Precise control method of SNCR in waste incineration based on mechanism model and dynamic prediction

Abstract

The embodiment of the present invention discloses a method for precise control of SNCR of waste incineration based on a mechanism model and dynamic prediction, which relates to the technical field of flue gas denitrification in waste incineration power plants. The method comprises: obtaining the operating parameters of the waste incinerator; constructing a mechanism model, i.e., a thermodynamic-combustion coupling model, using virtual sensing technology, based on the energy conservation equation, the radiation transfer equation, and CFD inverse solution, inverting the flue gas temperature and flow field velocity vector, and realizing three-dimensional flow field reconstruction and axial temperature field estimation; combining the mechanism model with measured data, optimizing the spray gun parameters using a particle swarm optimization algorithm, and realizing a multi-objective balance between denitrification efficiency and ammonia slip; and dynamically adjusting the spray gun flow rate, angle, and layout according to the optimized parameters. The method of the present invention has high denitrification efficiency and low reducing agent consumption.

Description

Technical Field

[0001] The present invention relates to the technical field of flue gas denitrification in waste incineration power plants, and in particular to a precise control method for SNCR of waste incineration based on a mechanism model and dynamic prediction. Background Art

[0002] In existing technologies, SNCR (Selective Non-Catalytic Reduction) denitrification systems in waste incinerators generally rely on direct high-temperature sensing technologies (such as fiber optic temperature measurement, acoustic wave velocity measurement, and TDLAS gas analysis) to obtain furnace temperature and flow field information. However, the waste incineration process faces the following technical bottlenecks:

[0003] 1. High temperature interference problem: The temperature inside the incinerator is as high as 1200℃. Traditional sensors are easily affected by heat radiation and corrosion, resulting in reduced measurement accuracy or even failure.

[0004] 2. Dynamic response lag: The waste composition is complex (water content and calorific value fluctuate by ±30%), and combustion conditions change frequently. The existing fixed parameter adjustment strategy is difficult to match the dynamic balance of optimal NOx generation and reduction in real time;

[0005] 3. Lack of spatial control accuracy: Single-point temperature measurement and fixed spray gun layout result in a reductant injection coverage rate of only 85%, and ammonia slip easily exceeds the standard (>8mg / Nm³);

[0006] 4. Limitations of model simplification: Existing mechanism models mostly assume ideal combustion conditions and do not integrate multi-physics field coupling, resulting in large prediction deviations.

[0007] Patent application CN116808804A discloses a SNCR control method based on the dynamic characteristics of NOx generation within a furnace. This method divides the NOx generation cycle into four phases (rapid drop, low level, rapid rise, and high level), adjusts the reducing agent dosage according to each phase, and establishes a dynamic control curve (e.g., the ratio between the maximum dosage Hh and the minimum dosage Hl). This method is applicable to the periodic NOx fluctuations of vibrating grate boilers.

[0008] The patent application has the following main shortcomings:

[0009] Strong furnace type specificity: Relying on the periodicity of the vibrating grate, it is not adaptable enough to non-periodic fluctuations (such as random fluctuations in the calorific value of garbage);

[0010] Simplified control parameters: The stages are divided only based on NOx concentration fluctuations, without integrating key reaction parameters such as fly ash catalytic effect and three-dimensional temperature field distribution, resulting in limited optimization of reducing agent dosage.

[0011] Patent application CN106362561A discloses a clustered SNCR control method based on the furnace flow field. This method groups boilers into groups and adjusts the reductant injection rate based on a fuzzy inference algorithm. The method dynamically allocates the spray gun flow rate based on the furnace temperature field (five temperature ranges), flue gas volume, and unit load. The control logic of this method is a fuzzy PID control method with fixed temperature partitioning.

[0012] The patent application has the following main shortcomings:

[0013] Rough division of temperature field: only divide into 5 groups according to average temperature (such as 900-1000℃), ignoring the influence of temperature gradient and local hot spots on reaction efficiency;

[0014] Simplified flow field simulation: The flow field simulation relies on the burner operation status and flue gas volume, and the accuracy of flow field parameters (such as velocity vector) is insufficient;

[0015] Control strategy lag: Feedback control based on outlet NOx concentration deviation lacks feedforward prediction and responds slowly to sudden changes in garbage components (such as sudden changes in calorific value).

