Combustion control method, system and device for volatile organic compounds

Abstract

The invention discloses a combustion control method, system and device for volatile organic compounds, and relates to the technical field of thermotechnical automatic control. Acquiring the gas inlet flow rate, the gas inlet concentration and the reversing period of the catalytic combustion reactor when the volatile organic compounds are combusted in the catalytic combustion reactor; inputting the gas inlet flow rate, the gas inlet concentration and the reversing period into the mechanism model to obtain an initial prediction value of the dynamic parameter of the volatile organic compound in the combustion process; analyzing the air inlet flow velocity, the air inlet concentration and the reversing period through an error compensation model to obtain a prediction deviation value of the dynamic parameter; correcting the initial prediction value through the prediction deviation value to obtain a prediction value of the dynamic parameter of the volatile organic compound in the combustion process; when the predicted value of the dynamic parameter does not meet the preset condition, generating a control signal; the control signal is used to adjust the state of the catalytic combustion reactor. The method can accurately control the catalytic combustion process of the volatile organic compounds.

Description

Technical Field

[0001] The invention relates to the technical field of thermal automatic control, and in particular to a combustion control method, system and device for volatile organic compounds. Background Art

[0002] Due to increasingly stringent environmental protection requirements, people have shown increasing interest in air pollution control technologies in recent decades. Among them, Volatile Organic Compounds (VOC) are one of the most common air pollutants. When the VOC content is low, catalytic combustion is a common technology for eliminating such pollutants. The industrial flue gas to be purified is usually close to room temperature and has a high flow rate. Depending on the properties of the VOC gas being treated and the catalyst used, the temperature range required for catalytic combustion varies from 200°C to 400°C. Therefore, the temperature needs to be precisely controlled to prevent the catalyst from deactivating. At the same time, a combustion scheme that can effectively recover the heat of reaction should be used to heat the inlet flue gas. Among the various schemes for achieving energy recovery, the flow conversion system has been widely studied.

[0003] However, the catalytic combustion process of VOC waste gas is complex, involving dynamic nonlinear interactions among phase flow, heat transfer, mass transfer and catalytic reactions. In addition, component and concentration fluctuations pose huge challenges to system safety and stable control of concentration and temperature fields. This makes it difficult to identify the dynamic parameters of the process in real time, and the corresponding active control methods are limited.

[0004] Therefore, there is an urgent need for a method that can accurately control the combustion process of volatile organic compounds. Summary of the invention

[0005] Based on this, it is necessary to provide a combustion control method, system and device for volatile organic compounds in order to address the above technical problems, which can accurately control the combustion process of volatile organic compounds.

[0006] The present invention adopts the following technical solutions:

[0007] The present invention provides a combustion control method for volatile organic compounds, comprising:

[0008] Obtaining the inlet gas flow rate, inlet gas concentration and reversing period of the catalytic combustion reactor when volatile organic matter is burned in the catalytic combustion reactor;

[0009] The intake air flow rate, intake air concentration and reversing period are input into the mechanism model to obtain the initial predicted values ​​of the dynamic parameters of volatile organic compounds during the combustion process; the mechanism model is a group of partial differential control equations based on mass conservation and energy conservation established according to the dynamic nonlinear interactive process of interphase flow, heat transfer, mass transfer and catalytic reaction in the catalytic combustion reactor, which is used to describe the combustion process of volatile organic compounds in the catalytic combustion reactor;

[0010] The intake air flow rate, intake air concentration and switching period are analyzed through the error compensation model to obtain the predicted deviation value of the dynamic parameters; the error compensation model is an online real-time correction machine learning algorithm constructed by learning historical data through a radial basis function neural network;

[0011] The initial prediction value is corrected by the prediction deviation value to obtain the prediction value of the dynamic parameters of volatile organic compounds during the combustion process;

[0012] When the predicted value of the dynamic parameter does not meet the preset condition, a control signal is generated; the control signal is used to adjust the state of the catalytic combustion reactor.

[0013] Preferably, the mechanism model includes a gas phase mass conservation equation, a gas phase energy conservation equation, a heat storage body mass conservation equation, a heat storage body energy conservation equation, a catalyst mass conservation equation, a catalyst energy conservation equation, and a reactor wall energy conservation equation;

[0014] The gas phase mass conservation equation is:

[0015]

[0016] The gas phase energy conservation equation is:

[0017]

[0018] The mass conservation equation of the thermal storage body is:

[0019] C g =C s

[0020] The energy conservation equation of the thermal storage body is:

[0021]

[0022] The catalyst mass conservation equation is:

[0023] h m α(C g -C s )=C s R i

[0024] The catalyst energy conservation equation is:

[0025]

[0026] The reactor wall energy conservation equation is expressed as:

[0027]

[0028] Among them, C g is the gas concentration in the gas phase, t is the reaction time, D eff is the effective diffusion coefficient of the bed axial direction, x is the axial distance along the bed, u0 is the superficial velocity of the bed, ε is the porosity of the bed, ρ g,0 is the gas density at room temperature, ρ g is the actual density of the gas, α is the specific surface area of ​​the bed, h m is the gas-solid phase mass transfer coefficient, C s is the gas concentration in the solid phase; T g is the gas phase temperature, k eff,g is the effective thermal conductivity of the gas phase in the axial direction, C p,g is the relative heat capacity of gas, h is the heat transfer coefficient between gas and solid phases, T s is the solid phase temperature, h w is the heat transfer coefficient between the gas phase and the catalytic combustion reactor wall, T w is the wall temperature of the catalytic combustion reactor, D r is the equivalent inner diameter of the catalytic combustion reactor; C s is the gas concentration in the solid phase; k eff,s is the solid phase axial effective thermal conductivity, C p,s is the solid specific heat capacity, ρ s is the actual density of the solid phase; R i is the gas phase chemical reaction rate; ΔH is the gas phase chemical reaction enthalpy change; k w is the thermal conductivity of the catalytic combustion reactor wall, C p,w is the specific heat capacity of the catalytic combustion reactor wall, ρ w is the actual density of the catalytic combustion reactor wall, d w is the wall thickness of the catalytic combustion reactor, d1 is the equivalent outer diameter of the catalytic combustion reactor wall, d2 is the equivalent outer diameter of the catalytic combustion reactor insulation material, and λ is the thermal conductivity of the insulation material.

