Intelligent control method, system, equipment and medium for steam production of waste incineration waste heat boiler

Abstract

The application relates to the technical field of solid waste incineration treatment and intelligent incineration control, and discloses an intelligent control method, system, equipment and medium for steam production of a waste incineration waste heat boiler. The method comprises the following steps: extracting typical stable combustion conditions of an incinerator to construct a typical condition set; performing three-dimensional numerical simulation on each typical condition, extracting thermal characteristic parameters at a preset reference cross section in a flue to construct a simulation data set; training a prediction model with the typical condition parameters as input and the thermal characteristic parameters as output; collecting current condition parameters, inputting the prediction model to obtain real-time thermal characteristic parameters and determining total heat release power of combustion; and adjusting the current condition parameters to minimize the deviation between the total heat release power and a target heat release power. The application can perceive the change trend of main steam flow in advance and perform advanced regulation and control, thereby effectively improving the stability of the main steam flow.

Description

Technical Field

[0001] This application relates to the field of solid waste incineration treatment and intelligent incineration control technology, specifically to an intelligent control method, system, equipment and medium for the steam output of a waste incineration waste heat boiler. Background Technology

[0002] Waste-to-energy incineration is one of the main methods for treating municipal solid waste. The stability of the main steam flow rate in the waste heat boiler directly affects the power generation efficiency, equipment lifespan, and grid-connected power quality of the steam turbine generator. However, due to the inherent characteristics of the waste entering the furnace, such as its complex composition, uneven calorific value, and discontinuous feeding, the heat load inside the incinerator changes drastically, resulting in frequent and significant fluctuations in the main steam flow rate of the waste heat boiler. This has been a long-standing technical challenge in the operation of waste-to-energy incineration systems.

[0003] Currently, a common approach to the stable control of main steam flow is to control the combustion ratio and air volume ratio entering the incinerator through negative feedback based on the difference between the main steam flow and the process setpoint.

[0004] However, the above-mentioned existing technical solutions still have the following shortcomings: First, relying on the feedback of the main steam flow for regulation is problematic because the heat generated by waste incineration needs to go through a complex heat exchange process to be converted into steam, and the incinerator and waste heat boiler system itself have a large thermal inertia. As a result, the feedback control has a significant time lag problem and cannot detect the changing trend of the main steam flow in advance.

[0005] Second, this scheme only begins to regulate after the steam flow rate changes, which is a passive response method and cannot fundamentally suppress fluctuations in the main steam flow rate, thus limiting the quality of control.

[0006] Third, the scheme does not take into account the impact of equipment aging, wear, coking and other changes on control parameters as the incinerator and waste heat boiler operate for a long time, which leads to the control effect gradually deteriorating as the operating time increases.

[0007] Therefore, there is an urgent need for an intelligent control scheme that can detect changes in the main steam flow rate in advance, achieve proactive regulation, and be able to adaptively update, so as to effectively improve the stability of the main steam flow rate. Summary of the Invention

[0008] To address the aforementioned issues, this application provides an intelligent control method, system, equipment, and medium for steam production in a waste incineration waste heat boiler. By extracting typical stable combustion conditions from historical operating parameters of the incinerator and constructing a typical operating condition set, and combining three-dimensional numerical simulation and machine learning, a predictive model is established that can predict the thermal characteristic parameters within the incinerator flue in real time from operating condition parameters. This model then calculates the total heat release power of the waste incineration process in real time and uses this total heat release power as the controlled variable for advance feedback regulation, thereby achieving early and stable control of the main steam flow.

[0009] The technical solution adopted in this application is as follows: Firstly, this application provides an intelligent control method for the steam output of a waste incineration waste heat boiler, comprising: Obtain historical operating parameters of the incinerator, extract typical stable combustion parameters from the historical operating parameters, and construct a typical operating condition set based on the typical stable combustion parameters; Three-dimensional numerical simulations of gas-solid coupled combustion in the incinerator were performed for each operating parameter in the typical operating condition set to generate three-dimensional numerical simulation results. The thermal characteristic parameters at the preset reference section in the flue of the incinerator were extracted from the three-dimensional numerical simulation results to construct a simulation dataset. Using the operating parameters from a typical set of operating conditions as input and the thermodynamic characteristic parameters corresponding to the operating parameters in the simulation dataset as output, machine learning training is performed to build a prediction model. The current operating parameters of the incinerator are collected, and the current operating parameters are input into the prediction model to obtain real-time thermodynamic characteristic parameters. The total heat release power of combustion is determined based on the real-time thermodynamic characteristic parameters. The target heat release power is determined based on the target main steam flow rate, and the current operating parameters are adjusted with the minimum deviation between the total combustion heat release power and the target heat release power as the target.

[0010] Secondly, this application also provides an intelligent control system for the steam output of a waste incineration waste heat boiler, comprising: The typical operating condition extraction unit is used to obtain historical operating condition parameters of the incinerator, extract typical stable combustion operating condition parameters from the historical operating condition parameters, and construct a typical operating condition set based on the typical stable combustion operating condition parameters. The simulation dataset construction unit is used to perform three-dimensional numerical simulation of gas-solid coupled combustion in the incinerator for each working parameter in the typical working condition set, generate three-dimensional numerical simulation results, extract the thermodynamic characteristic parameters at the preset reference section in the flue of the incinerator from the three-dimensional numerical simulation results, and construct the simulation dataset. The prediction model building unit is used to train a prediction model by taking the operating parameters in the typical operating condition set as input and the thermodynamic feature parameters in the simulation dataset corresponding to the operating condition parameters as output. The real-time computing unit is used to collect the current operating parameters of the incinerator, input the current operating parameters into the prediction model to obtain real-time thermodynamic characteristic parameters, and determine the total heat release power of combustion based on the real-time thermodynamic characteristic parameters. The operating condition control unit is used to determine the corresponding target heat release power based on the target main steam flow rate, and to adjust the operating condition parameters at the current moment with the goal of minimizing the deviation between the total combustion heat release power and the target heat release power.

[0011] Thirdly, this application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described intelligent control method for the steam production of a waste incineration waste heat boiler.

[0012] Fourthly, this application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the above-described intelligent control method for steam production of a waste incineration waste heat boiler.

[0013] The above-mentioned technical solution adopted in this application can achieve the following beneficial effects: This application constructs a set of operating conditions by extracting typical stable combustion conditions of the incinerator, and establishes a predictive model by combining three-dimensional numerical simulation and machine learning training. This model can directly and in real-time predict the thermal characteristic parameters within the incinerator flue from the operating condition parameters, thereby calculating the total heat release power of the incinerator combustion. Since the changing trend of the total heat release power can reflect the changing trend of the main steam flow in advance, using the total heat release power as the controlled variable for proactive feedback regulation can complete the control before the main steam flow actually changes, effectively overcoming the time lag problem of traditional feedback control and significantly improving the stability of the main steam flow.

[0014] This application, by selecting operating conditions that meet the fluctuation conditions from a massive amount of historical operating condition parameters and performing equal interval averaging, can cover the entire operating parameter range of the incinerator with fewer typical operating conditions, significantly reducing the computational resource consumption and time cost of subsequent three-dimensional numerical simulation, while avoiding errors caused by inaccurate measurement of single operating condition parameters.

[0015] This application performs three-dimensional numerical simulations of gas-solid coupled combustion in an incinerator for each operating parameter in a typical operating condition set, and adjusts the simulation process parameters to keep the simulation error within a preset range, thus ensuring the accuracy of the simulation dataset and providing a reliable data foundation for training the prediction model.

[0016] This application employs machine learning algorithms to train a prediction model, enabling the application of three-dimensional numerical simulation results to online real-time scenarios. This achieves millisecond-level rapid mapping from operating parameters to thermodynamic characteristic parameters, providing technical support for real-time calculation of total combustion heat release power and closed-loop control.

[0017] This application calculates the total heat release power of combustion by pre-selecting a preset reference section in the flue where the gas phase components are completely burned, and extracting the thermal power of the grid on the front wall of the section and the thermal parameters of the grid. This not only ensures the physical accuracy of the calculation, but also provides convenience for subsequent comparison and verification with on-site measurement data.

