Online tuning method and system for sintering waste heat power generation

Abstract

The invention discloses an online tuning method and system for sintering waste heat power generation, and solves the problems of unstable operation and limited power generation efficiency of a waste heat power generation system in the prior art. The method comprises the following steps: obtaining real-time operation data of a whole process of a sintering waste heat power generation process; on the basis of real-time data, the sensible heat potential of sinter entering the circular cooler is predicted online through a firing quantity prediction model; on the basis of the sensible heat potential and to-be-optimized process control parameters, the sintering waste heat power generation power is predicted on line through a full-process energy efficiency model; solving and determining an optimal process control parameter corresponding to the current sintering ore sensible heat potential by adopting a constrained programming algorithm by taking generated power maximization as a target; and finally, closed-loop tuning control is carried out on the sintering waste heat power generation system based on the optimal parameters. According to the method, the upstream waste heat fluctuation can be accurately sensed, the downstream operation parameters are collaboratively optimized in a feedforward manner, and the power generation efficiency and the operation stability of the non-afterburning type sintering waste heat power generation system are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of energy-saving technology in iron and steel metallurgy, and in particular to an online optimization control technology for sintering waste heat recovery power generation. Specifically, it is an online optimization method and system for sintering waste heat power generation based on a data-driven model for forward prediction and reverse optimization. Background Technology

[0002] Sintering is one of the major energy-consuming processes in steel production, accounting for approximately 9% to 12% of the industry's total energy consumption. Simultaneously, the sintering process generates a significant amount of waste heat, approximately 1.4 GJ of waste heat per ton of sinter produced. Therefore, the efficient recovery and utilization of waste heat from the sintering process, particularly through waste heat power generation, is a key technological pathway for the steel industry to achieve its energy conservation and emission reduction goals.

[0003] A typical sintering waste heat power generation process is as follows: Figure 1 As shown, the system mainly includes a waste heat recovery system and a waste heat utilization system. The waste heat recovery system uses a circulating fan to drive circulating air, forcibly penetrating the high-temperature sintered ore in the waste heat recovery section of the annular cooler to absorb its heat, forming two circulating air paths: one high-temperature and one low-temperature. These two paths of circulating air enter a dual-pressure waste heat boiler, heating feedwater to generate medium-pressure superheated steam and low-pressure superheated steam with specific pressure and temperature. The waste heat utilization system uses this superheated steam to drive a steam turbine, converting thermal energy into mechanical energy, which is then converted into electrical energy by a generator. To ensure the final cooling effect of the sintered ore, the cooling section of the annular cooler also uses a blower for forced air cooling, with exhaust gas directly discharged.

[0004] However, the sintering production process itself is volatile. Factors such as the start-up and shutdown of the sintering machine, adjustments to output, and changes in the composition of the sintered ore can all cause fluctuations in the temperature and quantity of the sintered ore entering the annular cooler, meaning the source supply of waste heat resources is unstable. This volatility poses a significant challenge to the stable and efficient operation of the subsequent waste heat power generation system.

[0005] To address the issue of fluctuating waste heat sources, existing technologies have proposed several solutions. For example, Chinese invention patent application CN200810118721.7 (publication date January 26, 2011) discloses a CO2 cycle and gas-assisted combustion method for sintering waste heat power generation, which stabilizes steam parameters by adjusting the amount of gas burned when the sinter temperature fluctuates through the installation of a gas-assisted combustion unit. Chinese invention patent CN201510243051.1 (publication date April 12, 2017) discloses a sintering waste heat power generation device and waste heat utilization method with an external superheated combustion unit, which stabilizes steam parameters by installing a superheated combustion unit. These solutions all employ the addition of an auxiliary heat source (combustion) to mitigate fluctuations. While this ensures the stability of power generation, it increases system complexity and operating costs, and is generally only applicable to new projects, making it difficult to apply to the numerous existing domestic sintering waste heat power generation systems without combustion assistance.