[0016] In summary, the existing technology relies on a single parameter, simplifies mechanism modeling, and lacks dynamic adaptability, resulting in low denitrification efficiency (40-60%) and high reducing agent consumption in the SNCR system. Summary of the Invention

[0017] In view of this, an embodiment of the present invention provides a precise SNCR control method for waste incineration based on a mechanism model and dynamic prediction, which has high denitrification efficiency and low reducing agent consumption.

[0018] A precise control method for SNCR of waste incineration based on mechanism model and dynamic prediction, including:

[0019] Step S101: Acquire operating parameters of a waste incinerator, wherein the waste incinerator is provided with a wall temperature sensor array and a pressure sensor array;

[0020] Step S102: Constructing a mechanism model, namely a thermodynamic-combustion coupling model, using virtual sensing technology, based on the energy conservation equation, radiation transfer equation, and CFD inverse solution, inverting the flue gas temperature and flow field velocity vector to achieve three-dimensional flow field reconstruction and axial temperature field estimation, wherein the thermodynamic-combustion coupling model includes a combustion and heat transfer sub-model and a NOx generation sub-model;

[0021] Step S103: combining the mechanism model and measured data, using the particle swarm optimization algorithm to optimize the spray gun parameters to achieve a multi-objective balance between denitrification efficiency and ammonia slip;

[0022] Step S104: Dynamically adjust the flow rate, angle and layout of the spray gun according to the optimized parameters.

[0023] Preferably, in step S101, the operating parameters include fuel calorific value, flue gas mass flow rate, combustion air mass flow rate, flue gas oxygen content, furnace air temperature, and fuel nitrogen content.

[0024] Preferably, in step S102, the combustion and heat transfer sub-models are as follows:

[0025] ,

[0026] ,

[0027] Among them, m fuel is the fuel mass flow rate, LHV is the fuel calorific value, Q comb is the effective heat released by combustion, Q loss is the heat loss of the furnace, m flue is the flue gas mass flow rate, C p is the constant pressure specific heat of flue gas, T flue is the flue gas temperature, T air is the temperature of the air entering the furnace.

[0028] Preferably, in step S102, the NOx generation sub-model is as follows:

[0029] ,

[0030] in, is the NOx generation rate, A is the pre-exponential factor, E is the activation energy, R is the gas constant, T is the combustion zone temperature, is the oxygen volume fraction, and [Fuel-N] is the fuel nitrogen content.

[0031] Preferably, the step S102 includes:

[0032] Based on the pressure sensor array, the continuity equation is solved by CFD inversely, the residual of the mass conservation equation is minimized, and the flow field velocity vector is inverted. ,

[0033] Among them, u is the flow field velocity vector, ρ is the flue gas density, which is calculated based on the flue gas temperature and pressure P , R S is the flue gas constant, S m is the mass source term, passing through the flue gas mass flow m flue and combustion air mass flow m air And the furnace volume V calibration, .

[0034] Preferably, the step S102 includes:

[0035] Using the radiation transfer equation, the axial temperature distribution of the furnace is fitted based on the wall temperature array.

[0036] According to the single point wall heat flux density ,

[0037] Inverse flue gas temperature ,

[0038] in, is the wall heat flux density, is the wall temperature, is the Stefan-Boltzmann constant, is the wall emissivity.

[0039] Preferably, the step S103 includes:

[0040] Step S1031: Automatically adjust the CFD calculation grid density according to the fuel calorific value wave;

[0041] Step S1032: Optimizing spray gun parameters with the goal of maximizing denitrification efficiency and minimizing ammonia slip;

[0042] Step S1033: Calculate the spray gun adjustment amount using an improved PID algorithm, wherein the improved PID algorithm introduces a mechanism model to predict the correction amount. and the dynamic compensation coefficient α.

[0043] Preferably, in step S1031, the grid resolution of the key area is increased to at least 1m.

[0044] Preferably, in step S1032, the spray gun parameter optimization formula is: ,

[0045] in, is a negative weight, Ammonia escape, is the weight coefficient;

[0046] Constraints: spray coverage ≥ 96%, ammonia nitrogen molar ratio ≤ 1.2, single spray gun flow fluctuation ≤ 20%.

[0047] Preferably, in step S1033, the improved PID algorithm formula is:

[0048] ,

[0049] in, is the flow adjustment value of the (i, j)-th spray gun, Kp, Ki, Kd are PID control parameters, e(t) is the NOx concentration deviation, which is equal to the measured value - the set value, α is the dynamic compensation coefficient, is the flow correction amount predicted by the mechanism model.