[0018] This application aims to minimize the deviation between the total heat release power and the target heat release power, and uses the range of values ​​of each operating parameter in a typical operating condition set as a constraint. It adopts an iterative optimization method based on a prediction model to adjust the operating parameters, which can quickly adjust the total heat release power to near the target value while ensuring that the operating parameters are within the historical safe operating range.

[0019] This application effectively overcomes the impact of equipment aging, pipe wall wear, coking and other state changes on control accuracy during the long-term operation of incinerators and waste heat boilers on control accuracy by monitoring the deviation between the predicted flue gas temperature output by the prediction model and the actual measured temperature in real time during the operation of the incinerator, and automatically triggering three-dimensional numerical simulation correction and incremental update of the prediction model when the deviation exceeds the warning threshold, thus realizing adaptive continuous optimization of the control system. Attached Figure Description

[0020] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 A schematic diagram of the waste incinerator and waste heat boiler system is shown. Figure 2 A flowchart illustrating an intelligent control method for steam production of a waste incineration waste heat boiler according to an embodiment of this application is shown. Figure 3 A schematic diagram of the structure of an intelligent control system for the steam output of a waste incineration waste heat boiler according to an embodiment of this application is shown; Figure 4 A schematic diagram of the structure of an electronic device according to an embodiment of this application is shown. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0022] Figure 1 A schematic diagram of the waste incinerator and waste heat boiler system is shown. (Refer to...) Figure 1 As shown, after the waste is burned above the grate in the incinerator, part of the heat generated is absorbed by the furnace wall and transferred to the steam-water system. The other part, along with the high-temperature flue gas, enters the waste heat boiler through the incinerator flue. In the waste heat boiler, after passing through the evaporator tube bundle, various superheaters, and the economizer, the heat is transferred to the water, causing the water to absorb the heat and generate steam. Due to the relatively long heat transfer path and the significant thermal inertia of the incinerator and waste heat boiler systems, the heat generated during waste incineration takes a certain amount of time to be completely converted into steam. Therefore, the changing trend of the total heat release during waste combustion in the incinerator can reflect the changing trend of the main steam flow rate in the waste heat boiler in advance.

[0023] However, limited by the detection capabilities of the equipment, the sensors can only acquire parameters such as temperature, pressure, and flow rate at a limited number of measuring points, and cannot directly measure the total heat release power of the incinerator. If regulation is based solely on feedback from the main steam flow, it can only respond passively after the actual steam flow changes, exhibiting significant lag and making it difficult to achieve proactive, precise, and stable control.

[0024] To address the aforementioned issues, the core control approach of this application is as follows: The temperature, pressure, velocity, composition, and density fields of the incinerator furnace and flue are reconstructed in real-time using numerical simulation and machine learning techniques, forming a full three-dimensional mesh. Based on this, the total heat release power of the incinerator is calculated in real-time, and feedback regulation is performed using the total heat release power as the controlled variable. Since changes in the total heat release power can reflect changes in steam flow rate in advance, by stabilizing the total heat release power near the target value, front-end regulation can be completed before actual changes in the main steam flow rate occur, thereby suppressing fluctuations in steam flow rate and achieving proactive and stable control of the main steam flow rate.

[0025] To achieve the above objectives, Figure 2 A flowchart illustrating an intelligent control method for steam production in a waste incineration waste heat boiler according to an embodiment of this application is shown. (Refer to...) Figure 2 As shown, this embodiment provides an intelligent control method for the steam output of a waste incineration waste heat boiler, including steps S210 to S250.

[0026] Step S210: Obtain historical operating parameters of the incinerator, extract typical stable combustion parameters from the historical operating parameters, and construct a typical operating condition set based on the typical stable combustion parameters.

[0027] During the long-term operation of incinerators, distributed control systems accumulate massive amounts of historical operating parameters. However, the complex composition, uneven calorific value, and discontinuous feeding of waste fed into incinerators result in a large number of transient disturbances in the historical data. Directly using all historical operating parameters for subsequent processing would not only significantly increase computational resource consumption but also affect the validity of the analysis results due to inconsistent data quality. Therefore, it is first necessary to select representative typical stable combustion conditions from the massive historical operating parameters to construct a typical operating condition set, providing a reliable data foundation for subsequent three-dimensional numerical simulations and machine learning model training.

[0028] In some optional implementations, step S210 involves acquiring historical operating parameters of the incinerator, extracting typical stable combustion parameters from the historical operating parameters, and constructing a typical operating condition set based on the typical stable combustion parameters. This includes: acquiring historical operating parameters of the incinerator within a preset historical time period at preset time intervals to form a historical operating parameter set; removing damaged operating parameters from the historical operating parameter set and removing operating parameters whose furnace temperature is lower than a preset temperature threshold to form a candidate operating parameter set; and selecting operating parameters that meet fluctuation conditions from the candidate operating parameter set to form a candidate operating parameter set. The fluctuation conditions include: the rate of change of flue gas temperature is less than a first threshold within a first time window; the rate of change of flue gas flow rate is less than a second threshold within a second time window; the deviation between the average furnace temperature and the instantaneous furnace temperature is less than a third threshold within a third time window; and the deviation between the average flue gas flow rate and the instantaneous flue gas flow rate is less than a fourth threshold within a fourth time window. The normal furnace temperature range of the incinerator is divided into a first preset number of temperature intervals. Operating parameters whose temperature deviation from each interval's dividing point is less than a fifth threshold are selected from the candidate operating parameter set and averaged to obtain representative temperature operating parameters for each temperature dividing point. The airflow range of each grate section of the incinerator is divided into a second preset number of flow intervals. Operating parameters whose flow deviation from each flow interval's dividing point is less than a sixth threshold are selected from the candidate operating parameter set and averaged to obtain representative flow parameters for each flow interval's dividing point. The operating speed range of each grate section of the incinerator is divided into a third preset number of speed intervals. Operating parameters whose speed deviation from each speed interval's dividing point is less than a seventh threshold are selected from the candidate operating parameter set and averaged to obtain representative speed operating parameters for each speed dividing point. The representative temperature operating parameters, representative flow parameters, and representative speed parameters are then summarized to form a typical operating condition set.

[0029] First, obtain the historical operating parameters of the incinerator.

[0030] Specifically, historical operating parameters of the incinerator over the past three years can be exported from the distributed control system's database every minute, forming a historical operating parameter set. The database of the distributed control system also stores historical measurement values ​​of key measuring points corresponding to these historical operating parameters, which can be retrieved together for screening typical stable combustion operating parameters. Operating parameters may include, but are not limited to: grate operating speed of each section, air flow rate of each grate section, primary air volume of each grate section, secondary air volume, waste feed rate, furnace multi-point temperature, flue gas temperature of each flue, flue gas flow rate, furnace negative pressure, flue gas oxygen content, lower heating value of the waste fed into the furnace, waste moisture content, main steam flow rate, and induced draft fan opening.

[0031] Secondly, candidate operating parameters are selected from historical operating parameters.

[0032] Because the operation of an incinerator involves issues such as shutdown, sensor malfunction, and data signal transmission, it is necessary to filter historical operating parameters. Damaged parameters (i.e., data with bad pixels are removed entirely) and parameters with furnace temperatures below a preset temperature threshold (e.g., 900℃) can be removed from the historical parameter set, forming a candidate set of operating parameters. This ensures that the selected candidate operating parameters are all within the normal operating temperature range of the incinerator furnace.

[0033] Then, candidate operating parameters are selected from the candidate operating parameters.

[0034] From the set of candidate operating condition parameters, those that meet the fluctuation conditions are selected to form a candidate operating condition parameter set. The fluctuation conditions include the following four aspects: First, the rate of change of flue gas temperature is less than the first threshold within the first time window.

[0035] Specifically, the first time window can be half an hour, and the first threshold can be 1%. This means selecting operating parameters where the flue gas temperature change rate is less than 1% within half an hour. Assume the historical flue gas temperatures measured at key measuring points within half an hour are T1, T2, T3, T4, T5, T6, and T7, and the flue gas temperature change rate within half an hour... The calculation formula is: , formula (1); in, This indicates taking the maximum value. This indicates taking the minimum value, average. This indicates taking the average value.

[0036] Second, the rate of change of flue gas flow rate is less than the second threshold within the second time window.

