Rotary kiln combustion intelligent scheduling and pressure balance control method and system considering coal gas calorific value fluctuation

Abstract

The application discloses a rotary kiln combustion intelligent scheduling and pressure balance control method and system considering coal gas calorific value fluctuation, through real-time acquisition of coal gas calorific value and filtering pretreatment, a heat value feedforward compensation amount is calculated to generate a compensated coal gas quantity reference value; a step response coefficient matrix model is constructed with a coal gas valve opening degree, a combustion air fan rotating speed, an induced draft fan rotating speed as a manipulated variable, and a kiln tail temperature, a kiln head pressure and a flue gas oxygen content as a controlled variable, a rolling optimization objective function is constructed based on the matrix; a dynamic pressure safety constraint boundary is calculated in real time according to the current coal gas flow and the heat value effective value and embedded as a hard constraint; in each control cycle, the feedback value of the controlled variable and the compensated coal gas quantity reference value are read, a quadratic programming problem with constraints is solved, and the coordinated control increments of the three manipulated variables are output to the actuator for execution. The application realizes heat value fluctuation advance compensation, combustion pressure coordinated stability and dynamic safety constraint guarantee.

Description

Technical Field

[0001] This invention relates to the field of industrial process control technology, and in particular to a method and system for intelligent scheduling and pressure balance control of rotary kiln combustion that takes into account fluctuations in the calorific value of coal gas, based on calorific value feedforward and multivariable predictive control. Background Technology

[0002] Rotary kilns are key thermal equipment in metallurgical lime production. Many steel companies utilize by-product gases (blast furnace gas and converter gas) as fuel to achieve comprehensive energy utilization. However, fluctuations in the calorific value of these by-product gases are a persistent problem hindering production. For example, the calorific value of blast furnace gas typically fluctuates between 700-900 kcal / Nm³, while that of converter gas fluctuates between 1600-2000 kcal / Nm³. When the two gases are mixed, the calorific value fluctuations are even more pronounced. These fluctuations not only affect the calcination temperature but also cause pressure fluctuations within the kiln due to changes in gas density and flow rate, posing significant challenges to stable control and safe production.

[0003] Similar existing technical solutions can be mainly divided into two categories: Temperature coordination control based on flow feedforward: This scheme, based on traditional temperature PID control, introduces changes in gas flow rate as a feedforward signal into the combustion air control, so that the blower follows the changes in gas flow rate.

[0004] Disadvantages: This scheme only considers changes in flow rate, but not changes in calorific value. When the calorific value fluctuates, even if the flow rate remains constant, the actual heat output changes, rendering the feedforward signal ineffective, and the system remains in a "blind adjustment" state.

[0005] Pressure feedback-based PID control: This scheme sets up a pressure transmitter at the kiln head or kiln tail, and adjusts the speed of the induced draft fan through a PID controller to maintain stable pressure inside the kiln.

[0006] Disadvantages: Pressure PID control is a pure feedback control, which only starts adjusting when the pressure has deviated from the set value, resulting in a lag for sudden fluctuations in calorific value / pressure. Furthermore, this scheme is not linked to combustion control, which can sometimes lead to interference between the induced draft fan and the blower, exacerbating pressure fluctuations.

[0007] In particular, rotary kilns in integrated iron and steel enterprises typically use blast furnace gas, converter gas, or mixtures thereof as fuel. The calorific value of these gases fluctuates frequently (with fluctuations exceeding ±15%) depending on the operating conditions of the upstream ironmaking and steelmaking processes. This results in the following deficiencies in existing control schemes: There is no effective compensation for calorific value fluctuations: Traditional combustion control only adjusts the gas flow rate. When the calorific value decreases, even if the flow rate remains unchanged, the actual heat input will decrease, leading to a drop in kiln temperature; conversely, when the calorific value increases, the temperature will overshoot. Operators can usually only passively adjust the control after the temperature changes, resulting in a serious lag.

[0008] Pressure fluctuations cause safety issues: Fluctuations in the calorific value of coal gas are often accompanied by pressure fluctuations. In traditional control systems, the coal gas regulating valve, combustion fan, and induced draft fan are controlled independently, lacking coordination. When the calorific value suddenly drops, operators may blindly open the coal gas valve to maintain the temperature, which may lead to a sudden drop in coal gas pressure (flameout) or a sudden rise in pressure (backfire). At the same time, the kiln head pressure fluctuates violently, which may result in smoke and dust emissions at best, and endanger combustion safety at worst.

[0009] Low combustion efficiency: Due to the unknown calorific value, the air-fuel ratio cannot be precisely matched. To ensure complete combustion, operators often conservatively use a larger air-fuel ratio, resulting in excess air carrying away a large amount of heat and increasing energy consumption. Summary of the Invention

[0010] The technical problem to be solved by this invention is to provide a method and system for intelligent scheduling and pressure balance control of rotary kiln combustion that takes into account the fluctuation of gas calorific value. This invention addresses the technical challenges of unstable combustion, drastic pressure fluctuations, and product quality fluctuations caused by frequent fluctuations in the calorific value of gas in rotary kilns for quicklime production in steel enterprises, which use low-calorific-value by-product gas (blast furnace gas, converter gas) as fuel.

