A sintering endpoint intelligent control method based on multi-objective optimization model

Abstract

The present invention provides a sintering endpoint intelligent control method based on a multi-objective optimization model, S1: calculate disturbance values ​​according to actual needs respectively; the disturbance values ​​have five items; S2: use the disturbance values ​​calculated in step S1 as variable values ​​of the prediction model to obtain a sintering endpoint position prediction model and compare them; S3: compare the endpoint position prediction model in step S2, and determine whether the endpoint position prediction model is still in the predicted endpoint area according to the actual situation changes in step S1; if it is still in the predetermined endpoint position area, reset the start command; if it is not in the predetermined endpoint position area, judge again according to the weight, and finally reset the start command after adjustment, perform cyclic calculation, and realize dynamic control of the sintering machine endpoint. This method is more automated and intelligent than the traditional sintering production endpoint control, and can accurately control the sintering endpoint position of the sintering machine according to different production tasks, ensuring the stability of sintering production.

Description

Technical Field

[0001] The present invention relates to the field of sintering endpoint control of a sintering machine in the iron and steel metallurgical industry, and in particular to a sintering endpoint intelligent control method based on a multi-objective optimization model. Background Art

[0002] The sintering production process is an important part of the iron and steel metallurgical industry. Its main task is to mix various raw materials such as ore powder, fuel, flux, return ore, etc. with water to produce qualified sintered ore for subsequent blast furnace ironmaking production. The main supporting equipment for the sintering process includes raw material feeding, mixing, material distribution, sintering machine, main exhaust fan, etc. The control process of the sintering endpoint is very complicated and is affected by many factors such as raw material conditions and production conditions.

[0003] Most of the existing endpoint control methods for sintering machines use single-target control. When there are too many changing parameters during use, manual intervention by operators is required, and automatic adjustment cannot be achieved according to changes in production conditions and tasks. Especially when raw material conditions change, frequent manual adjustments are required to achieve variable production conditions. Therefore, it is very necessary to develop a multi-objective optimization model based on different production tasks to achieve intelligent control of sintering endpoints. Summary of the invention

[0004] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a sintering endpoint intelligent control method based on a multi-objective optimization model.

[0005] To achieve the above purpose, the present invention adopts the following technical solution: a sintering endpoint intelligent control method based on a multi-objective optimization model: the specific steps of the method are as follows:

[0006] S1: Calculate the disturbance value according to actual needs;

[0007] There are five disturbance values, namely, the disturbance value of batching raw material change R1, the disturbance value of material layer permeability R2, the disturbance value of flue temperature R3, the disturbance value of ignition intensity R4 and the disturbance value of mixture moisture change R5;

[0008] When the market price of raw materials changes, the purchased raw material conditions change significantly, and the disturbance value R1 of the raw material change and the permeability disturbance value R2 of the material layer are obtained by calculation;

[0009] When the product is unsalable, the output of sintered ore needs to be reduced. At this time, the thickness of the material layer needs to be reduced, and the ignition intensity needs to be reduced. The flue temperature disturbance value R3 and the ignition intensity disturbance value R4 are obtained by calculation;

[0010] When the raw material working conditions change and the quality of the sintered ore products needs to be guaranteed, the material layer permeability disturbance value R2, flue temperature disturbance value R3, ignition intensity disturbance value R4 and mixture moisture change disturbance value R5 are obtained by calculation.

[0011] S2: Using the disturbance values ​​calculated in step S1 as variable values ​​of the prediction model, obtaining a sintering endpoint position prediction model, and performing comparison.

[0012] S3: Compare the terminal position prediction model in step S2, and determine whether the terminal position prediction model is still in the predicted terminal area according to the actual situation change in step S1;

[0013] If it is still in the predetermined end position area, reset the start command;

[0014] If it is not in the predetermined end position area, it will be judged again according to the weight, which includes the sintering machine speed adjustment, the main pumping frequency change and the adjustment of the material layer thickness;

[0015] Finally, in step S3, after all the endpoint position prediction models that are not in the predicted endpoint area have been adjusted, the start command is reset to end the prediction of the model.

[0016] Preferably, in step S2, the sintering end point position prediction model function is as follows:

[0017] Z=f(R1, R2, R3, R4, R5), the specific expression of function f is calculated based on production data;

[0018] The general expression of function f is: Z=k1*R1+k2*R2+k3*R3+k4*R4+k5*R5, where K1, k2, k3, k4, k5 are coefficients, which are calculated by the least quadratic method based on historical production data.

[0019] Preferably, in step S1, the calculation method of the batching raw material change disturbance value is as follows: raw material change disturbance value R1=[1-(actual alkalinity-reference alkalinity value) / reference alkalinity value]*[1-|actual coke powder amount-reference coke powder amount| / reference coke powder amount].