[0029] Preferably, the intake air flow rate, intake air concentration and switching period are analyzed by the error compensation model to obtain the predicted deviation value of the dynamic parameter, including:

[0030] Normalizing the intake air flow rate, intake air concentration and switching period respectively to obtain normalized intake air flow rate, intake air concentration and switching period;

[0031] The normalized intake air flow rate, intake air concentration and switching period are input into the error compensation model to obtain the predicted deviation normalized value of the dynamic parameter;

[0032] The prediction deviation normalized value is denormalized to obtain the prediction deviation value.

[0033] Preferably, the process of constructing the error compensation model includes:

[0034] Obtain the original data set, and normalize the original data set to obtain the training data set;

[0035] Iteratively train the radial basis function neural network through the training data set to obtain the loss function value of the radial basis function neural network in each training;

[0036] The parameters of the radial basis function neural network are updated according to the loss function value obtained from each training until the radial basis function neural network converges to obtain an error compensation model.

[0037] Preferably, the radial basis function neural network includes an input layer, a hidden layer and an output layer; the radial basis function neural network is iteratively trained through a training data set to obtain a loss function value of the radial basis function neural network for each training, including:

[0038] For each training, the mapping from the input layer to the hidden layer is represented Determine the output of the training data set in the hidden layer;

[0039] According to the mapping from hidden layer to output layer Determine sample prediction results;

[0040] Determine the loss function value based on the sample prediction results and the corresponding actual results;

[0041] Where X is the n-dimensional input vector (x1, x2, ..., x n ), α is the output of the m-dimensional hidden layer node (α1, α2, …, α m );C i is the center of the ith basis function, a vector of the same dimension as X; σ is the Gaussian kernel width, the number of hidden layer nodes m is determined by changing the number of nodes from small to large using the statistical F test method, and the position of the node center is determined by the K-means clustering method; Y is the r-dimensional output vector (y1, y2, …, y n );||XC i || is the vector XC i The norm of ω ij is the weight from the i-th hidden layer node to the j-th output.

[0042] Preferably, the initial prediction value is corrected by the prediction deviation value to obtain the prediction value of the dynamic parameter of the volatile organic compound during the combustion process, including:

[0043] Get the sum of the prediction deviation value and the initial prediction value;

[0044] The sum of the prediction deviation value and the initial prediction value is determined as the prediction value of the dynamic parameter.

[0045] Preferably, the dynamic parameters include bed temperature distribution and exhaust gas conversion rate; the predicted value of the dynamic parameter does not meet the preset conditions including: the predicted value of the bed temperature distribution is greater than the preset temperature threshold, or the predicted value of the exhaust gas conversion rate is less than the preset conversion threshold.

[0046] The present invention provides a combustion control system for volatile organic compounds, the system comprising a catalytic combustion reactor and a computer device;

[0047] A computer device is used to obtain the intake air flow rate, intake air concentration and reversing period of the catalytic combustion reactor when the volatile organic compounds are burned in the catalytic combustion reactor; and the intake air flow rate, intake air concentration and reversing period are respectively input into the mechanism model and the error compensation model to obtain the initial prediction value and prediction deviation value of the dynamic parameters of the volatile organic compounds during the combustion process; then the initial prediction value is corrected by the prediction deviation value to obtain the prediction value of the dynamic parameters of the volatile organic compounds during the combustion process, and finally, when the prediction value of the dynamic parameter does not meet the preset condition, a control signal is generated; the control signal is used to adjust the state of the catalytic combustion reactor;

[0048] The mechanism model is a set of partial differential control equations based on mass conservation and energy conservation, which is established according to the dynamic nonlinear interaction process of interphase flow, heat transfer, mass transfer and catalytic reaction in the catalytic combustion reactor. It is used to describe the combustion process of volatile organic matter in the catalytic combustion reactor; the error compensation model is an online real-time correction machine learning algorithm constructed by learning historical data through a radial basis function neural network.

[0049] Preferably, the system further comprises a reversing control valve, an online gas concentration measuring device, a gas flow meter and a temperature measuring device; the online gas concentration measuring device is respectively installed at the inlet and outlet pipes of the catalytic combustion reactor, the gas flow meter is installed at the inlet and outlet pipe positions of the catalytic combustion reactor, and the temperature measuring device is installed at the key temperature measuring point position of the catalytic combustion reactor; the reversing control valve is installed at the corresponding position of the system;

[0050] Gas concentration online measuring device, used to measure the gas concentration at the inlet and outlet of the catalytic combustion reactor and transmit the concentration signal to the computer equipment;

[0051] A gas flow meter, used to measure the inlet and outlet gas concentrations of the catalytic combustion reactor and transmit the gas flow signal to the computer device;

[0052] A temperature measuring device, used for measuring the temperature distribution signal in the bed layer and transmitting the temperature distribution signal in the bed layer to a computer device;

[0053] The reversing control valve is used to change the gas flow direction in the catalytic combustion reactor by adjusting the flow direction of volatile organic matter gas, and transmit the reversing signal to the computer equipment.

