Fly ash coal oxygen-enriched melting intelligent control method based on multi-mode coupling prediction

Abstract

The embodiment of the invention discloses a fly ash coal oxygen-enriched melting intelligent control method based on multi-mode coupling prediction, and belongs to the technical field of solid waste recycling and intelligent control crossing. The method comprises the steps that S101, multi-source data collected by a sensor is obtained; step S102, obtaining control parameters through an intelligent control algorithm according to the multi-source data; s103, according to the control parameters, controlling an oxygen-enriched spray gun, a slurry outlet and a material ratio; wherein the step S102 comprises the sub-steps of performing decoupling prediction on the temperature T, the hearth negative pressure P, the flue gas oxygen concentration CO and the molten slurry yield Qp by adopting a multivariable decoupling prediction model; carrying out fuzzy control on the hearth negative pressure P by adopting a fuzzy self-adaptive PID (Proportion Integration Differentiation) controller; and a multi-objective optimization algorithm is adopted to calculate and obtain control parameters. The control method is small in melting temperature fluctuation, high in hearth negative pressure stability, high in quartz sand utilization rate, low in energy consumption per ton of molten slurry and high in heat preservation cotton qualification rate.

Description

Technical Field

[0001] The present invention relates to the technical field of intersection between solid waste resource utilization and intelligent control, and in particular to an intelligent control method for oxygen-enriched melting of fly ash coal based on multimodal coupling prediction. Background Art

[0002] Fly ash primarily originates from industrial facilities such as coal-fired power plants and waste incineration plants. It contains large amounts of fine particulate matter and hazardous substances, such as heavy metals and dioxins. Improper handling can pose a serious threat to the environment and human health. Fly ash melting is an effective waste treatment technology that transforms fly ash into a stable vitreous product through high-temperature melting, achieving both harmlessness and resource utilization. This not only reduces harmful substances in fly ash but also transforms it into valuable building materials (such as insulation).

[0003] In the existing fly ash coal oxygen-enriched melting system for preparing thermal insulation cotton, traditional segmented PID control method and single-objective optimization control method are usually adopted, among which:

[0004] In the traditional segmented PID control method, the control strategy is: using segmented PID control to adjust single parameters such as melting temperature and furnace negative pressure (for example, setting PID controllers to control temperature and pressure separately); implementation method: through a fixed-parameter PID regulator, relying on the operator's experience to set PID parameters (such as proportional coefficient, integral time, etc.), to perform simple closed-loop control on the real-time data fed back by the sensor; applicable scenarios: under conditions where the fly ash composition is stable and the operating conditions change slightly, basic temperature and pressure control can be maintained.

[0005] The disadvantage of this control method is the problem of strong coupling of multiple variables: no coupling model of fuel quantity, oxygen concentration, material ratio and temperature, furnace negative pressure and slurry output has been established. The dynamic interaction between variables (such as changes in fuel quantity affect both temperature and negative pressure) leads to serious adjustment lag (the pure lag of temperature response is 16-28s), which easily leads to the risk of negative pressure exceeding the limit (>200Pa).

[0006] In the single-objective optimization control method, the control objective is to optimize a single indicator (such as energy consumption or output), for example, by adjusting the fuel amount and oxygen flow to minimize the unit energy consumption; the implementation method is to adjust the operating parameters through a single-objective optimization algorithm (such as the gradient descent method) based on a fixed mechanism model (such as the heat transfer equation Q=kAΔT) or an empirical formula; the applicable scenario is that local optimization can be achieved in scenarios with relatively stable operating conditions and a single objective.

[0007] The disadvantages of this control method are: (1) time-varying nonlinear system: the fluctuation of fly ash chlorine content (±15%) causes furnace slagging, changes the heat transfer coefficient k (fluctuation range ±20%), and the conventional model Q=kAΔT cannot be corrected in real time, resulting in temperature control deviation (±50°C); (2) multi-dimensional control target conflict: the slurry production Q must be satisfied at the same time. p ≥2t / d, insulation cotton fiber diameter (1-4μm qualified rate ≥90%), furnace negative pressure P∈(-200, 20Pa), traditional single-target control is difficult to balance. Summary of the Invention

[0008] In view of this, an embodiment of the present invention provides an intelligent control method for fly ash coal oxygen-enriched melting based on multimodal coupling prediction to solve the problems of multivariable strong coupling, time-varying nonlinear characteristics and multidimensional control target conflicts in the process of preparing thermal insulation cotton by oxygen-enriched melting of fly ash from waste incineration.

