A method for controlling the temperature of a coke oven vertical flue

Abstract

This invention relates to the field of ABC technology and provides a method for controlling the temperature of the vertical flue of a coke oven. The method includes: S10, data acquisition and preprocessing; S20, online soft measurement of coke cake maturity; S30, operating condition identification and state classification; S40, dynamic setting of target temperature; S50, multi-mode control and switching; S60, feedforward and feedback composite control; S70, control output and execution; and S80, closed-loop feedback and continuous optimization. By driving differentiated target temperature settings through online soft measurement of coke cake maturity, combined with multi-mode control and seamless switching, the accuracy and stability of the vertical flue temperature control are improved. Each carbonization chamber independently sets its target temperature according to its maturity progress, eliminating uneven heating and improving coke uniformity. Simultaneously, it reduces gas consumption per ton of coke and decreases thermal shock stress on the furnace body, extending the service life of the coke oven.

Description

Technical Field

[0001] The invention relates to the field of coking control technology, specifically to a method for controlling the temperature of the vertical flue of a coke oven. Background Technology

[0002] The temperature of the vertical flue of a coke oven is the most critical thermal parameter in the coking process, directly determining coke quality, heating gas consumption, and the service life of the coke oven. For a long time, coke oven temperature control has been a technical challenge in the field of metallurgical automation. This is fundamentally because the coke oven heating process exhibits high inertia, strong nonlinearity, multivariate coupling, and time-varying characteristics. Furthermore, the production process is driven by continuous variables such as temperature and flow rate, while also being affected by discrete events such as coal charging, coke pushing, and reversing. Conventional control methods are therefore insufficient to achieve ideal results.

[0003] Currently, the most widely used technologies in the field of coke oven temperature control have the following shortcomings:

[0004] First, existing technologies mostly use direct measurements of the flue temperature as the sole control target, without taking into account the actual maturity state of the coke cake. This fails to reflect the true coking process of the coke in each carbonization chamber, resulting in some coke being overburned while others are not yet mature.

[0005] Second, traditional methods often employ PID single-loop or cascade control. When operating conditions such as fluctuations in coal loading, changes in the calorific value of coal gas, and adjustments to the coking cycle change, a single control structure is difficult to adaptively adjust.

[0006] Third, existing coke oven temperature control schemes usually use the average temperature of the entire oven or the entire section as the basis for regulation, and cannot independently set the heating intensity according to the coking progress of each carbonization chamber. Summary of the Invention To address the shortcomings of existing technologies, this invention provides a method for controlling the temperature of the vertical flue of a coke oven, thereby solving the problems in existing coke oven temperature control technologies, such as the deviation of the target from mature requirements, the lack of adaptive strategies, and the inability of each chamber to be independently controlled according to the coking progress.

[0007] To achieve the above objectives, the invention provides the following technical solution:

[0008] A method for controlling the temperature of a coke oven vertical flue includes:

[0009] S10. Collect and preprocess process data, including the temperature of each flue, gas flow rate, air flow rate, flue suction, raw gas temperature and gas calorific value.

[0010] S20. Calculate the current maturity index of coke cake in each carbonization chamber using the soft measurement model of coke cake maturity;

[0011] S30. Based on the current maturity index, the fluctuation range of gas calorific value and the progress of coal charging and coking operations, the current production status is divided into corresponding working condition categories.

[0012] S40. Based on the operating condition category and the maturity index of each carbonization chamber, set differentiated target temperature values ​​for the vertical combustion channel for each combustion chamber.

[0013] S50. Select the current control mode from a variety of preset control modes according to the operating condition category, and perform a seamless switch when the operating condition changes;

[0014] S60. Feedback control is performed based on the deviation between the target temperature and the measured temperature. At the same time, feedforward compensation is performed based on the changing trend of the maturity index and the fluctuation of the gas calorific value as feedforward signals. The gas flow rate adjustment, air flow rate adjustment and flue gas suction adjustment are calculated.

[0015] S70. Output each adjustment value to the corresponding actuator to perform temperature regulation;

[0016] S80 collects the temperature of the vertical flue in real time and returns it to S60 to form a closed-loop control.

[0017] Optionally, in step S20, the soft measurement model for coke cake maturity is:

[0018] ,

[0019] in, Let be the coke cake maturity index of the i-th carbonization chamber at time t. Let be the vertical flue temperature of the combustion chamber corresponding to the i-th carbonization chamber. As the reference temperature, For the target temperature, The rate of temperature change of raw coal gas. Let t be the remaining coking time for the i-th carbonization chamber, and t be the time that coking has been in progress. To summarize the focal time, α, β, and γ are weighting coefficients, and α + β + γ = 1.

[0020] Optionally, in step S30, the operating conditions include the initial heating condition during coal charging, the steady-state condition during the middle stage of coking, the mature condition during the early stage of coking, the fluctuating calorific value of coal gas, and the abnormal disturbance condition.

