Method and system for blast furnace hot blast stove scheduling and blast furnace gas balance optimization based on data drive

Through data-driven blast furnace hot air grate scheduling and gas balance optimization methods, combined with real-time and historical data, the particle swarm optimization algorithm is used to optimize the duration of the hot air grate, which solves the uncertainty and fluctuation problems in blast furnace gas management, and improves operating efficiency and Safety and stabilize gas consumption.

CN118607890BActive Publication Date: 2024-11-15南京凯奥思数据技术有限公司
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411080682.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-08
Publication Date
2024-11-15
Estimated Expiration
2044-08-08

AI Technical Summary

Technical Problem

In blast furnace gas management, the steel industry faces problems such as uncertainty in gas consumption, fluctuations in hot air furnace fluctuations and difficult to accurately control gas balance, which affects the operation efficiency and production safety of blast furnaces, resulting in energy waste and environmental pollution.

Method used

The data-driven blast furnace hot air furnace scheduling and gas balance optimization method are adopted, combined with real-time and historical data science to arrange the hot air furnace replacement time and gas flow, and optimize the furnace and air supply time through particle swarm optimization algorithm to reduce the number of hot air furnaces. Stabilize gas consumption.

Benefits of technology

It improves the flexibility and efficiency of production scheduling, reduces the pressure of the gas pipeline network, stabilizes the gas consumption of blast furnaces, reduces the fluctuations and variance of the number of furnaces, and ensures the balance and safety of gas consumption.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118607890B_ABST
    Figure CN118607890B_ABST
Patent Text Reader

Abstract

The present invention discloses a method and system for scheduling hot blast furnaces and optimizing blast furnace gas balance based on data-driven. The method combines the real-time operation data and historical data of the blast furnace to calculate the initial state and duration of each hot blast furnace; then, according to the hot blast furnace replacement control mode of each hot blast furnace in the large blast furnace and / or the hot blast furnace replacement control mode of each hot blast furnace in the small blast furnace, respectively perform the furnace replacement scheduling time operation; based on the time of the hot blast furnace replacement scheduling, the future time per minute furnace burning state, the number of furnaces burned, and the variance of the number of furnaces burned are deduced; based on the hot blast furnace per minute furnace burning state comparison historical per minute furnace burning state statistical difference rate, as the overall furnace replacement execution rate; and then split to each blast furnace, as the execution rate of each blast furnace replacement. Scientifically, reasonably and accurately arrange the hot blast furnace replacement time and gas flow, improve the flexibility and efficiency of production scheduling, reduce the gas pipeline network pressure, and thus stabilize the blast furnace gas consumption.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of blast furnace hot blast stove replacement operations, and in particular to an optimization method and system based on combining blast furnace hot blast stove scheduling with blast furnace gas balance. Background Art

[0002] In the production process of the steel industry, blast furnace gas, as an important by-product, not only plays an important role in energy conservation and emission reduction, but is also a key factor in ensuring production continuity and stability.

[0003] The main problems faced by the steel industry in blast furnace gas management include uncertainty in gas consumption, fluctuations in hot blast furnace flue temperature, and difficulty in accurately controlling gas balance. These problems directly affect the operating efficiency and production safety of blast furnaces, and also lead to energy waste and environmental pollution. At present, most steel companies rely on experience to estimate gas volume, lack accurate predictions and scientific scheduling strategies, which makes it difficult to achieve optimal blast furnace operation and may even cause production accidents.

[0004] Therefore, how to select and optimize the time nodes for hot blast furnace replacement, reduce the number of hot blast furnaces burning simultaneously during the replacement process, and thus stabilize the consumption of blast furnace gas, is a technical problem that steel production enterprises urgently need to solve. Summary of the invention

[0005] The technical problem to be solved by the present invention is to develop a data-driven blast furnace hot blast stove scheduling and blast furnace gas balance optimization method and system, combining the real-time operation data and historical data of the blast furnace, scientifically, reasonably and accurately arrange the hot blast stove furnace change time and gas flow, improve the flexibility and efficiency of production scheduling, reduce the gas pipeline network pressure, and thus stabilize the blast furnace gas consumption.