[0050] The present invention has the following beneficial effects:

[0051] Mechanism-data dual-driven modeling: Combining the thermodynamic equations of waste incineration, this allows for dynamic calibration of theoretical predictions and measured data, addressing the issue of direct sensor failure in high-temperature areas.

[0052] Virtual sensing technology: Utilizing furnace pressure and wall temperature data collected by the newly added furnace pressure and wall multi-point temperature array (8 channels), CFD (fluid dynamics calculation software) is used to inversely solve the flow field continuity equation. Combined with the radiation transfer equation (RTE), the axial temperature distribution is inverted to achieve non-contact three-dimensional field reconstruction, improving the reliability and accuracy of parameter measurement under complex working conditions.

[0053] Adaptive injection optimization: Particle swarm optimization (PSO) is used to dynamically adjust the spray gun layout and injection parameters to improve the matching accuracy of the ammonia-nitrogen molar ratio, suppress ammonia escape, increase injection coverage, ensure sufficient mixing of the reducing agent and flue gas, and achieve high denitrification efficiency and low reducing agent consumption. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0055] Figure 1 Schematic diagram of the process of the waste incineration SNCR precise control method based on mechanism model and dynamic prediction of the present invention;

[0056] Figure 2 This is a system architecture diagram of the waste incineration SNCR precise control method based on mechanism model and dynamic prediction of the present invention;

[0057] Figure 3 A flowchart for implementing the virtual sensing technology in the method of the present invention;

[0058] Figure 4 This is a closed-loop control flow chart of the injection strategy optimization algorithm in the method of the present invention. DETAILED DESCRIPTION

[0059] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0060] It should be understood that the embodiments described are only a portion of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without creative work are within the scope of protection of the present invention.

[0061] The embodiment of the present invention provides a method for precise control of high-efficiency SNCR of waste incineration based on (multi-dimensional) mechanism model and dynamic prediction, such as Figure 1-2 Shown, including:

[0062] Step S101: Acquire operating parameters of a waste incinerator, wherein the waste incinerator is provided with a wall temperature sensor array and a pressure sensor array;

[0063] As an optional embodiment, in step S101, the operating parameters include fuel calorific value, flue gas mass flow rate, combustion air mass flow rate, flue gas oxygen content, furnace air temperature, and fuel nitrogen content.

[0064] In the specific implementation, in terms of hardware configuration, the following are added: pressure and furnace wall multi-point temperature array (8 channels, temperature measurement range 0℃~1200℃, accuracy 1℃; pressure measurement range -500Pa~500Pa, accuracy 0.1Pa); the original fuel calorific value (LHV) and flue gas mass flow rate (m flue )、Combustion air mass flow rate(m air ), flue gas oxygen content (ε), furnace air temperature (T air ), fuel nitrogen content ([Fuel-N]) and other parameters.

[0065] Step S102: Constructing a mechanism model, namely a thermodynamic-combustion coupling model, using virtual sensing technology, based on the energy conservation equation, radiation transfer equation, and CFD inverse solution, inverting the flue gas temperature and flow field velocity vector to achieve three-dimensional flow field reconstruction and axial temperature field estimation, wherein the thermodynamic-combustion coupling model includes a combustion and heat transfer sub-model and a NOx generation sub-model;

[0066] In this step, the thermodynamic-combustion coupling model includes two core sub-models, namely the combustion and heat transfer sub-model and the NOx generation sub-model.

[0067] As an optional embodiment, the combustion and heat transfer sub-model is as follows:

[0068] Based on the energy conservation equation:

[0069] in,

[0070] Parameter meaning:

[0071] m fuel : fuel mass flow rate (kg / s), measured by the feeding system;

[0072] LHV: fuel calorific value (kJ / kg);

[0073] Q comb : effective heat released by combustion (kJ / s);

[0074] Q loss : Furnace heat loss (kJ / s). According to the waste incinerator design manual, it is a fixed proportion of the fuel calorific value, usually 3%-5% of the fuel calorific value at rated load;

[0075] m flue : Flue gas mass flow rate (kg / s);

[0076] C p : Flue gas constant pressure specific heat ( ), according to the smoke composition ( etc.) calculations, which are related to temperature (check the thermodynamic properties table);

[0077] T flue : Flue gas temperature (K), inverted by wall temperature array (not directly detected, calculated by radiation heat transfer equation);

[0078] T air : Temperature of air entering the furnace (K).