[0037] Specifically, the first time window can be half an hour, and the second threshold can be 5%. This means selecting operating parameters where the flue gas flow rate change rate is less than 5% within half an hour. Assume the historical measured flue gas flow rates at key measuring points within half an hour are Q1, Q2, Q3, Q4, Q5, Q6, and Q7, and the flue gas flow rate change rate within half an hour... The calculation formula is: , formula (2).

[0038] Third, within the third time window, the deviation between the average furnace temperature and the instantaneous furnace temperature is less than the third threshold.

[0039] Specifically, the third time window can be half an hour before and after the time of a certain operating condition parameter, and the third threshold can be 1%. That is, if the difference between the instantaneous value of the historical measured furnace temperature at the key measuring point of a certain operating condition parameter and the average value of the historical measured furnace temperature within half an hour before and after the time of that operating condition parameter is less than 1%, then that operating condition parameter is retained. Assuming the instantaneous value of the historical measured furnace temperature at the key measuring point of a certain operating condition parameter is t4, and the instantaneous values ​​of the historical measured furnace temperature within half an hour before and after the time of that operating condition parameter are t1, t2, t3, t4, t5, t6, and t7 respectively, then the selection criteria are: ;Formula (3).

[0040] Fourth, the deviation between the average flue gas flow rate and the instantaneous flue gas flow rate within the fourth time window is less than the fourth threshold.

[0041] Specifically, the fourth time window can be half an hour before and after the time of a certain operating condition parameter, and the fourth threshold can be 5%. That is, if the difference between the instantaneous value of the historical measured flue gas flow at the key measuring point of a certain operating condition parameter and the average value of the historical measured flue gas flow within half an hour before and after the time of that operating condition parameter is less than 5%, then that operating condition parameter is retained. Assuming the instantaneous value of the historical measured flue gas flow at the key measuring point of a certain operating condition parameter is Q4, and the instantaneous values ​​of the historical measured flue gas flow within half an hour before and after the time of that operating condition parameter are Q1, Q2, Q3, Q4, Q5, Q6, and Q7 respectively, then the selection criteria are: ;Formula (4).

[0042] Next, typical stable combustion parameters are selected from the candidate operating parameters.

[0043] After the above screening, the number of operating parameters that meet the fluctuation conditions can reach several thousand or even tens of thousands. If all of them are used for three-dimensional numerical simulation, the computational resources and computation time consumption will be enormous. Therefore, it is necessary to further organize the candidate operating parameter set so that the number of operating parameters participating in the three-dimensional numerical simulation can be reduced while still covering a wider range of parameters.

[0044] The normal furnace temperature range of the incinerator is divided into a first preset number (e.g., 200) of temperature intervals. From the candidate operating parameters, those whose temperature deviation from each temperature interval's dividing point (201 dividing points) is less than a fifth threshold (e.g., within 0.5℃) are selected and averaged to obtain the representative operating parameters for each temperature dividing point. This reduces the number of operating parameters, broadens the coverage of operating parameters participating in the three-dimensional numerical simulation, and avoids measurement inaccuracies associated with single operating parameters through averaging.

[0045] For example: Suppose the furnace temperature range of a certain incinerator during normal operation is 900℃~1200℃. This range is now divided into 200 equal temperature intervals, each interval being 1.5℃ wide. The furnace temperature at the kth division point is: Assuming the 100th division point is 1050℃, we select operating parameters from the candidate operating parameters that include historical furnace temperatures at all key measuring points within the range of 1050±0.5℃ (i.e., 1049.5℃~1050.5℃). Assuming five operating parameters are selected, we average all characteristics of these five parameters to obtain the representative operating parameters for the 100th division point. By performing the above selection and averaging operations for each temperature interval division point, we obtain 201 representative operating parameters covering the entire temperature range. These operating parameters eliminate random measurement errors at single moments and ensure that the operating parameters participating in the simulation are uniformly distributed along the temperature dimension, providing a concise and representative input set for subsequent three-dimensional numerical simulations.

[0046] Similarly, the airflow range and operating speed range of each grate section (drying section, combustion section, and burnout section) need to be equally divided, screened, and averaged.

[0047] The air flow range of each grate section (drying section, combustion section, and burnout section) of the incinerator is divided into a second preset number (e.g., 100 equal sections per section) of flow intervals. From the candidate operating condition parameters, the operating condition parameters whose flow deviation from the equal division point of each flow interval is less than the sixth threshold (e.g., within 5%) are selected and averaged to obtain the representative operating condition parameters of the flow corresponding to each flow division point.

[0048] The operating speed range of each grate section (drying section, combustion section, and burnout section) of the incinerator is divided into three preset speed intervals (e.g., 100 intervals per section). From the candidate operating parameters, the operating parameters whose speed deviation from the dividing point of each speed interval is less than the seventh threshold (e.g., within 5%) are selected and averaged to obtain the representative operating parameters of the speed corresponding to each dividing point.

[0049] Finally, the representative operating parameters for each temperature, each flow rate, and each speed are summarized to form a typical operating condition set.

[0050] Step S220: Perform three-dimensional numerical simulation of gas-solid coupled combustion in the incinerator for each working parameter in the typical working condition set, generate three-dimensional numerical simulation results, extract the thermal characteristic parameters at the preset reference section in the flue of the incinerator from the three-dimensional numerical simulation results, and construct the simulation dataset.

[0051] After extracting typical operating conditions, relying solely on discrete data collected from limited key measurement points on-site cannot fully characterize the thermal distribution throughout the incinerator furnace and flue, making it difficult to accurately calculate the total heat release power of the entire waste incineration process. Therefore, this application utilizes CFD three-dimensional numerical simulation technology to conduct full-furnace gas-solid coupled combustion numerical simulations. For each typical operating condition, it recreates the complete physicochemical processes of solid-phase drying, pyrolysis, and combustion of waste in the grate, as well as gas-phase combustion and heat transfer in the flue. It outputs refined field data such as full-domain grid-level flue gas temperature, density, specific heat, velocity, and thermal power. Simultaneously, a pre-selected reference section for complete gas-phase combustion in the flue is used as a unified reference benchmark. All thermodynamic characteristic parameters associated with this section are extracted in batches from the simulation results and stored in a standardized manner to form a combustion simulation dataset that can support machine learning model training.

[0052] In some optional implementations, step S220 involves performing three-dimensional numerical simulations of gas-solid coupled combustion in the incinerator for each operating parameter in the typical operating condition set, generating three-dimensional numerical simulation results, and extracting the thermodynamic characteristic parameters at a preset reference section in the incinerator flue from the three-dimensional numerical simulation results to construct a simulation dataset. This includes: performing three-dimensional numerical simulations for each operating parameter in the typical operating condition set using a coupled solid-phase combustion simulation and gas-phase combustion simulation method to generate initial simulation results; wherein the three-dimensional numerical simulations of each operating parameter use the same mesh division; and calculating the simulation results of key measurement points in the initial simulation results. The simulation process parameters are adjusted based on the true error until the simulation error at all key measuring points is within the preset error threshold, thus obtaining the three-dimensional numerical simulation results. Thermodynamic characteristic parameters at the preset reference cross-section within the incinerator flue are extracted from the three-dimensional numerical simulation results. These thermodynamic characteristic parameters include: the thermal power of each wall grid in front of the preset reference cross-section, the flue gas temperature of each grid on the preset reference cross-section, the flue gas density of each grid on the preset reference cross-section, the specific heat of the flue gas of each grid on the preset reference cross-section, and the flue gas velocity perpendicular to the preset reference cross-section flowing through each grid on the preset reference cross-section. A combustion simulation dataset is constructed based on these thermodynamic characteristic parameters.

[0053] This application employs a coupled approach of solid-phase combustion simulation and gas-phase combustion simulation to perform three-dimensional numerical simulations on each operating parameter in a typical operating condition set.

[0054] In one specific implementation, a coupled approach of FLIC solid-phase combustion simulation and Fluent gas-phase combustion simulation is employed. FLIC is used for solid-phase combustion simulation to obtain a cloud map of the waste temperature distribution above the grate, as well as data on the changes in gas phase temperature, composition, and flow velocity above the waste bed along the grate's movement direction. The temperature, composition, and flow velocity above the waste bed calculated by the FLIC simulation are imported into the Fluent model as boundary conditions for gas phase combustion simulation. The same mesh generation is used for the three-dimensional numerical simulation of each operating parameter to ensure the comparability of simulation results between different operating parameters.