[0006] Another Chinese invention patent, CN202311239206.5 (publication date December 15, 2023), discloses a method to improve the load stability of sintering waste heat power generation. It proposes to introduce the sinter temperature of the sintering ring cooler into the waste heat power generation unit control system. However, this scheme does not explain how to use the sinter temperature to quantitatively analyze the impact of upstream process fluctuations, nor does it reveal the intrinsic correlation and synergistic optimization method between the sinter temperature and other key process parameters (such as circulating air volume, steam pressure, etc.).

[0007] In summary, the existing technology has the following problems: 1) For waste heat power generation systems without supplementary combustion, there is a lack of effective means to deal with fluctuations in waste heat at the source, which limits the power generation efficiency.

[0008] 2) Sintering ring cooling control and waste heat power generation control usually rely on manual experience, which has problems such as delayed response and untimely adjustment.

[0009] 3) Sintering plants and power plants are often independent units in terms of production management, resulting in data barriers and information silos. They lack cross-process collaborative optimization mechanisms, which restricts the further improvement of the potential for sintering waste heat utilization. Summary of the Invention

[0010] Purpose of the invention: This invention aims to solve the above-mentioned problems existing in the prior art and provide an online optimization method and system for sintering waste heat power generation. It breaks down the data barriers between the sintering and power generation processes, predicts the changes in the sensible heat potential of sintered ore online, and performs reverse optimization based on the whole process energy efficiency model to determine the optimal combination of process control parameters in real time. This maximizes the efficiency of sintering waste heat power generation while ensuring the cooling effect of sintered ore.

[0011] Technical Solution: An online optimization method for sintering waste heat power generation includes the following steps: acquiring real-time operating data of the entire sintering waste heat power generation process; based on the real-time operating data, predicting the sensible heat potential of the sinter entering the annular cooler online; based on the sensible heat potential of the sinter and the process control parameters to be optimized, predicting the sintering waste heat power generation output online; based on the predicted sintering waste heat power generation output, solving for and determining the optimal process control parameters corresponding to the current sensible heat potential of the sinter with the goal of maximizing power generation output; and optimizing the control of the sintering waste heat power generation system based on the optimal process control parameters. This solution constructs a closed-loop optimization framework of "prediction-optimization-control," changing the traditional control mode that relies on manual experience and delayed response, enabling the system to proactively respond to fluctuations in the upstream waste heat source, and to coordinately regulate various process parameters as a whole to maximize power generation efficiency.

[0012] Furthermore, the step of online prediction of the sensible heat potential of sinter entering the annular cooler includes: real-time acquisition of the operating parameters of the sintering system, which at least include the image information of the red-hot layer of the tail section acquired by an infrared thermal imaging device; based on the operating parameters of the sintering system, the sintering amount is calculated in real-time using a sintering amount prediction model; and based on the sintering amount and temperature, the sensible heat potential of the sinter is calculated. This solution solves the core problem of the difficulty in real-time online measurement of sintering amount. By introducing infrared thermal imaging as an information source and combining it with a prediction model, it achieves the advance and accurate quantification of the heat load entering the waste heat recovery system, providing a key and scientific decision-making basis for subsequent proactive and pre-regulation.

[0013] Furthermore, the sintering quantity prediction model is a gradient boosting decision tree model; the input features of the model include: sintering fabric thickness, sintering trolley width, sintering machine speed, and the maximum area of ​​the red-fired layer at the tail section calculated based on the image information of the red-fired layer at the tail section; the output of the model is the sintering quantity of the sintered ore. By employing a gradient boosting decision tree model with strong nonlinear fitting ability, and using the feature "maximum area of ​​the red-fired layer at the tail section," which is highly correlated with the sintering endpoint, as the model input, the accuracy and robustness of the sintering quantity prediction are improved, ensuring the reliability of the heat source potential prediction.

[0014] Furthermore, the calculation steps for the maximum area of ​​the red-fired layer at the tail section include: continuously acquiring multiple frames of infrared thermal imaging images during the material turning process of the sintering machine tail trolley; performing image recognition on each frame to calculate the contour area of ​​the red-fired layer; and selecting the maximum contour area calculated from all frames as the maximum area of ​​the red-fired layer at the tail section. This step provides a specific method for converting qualitative observations (red-fired layer topography) during the sintering process into precise quantitative data. It uses image processing technology to stably and reliably extract key feature parameters, providing high-quality data input for the establishment of a high-precision model.