[0079] As another optional embodiment, the NOx generation sub-model is as follows (combining the fuel nitrogen conversion mechanism and reaction kinetics):

[0080] However, considering that the main generation pathway of NOx in domestic waste incinerators is (>93.7% of the total NOx content), so the NOx generation (rate) sub-model of the waste incinerator is:

[0081]

[0082] Parameter meaning:

[0083] r NOx : NOx generation rate ;

[0084] A: Pre-exponential factor, reflecting the frequency and direction suitability of reactant molecular collisions, with a value of 10 12 ;

[0085] E: Activation energy (J / mol), the minimum energy required for the reaction to occur, with a value of 120,000 J / mol.

[0086] R: gas constant, value .

[0087] T is the combustion zone temperature (K), which is inverted by the wall temperature array and the radiation heat transfer equation (indirect measurement, indirect calculation);

[0088] ε is the oxygen volume fraction (%), corresponding to the oxygen content of flue gas, which is measured by DCS;

[0089] [Fuel-N] is the fuel nitrogen content (%), which is obtained through fuel composition analysis.

[0090] In this way, the above process can complete the construction of a multi-dimensional mechanism model, that is, a thermodynamic-combustion coupling model.

[0091] As another optional embodiment, step S102 includes:

[0092] Based on pressure sensor array (accuracy ), solve the continuity equation by CFD inversely, minimize the residual of the mass conservation equation, and invert the flow field velocity vector ,

[0093] Parameter meaning:

[0094] u: flow field velocity vector (m / s);

[0095] ρ: Flue gas density (kg / m³), calculated based on flue gas temperature and pressure P[ , R S is the flue gas constant, approximately ;

[0096] S m : Mass source term , which represents the mass input rate per unit volume, is used to calibrate the source term of mass conservation in the furnace, and reflects the mass inflow of flue gas and air during fuel combustion, through the flue gas mass flow rate m flue and combustion air mass flow m air And the furnace volume V calibration, .

[0097] In this way, the three-dimensional flow field reconstruction can be completed through this step.

[0098] As another optional embodiment, step S102 includes:

[0099] The radiative transfer equation (RTE) is used to fit the furnace axial temperature distribution based on the wall temperature array (8 channels). ,

[0100] Inverse flue gas temperature ,

[0101] The key parameters mean:

[0102] : Wall heat flux density, which indicates the amount of radiant heat received or lost by the wall per unit area, a typical value for a waste incinerator .

[0103] : Wall temperature, obtained through multi-point temperature array.

[0104] : Stefan-Boltzmann constant .

[0105] : Wall emissivity (0~1, depends on the wall material, about 0.8 for refractory bricks in waste incinerators).

[0106] In this way, the axial temperature field estimation can be completed through this step.

[0107] The three-dimensional flow field reconstruction and axial temperature field estimation can be done as follows: Figure 3 As shown in the figure, the non-contact three-dimensional field reconstruction technology solves the problem of direct sensing failure in high-temperature areas through wall sensors and inversion algorithms, reconstructs the furnace temperature field and flow field through non-contact methods, and improves the reliability and accuracy of parameter measurement under complex working conditions.

[0108] Step S103: combining the mechanism model and measured data, using the particle swarm optimization algorithm to optimize the spray gun parameters to achieve a multi-objective balance between denitrification efficiency and ammonia slip;

[0109] In this step, the DCS measured parameters (flue gas mass flow, oxygen content, etc.) are dynamically integrated with the predicted values ​​of the mechanism model to correct the non-directly detected parameters in the high temperature area such as Tflue and flow field velocity vector. Figure 4 As shown in the figure, the closed-loop control process of dynamic optimization of spray gun parameters is presented. It integrates the prediction of mechanism model and real-time feedback to achieve precise spray control. The process can be as follows:

[0110] 1. Input data driven: correction amount is predicted based on measured NOx concentration deviation e(t), fuel calorific value fluctuation (LHV±10%) and mechanism model. is the input that drives the control algorithm.

[0111] 2. Dual-algorithm collaborative optimization:

[0112] PSO algorithm: With the goal of maximizing denitrification efficiency and minimizing ammonia escape, it optimizes the flow rate, angle, and layout of the spray gun to meet the constraints of injection coverage ≥96% and ammonia-nitrogen molar ratio ≤1.2.