[0055] A three-dimensional numerical simulation is performed to obtain initial simulation results. The flue gas temperature at the corresponding key measurement points in the initial simulation results is compared with the historical measured flue gas temperature corresponding to the operating parameters, and the simulation error is calculated. If the simulation error is within 5% of the preset error threshold, the simulation is stopped, and the three-dimensional numerical simulation result is obtained. If the simulation error is outside 5% of the preset error threshold (including 5%), the simulation process parameters (such as turbulence model parameters, radiation model parameters, chemical reaction kinetic parameters, etc.) are adjusted, and the simulation process is repeated until the simulation error is within the preset error threshold, and the three-dimensional numerical simulation result is obtained.

[0056] Thermodynamic characteristic parameters at preset reference sections within the incinerator flue are extracted from the three-dimensional numerical simulation results. The selection of the preset reference section must ensure that, for any operating condition parameters, the gaseous components (CH4, CO, H2, etc.) within the furnace have been completely combusted before this preset reference section. Preferably, the preset reference section is selected as the flue section location corresponding to key measuring points in the on-site distributed control system, facilitating subsequent comparison and verification between the predicted values ​​output by the prediction model and the actual measured values ​​at the key measuring points, thereby achieving real-time monitoring and online correction of the model's prediction accuracy. For example, the preset reference section could be a section within the third flue.

[0057] Thermodynamic characteristic parameters include the following five categories: (1) Thermal power of each wall mesh in front of the preset reference section ;in, Indicates the face grid number on the front wall of the preset reference section; (2) Flue gas temperature of each grid on the preset reference section ;in, Indicates the grid number on the preset reference section; (3) Preset the flue gas density of each grid on the reference section ; (4) Specific heat of flue gas in each grid on the preset reference section ; (5) Flue gas velocity perpendicular to the preset reference section at each grid on the preset reference section. .

[0058] A combustion simulation dataset was constructed based on the aforementioned thermodynamic characteristic parameters. After completing the three-dimensional numerical simulation of all operating parameters for the typical operating condition set, approximately 800 sets of thermodynamic characteristic parameters were generated, forming the simulation dataset.

[0059] Step S230: Using the operating parameters from the typical operating condition set as input and the thermodynamic characteristic parameters corresponding to the operating parameters in the simulation dataset as output, machine learning training is performed to construct a prediction model.

[0060] After completing the extraction of typical operating conditions and the construction of simulation datasets, although high-precision simulation data under several typical operating condition parameters have been obtained, the long calculation time of a single three-dimensional numerical simulation makes it unsuitable for direct application to steam production control during incinerator operation. Therefore, it is necessary to establish an alternative model that can respond quickly and calculate thermodynamic characteristic parameters in real time. To this end, a prediction model is trained using typical operating condition parameters as input and corresponding simulated thermodynamic characteristic parameters as output, thereby achieving a rapid mapping from operating condition parameters to thermodynamic characteristic parameters and providing support for subsequent real-time calculation of total combustion heat release power.

[0061] In some optional implementations, step S230, using the operating parameters in the typical operating condition set as input and the thermodynamic characteristic parameters in the simulation dataset corresponding to the operating parameters as output, performs machine learning training to build a prediction model, including: using the operating parameters in the typical operating condition set as input features and the thermodynamic characteristic parameters in the simulation dataset corresponding to the operating parameters as output labels, performing iterative training with a machine learning algorithm, establishing a nonlinear mapping relationship from the operating parameters to the thermodynamic characteristic parameters, and obtaining the prediction model.

[0062] A machine learning model is established to connect operating parameters with thermodynamic characteristic parameters. This model can employ algorithms such as neural networks, support vector machines, and random forests. After training the model, real-time thermodynamic characteristic parameters at a preset reference section can be directly obtained by inputting real-time operating parameters into the prediction model.

[0063] Preferably, the prediction model may include two sub-models, both of which use a multi-output random forest regression model as the machine learning algorithm. The first sub-model is used to establish a nonlinear mapping relationship between the operating parameters and the thermal power of all walls in front of the preset reference section, and the second sub-model is used to establish a nonlinear mapping relationship between the operating parameters and the flue gas temperature, flue gas density, flue gas specific heat, and flue gas velocity of all grids on the preset reference section.

[0064] Multi-output random forests are composed of multiple decision trees, with the number of decision trees denoted as N (which can range from 100 to 500, for example, N=200). The difference between multi-output random forests and single-output random forests is that the leaf nodes of each tree do not store a scalar value, but a multi-dimensional output vector.

[0065] The specific training process for the first sub-model is as follows: The thermal power of all m grids on the front wall of the preset reference section is concatenated into a multi-dimensional output vector with dimension m. For each working parameter input feature X in the typical working condition set, there exists an m-dimensional output label Y in the corresponding simulation dataset, forming a set of training samples (X,Y). All training samples constitute the training set.

[0066] For each decision tree, a training subset with replacement, equal to the number of samples in the training set, is randomly selected from the training set using bootstrap sampling. Starting from the root node, the following operations are performed on the current node: Randomly select from all d input features (operating condition parameter dimensions) 1 feature as candidate feature The value is d / 3 rounded down.

[0067] For each candidate feature, obtain the value of all samples of the current node under that candidate feature, and use these values ​​after removing duplicates as the candidate splitting threshold. Try each candidate splitting threshold in turn, and divide the samples of the current node into left and right child nodes according to whether the value of the candidate feature is less than or equal to the candidate splitting threshold.

[0068] Calculate the reduction in multi-output mean squared error for each candidate partition. The multi-output mean squared error for the current node. The calculation formula is: , formula (5); in, This represents the total number of samples in the current node. Indicates the first The multidimensional output vector of each sample It represents the average value of the multidimensional output vector of all samples at the current node (calculated dimension by dimension). This represents the square of the L2 norm of a vector.

[0069] Reduction in mean square error of multi-output after partitioning for: , formula (6); in, This indicates the number of samples in the left child node. This indicates the number of samples in the right child node. This represents the mean square error of the multiple outputs of the left child node. This represents the mean square error of the multiple outputs of the right child node.

[0070] Among all candidate features and all candidate partitioning thresholds, the candidate feature and candidate partitioning threshold that maximize ΔMSE are selected as the optimal partition for the current node.

[0071] Recursively perform the above splitting operation on the left and right child nodes until any of the following stopping conditions are met: (a) the number of samples in the current node is less than the preset minimum number of samples threshold (e.g., 5); (b) the mean square error of the multi-output of the current node is less than the preset error threshold; (c) the maximum depth of the tree reaches the preset maximum depth threshold (e.g., 20 layers).

[0072] The node that reaches the stopping condition becomes a leaf node, and the average of the multidimensional output vectors of all samples of that node is stored as the prediction output of that leaf node.

[0073] After training all N decision trees, the real-time operating parameters are input into each decision tree. The tree structure is traversed from the root node to the leaf node to obtain the multi-dimensional output vector stored in the leaf node of each tree. The arithmetic mean of all N multi-dimensional output vectors is taken as the final prediction result. This prediction result is the real-time thermal power of all m grids on the front wall of the preset reference section.

[0074] The training method for the second sub-model is the same as that for the first sub-model, the only difference being the dimension of the output vector. The output label of the second sub-model is a 4n-dimensional vector composed of the flue gas temperature, flue gas density, flue gas specific heat, and flue gas velocity of all n grids on the preset reference section. The second sub-model is trained independently based on the same typical operating condition set and simulation dataset, following the same process described above. During real-time prediction, the real-time operating condition parameters are input into the second sub-model to obtain the real-time flue gas temperature, real-time flue gas density, real-time flue gas specific heat, and real-time flue gas velocity of each grid on the preset reference section.

[0075] Step S240: Collect the current operating parameters of the incinerator, input the current operating parameters into the prediction model to obtain real-time thermodynamic characteristic parameters, and determine the total heat release power of combustion based on the real-time thermodynamic characteristic parameters.

[0076] After offline training, the prediction model has established a nonlinear mapping relationship between operating parameters and thermodynamic characteristic parameters, enabling it to complete a single prediction within milliseconds. Therefore, the trained prediction model is deployed to an online operating environment. Based on the incinerator's current operating parameters, it quickly predicts the corresponding real-time thermodynamic characteristic parameters, thereby calculating the total heat release power of the incinerator and providing a real-time data foundation for subsequent operating condition control.