[0009] An intelligent control method for fly ash coal oxygen-enriched melting based on multi-modal coupling prediction, comprising:

[0010] Step S101: Acquire multi-source data collected by sensors;

[0011] Step S102: obtaining control parameters through an intelligent control algorithm according to the multi-source data;

[0012] Step S103: controlling the oxygen-enriched lance, the slurry outlet, and the material ratio according to the control parameters;

[0013] Wherein, the step S102 includes:

[0014] Step S1021: For temperature T, furnace negative pressure P, flue gas oxygen concentration C O and slurry production Q p , a multivariable decoupling prediction model is used for decoupling prediction;

[0015] Step S1022: For the furnace negative pressure P, a fuzzy adaptive PID controller is used for fuzzy control;

[0016] Step S1023: Calculate the control parameters using a multi-objective optimization algorithm.

[0017] Preferably, in step S101, the multi-source data collected by the sensor includes:

[0018] Temperature data collected by K-type thermocouples deployed at the bottom, top and outlet of the incinerator;

[0019] The furnace negative pressure data is collected by the pressure transmitter deployed at the center of the incinerator furnace waist;

[0020] Flue gas oxygen concentration data collected by a zirconia oxygen meter deployed at the flue gas cooler outlet;

[0021] Frequency data collected by a frequency converter deployed at the feed port of a screw conveyor for coal / fly ash / quartz sand / alumina.

[0022] Preferably, in step S1021, the multivariable decoupling prediction model uses the following state space equation to describe the melting process:

[0023] ;

[0024] Where X=[T,P,C O ,Q p ] T : state vector; T: temperature in melting furnace; P: negative pressure in furnace; C O : Flue gas oxygen concentration; Q p : slurry production;

[0025] U=[m c ,V O ,r] T : control vector; m c : Fuel coal feed amount; V O : oxygen flow rate; r: material ratio, i.e. fly ash: quartz sand: alumina, mass ratio;

[0026] Y: output vector, corresponding to sensor measurements; A: system matrix, describing the natural dynamic relationship between state variables; B: input matrix, characterizing the coupled effects of control variables on state variables; C: output matrix, mapping state variables to sensor measurements; W: process noise; V: measurement noise;

[0027] In the input matrix B, the coupling coefficient is defined based on the mechanism model , using LSTM to process the nonlinear and time-varying coupling characteristics in historical data and capture the control vector element u j With the state vector element x i The dynamic mapping relationship between them replaces the fixed coupling coefficient matrix in the traditional mechanism model and outputs the b ij Dynamic value.

[0028] Preferably, the step S1021 includes:

[0029] A multi-step prediction is performed every 5 seconds, and the LSTM model outputs the state of the next 30 seconds. The m is adjusted in advance through rolling optimization. c ;

[0030] And / or, calculate the prediction error E, and if E>5%, trigger the model online update.

[0031] Preferably, in step S1022, the fuzzy adaptive PID controller is a two-dimensional fuzzy controller, and the input is the furnace negative pressure deviation e=P set -P real With the deviation change rate ec=Δe / Δt, the output is ΔK p , ΔK i , ΔK d ;

[0032] The membership function adopts Gaussian type:

[0033] ;

[0034] Where, e: furnace negative pressure deviation; c: i : center of the universe; σ i : Gaussian function standard deviation, ΔK p : Proportional coefficient increment, ΔK i : Integral coefficient increment, ΔK d : differential coefficient increment, P set : Furnace negative pressure setting value, P real : Real-time measurement value of furnace negative pressure.

[0035] Preferably, the step S1022 includes:

[0036] When |e|>50Pa, start fuzzy control:

[0037] (1) Calculate e(t)=P set -P(t), ec(t)=[e(t)-e(t-1)] / Δt;

[0038] (2) Fuzzy reasoning to obtain ΔK p , ΔK i , ΔK d , update PID parameters: K p (t)=K p (t-1)+ΔK p , K i (t)=K i (t-1)+ΔK i , K d (t)=K d (t-1)+ΔK d ;

[0039] (3) Control output u(t)=K p e+K i ∑eΔt+K d Δe / Δt, ensure that the negative pressure fluctuation is ≤±10Pa.

[0040] Preferably, in step S1023, a dual objective function is defined:

[0041] Calculate the objective function 1 to minimize energy consumption:

[0042] ;

[0043] Among them, m c : Fuel coal feed amount; LHV: Lower heating value; V O : oxygen flow rate; C cost : Oxygen cost unit price; Q p : slurry production;

[0044] Calculate the objective function 2 for maximizing the component qualification rate:

[0045] ;

[0046] Among them, r: actual material ratio; r opt : optimal material ratio; k: slope parameter of S-type function.