[0054] The present invention provides a combustion control device for volatile organic compounds, characterized in that the device comprises:

[0055] An acquisition module, used to acquire the intake air flow rate, intake air concentration and reversing period of the catalytic combustion reactor when the volatile organic matter is burned in the catalytic combustion reactor;

[0056] An input module is used to input the intake air flow rate, intake air concentration and reversing period into the mechanism model to obtain the initial prediction value of the dynamic parameters of volatile organic compounds in the combustion process; the mechanism model is a group of partial differential control equations based on mass conservation and energy conservation established according to the dynamic nonlinear interactive process of interphase flow, heat transfer, mass transfer and catalytic reaction in the catalytic combustion reactor, so as to describe the combustion process of volatile organic compounds in the catalytic combustion reactor;

[0057] An analysis module is used to analyze the intake air flow rate, intake air concentration and commutation period through an error compensation model to obtain a predicted deviation value of the dynamic parameters; the error compensation model is an online real-time correction machine learning algorithm constructed by learning historical data through a radial basis function neural network;

[0058] A correction module, used to correct the initial prediction value by using the prediction deviation value to obtain the prediction value of the dynamic parameter of the volatile organic compound during the combustion process;

[0059] The control module is used to generate a control signal when the predicted value of the dynamic parameter does not meet the preset condition; the control signal is used to adjust the state of the catalytic combustion reactor.

[0060] The present invention provides a computer-readable storage medium, wherein the storage medium stores a computer program, and when the computer program is executed by a processor, the above-mentioned combustion control method of volatile organic compounds is implemented.

[0061] The present invention provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned combustion control method of volatile organic compounds when executing the program.

[0062] At least one of the above technical solutions adopted by the present invention can achieve the following beneficial effects:

[0063] In the present invention, the mechanism model is based on the partial differential control equations of mass conservation and energy conservation established according to the dynamic nonlinear interactive process of interphase flow, heat transfer, mass transfer and catalytic reaction in the catalytic combustion reactor, so as to describe the combustion process of volatile organic compounds in the catalytic combustion reactor, and the error compensation model is a machine learning algorithm based on radial basis function neural network for online correction; the mechanism model can describe the flow, chemical reaction and heat transfer process in the catalytic combustion reactor, which is helpful to accurately describe the catalytic combustion of volatile organic compounds in the catalytic combustion reactor, and the error compensation model is a model trained by pure data, and the predicted value output by the mechanism model is corrected by the predicted deviation value output by the error compensation model, that is, the mechanism model of pure chemical reaction is combined with the error compensation model driven by pure data, so that the dynamic parameters generated by the volatile organic compounds in the catalytic combustion process can be obtained more accurately, and an accurate prediction control logic can be formed, so as to accurately control the combustion process of the volatile organic compounds. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0065] Figure 1 A schematic diagram of the structure of a combustion control system for volatile organic compounds provided by the present invention;

[0066] Figure 2 A schematic flow chart of a combustion control method of volatile organic compounds provided by the present invention;

[0067] Figure 3 A schematic diagram of the structure of a radial basis function neural network provided by the present invention;

[0068] Figure 4 A structural schematic diagram of a hybrid model provided by the present invention;

[0069] Figure 5 A schematic diagram of a combustion control device for volatile organic compounds provided by the present invention;

[0070] Figure 6 A schematic diagram of a computer device for implementing a combustion control method of volatile organic compounds provided by the present invention.

[0071] Description of reference numerals:

[0072] 100. Combustion control system for volatile organic compounds; 101. Catalytic combustion reactor; 102. Computer equipment; 103. Reversing control valve; 104. Gas concentration online measuring device; 105. Gas flow meter; 106. Temperature measuring device. DETAILED DESCRIPTION

[0073] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the specific embodiments of the present invention and the corresponding drawings. Obviously, the described embodiments are only part 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 ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0074] The coal chemical industry is one of the important sources of air pollution. Pollutants such as sulfur oxides, nitrogen oxides, and smoke and dust have been controlled, but volatile organic compounds have not been effectively controlled. VOCs are important pollutants that affect the environment and human body, and are also important precursors of ozone, PM2.5, and secondary organic aerosols in the atmosphere. Thermal treatment is currently the mainstream technology for the treatment of industrial source VOCs, which is mainly divided into direct combustion (Thermal Oxidation, TO), regenerative thermal oxidizer (Regenerative Thermal Oxidizer, RTO), catalytic oxidation (Catalytic Oxidation, CO) and regenerative catalytic oxidation (Regenerative Catalytic Oxidation, RCO). Among them, catalytic oxidation technology, especially RCO, has the advantages of low operating temperature, low auxiliary fuel consumption, high safety, and low NOx emissions. It is the development direction of VOC treatment in industries such as coal chemical industry.

[0075] In the prior art, a control method for a catalytic cracking unit is proposed. By determining the process dynamic parameters such as the cyclone separator temperature and the regenerator coke content in the catalytic cracking unit, the optimal parameter setting value data set under variable conditions is established based on the process mechanism model, and a neural network prediction model is established based on this to achieve online prediction of the regenerator coke content. In another technology, a method for constructing a heterogeneous integrated prediction model for the CO emission concentration of the MSWI process is proposed. Offline training data is generated for the multi-condition mechanism model of the municipal solid waste incineration process, a mechanism mapping model is constructed, and an LSTM-based data-driven model is integrated to achieve online prediction of CO emission concentration. In another technology, a hybrid modeling method for a zinc oxide volatilization kiln based on mechanism and support vector machine is proposed. A hybrid model of the zinc oxide volatilization kiln conservation equation group and the support vector machine regression algorithm fitting dynamic parameters is established. While characterizing the operating mechanism and reaction process in the kiln, the dynamic characteristics of the volatilization kiln production process are accurately reflected. The catalytic combustion process of volatile organic compounds has high nonlinearity, and the catalyst is easily deactivated by over-temperature. Conventional control methods are difficult to maintain safe, stable and long-term operation of the reactor. At present, there is no reasonable and feasible online prediction method for VOC catalytic combustion. Compared with the mechanism + data-driven hybrid model, the pure data-driven model established based on the mechanism model data set lacks the mechanistic mathematical representation of the reaction process and is prone to non-physical results. From the above technology, it can be seen that the prediction of the process dynamic parameters of the VOC combustion process in the prior art is based on a pure mechanism model as a prediction method, a pure data-driven model as a prediction method, or a conventional hybrid model method.

