Coking chamber coking process temperature field collaborative control method based on digital twinning

Abstract

The application discloses a coke oven carbonization chamber coking process temperature field cooperative control method based on digital twinning, which comprises the following steps: constructing a heat supply correlation topology digital twin body between the combustion chamber and the carbonization chamber, generating an asynchronous degree vector of the whole oven coking process by calculating the remaining coking time, and establishing an asynchronous degree propagation dynamics model; constructing an optimization framework with the asynchronous degree vector and the total energy consumption of the whole oven as targets, solving the incremental instruction of the combustion chamber through an online Pareto negotiation mechanism; superimposing the incremental instruction and the current heat supply instruction and then issuing and executing them, and taking the change rate of the asynchronous degree vector module length as a trigger mechanism, dynamically adjusting the control period and performing closed-loop iteration to the zero neighborhood. Through the asynchronous degree propagation dynamics modeling, the double-target Pareto online negotiation and the asynchronous degree change rate dynamic trigger control period adjustment, the application effectively improves the cooperative control quality of the whole oven temperature field, and solves the problem of giving consideration to the synchronous maturation of the carbonization chamber coking process and the optimization of energy consumption.

Description

Technical Field

[0001] This invention relates to the field of process control technology in the coking industry, specifically to a method for coordinated control of the temperature field during the coking process in the coking chamber of a coke oven based on digital twins. Background Technology

[0002] Coke ovens are core thermal equipment in the metallurgical industry used to convert coal into coke through high-temperature dry distillation under air-isolated conditions. In actual production, a coke oven typically contains dozens to hundreds of carbonization chambers. To ensure coke quality and minimize energy consumption, precise control of the heat supply to each combustion chamber is necessary to ensure that the temperature field of the coal in each carbonization chamber evolves along a predetermined trajectory, ultimately achieving uniform maturation within the planned coking cycle. However, in actual production, due to different coking sequences, each carbonization chamber is always in a spatiotemporally asynchronous state of coking process, resulting in misaligned heat demands over time. Further complicating matters, when the heat supply to one combustion chamber is adjusted to improve the coking process of its associated carbonization chambers, this adjustment propagates to adjacent carbonization chambers through heating topology coupling, triggering a chain propagation effect of asynchrony. Current technologies do not quantitatively model and control this propagation effect, making it difficult to accurately predict its global impact in control decisions.

[0003] A search revealed that relevant prior art, such as the furnace temperature control device, furnace temperature control method, and coke manufacturing method disclosed in patent number WO2026094568A1, includes: a storage unit storing at least one of a first model and a second model; and a control unit. The control unit acquires operational information including temperature measurements of multiple combustion chambers, temperature measurements of multiple carbonization chambers, actual values ​​of heat supplied to multiple combustion chambers, and actual and planned values ​​of the state of coal filling multiple carbonization chambers; calculates a predicted temperature value for each of the multiple combustion chambers by inputting information included in the operational information into a first volume model, or by inputting information included in the operational information into a second volume model; and calculates the heat supplied to each of the multiple combustion chambers based on the target temperature of the combustion chamber and the calculated predicted temperature value of the combustion chamber, or based on the target temperature of the carbonization chamber and the calculated predicted temperature value of the carbonization chamber. Patent CN120699641A discloses a coke oven heating control system, method, apparatus, electronic equipment, and storage medium, including: a coke pushing information acquisition module for acquiring the position of the coke pushing rod and the corresponding temperature value in the coke pushing rod; a gas regulation module connected to the gas pipelines of different combustion chambers in the coke oven for regulating the gas intake in each combustion chamber; and a central control module electrically connected to the coke pushing information acquisition module and the gas regulation module for acquiring the position and temperature value of the coke pushing rod and determining the combustion chamber based on the position of the coke pushing rod; and controlling the gas regulation module to regulate the gas intake in the combustion chamber based on the temperature value and the combustion chamber. This achieves a more uniform degree of coke maturity in the combustion chamber, improving coke quality. Patent CN120524722B discloses a building energy efficiency dynamic simulation method and system based on digital twins. This method designs a digital twin model of building energy efficiency using a neural symbolic method, and can autonomously generate energy efficiency optimization strategies based on real-time data and external changes through interactive simulation using reinforcement learning. To ensure model accuracy, all data is assigned a unified semantic framework and knowledge associations, and machine learning algorithms are used to achieve automated synchronous updates and full-dimensional adaptive corrections of the model, including adjustments to state variables, model parameters, and structural information. It can also predict future energy consumption, generate future control action sequences, and use event-driven MPC algorithms to generate high-level instructions for key events to achieve intelligent control, ultimately minimizing building energy consumption while meeting various constraints. This method fully leverages the advantages of data-driven and model-driven approaches, enhancing the intelligence level of building energy efficiency management.

[0004] The aforementioned existing technologies establish a simplified mathematical model of the coke oven thermal process and perform multivariate coordinated optimization of the heat supply to each combustion chamber in order to reduce energy consumption while meeting the tracking accuracy of the temperature curves of each carbonization chamber. However, its optimization objective is still the single indicator of minimizing the temperature deviation of each carbonization chamber. In addition, the control cycle of the existing technologies is usually fixed and cannot be dynamically adjusted according to the actual convergence state of the coking process. Therefore, the existing technologies fail to fundamentally solve the core problem of how to enable the asynchronous carbonization chambers to work together to achieve coking maturity on a whole-furnace scale under the constraints of physical coupling of heating. Summary of the Invention

[0005] The purpose of this invention is to provide a method for coordinated control of the temperature field in the coking process of coking chambers in coke ovens based on digital twins. This method overcomes the global control deviation caused by neglecting the propagation of asynchronous degree in the prior art, avoids paying too high an energy cost for pursuing minor uniformity improvements, and solves the problem of coordinated control of coking processes in multiple coking chambers under the conditions of physical coupling and asynchronous process conflict.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A method for coordinated control of the temperature field during the coking process in the coking chamber of a coke oven based on digital twins includes the following steps:

[0008] S1: Construct a digital twin of the heating association topology between the combustion chamber and the carbonization chamber, calculate the remaining coking time of each carbonization chamber under the current heating command of the combustion chamber in real time, generate the asynchronous degree vector of the coking process of the whole furnace, and establish an asynchronous degree propagation dynamic model based on the heating association topology;

[0009] S2: Construct an optimization architecture with asynchronous degree vector and total furnace energy consumption as objectives, and use the asynchronous degree propagation dynamics model as dynamic constraint to solve the incremental command of the combustion chamber through an online Pareto negotiation mechanism;

[0010] S3: The incremental command is superimposed with the current heating command and then issued for execution. The control cycle is dynamically adjusted and closed-loop iteration is performed using the rate of change of the asynchronous degree vector magnitude as the trigger mechanism, driving the asynchronous degree vector to converge to the zero neighborhood.