[0011] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: A method for intelligent scheduling and pressure balance control of rotary kiln combustion considering fluctuations in the calorific value of coal gas includes the following steps: S1. Real-time acquisition of gas calorific value signal and filtering preprocessing to obtain effective calorific value; S2. Based on the target total heat input and the effective value of the calorific value, calculate the reference gas flow rate and the gas flow rate feedforward compensation amount when the calorific value changes, and generate a reference value of the gas flow rate after compensation. S3. Construct a multivariate predictive control model with the gas valve opening, combustion fan speed, and induced draft fan speed as manipulated variables, and the kiln tail temperature, kiln head pressure, and flue gas oxygen content as controlled variables. The model includes the large lag characteristic of the gas valve opening on the kiln tail temperature, the rapid coupling characteristic of the gas valve opening on the kiln head pressure, the coupling characteristic of the combustion air volume on the kiln tail temperature and the kiln head pressure, and the direct effect characteristic of the induced draft fan speed on the kiln head pressure. S4. Calculate the safety constraint boundary of the kiln head pressure in real time based on the current gas flow rate and effective calorific value, and add the safety constraint boundary as a hard constraint to the multivariate predictive control model. S5. In each control cycle, read the feedback value of the controlled variable and the reference value of the compensated gas flow rate, solve the constrained quadratic programming problem through rolling optimization, output the collaborative control increment of the manipulated variable, and send it to the gas regulating valve, the combustion fan frequency converter and the induced draft fan frequency converter for execution.

[0012] By employing the above method, the present invention can detect fluctuations in the calorific value of coal gas in advance, dynamically predict its impact on combustion and pressure, and coordinate the control of coal gas flow, combustion air volume and induced draft volume, thereby achieving stable combustion and pressure balance with anti-interference capabilities, and improving product quality and energy utilization efficiency.

[0013] In the above technical solution, the gas calorific value signal in step S1 is acquired in real time by an online gas calorific value meter, or predicted by a soft measurement model based on the upstream process conditions.

[0014] In the above technical solution, the filtering and preprocessing of the gas calorific value signal includes smoothing the calorific value signal and eliminating measurement noise.

[0015] In the above technical solution, the calorific value feedforward compensation amount in step S2 is calculated based on the change in the effective calorific value and the compensation intensity coefficient, and the reference value of the compensated gas volume is equal to the sum of the reference gas flow rate and the calorific value feedforward compensation amount.

[0016] In the above technical solution, the safety constraint boundary mentioned in step S4 includes the lower limit of the deflaming critical pressure and the upper limit of the tempering critical pressure calculated in real time based on the current gas flow rate and burner characteristics.

[0017] In the above technical solution, the objective of the rolling optimization solution in step S5 is to make the controlled variable track the set value, minimize the change of the manipulated variable, and prioritize satisfying the safety constraint boundary when the kiln head pressure is predicted to exceed the limit.

[0018] Specifically, when solving the rolling optimization objective function, if the kiln head pressure is predicted to exceed the limit, the dynamic pressure constraint boundary is automatically satisfied first, and the temperature adjustment rate is appropriately sacrificed.

[0019] In the above technical solution, the rolling optimization in step S5 includes: A step response coefficient matrix model is constructed as a multivariable predictive control model. Based on this matrix, a predictive output equation and a rolling optimization objective function containing deviation terms, smoothing terms and baseline deviation terms are constructed. In each control cycle, the feedback value of the controlled variable and the reference value of the gas quantity after compensation are read, and a constrained quadratic programming problem is solved to output the coordinated control increment of the three manipulated variables.

[0020] In the above technical solution, the actuator includes a gas regulating valve, a combustion fan frequency converter, and an induced draft fan frequency converter.

[0021] A rotary kiln combustion intelligent scheduling and pressure balance control system that considers fluctuations in the calorific value of coal gas includes: The calorific value sensing module is used to collect the calorific value signal of coal gas in real time and perform filtering preprocessing; The calorific value feedforward compensator is used to calculate the reference gas flow rate and the gas flow rate feedforward compensation amount based on the target total heat input and the effective calorific value, and generate a reference value of gas flow rate after compensation. A multivariate predictive controller is used to construct a predictive control model with gas valve opening, combustion fan speed, and induced draft fan speed as manipulated variables, and kiln tail temperature, kiln head pressure, and flue gas oxygen content as controlled variables. It calculates the safety constraint boundary of the kiln head pressure in real time based on the current gas flow rate and effective calorific value, and outputs the collaborative control increment through rolling optimization. The feedback signal acquisition module is used to read the feedback values ​​of the current kiln tail temperature T, flue gas oxygen content O2, and kiln head pressure P. The actuators, including the gas regulating valve, the combustion fan frequency converter, and the induced draft fan frequency converter, are used to receive and execute the cooperative control increments.