[0076] Based on this, the present invention provides a combustion control method, system and device for volatile organic compounds. The method is based on a mechanism model and a neural network model to quickly and accurately identify the dynamic parameters of VOCs during the combustion process, form a predictive control logic, and avoid high-temperature burning of the catalyst and excessive emission of unburned VOC waste gas caused by control lag.

[0077] The technical solutions provided by various embodiments of the present invention are described in detail below in conjunction with the accompanying drawings.

[0078] The present invention provides a combustion control system for volatile organic compounds, such as Figure 1 As shown, the combustion control system 100 of volatile organic compounds includes a catalytic combustion reactor 101 and a computer device 102; the combustion control system 100 of volatile organic compounds also includes a reversing control valve 103, a gas concentration online measuring device 104, a gas flow meter 105 and a temperature measuring device 106; the gas concentration online measuring device 104 is respectively installed at the inlet and outlet pipelines of the catalytic combustion reactor 102, the gas flow meter 105 is installed at the inlet and outlet pipeline positions of the catalytic combustion reactor 102, and the temperature measuring device 106 is installed at the key temperature measuring point position of the catalytic combustion reactor 102; the reversing control valve 103 is installed at the corresponding position of the system.

[0079] The computer device 102 is used to obtain the intake air flow rate, intake air concentration and reversing period of the catalytic combustion reactor 102 when the volatile organic compounds are burned in the catalytic combustion reactor 102; and input the intake air flow rate, intake air concentration and reversing period into the mechanism model and the error compensation model respectively to obtain the initial predicted value and predicted deviation value of the dynamic parameters of the volatile organic compounds during the combustion process; then correct the initial predicted value by the predicted deviation value to obtain the predicted value of the dynamic parameter of the volatile organic compounds during the combustion process, and finally generate a control signal when the predicted value of the dynamic parameter does not meet the preset conditions; the control signal is used to adjust the state of the catalytic combustion reactor 102.

[0080] The mechanism model is a set of partial differential control equations based on mass conservation and energy conservation, which is established according to the dynamic nonlinear interaction process of interphase flow, heat transfer, mass transfer and catalytic reaction in the catalytic combustion reactor 102, and is used to describe the combustion process of volatile organic matter in the catalytic combustion reactor 102; the error compensation model is an online real-time correction machine learning algorithm constructed by learning historical data through a radial basis function neural network.

[0081] The gas concentration online measuring device 104 is used to measure the inlet and outlet gas concentrations of the catalytic combustion reactor 102, and transmit the concentration signal to the computer device 102; the gas flow meter 105 is used to measure the inlet and outlet gas concentrations of the catalytic combustion reactor 102, and transmit the gas flow signal to the computer device 102; the temperature measuring device 106 is used to measure the temperature distribution signal in the bed layer, and transmit the temperature distribution signal in the bed layer to the computer device 102; the reversing control valve 103 is used to change the gas flow direction in the catalytic combustion reactor 102 by adjusting the flow direction of the volatile organic gas, and transmit the reversing signal to the computer device 102.

[0082] Specifically, the reversing control valve 103 is installed at a corresponding position of the pipeline system, and the VOC waste gas flow direction is changed by adjusting the opening and closing of each valve to change the gas flow direction in the catalytic combustion reactor 102, and the reversing signal is transmitted to the computer 5.

[0083] The computer device 102 reads the temperature, intake air concentration, intake air flow rate and switching period, and performs online prediction calculation through a hybrid model composed of a built-in mechanism model and an error compensation model to obtain the bed temperature distribution and exhaust gas conversion rate in the catalytic combustion reactor 102 at a certain time in the future. When the calculated bed temperature in the catalytic combustion reactor 102 exceeds a preset temperature threshold, or the exhaust gas conversion rate is lower than a preset conversion threshold, the computer device 102 outputs a corresponding valve control signal (control signal) to adjust the state of the catalytic combustion reactor 102 to achieve predictive control.

[0084] The temperature measuring device 106 is installed at the key temperature measuring point of the catalytic combustion reactor 102 to measure the temperature distribution in the bed and transmit the temperature signal to the computer device 102. The key temperature measuring points of the catalytic combustion reactor 102 are mainly the corresponding positions where the catalyst bed is prone to overheating and damage. It should be noted that after the computer device 102 receives the temperature signal (temperature distribution in the bed) sent by the temperature measuring device 106, it compares the temperature distribution in the bed with a preset temperature threshold. After the temperature distribution in the bed is greater than the preset temperature threshold, a control signal is generated to adjust the state of the catalytic combustion reactor 102 through the control signal.

[0085] In this way, the process dynamic parameters of the catalytic combustion reactor predicted by the hybrid model and the temperature actually measured by the temperature measuring device are used to comprehensively control the state of the catalytic combustion reactor, thereby achieving precise control of the catalytic combustion reactor.

[0086] Based on the above system, the present invention also provides a combustion control method for volatile organic compounds, which is applied to a computer device in any of the above embodiments, such as Figure 2 As shown, the specific steps include:

[0087] S201, obtaining the intake air flow rate, intake air concentration and reversing period of the catalytic combustion reactor when volatile organic compounds are combusted in the catalytic combustion reactor.

[0088] An online gas concentration measuring device is installed on the inlet and outlet pipes of the catalytic combustion reactor to measure the intake air concentration of the catalytic combustion reactor. After the online gas concentration measuring device measures the intake air concentration, the intake air concentration will be sent to the computer device; similarly, a gas flow meter is installed at the inlet and outlet pipes of the catalytic combustion reactor. After the gas flow meter measures the intake air flow rate of the catalytic combustion reactor, the intake air concentration will be sent to the computer device.