[0011] Furthermore, the heating-related topology digital twin constructed in S1 is a two-way coupled model:

[0012] On the carbonization chamber side, the coking process of the coal is controlled by a set of partial differential equations describing the coupled evolution of the coal temperature field and the volatile matter analysis process; among them, the temperature field of the coal inside the carbonization chamber satisfies the unsteady-state heat conduction equation containing an internal heat source, expressed as:

[0013]

[0014] In the formula, The temperature field of the coal in the carbonization chamber is a function of spatial coordinates and time; This indicates the apparent density of the coal. This indicates the specific heat capacity of the coal. Indicates the thermal conductivity of the coal; This represents the thermal conductivity term, which describes the process of heat diffusion within the coal. This represents the pyrolysis reaction source term, that is, the contribution of heat absorption or release during the pyrolysis of coal to the temperature field;

[0015] The pyrolysis reaction source term is directly related to the volatile matter precipitation rate. The volatile matter precipitation process is described by the equation of state of the residual volatile matter in the coal, and its evolution follows a first-order kinetic model, expressed as:

[0016]

[0017] In the formula, It indicates the residual volatile matter in coal, that is, the mass of volatile matter that has not yet been released from a unit mass of coal; This indicates the rate of change of residual volatile matter over time. The pre-exponential factor representing the pyrolysis reaction; This represents the activation energy of a pyrolysis reaction. Represents the universal gas constant; Indicates the temperature of the coal;

[0018] On the combustion chamber side, the heat generated by the combustion of gas is transferred to the carbonization chamber through the furnace wall. The furnace wall serves as a heat transfer channel between the combustion chamber and the carbonization chamber, and its heat transfer characteristics are characterized by the dynamic thermal resistance of the furnace wall. That is, the thermal resistance changes dynamically with the operating state of both sides of the furnace wall. A one-dimensional unsteady-state heat conduction control equation along the thickness direction inside the furnace wall is established:

[0019]

[0020] In the formula, Indicates the density of the furnace wall material; This indicates the specific heat capacity of the furnace wall material; The temperature field inside the furnace wall is represented by the furnace wall thickness coordinates as a function of time. Indicates the thermal conductivity of the furnace wall material; The coordinates represent the direction of furnace wall thickness.

[0021] Furthermore, S1 also includes real-time correction of the heating-related topology digital twin using a data assimilation algorithm; wherein, the remaining coking time is the difference between the current time and the time when the carbonization chamber in the heating-related topology digital twin reaches a mature state; for the first One carbonization chamber, remaining coking time Represented as:

[0022]

[0023] In the formula, Indicates the first The virtual moment when a carbonization chamber reaches a mature state in the digital twin simulation; Indicates the current moment;

[0024] After obtaining the remaining coking time for each carbonization chamber, calculate the corresponding degree of asynchrony; for the first... For each carbonization chamber, the asynchrony is defined as the deviation between the remaining coking time and the planned remaining coking time. The planned remaining coking time is the total planned coking time of the carbonization chamber from the coal loading time to the predetermined coking push time, minus the actual coking time elapsed from the coal loading time to the current time. It is expressed as:

[0025]

[0026] In the formula, Indicates the first The remaining coking time for each carbonization chamber; Indicates the first The total planned coking time for each carbonization chamber from the time of coal loading to the scheduled time of coking push; Indicates the first The actual coking time that has elapsed from the moment the coal was loaded into the carbonization chamber to the present moment;

[0027] Based on the difference between the remaining coking time and the planned remaining coking time, the asynchrony of each carbonization chamber is obtained, and the entire furnace is then... The asynchrony of each carbonization chamber within a coking cycle is arranged according to its serial number, forming a... Asynchronous degree vector of the whole furnace coking process .

[0028] Furthermore, the construction process of the asynchronous degree propagation dynamics model in S1 is as follows:

[0029] S101: Perform sensitivity analysis on the digital twin of the heating-related topology and obtain the sensitivity matrix;

[0030] S102: Based on sensitivity analysis, construct the asynchronous propagation matrix A, and then establish the asynchronous propagation dynamic equation, expressed as:

[0031]

[0032] in, This represents the asynchrony vector of the k-th control cycle; This indicates the change in asynchrony directly caused by the adjustment of the heating supply; Indicates the asynchronous propagation term;

[0033] The relationship between the change in asynchrony and the adjustment of heating supply is as follows:

[0034]

[0035] In the formula, This represents the transformation matrix.

[0036] Furthermore, the optimization architecture in S2 includes a first optimization objective. Second optimization objective Among them, the first optimization objective Defined as minimizing the modulus of the asynchronous degree vector prediction after applying heat supply adjustment, expressed as:

[0037]

[0038] In the formula, This represents the magnitude of the predicted value of the adjusted asynchronous degree vector;

[0039] Second optimization objective Defined as minimizing the total energy consumption of the entire furnace after applying heat supply adjustment, the total energy consumption of the entire furnace is characterized by the total gas consumption of each combustion chamber, after applying heat supply adjustment. Then, the predicted value of the total energy consumption of the entire furnace. Represented as:

[0040]

[0041] In the formula, This represents the total gas consumption of all combustion chambers in the entire furnace during the current control cycle; Indicates the first The change in heat supply to each combustion chamber; then the second optimization objective Represented as:

[0042] .

[0043] Furthermore, the method for obtaining incremental instructions in S2 is as follows:

[0044] S201: The second optimization objective Convert to constraints, and set M energy consumption constraint levels. ;

[0045] S202: Based on each energy consumption constraint level To minimize the primary optimization objective, additional constraints are added. Obtain M non-dominated solutions;

[0046] S203: Arrange the M non-dominated solutions in the objective space according to... and The values ​​are arranged to form the Pareto front curve of the current control cycle;

[0047] S204: Based on the principle of maximizing marginal benefits, the optimal operating point is selected, expressed as:

[0048]

[0049] In the formula, This indicates the selected optimal operation point; Let represent the set of all nondominated solutions on the Pareto front; This indicates that the constraint level is positive; the change in heat supply corresponding to the selected optimal operating point is used as the incremental instruction. .

[0050] Furthermore, in S3, the incremental command is superimposed with the current heating command of each combustion chamber to obtain the updated overall furnace heating command vector. , is represented as:

[0051]

[0052] Expanding to each combustion chamber, the first The updated heating command for each combustion chamber is expressed as follows:

[0053]

[0054] After a control cycle is completed, calculate the rate of change of the asynchronous degree vector magnitude. :

[0055]

[0056] In the formula, This is the magnitude of the asynchronous degree vector for this control cycle. The magnitude of the asynchronous degree vector of the previous control cycle;

[0057] By setting three threshold parameters, namely the upper limit threshold... Lower threshold and adjustment ratio factor Combined with the rate of change of the asynchronous degree vector magnitude Dynamic adjustment of the execution control cycle.