[0022] In the above technical solution, the calorific value sensing module includes an online gas calorific value meter or a soft measurement model based on the upstream process conditions.

[0023] In the above technical solution, the predictive control model of the multivariable predictive controller includes the large lag characteristic of the gas valve opening on the kiln tail temperature, the rapid coupling characteristic of the gas valve opening on the kiln head pressure, the coupling characteristic of the combustion air volume on the kiln tail temperature and the kiln head pressure, and the direct effect characteristic of the induced draft fan speed on the kiln head pressure.

[0024] In the above technical solution, the safety constraint boundary includes the lower limit of the deflaming critical pressure and the upper limit of the tempering critical pressure, which are calculated in real time based on the current gas flow rate and burner characteristics.

[0025] In the above technical solution, the multivariate predictive controller reads the feedback value of the controlled variable and the reference value of the compensated gas flow rate in each control cycle, and solves the constrained quadratic programming problem through rolling optimization to make the controlled variable track the set value and minimize the change of the manipulated variable. When the kiln head pressure is predicted to exceed the limit, the safety constraint boundary is satisfied first.

[0026] Based on the above-described method and system, the present invention also provides a computer-readable storage medium.

[0027] A computer-readable storage medium storing a program that, when executed by a processor, implements the steps of any of the methods described above.

[0028] In summary, this invention proposes a method and system for intelligent scheduling and pressure balance control of rotary kiln combustion, based on calorific value feedforward and multivariate predictive control, taking into account fluctuations in the calorific value of coal gas. By acquiring coal gas calorific value information in real-time or near real-time, a calorific value fluctuation feedforward compensation model is constructed. Model predictive control (MPC) is used to perform multivariate coordinated control of coal gas flow rate, combustion air volume, and induced draft volume, thereby eliminating the impact of calorific value fluctuations on temperature and pressure in advance, and achieving integrated control of combustion stability and pressure balance.

[0029] This invention acquires the calorific value of coal gas in real time and performs filtering preprocessing to calculate the calorific value feedforward compensation amount to generate a reference value for the compensated coal gas volume. It constructs a step response coefficient matrix model with the coal gas valve opening, combustion fan speed, and induced draft fan speed as manipulated variables, and the kiln tail temperature, kiln head pressure, and flue gas oxygen content as controlled variables. Based on this matrix, it constructs a predictive output equation and a rolling optimization objective function containing deviation terms, smoothing terms, and a baseline deviation term. It calculates the dynamic pressure safety constraint boundary in real time based on the current coal gas flow rate and effective calorific value and embeds it as a hard constraint. In each control cycle, it reads the feedback value of the controlled variables and the reference value of the compensated coal gas volume, solves a constrained quadratic programming problem, and outputs the collaborative control increment of the three manipulated variables to the actuator for execution.

[0030] This invention introduces a dynamic compensation mechanism based on calorific value feedforward: an online gas calorific value meter (or a soft measurement model based on operating conditions) is used to acquire calorific value changes in real time. A nonlinear mapping model between calorific value and heat is constructed. When a decrease in calorific value is detected, the required increase in gas flow rate is automatically calculated in advance to achieve a dynamic balance of "calorific value decrease, flow rate increase," maintaining a constant total input heat. This mechanism acts before temperature changes, overcoming the large lag.

[0031] This invention introduces a combustion-pressure multivariate predictive collaborative control model: a multivariate predictive control model is established with gas flow rate, combustion air volume, and induced draft volume as manipulated variables (MV), and calcination zone temperature (or kiln tail temperature), kiln head pressure, and flue gas oxygen content as controlled variables (CV). Utilizing the rolling optimization function of MPC, the collaborative adjustment of the three manipulated variables is automatically calculated to achieve multi-objective collaboration in temperature tracking, pressure balance, and air-fuel ratio optimization. The key is that the model internally incorporates the coupling relationship between calorific value changes and temperature, as well as the coupling relationship between valve actions and pressure; decoupling calculations avoid action conflicts.

[0032] This invention creatively sets an adaptive constraint on pressure safety boundaries: a dynamic pressure constraint boundary is introduced into the MPC controller. Based on the current gas flow rate and calorific value, the upper and lower limits of safe kiln head pressure (to prevent flameout and backfire) are calculated in real time, and these boundaries are added as hard constraints to the optimization problem. When a potential pressure exceedance is predicted, the controller automatically prioritizes pressure safety, appropriately sacrificing the temperature regulation rate to ensure combustion safety.