[0077] In some optional implementations, step S240 includes: collecting the current operating parameters of the incinerator, inputting the current operating parameters into a prediction model to obtain real-time thermal characteristic parameters, and determining the total heat release power of combustion based on the real-time thermal characteristic parameters, including: collecting the current operating parameters of the incinerator; inputting the current operating parameters into a prediction model, and obtaining, through the prediction model, the real-time thermal power of each wall grid in front of the preset reference section, the real-time flue gas temperature of each grid on the preset reference section, the real-time flue gas density of each grid on the preset reference section, the real-time flue gas specific heat of each grid on the preset reference section, and the real-time flue gas flow perpendicular to the preset reference section through each grid on the preset reference section. Airflow velocity; the total real-time heat power of all wall grids in front of the preset reference section is determined by summing the real-time heat power of each wall grid in front of the preset reference section; for each grid on the preset reference section, the real-time flue gas temperature, real-time flue gas density, real-time flue gas specific heat, real-time flue gas velocity, and cross-sectional area where the grid intersects with the preset reference section are multiplied to obtain the real-time heat power of the grid; the real-time heat power of each grid on the preset reference section is summed to obtain the total real-time heat power flowing through the preset reference section; the total real-time heat power of all wall grids in front of the preset section is added to the total real-time heat power flowing through the preset reference section to obtain the total heat release power of the incinerator.

[0078] During the operation of the incinerator, the current operating parameters are read in real time from the database of the distributed control system.

[0079] The current operating parameters are input into the prediction model, and real-time thermal characteristic parameters are obtained through the prediction model. As shown above, when the prediction model is a multi-output random forest regression model including a first sub-model and a second sub-model, the current operating parameters are input into the first sub-model and the second sub-model respectively. The first sub-model outputs the real-time thermal power of each wall grid in front of the preset reference section, and the second sub-model outputs the real-time flue gas temperature, real-time flue gas density, real-time flue gas specific heat, and real-time flue gas velocity of each grid on the preset reference section.

[0080] The real-time total heat power of all wall meshes in front of the preset reference section is calculated using the following formula (7). : , formula (7); in, This represents the real-time thermal power of each wall mesh in front of the preset reference section. This indicates the total number of all grids on the front wall of the preset reference section.

[0081] The real-time total heat power flowing through the preset reference section is calculated according to the following formula (8). : , formula (8); in, This indicates the real-time flue gas temperature of each grid on the preset reference section. This represents the real-time flue gas density of each grid on the preset reference section. This represents the real-time specific heat of flue gas in each grid on the preset reference section. This represents the real-time flue gas velocity perpendicular to the reference cross-section at each grid point on the preset reference cross-section. This represents the cross-sectional area of ​​the portion where each grid on the preset reference section intersects with the preset reference section. This indicates the total number of all grids on the preset reference section.

[0082] The total real-time heat power of all wall grids in front of the preset cross-section is added to the total real-time heat power flowing through the preset reference cross-section to obtain the total heat release power of the incinerator. : , formula (9).

[0083] Step S250: Determine the corresponding target heat release power based on the target main steam flow rate, and adjust the current operating parameters with the minimum deviation between the total combustion heat release power and the target heat release power as the target.

[0084] Once the real-time total heat release power of the incinerator is obtained, the foundation for closed-loop control is established. As mentioned earlier, the changing trend of the total heat release power during waste incineration can reflect the changing trend of the main steam flow rate in advance. Therefore, if the total heat release power can be stabilized near a certain target value, the main steam flow rate can be indirectly stabilized in advance without waiting for the actual change in steam flow rate before responding passively. To this end, it is necessary to calculate and determine the corresponding target heat release power based on the desired main steam flow rate, and optimize the controllable current operating parameters with the goal of minimizing the deviation between the total heat release power and the target heat release power.

[0085] In some optional implementations, step S250, determining the corresponding target heat release power based on the target main steam flow rate, and adjusting the current operating parameters with the goal of minimizing the deviation between the total combustion heat release power and the target heat release power, includes: determining the target heat release power based on the target main steam flow rate through a preset mapping relationship; using the current operating parameters as the initial point, with the goal of minimizing the deviation between the total combustion heat release power and the target heat release power, and with the interval formed by the minimum and maximum values ​​of each operating parameter in the typical operating condition set as the boundary constraint, iteratively optimizing the operating parameters using a gradient descent method based on a prediction model; wherein, after each iteration, it is determined whether the deviation between the total combustion heat release power and the target heat release power under the candidate operating parameters is within a preset error range; if yes, the current operating parameters are adjusted to the candidate operating parameters; if not, the iteration continues until the deviation between the total combustion heat release power and the target heat release power under the candidate operating parameters is within a preset error range or the maximum number of iterations is reached.

[0086] Based on the expected main steam flow rate of the incinerator waste heat boiler system, the total heat release power of the incinerator is determined through a preset mapping relationship. This mapping relationship can be obtained through thermodynamic calculations or regression analysis of historical data. For example, the mapping relationship is as follows: based on the heat balance model of the waste heat boiler, the target main steam flow rate is converted into the required heat power input value, taking into account the thermal efficiency between the incinerator and the waste heat boiler during the conversion, i.e., target heat release power = target main steam flow rate × steam enthalpy / overall thermal efficiency. This mapping relationship can be pre-stored in the control system.

[0087] Taking the current operating parameters as the initial point, the objective function is to minimize the deviation between the total heat release power of combustion and the target heat release power. The boundary constraint is the interval formed by the minimum and maximum values ​​of each operating parameter in the typical operating condition set. The gradient descent method based on the prediction model is used to iteratively optimize the operating parameters.

[0088] The objective function can be expressed as: , formula (10); in, Indicates operating parameters The total heat release power of combustion, This indicates the target heat release power.

[0089] Since the total heat release power of real-time combustion is calculated through the prediction model and cannot be directly differentiated using analytical expressions, the gradient is approximated using a numerical differentiation method: For each adjustable parameter in the current operating parameters, a small perturbation δ (e.g., δ = 0.1% to 1% of the current value) is applied near its current value. The perturbed operating parameters are then input into the prediction model, and the total heat release power of combustion after the perturbation is calculated. The gradient component of the perturbed parameter is obtained through difference approximation. The gradient components of all parameters together constitute the complete gradient vector at the current operating parameters. Subsequently, the operating parameters are updated according to the standard iterative formula of the gradient descent method.

[0090] After obtaining candidate operating condition parameters in each iteration, these parameters are input into the prediction model to calculate the corresponding candidate total heat release power. The model then determines whether the deviation between the candidate total heat release power and the target heat release power is within a preset error range (e.g., absolute deviation ≤ 3%). If yes, the current operating condition parameters are adjusted to the candidate operating condition parameters; otherwise, iteration continues until the deviation between the total heat release power and the target heat release power under the candidate operating condition parameters is within the preset error range or the maximum number of iterations (e.g., 100) is reached. If the deviation requirement is still not met when the maximum number of iterations is reached, the candidate operating condition parameter with the smallest deviation in the historical iterations is selected as the adjustment result output.

[0091] The finalized candidate operating parameters are sent to the distributed control system, which then executes the corresponding adjustment actions.

[0092] In some optional embodiments, the above method further includes: during the operation of the incinerator, acquiring verification operating condition parameters within a preset time period prior to the current moment, extracting typical stable combustion operating condition parameters from the verification operating condition parameters, and constructing a verification operating condition set based on the typical stable combustion operating condition parameters; inputting each operating condition parameter in the verification operating condition set into a prediction model to obtain the predicted thermodynamic characteristic parameters corresponding to each operating condition parameter, and extracting the predicted flue gas temperature of a specific grid on a preset reference section from each predicted thermodynamic characteristic parameter; acquiring the actual flue gas temperature of each actual measuring point corresponding to the specific grid, calculating the deviation between each predicted flue gas temperature and each actual flue gas temperature; and statistically analyzing the predicted flue gas temperature and the actual flue gas temperature. The number of operating parameters whose deviations exceed a preset deviation threshold is used to trigger a correction process when the number exceeds a preset warning threshold. The correction process includes: performing a three-dimensional numerical simulation of gas-solid coupled combustion in the incinerator for each operating parameter in the verification operating condition set whose deviation exceeds the preset deviation threshold, and generating corresponding three-dimensional numerical simulation results; extracting verification thermodynamic characteristic parameters at a preset reference section in the incinerator flue from the generated three-dimensional numerical simulation results, and constructing a verification simulation dataset based on the verification thermodynamic characteristic parameters; and incrementally updating the prediction model using the operating parameters in the verification operating condition set whose deviations exceed the preset deviation threshold and the verification thermodynamic characteristic parameters in the verification simulation dataset as new samples.