[0021] Optionally, the criterion for determining the initial heating condition during coal loading is: coke cake maturity index. Furthermore, the time after coal loading is less than or equal to the first 20% of the coking cycle;

[0022] The criteria for determining the steady-state condition during the coking process are: coke cake maturity index. ;

[0023] The criteria for determining the maturity condition in the early stage of coke pushing are: coke cake maturity index. And the remaining coking time is less than or equal to the last 20% of the coking cycle.

[0024] Optionally, in step S40, the dynamic setting of the target temperature specifically involves:

[0025] Under the initial heating condition during coal loading, the target temperature of the combustion chamber corresponding to the carbonization chamber is set to the standard target temperature plus a first offset value, where the first offset value varies with the coke cake maturity index. As it increases, it decreases linearly to zero;

[0026] Under steady-state conditions during the mid-coking process, the target temperature is set to the standard target temperature.

[0027] Under the mature coke-pushing condition in the early stage, the target temperature of the combustion chamber corresponding to the carbonization chamber is set as the standard target temperature minus a second offset value, where the second offset value varies with the coke cake maturity index. It increases linearly with the increase of ;

[0028] Among them, different carbonization chambers independently calculate the target temperature based on their respective coke cake maturity index, thereby achieving differentiated settings for each combustion chamber.

[0029] Optionally, in step S50, the preset multiple control modes include generalized predictive control mode, fuzzy control mode, feedforward expert control mode, and PID cascade control mode.

[0030] Optionally, in step S60, the feedforward compensation includes:

[0031] Commutation feedforward compensation, when a commutation operation signal is detected, before the commutation occurs. The gas flow rate adjustment within seconds is superimposed with the first feedforward compensation value that dynamically changes with the reversal process;

[0032] Calorific value feedforward compensation: When the calorific value fluctuation of the gas exceeds a preset threshold, a second feedforward compensation value is calculated based on the calorific value change ΔH. This is then added to the gas flow rate regulation, where K is the compensation coefficient;

[0033] Maturity feedforward compensation: Maturity feedforward compensation: When the rate of change exceeds a preset threshold, a third feedforward compensation value is calculated based on the rate of change and superimposed on the gas flow rate adjustment amount of the corresponding combustion chamber.

[0034] Optionally, step S80 further includes: recording temperature data, maturity index and control deviation during the control process into a database, and using the recursive least squares method to update the weight coefficients α, β and γ of the soft sensor model online with the goal of minimizing the sum of squared deviations between the actual temperature value and the target value.

[0035] Optionally, the sampling period of the data acquisition is less than or equal to 10 seconds, and the control period of the closed-loop control is less than or equal to 30 seconds.

[0036] Optionally, the temperature control accuracy of the vertical flue reaches ±5℃, and the temperature stability coefficient of the entire furnace reaches 0.90 or higher.

[0037] Compared with existing technologies, the beneficial effects of the invention are:

[0038] By organically combining differentiated target temperature settings driven by coke cake maturity, multi-mode intelligent switching, and feedforward feedback composite control, the temperature control accuracy of the vertical flue can reach within ±5℃, and the overall furnace straight-line temperature stability coefficient is consistently above 0.90, an improvement of more than 50% compared to traditional PID cascade control. Furthermore, each carbonization chamber independently sets its target temperature according to its own maturity level, solving the problem of uneven heating where mature chambers are over-fired and immature chambers are under-fired, further improving the uniformity of coke quality. Simultaneously, proactive cooling in the early stage of coke pushing prevents over-burning of coke, and on-demand heating in each chamber reduces ineffective heating, reducing gas consumption per ton of coke by more than 3%. In addition, stable temperature control can reduce thermal shock stress on the furnace body, which helps to slow down the deterioration process of the silica bricks in the furnace walls and extend the service life of the coke oven. Attached Figure Description

[0039] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0040] Figure 1 This is a schematic diagram of the process steps in the embodiments of the invention. Detailed Implementation

[0041] In the following description, only certain exemplary embodiments are briefly described. As those skilled in the art will recognize, the described embodiments can be modified in various ways without departing from the spirit or scope of the invention. Therefore, the drawings and description are considered to be exemplary in nature and not restrictive.

[0042] The following disclosure provides numerous different embodiments or examples for implementing various structures of the invention. To simplify the disclosure, specific examples of components and arrangements are described below. These are, of course, merely illustrative and not intended to limit the invention. Furthermore, reference numerals and / or letters may be repeated in different examples; such repetition is for simplification and clarity and does not in itself indicate a relationship between the various embodiments and / or arrangements discussed. In addition, examples of various specific processes and materials are provided in the invention; however, those skilled in the art will recognize the application of other processes and / or the use of other materials.

[0043] The embodiments of the invention will now be described in detail with reference to the accompanying drawings.

[0044] Example:

[0045] See Figure 1 This invention provides a method for controlling the temperature of the vertical combustion chamber in a coke oven. This method uses the actual maturity of the coke cake as the control target, dynamically sets differentiated target temperatures for each combustion chamber by online soft-sensor measurement of the coke cake maturity index, and achieves adaptive temperature regulation under all operating conditions based on a multi-mode control strategy. The overall control objective of the method is: the vertical combustion chamber temperature control accuracy reaches ±5℃, and the overall straight-line temperature stability coefficient of the entire oven reaches 0.90 or higher. The method specifically includes the following steps executed in sequence:

[0046] S10. Data Acquisition and Preprocessing.