[0089] The computer device controls the switching period of the catalytic combustion reactor, so the computer device can directly obtain the switching period of the catalytic combustion reactor from the memory.

[0090] S202, inputting the intake air flow rate, intake air concentration and reversing period into the mechanism model to obtain the initial predicted values ​​of the dynamic parameters of the volatile organic compounds during the combustion process; the mechanism model is a group of partial differential control equations based on mass conservation and energy conservation established according to the dynamic nonlinear interactive process of interphase flow, heat transfer, mass transfer and catalytic reaction in the catalytic combustion reactor, so as to describe the combustion process of the volatile organic compounds in the catalytic combustion reactor.

[0091] The mechanism model includes the gas phase mass conservation equation, the gas phase energy conservation equation, the heat storage body mass conservation equation, the heat storage body energy conservation equation, the catalyst mass conservation equation, the catalyst energy conservation equation, and the reactor wall energy conservation equation.

[0092] Specifically, the gas phase mass conservation equation is shown in formula (1).

[0093]

[0094] Among them, C g is the gas concentration in the gas phase, t is the reaction time, D eff is the effective diffusion coefficient of the bed axial direction, x is the axial distance along the bed, u0 is the superficial velocity of the bed, ε is the porosity of the bed, ρ g,0 is the gas density at room temperature, ρ g is the actual density of the gas, α is the specific surface area of ​​the bed, h m is the gas-solid phase mass transfer coefficient, C s is the gas concentration in the solid phase.

[0095] The gas phase energy conservation equation can be expressed as:

[0096]

[0097] Among them, T g is the gas phase temperature, k eff,g is the effective axial thermal conductivity of the gas phase, C p,g is the relative heat capacity of gas, h is the heat transfer coefficient between gas and solid phases, T s is the solid phase temperature, h w is the heat transfer coefficient between the gas phase and the catalytic combustion reactor wall, T w is the wall temperature of the catalytic combustion reactor, D r is the equivalent inner diameter of the catalytic combustion reactor.

[0098] The mass conservation equation of the heat storage body can be expressed as:

[0099] C g =C s (3)

[0100] Among them, C s is the gas concentration in the solid phase.

[0101] The energy conservation equation of the thermal storage body can be expressed as:

[0102]

[0103] Among them, k eff,s is the solid phase axial effective thermal conductivity, C p,s is the solid specific heat capacity, ρ s is the actual density of the solid phase.

[0104] The catalyst mass conservation equation can be expressed as:

[0105] h m α(C g -C s )=C s R i (5)

[0106] where R i is the gas-phase chemical reaction rate.

[0107] The catalyst energy conservation equation can be expressed as:

[0108]

[0109] where ΔH is the enthalpy change of the gas-phase chemical reaction.

[0110] The reactor wall energy conservation equation can be expressed as:

[0111]

[0112] where k w is the thermal conductivity of the catalytic combustion reactor wall, C p,w is the specific heat capacity of the catalytic combustion reactor wall, ρ w is the actual density of the catalytic combustion reactor wall, d w is the thickness of the catalytic combustion reactor wall, d1 is the equivalent outer diameter of the catalytic combustion reactor wall, d2 is the equivalent outer diameter of the thermal insulation material of the catalytic combustion reactor, and λ is the thermal conductivity of the thermal insulation material.

[0113] It should be noted that in the input of the mechanism model, the inlet gas velocity is the apparent velocity u0 in the energy conservation equation of the mechanism model, and the inlet gas concentration is the gas concentration C g at x = 0 corresponding value C g0 . The commutation period T0 is the solution boundary of time t in the mechanism model, that is, the solution interval is 0 < t < T0. In the output of the mechanism model, the bed temperature distribution is the distribution Ts(x) of the solid phase temperature T s along the position x of the catalytic combustion reactor; the waste gas conversion rate is the ratio of the difference between the inlet gas concentration and the exhaust gas concentration obtained during the operation of the mechanism model to the inlet gas concentration, that is, the waste gas conversion rate = (C g0 -C goutlet ) / C g0 , where C goutlet is the exhaust gas concentration.

[0114] Therefore, the intake air flow rate, intake air concentration and reversing period are input into the mechanism model to obtain the initial predicted values ​​of the dynamic parameters of volatile organic compounds during the combustion process; the initial predicted values ​​include the initial predicted bed temperature distribution and the initial predicted exhaust gas conversion rate.

[0115] Based on the established mechanism model, preliminary prediction results of dynamic parameters such as temperature distribution and exhaust gas conversion rate of the VOC catalytic combustion reactor in accordance with basic physical laws under different operating input conditions can be obtained.

[0116] S203, analyzing the intake air flow rate, intake air concentration and switching period through an error compensation model to obtain a predicted deviation value of the dynamic parameter; the error compensation model is an online real-time correction machine learning algorithm constructed by learning historical data through a radial basis function neural network.

[0117] Among them, the error compensation model can be constructed using a radial basis function neural network (RBFNN), and its input parameters are the inlet gas component concentration of the VOC gas in the low-temperature catalytic combustion process flowing to the catalytic combustion reactor (Regenerative Catalytic Oxidation, RCO), the gas superficial flow rate in the catalytic combustion reactor, and the catalytic combustion reactor reversing period. The model output parameter is the deviation between the calculated results of the mechanism model (temperature distribution in the catalytic combustion reactor and VOC exhaust gas conversion rate) and the actual value. The schematic diagram of the radial basis function neural network structure is shown in Figure 3 shown.