[0015] Furthermore, the step of online prediction of sintering waste heat power generation is achieved through a full-process energy efficiency model. The input features of the full-process energy efficiency model include: the sensible heat potential of the sintered ore, the operating parameters of the annular cooler, the operating parameters of the circulating air system, the operating parameters of the waste heat boiler, and the operating parameters of the steam system. The output of the full-process energy efficiency model is the sintering waste heat power generation. By constructing an energy efficiency model covering the entire process from heat source to power conversion, the data barriers and information silos between the two production units of sintering and power generation are broken down, and a digital end-to-end mapping relationship is established. This enables the system to quantitatively evaluate the impact of any set of process parameters on the final power generation, which is the foundation for achieving global optimization.

[0016] Furthermore, the step of solving and determining the optimal process control parameters includes: establishing an optimization model with maximizing power generation as the objective function; using the process control parameters to be optimized as decision variables, and the safe operating range of the equipment as constraints; and using a constraint programming algorithm to solve the optimization model to obtain the optimal process control parameters. The process control parameters to be optimized include at least: the speed of the annular cooler, the frequency of the circulating fan, the medium-pressure steam pressure, and the low-pressure steam pressure. This step transforms the predictive model into a guiding optimization tool. Through mathematical programming, it can scientifically calculate the combination of multivariate control parameters that enables the power generation to reach its theoretical maximum value, while satisfying all safety and process constraints. This replaces manual trial and error and local optimization methods, achieving scientific and optimized control decisions.

[0017] Furthermore, the method also includes: dividing the sensible heat potential of the sinter into multiple preset intervals; for each preset interval, pre-calculating and storing a set of optimal process control parameters to form an operating condition-parameter database; the step of optimizing the control of the sinter waste heat power generation system specifically involves: querying the corresponding optimal process control parameters from the operating condition-parameter database based on the preset interval to which the currently calculated sensible heat potential of the sinter belongs, and sending them to the control system for execution. This step transforms the complex online optimization calculation problem into a fast table lookup problem, reducing the resource consumption and time delay of online calculations, and facilitating deployment and application in industrial settings.

[0018] This solution also relates to an online optimization system for sintering waste heat power generation, characterized by comprising: a data acquisition module for acquiring real-time operating data of the entire sintering waste heat power generation process; a processing module connected to the data acquisition module for: predicting the sensible heat potential of the sinter entering the annular cooler online based on the real-time operating data; predicting the sintering waste heat power generation power online based on the sensible heat potential of the sinter and the process control parameters to be optimized; solving and determining the optimal process control parameters corresponding to the current sensible heat potential of the sinter based on the predicted sintering waste heat power generation power, with the goal of maximizing power generation; and a control module connected to the processing module for generating control commands based on the optimal process control parameters to optimize the control of the sintering waste heat power generation system.

[0019] Furthermore, the processing module's function of predicting the sensible heat potential of the sinter entering the annular cooler online is achieved by running a firing amount prediction model. Specifically, the processing module is used to: acquire the operating parameters of the sintering system from the data acquisition module, the operating parameters of the sintering system including at least the red flame layer image information of the tail section obtained by infrared thermal imaging equipment; calculate the firing amount of the sinter in real time based on the operating parameters of the sintering system through the firing amount prediction model; and calculate the sensible heat potential of the sinter based on the firing amount and the temperature of the sinter.

[0020] Furthermore, the sintering quantity prediction model is a gradient boosting decision tree model; the input features of the model include: sintering material thickness, sintering trolley width, sintering machine speed, and the maximum area of ​​the red-fired layer at the tail section calculated based on the red-fired layer image information at the tail section; the output of the model is the sintering quantity of the sintered ore.

[0021] Furthermore, the processing module is also used to perform the following steps to calculate the maximum area of ​​the red-fired layer at the tail section: continuously acquiring multiple frames of infrared thermal imaging images during the material turning process of the sintering machine tail trolley; performing image recognition on each frame of image to calculate the outline area of ​​the red-fired layer; and selecting the maximum outline area calculated in all frames as the maximum area of ​​the red-fired layer at the tail section.