[0047] Step S10 is responsible for providing accurate and reliable input information for subsequent soft measurement of coke cake maturity, operating condition identification, target temperature setting, and feedback control. Specifically, process data is collected in real time by sensor groups deployed at various measuring points in the coke oven. The process data includes the temperature of each flue, gas flow rate, air flow rate, flue suction, raw gas temperature, and gas calorific value.

[0048] The temperature of each flue is obtained through an array of thermocouples installed at the observation holes on the furnace top. This array is spaced out along the coke oven side from the machine side to the coke side, covering all combustion chambers of the entire furnace. The gas flow rate is obtained through an orifice plate flow meter installed on the gas branch pipe in conjunction with a differential pressure transmitter. The air flow rate is obtained through a hot-wire flow meter installed in the air inlet box. The suction force of the branch flue is collected by a pressure transmitter installed at the flue gas outlet of the branch flue. The raw gas temperature is measured by a sheathed thermocouple inserted in the raw gas outlet pipe. The calorific value of the gas is detected in real time by an online calorific value analyzer installed on the main gas pipeline.

[0049] The raw signals collected by the aforementioned sensors are amplified and filtered by the signal conditioning circuit, then converted into digital signals by the analog-to-digital converter module, and subsequently sent to the industrial control computer. The industrial control computer performs data preprocessing on the received digital signals.

[0050] The preprocessing includes: using a moving average filter to eliminate random noise in the signal; using an amplitude limiting filter to remove abnormal pulse values ​​caused by occasional sensor malfunctions; and using linear interpolation to fill in data gaps caused by momentary communication interruptions. This preprocessing ensures that the data entering subsequent calculations has sufficient accuracy and continuity.

[0051] In step S10, considering the large inertial characteristics of coke oven temperature changes and the timeliness of the control system's response to disturbances, in order to ensure the real-time performance and response speed of the control system, the data acquisition sampling period is set to less than or equal to 10 seconds, ensuring that the data fully reflects the process dynamics while avoiding data redundancy.

[0052] Step S10 provides reliable data for the entire control method through comprehensive and accurate data acquisition and preprocessing.

[0053] S20, online soft measurement of coke cake maturity.

[0054] Step S20 can solve the technical problem that process indicators (coke cake maturity) are difficult to measure directly online. Specifically, based on auxiliary variables such as the temperature distribution of each flue, the rate of change of raw coal gas temperature, and coking time, the current maturity index of coke cake in each carbonization chamber is calculated through the constructed soft measurement model of coke cake maturity.

[0055] Coke cake maturity refers to the degree of maturity reached by the coke cake during the coking process within the carbonization chamber. Its value is normalized to between 0 and 1, where 0 indicates that the coke cake has just been loaded with coal and has not yet begun coking, and 1 indicates that the coke cake is fully mature and ready for coking. Although the core temperature of the coke cake is the most direct indicator of maturity, thermocouples cannot survive and operate continuously for extended periods in the harsh environment of the carbonization chamber—high temperature, strong reducing atmosphere, and coking operation—making this indicator impossible to obtain directly from hardware sensors. Therefore, soft measurement techniques must be used for indirect estimation.

[0056] The soft measurement model for the maturity of caramel cake is as follows:

[0057] The soft measurement model for coke cake maturity is as follows:

[0058] ,

[0059] in, Let be the coke cake maturity index of the i-th carbonization chamber at time t. Let be the vertical flue temperature of the combustion chamber corresponding to the i-th carbonization chamber. As the reference temperature, For the target temperature, The rate of temperature change of raw coal gas. Let t be the remaining coking time for the i-th carbonization chamber, and t be the time that coking has been in progress. To summarize the focal time, α, β, and γ are weighting coefficients, and satisfy α+β+γ=1.

[0060] The soft measurement model for coke cake maturity consists of a weighted average of three terms: temperature, rate of temperature rise, and time. The physical origins and derivation basis of these three terms are as follows:

[0061] First, the temperature term ( Based on the basic principles of heat transfer, during the coke oven heating process, heat is transferred from the combustion flame in the combustion chamber to the carbonization chamber through the furnace wall via heat conduction and radiation. The driving force for heat transfer is the difference between the temperature of the fire channel and the temperature of the coke cake.

[0062] According to Fourier's law of heat transfer and a simplified one-dimensional steady-state heat conduction model, the heat flux density q can be expressed as: Where q is the heat flux density transferred to the carbonization chamber through the furnace wall, λ is the equivalent thermal conductivity of the furnace wall refractory material, and δ is the furnace wall thickness. Under normal coke oven production conditions, λ and δ can be considered constants over a short period of time. Therefore, the heat transferred to the carbonization chamber is proportional to the difference between the flue temperature and the coke cake temperature.