[0006] The technical solution adopted by the present invention is:

[0007] A data-driven method for scheduling hot blast furnaces and optimizing blast furnace gas balance comprises at least the following steps:

[0008] Calculation steps for the initial state and duration of each hot blast furnace: obtain the initial state of each hot blast furnace and the measurement point data per minute within the set hour before the current time through the OPC interface, remove the abnormal data of state jump through the forward filling method, and count the duration of each state at the current time; the initial state of each hot blast furnace includes one of the following: furnace burning, air supply, and furnace stuffing;

[0009] According to the hot blast furnace change control mode of each hot blast furnace in the large blast furnace and / or the hot blast furnace change control mode of each hot blast furnace in the small blast furnace, the furnace change scheduling time operation is performed respectively; according to the initial state of each hot blast furnace in the large blast furnace or the small blast furnace, the number of furnace changes, the current clock or the deduced start time, and the preset furnace burning, air supply, and furnace change operation duration at each furnace change; output or deduce the furnace change state and start and end time of each hot blast furnace;

[0010] Based on the time of hot blast furnace replacement scheduling, the furnace burning status, number of furnaces burned, and variance of the number of furnaces burned per minute in the future are deduced;

[0011] The statistical difference rate between the hot blast furnace's burning status per minute and the historical burning status per minute is used as the overall furnace change execution rate; then it is divided into each blast furnace as the execution rate of each blast furnace change.

[0012] In the above technical solution, the control mode of changing the hot blast stoves in the large blast furnace includes the following steps:

[0013] A counts the hot air furnace numbers with the longest burning time and the longest air supply time based on the initial state duration. These two hot air furnaces are the hot air furnaces for the next furnace replacement;

[0014] B calculates the waiting time for the next furnace change based on the preset furnace burning time and the furnace burning time in the current state. If the waiting time is less than 0, the waiting time is set to 0 to prevent overlapping furnace changes in the blast furnace;

[0015] C records the start time of this furnace change as "current clock + waiting time", and the end time as "current clock + waiting time + furnace change operation time";

[0016] D updates the current clock according to the clock update logic "current clock = current clock + waiting time + furnace change operation time";

[0017] E. Update the current status based on the number of hot blast stoves involved in firing and air supply in the large blast furnace and the benchmark duration of the firing and air supply status of each hot blast stove, and update the duration to add "waiting time + furnace changing operation time";

[0018] F returns to step A to deduce the next furnace change time.

[0019] In the above technical solution, in step E, if the current state is "two burning and two sending", the states of the two hot blast furnaces in step 1) are switched, from burning to sending, and from sending to burning, respectively, and the state duration is set to 0, the states of other hot blast furnaces remain unchanged, and the duration is increased by the waiting time + the furnace changing operation time; if the current state is "three burning and one sending", the hot blast furnace in the burning state in step (1) is switched to the sending state, and the state duration is set to 0, the states of other hot blast furnaces remain unchanged, and the duration is increased by the waiting time + the furnace changing operation time; if the current state is "one burning and three sending" or all four are in the sending state, the hot blast furnace in the sending state in step (1) is switched to the burning state, and the state duration is set to 0, the states of other hot blast furnaces remain unchanged, and the duration is increased by "the waiting time + the furnace changing operation time".

[0020] In the above technical solution, the control mode of changing the hot blast stoves in the small blast furnace includes the following steps:

[0021] AA counts the hot air furnace numbers with the longest burning time and the longest air supply time based on the initial state duration. These two hot air furnaces are the hot air furnaces that may be replaced next time.

[0022] BB calculates the waiting time for the next furnace change based on the difference between the preset furnace burning time and the furnace burning time in the current state. If the waiting time is less than 0, the waiting time is set to 0 to prevent overlapping furnace changes in the blast furnace.