[0047] Preferably, the step S1023 includes:

[0048] Run the optimization algorithm every 10 minutes:

[0049] (1) Initialization population: 50 individuals, coded as [m c ,V O ,r];

[0050] (2) Calculate the fitness f1, f2, and perform non-dominated sorting and congestion calculation;

[0051] (3) Selection, crossover, and mutation to generate offspring;

[0052] (4) Retain the top 20% Pareto optimal solutions and output the optimal operating point ( , , ).

[0053] Preferably, a hydraulic push rod system is provided at the slurry outlet, and the hydraulic push rod system is linked with the slurry outlet opening α to satisfy the flow equation Q of the incompressible fluid. p =k v αΔp 0.5 , where k v is the flow coefficient; Δp is the pressure difference between the upstream and downstream of the slurry outlet.

[0054] Preferably, the step S103 includes:

[0055] Step S104: Establish a three-dimensional thermodynamic model of the melting furnace, map the temperature field and pressure distribution in real time, and set a three-level alarm mechanism.

[0056] The present invention aims to address technical bottlenecks such as strong multivariable coupling, time-varying nonlinear characteristics, and multidimensional control objective conflicts in the production of thermal insulation wool from fly ash from waste incineration using oxygen-enriched melting. By coupling multi-source data fusion with intelligent algorithms, this method provides a highly efficient (increased production capacity), intelligent (fully automated control), and environmentally friendly (low energy consumption and low emissions) fully automated control method. This method addresses the core shortcomings of traditional control technologies and improves key performance indicators. This intelligent control method for oxygen-enriched melting of fly ash and coal, based on multimodal coupling prediction, achieves minimal melting temperature fluctuations, strong furnace negative pressure stability, high quartz sand utilization, low energy consumption per ton of slurry, and a high insulation wool qualification rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] 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.

[0058] Figure 1 Schematic diagram of the flow of the intelligent control method for fly ash coal oxygen-enriched melting based on multi-modal coupling prediction of the present invention;

[0059] Figure 2 This is a system architecture diagram of the intelligent control method for fly ash-coal oxygen-enriched melting based on multimodal coupling prediction of the present invention. It shows the data flow and control flow of the four-layer architecture, highlighting the closed-loop logic of "data acquisition → algorithm processing → execution control → monitoring feedback";

[0060] Figure 3 This is a flow chart of fuzzy control rule reasoning in the method of the present invention (for furnace negative pressure control);

[0061] Figure 4 This is a multi-objective optimization flow chart of NSGA-II in the method of the present invention (coordinated optimization of energy consumption and qualified rate). DETAILED DESCRIPTION

[0062] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0063] 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.

[0064] This invention relates to a fully automatic control method for a "fly ash and coal oxygen-enriched melting system for producing thermal insulation wool." This method addresses the high-temperature melting process (1300-1500°C) of waste incineration fly ash (processing capacity 1.80 t / d) and auxiliary materials (quartz sand 0.65 t / d, alumina 0.12 t / d). By coupling multi-source data fusion with intelligent algorithms, it addresses the complex nonlinear control challenges of oxygen-enriched combustion (oxygen flow rate 250 Nm³ / h), material proportioning (coal 2.45 t / d), and slurry conditioning (slurry output ≥ 2 t / d), achieving efficient, intelligent, and environmentally friendly thermal insulation wool production.

[0065] The embodiment of the present invention provides an intelligent control method for fly ash coal oxygen-enriched melting (for preparing thermal insulation cotton) based on multi-modal coupling prediction, such as Figure 1 Shown, including:

[0066] Step S101: Acquire multi-source data collected by sensors;

[0067] This step corresponds to Figure 2 The (multi-source) data acquisition layer in .

[0068] As an optional embodiment, in step S101, the multi-source data collected by the sensor includes:

[0069] Temperature data collected by K-type thermocouples deployed at the bottom, top and outlet of the incinerator;

[0070] The furnace negative pressure data is collected by the pressure transmitter deployed at the center of the incinerator furnace waist;

[0071] Flue gas oxygen concentration data collected by a zirconia oxygen meter deployed at the flue gas cooler outlet;

[0072] Frequency data collected by a frequency converter deployed at the feed port of a screw conveyor for coal / fly ash / quartz sand / alumina.

[0073] In specific implementation, the hardware configuration of this step can be as shown in Table 1:

[0074] Table 1

[0075]

[0076] Step S102: obtaining control parameters through an intelligent control algorithm according to the multi-source data;

[0077] This step corresponds to Figure 2 The intelligent control algorithm layer in .