[0063] Caramel temperature As the coking process gradually increases, its temperature correlates with the flue temperature. There is an approximately linear relationship between them, therefore, the position of the flue temperature relative to the target temperature range is used to characterize the contribution of temperature to maturity. When the temperature term contributes zero, it indicates that the flue temperature is at the lower limit of effective heating, and the heat input is insufficient to promote coking; when At this point, the temperature term contributes 1, indicating that the flue temperature has reached the ideal setpoint under the current operating conditions, and the heat input meets the normal coking requirements. The temperature term reflects the impact of heating intensity on the coke cake maturation process.

[0064] The rate of temperature rise ( Based on the fire-fall management theory, fire-fall management is a classic method used in the coking industry to judge the maturity of coke for a long time. Its core is that the coke cake continuously releases raw coal gas during the coking process. The temperature and composition of the raw coal gas change regularly with the coking process. When the coke cake reaches maturity, the temperature of the raw coal gas shows a characteristic "fire-fall" inflection point.

[0065] Specifically, in the early stage of coking, a large amount of volatile matter rapidly escapes from the coal and burns and releases heat in the combustion chamber, causing the temperature of the raw coal gas to rise sharply. The values ​​are relatively large. Entering the mid-stage of coking, the volatile matter emission rate tends to stabilize, and the temperature of the raw coal gas continues to rise, but the rate slows down. Furthermore, the value decreases. Near the point of impact, the volatile matter has essentially been completely released, and the temperature of the raw coal gas reaches its peak. After the initial spark, the unburnt gas, lacking any combustible components, continues to burn and release heat, causing the temperature to drop. .

[0066] This shows the rate of temperature change of the raw coal gas. This is a strong indicator of maturity. Compare this rate with the remaining coking time. The physical meaning of multiplication is that the same rate of temperature rise contributes differently to the judgment of maturity in the early and late stages of coking. The same rate change in the late stage means that the maturity process is closer, and therefore the sensitivity is higher.

[0067] Time item ( Based on the coking index model, the coking index... It is a widely used quantitative indicator of maturity in the field of coking engineering, where τC represents the carbonization time already completed. This is the standard time elapsed from coal loading to the occurrence of fire. This coking essentially characterizes the time completion of the coking process. This application will... Replace with summary focus time, which is more easily obtained in industrial settings. This allows the time term to be calculated directly from production plan parameters, avoiding the problem of inaccurate measurement of the fire-fall time. The time term reflects the natural progression of the coking reaction over time; even under completely constant temperature conditions, the maturity of the coke cake will monotonically increase with time.

[0068] The temperature, temperature rise rate, and time terms described above characterize the maturity of the coke cake from three different dimensions: thermodynamic driving force, fire-fall characteristic signal, and reaction kinetic process, respectively. A comprehensive judgment of multi-source information is achieved through linear weighted combination.

[0069] The weighting coefficients α, β, and γ in the soft measurement model for coke cake maturity are determined using multi-linear regression, which includes the following steps:

[0070] S21. Collect n sets of historical production data. Each set of data includes:

[0071] Independent variable: , , .

[0072] Dependent variable: The actual maturity of the coke cake in this oven (obtained through manual experience assessment or sampling testing after coke pushing).

[0073] S22. Establish the regression equation:

[0074] ,

[0075] S23. The objective function is to minimize the sum of squared residuals.

[0076] ,

[0077] S24. Given the constraint α+β+γ=1, solve for the optimal weight coefficients using the Lagrange multiplier method.

[0078] This step S20 transforms the coke cake maturity, which was originally not directly measurable, into a real-time calculable index by running the aforementioned soft measurement model online in the control system. This provides a quantitative basis for subsequent operating condition identification and target temperature differentiation setting, and realizes the transformation of the control target from controlling the flue temperature to controlling the coke cake maturity.

[0079] S30, Operating condition identification and state classification.

[0080] Step S30 uses the coke cake maturity index of each carbonization chamber calculated in step S20, combined with multi-source information such as the fluctuation range of gas calorific value and the progress of coal charging and coking operations, to classify the current production status into corresponding operating condition categories. The purpose of operating condition classification is that the coke oven production process exhibits drastically different dynamic characteristics at different stages, and a single control strategy cannot achieve ideal control effects under all operating conditions. Therefore, operating condition identification provides a basis for the selection and switching of subsequent control modes.

[0081] In step S30, the operating conditions include the initial heating condition during coal charging, the steady-state condition during the middle stage of coking, the mature condition during the early stage of coking, the gas calorific value fluctuation condition, and the abnormal disturbance condition. The criteria and judgment conditions for classifying each type of operating condition are as follows:

[0082] The criteria for determining the initial temperature rise condition during coal loading are: coke cake maturity index. Furthermore, the time after coal loading is less than or equal to the first 20% of the coking cycle. A typical characteristic of this condition is that the temperature of the newly loaded coal is much lower than the furnace wall temperature of the carbonization chamber, with a large amount of heat used to heat the cold coal. This poses a significant risk of a sharp drop in furnace temperature, constituting a period of strong disturbance. During this stage, the primary control objective is to quickly restore the furnace temperature and prevent excessive temperature drops.

