Intelligent dynamic regulation and control method for air temperature of blast furnace hot blast stove of iron making plant
By combining data acquisition and multi-objective function optimization with Kalman filtering and LSTM network, the response lag and multi-parameter coupling problems of blast furnace hot blast stove temperature regulation were solved, achieving efficient blast temperature control and reducing energy consumption and NOx emissions.
Patent Information
- Application Number
- CN202511310483.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-09-15
- Publication Date
- 2025-11-07
AI Technical Summary
The existing blast furnace hot blast stove temperature regulation relies on manual experience or a single PID control, which has problems such as response lag, poor coupling of multiple parameters, energy waste and equipment wear and tear, and lacks the ability to fuse and analyze multi-source heterogeneous data.
A multi-objective function combining data acquisition, Kalman filtering algorithm, and LSTM network is used to achieve intelligent dynamic adjustment of wind temperature through sensing module, main control unit, and actuator. The Kalman filtering algorithm is used to eliminate noise, the LSTM network predicts future operating conditions, and a multi-objective function is constructed for optimization. Model predictive control is used for fault diagnosis.
It has reduced the air temperature fluctuation range from ±15℃ to ±3℃, reduced energy consumption per ton of iron by 5-12%, reduced NOx emissions by more than 20%, and improved the accuracy and adaptability of air temperature regulation.
Smart Images
Figure CN120905460A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of automatic control of metallurgical industry equipment, and particularly relates to a method for intelligently dynamically adjusting and controlling the air temperature of a blast furnace hot blast stove in an iron-making plant. BACKGROUND
[0002] In normal production of a blast furnace, hot blast with stable temperature and flow rate is needed to provide heat for the blast furnace. The existing hot blast stove for supplying hot blast to the blast furnace is generally a regenerative hot blast stove. The regenerative hot blast stove comprises a combustion chamber and a regenerative chamber, and adopts a switching cycle working mode of combustion heating and blast supply. The combustion heating is generated by burning gas in the combustion chamber to heat the heat storage material in the regenerative chamber. When the dome temperature of the combustion chamber reaches 1350 DEG C and the flue gas temperature reaches 400 DEG C, the combustion heating is stopped and the blast supply is switched. In the blast supply, the fan of the power source drives the cold blast to pass through the regenerative chamber to absorb the heat of the heat storage material, so that the cold blast becomes hot blast and is then supplied to the blast furnace. Generally, four hot blast stoves are needed for one blast furnace. Among them, the blast furnace with a diameter of 1350 m 3 The blast volume of the blast furnace is as high as 3500 m 3 / min, and the air temperature is required to be above 1200 DEG C or even 1250 DEG C. The hot blast stove operator manually adjusts or performs fuzzy adjustment by single PID control according to the instructions of the foreman.
[0003] In the process of blast furnace iron-making, the air temperature of the hot blast stove is a core parameter affecting the iron-making efficiency, energy consumption and quality of pig iron. The traditional air temperature adjustment relies on manual experience or single PID control, and has the following defects:
[0004] 1. Response lag: manual adjustment cannot track the changes of the blast furnace conditions in real time, resulting in large fluctuations of the air temperature.
[0005] 2. Poor multi-parameter coupling: the dynamic relationship among parameters such as gas pressure, air flow, combustion efficiency and the like is not comprehensively considered.
[0006] 3. Energy waste: excessive or insufficient blast supply leads to low heat energy utilization rate and increased NOx emission.
[0007] 4. Equipment wear: frequent and intense adjustment shortens the service life of the hot blast pipeline and the burner.
[0008] The existing technology adopts a fuzzy control algorithm, but lacks the fusion analysis ability of multi-source heterogeneous data. Therefore, it is necessary to design a method for intelligently dynamically adjusting and controlling the air temperature of a blast furnace hot blast stove in an iron-making plant based on multi-parameter feedback, so as to solve the conflict between the dynamic response of the existing actuator and the process constraint. SUMMARY
[0009] In view of the problems in the prior art, the present application aims to provide a method for intelligently dynamically adjusting and controlling the air temperature of a blast furnace hot blast stove in an iron-making plant.
[0010] The technical scheme adopted by the present application to solve its technical problems is: a kind of for iron mill blast furnace hot blast stove air temperature intelligent dynamic regulation control method, comprising the following steps:
[0011] S1, data acquisition: the running data of blast furnace hot blast stove is collected by sensing module, and is transmitted to main control unit, and main control unit controls operating parameter by actuator;
[0012] S2, data fusion: main control unit transmits data to blast furnace MES system, and Kalman filter algorithm and LSTM network are built in blast furnace MES system;
[0013] S3, construct multi-objective function: blast furnace hot blast stove air temperature regulation proportion:
[0014] K= ω1 * ΔT + ω2 * Fuel_cons + ω3 * NOx_emit;
[0015] S4, model predictive control: carry out fault diagnosis, real-time monitoring is carried out to the combustion state in blast furnace, when detecting pipe blockage or combustion oscillation, it is automatically switched to safety mode.