[0118] Specifically, the construction process of the error compensation model includes: obtaining an original data set and normalizing the original data set to obtain a training data set; iteratively training the radial basis function neural network through the training data set to obtain the loss function value of the radial basis function neural network for each training; updating the parameters of the radial basis function neural network according to the loss function value obtained from each training until the radial basis function neural network converges to obtain the error compensation model.

[0119] The radial basis function neural network training process adopts the orthogonal design method. Under multiple sets of mutually orthogonal operating conditions, the VOC low-temperature catalytic combustion process fluid dynamics (Computational Fluid Dynamics, CFD) is used to calculate and simulate the actual catalytic combustion reactor operation process, fully stimulating the bed temperature distribution characteristics and reaction mechanism characteristics, generating catalytic combustion process operation characteristic information with good orthogonality and a large range of operating point changes, and combining with actual industrial data to construct the catalytic combustion reactor VOC low-temperature catalytic combustion pseudo-state bed temperature distribution and VOCs waste gas conversion rate RBFNN model training samples, and the training data set has both orthogonality and completeness. That is, the original data set includes experimental data, simulation results and industrial data.

[0120] In the original data set, the input parameters and output parameters of the error compensation model in the original data set are normalized, as shown in formula (8).

[0121]

[0122] Among them, x * is the normalized new value, x is the input / output data, and Δ is k0(x max -x min ), x max 、x min are the maximum and minimum values ​​of the input / output data, respectively, and k0 is the proportionality coefficient. By adjusting the proportionality coefficient k0 corresponding to the input / output parameter, the model tracking capability can be improved and the model performance can be improved. Optionally, in a specific implementation, the proportionality coefficient k0 ranges from 1.05 to 1.15.

[0123] Optionally, the radial basis function neural network includes an input layer, a hidden layer and an output layer; the radial basis function neural network is iteratively trained through a training data set to obtain a loss function value of the radial basis function neural network for each training, including: for each training, a mapping representation from the input layer to the hidden layer is used Determine the output of the training data set in the hidden layer; represent it according to the mapping from the hidden layer to the output layer Determine the sample prediction results.

[0124] Where X is the n-dimensional input vector (x1, x2, ..., x n ), α is the output of the m-dimensional hidden layer node (α1, α2, …, α m );C i is the center of the ith basis function, a vector of the same dimension as X; σis the Gaussian kernel width, the number of hidden layer nodes m is determined by changing the number of nodes from small to large using the statistical F test method, and the position of the node center is determined by the K-means clustering method; Y is the r-dimensional output vector (y1, y2, …, y n );||XC i || is the vector XC i The norm of ω ij is the weight from the i-th hidden layer node to the j-th output; the weight update adopts the gradient descent method. Optionally, the value range of the number of hidden layer nodes m can be 15 to 20.

[0125] Determine the loss function value based on the sample prediction results and the corresponding actual results;

[0126] The radial basis function of the radial basis function neural network is a Gaussian function. The input layer X is a training data set consisting of the intake concentration, intake flow rate and switching period of the VOC exhaust low-temperature catalytic combustion reactor. The sample prediction result is the output of the hidden layer. The actual result is the corresponding bed temperature distribution and VOCs exhaust gas conversion rate in the training data set.

[0127] Substitute the sample prediction results and the actual results into the preset loss function to obtain the loss function value; then update the radial basis function neural network according to the loss function value through the gradient descent method or the back propagation method.

[0128] The condition for the radial basis function neural network to converge may be that the number of iterative training times of the radial basis function neural network reaches a preset number threshold, or that the accuracy of the radial basis function neural network reaches a preset accuracy threshold; after each training of the radial basis function neural network is completed, the radial basis function neural network can be evaluated through a test data set to determine the accuracy of the radial basis function neural network.

[0129] In an exemplary embodiment, the intake flow rate, intake concentration and switching period are analyzed by an error compensation model to obtain a predicted deviation value of the dynamic parameter, including: normalizing the intake flow rate, intake concentration and switching period respectively to obtain normalized intake flow rate, intake concentration and switching period; inputting the normalized intake flow rate, intake concentration and switching period into the error compensation model to obtain a predicted deviation normalized value of the dynamic parameter; and denormalizing the predicted deviation normalized value to obtain a predicted deviation value.

[0130] The intake air flow rate, intake air concentration and switching period can be normalized by formula (8), where Δ in formula (8) is k0(x max -x min ), x max 、x minare the maximum and minimum values ​​of the corresponding input / output data in the training data set, respectively. In this way, the intake flow rate, intake concentration and switching period can be directly substituted into formula (8) to obtain the normalized intake flow rate, intake concentration and switching period.

[0131] Then the normalized intake flow rate, intake concentration and commutation period are input into the error compensation model to obtain the normalized value of the predicted deviation of the dynamic parameters. The normalized value of the predicted deviation is the normalized value of the deviation between the initial predicted value and the true value. The normalized value of the predicted deviation includes the normalized value of the deviation between the initial predicted bed temperature distribution and the true bed temperature distribution, and the normalized value of the deviation between the initial predicted exhaust gas conversion rate and the true exhaust gas conversion rate.

[0132] The normalized value of the predicted deviation is denormalized, that is, the data is converted into a uniform range. It can also be denormalized using formula (8), that is, the normalized value of the predicted deviation is substituted into formula (8) as a new normalized value to obtain the predicted deviation value; that is, the normalized value of the deviation between the initial predicted bed temperature distribution and the actual bed temperature distribution, and the normalized value of the deviation between the initial predicted exhaust gas conversion rate and the actual exhaust gas conversion rate are substituted into formula (8) to obtain the deviation value between the initial predicted bed temperature distribution and the actual bed temperature distribution, and the deviation value between the initial predicted exhaust gas conversion rate and the actual exhaust gas conversion rate.

[0133] S204, correcting the initial prediction value by using the prediction deviation value to obtain the prediction value of the dynamic parameter of the volatile organic compound during the combustion process.