[0022] Furthermore, the processing module's function of predicting the power generation from sintering waste heat online is achieved by running a full-process energy efficiency model; the input features of the full-process energy efficiency model include: the sensible heat potential of the sintered ore, the operating parameters of the annular cooler, the operating parameters of the circulating air system, the operating parameters of the waste heat boiler, and the operating parameters of the steam system; the output of the full-process energy efficiency model is the power generation from sintering waste heat.

[0023] Furthermore, the processing module's function of solving for and determining the optimal process control parameters is achieved by running an optimization model. Specifically, the processing module is used to: establish an optimization model with maximizing power generation as the objective function; use the process control parameters to be optimized as decision variables and the equipment safe operating range of the decision variables as constraints; and use a constraint programming algorithm to solve the optimization model to obtain the optimal process control parameters. The process control parameters to be optimized include at least: the speed of the annular cooler, the frequency of the circulating fan, the medium-pressure steam pressure, and the low-pressure steam pressure.

[0024] Furthermore, the system also includes a storage module for storing a condition-parameter database; the condition-parameter database stores multiple preset sensible heat potential ranges of sinter and optimal process control parameters corresponding to each range; the control module is specifically used to: obtain the current sensible heat potential of sinter from the processing module, query the condition-parameter database based on the potential, obtain the corresponding optimal process control parameters, and generate control instructions accordingly.

[0025] This solution also relates to a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method described above.

[0026] Beneficial effects: Compared with the prior art, the present invention has the following beneficial effects: By collecting the full-process time-series data of sintering ring cooling, circulating fan, waste heat boiler and steam turbine unit, and constructing a unified energy efficiency model, the two production links of sintering and power generation, which were originally isolated from each other, are closely linked at the data level and control logic, and a two-way data channel is established, laying the foundation for achieving global optimization. Attached Figure Description

[0027] Figure 1 This is a typical process flow diagram for sintering waste heat power generation. Figure 2 This is a flowchart of the method of the present invention; Figure 3 This is a schematic diagram of the red-fired layer at the tail of the sintering machine. Figure 4 for Figure 3 Corresponding contour image diagram; Figure 5 Schematic diagram of pre-defined intervals for the sensible heat potential of sintered ore; Figure 6 This is a system structure diagram of the present invention. Detailed Implementation

[0028] The present invention will be further illustrated below with reference to the accompanying drawings and specific embodiments. These embodiments are for illustrative purposes only and are not intended to limit the scope of the invention.

[0029] This invention provides an online optimization method, system, and computer-readable storage medium for sinter waste heat power generation. The aim is to maximize power generation by accurately predicting the waste heat potential of sintered ore and optimizing key process parameters of the waste heat power generation system based on this prediction.

[0030] This embodiment provides a specific process for an online optimization method for sintering waste heat power generation, such as... Figure 2 As shown, this method can dynamically adapt to changes in sintering conditions, especially fluctuations in the sensible heat potential of the sinter entering the annular cooler, thereby optimizing power generation efficiency in real time.

[0031] S1: Obtain real-time operating data for the entire sintering waste heat power generation process; In this embodiment, a comprehensive data acquisition system needs to be built first. Through a distributed control system (DCS), a programmable logic controller (PLC), and various sensors, the process data of the entire sintering waste heat power generation process can be acquired in real time and continuously.

[0032] The specific data collected includes, but is not limited to: 1) Sintering system operating parameters: sintering machine speed (m / min), sintering material thickness (mm), sintering trolley width (m), etc.

[0033] 2) Tail-end infrared thermal imaging information: In the tail-end trolley unloading area of ​​the sintering machine, one or more infrared thermal imaging devices are installed to continuously capture infrared thermal images of the sinter during the unloading process.

[0034] 3) Operating parameters of the annular cooler: machine speed (r / min), fan opening degree or frequency (% or Hz) of each section, inlet and outlet material temperature, etc.