[0058] Furthermore, the method for obtaining the sensitivity matrix in S101 is as follows: Set vector This represents the current heating command for all combustion chambers in the entire furnace, vector. This represents the actual heat received by each carbonization chamber in the entire furnace under the current heating command; at the current operating point of the digital twin, a small positive perturbation is applied to the heat supply of each combustion chamber, and the change in heat received by each carbonization chamber after the perturbation is transferred through the furnace wall heat transfer and coal heat absorption links is calculated; the first... Changes in heat supply to each combustion chamber The cause of the first Each carbonization chamber is affected by heat changes The ratio is defined as the partial derivative. This partial derivative quantitatively describes the first... The combustion chamber for the first The degree of impact on the heating of each carbonization chamber, traversing the entire process. One combustion chamber and Each carbonization chamber, with its partial derivatives arranged in matrix form, yields a... OK Sensitivity matrix of columns ;

[0059] For the whole furnace One combustion chamber and One carbonization chamber, asynchronous propagation matrix For one A square matrix of dimensions.

[0060] Furthermore, the transformation matrix in S102 The method for obtaining this information is as follows: Define the change in heat received by each carbonization chamber. The change in the asynchrony of the carbonization chamber They are negatively correlated, and the proportionality coefficient is set to... , indicating the first The change in asynchrony caused by the change in heat received by a unit carbonization chamber, with dimensions of time divided by heat, is the amount of heat received by each carbonization chamber in the entire furnace. The transformation matrix is ​​obtained by arranging the elements. .

[0061] Furthermore, the dynamic adjustment rules for the control cycle are as follows:

[0062] when This indicates that the asynchrony is converging rapidly, the current control direction is correct and the effect is significant. At this point, the current control cycle should be maintained to continue the momentum of rapid convergence.

[0063] when This indicates that the asynchrony is converging at a moderate rate, the control effect is normal, and the current control cycle remains unchanged.

[0064] when This indicates that the improvement effect of asynchrony has become weak and has approached or reached the steady-state neighborhood under the current control precision. The marginal benefit of continuing to control at the same frequency is decreasing, so the control period can be extended according to the preset adjustment ratio factor.

[0065] when This indicates that the asynchronous modulus has rebounded and increased, which may be due to incorrect control direction, digital twin model mismatch, or abnormal disturbances in production conditions. The control cycle can be shortened by a preset adjustment ratio factor.

[0066] Compared with the prior art, the beneficial effects of the present invention are:

[0067] The present invention provides a digital twin-based method for coordinated control of the temperature field in the coking process of a coke oven carbonization chamber. Through asynchronous degree propagation dynamic modeling, bi-objective Pareto online negotiation, and dynamic triggering control cycle adjustment of asynchronous degree change rate, it effectively improves the quality of coordinated control of the temperature field of the entire furnace and solves the problem of balancing the synchronous maturation of the coking process in the carbonization chamber with energy consumption optimization. Attached Figure Description

[0068] Figure 1 This is a flowchart of the collaborative control method according to an embodiment of the present invention;

[0069] Figure 2 This is a flowchart of the method in step 1 of this embodiment of the invention;

[0070] Figure 3 This is a flowchart of step 2 in an embodiment of the present invention. Detailed Implementation

[0071] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0072] Please see Figure 1-3 The method for coordinated temperature field control of the coking process in the coking chamber of a coke oven based on digital twins, provided in this embodiment of the invention, mainly includes the following steps:

[0073] Step 1: Construct a digital twin of the heating association topology between the combustion chamber and the carbonization chamber, calculate the remaining coking time of each carbonization chamber under the current heating command of the combustion chamber in real time, and generate the asynchronous degree vector of the coking process of the whole furnace; further, based on the heating association topology, establish an asynchronous degree propagation dynamic model.

[0074] This step is the foundation of the entire collaborative control method. Its purpose is to establish a high-fidelity online simulation environment that can accurately map the thermal state of the entire coke oven and the coking process of each carbonization chamber in real time. Based on this, a direct controlled variable for subsequent collaborative control, the asynchronous degree vector of the coking process of the entire oven, is defined and calculated. The specific method is as follows:

[0075] First, a digital twin of the heating-related topology is constructed, which is a two-way coupled model. At the physical mechanism level, a combustion chamber in a coke oven usually provides heat to the adjacent carbonization chambers on both sides simultaneously, while a carbonization chamber is also heated by the combustion chambers on both sides, forming an interwoven mesh heating topology. Based on this topology, the digital twin establishes a correlation matrix between the combustion chamber and the carbonization chamber, explicitly quantifying the heating and heat-receiving relationship between the combustion chamber and the carbonization chamber of the entire furnace.

[0076] In this embodiment, the bidirectional coupling characteristic of the digital twin is manifested in two aspects: on the one hand, the heat generated by the combustion of gas in the combustion chamber is transferred to the carbonization chamber through the furnace wall, driving the coal to heat up and undergo pyrolysis; on the other hand, the pyrolysis process of the coal in the carbonization chamber, in turn, affects the heat transfer state of the furnace wall. Specifically, as the coking process proceeds, volatile matter in the coal is continuously released, and the thermal conductivity, density, and volume of the coal are dynamically updated with changes in temperature and residual volatile matter. These changes in the physical properties of the coal will alter the heat transfer conditions on the surface of the furnace wall on the carbonization chamber side, thereby affecting the temperature inside the furnace wall. The digital twin of this invention incorporates the gradient distribution of heat transfer and the heat flux density transferred from the combustion chamber to the carbonization chamber. By introducing the dynamic thermal resistance of the furnace wall, which changes in real time with the coking process, the digital twin couples the coal pyrolysis kinetic domain in the carbonization chamber with the turbulent combustion domain in the combustion chamber, and simultaneously calculate the state evolution of both. Compared with the traditional coke oven thermal model, which only considers the unidirectional heat transfer from the combustion chamber to the carbonization chamber, this bidirectional coupling mechanism enables the digital twin to realistically reproduce the physical nature of the interaction between the carbonization chamber and the combustion chamber in the actual coke oven, providing a fidelity basis for subsequent co-calculation.