[0134] In an exemplary embodiment, the initial prediction value is corrected by the prediction deviation value to obtain the prediction value of the dynamic parameter of the volatile organic compound during the combustion process, including: obtaining the sum of the prediction deviation value and the initial prediction value; and determining the sum of the prediction deviation value and the initial prediction value as the prediction value of the dynamic parameter, as shown in formula (9).

[0135]

[0136] Among them, α is the predicted value of the dynamic parameter, α0 is the initial predicted value, is the predicted deviation value; accordingly, according to formula (9), the initial predicted bed temperature distribution and the sum of the deviation values ​​between the initial predicted bed temperature distribution and the actual bed temperature distribution are calculated to determine the predicted value of the bed temperature distribution, the initial predicted exhaust gas conversion rate and the sum of the deviation values ​​between the initial predicted exhaust gas conversion rate and the actual exhaust gas conversion rate are calculated to determine the predicted value of the exhaust gas conversion rate.

[0137] S205, when the predicted value of the dynamic parameter does not meet the preset condition, a control signal is generated; the control signal is used to adjust the state of the catalytic combustion reactor.

[0138] Specifically, the dynamic parameters include bed temperature distribution and exhaust gas conversion rate; the predicted value of the dynamic parameter does not meet the preset conditions including: the predicted value of the bed temperature distribution is greater than the preset temperature threshold, or the predicted value of the exhaust gas conversion rate is less than the preset conversion threshold. That is, when the predicted value of the bed temperature distribution is greater than the preset temperature threshold, or the predicted value of the exhaust gas conversion rate is less than the preset conversion threshold, a control signal is generated.

[0139] In the combustion control method of volatile organic compounds provided by the present invention, the mechanism model can realize the rapid calculation of the bed temperature distribution and VOC waste gas conversion rate of the catalytic combustion reactor, obtain the calculation results with preliminary accuracy in accordance with the basic physical and chemical principles, and realize the error compensation correction of dynamic parameters by fitting the calculation error of the mechanism model through the radial basis function neural network, so as to realize the rapid prediction and accurate identification of the VOC low-temperature conversion process.

[0140] In an exemplary embodiment, the present invention further provides a hybrid model, the structure of the hybrid model is as follows: Figure 4 As shown, the hybrid model is a coupling model of a mechanism model that provides the original prediction results and a radial basis function neural network that corrects the deviation between the original prediction results and the actual data. The coupling process of the model includes the following steps: establishing a simplified mechanism model of the VOC low-temperature and high-efficiency catalytic combustion reactor, submitting the model to the solver to obtain the bed temperature distribution of the VOC low-temperature and high-efficiency catalytic combustion reactor and the VOC exhaust gas conversion rate calculation results. The model input parameters include the intake flow rate u, intake air concentration c and switching period t, and the output parameters are the bed temperature distribution and exhaust gas conversion rate and other key parameters α0; establishing a data-driven error compensation model based on the radial basis function neural network, and submitting the model to the solver to obtain the deviation of the bed temperature distribution of the VOC low-temperature and high-efficiency catalytic combustion reactor and the VOC exhaust gas conversion rate calculation results α0 from the actual value. The error compensation model is trained through experimental data, simulation results and industrial data, and the deviation between the result α0 calculated by the mechanism model and the true value The coupling mechanism model and the error compensation model are used to calculate the predicted values ​​of key parameters such as the bed temperature distribution and the exhaust gas conversion rate of the catalytic combustion reactor, α, as

[0141] Among them, when the hybrid model is put into actual industrial use, the error compensation model can be corrected online based on real-time industrial data to improve the prediction accuracy of the model under a specific implementation method.

[0142] The hybrid model takes the real-time monitoring data of the VOC low-temperature catalytic reactor as input, and takes the sum of the original prediction result of the mechanism model and the deviation prediction value of the neural network model as output. The combustion control system of volatile organic compounds is installed on the VOC low-temperature catalytic high-efficiency conversion process. The combustion control system of volatile organic compounds includes a gas flow meter, an infrared flue gas analyzer, a temperature measuring thermocouple, and a computer device. The gas flow meter and the infrared flue gas analyzer are installed at the inlet and outlet positions of the catalytic combustion reactor, and the temperature measuring thermocouple is installed on the catalytic combustion reactor. The computer device is used to read the input data of the hybrid model, perform prediction calculations, output prediction results and control signals. The input data includes the reversing cycle, intake flow rate, and intake concentration of the catalytic combustion reactor, and the output prediction results include the catalytic combustion reactor bed temperature distribution, maximum temperature, and exhaust concentration. The combustion control method and system of the present invention can realize the rapid and accurate identification of key process parameters such as the temperature of the catalytic combustion reactor and the exhaust gas conversion rate in the VOC low-temperature conversion process, form a prediction control logic, and avoid the high-temperature burning of the catalyst and the excessive emission of unburned VOC exhaust gas caused by control lag.

[0143] In the present invention, a hybrid model composed of a simplified VOCs low-temperature catalytic oxidation process mechanism model and a radial basis function neural network can realize rapid identification of key parameters of the VOCs catalytic combustion process while satisfying high-precision prediction of bed temperature distribution and VOCs conversion rate, thereby forming a predictive control logic; the training mode of the radial basis function neural network can supplement the training set, realize online correction of the hybrid model, and improve the adaptability and accuracy of the online prediction method; the predictive control logic formed based on the hybrid model can realize early control of reactor overtemperature, protect the catalyst from overheating damage, and effectively extend the catalyst life.

[0144] When applying the combustion control method of volatile organic compounds provided by the present invention, it is not necessary to follow Figure 2 The steps are executed in the order shown. The specific execution order of the steps can be determined according to needs, and the present invention does not limit this.