[0113] Improved PID algorithm: Introducing mechanism model prediction correction And the dynamic compensation coefficient α (0.3-0.8) solves the problem of delayed response of traditional PID to complex working conditions.

[0114] 3. Optimized parameter-driven spray gun adjustment: Combined with dynamic meshing (1m grid resolution in key areas), spatial coverage accuracy is improved, and measured emission data is fed back into the model to form an iterative optimization closed loop.

[0115] That is, as an optional embodiment, the step S103 adopts an adaptive injection strategy, including:

[0116] Step S1031 (dynamic meshing): automatically adjust the CFD calculation mesh density according to the fuel calorific value wave;

[0117] In this step, the CFD calculation grid density can be automatically adjusted according to the fuel calorific value fluctuation (LHV ± 10%), and the grid resolution in key areas (such as burners and spray gun layers) is increased to 1m.

[0118] Step S1032 (Particle Swarm Optimization (PSO)): Optimize the spray gun parameters with the goal of maximizing denitrification efficiency and minimizing ammonia slip;

[0119] In this step, the spray gun parameters are optimized with the goal of maximizing denitrification efficiency and minimizing ammonia escape: ,

[0120] That is, in actual optimization, Converting to negative weights ensures that the algorithm improves denitrification efficiency and reduces ammonia escape by reducing the objective function value. :Directly aiming at minimizing ammonia escape, the weight coefficients ω1 and ω2 are dynamically configured according to engineering requirements (for example, when denitrification efficiency is prioritized, ω1>ω2, such as: ω1=0.7, ω2=0.3).

[0121] Constraints: spray coverage ≥ 96%; ammonia nitrogen molar ratio ≤ 1.2; single spray gun flow fluctuation ≤ 20%.

[0122] Step S1033: Calculate the spray gun adjustment amount using an improved PID algorithm, wherein the improved PID algorithm introduces a mechanism model to predict the correction amount. and the dynamic compensation coefficient α.

[0123] In this step, the gun adjustment amount is calculated using an improved PID algorithm, which introduces a mechanism model to predict the correction amount based on traditional PID control. And the dynamic compensation coefficient α solves the problem of delayed response of pure data-driven PID to complex working conditions:

[0124]

[0125] Where:

[0126] : Flow rate adjustment of the (i, j)-th spray gun (L / h);

[0127] Kp, Ki, Kd: PID control parameters, Kp is the proportional coefficient, Ki is the integral coefficient, and Kd is the differential coefficient;

[0128] e(t) is the NOx concentration deviation (mg / Nm³), which is equal to the measured value minus the set value;

[0129] α is the dynamic compensation coefficient (adaptively adjusted according to the fluctuation rate of fuel calorific value, ranging from 0.3-0.8);

[0130] The flow rate correction value predicted by the mechanism model is based on the combustion and heat transfer sub-model and the NOx generation sub-model. There is a close feedforward-feedback coupling relationship in the calculation, and the NOx generation rate sub-model is one of the core bases for realizing "mechanism prediction-driven dynamic adjustment of the spray gun".

[0131] For example: When When the pressure rises, the mechanism model will calculate the need to increase the amount of reducing agent injection >0, otherwise it decreases.

[0132] Step S104: Dynamically adjust the flow rate, angle and layout of the spray gun according to the optimized parameters.

[0133] The system architecture diagram (four-layer architecture) corresponding to the present invention can be as follows Figure 2 The figure shows the overall architecture of the SNCR precision control system for waste incinerators. It uses a hierarchical design to implement a closed-loop process from data acquisition, model calculation, intelligent decision-making to execution control, specifically including:

[0134] 1. Data acquisition layer: The operating parameters obtained through the DCS interface are combined with the newly added wall temperature array and pressure sensor array to collect key data in real time.

[0135] 2. Model calculation layer: Construct a thermodynamic-combustion coupling model (combustion and heat transfer sub-model, NOx generation sub-model) and a virtual sensing module. Based on the energy conservation equation, the radiation transfer equation (RTE) and CFD inverse solution, invert the indirect detection parameters such as flue gas temperature and flow field velocity vector.

[0136] 3. Intelligent decision-making layer: The mechanism model is calibrated with measured data, and the particle swarm optimization algorithm (PSO) is used to optimize the spray gun parameters to achieve a multi-objective balance between denitrification efficiency and ammonia escape.