[0033] Compared with the prior art, the beneficial effects of this invention are: (1) This invention fundamentally eliminates interference from calorific value fluctuations and achieves proactive compensation. By collecting the calorific value of the gas in real time and performing filtering preprocessing, the calorific value feedforward compensation amount is calculated, and a reference value of the gas volume after compensation is generated. This feedforward mechanism operates before the temperature changes, realizing a leap from passively waiting for temperature changes to actively compensating for heat changes. Compared with the blind adjustment state in the prior art that only uses the gas flow rate as a feedforward signal and cannot sense calorific value fluctuations, this invention can control the temperature fluctuation of the calcination zone within ±5℃ when the calorific value fluctuation is ±10%, which is more than 60% lower than the traditional control.

[0034] (2) This invention uses a multivariate predictive controller (MPC) to coordinate the scheduling of the gas regulating valve, the combustion fan frequency converter, and the induced draft fan frequency converter. The model explicitly includes the rapid positive pressure coupling characteristics of the gas valve opening on the kiln head pressure, the positive pressure coupling characteristics of the combustion fan speed on the kiln head pressure, and the rapid negative pressure regulation characteristics of the induced draft fan speed on the kiln head pressure. Decoupling calculations are used to avoid action conflicts. When the calorific value decreases and the gas valve needs to be opened wider and the combustion fan speed increased, the MPC synchronously calculates that the induced draft fan should increase its suction in advance, so that the kiln head pressure remains stable after the three actions are coordinated. Compared with the drastic pressure fluctuations caused by the independent control of each actuator in the prior art, the kiln head pressure fluctuation range of this invention can be reduced from ±50Pa to within ±10Pa, eliminating safety hazards such as smoke and backfire.

[0035] (3) This invention introduces a dynamic pressure constraint boundary in the MPC controller. Based on the current gas flow rate and effective calorific value, combined with the characteristics of burner deflaming and tempering limit flow rates, the lower limit of the deflaming critical pressure and the upper limit of the tempering critical pressure are calculated in real time. This boundary is embedded as a hard constraint in the rolling optimization objective function. When it is predicted that the pressure may exceed the limit, the controller automatically prioritizes meeting the pressure safety constraint and appropriately sacrifices the temperature regulation rate. Compared with the passive protection methods of fixed threshold alarms or interval control in the prior art, the dynamic boundary of this invention changes in real time with the operating conditions, realizing the unity of safety and adjustability. No abnormal operating condition alarms such as deflaming or tempering occurred during operation.

[0036] (4) Since the calorific value is known and the air-fuel ratio can be precisely matched, the control quantity benchmark deviation term in the MPC rolling optimization objective function is based on the compensated gas volume reference value, so that the combustion blower speed and gas flow rate are precisely matched. Under the premise of ensuring complete combustion, the excess air coefficient is reduced to near the theoretical value, reducing the heat loss carried away by excess air. Compared with the existing technology where operators conservatively use an experience-based adjustment method with a relatively large air-fuel ratio, this invention can reduce the excess air coefficient to near the theoretical value under the premise of ensuring complete combustion, reducing flue gas heat loss, and the coal consumption per ton of ash is expected to be reduced by 2%-4%. At the same time, dynamic pressure constraints ensure that combustion is always within a safe window.

[0037] (5) The system automatically completes the entire process of calorific value acquisition, filtering preprocessing, feedforward compensation calculation, feedback value reading, constraint boundary setting, rolling optimization solution, and control command output in each control cycle without manual intervention. The system can autonomously cope with frequent disturbances in the gas system, greatly reducing the frequency of operator adjustments and labor intensity, and realizing unmanned and intelligent operation of the rotary kiln combustion system.

[0038] This invention is particularly applicable to rotary kilns for quicklime production in steel enterprises that use low-calorific-value by-product gas as fuel. It addresses the problems of unstable combustion, drastic pressure fluctuations in the kiln, and product quality fluctuations caused by frequent fluctuations in the calorific value of the gas. Through online calorific value sensing, dynamic compensation with calorific value feedforward, and multivariate predictive collaborative control, it achieves temperature stability, pressure balance, and optimized combustion efficiency in the rotary kiln combustion process. Attached Figure Description

[0039] The present invention will be further described below with reference to the accompanying drawings and embodiments. In the accompanying drawings: Figure 1 This is a schematic diagram of the intelligent scheduling and pressure balance control system for rotary kiln combustion that takes into account fluctuations in the calorific value of coal gas, as described in this invention.

[0040] Figure 2 This is a control flowchart of the intelligent scheduling and pressure balance control method for rotary kiln combustion that takes into account the fluctuation of gas calorific value in this invention. Detailed Implementation

[0041] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0042] Example 1 Figure 1 This is a schematic diagram of the intelligent scheduling and pressure balance control system for rotary kiln combustion that takes into account fluctuations in the calorific value of coal gas, as described in this invention.