[0093] During the long-term operation of the incinerator, the accuracy of the prediction model may gradually decrease due to factors such as the incinerator's aging and pipe wall wear. If the accuracy of the prediction model is not corrected, it will affect the accuracy of the calculation of the total heat release power of subsequent combustion, thereby weakening the overall system control effect. Therefore, an online correction method for the prediction model is also provided.

[0094] During incinerator operation, verification operating condition parameters for a preset time period prior to the current moment are periodically acquired, either in response to verification trigger commands or periodically. The preset time period for selecting verification operating condition parameters is no earlier than the time the prediction model was deployed to the online operating environment.

[0095] Extract parameters that satisfy typical stable combustion conditions from the validation operating condition parameters, and construct a validation operating condition set based on these parameters. The method for constructing the validation operating condition set can be the same as that for constructing the typical operating condition set, and will not be described again.

[0096] Each operating parameter in the validation set is input into the prediction model to obtain the predicted thermodynamic characteristic parameters corresponding to each operating parameter. The predicted flue gas temperature of a specific grid on a preset reference section is then extracted from the predicted thermodynamic characteristic parameters. Simultaneously, the actual flue gas temperature measured at the actual measuring point corresponding to that specific grid location is obtained through the on-site distributed control system. .in, This indicates the parameter number in the verification test case set.

[0097] Calculate the deviation between the predicted flue gas temperature and the actual flue gas temperature. : %, formula (11).

[0098] This deviation serves as a quantitative indicator to measure the difference between the predicted and actual values.

[0099] statistics The number of operating parameters exceeding a preset deviation threshold will trigger a correction process when it exceeds a preset warning threshold (e.g., 10%). If the number does not exceed the preset warning threshold, the accuracy of the prediction model is considered to still meet the requirements, and the correction process is not triggered.

[0100] Once the calibration process is triggered, the gas-solid combustion process in the incinerator is re-simulated in three dimensions for each operating parameter in the verification operating condition set whose deviation exceeds the preset deviation threshold, generating three-dimensional numerical simulation results corresponding to the operating parameters of the verification operating condition set. The simulation process is the same as described above and will not be repeated here.

[0101] From the regenerated three-dimensional numerical simulation results, the verification thermodynamic characteristic parameters at the preset reference section in the incinerator flue are extracted in exactly the same way as in the previous steps, and a verification simulation dataset is constructed based on the verification thermodynamic characteristic parameters.

[0102] Operating parameters whose deviations exceed a preset deviation threshold in the verification operating condition set, along with the corresponding verification thermodynamic feature parameters from the verification simulation dataset, are added incrementally to the original training samples. Based on the original prediction model, the prediction model is retrained using the updated training samples. This incremental update process retains the existing knowledge covered in the original offline training while incorporating new samples reflecting the latest incinerator and boiler conditions.

[0103] After the prediction model is updated, the new prediction model is used as the model for subsequent real-time predictions. Through the above correction mechanism, during the long-term operation of the incinerator and waste heat boiler, whenever the prediction accuracy decreases, the correction process is automatically triggered to regenerate high-precision simulation data and update the prediction model. This ensures that the accuracy of the prediction model does not decrease significantly due to the long-term operation of the system, thus achieving the self-adaptation and continuous optimization of the control system.

[0104] Figure 3 An intelligent control system for the steam production of a waste incineration waste heat boiler according to one embodiment of this application is shown. (Refer to...) Figure 3 As shown, this embodiment provides an intelligent control system 300 for the steam output of a waste incineration waste heat boiler, comprising: Typical operating condition extraction unit 310 is used to obtain historical operating condition parameters of the incinerator, extract typical stable combustion operating condition parameters from the historical operating condition parameters, and construct a typical operating condition set based on the typical stable combustion operating condition parameters. The simulation dataset construction unit 320 is used to perform three-dimensional numerical simulation of gas-solid coupled combustion in the incinerator for each working parameter in the typical working condition set, generate three-dimensional numerical simulation results, extract the thermal characteristic parameters at the preset reference section in the flue of the incinerator from the three-dimensional numerical simulation results, and construct the simulation dataset. The prediction model building unit 330 is used to train a prediction model by taking the operating parameters in the typical operating condition set as input and the thermodynamic characteristic parameters in the simulation dataset corresponding to the operating parameters as output. The real-time computing unit 340 is used to collect the current operating parameters of the incinerator, input the current operating parameters into the prediction model to obtain real-time thermodynamic characteristic parameters, and determine the total heat release power of combustion based on the real-time thermodynamic characteristic parameters. The operating condition control unit 350 is used to determine the corresponding target heat release power based on the target main steam flow rate, and to adjust the operating condition parameters at the current moment with the minimum deviation between the total combustion heat release power and the target heat release power.

[0105] In some optional embodiments, in the above system, the typical operating condition extraction unit 310 is used to: acquire historical operating condition parameters of the incinerator within a preset historical time period at preset time intervals, forming a historical operating condition parameter set; remove damaged operating condition parameters from the historical operating condition parameter set, and remove operating condition parameters whose furnace temperature is lower than a preset temperature threshold, forming a candidate operating condition parameter set; filter operating condition parameters that meet fluctuation conditions from the candidate operating condition parameter set, forming a candidate operating condition parameter set; wherein, the fluctuation conditions include: the flue gas temperature change rate is less than a first threshold in a first time window; and the flue gas flow rate change rate is less than a second threshold in a second time window; and the deviation between the average furnace temperature and the instantaneous furnace temperature is less than a third threshold in a third time window; and the deviation between the average flue gas flow rate and the instantaneous flue gas flow rate is less than a fourth threshold in a fourth time window; and divide the normal furnace temperature range of the incinerator into a first preset... The following steps are performed to obtain representative temperature parameters for each temperature interval: First, the airflow range of each grate section of the incinerator is divided into a second preset number of flow intervals. Then, the parameters selected from the candidate operating condition parameter set whose temperature deviation from each interval's dividing point is less than a fifth threshold are averaged to obtain representative flow parameters for each flow interval's dividing point. Second, the operating speed range of each grate section of the incinerator is divided into a third preset number of speed intervals. Then, the parameters selected from the candidate operating condition parameter set whose speed deviation from each speed interval's dividing point is less than a seventh threshold are averaged to obtain representative speed parameters for each speed interval's dividing point. Finally, the representative temperature parameters, representative flow parameters, and representative speed parameters are summarized to form a typical operating condition set.

[0106] In some optional implementations, in the above system, the simulation dataset construction unit 320 is used to: perform three-dimensional numerical simulations on each operating parameter in the typical operating condition set using a coupled solid-phase combustion simulation and gas-phase combustion simulation method to generate initial simulation results; wherein, the three-dimensional numerical simulation of each operating parameter uses the same mesh division; calculate the simulation error of key measuring points in the initial simulation results, adjust the simulation process parameters according to the simulation error until the simulation error of all key measuring points is within a preset error threshold, and obtain the three-dimensional numerical simulation results; extract the thermodynamic characteristic parameters at the preset reference section in the incinerator flue from the three-dimensional numerical simulation results; wherein, the thermodynamic characteristic parameters include: the thermal power of each wall mesh in front of the preset reference section, the flue gas temperature of each mesh on the preset reference section, the flue gas density of each mesh on the preset reference section, the specific heat of the flue gas of each mesh on the preset reference section, and the flue gas velocity perpendicular to the preset reference section flowing through each mesh on the preset reference section; and construct a combustion simulation dataset based on the thermodynamic characteristic parameters.

[0107] In some optional implementations, in the above system, the prediction model building unit 330 is used to: use the operating parameters in the typical operating condition set as input features, use the thermodynamic feature parameters in the simulation dataset corresponding to the operating parameters as output labels, use machine learning algorithms for iterative training, establish a nonlinear mapping relationship from the operating parameters to the thermodynamic feature parameters, and obtain a prediction model.