[0077] Specifically, on the carbonization chamber side, the coking process of the coal is controlled by a set of partial differential equations describing the coupled evolution of the coal temperature field and the volatile matter analysis process. The temperature field of the coal inside the carbonization chamber satisfies the unsteady-state heat conduction equation containing an internal heat source, expressed as:

[0078]

[0079] in, The temperature field of the coal in the carbonization chamber is a function of spatial coordinates and time; This indicates the apparent density of the coal. This indicates the specific heat capacity of the coal. Indicates the thermal conductivity of the coal; This represents the thermal conductivity term, which describes the process of heat diffusion within the coal. This represents the pyrolysis reaction source term, that is, the contribution of heat absorption or release during the pyrolysis of coal to the temperature field;

[0080] The pyrolysis reaction source term is directly related to the volatile matter release rate. The volatile matter release process is described by the equation of state of the residual volatile matter in the coal, and its evolution follows a first-order kinetic model, expressed as:

[0081]

[0082] in, It indicates the residual volatile matter in coal, that is, the mass of volatile matter that has not yet been released from a unit mass of coal; This indicates the rate of change of residual volatile matter over time. The pre-exponential factor representing the pyrolysis reaction; This represents the activation energy of a pyrolysis reaction. Represents the universal gas constant; Indicates the temperature of the coal;

[0083] The two formulas above constitute the coupled state equations of coal volatile matter residue and temperature. The coupling relationship between the two is reflected in the fact that the temperature field drives the precipitation rate of volatile matter, while the thermal effect in the volatile matter precipitation process reacts on the temperature field through the pyrolysis reaction source term.

[0084] On the combustion chamber side, the heat generated by the combustion of gas is transferred to the carbonization chamber through the furnace wall. The furnace wall serves as a heat transfer channel between the combustion chamber and the carbonization chamber, and its heat transfer characteristics are characterized by the dynamic thermal resistance of the furnace wall. This thermal resistance is not a constant value, but changes dynamically with the operating state of both sides of the furnace wall. The one-dimensional unsteady-state heat conduction control equation along the thickness direction inside the furnace wall is expressed as:

[0085]

[0086] in, Indicates the density of the furnace wall material; This indicates the specific heat capacity of the furnace wall material; The temperature field inside the furnace wall is represented by the furnace wall thickness coordinates as a function of time. Indicates the thermal conductivity of the furnace wall material; Coordinates indicating the direction of furnace wall thickness;

[0087] The boundary conditions on both sides of the furnace wall are coupled to the combustion chamber and the carbonization chamber, respectively. On the combustion chamber side, the furnace wall surface receives convective and radiative heat transfer from the high-temperature flue gas in the combustion chamber. On the carbonization chamber side, the furnace wall surface transfers heat to the coal. This double-sided coupled boundary condition makes the dynamic thermal resistance of the furnace wall a real-time variable connecting the heating state of the combustion chamber and the heat absorption state of the carbonization chamber.

[0088] To ensure that the simulation state of the digital twin in this embodiment of the invention maintains high fidelity consistency with the actual coking process, a data assimilation algorithm is used for real-time correction. The measured values ​​of the furnace wall temperature at multiple points are compared with the temperature output values ​​of the digital twin at the corresponding spatial locations to obtain a residual sequence. Based on this residual sequence, an online ensemble Kalman filter algorithm is used to absorb the system state information contained in the observations and output unbiased correction values ​​for the key physical property parameters of the model. These key physical property parameters include the virtual thermal diffusivity of the coal and the interfacial heat transfer coefficient of the furnace wall. The virtual thermal diffusivity of the coal determines the speed of heat propagation within the coal, directly affecting the accuracy of the temperature field extrapolation. The interfacial heat transfer coefficient of the furnace wall determines the heat transfer efficiency between the combustion chamber side furnace wall and the carbonization chamber side furnace wall, affecting the mapping relationship between heating and receiving. Through continuous online correction of the above parameters, the output of the digital twin gradually converges to a state consistent with the actual furnace wall temperature sequence, thereby constructing an online simulation environment that accurately maps the actual coking process.

[0089] Based on the high-fidelity online simulation environment already constructed in the digital twin, the asynchronous degree vector of the coking process in each carbonization chamber of the entire furnace is calculated in real time. For each carbonization chamber in the coking cycle, the output state of the online simulation environment at the current moment of the carbonization chamber is used as the starting point, and a forward extrapolation is performed in the virtual spacetime. The three-dimensional temperature field distribution and volatile matter residue spatial distribution of the carbonization chamber at the current moment are used as the initial state, and the current heating command continuation of each combustion chamber associated with the carbonization chamber is used as the boundary condition. The physicochemical equations of the coking process are integrated along the time axis by the digital twin. The extrapolation continues until the carbonization chamber is in the coking cycle. The calculated state of the carbonization chamber simultaneously satisfies the preset maturity critical condition; wherein, the maturity critical condition is a dual-condition coupled criterion: first, the temperature in the central region of the coal in the carbonization chamber reaches the preset coke maturity temperature threshold; second, the residual volatile matter in this region decreases to the preset volatile matter removal completion threshold; only when both the temperature condition and the component condition are simultaneously satisfied is the carbonization chamber determined to have reached the coking maturity state; the moment when the carbonization chamber reaches the maturity state in the simulation is recorded as the virtual maturity moment, and the difference between it and the current moment is used to obtain the remaining coking time that the carbonization chamber needs to undergo to complete coking under the current heating command; for the first One carbonization chamber, remaining coking time Represented as:

[0090]

[0091] in, Indicates the first The virtual moment when a carbonization chamber reaches a mature state in the digital twin simulation; Indicates the current moment;

[0092] After obtaining the remaining coking time for each carbonization chamber, calculate the corresponding degree of asynchrony; for the first... For each carbonization chamber, the asynchrony is defined as the deviation between the remaining coking time and the planned remaining coking time. The planned remaining coking time is the total planned coking time of the carbonization chamber from the coal loading time to the predetermined coking push time, minus the actual coking time elapsed from the coal loading time to the current time. It is expressed as:

[0093]

[0094] in, Indicates the first The remaining coking time for each carbonization chamber; Indicates the first The total planned coking time for each carbonization chamber from the time of coal loading to the scheduled time of coking push; Indicates the first The actual coking time that has elapsed from the moment the coal was loaded into the carbonization chamber to the present moment;

[0095] Based on this, the first Asynchrony of each carbonization chamber It is expressed as the difference between the remaining coking time and the planned remaining coking time.

[0096] In the above embodiments, the physical meaning of asynchrony is specifically as follows: when When the value is positive, it indicates that the carbonization chamber is expected to reach maturity in a longer time than planned under the current heating conditions. The carbonization chamber is in a state of coking lag and requires increased heating to accelerate its coking process. When the value is negative, it indicates that the carbonization chamber is expected to reach maturity in a shorter time than planned under the current heating conditions. This means the carbonization chamber is in an advanced coking state, and heating can be appropriately reduced to save energy and transfer heat to lagging carbonization chambers. When the value approaches zero, it indicates that the coking process in the carbonization chamber is basically synchronized with the coking plan.

[0097] The whole furnace The asynchrony of each carbonization chamber within a coking cycle is arranged according to its serial number, forming a... The asynchronous degree vector of the whole furnace coking process The asynchronous degree vector condenses the spatiotemporal incoordination of the coking process in each carbonization chamber of the whole furnace into a quantitative index under a unified time dimension. The larger the magnitude of the asynchronous degree vector, the more serious the asynchronous degree of the coking process between each carbonization chamber of the whole furnace, and the more necessary it is to implement coordinated control.