[0083] The criteria for determining the steady-state condition during the mid-stage of coking are: coke cake maturity index. The typical characteristics of this operating condition are that the coke cake in the carbonization chamber is in a stable coking stage, the raw coal gas evolution rate is relatively stable, and the heat load changes slowly. This is the longest stage in the entire coking cycle. During this stage, the primary control objective is to maintain temperature stability and achieve high-precision temperature tracking.

[0084] The criteria for determining the maturity condition in the early stage of coke pushing are: coke cake maturity index. Furthermore, the remaining coking time is less than or equal to the last 20% of the coking cycle. A typical characteristic of this condition is that the coke cake is nearing or has reached maturity; continuing to heat at full load would result in over-burning of the coke and energy waste. At this stage, the control objective is to moderately reduce the heating intensity to achieve energy-saving control by cooling down as the coke matures.

[0085] The criterion for determining the fluctuating calorific value of coal gas is: the change in the detected calorific value of coal gas within a unit of time exceeds a preset threshold (usually ±3% of the set value). This condition can coexist with the other three time-scale conditions and is considered an overlay condition. Its typical characteristic is that the calorific value of the heating coal gas fluctuates due to changes in the operating conditions of the upstream gas generator or adjustments in the gas ratio, causing changes in the actual heat supply while the gas flow rate remains constant. In this case, calorific value feedforward compensation needs to be introduced into the control system.

[0086] The criteria for determining abnormal disturbance conditions are: any of the above process parameters exceeds the corresponding normal operating range, or the rate of change exceeds the allowable limit. This condition is a fault-prone condition and requires a conservative stabilization control strategy.

[0087] Through the above-mentioned multi-dimensional and multi-threshold working condition classification, step S30 decomposes the complex and ever-changing coke oven production process into several typical working conditions with different control requirements, providing accurate working condition labels for the control mode selection in the subsequent step S50, so that the subsequent control decisions can have clear objectives.

[0088] S40, Target Temperature Dynamic Setting.

[0089] Step S40 sets differentiated target temperatures for each combustion chamber based on the operating condition category identified in step S30 and the coke cake maturity index calculated in step S20, allowing the target temperature to be dynamically adjusted according to changes in coke cake maturity. Unlike traditional control methods that uniformly set the target temperature for the entire furnace, step S40 enables different coking chambers to have their own independent target temperatures based on their respective coking stages and maturity states.

[0090] Specifically, the target temperature setting is handled in the following three ways:

[0091] During the initial heating phase of coal charging, the target temperature of the combustion chamber corresponding to the carbonization chamber is set to the standard target temperature plus a first offset value. The standard target temperature is a baseline value pre-calculated based on the coking time and coal blending ratio of the carbonization chamber, representing the optimal heating temperature under steady-state conditions during the mid-coking phase. The first offset value is applied immediately after coal charging to address the sudden drop in furnace temperature caused by the large amount of heat absorbed by the newly charged coal. Its initial value is typically +20℃ to +40℃, varying with the coke cake maturity index. As it increases, it decreases linearly to zero. When When the value reaches 0.2, the first offset value returns to zero, and the target temperature returns to the standard target temperature. The design logic for this linear reduction is as follows: as the coal temperature gradually increases, its heat absorption rate gradually decreases, and the need for additional heating also decreases accordingly. Therefore, the offset should be inversely related to the maturity progress.

[0092] Under steady-state conditions during the mid-coking process, the target temperature is set to the standard target temperature. At this point, the coke cake is in a stable coking stage, and the heat transfer process has established a dynamic equilibrium. No additional temperature offset is required, and the main task of control is to maintain the stability of the target temperature.

[0093] Under the mature coke-pushing condition in the early stage, the target temperature of the combustion chamber corresponding to the carbonization chamber is set to the standard target temperature minus a second offset value. The second offset value is applied when the coke cake is nearing maturity, used to proactively reduce heating intensity before coke pushing, preventing over-burning of the coke and saving gas consumption. The second offset value varies with the coke cake maturity index. It increases linearly with the increase of , when As the value increases from 0.8 to 1.0, the second offset value increases linearly from 0 to its maximum value (typically -10°C to -30°C). The design logic for this linear increase is that the closer the cake is to full maturity, the less necessary it is to continue heating, and the cooling rate should be increased accordingly.

[0094] The target temperature settings mentioned above are calculated independently for each carbonization chamber, with each chamber based on its own coke cake maturity index. The specific target temperature is determined by the operating condition category and corresponding offset rules. Due to differences in coal charging time and coke pushing plans, the coking process of different carbonization chambers in the same coke oven is not synchronized. By setting independent settings, the heating intensity of each carbonization chamber can be precisely matched with its actual needs, which not only ensures the uniformity of coke quality in each oven, but also achieves the energy-saving goal of on-demand heating.

[0095] In step S40, the coke cake maturity index calculated in step S20 is converted into an executable control target value through dynamic differential target temperature setting.

[0096] S50, multi-mode control and switching.

[0097] In step S50, based on the current operating condition category identified in step S30, one of the preset control modes is selected and temperature control is executed. At the same time, when the operating condition changes, the control mode is switched seamlessly.