[0078] Wherein, the step S102 includes:

[0079] Step S1021: For temperature T, furnace negative pressure P, flue gas oxygen concentration CO and slurry production Q p , a multivariable decoupling prediction model is used for decoupling prediction;

[0080] During the oxygen-enriched melting process of fly ash coal, the temperature T, furnace negative pressure P, flue gas oxygen concentration C O and slurry production Q p There are complex multivariable coupling relationships between them. Traditional PID control has difficulty dealing with such strongly coupled, nonlinear dynamic systems. The core goal of the multivariable decoupling prediction model is to achieve accurate prediction and control by constructing state-space equations, combining mechanism models with data-driven methods, and decoupling the interactions between variables. As an optional embodiment, the multivariable decoupling prediction model uses the following state-space equations to describe the melting process:

[0081] ;

[0082] in:

[0083] X=[T,P,C O ,Q p ] T : State vector, containing four core parameters.

[0084] U=[m c ,V O ,r] T : Control vectors, representing the fuel coal feed rate, oxygen flow rate, and material ratio respectively.

[0085] Y: Output vector, corresponding to the sensor measurement value (such as temperature, pressure, etc.).

[0086] A: System matrix, which describes the natural dynamic relationship between state variables (such as the natural decay of temperature over time).

[0087] B: Input matrix, characterizing the coupled effects of the control variables on the state variables (key decoupling parameters).

[0088] C: Output matrix, mapping state variables to sensor measurements.

[0089] W: Process noise (such as fly ash composition fluctuations, environmental interference).

[0090] V: Measurement noise (sensor error).

[0091] For the state vector X=[T,P,C O ,Q p ] T

[0092] T: Temperature inside the melting furnace (core control parameter), temperature of the molten pool and flue gas above the furnace, designed operating range 1300℃~1500℃;

[0093] P: Furnace negative pressure (key parameter to prevent harmful gas escape), real-time value of the pressure measuring point at the center of the furnace waist, designed control range -2000 Pa ~200Pa;

[0094] C O : Flue gas oxygen concentration (oxygen-enriched combustion efficiency index), the detection value of the zirconia oxygen meter at the flue gas cooler outlet, reflects the utilization rate of oxygen-enriched air;

[0095] Q p : Slurry output (core production capacity indicator), the amount of slurry discharged from the melting furnace every day, the design scale is ≥2t / d;

[0096] For the control vector U=[m c ,V O ,r] T

[0097] m c : Fuel coal feed rate is controlled by coal conveying screw, with a rated feed rate of 102kg / h (2.45t / d);

[0098] V O : Oxygen flow rate (core parameter of oxygen-enriched combustion system), oxygen production system outlet flow rate, design capacity 250Nm³ / h;

[0099] r: material ratio (fly ash: quartz sand: alumina), mass ratio, the design value is 1.80t / d:0.65t / d:0.12t / d≈15:5.4:1;

[0100] For the input matrix B

[0101] Defining coupling coefficients based on the mechanism model (b ij That is, each element in the input matrix B), where x i is the i-th element of the state vector X (i=1,2,3,4, corresponding to T,P,C O ,Q p ), u j is the jth element of the control vector U (j=1,2,3, corresponding to m c ,V O ,r). The input matrix B is a 4×3 matrix, whose i-th row and j-th column element b ij Represents the control vector element u j For the state vector element x i The coupling coefficient of the control vector u is captured by using the LSTM model to process the nonlinear and time-varying coupling characteristics in the historical data (such as the difference in the impact of fuel volume on temperature under different chlorine contents). j With the state vector element xi The dynamic mapping relationship between them replaces the fixed coupling coefficient matrix in the traditional mechanism model and outputs the b ij Dynamic value.

[0102] The complete structure of the input matrix B is as follows:

[0103] ;

[0104] Each row in the matrix corresponds to a state variable, and each column corresponds to a control variable. For example:

[0105] Line 1 [b T-mc , b T-VO , b T-r ] represents the coupled effects of fuel quantity, oxygen flow rate, and material ratio on temperature;

[0106] Column 2 [b T-VO , b P-VO , b CO-VO , b Qp-VO ] represents the coupled effect of oxygen flow on temperature, negative pressure, oxygen concentration and output.

[0107] For example, when the fly ash chlorine content increases from 2.5% to 5.8%, LSTM learns b T-mc (b T-mc The rate of change in unit fuel coal feed rate on the melting temperature increases from 3.0℃ / (t / d) to 4.2℃ / (t / d), and b is updated adaptively. T-mc , avoiding the insufficient lag correction caused by the fixed k in the conventional model.

[0108] For the output matrix C (its corresponding sensor mapping relationship)

[0109] The output matrix maps state variables to sensor measurements:

[0110] ;

[0111] Ensure that each state variable corresponds to the real-time monitoring data of a unique sensor.

[0112] Model advantages and decoupling mechanism

[0113] Dynamic decoupling: Quantify the coupling relationship between variables through matrix B and decompose the multivariable system into approximately independent subsystems. For example, adjust m c When the coupling effect is large, the model can predict its joint impact on T and P, and compensate for the coupling effect through the control algorithm.