[0098] Further, based on the heating-related topology, an asynchronous propagation dynamic model is established in the above steps. This model quantitatively describes a physical law: when the heating supply is adjusted to any combustion chamber, the adjustment not only changes the asynchronous degree of its directly associated carbonization chamber, but also propagates the change in asynchronous degree to the indirectly associated carbonization chamber through heating topology coupling, forming a chain propagation effect. The specific model construction process is as follows:

[0099] First, based on the sensitivity analysis of the heating-related topology digital twin, an asynchronous degree propagation matrix is ​​constructed; then, a vector is defined. This represents the current heating command for all combustion chambers in the entire furnace, vector. This represents the actual heat received by each carbonization chamber in the entire furnace under the current heating command; at the current operating point of the digital twin, a small positive perturbation is applied to the heat supply of each combustion chamber, and the change in heat received by each carbonization chamber after the perturbation is transferred through the furnace wall heat transfer and coal heat absorption links is calculated; the first... Changes in heat supply to each combustion chamber The cause of the first Each carbonization chamber is affected by heat changes The ratio is defined as the partial derivative. This partial derivative quantitatively describes the first... The combustion chamber for the first The degree of impact on the heating of each carbonization chamber, traversing the entire process. One combustion chamber and Each carbonization chamber, by arranging the above partial derivatives in matrix form, yields a... OK Sensitivity matrix of columns ;

[0100] For the whole furnace One combustion chamber and One carbonization chamber, asynchronous propagation matrix It is A square matrix of dimensionality describes the mutual influence between the asynchronous degrees of each carbonization chamber; the matrix The Line number Column elements Indicates the first When a unit asynchronous change occurs in the carbonization chamber, the heat transfer coupling through the shared combustion chamber affects the first... The propagation effect of the asynchronous degree of each carbonization chamber; the calculation method of the asynchronous degree matrix is ​​as follows: apply virtual heat supply disturbance to the combustion chamber associated with each carbonization chamber in sequence, simulate the steady-state effect of the disturbance on the asynchronous degree of all carbonization chambers through digital twin simulation, and fill the corresponding positions of the matrix after normalizing each influence coefficient. The main diagonal elements are zero (the asynchronous degree change of the carbonization chamber itself has been directly calculated), while the off-diagonal elements reflect the asynchronous degree propagation coupling strength between adjacent carbonization chambers in the heating topology;

[0101] Based on this, a discrete-time form of the asynchronous propagation dynamics equation is established, in the first... Each control cycle adjusts the heat supply applied to each combustion chamber. Subsequently, the expected evolution of the asynchronous degree vector of the entire furnace follows the following propagation dynamics equation, expressed as:

[0102]

[0103] in, This represents the asynchrony vector of the k-th control cycle; This indicates the change in the asynchrony of each carbonization chamber directly caused by the adjustment of the heat supply. This represents the asynchronous propagation term, i.e. The direct asynchronous degree change included is propagated to the adjacent carbonization chamber through the heating topology, resulting in a secondary asynchronous degree change.

[0104] The relationship between the change in asynchrony and the adjustment of heating supply is as follows:

[0105]

[0106] in, The transformation matrix is ​​represented by the following: Based on the residual coking time derivation mechanism, the residual coking time of the carbonization chamber is closely related to its heat input. When the heat input to the carbonization chamber increases, the coal heating rate accelerates, the time required to reach maturity is shortened, and the asynchrony decreases in the negative direction; conversely, when the heat input decreases, the asynchrony increases in the positive direction. This relationship is expressed as the partial derivative of the asynchrony vector with respect to the heat input vector. Under the first-order approximation, the change in heat input to each carbonization chamber is... The change in the asynchrony of the carbonization chamber They are negatively correlated, and the proportionality coefficient is set to... , indicating the first The change in asynchrony caused by the change in heat received by a unit carbonization chamber, with dimensions of time divided by heat, is the amount of heat received by each carbonization chamber in the entire furnace. Arranged as a transformation matrix .

[0107] The physical significance of the propagation dynamics model lies in the fact that when a lagging carbonization chamber accelerates coking (reducing asynchrony) due to increased heat supply from its associated combustion chamber, the increased heat supply from this combustion chamber simultaneously causes its adjacent carbonization chamber on the other side to also receive more heat, resulting in a change in the asynchrony of the adjacent carbonization chamber. If the adjacent carbonization chamber itself is in a leading state, this propagation effect will exacerbate its leading degree; if the adjacent carbonization chamber is also in a lagging state, it will accelerate its convergence. Therefore, the effect of heat supply adjustment is not an isolated event, but a dynamic process propagating in the heating topology. The propagation dynamics model explicitly quantifies this complex chain propagation effect into mathematical constraints, providing dynamic constraints that truly reflect the physical coupling characteristics.

[0108] In addition, it is necessary to calculate the total energy consumption of the entire furnace, which is defined as the total energy consumption of the entire furnace during the current control cycle. The total gas consumption of each combustion chamber is obtained by summing the heat supplied by each combustion chamber.

[0109] Step 2: Construct an optimization architecture with the asynchronous degree vector and total furnace energy consumption as objectives, and use the asynchronous degree propagation dynamics model as dynamic constraints to solve the incremental commands of the combustion chamber through an online Pareto negotiation mechanism.

[0110] This step is the core decision-making link of the entire collaborative control method. Unlike existing technologies that only minimize temperature deviation or a single asynchronous index as the optimization objective, this step takes the total gas consumption of all combustion chambers in the entire furnace as an optimization objective of equal importance to the asynchronousness of the coking process of the entire furnace. It constructs a dual-objective optimization architecture of asynchronousness and energy consumption, and solves the incremental commands of each combustion chamber through an online Pareto negotiation mechanism under the dynamic constraints of the asynchronousness propagation dynamics model. The specific method is as follows:

[0111] First, construct a dual-objective optimization architecture and define the first optimization objective. To minimize the predicted modulus of the asynchronous degree vector after applying a heat supply adjustment; unlike existing technologies that only consider the direct impact of heat supply adjustment on the asynchronous degree of each carbonization chamber, this step uses the asynchronous degree propagation kinetic model as the dynamic constraint for this objective. According to this model, when a heat supply adjustment is applied to each combustion chamber, the predicted value of the adjusted asynchronous degree vector not only includes the direct effect of the heat supply adjustment, but also the secondary effect generated after the effect is propagated to the adjacent carbonization chambers through the heat supply topology. The direct effect is obtained by mapping the heat supply adjustment through sensitivity and transformation matrix.