[0108] In some optional embodiments, in the above system, the real-time computing unit 340 is used to: collect the current operating parameters of the incinerator; input the current operating parameters into a prediction model, and obtain, through the prediction model, the real-time thermal power of each wall grid in front of the preset reference section, the real-time flue gas temperature of each grid on the preset reference section, the real-time flue gas density of each grid on the preset reference section, the real-time flue gas specific heat of each grid on the preset reference section, and the real-time flue gas velocity perpendicular to the preset reference section flowing through each grid on the preset reference section; and determine the total real-time thermal power of each wall grid in front of the preset reference section. The real-time total thermal power of all wall grids before the preset cross-section is calculated. For each grid on the preset reference cross-section, the real-time flue gas temperature, real-time flue gas density, real-time flue gas specific heat, real-time flue gas velocity, and cross-sectional area where the grid intersects with the preset reference cross-section are multiplied to obtain the real-time thermal power of the grid. The real-time thermal power of each grid on the preset reference cross-section is summed to obtain the real-time total thermal power flowing through the preset reference cross-section. The real-time total thermal power of all wall grids before the preset cross-section is added to the real-time total thermal power flowing through the preset reference cross-section to obtain the total heat release power of the incinerator.

[0109] In some optional embodiments, in the above system, the operating condition control unit 350 is used to: determine the target heat release power according to the target main steam flow rate through a preset mapping relationship; take the current operating condition parameters as the initial point, take the minimum deviation between the total combustion heat release power and the target heat release power as the objective function, and take the interval formed by the minimum and maximum values ​​of each operating condition parameter in the typical operating condition set as the boundary constraint, and use the gradient descent method based on the prediction model to iteratively optimize the operating condition parameters; wherein, after each iteration, it is determined whether the deviation between the total combustion heat release power and the target heat release power under the candidate operating condition parameters is within a preset error range; if so, the current operating condition parameters are adjusted to the candidate operating condition parameters; if not, the iteration continues until the deviation between the total combustion heat release power and the target heat release power under the candidate operating condition parameters is within a preset error range or the maximum number of iterations is reached.

[0110] In some optional embodiments, the system further includes: an online update unit, configured to, during incinerator operation, acquire verification operating condition parameters within a preset time period prior to the current moment, extract typical stable combustion operating condition parameters from the verification operating condition parameters, construct a verification operating condition set based on the typical stable combustion operating condition parameters; input each operating condition parameter in the verification operating condition set into a prediction model to obtain the predicted thermodynamic characteristic parameters corresponding to each operating condition parameter, and extract the predicted flue gas temperature of a specific grid on a preset reference section from each predicted thermodynamic characteristic parameter; acquire the actual flue gas temperature of each actual measurement point corresponding to the specific grid, calculate the deviation between each predicted flue gas temperature and each actual flue gas temperature; and statistically analyze the difference between the predicted flue gas temperature and the actual flue gas temperature. The number of operating parameters whose flue gas temperature deviation exceeds a preset deviation threshold is used to trigger a correction process when the number exceeds a preset warning threshold. The correction process includes: performing a three-dimensional numerical simulation of gas-solid coupled combustion in the incinerator for each operating parameter in the verification operating condition set whose deviation exceeds the preset deviation threshold, generating corresponding three-dimensional numerical simulation results; extracting verification thermodynamic characteristic parameters at a preset reference section in the incinerator flue from the generated three-dimensional numerical simulation results, and constructing a verification simulation dataset based on the verification thermodynamic characteristic parameters; and incrementally updating the prediction model using the operating parameters in the verification operating condition set whose deviation exceeds the preset deviation threshold and the verification thermodynamic characteristic parameters in the verification simulation dataset as new samples.

[0111] It should be noted that the aforementioned intelligent control system 300 for the steam output of the waste incineration waste heat boiler can realize the aforementioned intelligent control method for the steam output of the waste incineration waste heat boiler, which will not be elaborated further.

[0112] Figure 4 This invention illustrates a schematic diagram of the structure of an electronic device according to an embodiment of the present application. Figure 4 As shown, the electronic device includes a processor, internal memory, a network interface, and a non-volatile storage medium connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile and / or volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The network interface is used for communication with external devices via a network connection. When the computer program is executed by the processor, it implements the functions or steps of an intelligent control method for the steam production of a waste incineration waste heat boiler.

[0113] In one embodiment, the electronic device provided in this application includes a memory and a processor. The memory stores a database and a computer program that can run on the processor. When the processor executes the computer program, it implements the steps of an intelligent control method for the steam production of a waste incineration waste heat boiler.

[0114] The above is as stated in this application. Figure 3 The method for executing the intelligent control system for steam production in a waste incineration waste heat boiler, as disclosed in the illustrated embodiment, can be applied to a processor or implemented by a processor. During implementation, each step of the above method can be completed by integrated logic circuits in the processor's hardware or by instructions in software form. The processor can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. The steps of the method disclosed in the embodiments of this application can be directly embodied as being executed by a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software modules can reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. This storage medium is located in memory, and the processor reads information from the memory and, in conjunction with its hardware, completes the steps of the above method.

[0115] In one embodiment, a computer-readable storage medium is also provided, on which a computer program is stored, the steps of which, when executed by a processor, implements a method for intelligent control of steam production from a waste incineration waste heat boiler.

[0116] It should be noted that the functions or steps that the above-mentioned electronic devices or computer-readable storage media can achieve can be referred to the relevant descriptions in the foregoing method embodiments. To avoid repetition, they will not be described one by one here.

[0117] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), Synchlink, DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM), etc.

[0118] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0119] It should be noted that any AI models, software tools, or components not belonging to this company appearing in the embodiments of this application are merely illustrative examples and do not represent actual use. All user personal information involved in the embodiments of this application has been authorized (with the knowledge and consent) by the relevant parties or has been fully authorized by all parties, and the executing entity may obtain it through various legal and compliant means. The collection, storage, use, processing, transmission, provision, and disclosure of the information, data, and signals involved all comply with relevant laws and regulations and do not violate public order and good morals.

[0120] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method for intelligent control of steam output from a waste incineration waste heat boiler, characterized in that, include: Obtain historical operating parameters of the incinerator, extract typical stable combustion parameters from the historical operating parameters, and construct a typical operating condition set based on the typical stable combustion parameters; Three-dimensional numerical simulations of gas-solid coupled combustion in the incinerator were performed for each operating parameter in the typical operating condition set to generate three-dimensional numerical simulation results. The thermal characteristic parameters at the preset reference section in the flue of the incinerator were extracted from the three-dimensional numerical simulation results to construct a simulation dataset. Using the operating parameters from a typical set of operating conditions as input and the thermodynamic characteristic parameters corresponding to the operating parameters in the simulation dataset as output, machine learning training is performed to build a prediction model. The current operating parameters of the incinerator are collected, and the current operating parameters are input into the prediction model to obtain real-time thermodynamic characteristic parameters. The total heat release power of combustion is determined based on the real-time thermodynamic characteristic parameters. The target heat release power is determined based on the target main steam flow rate, and the current operating parameters are adjusted with the minimum deviation between the total combustion heat release power and the target heat release power as the target.

2. The intelligent control method for steam output of a waste incineration waste heat boiler according to claim 1, characterized in that, The process of acquiring historical operating parameters of the incinerator, extracting typical stable combustion parameters from the historical operating parameters, and constructing a typical operating condition set based on the typical stable combustion parameters includes: Historical operating parameters of the incinerator within a preset historical time period are obtained at preset time intervals to form a set of historical operating parameters. Damaged operating parameters are removed from the historical operating parameter set, and operating parameters with furnace temperatures below the preset temperature threshold are also removed, forming a set of candidate operating parameters. The working condition parameters that meet the fluctuation conditions are selected from the set of candidate working condition parameters to form a set of candidate working condition parameters; The fluctuation conditions include: Within the first time window, the rate of change of flue gas temperature is less than the first threshold. Furthermore, the rate of change of flue gas flow rate is less than the second threshold within the second time window; Furthermore, the deviation between the average furnace temperature and the instantaneous furnace temperature within the third time window is less than the third threshold. Furthermore, the deviation between the average flue gas flow rate and the instantaneous flue gas flow rate within the fourth time window is less than the fourth threshold. The normal furnace temperature range of the incinerator is divided into a first preset number of temperature intervals. The operating parameters with temperature deviations less than the fifth threshold from the candidate operating parameters set are selected and averaged to obtain the temperature representative operating parameters corresponding to each temperature interval. The air flow range of each grate section of the incinerator is divided into a second preset number of flow intervals. The operating parameters with flow deviations less than the sixth threshold from the candidate operating parameters set are selected and averaged to obtain the representative operating parameters of flow corresponding to each flow interval. The operating speed range of each grate section of the incinerator is divided into three preset speed intervals. The operating parameters with speed deviations less than the seventh threshold from the candidate operating parameters set are selected and averaged to obtain the representative operating parameters of speed corresponding to each speed interval. Summarize the representative operating parameters for each temperature, each flow rate, and each speed to form a typical operating condition set.