[0114] Rolling optimization: Execute a multi-step forecast (for the next 30 seconds) every 5 seconds, and use real-time data to correct forecast errors (if the error is greater than 5%, a model update is triggered), improving dynamic response speed.

[0115] Data-driven enhancement: LSTM training supplements the deficiencies of the mechanism model and adapts to complex working conditions such as fluctuations in fly ash composition (such as chlorine content ±15%).

[0116] As another optional embodiment, step S1021 includes:

[0117] A multi-step prediction is performed every 5 seconds, and the LSTM model outputs the state of the next 30 seconds. The m is adjusted in advance through rolling optimization. c ;

[0118] And / or, calculate the prediction error E, and if E>5%, trigger the model online update.

[0119] In specific implementation, the coupled model prediction (rolling optimization) performs a prediction every 5 seconds:

[0120] Input current U(t)=[m c (t),V O (t),r(t)]

[0121] 1. The LSTM model outputs the state X(t+1), ..., X(t+6) for the next 30 seconds (time step 5 seconds). m is adjusted in advance through rolling optimization. c (For example, when it is predicted that the temperature will overshoot, the fuel amount will be reduced 10 seconds in advance), shortening the response time of the traditional PID from >16 seconds to within 15 seconds.

[0122] 2. Calculate the prediction error , if E>5%, the model online update is triggered.

[0123] Step S1022: For the furnace negative pressure P, a fuzzy adaptive PID controller is used for fuzzy control;

[0124] This step uses a fuzzy adaptive PID controller (for furnace negative pressure control).

[0125] As an optional embodiment, a two-dimensional fuzzy controller is designed, and the input is the negative pressure deviation e=P set -P real With the deviation change rate ec=Δe / Δt, the output is ΔK p , ΔK i , ΔK d .

[0126] ΔK p : Proportional coefficient increment, used to dynamically adjust the proportional gain of the PID controller to improve the system response speed;

[0127] ΔK i : Integral coefficient increment, used to adjust the strength of the integral action and eliminate static errors;

[0128] ΔK d : Differential coefficient increment, used to adjust the intensity of the differential action and suppress system overshoot.

[0129] The membership function adopts Gaussian type: ;

[0130] in:

[0131] e: furnace negative pressure deviation, e=P set -P real , the set value is the target value within the range of -200~20Pa;

[0132] P set : Furnace negative pressure set value (control target value);

[0133] P real : Real-time measurement value of furnace negative pressure (collected through pressure transmitter).

[0134] c i : The center of the universe (fuzzy subset core value), take NB (negative large) = -200Pa, NM (negative medium) = -100Pa, ZO (zero) = 0Pa, PM (positive medium) = 100Pa, PB (positive large) = 200Pa;

[0135] σ i : Gaussian function standard deviation (determines the width of membership distribution), take 50Pa, ensure the overlap of adjacent fuzzy subsets is about 30%, and improve control smoothness;

[0136] The fuzzy rule matrix (partial examples) can be shown in Table 2:

[0137] Table 2

[0138]

[0139] As another optional embodiment, for fuzzy PID dynamic adjustment, when |e|>50Pa, fuzzy control is started:

[0140] 1. Calculate e(t)=P set -P(t), ec(t)=[e(t)-e(t-1)] / Δt;

[0141] 2. Obtain ΔK by fuzzy reasoning p , ΔK i, ΔK d , update PID parameters: K p (t)=Kp (t-1)+ΔK p , K i (t)=K i (t-1)+ΔK i , K d (t)=K d (t-1)+ΔK d ;

[0142] 3. Control output u(t)=K p e+K i ∑eΔt+K d Δe / Δt, ensure that the negative pressure fluctuation is ≤±10Pa.

[0143] The execution process of this step can be as follows Figure 3 As shown, Figure 3 It fully presents the closed-loop process from negative pressure acquisition to PID parameter adjustment, highlighting the core steps of fuzzy logic "input fuzzification → rule reasoning → defuzzification", the core of which is the application of two-dimensional fuzzy rule matrix and Gaussian membership function.

[0144] Step S1023: Calculate the control parameters using a multi-objective optimization algorithm.

[0145] This step can specifically adopt the NSGA-Ⅱ multi-objective optimization algorithm, which aims to coordinately optimize energy consumption and pass rate.