[0043] The system consists of two main parts: the left-side rotary kiln body and detection and execution unit, and the right-side calorific value feedforward compensation and MPC multivariable prediction and control unit. These parts are connected by signal lines to form a closed-loop control system. This constitutes the complete technical solution of this invention, from "calorific value sensing → feedforward compensation → multivariable collaborative prediction and control → actuator linkage," reflecting the hardware implementation architecture of the two core innovative points: calorific value feedforward and MPC multivariable decoupling.

[0044] The tilted, rotating cylindrical kiln body is used for calcining quicklime; the gas pipeline (with a calorimeter and regulating valve) connects to the rotary kiln body and supplies blast furnace gas / converter gas into the kiln; an online calorimeter is installed on the pipeline to monitor the calorific value of the gas in real time, and a gas regulating valve is installed to receive control commands and adjust the opening degree; the combustion air blower blows combustion air into the rotary kiln body to maintain the oxygen required for combustion; the induced draft fan extracts flue gas from the tail of the rotary kiln to maintain the negative pressure balance inside the kiln.

[0045] The calorific value feedforward compensator receives the gas calorific value signal from the calorific value meter, calculates the gas flow feedforward compensation amount when the calorific value fluctuates, and outputs the compensated gas flow reference value to the MPC multivariate predictive controller. The MPC multivariable predictive controller is the core control unit of the system. It receives three input signals: (1) feedback signals of temperature, pressure and oxygen content; (2) reference value of gas volume after compensation; and (3) internally set dynamic pressure constraint boundary. Through rolling optimization solution, it outputs three control commands.

[0046] The gas regulating valve receives control commands from the MPC and regulates the gas flow rate; the combustion fan frequency converter receives control commands from the MPC and regulates the combustion fan speed; the induced draft fan frequency converter receives control commands from the MPC and regulates the induced draft fan speed.

[0047] The calorimeter's detection signal is output to the calorific value feedforward compensator; the temperature, pressure, and oxygen content feedback signals are output to the MPC multivariable predictive controller.

[0048] Table 1 Coupling relationships between components

[0049] Referring to Table 1, the coupling relationships of each component illustrate the core of MPC decoupling as follows: When the calorific value decreases and it is necessary to open the gas valve (positive pressure impact) and increase the combustion air (positive pressure impact), the MPC synchronously calculates that the induced draft fan should increase its suction force in advance so that the kiln head pressure remains stable after the three work together.

[0050] Compared to existing gas flow feedforward structures, this invention adopts gas calorific value feedforward, and the controller structure is replaced by an MPC multivariable predictive controller (three inputs and three outputs unified optimization) instead of a cascaded PID + independent feedforward module; in terms of pressure control, the dynamic pressure constraint of this invention is embedded in the MPC (hard constraint priority). Furthermore, the actuators are not adjusted independently, but rather the three frequency converters are coordinated and decoupled.

[0051] The safety mechanism is no longer a simple over-limit alarm, but a predictive over-limit superposition and automatic priority pressure protection; The aforementioned advantages constitute the innovative technical solution of this invention, which involves "calorific value sensing → feedforward compensation → multi-variable collaborative predictive control → actuator linkage".

[0052] Figure 2 This is a control flowchart of the intelligent scheduling and pressure balance control method for rotary kiln combustion that considers fluctuations in the calorific value of coal gas, as described in this invention. The entire process is a cyclic execution structure, sequentially executing calorific value acquisition, filtering preprocessing, feedforward compensation calculation, feedback value reading, MPC constraint setting, rolling optimization solution, and control command output, finally returning to the loop starting point to form a closed-loop periodic control.

[0053] Specifically, it includes: (1) Control cycle start node, initialize or enter the next cycle; (2) Real-time acquisition of gas calorific value Q(t): Obtain the current gas calorific value signal from the online calorific value meter or soft measurement model on the gas pipeline; (3) Calorific value filtering and preprocessing: The acquired calorific value signal is smoothed and filtered to remove measurement noise and obtain the effective calorific value; (4) Calculate the feedforward compensation amount of calorific value: Based on the target total heat input and the effective value of calorific value, calculate the reference gas flow rate and the feedforward compensation amount of gas flow rate, and generate the reference value of gas flow rate after compensation. (5) Collect real-time feedback signals of kiln tail temperature (T), kiln head pressure (P), and flue gas oxygen content (O2); (6) MPC reads set values ​​(T_set, P_set, O2_set) and dynamic pressure constraint boundaries: The MPC controller reads the set values ​​of the three controlled variables and calculates the dynamic safety constraint boundaries of the kiln head pressure (lower limit of deflaming critical pressure and upper limit of tempering critical pressure) in real time based on the current gas flow rate and effective calorific value. (7) MPC rolling optimization to solve the control increment of gas valve, combustion fan and induced draft fan: core calculation steps: construct the prediction output equation based on the step response coefficient matrix model, establish the rolling optimization objective function containing deviation term, smoothing term and benchmark deviation term, solve the constrained quadratic programming problem, and output the coordinated control increment of the three manipulated variables; (8) Output control commands to the actuators: Send the control increments obtained from the solution to the gas regulating valve, the combustion fan frequency converter, and the induced draft fan frequency converter for execution; (9) Return to start (loop execution): The control cycle ends and waits for the next cycle to be triggered, forming a periodic closed-loop control.