[0098] The dynamic characteristics of the coke oven heating process change significantly with the operating conditions. It is impossible to maintain optimal performance under all operating conditions through a single control algorithm. Therefore, it is necessary to build a multi-mode control strategy library and select it in real time according to the operating conditions.

[0099] The various control modes preset in step S50 include generalized predictive control mode, fuzzy control mode, feedforward expert control mode, and PID cascade control mode.

[0100] In step S50, the applicable scope and control characteristics of each mode are as follows:

[0101] Generalized predictive control (GDC) is based on the classic framework of predictive models, rolling optimization, and feedback correction. It calculates control variables through multi-step prediction and quadratic performance index optimization, offering advantages such as relatively relaxed requirements on model accuracy and strong adaptability to time-delay systems. It is suitable for high-precision temperature tracking control under steady-state conditions during the mid-coking process. Under these conditions, the process dynamics are relatively stable, allowing GDC to fully leverage its predictive advantages and achieve temperature control accuracy within ±5℃.

[0102] Fuzzy control, based on fuzzy logic reasoning, quantifies the operator's control experience into a fuzzy rule base. It calculates the control quantity through three steps: fuzzification, rule reasoning, and defuzzification. It does not require a precise mathematical model of the controlled object and exhibits a certain degree of robustness to parameter changes and external disturbances. It is suitable for temperature regulation during the transition period in the initial heating stage of coal loading and the early mature stage of coking. In these two transitional conditions, the system dynamics change drastically, making it difficult to establish a precise model. Fuzzy control can achieve effective control by utilizing expert experience.

[0103] The feedforward expert control mode summarizes the operational experience and rules of process experts into a generative rule set, which can take control actions in advance based on predictable disturbance signals. It is suitable for advance compensation control before reversing operations and under conditions of fluctuating gas calorific value.

[0104] The PID cascade control mode is based on the traditional PID algorithm. It uses a secondary loop to quickly eliminate fluctuations in intermediate variables such as gas flow rate, while the main loop slowly adjusts the temperature of the vertical flue. It has a simple structure, high reliability, and is suitable for stable regulation under abnormal interference conditions. It ensures that the system does not become unstable under special circumstances.

[0105] In step S50, the mode selection method is as follows: when the operating condition is a steady-state condition in the middle stage of coking, the generalized predictive control mode is selected; when the operating condition is a heating condition in the early stage of coal charging or a mature condition in the early stage of coking, the fuzzy control mode is selected; when a reversing operation signal or a gas calorific value fluctuation condition is detected, the feedforward expert control mode is superimposed on the original control mode for feedforward compensation without changing the main control mode; when the operating condition is an abnormal disturbance condition, the mode is forcibly switched to the PID cascade control mode.

[0106] S60, feedforward and feedback composite control.

[0107] Based on the current control mode selected in step S50, feedback control is performed using the deviation between the target temperature set in step S40 and the measured temperature. At the same time, feedforward compensation is performed using the changing trend of the coke cake maturity index and the fluctuation of the gas calorific value as feedforward signals. The gas flow rate regulation, air flow rate regulation, and flue gas suction regulation are calculated in a comprehensive manner.

[0108] The feedback control section is executed by the currently selected control mode. Taking the generalized predictive control mode as an example, its control law uses the target temperature set in step S40 as the setpoint r(t), and the measured temperature of the flue gas collected in step S10 as the controlled variable y(t). The deviation e(t) = r(t) − y(t) is calculated, and the output for several future steps is predicted by the predictive model to minimize the following performance index:

[0109] ,

[0110] in, and These represent the start and end times of the prediction time domain, respectively. To control the time domain, The weighting coefficients are used to control the system, and Δu represents the control increment. The optimal control increment sequence is obtained by solving this optimization problem online, and the first control increment is applied. The fuzzy control mode, on the other hand, calculates the control quantity directly from the deviation e and the rate of change of deviation ec through fuzzy rule inference.

[0111] In addition, the feedforward compensation section sets up compensation channels for the three main types of predictable interference sources:

[0112] Feedforward compensation during reversal. The coke oven performs a reversal operation every 20 to 30 minutes. During this reversal, the flow direction of the heating gas and the gas flow changes, causing a regular temporary drop in combustion chamber temperature. This step, upon detecting a reversal operation signal, calculates the feedforward compensation based on historical reversal temperature patterns in the vertical flue before the reversal operation occurs. The gas flow rate adjustment is superimposed on the first feedforward compensation value within seconds, and this compensation value changes dynamically with the reversal process. Specifically, the gas flow rate is increased starting 5 to 10 seconds before the reversal, the compensation value is maintained during the reversal, and the compensation is gradually removed after the reversal is completed, thereby offsetting the temperature drop caused by combustion interruption during the reversal.