[0035] 4) Operating parameters of the circulating air system: circulating fan frequency (Hz), circulating air temperature (°C), circulating air pressure (Pa), etc.

[0036] 5) Waste heat boiler operating parameters: boiler feedwater flow rate (t / h), feedwater temperature (°C), feedwater pressure (MPa), boiler pressure (MPa), steam temperature at the outlet of each superheater (°C), steam pressure (MPa), etc.

[0037] 6) Steam system operating parameters: flow rate (t / h), pressure (MPa), and temperature (°C) of the main steam (medium-pressure superheated steam) and makeup steam (low-pressure superheated steam) entering the steam turbine.

[0038] 7) Power generation system data: Real-time power generation of generator sets (MW).

[0039] S2: Based on the real-time operating data, predict the sensible heat potential of the sinter entering the annular cooler online: The sensible heat potential of sintered ore is the most critical upstream factor determining waste heat power generation, but it is difficult to measure directly. This embodiment uses a model-based prediction method to calculate this potential value online, specifically including: S2.1 Real-time calculation of sintering amount: First, the maximum area of ​​the red-hot layer at the tail section of the computer is required, specifically including: 1) By using an infrared thermal imaging device installed at the tail of the sintering machine, multiple frames of infrared thermal imaging images are continuously acquired during the material turning process of the tail trolley of the sintering machine, such as... Figure 3 As shown, this is one of the grayscale images from 10-20 frames captured during a material turning process on a trolley.

[0040] 2) Perform image recognition on each frame, identifying the red-hot areas by setting a temperature threshold (e.g., 600℃) and calculating their contour area. Figure 4 As shown, the contour image is in a computable state after processing.

[0041] 3) Select the maximum contour area calculated in all frames as the "maximum area of ​​the red fire layer on the tail section" to characterize the sintering effect at that moment.

[0042] Secondly, the calculated "maximum area of ​​the red-fired layer at the tail section", along with parameters such as the thickness of the sintering fabric, the width of the sintering trolley, and the speed of the sintering machine, are used as input features and fed into a pre-trained sintering amount prediction model.

[0043] In this embodiment, the preferred sintering rate prediction model is a Gradient Boosting Decision Tree (GBDT) model. This model is trained using historical production data and can accurately fit the nonlinear relationship between the input features and the sintering rate (in tons / hour). The model's output is the real-time predicted sintering rate.

[0044] S2.2 Calculation of the sensible heat potential of sinter: According to the principle of energy conservation, the sensible heat potential of sintered ore can be calculated using the following formula: Q0=m*(t0-t) z )*c In the formula, m is the sintering rate (t / h) predicted by the above model.

[0045] t0 is the initial temperature (°C) of the sinter entering the annular cooler, which can be measured by an infrared thermometer.

[0046] t z The target discharge temperature (°C) after the sinter is cooled is set according to process requirements. For example, to ensure the safety of subsequent belt conveyor transport, it is usually required to be below 150°C.

[0047] c represents the average specific heat capacity of the sinter (kJ / (kg·℃)), which can be taken as an empirical value based on metallurgical process data, for example, about 0.95 kJ / (kg·℃).

[0048] This calculation yields the sensible heat potential Q0 of the sinter, which fluctuates in real time with the operating conditions (unit: MJ / h).

[0049] S3: Online prediction of sintering waste heat power generation Based on obtaining the sensible heat potential of sinter, it is necessary to further predict the final power generation of the system under different combinations of downstream process parameters.

[0050] This embodiment achieves this prediction function through a full-process energy efficiency model. This model is a more complex mechanism- and data-driven model, and its input features include: 1) The apparent heat potential of the sinter calculated in the previous step.

[0051] 2) Process control parameters to be optimized, such as: annular cooler speed, circulating fan frequency, medium-pressure steam pressure, and low-pressure steam pressure.

[0052] 3) Operating parameters of other related annular coolers, circulating air, waste heat boilers and steam systems.

[0053] The model outputs the predicted power generation from sintering waste heat. This model can be built based on a deep neural network (DNN) trained on historical data or another set of GBDT models, which learns the complex relationship between the energy conversion efficiency and various process parameters throughout the entire process from heat source input to power output.