[0016] Specifically, the sensing module in step S1 includes thermocouple, pressure transmitter, mass flowmeter and infrared thermal imager, the thermocouple is installed at the top of blast furnace to measure dome temperature, the pressure transmitter is installed on the coal gas delivery pipeline to monitor the coal gas pressure, the mass flowmeter is installed on the air delivery pipeline and the coal gas delivery pipeline, the mass flowmeter monitors air flow and coal gas flow, and the infrared thermal imager is installed in the hot blast pipeline area, and the infrared thermal imager monitors the temperature field distribution of the hot blast pipeline.
[0017] Specifically, the main control unit in step S1 is equipped with edge computing chip, and runs multi-modal control algorithm, and the main control unit supports OPC UA protocol and communicates with blast furnace MES system.
[0018] Specifically, the actuator in step S1 includes electric regulating valve, variable frequency fan and burner inclination motor, the electric regulating valve is installed on the air delivery pipeline and the coal gas delivery pipeline, and the electric regulating valve controls air flow and coal gas flow;Variable frequency fan is installed on the air inlet pipe of hot blast stove, and variable frequency fan controls air supply adjustment;Burner inclination motor is installed on the angle adjustment platform of burner, and burner inclination motor controls the optimal combustion position of burner.
[0019] Specifically, the Kalman filter algorithm in step S2 eliminates the noise of all sensor transmission data of sensing module, and combines LSTM network to predict future 30 seconds working condition, and the working condition includes but is not limited to blast furnace permeability change out threshold range control variable frequency fan action.
[0020] Specific is, the blast furnace hot blast stove air temperature adjustment formula in the step S3 ΔT is the air temperature fluctuation rate, Fuel_cons is the fuel consumption rate, NOx_emit is the pollutant emission rate, ω1, ω2, ω3 are weight coefficients, and the weight coefficients are dynamically adjusted by Q-learning algorithm reinforcement learning.
[0021] Specific is, the model predictive control in the step S4 is periodically rolled optimization adjustment instruction with 0.5 seconds, and the fault diagnosis is calculated combustion state by PHM algorithm.
[0022] The present application has the following beneficial effects:
[0023] The present application has the following beneficial effects: BRIEF DESCRIPTION OF DRAWINGS
[0024] Fig. 1 It is a flow chart for the intelligent dynamic regulation and control method of blast furnace hot blast stove air temperature for ironworks.
[0025] Fig. 2 It is a data acquisition and processing flow chart for the intelligent dynamic regulation and control method of blast furnace hot blast stove air temperature for ironworks. DETAILED DESCRIPTION
[0026] The technical solutions in the embodiments of the present application will be further clearly and completely explained in detail with reference to the accompanying drawings in the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of the present application.
[0027] As shown in Figs. 1-2 A kind of intelligent dynamic regulation and control method of blast furnace hot blast stove air temperature for ironworks, comprising the following steps:
[0028] 1, data acquisition: the running data of blast furnace hot blast stove are collected by sensing module, and are transmitted to main control unit, and main control unit controls operating parameter by actuator.
[0029] The sensing module includes a thermocouple, a pressure transmitter, a mass flow meter and an infrared thermal imager, the thermocouple is installed at the top of the blast furnace to measure the dome temperature, the pressure transmitter is installed on the coal gas conveying pipeline to monitor the coal gas pressure, the mass flow meter is installed on the air conveying pipeline and the coal gas conveying pipeline, the mass flow meter monitors the air flow and the coal gas flow, and the infrared thermal imager is installed in the hot blast pipeline area, and the infrared thermal imager monitors the temperature field distribution of the hot blast pipeline.
[0030] The main control unit is provided with an edge computing chip and runs a multi-modal control algorithm, and the main control unit supports an OPC UA protocol and communicates with a blast furnace MES system.
[0031] The actuator includes an electric regulating valve, a variable frequency fan and a burner inclination motor, the electric regulating valve is installed on the air conveying pipeline and the coal gas conveying pipeline, the electric regulating valve controls the air flow and the coal gas flow, the variable frequency fan is installed on the air inlet pipeline of the hot blast furnace, and the variable frequency fan controls the air supply amount adjustment, and the burner inclination motor is installed on the angle adjusting platform of the burner, and the burner inclination motor controls the optimal combustion position of the burner.
[0032] 2. Data fusion: the main control unit transmits data to the blast furnace MES system, and a Kalman filter algorithm and an LSTM network are built in the blast furnace MES system.
[0033] The Kalman filter algorithm eliminates the noise of the data transmitted by all sensors of the sensing module, and the LSTM network is combined to predict the future 30-second working condition, and the working condition is controlled to act on the variable frequency fan when the blast furnace permeability changes out of the threshold range.
[0034] 3. Multi-objective function is constructed: blast furnace hot blast furnace air temperature regulation ratio:
[0035] K = ω1 * ΔT + ω2 * Fuel_cons + ω3 * NOx_emit.
[0036] In the formula, ΔT is the air temperature fluctuation rate, Fuel_cons is the fuel consumption rate, NOx_emit is the pollutant emission rate, ω1, ω2 and ω3 are weight coefficients, and the weight coefficients are dynamically adjusted through Q-leaming algorithm reinforcement learning.