[0113] Calorific value feedforward compensation. The calorific value of the heating gas changes frequently due to fluctuations in the gas source, which is a significant disturbance affecting the stability of the furnace temperature. When the detected fluctuation in the calorific value of the gas exceeds a preset threshold (usually ±3%), a second feedforward compensation value ΔQ = K•ΔH is calculated based on the change in calorific value ΔH and added to the gas flow rate adjustment. Here, K is the compensation coefficient, which physically represents the gas flow rate required to compensate for a unit change in calorific value. This coefficient is obtained through dynamic decoupling experiments of calorific value, flow rate, and temperature. This compensation causes the gas flow rate to change inversely with the calorific value: the gas flow rate decreases when the calorific value increases and increases when the calorific value decreases, ensuring that the total heat input to the combustion chamber remains constant.

[0114] Maturity feedforward compensation. When When the rate of change exceeds a preset threshold, it indicates an abnormal trend of accelerated or decelerated coke cake maturation, requiring pre-adjustment of the heating intensity. A third feedforward compensation value is calculated based on the rate of change in maturity and superimposed on the corresponding combustion chamber's gas flow regulation, enabling the control system to adjust the heating intensity before the actual temperature change, thus achieving proactive control.

[0115] After the above three feedforward compensation values ​​are superimposed with the output value of the feedback controller, and after dynamic characteristic compensation and amplitude limiting processing of the actuator, the setpoints of each control quantity are finally generated and sent to the process control level. This step S50, through the organic combination of feedforward and feedback, takes into account both the robustness of feedback control and the timeliness of feedforward control, significantly improving the overall suppression capability of the control system against dynamic forces.

[0116] S70, Control Output and Execution.

[0117] Step S70 outputs the adjustment values ​​calculated in step S60 to the corresponding field actuators, completing the physical-level temperature regulation operation. Specifically:

[0118] The gas flow regulation is output as a standard current signal of 4mA~20mA to the electric regulating valve installed on the gas branch pipe of each combustion chamber. The gas flow into each combustion chamber is regulated by changing the valve opening. The electric regulating valve is equipped with a valve position feedback transmitter, which transmits the valve position opening signal back to the control system in real time to form a valve position closed loop, ensuring the accurate execution of valve position commands.

[0119] The air flow rate regulation output is sent to the air damper actuator installed at the air inlet box, which changes the amount of combustion air supplied by adjusting the air inlet area. The air flow rate regulation follows the principle of coordination with the gas flow rate, and the two change in tandem according to the set air-fuel ratio curve to ensure that the gas is fully combusted in the combustion chamber and avoid incomplete combustion or excess air caused by air-fuel ratio imbalance.

[0120] The suction adjustment of the sub-flue is output to the sub-flue flap actuator. By adjusting the flap opening of the sub-flue flue gas outlet, the suction regime in the furnace is changed, maintaining the pressure at the bottom of the carbonization chamber within a slightly positive pressure range. This ensures smooth flue gas discharge from the coke oven heating system and prevents excessively low furnace head flue temperature and air leakage caused by excessive suction.

[0121] S80, closed-loop feedback and continuous optimization.

[0122] Step S80 completes the closed-loop and self-optimization functions of the entire control method. On one hand, it collects the measured temperature of each flue in real time and returns it to step S60, using it as the input signal for feedback control to participate in the deviation calculation of the next control cycle, forming a complete closed-loop control structure, thereby ensuring the stability of the control. On the other hand, it records the temperature data, maturity index, and control deviation during the control process into the real-time database and the historical database for continuous optimization of the parameters of the soft sensor model.

[0123] In step S80, the control cycle of the closed-loop control is set to be less than or equal to 30 seconds. The time constant of coke oven temperature changes is usually on the order of several minutes to tens of minutes. A control cycle of 30 seconds can capture temperature changes and respond in a timely manner, while also avoiding frequent actuator movements caused by an excessively short control cycle.

[0124] The online update of the weight coefficients of the soft measurement model is performed as follows: taking the minimum sum of squared deviations between the actual value and the target value of the flue temperature within the control cycle as the objective function, the weight coefficients α, β, and γ are updated online using the recursive least squares method.

[0125] In this embodiment, by organically combining differentiated target temperature settings driven by coke cake maturity, multi-mode intelligent switching, and feedforward feedback composite control, the temperature control accuracy of the vertical flue can reach within ±5℃, and the overall furnace straight-line temperature stability coefficient is stably above 0.90, which is more than 50% higher than traditional PID cascade control. Furthermore, each carbonization chamber independently sets its target temperature according to its own maturity, solving the problem of uneven heating where mature chambers are over-fired and immature chambers are under-fired, further improving the uniformity of coke quality. Simultaneously, proactive cooling in the early stage of coke pushing prevents over-burning of coke, and on-demand heating in each chamber reduces ineffective heating, reducing gas consumption per ton of coke by more than 3%. In addition, stable temperature control can reduce thermal shock stress on the furnace body, which helps to slow down the deterioration process of the silica bricks in the furnace walls and extend the service life of the coke oven.

[0126] The above embodiments are merely preferred embodiments of the invention and are not intended to limit the technical solutions of the invention. Any technical solution that can be implemented based on the above embodiments without creative effort should be considered to fall within the scope of protection of the invention patent.