[0054] S4: Solve for the optimal process control parameters with the goal of maximizing power generation.

[0055] This step is the core of the optimization control. Based on the full-process energy efficiency model established in the previous step, an optimization problem is constructed.

[0056] 1) Objective function: Maximize the predicted power generation P from sintering waste heat. gen .

[0057] Max(P) gen ) = f(Q potential ,x1,x2,...,x n ) Where f is the full-process energy efficiency model, Q potential Let x1, x2, ..., x be the sensible heat potential of the sinter at the current moment. n These are the process control parameters to be optimized.

[0058] Decision variables: process control parameters to be optimized, which in this embodiment include at least: annular cooler speed, circulating fan frequency, medium-pressure steam pressure and low-pressure steam pressure.

[0059] Constraints: To ensure equipment safety and stable production, it is necessary to set the allowable operating range for each decision variable. For example: v min ≤ Ring cooler speed ≤ v max ; f min ≤ Circulating fan frequency ≤ f max ; p mid_min ≤medium-pressure steam pressure≤p mid_max ; p low_min ≤low-pressure steam pressure≤p low_max ; Solution Algorithm: The optimization model is solved using constraint programming or other advanced optimization algorithms (such as genetic algorithms and particle swarm optimization). Under the premise of satisfying all constraints, the algorithm finds a set of decision variable values ​​that maximizes the objective function (power generation). The solution obtained is the optimal combination of process control parameters corresponding to the current sensible heat potential of the sinter.

[0060] S5: Based on the optimal process control parameters, optimize the control of the sintering waste heat power generation system.

[0061] After the optimizer calculates the optimal parameter combination, the system sends these parameter values ​​(setpoints, SP) to the factory's DCS or PLC system through the communication interface. The DCS / PLC then automatically adjusts the corresponding actuators (such as frequency converters and regulating valves) to achieve closed-loop optimization control of the circulator speed, circulating fan frequency, steam pressure, etc.

[0062] This embodiment provides another specific method for optimizing control, aiming to reduce the complexity and response time of online calculations. Based on Embodiment 1, this embodiment adds a working condition-parameter database.

[0063] The method includes: 1) Establish an offline operating condition-parameter database: First, the possible range of sensible heat potential of sinter is divided into several preset intervals, such as... Figure 5 As shown, the two-dimensional state space formed by the sensible heat potential Q0 and the initial temperature t0 of the sinter is divided into grids. Each rectangular region defined by the potential interval ∆Q0 and the temperature interval ∆t0 is considered as an independent working condition group.

[0064] Then, for each preset interval's typical value (such as the median), the optimization model in S4 of Example 1 is run to calculate the optimal combination of process control parameters under that potential value.

[0065] The potential range and the calculated optimal parameter combination are stored in the database as a correspondence, forming a "condition-parameter" knowledge base.

[0066] 2) Online query and control: In actual operation, the system first calculates the current sensible heat potential of the sinter in real time according to S2 of Example 1.

[0067] Then, the system queries the working condition-parameter database based on the calculated potential value and matches it to the preset range to which it belongs.

[0068] The pre-calculated optimal process control parameters corresponding to the interval are directly extracted from the database.

[0069] Finally, these parameters are sent to the control system for execution.

[0070] This approach shifts complex optimization calculations to offline execution, requiring only rapid table lookup operations during online operation. This significantly improves the system's response speed and robustness, making it particularly suitable for scenarios with limited computing resources or extremely high requirements for control response time. Example

[0071] To achieve the above method, the present invention also provides an online optimization system for sintering waste heat power generation. This system can be a software system deployed on an industrial computer or server, such as... Figure 6 As shown, its structure includes: Data acquisition module 201: responsible for communicating with the factory's DCS / PLC and other underlying control systems 205, as well as various sensors 206 such as infrared thermal imagers, to acquire all the real-time operating data described in the method embodiment.

[0072] Processing module 202: This is the core of the system, integrating and running internally. 1) Firing quantity prediction model, used to perform S2 calculation and output the sensible heat potential of sintered ore.