[0054] Steps (2)-(3)-(4) constitute the calorific value feedforward channel. Calorific value fluctuation → immediate calculation of compensation amount → MPC advance action (advance); the compensation amount calculation is completed before the temperature changes, realizing the leap from passive feedback to active feedforward.

[0055] The three-in-one approach of MPC: prediction, constraint, and optimization. Prediction: Step (6) reads the set value, Step (7) performs rolling optimization to solve for future multi-step control increments; Constraints: Step (6) embeds the dynamic pressure constraint boundary as a hard constraint into the optimization problem; Optimization: Step (7) simultaneously optimizes temperature tracking, pressure balance, oxygen content control and actuator smoothness.

[0056] The key to step (7) is that it is not a separate calculation of the three independent loops, but a unified optimization solution. When the calorific value of the gas decreases, the multivariate predictive controller's coordinated scheduling logic for the three actuators is as follows: The gas regulating valve is opened more to compensate for the reduced heat input due to the decrease in calorific value, thus maintaining a constant total heat input.

[0057] The combustion blower speed is increased to match the increased gas flow, ensuring a sufficient oxygen supply and maintaining the optimal air-fuel ratio required for complete combustion.

[0058] The increased speed of the induced draft fan enhances the suction force, offsetting the sudden increase in positive pressure at the kiln head caused by the opening of the gas valve and the acceleration of the combustion fan, thus achieving a pressure balance without disturbance.

[0059] The adjustment increments of the three actuators mentioned above are solved simultaneously by a multivariate predictive controller in a unified optimization problem, rather than being calculated separately by three independent control loops. This optimization problem has the control objectives of tracking the kiln tail temperature setpoint, maintaining stable kiln head pressure, and ensuring optimal oxygen content in the flue gas. The variables to be solved are the gas valve opening, combustion fan speed, and induced draft fan speed, with a dynamic pressure safety boundary as a hard constraint. When a potential exceedance of the kiln head pressure is predicted, the controller automatically prioritizes meeting the pressure safety constraint, appropriately sacrificing the temperature regulation rate to ensure the combustion system always operates within a safe operating window.

[0060] Example 2 This implementation case applies to a rotary kiln for quicklime in a steel company, specifically a rotary kiln for quicklime with a daily output of 600 tons that uses blast furnace gas.

[0061] System architecture: The system consists of three parts: a calorific value sensing module, a multivariable predictive controller (MPC), and actuators (gas regulating valve, combustion fan frequency converter, and induced draft fan frequency converter).

[0062] Specific steps of the control method: Step S1: Calorific value sensing and preprocessing.

[0063] a. Real-time calorific value acquisition: The current calorific value Q_gas (t) of the gas is read in real time from the online calorific value meter on the gas pipeline (or calculated based on the gas composition analyzer).

[0064] b. Calorific value prediction: For cases where the calorific value meter has a large response delay, a short-term calorific value prediction model based on the upstream process conditions (such as blast furnace air volume and converter blowing stage) can be established to predict the calorific value change trend 1-2 minutes in advance.

[0065] c. Heat value smoothing filtering: The heat value signal is filtered to remove measurement noise and obtain the effective heat value Q_eff for control.

[0066] Step S2: Calculation of calorific value feedforward compensation.

[0067] a. Set the current target total heat input to H_set (determined by the upper-level process requirements).

[0068] b. Based on the current calorific value Q_eff, calculate the baseline gas flow rate F_fuel_base = H_set / Q_eff required to maintain the total heat.

[0069] c. When the calorific value changes, the feedforward compensator calculates the gas flow compensation amount ΔF_fuel_ff: ΔF_fuel_ff = K × (1 / Q_eff - 1 / Q_eff_prev) × H_set, where K is the compensation intensity coefficient and Q_eff_prev is the calorific value at the previous moment.

[0070] d. The final gas flow feedforward reference value fed into the MPC is F_fuel_ff_ref = F_fuel_base + ΔF_fuel_ff.

[0071] Step S3: Multivariate predictive collaborative control.

[0072] S3.1 Constructing the MPC prediction model: Establish a 3x3 dynamic matrix model with the gas valve opening (MV1), combustion fan speed (MV2), and induced draft fan speed (MV3) as inputs, and kiln tail temperature (CV1, representing calcination intensity), kiln head pressure (CV2), and flue gas oxygen content (CV3) as outputs. The model needs to include the following key dynamic characteristics: The hysteresis characteristic of MV1 to CV1 (large time constant); that is, after the gas valve is opened, the kiln tail temperature will take a long time to rise. The rapid coupling of MV1 to CV2 (the instantaneous opening of the gas valve will cause positive pressure at the kiln head); that is, the instant the gas valve is opened, the kiln head pressure immediately increases to positive pressure. MV2 is coupled to CV1 and CV2 (increasing airflow will help combustion and increase temperature, while also increasing pressure); however, this positive gain decreases as the air-fuel ratio exceeds a preset threshold.