[0020] Generally, the benchmark alkalinity value and benchmark coke powder of each manufacturing enterprise are production-set values. The benchmark alkalinity value is between 1.8 and 2.0, and the benchmark coke powder content is between 4 and 6%.

[0021] Preferably, in step S1, the flue temperature disturbance value is calculated as follows:

[0022] Flue temperature disturbance value R2 = [1-|actual temperature-reference temperature| / reference temperature]. Generally, the flue reference temperature of each production enterprise is a production fixed value, and the reference temperature value is 110~140℃.

[0023] Preferably, in step S1, the calculation method of the ignition intensity disturbance value is as follows: ignition intensity disturbance value R3 = 1 - |actual temperature - reference temperature| / reference temperature. Generally, the reference temperature of the flue of each production enterprise is a production fixed value, and the reference temperature value is 1100°C.

[0024] Preferably, in step S1, the calculation method of the material layer air permeability disturbance value is as follows: material layer air permeability disturbance value R4 ;

[0025] Where: Q is the air volume of the sintering machine; A is the area of ​​the sintering machine; H is the height of the sintering material, ∆p is the material layer resistance, and n is the coefficient, which is generally taken as 0.6.

[0026] Preferably, in step S1, the calculation method of the disturbance value of the mixture moisture change is as follows:

[0027] Moisture disturbance R5 = reference value - total water supply / total material weight * correction factor. Generally, the moisture reference value of each production enterprise is a production fixed value, which is about 7.2%. The total water supply is read from the water pipe meter, and the total material weight is read from the belt scale. The correction factor is calculated by fitting the least quadratic method based on the production history data.

[0028] Preferably, in step S3, the sintering endpoint position prediction model is used to calculate and compare the target position to determine whether the difference should be adjusted, and if necessary, adjustment is made;

[0029] According to the speed adjustment amount of the sintering machine, determine whether the speed adjustment amount of the sintering machine exceeds the adjustment range, if so, reset the start command; otherwise, adjust the speed of the sintering machine and then reset the start command;

[0030] According to the change of the main pumping frequency, it is judged whether the adjustment amount of the main pumping frequency exceeds the range, and if so, the start command is reset; otherwise, the main pumping frequency is adjusted, and then the start command is reset;

[0031] According to the adjusted material layer thickness, determine whether the material layer thickness exceeds the range, if so, reset the start command; otherwise, adjust the roller speed and then reset the start command;

[0032] When the sintering machine endpoint control is no longer performed, exit this prediction model and end.

[0033] Compared with the prior art, the beneficial effects of the present invention are as follows: this method is more automated and intelligent than traditional sintering production endpoint control. While reducing the workload of operators, it can accurately control the sintering endpoint position of the sintering machine according to different production tasks to ensure the stability of sintering production; the sintering endpoint position prediction model has a built-in machine self-learning function. By accumulating learning and adjustment experience, and optimizing the adjustment of the sintering machine speed adjustment amount, the main pumping frequency change amount, and the adjustment of the material layer thickness, an optimal production database under different target conditions is established to achieve the goal of rapid response and production adjustment. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 Schematic diagram of the overall process of this method. Implementation

[0035] In order to provide a further understanding of the purpose, structure, features, and functions of the present invention, the following detailed description is given in conjunction with the embodiments.

[0036] Please refer to Figure 1 The present invention provides a sintering endpoint intelligent control method based on a multi-objective optimization model.

[0037] The production process of the sintering machine needs to be adjusted according to the actual situation of the enterprise. When the enterprise is operating at a loss, it is necessary to control the production cost. At this time, the raw material conditions of the sintering machine become worse. The production task is mainly to produce qualified sintered ore under low-cost raw material conditions. The main goal is low cost and high quality. At this time, it is necessary to reduce the thickness of the sintering material layer, reduce the frequency of the main exhaust fan, and reduce the ignition intensity.

[0038] When the steel market demand is strong and production is profitable, the production task of the sintering machine will shift to output as the main goal, and the control target of production cost will be weakened. At this time, nearly full-load production is required, the material layer thickness is close to the maximum value, the main exhaust fan frequency is increased, and the ignition intensity is improved.

[0039] When product sales are average, quality control is the goal. At this time, the production of the sintering machine is mainly based on stability to avoid fluctuations in product quality indicators.

[0040] When raw material conditions or operating conditions fluctuate abnormally, it is necessary to adjust the operating parameters of the sintering machine and supporting equipment in a timely manner to adapt to the adverse effects of the fluctuations.