[0073] 2) The full-process energy efficiency model is used to perform S3 calculations and output the power generation capacity of sintering waste heat.

[0074] 3) Optimize the model solver by using algorithms such as constraint programming to perform optimization calculations for S4 and output the optimal process control parameters.

[0075] Control module 203: Responsible for generating control commands from the optimal parameter combination calculated by the processing module and sending them to the underlying control system 205 such as DCS / PLC for execution.

[0076] Storage module 204: This is an optional module used to store the working condition-parameter database described in the second embodiment of the method, as well as to store historical data and model files, and to interact with the processing module 202.

[0077] The present invention also provides a computer-readable storage medium having a computer program stored thereon. When the program is executed by a processor, it can perform the steps described in any of the above method embodiments. The storage medium can be any type of non-volatile memory, such as a hard disk, a solid-state drive (SSD), a USB flash drive, or an optical disk.

[0078] In summary, by introducing online prediction and optimization technology based on machine vision and data-driven models, this invention can accurately sense fluctuations in upstream sintering waste heat and adjust the operating parameters of the downstream waste heat recovery and power generation system in a feedforward manner, ensuring that the system always operates at its highest efficiency point under the current operating conditions, thus demonstrating significant energy-saving and consumption-reducing benefits and application value.

Claims

1. A method for online optimization of sintering waste heat power generation, characterized in that, Includes the following steps: Obtain real-time operating data for the entire sintering waste heat power generation process; Based on the real-time operating data, the sensible heat potential of the sinter entering the annular cooler is predicted online. Based on the sensible heat potential of the sinter and the process control parameters to be optimized, the power generation capacity of the sinter waste heat is predicted online. Based on the predicted power generation from sintering waste heat, with the goal of maximizing power generation, the optimal process control parameters corresponding to the current sensible heat potential of the sintered ore are solved and determined. Based on the optimal process control parameters, the sintering waste heat power generation system is optimized and controlled.

2. The method according to claim 1, characterized in that, The step of predicting the sensible heat potential of sinter entering the annular cooler online includes: Real-time acquisition of the operating parameters of the sintering system, wherein the operating parameters of the sintering system include at least the image information of the red fire layer of the tail section acquired by the infrared thermal imaging device; Based on the operating parameters of the sintering system, the sintering amount of the sintered ore is calculated in real time using the sintering amount prediction model. The sensible heat potential of the sinter is calculated based on the sintering amount and sintering temperature.

3. The method according to claim 2, characterized in that, The firing amount prediction model is a gradient boosting decision tree model; The input features of the model include: sintering fabric thickness, sintering trolley width, sintering machine speed, and the maximum area of ​​the red-fired layer at the tail section calculated based on the image information of the red-fired layer at the tail section. The output of the model is the sintering amount of the sintered ore.

4. The method according to claim 3, characterized in that, The calculation steps for the maximum area of ​​the red-hot layer at the tail section include: Continuously acquire multiple frames of infrared thermal imaging images during the material turning process at the tail trolley of the sintering machine; Image recognition is performed on each frame of the image to calculate the outline area of ​​the red flame layer; The maximum contour area calculated in all frames is selected as the maximum area of ​​the red flame layer of the tail section.

5. The method according to claim 1, characterized in that, The step of predicting the power generation from sintering waste heat online is achieved through a full-process energy efficiency model; The input features of the full-process energy efficiency model include: the sensible heat potential of the sinter, the operating parameters of the annular cooler, the operating parameters of the circulating air system, the operating parameters of the waste heat boiler, and the operating parameters of the steam system. The output of the full-process energy efficiency model is the power generated by the waste heat from sintering.

6. The method according to claim 1 or 5, characterized in that, The steps for solving and determining the optimal process control parameters include: Establish an optimization model with the objective function of maximizing power generation; The process control parameters to be optimized are used as decision variables, and the safe operating range of the equipment for the decision variables is used as a constraint. The optimal process control parameters are obtained by solving the optimization model using a constraint programming algorithm. The process control parameters to be optimized include at least: the speed of the annular cooler, the frequency of the circulating fan, the medium-pressure steam pressure, and the low-pressure steam pressure.