[0073] The heating effect is obvious when the air volume is moderate; if the air volume is too large, too much air will carry away heat, thus weakening the heating effect and reducing the temperature rise gain per unit air volume.

[0074] MV3 has a direct and rapid effect on CV2 (draft adjustment). That is, after increasing the induced draft fan, the kiln head pressure immediately increases negatively (positively) (decreases).

[0075] Specifically, a step response coefficient matrix model, a predicted output equation, and a rolling optimization objective function are constructed. In the matrix: the gas valve opening to the kiln tail temperature channel has a large hysteresis characteristic; the gas valve opening to the kiln head pressure channel has a fast positive pressure coupling characteristic; the combustion fan speed to the kiln tail temperature channel has a coupling characteristic where the gain decreases as the air-fuel ratio exceeds the limit; the combustion fan speed to the kiln head pressure channel has a fast positive pressure coupling characteristic; and the induced draft fan speed to the kiln head pressure channel has a fast negative pressure regulation characteristic.

[0076] S3.2 Set dynamic constraint boundaries: Pressure safety boundary: Based on the current gas flow rate F_fuel and burner characteristics, the deflaming critical pressure P_min and tempering critical pressure P_max are calculated in real time. The constraint on the kiln head pressure CV2 is set as [P_min + Margin, P_max - Margin], which serves as a hard constraint for MPC optimization.

[0077] Oxygen content boundaries: Set the upper limit (energy saving) and lower limit (to prevent incomplete combustion) of the oxygen content CV3 in flue gas.

[0078] S3.3 Scrolling optimization solution: In each control cycle (e.g., T=10 seconds), MPC reads the current CVs feedback values ​​(T_act, P_act, O2_act).

[0079] Read the calorific value feedforward reference value F_fuel_ff_ref calculated in step S2 and use it as part of the target reference trajectory of MV1 (gas quantity).

[0080] MPC solves a constrained quadratic programming optimization problem with the goal of making the CVs track the setpoints (T_set, P_set, O2_act) while making the changes in MVs as gradual as possible. The optimization result outputs a sequence of MVs increments for the next N steps, but only the increment for the first step is distributed.

[0081] The optimized calculations of ΔMV1, ΔMV2, and ΔMV3 are mutually coordinated.

[0082] For example, when the calorific value decreases, ΔMV1 is positive (the gas valve is opened wider), ΔMV2 automatically calculates the matching air volume increment, and ΔMV3 is automatically adjusted according to the pressure change to ensure that the positive pressure caused by the moment the gas valve is opened is offset by the advance action of the induced draft fan, thus achieving pressure undisturbed operation.

[0083] Step S4: Execution and Feedback.

[0084] The gas regulating valve, the combustion fan frequency converter, and the induced draft fan frequency converter respectively receive and execute control commands.

[0085] The system enters the next control cycle, forming a closed loop.

[0086] Comparison of results before and after using the control method of this invention: Operating conditions: The kiln uses blast furnace gas, whose calorific value fluctuates frequently between 700-900 kcal / Nm³, with irregular fluctuation cycles.

[0087] Implementation status: The system has been running continuously and stably for 2 months, experiencing multiple periods of sudden increases and decreases in calorific value.

[0088] Temperature stability: The standard deviation of the kiln tail temperature decreased from ±12℃ before commissioning to ±4℃.

[0089] Pressure stability: The pressure fluctuation range at the kiln head has been reduced from ±60Pa to ±15Pa, and smoke emission has been basically eliminated.

[0090] Energy consumption: Gas consumption per ton of lime decreased by an average of 3.1%.

[0091] Safety: No alarms were triggered for abnormal operating conditions such as flameout or backfire during operation.

[0092] Conclusion: Experiments have shown that this method can effectively cope with fluctuations in the calorific value of coal gas and achieve synergistic control of combustion stability and pressure balance, thus possessing significant value for widespread application.

[0093] It should be understood that those skilled in the art can make improvements or modifications based on the above description, and all such improvements and modifications should fall within the protection scope of the appended claims.