[0041] Based on the above problems, the present invention discloses a sintering endpoint intelligent control method based on a multi-objective optimization model: when the raw material market price changes and the purchased raw material conditions change significantly, the model calculates the raw material change disturbance value R1 and the material layer permeability disturbance value R2, and compares the calculation of the sintering endpoint position prediction model with the target position to determine whether the difference needs to be adjusted. If adjustment is required, the sintering machine speed, main exhaust fan frequency, and material layer thickness are calculated according to the weight value to determine whether they are within the adjustment range. If they are within the range, an adjustment instruction is issued, otherwise it returns to recalculate;

[0042] When the product is unsalable, the output of sintered ore needs to be reduced. At this time, the thickness of the material layer needs to be reduced and the ignition intensity needs to be reduced. The model calculates the flue temperature disturbance value R3 and the ignition intensity disturbance value R4. The sintering end point position prediction model calculates and compares the target position to determine whether the difference needs to be adjusted. If adjustment is required, the sintering machine speed, main exhaust fan frequency, and material layer thickness are calculated according to the weights to determine whether they are within the adjustment range. If they are within the range, an adjustment command is issued, otherwise it returns to recalculate.

[0043] When the raw material working conditions change, the quality of sintered ore products needs to be guaranteed. The model calculates the material layer permeability disturbance value R2, flue temperature disturbance value R3, ignition intensity disturbance value R4 and mixture moisture change disturbance value R5. The sintering end point position prediction calculation model is compared with the target position to determine whether the difference needs to be adjusted. If adjustment is required, the sintering machine speed, main exhaust fan frequency and material layer thickness are calculated according to the weight value to determine whether they are within the adjustment range. If they are within the range, an adjustment command is issued, otherwise it returns to recalculate.

[0044] In addition, the sintering endpoint position prediction model has a built-in machine self-learning function. By accumulating learning and adjustment experience, and optimizing the weighted sintering machine speed adjustment amount, main pumping frequency change amount, and material layer thickness adjustment, an optimal production database under different target conditions is established to achieve the goal of rapid response and production adjustment.

[0045] The present invention has been described by the above-mentioned relevant embodiments, however, the above-mentioned embodiments are only examples for implementing the present invention. It must be pointed out that the disclosed embodiments do not limit the scope of the present invention. On the contrary, changes and modifications made without departing from the spirit and scope of the present invention are all within the scope of patent protection of the present invention.

Claims

1. A sintering endpoint intelligent control method based on a multi-objective optimization model, characterized by: The specific steps of the method are as follows: S1: Calculate the disturbance value according to actual needs; There are five disturbance values, namely, the disturbance value of batching raw material change R1, the disturbance value of material layer permeability R2, the disturbance value of flue temperature R3, the disturbance value of ignition intensity R4 and the disturbance value of mixture moisture change R5; Raw material change disturbance value R1=[1-(actual alkalinity-reference alkalinity value) / reference alkalinity value]*[1-|actual coke powder amount-reference coke powder amount| / reference coke powder amount]; The calculation method of the permeability disturbance value of the material layer is as follows: permeability disturbance value of the material layer ; Wherein, Q is the air volume of the sintering machine; A is the area of ​​the sintering machine; H is the height of the sintering material, ∆p is the material layer resistance, and n is the coefficient; The calculation method of the flue temperature disturbance value is as follows: flue temperature disturbance value R2=[1-|actual temperature-reference temperature| / reference temperature]; The calculation method of the ignition intensity disturbance value is as follows: ignition intensity disturbance value R3=1-|measured temperature-reference temperature| / reference temperature; The calculation method of the disturbance value of the moisture change of the mixture is as follows: moisture disturbance R5 = reference value - total water supply / total material weight * correction factor; the correction factor is calculated by fitting the least quadratic method based on the production history data; S2: using the disturbance values ​​calculated in step S1 as variable values ​​of the prediction model to obtain a sintering endpoint position prediction model, and making a comparison; S3: Compare the terminal position prediction model in step S2, and determine whether the terminal position prediction model is still in the predicted terminal area according to the actual situation change in step S1; If it is still in the predetermined end position area, reset the start command; If it is not in the predetermined end position area, it will be judged again according to the weight, which includes the sintering machine speed adjustment amount, the main pumping frequency change amount and the adjustment of the material layer thickness.

2. The sintering endpoint intelligent control method based on a multi-objective optimization model according to claim 1, characterized in that: In the step S2, the sintering endpoint position prediction model function is as follows: Z=f(R1, R2, R3, R4, R5), and the specific expression of function f is calculated based on production data.

3. The sintering endpoint intelligent control method based on a multi-objective optimization model according to claim 1, characterized in that: In step S3, the sintering endpoint position prediction model is used to calculate and compare the target position to determine whether the difference is adjusted.

Images