[0137] 4. Execution layer: Dynamically adjust the spray gun flow, angle, and layout based on the optimized parameters, and improve the reductant injection coverage through the atomization system, forming a closed-loop control of "data collection-model calculation-intelligent decision-making-execution feedback".

[0138] In summary, the present invention has the following beneficial effects:

[0139] Mechanism-data dual-driven modeling: Combining the thermodynamic equations of waste incineration, this allows for dynamic calibration of theoretical predictions and measured data, addressing the issue of direct sensor failure in high-temperature areas.

[0140] Virtual sensing technology: Utilizing furnace pressure and wall temperature data collected by the newly added furnace pressure and wall multi-point temperature array (8 channels), CFD (fluid dynamics calculation software) is used to inversely solve the flow field continuity equation. Combined with the radiation transfer equation (RTE), the axial temperature distribution is inverted to achieve non-contact three-dimensional field reconstruction, improving the reliability and accuracy of parameter measurement under complex working conditions.

[0141] Adaptive injection optimization: Particle swarm optimization (PSO) is used to dynamically adjust the spray gun layout and injection parameters to improve the matching accuracy of the ammonia-nitrogen molar ratio, suppress ammonia escape, increase injection coverage, ensure sufficient mixing of the reducing agent and flue gas, and achieve high denitrification efficiency and low reducing agent consumption.

[0142] The present invention specifically relates to a SNCR dynamic control method that integrates a thermodynamic-combustion coupling model, virtual sensing technology, and an intelligent optimization algorithm. The method is suitable for removing nitrogen oxides (NOx) from flue gas under complex operating conditions in waste incinerators. Through multi-dimensional modeling and intelligent optimization, the present invention fills the gap in control accuracy and robustness under complex operating conditions. It is particularly suitable for scenarios where high-temperature sensors fail and waste composition fluctuates.

[0143] Implementation case (a waste incineration power plant in Zhejiang):

[0144] Processing scale: 750t / d domestic waste;

[0145] Inlet NOx concentration: 300mg / Nm³ (standard state, );

[0146] Daily average of export indicators , the process detection value is ;

[0147] Reducing agent: 20% ammonia water;

[0148] (1) Calculation of mechanism model application

[0149] 1) Combustion heat and smoke temperature inversion

[0150] Fuel calorific value (LHV): Assuming the calorific value of garbage is 6000kJ / kg,

[0151] Fuel mass flow rate ,

[0152] Heat of combustion: ,

[0153] Wall heat flux density:

[0154] Smoke temperature inversion: The wall temperature array measures the combustion zone's corresponding wall temperature of 950°C (1223K), and combines this with the radiation heat transfer equation to calculate (Correction for heat dissipation loss is taken into account).

[0155] 2) NOx generation rate calculation

[0156] Oxygen volume fraction (ε): Measured at 5% (standard state), ε=0.05

[0157] The nitrogen content of the fuel is 1.5%, then [Fuel-N]=0.015;

[0158]

[0159] 3) Calculation of spray gun adjustment amount

[0160] NOx concentration deviation: ;

[0161] Dynamic compensation coefficient: Assuming the fuel calorific value fluctuates by +10%, α=0.6;

[0162] Mechanistic model prediction correction: (Indicates that the reducing agent is predicted to be excessive and the injection needs to be reduced)

[0163]

[0164] (K p =0.8, K i =0.05, K d =0.2, assuming the differential term is 0 in steady state)

[0165] , the integral gradually adjusts to equilibrium over time.

[0166] (2) Virtual sensing and injection optimization

[0167] Three-dimensional flow field reconstruction: The furnace negative pressure of -100Pa was detected by a pressure sensor array (accuracy 0.1Pa). Combined with the flue gas temperature of 1400K, the flow velocity vector was obtained by CFD inverse solution. The spray gun layout was adjusted to the high flow velocity area, and the spray coverage rate was increased from 85% to 96%, ensuring that the reducing agent and flue gas were fully mixed.

[0168] Dynamic meshing: When the fuel calorific value fluctuates by +10%, the mesh resolution in key areas is increased to 1m, optimizing the reducing agent atomization path and reducing ammonia escape.