[0023] The CC records the start time of this furnace change as "current clock + waiting time", and the end time as "current clock + waiting time + furnace change operation time";

[0024] DD updates the current clock to "current clock + waiting time + furnace change operation time";

[0025] EE updates the current state. If the current state is two burning and one sending, the hot blast furnace in the burning state in step 1) is switched to the sending state, and the state duration is set to 0. The states of other hot blast furnaces remain unchanged, and the duration is increased by the waiting time + the furnace changing operation time. If the current state is one burning and two sending or three sending, the hot blast furnace in the sending state in step 1) is switched to the burning state, and the state duration is set to 0. The states of other hot blast furnaces remain unchanged, and the duration is increased by "waiting time + furnace changing operation time".

[0026] Return to step AA to deduce the next furnace change time.

[0027] In the above technical solution, the furnace burning status, furnace burning number, and the variance of the furnace burning number per minute in the future are deduced based on the time of the hot blast furnace replacement schedule according to the following steps:

[0028] S1) Set the current clock and deduction duration, calculate the status of each hot blast furnace every minute according to the furnace replacement schedule obtained by the hot blast furnace deduction logic;

[0029] S2) Calculate the number of furnaces burning simultaneously per minute based on the status of the hot air furnace per minute;

[0030] S3) Calculate the variance of the number of furnaces fired during this simulation period based on the number of furnaces fired simultaneously per minute.

[0031] In the above technical solution, particle swarm optimization is used to optimize the operation time of the hot blast furnace, specifically, the burning and air supply time of each furnace change preset in the optimization input parameters. Other neural network optimization methods can also be used to optimize the operation time.

[0032] In the above technical solution, the steps of using particle swarm optimization to optimize the operation time of the hot blast furnace are as follows:

[0033] S11 constructs an initial particle group according to the mean of the historical burning and air supply time of each hot blast stove, and the particle group size is set to 50 particles;

[0034] S12 sets the upper and lower bounds of each particle, i.e., the fluctuation range of the time length [-10, 10]; sets the optimal position of each particle, i.e., the current position; calculates the global optimal solution of the current particle swarm, i.e., the particle with the optimal position;

[0035] S13 constructs an optimization objective function, taking the number of furnaces fired per minute and the fluctuation variance of the number of furnaces fired in module 1 as the target;

[0036] S14 updates the speed and position of the particle according to the optimal position of each particle within the range of the upper and lower bounds of the particle;

[0037] S15 substitutes all updated particles into the objective function, calculates the objective value of each particle, and if the objective value is smaller than the objective value of the optimal position of the particle, updates the optimal position of the particle; if the objective value is smaller than the objective value of the current global optimal solution, updates the global optimal solution;

[0038] S16 sets the number of iterations, repeats steps S14-S15, and outputs the global optimal solution, i.e., the duration of each hot blast furnace firing and air supply;

[0039] S17 substitutes the global optimal solution into the blast furnace change time deduction function in module 1 to obtain the blast furnace change time arrangement of each hot blast furnace after optimizing the peak shifting.

[0040] In the above technical solution, the execution rate of each blast furnace replacement is calculated as follows:

[0041] .

[0042] The present invention also provides a data-driven blast furnace hot blast stove scheduling and blast furnace gas balance optimization system for implementing the above method.

[0043] At the same time, the present invention also provides a storage medium storing a computer program, which is used to execute the steps included in the above method when the program is executed.

[0044] In summary, the present invention aims to solve key problems such as the stable consumption of hot blast furnace gas by establishing a series of data-driven operations optimization models, in order to achieve the following objectives: improve the flexibility and efficiency of production scheduling and reduce the pressure of gas pipeline network through a scientific hot blast furnace scheduling model. The core issue of the study is how to combine the real-time operation data and historical data of the blast furnace to develop a set of hot blast furnace replacement time node scheduling strategies, in order to reduce the number of hot blast furnaces burning at the same time, thereby stabilizing the consumption of blast furnace gas.