[0037] A code implementation case of a multi-objective function of the application is as follows:
[0038] $$\min\left(\omega_1\cdot\Delta T+\omega_2\cdot\text{Fuel}_{cons}+\omega_3\cdot\text{NOx}_{emit}\right)$$.
[0039] 4. Model predictive control: fault diagnosis, real-time monitoring of the combustion state in the blast furnace, automatic switching to a safe mode when pipe blockage or combustion oscillation is detected.
[0040] The model predictive control (MPC) performs periodic rolling optimization of the adjustment command at 0.5 seconds, and the fault diagnosis is calculated by the PHM algorithm to calculate the combustion state.
[0041] In practical application of the blast furnace hot blast furnace air temperature intelligent adjusting device designed by the present application, through multi-source data fusion, dynamic multi-target optimization and digital twin technology, the precision and adaptability of air temperature adjustment are realized. Compared with the prior art, the air temperature fluctuation range is reduced from ±15℃ to ±3℃, the ton iron energy consumption is reduced by 5-12%, and the NOx emission is reduced by more than 20%.
[0042] The present application is not limited to the above-mentioned embodiments, and any person should know that the structural changes made under the inspiration of the present application fall within the protection scope of the present application.
[0043] The technologies, shapes and structures not described in detail in the present application are all known technologies.
Claims
1. A method for intelligent dynamic regulation and control of blast furnace hot blast stove temperature in an iron making plant, characterized by, The method comprises the following steps: S1, data acquisition: collecting the operation data of the blast furnace hot blast stove through the sensing module and transmitting to the main control unit, and the main control unit controls the operation parameters through the actuator; S2, data fusion: the main control unit transmits the data to the blast furnace MES system, and the Kalman filter algorithm and the LSTM network are built in the blast furnace MES system; S3, construction of multi-objective function: the blast furnace hot blast stove air temperature regulation ratio: K=ω1*ΔT+ω2*Fuel_cons+ω3*NOx_emit; S4, model predictive control: fault diagnosis, real-time monitoring of the combustion state in the blast furnace, automatic switching to the safety mode when detecting pipe blockage or combustion oscillation.
2. The method for intelligent dynamic regulation and control of blast temperature of blast furnace hot blast stove of iron making plant as claimed in claim 1 wherein, The sensing module in step S1 includes thermocouples, pressure transmitters, mass flow meters and infrared thermographs, the thermocouples are installed at the top of the blast furnace to measure the dome temperature, the pressure transmitter is installed on the coal gas conveying pipeline to monitor the coal gas pressure, the mass flow meter is installed on the air conveying pipeline and the coal gas conveying pipeline, the mass flow meter monitors the air flow and the coal gas flow, and the infrared thermograph is installed in the hot blast pipeline area, and the infrared thermograph monitors the temperature field distribution of the hot blast pipeline.
3. The method for intelligent dynamic regulation and control of blast temperature of blast furnace hot blast stove of iron making plant as claimed in claim 1 wherein, The main control unit in step S1 is equipped with an edge computing chip and runs a multi-modal control algorithm, and the main control unit supports OPC UA protocol and communicates with the blast furnace MES system.
4. The intelligent dynamic regulation control method for blast furnace hot blast stove temperature of ironworks blast furnace according to claim 1, characterized in that, The actuator in step S1 includes electric regulating valves, variable frequency fans and burner inclination motors, the electric regulating valves are installed on the air conveying pipeline and the coal gas conveying pipeline, the electric regulating valves control the air flow and the coal gas flow, the variable frequency fan is installed on the air inlet pipeline of the hot blast stove, and the variable frequency fan controls the air supply adjustment, and the burner inclination motor is installed on the angle adjustment platform of the burner, and the burner inclination motor controls the optimal combustion position of the burner.
5. The intelligent dynamic regulation control method for blast furnace hot blast stove temperature of ironworks blast furnace according to claim 4, characterized in that, The Kalman filter algorithm in step S2 eliminates the noise of all sensor transmission data of the sensing module, and combines the LSTM network to predict the future 30 seconds working condition, which includes but is not limited to the blast furnace permeability change out of the threshold range to control the variable frequency fan action.
6. The intelligent dynamic regulation control method for blast furnace hot blast stove temperature of ironworks blast furnace according to claim 1, characterized in that, In the formula of the blast furnace hot blast stove air temperature regulation ratio in step S3, ΔT is the air temperature fluctuation rate, Fuel_cons is the fuel consumption rate, NOx_emit is the pollutant emission rate, ω1, ω2 and ω3 are weight coefficients, and the weight coefficients are dynamically adjusted through Q-leaming algorithm reinforcement learning.
7. The method for intelligent dynamic regulation and control of blast temperature of blast furnace hot blast stove of iron making plant as claimed in claim 1 wherein, In step S4, the model predictive control is a periodical rolling optimization adjustment instruction with a cycle of 0.5 seconds, and the PHM algorithm is used for fault diagnosis to calculate the combustion state.