7. The method according to claim 1, characterized in that, The method further includes: The sensible heat potential of sintered ore is divided into multiple preset ranges; For each preset interval, a set of optimal process control parameters are pre-calculated and stored to form a working condition-parameter database; The specific steps for optimizing the control of the sintering waste heat power generation system are as follows: based on the preset range to which the sensible heat potential of the sintered ore belongs, the corresponding optimal process control parameters are queried from the operating condition-parameter database and sent to the control system for execution.

8. An online optimization system for sintering waste heat power generation, characterized in that, include: The data acquisition module is used to acquire real-time operating data of the entire sintering waste heat power generation process; The processing module, connected to the data acquisition module, is used for: Based on the real-time operating data, the sensible heat potential of the sinter entering the annular cooler is predicted online. Based on the sensible heat potential of the sinter and the process control parameters to be optimized, the power generation capacity of the sinter waste heat is predicted online. Based on the predicted power generation from sintering waste heat, with the goal of maximizing power generation, the optimal process control parameters corresponding to the current sensible heat potential of the sintered ore are solved and determined. The control module, connected to the processing module, is used to generate control commands based on the optimal process control parameters to optimize the control of the sintering waste heat power generation system.

9. The system according to claim 8, characterized in that, The processing module is used to predict the sensible heat potential of sinter entering the annular cooler online. This is achieved by running a sintering rate prediction model. Specifically, the processing module is used for: The operating parameters of the sintering system are obtained from the data acquisition module. The operating parameters of the sintering system include at least the image information of the red fire layer of the tail section obtained by the infrared thermal imaging device. Based on the operating parameters of the sintering system, the sintering amount of the sintered ore is calculated in real time using the sintering amount prediction model. The sensible heat potential of the sinter is calculated based on the sintering amount and sintering temperature.

10. The system according to claim 9, characterized in that, The firing amount prediction model is a gradient boosting decision tree model; The input features of the model include: sintering fabric thickness, sintering trolley width, sintering machine speed, and the maximum area of ​​the red-fired layer at the tail section calculated based on the image information of the red-fired layer at the tail section. The output of the model is the sintering amount of the sintered ore.

11. The system according to claim 10, characterized in that, The processing module is also configured to perform the following steps to calculate the maximum area of ​​the red-fire layer on the tail section: Continuously acquire multiple frames of infrared thermal imaging images during the material turning process at the tail trolley of the sintering machine; Image recognition is performed on each frame of the image to calculate the outline area of ​​the red flame layer; The maximum contour area calculated in all frames is selected as the maximum area of ​​the red flame layer of the tail section.

12. The system according to claim 8, characterized in that, The processing module's function of predicting the power generation from sintering waste heat online is achieved by running a full-process energy efficiency model; The input features of the full-process energy efficiency model include: the sensible heat potential of the sinter, the operating parameters of the annular cooler, the operating parameters of the circulating air system, the operating parameters of the waste heat boiler, and the operating parameters of the steam system. The output of the full-process energy efficiency model is the power generated by the waste heat from sintering.

13. The system according to claim 8 or 12, characterized in that, The processing module, which is used to solve for and determine the optimal process control parameters, achieves this by running an optimization model. Specifically, the processing module is used for: Establish an optimization model with the objective function of maximizing power generation; The process control parameters to be optimized are used as decision variables, and the safe operating range of the equipment for the decision variables is used as a constraint. The optimal process control parameters are obtained by solving the optimization model using a constraint programming algorithm. The process control parameters to be optimized include at least: the speed of the annular cooler, the frequency of the circulating fan, the medium-pressure steam pressure, and the low-pressure steam pressure.

14. The system according to claim 8, characterized in that, The system also includes a storage module for storing a working condition-parameter database; The operating condition-parameter database stores multiple preset sensible heat potential ranges for sintered ore, as well as the optimal process control parameters corresponding to each range. The control module is specifically used to: obtain the current sensible heat potential of the sinter from the processing module, query the operating condition-parameter database based on the potential, obtain the corresponding optimal process control parameters, and generate control instructions accordingly.

15. A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method as described in any one of claims 1 to 7.

Images