[0146] As an optional embodiment, in this step, a dual objective function is defined:

[0147] Objective function 1 (minimizing energy consumption):

[0148] ;

[0149] Numerator: total daily fuel cost (coal calorific value consumption + oxygen procurement cost);

[0150] Denominator: daily slurry production;

[0151] Objective: Minimize the unit slurry energy cost;

[0152] in,

[0153] m c : Fuel coal feed rate (same as state vector definition), unit: t / d;

[0154] LHV=25MJ / kg (the conventional value in the industry, the lower heating value of coal corresponds to 25000kJ / kg);

[0155] V O : oxygen flow rate (same as state vector definition), unit: Nm³ / h;

[0156] C cost : The unit price of oxygen cost is estimated to be 0.8 yuan / Nm³ based on the market price of industrial liquid oxygen;

[0157] Q p : magma production (same as state vector definition), unit: t / d;

[0158] Objective function 2 (maximizing component qualification rate):

[0159] ;

[0160] When the actual ratio r approaches the optimal ratio r opt When , f2 approaches 1 (qualified rate 100%);

[0161] The nonlinear mapping relationship reflects the sensitivity of the insulation cotton composition (such as the Al2O3 / SiO2 ratio) to the ratio fluctuation;

[0162] in:

[0163] r: actual material ratio (mass ratio of fly ash: quartz sand: alumina) calculated in real time, dynamically determined by the feed rate of each material;

[0164] r opt : Optimal material ratio (test calibration value), fly ash: quartz sand: alumina = 15:5.4:1;

[0165] k: S-type function slope parameter (empirical value k=10), used to amplify the impact of ratio deviation on the qualified rate and improve optimization sensitivity.

[0166] As another optional embodiment, for multi-objective optimization solution (NSGA-II implementation), the optimization algorithm is run every 10 minutes:

[0167] 1. Initialize the population (50 individuals, coded as [m c ,V O ,r]);

[0168] 2. Calculate the fitness f1, f2, and perform non-dominated sorting and congestion calculation;

[0169] 3. Select, crossover (single-point crossover, probability 0.8), and mutation (Gaussian mutation, standard deviation 0.05) to generate offspring;

[0170] 4. Keep the top 20% Pareto optimal solutions and output the optimal operating point ( , , ), fuel efficiency is improved, reducing energy consumption per ton of slurry, including:

[0171] : Optimal fuel coal feed rate (unit: t / d), the optimal value of fuel amount obtained by NSGA-Ⅱ algorithm optimization;

[0172] : Optimal oxygen flow (unit: Nm³ / h), the optimal air supply rate for the oxygen-enriched combustion system;

[0173] : Optimal material ratio (mass ratio of fly ash: quartz sand: alumina).

[0174] The execution process of this step can be as follows Figure 4 As shown, Figure 4 Demonstrates how the non-dominated sorting genetic algorithm (NSGA-Ⅱ) solves the fuel quantity m through selection, crossover, and mutation operations. c , oxygen flow V O , the Pareto optimal solution of the material ratio r corresponds to the dual objectives of "minimizing the unit slurry energy consumption cost" and "maximizing the composition qualification rate".

[0175] Step S103: controlling the oxygen-enriched lance, the slurry outlet, and the material ratio according to the control parameters;

[0176] This step corresponds to Figure 2 The actuator layer in the controller is used to achieve precise control.

[0177] As an optional embodiment, the control in this step mainly includes:

[0178] 8 sets of oxygen-enriched spray gun regulating valves (DN20): pneumatic regulating valves are used, with flow control accuracy of ±1% and response time of <1s;

[0179] Hydraulic push rod system: stroke 1600mm, linked with the slurry outlet opening α, meeting the flow equation Q of incompressible fluid p =k v αΔp 0.5 (k v is the flow coefficient; it characterizes the flow characteristics of the valve or slurry outlet and is obtained through on-site calibration; Δp is the pressure difference between the upstream and downstream of the slurry outlet, which is the power source driving the slurry flow);

[0180] Screw conveyor frequency converter: speed regulation range 0-50Hz, realizing quantitative feeding of coal (102kg / h±15%) and fly ash (75kg / h±10%).

[0181] As another optional embodiment, the step S103 includes:

[0182] Step S104: Establish a three-dimensional thermodynamic model of the melting furnace, map the temperature field and pressure distribution in real time, and set a three-level alarm mechanism.

[0183] This step corresponds to Figure 2 Remote monitoring layer (digital twin) in.

[0184] In specific implementation, a three-dimensional thermodynamic model of the melting furnace is established to map the temperature field (resolution ±50mm) and pressure distribution (accuracy ±5Pa) in real time;

[0185] Set up a three-level alarm mechanism:

[0186] Early warning (deviation ≤ 10%): sound and light prompts, automatic adjustment of control parameters;

[0187] Over limit (deviation 10%-30%): trigger the backup spray gun and start the emergency feeding mode;

[0188] Fault (deviation > 30%): The DCS system is shut down and the fault code is recorded (based on Bayesian network diagnosis, accuracy ≥ 95%).