[0112] According to the asynchronous propagation dynamics equation, the predicted value E of the adjusted asynchronous degree vector is the current asynchronous degree vector. direct effects With propagation effect The combination of the three factors leads to the first optimization objective. Represented as:

[0113]

[0114] Define the second optimization objective To minimize the total energy consumption of the entire furnace after applying heating supply adjustments, the total energy consumption of the entire furnace is characterized by the total gas consumption of each combustion chamber. Then, the predicted value of the total energy consumption of the entire furnace. Represented as:

[0115]

[0116] in, This represents the total gas consumption of all combustion chambers in the entire furnace during the current control cycle; Indicates the first The change in heat supply to each combustion chamber; therefore, the second optimization objective Represented as:

[0117]

[0118] The constraints of this dual-objective optimization architecture include two categories. The first category is process safety constraints, which specifically include: the temperature of each combustion chamber does not exceed the preset upper limit of the combustion chamber temperature; the coke maturity of each carbonization chamber in the predicted final state is not lower than the preset lower limit of maturity; and the rate of change of heat supply between adjacent combustion chambers does not exceed the preset upper limit of the adjustment rate to prevent excessive thermal stress on the furnace wall. The second category is dynamic constraints, namely the asynchronous degree propagation dynamic equation. This constraint ensures that the optimization solution process truly reflects the physical laws of asynchronous degree propagation in the heating topology.

[0119] After the optimized architecture is built, the incremental commands for the combustion chamber are solved through an online Pareto negotiation mechanism. The core principle of this mechanism is to calculate online within each control cycle. and The Pareto front between the two objectives is identified, and then, based on preset negotiation rules, the optimal operating point is automatically selected to achieve the best balance between improving asynchronicity and controlling energy consumption. The specific process is as follows:

[0120] In the current control cycle, taking the current asynchrony vector and current total energy consumption as starting points, and the change in heat supply of each combustion chamber as decision variables, the bi-objective optimization problem is transformed into a series of single-objective optimization sub-problems for solution: the second optimization objective... Convert to constraints and set a series of different energy consumption constraint levels. , , For each Pareto front sampling point, the preset number of sampling points is used. To minimize To optimize the objective, energy consumption constraints are added to the existing process safety constraints and asynchronous propagation dynamics constraints. The optimal solution to this subproblem corresponds to a non-dominated solution. At this solution, it is impossible to further improve the asynchronicity without increasing energy consumption. (This can be achieved by traversing...) At different energy consumption constraint levels, we obtain A non-dominated solution, The non-dominated solutions are arranged in the objective space according to... and The values ​​are arranged to form the Pareto front curve for the current control cycle.

[0121] The Pareto front curve visually illustrates the trade-off between improved asynchronicity and energy consumption control. Each point on the front represents a feasible heating adjustment scheme; the closer to the front, the more feasible the scheme. One end of the shaft indicates a better improvement in asynchrony, but the total energy consumption is relatively higher; the closer to the shaft, the better. One end of the axis indicates that the total energy consumption is low but the improvement effect of asynchrony is relatively limited. There is no absolutely optimal solution on the front that is better than other points in both objectives at the same time. The selection of the optimal operation point depends on the trade-off preference between the two objectives.

[0122] After obtaining the Pareto frontier, the optimal operating point is selected according to the pre-set negotiation rules, specifically the principle of maximizing marginal benefits. That is, the operating point on the Pareto frontier that maximizes the ratio of the decrease in the asynchronous degree vector magnitude to the increase in total energy consumption is selected, expressed as:

[0123]

[0124] in, This represents the index of the optimal operation point selected during the negotiation on the Pareto front. Let represent the set of all nondominated solutions on the Pareto front; To represent positive numbers, to prevent the denominator from being zero;

[0125] The cost-effectiveness of each candidate operation point is evaluated one by one on the Pareto front. The solution that brings the greatest improvement in asynchrony with the increase in energy consumption per unit is given priority. When there is an operation point on the Pareto front whose increase in energy consumption is negative (that is, it can improve asynchrony while reducing energy consumption), the ratio of this point will be very large, and the negotiation mechanism will give priority to the solution that can reduce energy consumption and achieve coordination. When all solutions that can improve asynchrony require an increase in energy consumption, the negotiation mechanism selects the solution with the highest cost-effectiveness to obtain the greatest improvement in asynchrony with the least energy consumption cost. When the asynchrony is close to the zero neighborhood and the improvement space is limited, the negotiation mechanism will automatically tend to select the operation point with lower energy consumption to avoid paying unnecessary energy consumption costs for small improvements in asynchrony.

[0126] Finally, the change in heat supply corresponding to the optimal operating point selected through negotiation is used as the incremental command for each combustion chamber. .

[0127] Step 3: The incremental command is superimposed with the current heating command and then sent out for execution. The control cycle is dynamically adjusted and closed-loop iteration is performed using the change rate of the asynchronous degree vector as a trigger mechanism to drive the asynchronous degree vector to converge to the zero neighborhood.

[0128] This step superimposes and executes the incremental command with the current heating command, while introducing the rate of change of the asynchronous degree vector magnitude as a dynamic trigger condition for the control cycle, replacing the fixed control cycle approach in existing technologies. This dynamic triggering mechanism can adaptively adjust the control frequency according to the actual convergence speed of the asynchronous degree of the entire furnace coking process. When the asynchronous degree is large and the convergence space is sufficient, it can quickly adjust with a shorter cycle to accelerate convergence; when the asynchronous degree is close to steady state, it can automatically extend the cycle to avoid over-control and oscillation, reducing equipment wear and additional energy consumption caused by frequent actions of the actuator. The specific method is as follows:

[0129] First, the incremental command is superimposed with the current heating command of each combustion chamber to obtain the updated whole-furnace heating command vector. , is represented as:

[0130]

[0131] Expanding to each combustion chamber, the first The updated heating command for each combustion chamber is expressed as follows:

[0132]

[0133] After generating the updated overall furnace heating command, it is sent to the actuators of each combustion chamber, driving the gas regulating valves of each combustion chamber to perform corresponding actions. The heating supply of each combustion chamber is adjusted accordingly, and the heating state of each carbonization chamber in the whole furnace changes accordingly. The coking process continues to evolve under the new heating conditions.

[0134] After a control cycle is completed, the rate of change of the asynchronous degree vector magnitude is calculated, which serves as the trigger for whether to adjust in the next control cycle. Specifically:

[0135] After the current control cycle is completed, step 1 is executed again. The digital twin recalculates the remaining coking time of each carbonization chamber based on the updated overall furnace heating status, generating a new overall furnace asynchronous vector. The magnitude of the asynchronous vector for this control cycle is denoted as... Let the magnitude of the asynchronous degree vector of the previous control cycle be denoted as Rate of change of the magnitude of the asynchronous degree vector Represented as:

[0136]

[0137] when When the value is negative, it indicates that the adjustment of the heating supply has reduced the asynchronous modulus of the entire furnace compared with the previous cycle, the adjustment direction is correct and the effect is significant, and the larger the absolute value, the faster the convergence speed.