[0045] Compared with the prior art, the beneficial effects of the present invention are:

[0046] A creative control mode for switching multiple hot blast furnaces in a large blast furnace and multiple hot blast furnaces in a small blast furnace was proposed. The furnace firing status in the future was estimated based on the time of the hot blast furnace switching schedule. The furnace firing time and furnace switching time model were optimized based on the particle swarm algorithm. The peak and valley values ​​of the number of furnaces fired are determined by the furnace conditions of each blast furnace and their respective operating time. The model can reduce the fluctuation caused by the operating time.

[0047] The simulated firing status of each hot blast furnace is compared with the actual firing status to calculate the scheduling execution rate. The furnace replacement time node obtained after model simulation and optimization makes the number of furnaces fired at the same time have a smaller fluctuation range and smaller variance than the historical actual value;

[0048] It is difficult for the model to reduce the peak number of furnace firings, but it is easier to reduce the valley number;

[0049] The mean value of the hot blast stove gas consumption calculated based on the model is roughly equal to the historical consumption mean value, and the variance is smaller than the historical consumption, indicating that executing the furnace change time according to the model will stabilize the gas consumption of the hot blast stove. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] The present invention will be further described below with reference to the accompanying drawings and embodiments, in which:

[0051] Figure 1 The invention discloses a data-driven hot blast furnace scheduling and blast furnace gas balance optimization method (a hot blast furnace scheduling algorithm based on particle swarm optimization). DETAILED DESCRIPTION

[0052] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0053] like Figure 1 As shown, in this embodiment, 14 hot blast furnaces of 4 blast furnaces of a steel enterprise are taken as an example, among which No. 1 and No. 2 blast furnaces are large blast furnaces, each with 4 hot blast furnaces, and No. 3 and No. 4 blast furnaces are small blast furnaces, each with 3 hot blast furnaces. The total gas consumption of all hot blast furnaces accounts for about 1 / 3 of the total consumption of the gas pipeline network, and only the hot blast furnaces in the furnace burning state consume gas, and the air supply state does not consume gas. The state switching of the hot blast furnace is determined by the flue gas temperature. Therefore, the gas consumption of the hot blast furnace is related to the number of hot blast furnaces in the furnace burning state, and the hot blast furnace in the furnace burning state at a certain moment is determined by the state of the hot blast furnace at the previous moment and the flue gas temperature when the furnace is changed.

[0054] Therefore, the number of furnace changes at the same time is determined by two factors:

[0055] (1) The initial operating conditions of all hot blast furnaces (state, temperature, and duration of the state);

[0056] (2) The flue gas temperature (furnace change duration) is determined each time a furnace is changed. The flue gas temperature of the hot blast furnaces of No. 1 and No. 2 blast furnaces is controlled between 340°C and 380°C, and the hot blast temperature is between 1200°C and 1230°C. The flue gas temperature of No. 3 and No. 4 blast furnaces is 330°C to 360°C, and the hot blast temperature is 1200°C.

[0057] However, the initial operating condition is the actual hot blast furnace state, but the flue gas temperature when changing the furnace determines the duration of the state and ultimately affects the number of furnace changes at the same time, which is something we can decide.

[0058] Therefore, the focus of this model is how to calculate a reasonable state duration for each hot blast furnace each time the furnace is changed, and to achieve the purpose of stabilizing the number of furnace changes through the change of state duration, thereby achieving the purpose of stabilizing the gas consumption of the hot blast furnace.

[0059] Model composition:

[0060] Module 1 Hot blast furnace scheduling time node calculation

[0061] Calculate the initial state and duration of each hot blast furnace; obtain the measurement point data of each hot blast furnace's initial state (burning furnace, air supply, and stuffy furnace) every minute within 4 hours before the current time through the OPC interface, remove the abnormal data of state jump through the forward filling method, and count the duration of each state at the current time;

[0062] The logic or control mode of the 4 hot blast stoves in the large blast furnace and the 3 hot blast stoves in the small blast furnace is as follows; the deduction logic or control mode of the 4 hot blast stoves in the large blast furnace is as follows:

[0063] Input the initial state of each hot blast furnace in a large blast furnace, the number of furnace changes, the current clock (start time of deduction), and the preset duration of furnace burning, air supply, and furnace change each time the furnace is changed; output the furnace change state and start and end time of each hot blast furnace. The specific calculation logic is as follows:

[0064] 1) Based on the initial state duration, the hot air furnace numbers of the longest burning time and the longest air supply time are counted. These two hot air furnaces are the hot air furnaces for the next furnace replacement;

[0065] 2) Based on the preset furnace burning time and the current state of furnace burning time, the waiting time for the next furnace change is calculated. If the waiting time is less than 0, the waiting time is set to 0 to prevent overlapping furnace changes in the blast furnace;

[0066] 3) The start time of this furnace change is recorded as the current clock + waiting time, and the end time is recorded as the current clock + waiting time + furnace change time;

[0067] 4) Update current clock = current clock + waiting time + furnace change operation time;

[0068] 5) Update the current state. If the current state is two burning and two sending, the two hot blast furnaces in step 1) are switched (burning furnace is switched to sending air, and sending air is switched to burning furnace), and the state duration is set to 0. The states of other hot blast furnaces remain unchanged, and the duration is increased by the waiting time + furnace changing operation time; if the current state is three burning and one sending, the hot blast furnace in the burning state in step 1) is switched to the sending air state, and the state duration is set to 0. The states of other hot blast furnaces remain unchanged, and the duration is increased by the waiting time + furnace changing operation time; if the current state is one burning and three sending or all four are in the sending air state, the hot blast furnace in the sending air state in step 1) is switched to the burning state, and the state duration is set to 0. The states of other hot blast furnaces remain unchanged, and the duration is increased by the waiting time + furnace changing operation time;

[0069] 6) Return to step 1) to deduce the next furnace change time;

[0070] The deduction logic or control mode of the three hot blast stoves in a small blast furnace:

[0071] Input the initial state of each hot blast furnace in a small blast furnace, the number of furnace changes, the current clock (start time of deduction), and the preset duration of furnace burning, air supply, and furnace change each time the furnace is changed; output the furnace change state and start and end time of each hot blast furnace. The specific calculation logic is as follows:

[0072] 1) Based on the initial state duration, the hot air furnace numbers of the longest burning time and the longest air supply time are counted. These two hot air furnaces are the hot air furnaces that may be replaced next time;

[0073] 2) Based on the preset furnace burning time and the current state of furnace burning time, the waiting time for the next furnace change is calculated. If the waiting time is less than 0, the waiting time is set to 0 to prevent overlapping furnace changes in the blast furnace;

[0074] 3) The start time of this furnace change is recorded as the current clock + waiting time, and the end time is recorded as the current clock + waiting time + furnace change time;

[0075] 4) Update current clock = current clock + waiting time + furnace change operation time;

[0076] 5) Update the current state. If the current state is two burning and one sending, the hot blast furnace in the burning state in step 1) is switched to the sending state, and the state duration is set to 0. The states of other hot blast furnaces remain unchanged, and the duration is increased by the waiting time + the furnace changing operation time; if the current state is one burning and two sending or three sending, the hot blast furnace in the sending state in step 1) is switched to the burning state, and the state duration is set to 0. The states of other hot blast furnaces remain unchanged, and the duration is increased by the waiting time + the furnace changing operation time;

[0077] 6) Return to step 1) to deduce the next furnace change time;

[0078] Based on the time of the hot blast furnace replacement schedule, the furnace burning status, the number of furnaces burned, and the variance of the number of furnaces burned in the future are calculated every minute; 1) Set the current clock and the deduction time, and calculate the status of each hot blast furnace every minute according to the furnace replacement time schedule obtained by the hot blast furnace deduction logic;

[0079] 2) According to the status of the hot air furnace every minute, calculate the number of furnaces burning simultaneously every minute;

[0080] 3) Based on the number of furnaces burning simultaneously per minute, calculate the variance of the number of furnaces burning during this period of simulation

[0081] Module 2 Particle Swarm Optimization of Hot Blast Furnace Operation Time

[0082] The main function of this module is to optimize the input parameters in module 1 - the preset furnace burning and air supply time for each furnace change, so as to achieve the purpose of staggered furnace change. The specific logic is as follows:

[0083] Construct an initial particle swarm based on the mean of the historical firing and air supply duration of each hot blast furnace, and set the particle swarm size to 50 particles; set the upper and lower bounds of each particle, that is, the fluctuation range of the duration [-10,10]; set the optimal position of each particle, that is, the current position; calculate the global optimal solution of the current particle swarm, that is, the particle with the optimal position; construct an optimization objective function, with the number of furnaces fired per minute and the fluctuation variance of the number of furnaces fired in module one as the target; within the range of the upper and lower bounds of the particles, update the speed and position of the particles according to the optimal position of each particle; substitute all updated particles into the objective function, calculate the target value of each particle, if the target value is smaller than the target value of the optimal position of the particle, update the optimal position of the particle; if the target value is smaller than the target value of the current global optimal solution, update the global optimal solution; set the number of iterations, repeat steps 4-5, and output the global optimal solution, that is, the duration of each firing and air supply of each hot blast furnace; substitute the global optimal solution into the blast furnace change time deduction function in module one, and obtain the time arrangement of each hot blast furnace change after optimizing the peak shift. Module three Evaluation and analysis of the execution rate of each blast furnace change

[0084] The statistical difference rate of the hot blast furnace burning status per minute compared with the historical burning status per minute introduced by the previous model is used as the overall furnace change execution rate; then it is divided into each blast furnace as the execution rate of each blast furnace change; the calculation formula is as follows

[0085]

[0086] It should be understood that those skilled in the art can make improvements or changes based on the above description, and all these improvements and changes should fall within the protection scope of the present invention.

Claims

1. A data-driven blast furnace hot blast stove scheduling and blast furnace gas balance optimization method, characterized in that At least the following steps are included: Calculation steps for the initial state and duration of each hot air furnace: obtain the initial state of each hot air furnace and the measurement point data per minute within the set hour before the current time through the OPC interface, remove the abnormal data of state jump through the forward filling method, and count the duration of each state at the current time; The initial state of each hot blast furnace includes one of furnace burning, air supply, and furnace stuffing; According to the hot blast furnace change control mode of each hot blast furnace in the large blast furnace and / or the hot blast furnace change control mode of each hot blast furnace in the small blast furnace, the furnace change scheduling time operation is performed respectively; according to the initial state of each hot blast furnace in the large blast furnace or the small blast furnace, the number of furnace changes, the current clock or the deduced start time, and the preset furnace burning, air supply, and furnace change operation duration at each furnace change; output or deduce the furnace change state and start and end time of each hot blast furnace; The particle swarm optimization is used to optimize the operation time of the hot blast furnace, specifically, the furnace burning and air supply time of each furnace change preset in the optimization input parameters; Based on the time of hot blast furnace replacement scheduling, the furnace burning status, number of furnaces burned, and variance of the number of furnaces burned per minute in the future are deduced; The statistical difference rate of the hot blast furnace's per-minute burning status compared with the historical per-minute burning status is used as the overall furnace change execution rate; then it is divided into each blast furnace as the execution rate of each blast furnace change; Among them, the hot blast stove replacement control mode in the large blast furnace includes the following steps: A counts the hot air furnace numbers with the longest burning time and the longest air supply time based on the initial state duration. These two hot air furnaces are the hot air furnaces for the next furnace replacement; B calculates the waiting time for the next furnace change based on the preset furnace burning time and the furnace burning time in the current state. If the waiting time is less than 0, the waiting time is set to 0 to prevent overlapping furnace changes in the blast furnace; C records the start time of this furnace change as "current clock + waiting time", and the end time as "current clock + waiting time + furnace change operation time"; D Update the current clock according to the clock update logic "current clock = current clock + waiting time + furnace change operation time"; E. Update the current status based on the number of hot blast stoves involved in firing and air supply in the large blast furnace and the benchmark duration of the firing and air supply status of each hot blast stove, and update the duration to add "waiting time + furnace changing operation time"; F returns to step A to deduce the next furnace change time; The control mode for changing the hot blast stoves in the small blast furnace comprises the following steps: AA counts the hot air furnace numbers with the longest burning time and the longest air supply time based on the initial state duration. These two hot air furnaces are the hot air furnaces that may be replaced next time. BB calculates the waiting time for the next furnace change based on the difference between the preset furnace burning time and the furnace burning time in the current state. If the waiting time is less than 0, the waiting time is set to 0 to prevent overlapping furnace changes in the blast furnace. The CC records the start time of this furnace change as "current clock + waiting time", and the end time as "current clock + waiting time + furnace change operation time"; DD updates the current clock to "current clock + waiting time + furnace change operation time"; EE updates the current state. If the current state is two burning and one sending, the hot blast furnace in the burning state in step AA is switched to the sending state, and the state duration is set to 0. The states of other hot blast furnaces remain unchanged, and the duration is increased by the waiting time + the furnace changing operation time; if the current state is one burning and two sending or three sending, the hot blast furnace in the sending state in step AA is switched to the burning state, and the state duration is set to 0. The states of other hot blast furnaces remain unchanged, and the duration is increased by "waiting time + furnace changing operation time"; Return to step AA to deduce the next furnace change time.