[0189] The results of the method of the present invention are shown in Table 3:

[0190] Table 3

[0191]

[0192] In summary, the present invention aims to address technical bottlenecks such as strong multivariable coupling, time-varying nonlinear characteristics, and multidimensional control objective conflicts in the process of preparing thermal insulation cotton from fly ash from waste incineration using oxygen-enriched melting. By coupling multi-source data fusion with intelligent algorithms, the present invention provides a highly efficient (increased production capacity), intelligent (fully automatic control), and environmentally friendly (low consumption and low emissions) fully automatic control method, addressing the core shortcomings of traditional control technologies and improving key performance indicators. The present invention's intelligent control method for oxygen-enriched melting of fly ash and coal based on multimodal coupling prediction achieves minimal melting temperature fluctuations, strong furnace negative pressure stability, high quartz sand utilization, low energy consumption per ton of slurry, and a high insulation cotton qualification rate.

[0193] Implementation Cases:

[0194] Project Background (A Waste Incineration Plant in Guangxi)

[0195] Core equipment: oxygen-enriched melting furnace (melt pool size 1500×800×600mm), four-roller centrifuge (power 30kW);

[0196] Difficulties in control: Fluctuations in fly ash chlorine content (average 4.2%, range 2.5%-5.8%), and air humidity >85% during the rainy season, which affects material fluidity;

[0197] On-site verification of key formulas:

[0198] 1). Multivariable decoupling model parameter calibration

[0199] Get the coupling coefficient through step response test:

[0200] The fuel quantity increases by 10% (2.45t / d→2.695t / d), and the melting temperature increases by 35℃ (1300℃→1335℃). T-mc =3.5℃ / (t / d);

[0201] The oxygen flow rate increases by +5% (250Nm³ / h→262.5Nm³ / h), and the furnace negative pressure decreases by 20Pa. P-VO =-4Pa / (Nm3 / h);

[0202] 2). Fuzzy PID control parameter tuning

[0203] Determine the initial PID parameters based on on-site debugging:

[0204] K p 0=1.2, K i 0=0.05, K d 0=0.8;

[0205] Fuzzy adjustment range: K p ∈[0.8,1.5], K i ∈[0.03,0.1], K d ∈[0.5,1.2];

[0206] Typical working conditions: When air leakage at the feed inlet causes the negative pressure to drop suddenly to -150Pa, the fuzzy controller stabilizes the negative pressure to -10Pa within 15s, which is 25s faster than the conventional PID.

[0207] 3). Multi-objective optimization calculation example

[0208] The fly ash composition on a certain day is: chlorine 3.8%, heavy metals (Pb+Cd) 0.08%, set Q p =2.1t / d;

[0209] Initial parameter: m c =2.5t / d, V O =240Nm³ / h, r=2.8:1:0.15;

[0210] Objective function values: f1=12.5 yuan / t, f2=0.82;

[0211] After NSGA-Ⅱ optimization: m c =2.38t / d, V O =255Nm³ / h, r=2.7:1:0.18;

[0212] Optimization effect: f1 = 10.2 yuan / t (↓18.4%), f2 = 0.91 (↑10.9%). At the same time, the furnace negative pressure fluctuation was reduced from ±80Pa to ±15Pa, and the insulation cotton qualification rate was increased from 82% to 95%.

[0213] 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. An intelligent control method for fly ash coal oxygen-enriched melting based on multimodal coupling prediction, characterized in that: include: Step S101: Acquire multi-source data collected by sensors; Step S102: obtaining control parameters through an intelligent control algorithm according to the multi-source data; Step S103: controlling the oxygen-enriched lance, the slurry outlet, and the material ratio according to the control parameters; Wherein, the step S102 includes: Step S1021: For temperature T, furnace negative pressure P, flue gas oxygen concentration C O and slurry production Q p , a multivariable decoupling prediction model is used for decoupling prediction; Step S1022: For the furnace negative pressure P, a fuzzy adaptive PID controller is used for fuzzy control; Step S1023: Calculate the control parameters using a multi-objective optimization algorithm.

2. The intelligent control method for fly ash coal oxygen-enriched melting based on multimodal coupling prediction according to claim 1 is characterized in that: In step S101, the multi-source data collected by the sensor includes: Temperature data collected by K-type thermocouples deployed at the bottom, top and outlet of the incinerator; The furnace negative pressure data is collected by the pressure transmitter deployed at the center of the incinerator furnace waist; Flue gas oxygen concentration data collected by a zirconia oxygen meter deployed at the flue gas cooler outlet; Frequency data collected by a frequency converter deployed at the feed port of a screw conveyor for coal / fly ash / quartz sand / alumina.