Claims

1. A rotary kiln combustion intelligent scheduling and pressure balance control method considering coal gas heat value fluctuation, characterized in that, Includes the following steps: S1. Real-time acquisition of gas calorific value signal and filtering preprocessing to obtain effective calorific value; S2. Based on the target total heat input and the effective value of calorific value, calculate the reference gas flow rate and the gas flow rate feedforward compensation amount when the calorific value changes, and generate the reference value of gas flow rate after compensation. S3. Construct a multivariate predictive control model with the gas valve opening, combustion fan speed, and induced draft fan speed as manipulated variables, and the kiln tail temperature, kiln head pressure, and flue gas oxygen content as controlled variables. The model includes the large lag characteristic of the gas valve opening on the kiln tail temperature, the rapid coupling characteristic of the gas valve opening on the kiln head pressure, the coupling characteristic of the combustion air volume on the kiln tail temperature and the kiln head pressure, and the direct effect characteristic of the induced draft fan speed on the kiln head pressure. S4. Based on the current gas flow rate and effective calorific value, calculate the safety constraint boundary of the kiln head pressure in real time, and add the safety constraint boundary as a hard constraint to the multivariate predictive control model; the safety constraint boundary includes the lower limit of the de-flaming critical pressure and the upper limit of the tempering critical pressure calculated in real time based on the current gas flow rate and burner characteristics. S5. In each control cycle, read the feedback value of the controlled variable and the reference value of the compensated gas flow rate, solve the constrained quadratic programming problem through rolling optimization, output the collaborative control increment of the manipulated variable, and send it to the gas regulating valve, combustion fan frequency converter and induced draft fan frequency converter for execution. The rolling optimization includes: constructing a step response coefficient matrix model as a multivariable predictive control model; constructing a predictive output equation and a rolling optimization objective function containing a deviation term, a smoothing term, and a baseline deviation term based on the matrix; reading the feedback value of the controlled variable and the reference value of the gas quantity after compensation in each control cycle; solving a constrained quadratic programming problem; and outputting the coordinated control increment of the three manipulated variables.

2. The intelligent scheduling and pressure balancing control method of a rotary kiln combustion considering fluctuations in the calorific value of the gas according to claim 1, characterized in that, The gas calorific value signal mentioned in step S1 is acquired in real time by an online gas calorific value meter, or predicted by a soft measurement model based on the upstream process conditions.

3. The method of claim 1, wherein, In step S2, the calorific value feedforward compensation amount is calculated based on the change in the effective calorific value and the compensation intensity coefficient. The reference value of the gas volume after compensation is equal to the sum of the reference gas flow rate and the calorific value feedforward compensation amount.

4. The intelligent scheduling and pressure balance control method for rotary kiln combustion considering gas calorific value fluctuations according to claim 1, characterized in that, The objective of the rolling optimization solution in step S5 is to make the controlled variable track the set value, minimize the change of the manipulated variable, and prioritize satisfying the safety constraint boundary when the kiln head pressure is predicted to exceed the limit.

5. A rotary kiln combustion intelligent scheduling and pressure balance control system considering fluctuations in the calorific value of coal gas, characterized in that, The steps for implementing the method according to any one of claims 1-4 include: The calorific value sensing module is used to collect the calorific value signal of coal gas in real time and perform filtering preprocessing; The calorific value feedforward compensator is used to calculate the reference gas flow rate and the gas flow rate feedforward compensation amount based on the target total heat input and the effective calorific value, and generate a reference value of gas flow rate after compensation. A multivariate predictive controller is used to construct a predictive control model with gas valve opening, combustion fan speed, and induced draft fan speed as manipulated variables, and kiln tail temperature, kiln head pressure, and flue gas oxygen content as controlled variables. It calculates the safety constraint boundary of the kiln head pressure in real time based on the current gas flow rate and effective calorific value, and outputs the collaborative control increment through rolling optimization. The feedback signal acquisition module is used to read the feedback values ​​of the current kiln tail temperature, kiln head pressure, and flue gas oxygen content. The actuators, including the gas regulating valve, the combustion fan frequency converter, and the induced draft fan frequency converter, are used to receive and execute the cooperative control increments.

6. The intelligent scheduling and pressure balance control system for rotary kiln combustion considering fluctuations in the calorific value of coal gas as described in claim 5, characterized in that, The predictive control model of the multivariable predictive controller includes the large lag characteristic of the gas valve opening on the kiln tail temperature, the rapid coupling characteristic of the gas valve opening on the kiln head pressure, the coupling characteristic of the combustion air volume on the kiln tail temperature and the kiln head pressure, and the direct effect of the induced draft fan speed on the kiln head pressure.

7. The intelligent scheduling and pressure balance control system for rotary kiln combustion considering fluctuations in the calorific value of coal gas as described in claim 5, characterized in that, The multivariate predictive controller reads the feedback value of the controlled variable and the reference value of the compensated gas flow rate in each control cycle. It solves the constrained quadratic programming problem through rolling optimization to make the controlled variable track the set value and minimize the change of the manipulated variable. When the kiln head pressure is predicted to exceed the limit, the safety constraint boundary is satisfied first.

8. A computer-readable storage medium storing a program thereon, characterized in that: When the program is executed by a processor, it implements the steps of the method described in any one of claims 1-4.

Images