2. The method for scheduling hot blast furnace and optimizing blast furnace gas balance based on data drive according to claim 1, characterized in that In step E, if the current state is "two burning and two sending", the states of the two hot blast furnaces in step A are switched, from burning to sending, and from sending to burning, respectively, and the state duration is set to 0, the states of other hot blast furnaces remain unchanged, and the duration is increased by the waiting time + the furnace changing operation time; if the current state is "three burning and one sending", the hot blast furnace in the burning state in step A is switched to the sending state, and the state duration is set to 0, the states of other hot blast furnaces remain unchanged, and the duration is increased by the waiting time + the furnace changing operation time; If the current state is "one burning and three sending" or all four are in the air sending state, the hot blast furnace in the air sending state in step A is switched to the burning state, and the state duration is set to 0. The states of other hot blast furnaces remain unchanged, and the duration is increased by "waiting time + furnace changing operation time".

3. The method for scheduling hot blast furnace and optimizing blast furnace gas balance based on data drive according to claim 1, characterized in that The steps of using particle swarm optimization to optimize the operation time of hot blast furnace are as follows: S11 constructs an initial particle group according to the mean of the historical burning and air supply time of each hot blast stove, and the particle group size is set to 50 particles; S12 sets the upper and lower bounds of each particle, i.e., the fluctuation range of the time length [-10, 10]; sets the optimal position of each particle, i.e., the current position; calculates the global optimal solution of the current particle swarm, i.e., the particle with the optimal position; S13 constructs an optimization objective function, taking the number of furnaces fired per minute and the fluctuation variance of the number of furnaces fired in module 1 as the target; S14 updates the speed and position of the particle according to the optimal position of each particle within the range of the upper and lower bounds of the particle; S15 substitutes all updated particles into the objective function, calculates the objective value of each particle, and if the objective value is smaller than the objective value of the optimal position of the particle, updates the optimal position of the particle; if the objective value is smaller than the objective value of the current global optimal solution, updates the global optimal solution; S16 sets the number of iterations, repeats steps S14-S15, and outputs the global optimal solution, i.e., the duration of each hot blast furnace firing and air supply; S17 substitutes the global optimal solution into the blast furnace change time deduction function in module 1 to obtain the blast furnace change time arrangement of each hot blast furnace after optimizing the peak shifting.

4. The method for scheduling hot blast furnaces and optimizing blast furnace gas balance based on data drive according to claim 1, characterized in that The execution rate of each blast furnace replacement; the calculation formula is as follows: 。 5. A data-driven blast furnace hot blast stove scheduling and blast furnace gas balance optimization system, characterized in that Used to implement the method described in any one of claims 1 to 4 above.

6. A storage medium storing a computer program, characterized in that When the program is executed, it is used to perform the steps included in the method according to any one of claims 1 to 4.