mutation position in the individuals of the population is determined, a mutation operation is performed through a basic bit mutation method, and the original gene at the gene mutation position is inverted according to a preset probability.
6. A computer system, characterized in that: The computer system is configured to execute the blast furnace charge trajectory model building method according to any one of claims 1 to 5.
7. A computer-readable storage medium, characterized in that: It includes a memory, in which a computer program is stored. When the computer program is executed by a processor, the blast furnace charge trajectory model construction method according to any one of claims 1 to 5 is implemented.
Citation Information
Patent Citations
Blast furnace burden distribution value simulation method based on intelligent algorithm
CN102629286A
Characterization method, system, equipment and storage medium for blast furnace burden surface shape
CN110551861A