[0138] when When the value is a small negative value close to zero or zero, it indicates that the change in the asynchronous modulus is weak and may have approached the steady-state neighborhood. The marginal benefit of continuing to regulate at the same frequency is diminishing.

[0139] when When the value is positive, it indicates that the asynchronous magnitude has rebounded and increased, which may indicate a deviation in the control direction, model mismatch, or external disturbance, requiring faster detection and response.

[0140] Based on the rate of change of the asynchronous degree vector magnitude, the control cycle is dynamically adjusted; three core threshold parameters are set: upper limit threshold... Lower threshold and adjustment ratio factor ; To determine whether the asynchronous degree is converging rapidly, a small negative number is usually used, such as -0.1, which indicates that the asynchronous degree magnitude has decreased by more than 10% compared to the previous period, indicating rapid convergence. Used to determine whether the change in asynchrony has entered a weak region, it is usually taken as a negative number close to zero, such as -0.02, indicating that when the magnitude decrease is less than 2%, convergence is considered to have leveled off; adjusting the scaling factor The adjustment range used to control the cycle is usually between 1.2 and 1.5 to ensure a smooth transition during the adjustment process.

[0141] The specific rules for the dynamic adjustment of the above control cycle are as follows:

[0142] when This indicates that the asynchrony is converging rapidly, the current control direction is correct and the effect is significant. At this point, the current control cycle should be maintained to continue the momentum of rapid convergence.

[0143] when This indicates that the asynchrony is converging at a moderate rate, the control effect is normal, and the current control cycle remains unchanged.

[0144] when This indicates that the improvement effect of asynchrony has become weak and has approached or reached the steady-state neighborhood under the current control accuracy. The marginal benefit of continuing to control at the same frequency is decreasing. At this time, the control cycle is extended according to the preset adjustment ratio factor to reduce unnecessary heat supply fine-tuning actions, reduce equipment wear and regulation energy consumption caused by frequent operation of gas regulating valve, and avoid system oscillation caused by excessive control.

[0145] when This indicates that the asynchronous degree modulus has rebounded and increased, which may be due to incorrect control direction, mismatch of digital twin model, or abnormal disturbance of production conditions. In this case, the control cycle is shortened by the preset adjustment ratio factor so that the state change can be detected quickly, the control direction can be recalculated and corrected in a shorter cycle, and the response capability to abnormal conditions can be enhanced.

[0146] The mechanism based on the dynamic triggering control cycle of the asynchronous degree change rate in this embodiment has significant engineering advantages. In the early and middle stages of the coking process, the asynchronous degree of each carbonization chamber is relatively large, and the adjustment of the heat supply has a significant effect on improving the asynchronous degree. The value is usually a negative number with a large absolute value. High-frequency regulation with a short control cycle quickly drives the asynchronous degree vector to converge to zero. As the coking process progresses and the asynchronous degree continues to decrease, the marginal effect of adjusting the heat supply diminishes. As the value gradually approaches zero, the control cycle is automatically extended, entering a steady-state maintenance phase of low-frequency fine-tuning to avoid over-control and oscillations. Throughout the entire coking cycle, the adaptability of the control cycle ensures the efficiency and stability of the regulation process.

[0147] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for coordinated control of the temperature field during the coking process in a coke oven carbonization chamber based on digital twins, characterized in that, Includes the following steps: S1: Construct a digital twin of the heating association topology between the combustion chamber and the carbonization chamber, calculate the remaining coking time of each carbonization chamber under the current heating command of the combustion chamber in real time, generate the asynchronous degree vector of the coking process of the whole furnace, and establish an asynchronous degree propagation dynamic model based on the heating association topology; S2: Construct an optimization architecture with asynchronous degree vector and total furnace energy consumption as objectives, and use the asynchronous degree propagation dynamics model as dynamic constraint to solve the incremental command of the combustion chamber through an online Pareto negotiation mechanism; S3: The incremental command is superimposed with the current heating command and then issued for execution. The control cycle is dynamically adjusted and closed-loop iteration is performed using the rate of change of the asynchronous degree vector magnitude as the trigger mechanism, driving the asynchronous degree vector to converge to the zero neighborhood.

2. The method for coordinated control of temperature field in the coking process of a coke oven carbonization chamber based on digital twin as described in claim 1, characterized in that: The heating-related topology digital twin constructed in S1 is a two-way coupled model: On the carbonization chamber side, the coking process of the coal is controlled by a set of partial differential equations describing the coupled evolution of the coal temperature field and the volatile matter analysis process; among them, the temperature field of the coal inside the carbonization chamber satisfies the unsteady-state heat conduction equation containing an internal heat source, expressed as: In the formula, The temperature field of the coal in the carbonization chamber is a function of spatial coordinates and time; Indicates the apparent density of coal; This indicates the specific heat capacity of the coal. Indicates the thermal conductivity of the coal; This represents the thermal conductivity term, which describes the process of heat diffusion within the coal. This represents the pyrolysis reaction source term, that is, the contribution of heat absorption or release during the pyrolysis of coal to the temperature field; The pyrolysis reaction source term is directly related to the volatile matter precipitation rate. The volatile matter precipitation process is described by the equation of state of the residual volatile matter in the coal, and its evolution follows a first-order kinetic model, expressed as: In the formula, It indicates the residual volatile matter in coal, that is, the mass of volatile matter that has not yet been released from a unit mass of coal; This indicates the rate of change of residual volatile matter over time. The pre-exponential factor representing the pyrolysis reaction; This represents the activation energy of a pyrolysis reaction. Represents the universal gas constant; Indicates the temperature of the coal; On the combustion chamber side, the heat generated by the combustion of gas is transferred to the carbonization chamber through the furnace wall. The furnace wall serves as a heat transfer channel between the combustion chamber and the carbonization chamber, and its heat transfer characteristics are characterized by the dynamic thermal resistance of the furnace wall. That is, the thermal resistance changes dynamically with the operating state of both sides of the furnace wall. A one-dimensional unsteady-state heat conduction control equation along the thickness direction inside the furnace wall is established: In the formula, Indicates the density of the furnace wall material; This indicates the specific heat capacity of the furnace wall material; The temperature field inside the furnace wall is represented by the furnace wall thickness coordinates as a function of time. Indicates the thermal conductivity of the furnace wall material; The coordinates represent the direction of furnace wall thickness.