[0169] (3) Effect verification

[0170] Spatial coverage: spray coverage rate of spray gun is increased to 96% (traditional solution is 85%);

[0171] Emission performance: NOx emission hourly average ≤150mg / Nm³, daily average ≤100mg / Nm³ (reduced by 40%), ammonia slip ≤8mg / Nm³ (reduced by 60%);

[0172] Economical: Denitrification efficiency exceeds 70%, the amount of reducing agent 20% ammonia water is saved by 18.3%, and the compressed air energy consumption is reduced by 15.1%.

[0173] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included in the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A precise control method for SNCR of waste incineration based on mechanism model and dynamic prediction, characterized in that: include: Step S101: Acquire operating parameters of a waste incinerator, wherein the waste incinerator is provided with a wall temperature sensor array and a pressure sensor array; Step S102: Constructing a mechanism model, namely a thermodynamic-combustion coupling model, using virtual sensing technology, based on the energy conservation equation, radiation transfer equation, and CFD inverse solution, inverting the flue gas temperature and flow field velocity vector to achieve three-dimensional flow field reconstruction and axial temperature field estimation, wherein the thermodynamic-combustion coupling model includes a combustion and heat transfer sub-model and a NOx generation sub-model; Step S103: combining the mechanism model and measured data, using the particle swarm optimization algorithm to optimize the spray gun parameters to achieve a multi-objective balance between denitrification efficiency and ammonia slip; Step S104: Dynamically adjust the flow rate, angle and layout of the spray gun according to the optimized parameters; Wherein, the step S102 includes: Based on the pressure sensor array, the continuity equation is solved by CFD inversely, the residual of the mass conservation equation is minimized, and the flow field velocity vector is inverted. , Where u is the flow field velocity vector, ρ is the flue gas density, calculated based on the flue gas temperature and pressure P ，R S is the flue gas constant, S m is the mass source term, through the flue gas mass flow m flue and combustion air mass flow m air And the furnace volume V calibration, .

2. The method according to claim 1, characterized in that In step S101, the operating parameters include fuel calorific value, flue gas mass flow rate, combustion air mass flow rate, flue gas oxygen content, furnace air temperature, and fuel nitrogen content.

3. The method according to claim 1, characterized in that In step S102, the combustion and heat transfer sub-models are as follows: ， ， Among them, m fuel is the fuel mass flow rate, LHV is the fuel calorific value, Q comb is the effective heat released by combustion, Q loss is the heat loss of the furnace, m flue is the flue gas mass flow rate, C p is the constant pressure specific heat of flue gas, T flue is the flue gas temperature, T air is the temperature of the air entering the furnace.

4. The method according to claim 1, wherein In step S102, the NOx generation sub-model is as follows: ， in, is the NOx formation rate, A is the pre-exponential factor, E is the activation energy, R is the gas constant, T is the combustion zone temperature, ε is the oxygen volume fraction, and [Fuel-N] is the fuel nitrogen content.

5. The method according to claim 1, wherein The step S102 includes: Using the radiation transfer equation, the axial temperature distribution of the furnace is fitted based on the wall temperature array. According to the single point wall heat flux density , Inverse flue gas temperature , in, is the wall heat flux density, is the wall temperature, is the Stefan-Boltzmann constant, is the wall emissivity.

6. The method according to claim 1, characterized in that The step S103 includes: Step S1031: Automatically adjust the CFD calculation grid density according to the fuel calorific value wave; Step S1032: Optimizing spray gun parameters with the goal of maximizing denitrification efficiency and minimizing ammonia slip; Step S1033: Calculate the spray gun adjustment amount using an improved PID algorithm, wherein the improved PID algorithm introduces a mechanism model to predict the correction amount. and the dynamic compensation coefficient α.

7. The method according to claim 6, characterized in that In step S1031, the grid resolution of the key area is increased to at least 1 meter.

8. The method according to claim 6, characterized in that In step S1032, the spray gun parameter optimization formula is: , in, is a negative weight, Ammonia escape, is the weight coefficient; Constraints: spray coverage ≥ 96%, ammonia nitrogen molar ratio ≤ 1.2, single spray gun flow fluctuation ≤ 20%.

9. The method according to claim 6, characterized in that In step S1033, the improved PID algorithm formula is: , in, is the flow adjustment value of the (i, j)-th spray gun, Kp, Ki, Kd are PID control parameters, e(t) is the NOx concentration deviation, which is equal to the measured value - the set value, α is the dynamic compensation coefficient, is the flow correction amount predicted by the mechanism model.

Images