Claims

1. A method for controlling the temperature of a coke oven vertical flue, characterized in that, include: S10. Collect and preprocess process data, including the temperature of each flue, gas flow rate, air flow rate, flue suction, raw gas temperature and gas calorific value. S20. Calculate the current maturity index of coke cake in each carbonization chamber using the soft measurement model of coke cake maturity; S30. Based on the current maturity index, the fluctuation range of gas calorific value and the progress of coal charging and coking operations, the current production status is divided into corresponding working condition categories. S40. Based on the operating condition category and the maturity index of each carbonization chamber, set differentiated target temperature values ​​for the vertical combustion channel for each combustion chamber. S50. Select the current control mode from a variety of preset control modes according to the operating condition category, and perform a seamless switch when the operating condition changes; S60. Feedback control is performed based on the deviation between the target temperature and the measured temperature. At the same time, feedforward compensation is performed based on the changing trend of the maturity index and the fluctuation of the gas calorific value as feedforward signals. The gas flow rate adjustment, air flow rate adjustment and flue gas suction adjustment are calculated. S70. Output each adjustment value to the corresponding actuator to perform temperature regulation; S80 collects the temperature of the vertical flue in real time and returns it to S60 to form a closed-loop control.

2. The method for controlling the temperature of the coke oven vertical flue according to claim 1, characterized in that, In step S20, the soft measurement model for the maturity of the coke cake is as follows: ， in, Let be the coke cake maturity index of the i-th carbonization chamber at time t. Let be the vertical flue temperature of the combustion chamber corresponding to the i-th carbonization chamber. As the reference temperature, For the target temperature, The rate of temperature change of the raw coal gas. Let t be the remaining coking time for the i-th carbonization chamber, and t be the time that coking has been in progress. To summarize the focal time, α, β, and γ are weighting coefficients, and α + β + γ = 1.

3. The method for controlling the temperature of the coke oven vertical flue according to claim 2, characterized in that, In step S30, the operating conditions include the initial heating condition during coal loading, the steady-state condition during the middle stage of coking, the mature condition during the early stage of coking, the fluctuating calorific value of coal gas, and the abnormal disturbance condition.

4. The method for controlling the temperature of the coke oven vertical flue according to claim 3, characterized in that, The criteria for determining the initial heating condition during coal charging are: coke cake maturity index. Furthermore, the time after coal loading is less than or equal to the first 20% of the coking cycle; The criteria for determining the steady-state condition during the coking process are: coke cake maturity index. ; The criteria for determining the maturity condition in the early stage of coke pushing are: coke cake maturity index. And the remaining coking time is less than or equal to the last 20% of the coking cycle.

5. The method for controlling the temperature of the coke oven vertical flue according to claim 2, characterized in that, In step S40, the target temperature is dynamically set as follows: Under the initial heating condition during coal loading, the target temperature of the combustion chamber corresponding to the carbonization chamber is set to the standard target temperature plus a first offset value, where the first offset value varies with the coke cake maturity index. As it increases, it decreases linearly to zero; Under steady-state conditions during the mid-coking process, the target temperature is set to the standard target temperature. Under the mature coke-pushing condition in the early stage, the target temperature of the combustion chamber corresponding to the carbonization chamber is set as the standard target temperature minus a second offset value, where the second offset value varies with the coke cake maturity index. It increases linearly with the increase of ; Among them, different carbonization chambers independently calculate the target temperature based on their respective coke cake maturity index, thereby achieving differentiated settings for each combustion chamber.

6. The method for controlling the temperature of the coke oven vertical flue according to claim 2, characterized in that, In step S50, the preset control modes include generalized predictive control mode, fuzzy control mode, feedforward expert control mode, and PID cascade control mode.

7. The method for controlling the temperature of the coke oven vertical flue according to claim 2, characterized in that, In step S60, feedforward compensation includes: Commutation feedforward compensation, when a commutation operation signal is detected, before the commutation occurs. The gas flow rate adjustment within seconds is superimposed with the first feedforward compensation value that dynamically changes with the reversal process; Calorific value feedforward compensation: When the calorific value fluctuation of the gas exceeds a preset threshold, a second feedforward compensation value is calculated based on the calorific value change ΔH. This is then added to the gas flow rate regulation, where K is the compensation coefficient; Maturity feedforward compensation: Maturity feedforward compensation: When the rate of change exceeds a preset threshold, a third feedforward compensation value is calculated based on the rate of change and superimposed on the gas flow rate adjustment amount of the corresponding combustion chamber.

8. The method for controlling the temperature of the coke oven vertical flue according to claim 2, characterized in that, Step S80 further includes: recording temperature data, maturity index and control deviation during the control process into the database, and using the recursive least squares method to update the weight coefficients α, β and γ of the soft sensor model online with the goal of minimizing the sum of squared deviations between the actual temperature value and the target value.

9. The method for controlling the temperature of the coke oven vertical flue according to claim 2, characterized in that, The sampling period for data acquisition is less than or equal to 10 seconds, and the control period for closed-loop control is less than or equal to 30 seconds.

10. The method for controlling the temperature of the coke oven vertical flue according to claim 2, characterized in that, The temperature control accuracy of the vertical flue reaches ±5℃, and the temperature stability coefficient of the entire furnace reaches over 0.90.

Images