3. The method for coordinated control of temperature field in the coking process of a coke oven carbonization chamber based on digital twin as described in claim 2, characterized in that: S1 also includes real-time correction of the heating-related topology digital twin using a data assimilation algorithm; wherein, the remaining coking time is the difference between the current time and the time when the carbonization chamber in the heating-related topology digital twin reaches a mature state; for the first One carbonization chamber, remaining coking time Represented as: In the formula, Indicates the first The virtual moment when a carbonization chamber reaches a mature state in the digital twin simulation; Indicates the current time; After obtaining the remaining coking time for each carbonization chamber, calculate the corresponding degree of asynchrony; for the first... For each carbonization chamber, the asynchrony is defined as the deviation between the remaining coking time and the planned remaining coking time. The planned remaining coking time is the total planned coking time of the carbonization chamber from the coal loading time to the predetermined coking push time, minus the actual coking time elapsed from the coal loading time to the current time. It is expressed as: In the formula, Indicates the first The remaining coking time for each carbonization chamber; Indicates the first The total planned coking time for each carbonization chamber from the time of coal loading to the scheduled time of coking push; Indicates the first The actual coking time that has elapsed from the moment the coal was loaded into the carbonization chamber to the present moment; Based on the difference between the remaining coking time and the planned remaining coking time, the asynchrony of each carbonization chamber is obtained, and the entire furnace is then... The asynchrony of each carbonization chamber within a coking cycle is arranged according to its serial number, forming a... Asynchronous degree vector of the whole furnace coking process .

4. The method for coordinated control of temperature field in the coking process of a coke oven carbonization chamber based on digital twin as described in claim 3, characterized in that: The specific process of constructing the asynchronous propagation dynamics model in S1 is as follows: S101: Perform sensitivity analysis on the digital twin of the heating-related topology and obtain the sensitivity matrix; S102: Based on sensitivity analysis, construct the asynchronous propagation matrix A, and then establish the asynchronous propagation dynamic equation, expressed as: in, This represents the asynchrony vector of the k-th control cycle; This indicates the change in asynchrony directly caused by the adjustment of the heating supply; Indicates the asynchronous propagation term; The relationship between the change in asynchrony and the adjustment of heating supply is as follows: In the formula, This represents the transformation matrix.

5. The method for coordinated control of temperature field in the coking process of a coke oven carbonization chamber based on digital twin as described in claim 4, characterized in that: The optimized architecture in S2 includes the first optimization objective. Second optimization objective Among them, the first optimization objective Defined as minimizing the modulus of the asynchronous degree vector prediction after applying heat supply adjustment, expressed as: In the formula, This represents the magnitude of the predicted value of the adjusted asynchronous degree vector; Second optimization objective Defined as minimizing the total energy consumption of the entire furnace after applying heat supply adjustment, the total energy consumption of the entire furnace is characterized by the total gas consumption of each combustion chamber, after applying heat supply adjustment. Then, the predicted value of the total energy consumption of the entire furnace. Represented as: In the formula, This represents the total gas consumption of all combustion chambers in the entire furnace during the current control cycle; Indicates the first The change in heat supply to each combustion chamber; then the second optimization objective Represented as: 。 6. The method for coordinated control of temperature field in the coking process of a coke oven carbonization chamber based on digital twin as described in claim 5, characterized in that: The method for obtaining incremental instructions in S2 is as follows: S201: The second optimization objective Convert to constraints, and set M energy consumption constraint levels. ; S202: Based on each energy consumption constraint level To minimize the primary optimization objective, additional constraints are added. Obtain M non-dominated solutions; S203: Arrange the M non-dominated solutions in the objective space according to... and The values ​​are arranged to form the Pareto front curve of the current control cycle; S204: Based on the principle of maximizing marginal benefits, the optimal operating point is selected, expressed as: In the formula, This indicates the selected optimal operation point; Let represent the set of all nondominated solutions on the Pareto front; This indicates that the constraint level is positive; the change in heat supply corresponding to the selected optimal operating point is used as the incremental instruction. .

7. The method for coordinated control of temperature field in the coking process of a coke oven carbonization chamber based on digital twin as described in claim 6, characterized in that: Specifically, in S3, the incremental command is superimposed with the current heating command of each combustion chamber to obtain the updated overall furnace heating command vector. , is represented as: Expanding to each combustion chamber, the first The updated heating command for each combustion chamber is expressed as follows: After a control cycle is completed, calculate the rate of change of the asynchronous degree vector magnitude. : In the formula, This is the magnitude of the asynchronous degree vector for this control cycle. The magnitude of the asynchronous degree vector of the previous control cycle; By setting three threshold parameters, namely the upper limit threshold... Lower threshold and adjustment ratio factor Combined with the rate of change of the asynchronous degree vector magnitude Dynamic adjustment of the execution control cycle.

8. The method for coordinated control of temperature field in the coking process of a coke oven carbonization chamber based on digital twin as described in claim 4, characterized in that: The method for obtaining the sensitivity matrix in S101 is as follows: Set the vector This represents the current heating command for all combustion chambers in the entire furnace, vector. This represents the actual heat received by each carbonization chamber of the entire furnace under the current heating command; at the current operating point of the digital twin, a small positive perturbation is applied to the heat supply of each combustion chamber, and the change in heat received by each carbonization chamber after the perturbation is transferred through the furnace wall heat transfer and coal heat absorption links is calculated; the first... Changes in heat supply to each combustion chamber The cause of the first Each carbonization chamber is affected by heat changes The ratio is defined as the partial derivative. This partial derivative quantitatively describes the first... The combustion chamber for the first The degree of impact on the heating of each carbonization chamber, traversing the entire... One combustion chamber and Each carbonization chamber, with its partial derivatives arranged in matrix form, yields a... OK Sensitivity matrix of columns ; For the whole furnace One combustion chamber and One carbonization chamber, asynchronous propagation matrix For one A square matrix of dimensions.

9. The method for coordinated control of temperature field in the coking process of a coke oven carbonization chamber based on digital twin as described in claim 4, characterized in that: Transformation matrix in S102 The method for obtaining this information is as follows: Define the change in heat received by each carbonization chamber. The change in the asynchrony of the carbonization chamber They are negatively correlated, and the proportionality coefficient is set to... , indicating the first The change in asynchrony caused by the change in heat received by a unit carbonization chamber, with dimensions of time divided by heat, is the amount of heat received by each carbonization chamber in the entire furnace. The transformation matrix is ​​obtained by arranging the elements. .

10. The method for coordinated control of temperature field in the coking process of a coke oven carbonization chamber based on digital twin as described in claim 7, characterized in that: The dynamic adjustment rules for the control cycle are as follows: when This indicates that the asynchrony is converging rapidly, the current control direction is correct and the effect is significant. At this point, the current control cycle should be maintained to continue the momentum of rapid convergence. when This indicates that the asynchrony is converging at a moderate rate, the control effect is normal, and the current control cycle remains unchanged. when This indicates that the improvement effect of asynchrony has become weak and has approached or reached the steady-state neighborhood under the current control precision. The marginal benefit of continuing to control at the same frequency is decreasing, so the control period can be extended according to the preset adjustment ratio factor. when This indicates that the asynchronous modulus has rebounded and increased, which may be due to incorrect control direction, digital twin model mismatch, or abnormal disturbances in production conditions. The control cycle can be shortened by a preset adjustment ratio factor.

Images