3. The intelligent control method for steam output of a waste incineration waste heat boiler according to claim 1, characterized in that, The method involves performing three-dimensional numerical simulations of gas-solid coupled combustion within the incinerator for each operating parameter in the typical operating condition set, generating three-dimensional numerical simulation results, and extracting thermodynamic characteristic parameters at a preset reference section within the incinerator flue from the three-dimensional numerical simulation results to construct a simulation dataset, including: A combination of solid-phase combustion simulation and gas-phase combustion simulation was adopted to perform three-dimensional numerical simulations on each operating parameter in the typical operating condition set to generate initial simulation results; the same mesh division was used for the three-dimensional numerical simulation of each operating parameter. Calculate the simulation error of key measurement points in the initial simulation results, adjust the simulation process parameters according to the simulation error, until the simulation error of all key measurement points is within the preset error threshold, and obtain the three-dimensional numerical simulation results. The thermal characteristic parameters at the preset reference section in the incinerator flue were extracted from the three-dimensional numerical simulation results. The thermal characteristic parameters include: thermal power of each wall grid in front of the preset reference section, flue gas temperature of each grid on the preset reference section, flue gas density of each grid on the preset reference section, flue gas specific heat of each grid on the preset reference section, and flue gas velocity perpendicular to the preset reference section flowing through each grid on the preset reference section. A combustion simulation dataset was constructed based on thermodynamic characteristic parameters.

4. The intelligent control method for steam output of a waste incineration waste heat boiler according to claim 1, characterized in that, The process of using operating condition parameters from a typical operating condition set as input and thermodynamic feature parameters corresponding to the operating condition parameters in the simulation dataset as output for machine learning training to construct a prediction model includes: Using the operating parameters in a typical operating condition set as input features and the thermodynamic feature parameters corresponding to the operating parameters in the simulation dataset as output labels, a machine learning algorithm is used for iterative training to establish a nonlinear mapping relationship between the operating parameters and the thermodynamic feature parameters, thereby obtaining a prediction model.

5. The intelligent control method for steam output of a waste incineration waste heat boiler according to claim 1, characterized in that, The process of collecting the current operating parameters of the incinerator, inputting these parameters into a prediction model to obtain real-time thermal characteristic parameters, and determining the total heat release power of combustion based on these parameters includes: Collect the current operating parameters of the incinerator; Input the current operating parameters into the prediction model, and obtain the real-time thermal power of each wall grid in front of the preset reference section, the real-time flue gas temperature of each grid on the preset reference section, the real-time flue gas density of each grid on the preset reference section, the real-time flue gas specific heat of each grid on the preset reference section, and the real-time flue gas velocity perpendicular to the preset reference section flowing through each grid on the preset reference section through the prediction model. The total real-time thermal power of all wall grids in front of the preset cross section is determined by the sum of the real-time thermal power of each wall grid in front of the preset reference cross section. For each grid on the preset reference section, the real-time flue gas temperature, real-time flue gas density, real-time flue gas specific heat, real-time flue gas velocity, and cross-sectional area where the grid intersects with the preset reference section are multiplied to obtain the real-time thermal power of the grid. The real-time thermal power of each grid on the preset reference section is summed to obtain the real-time total thermal power flowing through the preset reference section. The total real-time heat power of all wall grids in front of the preset cross section is added to the total real-time heat power flowing through the preset reference cross section to obtain the total heat release power of the incinerator.

6. The intelligent control method for steam output of a waste incineration waste heat boiler according to claim 1, characterized in that, The step of determining the corresponding target heat release power based on the target main steam flow rate, and using the minimum deviation between the total combustion heat release power and the target heat release power as the target for adjusting the operating parameters at the current moment, includes: The target heat release power is determined based on the target main steam flow rate through a preset mapping relationship; Taking the current operating parameters as the initial point, the objective function is to minimize the deviation between the total heat release power of combustion and the target heat release power. The boundary constraint is the interval formed by the minimum and maximum values ​​of each operating parameter in the typical operating condition set. The gradient descent method based on the prediction model is used to iteratively optimize the operating parameters. After each iteration, it is determined whether the deviation between the total heat release power of combustion under the candidate operating condition parameters and the target heat release power is within the preset error range. If yes, the operating condition parameters at the current moment are adjusted to the candidate operating condition parameters. If no, the iteration continues until the deviation between the total heat release power of combustion under the candidate operating condition parameters and the target heat release power is within the preset error range or the maximum number of iterations is reached.

7. The intelligent control method for steam output of a waste incineration waste heat boiler according to claim 1, characterized in that, The method further includes: During the operation of the incinerator, the verification operating condition parameters within a preset time period before the current moment are obtained, typical stable combustion operating condition parameters are extracted from the verification operating condition parameters, and a verification operating condition set is constructed based on the typical stable combustion operating condition parameters. Each operating condition parameter in the verification operating condition set is input into the prediction model to obtain the predicted thermodynamic characteristic parameters corresponding to each operating condition parameter, and the predicted flue gas temperature of a specific grid on a preset reference section is extracted from each predicted thermodynamic characteristic parameter. Obtain the actual flue gas temperature at each actual measurement point corresponding to a specific grid, and calculate the deviation between each predicted flue gas temperature and each actual flue gas temperature; The system counts the number of operating parameters whose deviations from the predicted flue gas temperature and the actual flue gas temperature exceed a preset deviation threshold. When the number exceeds a preset warning threshold, a correction process is triggered. The calibration process includes: For each working parameter in the verification working condition set whose deviation exceeds the preset deviation threshold, a three-dimensional numerical simulation of gas-solid coupled combustion in the incinerator is performed to generate the corresponding three-dimensional numerical simulation results. The verification thermodynamic characteristic parameters at the preset reference section in the incinerator flue are extracted from the generated three-dimensional numerical simulation results, and a verification simulation dataset is constructed based on the verification thermodynamic characteristic parameters. The prediction model is incrementally updated using operating condition parameters in the verification operating condition set whose deviations exceed a preset deviation threshold and verification thermodynamic characteristic parameters in the verification simulation dataset as new samples.

8. An intelligent control system for the steam output of a waste incineration waste heat boiler, characterized in that, include: The typical operating condition extraction unit is used to obtain historical operating condition parameters of the incinerator, extract typical stable combustion operating condition parameters from the historical operating condition parameters, and construct a typical operating condition set based on the typical stable combustion operating condition parameters. The simulation dataset construction unit is used to perform three-dimensional numerical simulation of gas-solid coupled combustion in the incinerator for each working parameter in the typical working condition set, generate three-dimensional numerical simulation results, extract the thermodynamic characteristic parameters at the preset reference section in the flue of the incinerator from the three-dimensional numerical simulation results, and construct the simulation dataset. The prediction model building unit is used to train a prediction model by taking the operating parameters in the typical operating condition set as input and the thermodynamic feature parameters in the simulation dataset corresponding to the operating condition parameters as output. The real-time computing unit is used to collect the current operating parameters of the incinerator, input the current operating parameters into the prediction model to obtain real-time thermodynamic characteristic parameters, and determine the total heat release power of combustion based on the real-time thermodynamic characteristic parameters. The operating condition control unit is used to determine the corresponding target heat release power based on the target main steam flow rate, and to adjust the operating condition parameters at the current moment with the goal of minimizing the deviation between the total combustion heat release power and the target heat release power.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the intelligent control method for steam production of waste incineration waste heat boilers as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the intelligent control method for steam production of waste incineration waste heat boiler as described in any one of claims 1 to 7.

Images