3. The intelligent control method for fly ash coal oxygen-enriched melting based on multi-modal coupling prediction according to claim 1 is characterized in that: In step S1021, the multivariable decoupling prediction model uses the following state space equation to describe the melting process: ； in, : state vector; T: temperature in melting furnace; P: negative pressure in furnace; C O : Flue gas oxygen concentration; Q p : slurry production; : control vector; m c : Fuel coal feed amount; V O : oxygen flow rate; r: material ratio, i.e. fly ash: quartz sand: alumina, mass ratio; Y: output vector, corresponding to sensor measurements; A: system matrix, describing the natural dynamic relationship between state variables; B: input matrix, characterizing the coupled effects of control variables on state variables; C: output matrix, mapping state variables to sensor measurements; W: process noise; V: measurement noise; In the input matrix B, the coupling coefficient is defined based on the mechanism model , using LSTM to process the nonlinear and time-varying coupling characteristics in historical data and capture the control vector element u j With the state vector element x i The dynamic mapping relationship between them replaces the fixed coupling coefficient matrix in the traditional mechanism model and outputs the b ij Dynamic value.

4. The intelligent control method for fly ash coal oxygen-enriched melting based on multi-modal coupling prediction according to claim 3 is characterized in that: The step S1021 includes: A multi-step prediction is performed every 5 seconds, and the LSTM model outputs the state of the next 30 seconds. The m is adjusted in advance through rolling optimization. c ; And / or, calculate the prediction error E, and if E>5%, trigger the model online update.

5. The intelligent control method for fly ash coal oxygen-enriched melting based on multi-modal coupling prediction according to claim 1 is characterized in that: In step S1022, the fuzzy adaptive PID controller is a two-dimensional fuzzy controller, and the input is the furnace negative pressure deviation e=P set -P real With the deviation change rate ec=Δe / Δt, the output is ΔK p , ΔK i , ΔK d ; The membership function adopts Gaussian type: ； Where, e: furnace negative pressure deviation; c: i : center of the universe; σ i : Gaussian function standard deviation, ΔK p : Proportional coefficient increment, ΔK i : Integral coefficient increment, ΔK d : differential coefficient increment, P set : Furnace negative pressure setting value, P real : Real-time measurement value of furnace negative pressure.

6. The intelligent control method for fly ash coal oxygen-enriched melting based on multi-modal coupling prediction according to claim 5 is characterized in that: The step S1022 includes: When |e|>50Pa, start fuzzy control: (1) Calculate e(t)=P set -P(t), ec(t)=[e(t)-e(t - 1)] / Δt; (2) Fuzzy reasoning to obtain ΔK p , ΔK i , ΔK d , update PID parameters: K p (t)=K p (t-1)+ΔK p , K i (t)=K i (t-1)+ΔK i , K d (t)=K d (t-1)+ΔK d ; (3) Control output u(t)=K p e+K i ∑eΔt+K d Δe / Δt, ensure that the negative pressure fluctuation is ≤±10Pa.

7. The intelligent control method for fly ash coal oxygen-enriched melting based on multi-modal coupling prediction according to claim 1 is characterized in that: In step S1023, a dual objective function is defined: Calculate the objective function 1 to minimize energy consumption: ； Among them, m c : Fuel coal feed amount; LHV: Lower heating value; V O : oxygen flow rate; C cost : Oxygen cost unit price; Q p : slurry production; Calculate the objective function 2 for maximizing the component qualification rate: ； Among them, r: actual material ratio; r opt : optimal material ratio; k: slope parameter of S-type function.

8. The intelligent control method for fly ash coal oxygen-enriched melting based on multi-modal coupling prediction according to claim 7 is characterized in that: The step S1023 includes: Run the optimization algorithm every 10 minutes: (1) Initialization population: 50 individuals, coded as [m c ,V O ,r]; (2) Calculate the fitness f1, f2, and perform non-dominated sorting and congestion calculation; (3) Selection, crossover, and mutation to generate offspring; (4) Retain the top 20% Pareto optimal solutions and output the optimal operating point ( , , ).

9. The intelligent control method for fly ash coal oxygen-enriched melting based on multi-modal coupling prediction according to claim 1 is characterized in that: The outlet is equipped with a hydraulic push rod system, which is linked to the outlet opening α to meet the flow equation Q of the incompressible fluid. p =k v αΔp 0.5 , where k v is the flow coefficient; Δp is the pressure difference between the upstream and downstream of the slurry outlet.

10. The intelligent control method for fly ash coal oxygen-enriched melting based on multi-modal coupling prediction according to claim 1 is characterized in that: The step S103 includes: Step S104: Establish a three-dimensional thermodynamic model of the melting furnace, map the temperature field and pressure distribution in real time, and set a three-level alarm mechanism.

Images