[0145] The above is a combustion control method of volatile organic compounds provided by one or more embodiments of the present invention. Based on the same idea, the present invention also provides a corresponding combustion control device of volatile organic compounds, such as Figure 5 shown.

[0146] Figure 5 A schematic diagram of a combustion control device for volatile organic compounds provided by the present invention, the device 500 comprises:

[0147] An acquisition module 501 is used to acquire the intake air flow rate, intake air concentration and reversing period of the catalytic combustion reactor when the volatile organic matter is burned in the catalytic combustion reactor;

[0148] An input module 502 is used to input the intake air flow rate, intake air concentration and reversing period into the mechanism model to obtain the initial prediction value of the dynamic parameters of the volatile organic compounds in the combustion process; the mechanism model is a group of partial differential control equations based on mass conservation and energy conservation established according to the dynamic nonlinear interactive process of interphase flow, heat transfer, mass transfer and catalytic reaction in the catalytic combustion reactor, so as to describe the combustion process of the volatile organic compounds in the catalytic combustion reactor;

[0149] The analysis module 503 is used to analyze the intake air flow rate, intake air concentration and switching period through an error compensation model to obtain a predicted deviation value of the dynamic parameter; the error compensation model is an online real-time correction machine learning algorithm constructed by learning historical data through a radial basis function neural network;

[0150] A correction module 504 is used to correct the initial prediction value by using the prediction deviation value to obtain the prediction value of the dynamic parameter of the volatile organic compound during the combustion process;

[0151] The control module 505 is used to generate a control signal when the predicted value of the dynamic parameter does not meet the preset condition; the control signal is used to adjust the state of the catalytic combustion reactor.

[0152] The specific definition of the combustion control device for volatile organic compounds can be found in the definition of the combustion control method for volatile organic compounds mentioned above, which will not be repeated here. Each module in the above-mentioned combustion control device for volatile organic compounds can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0153] The present invention also provides a computer-readable storage medium, which stores a computer program, which can be used to execute the above Figure 2 A combustion control method for volatile organic compounds is provided.

[0154] The present invention also provides Figure 6 The structural diagram of the computer device shown in FIG. Figure 6 As shown in the figure, at the hardware level, the computer device includes a processor, an internal bus, a network interface, a memory, and a non-volatile memory, and may also include other hardware required for the business. The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs it to achieve the above Figure 2 A combustion control method for volatile organic compounds is provided.

[0155] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided by the present invention can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory or optical memory, etc. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0156] The technical features of the above embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of the present invention.

Claims

1. A method for controlling the combustion of volatile organic compounds, characterized in that: The method comprises: Obtaining the intake air flow rate, intake air concentration and reversing period of the catalytic combustion reactor when volatile organic matter is burned in the catalytic combustion reactor; The intake air flow rate, the intake air concentration and the reversing period are input into a mechanism model to obtain initial prediction values ​​of dynamic parameters of the volatile organic matter during the combustion process; the mechanism model is a group of partial differential control equations based on mass conservation and energy conservation established according to the dynamic nonlinear interactive process of interphase flow, heat transfer, mass transfer and catalytic reaction in a catalytic combustion reactor, so as to describe the combustion process of the volatile organic matter in the catalytic combustion reactor; The intake air flow rate, the intake air concentration and the commutation period are analyzed by an error compensation model to obtain a predicted deviation value of a dynamic parameter; the error compensation model is an online real-time corrected machine learning algorithm constructed by learning historical data through a radial basis function neural network; Correcting the initial prediction value by using the prediction deviation value to obtain a prediction value of a dynamic parameter of the volatile organic compound during the combustion process; When the predicted value of the dynamic parameter does not meet the preset condition, a control signal is generated; the control signal is used to adjust the state of the catalytic combustion reactor.

2. The method according to claim 1, characterized in that The mechanism model includes a gas phase mass conservation equation, a gas phase energy conservation equation, a heat storage body mass conservation equation, a heat storage body energy conservation equation, a catalyst mass conservation equation, a catalyst energy conservation equation and a reactor wall energy conservation equation; The gas phase mass conservation equation is: The gas phase energy conservation equation is: The mass conservation equation of the thermal storage body is: C g =C s The energy conservation equation of the thermal storage body is: The catalyst mass conservation equation is: h m α(C g -C s )=C s R i The catalyst energy conservation equation is: The reactor wall energy conservation equation is expressed as: Among them, C g is the gas concentration in the gas phase, t is the reaction time, D eff is the effective diffusion coefficient of the bed axial direction, x is the axial distance along the bed, u0 is the superficial velocity of the bed, ε is the porosity of the bed, ρ g,0 is the gas density at room temperature, ρ g is the actual density of the gas, α is the specific surface area of ​​the bed, h m is the gas-solid phase mass transfer coefficient, C s is the gas concentration in the solid phase; T g is the gas phase temperature, k eff,g is the effective thermal conductivity of the gas phase in the axial direction, C p,g is the relative heat capacity of gas, h is the heat transfer coefficient between gas and solid phases, T s is the solid phase temperature, h w is the heat transfer coefficient between the gas phase and the catalytic combustion reactor wall, T w is the wall temperature of the catalytic combustion reactor, D r is the equivalent inner diameter of the catalytic combustion reactor; C s is the gas concentration in the solid phase; k eff,s is the solid phase axial effective thermal conductivity, C p,s is the solid specific heat capacity, ρ s is the actual density of the solid phase; R i is the gas phase chemical reaction rate; ΔH is the gas phase chemical reaction enthalpy change; k w is the thermal conductivity of the catalytic combustion reactor wall, C p,w is the specific heat capacity of the catalytic combustion reactor wall, ρ w is the actual density of the catalytic combustion reactor wall, d w is the wall thickness of the catalytic combustion reactor, d1 is the equivalent outer diameter of the catalytic combustion reactor wall, d2 is the equivalent outer diameter of the catalytic combustion reactor insulation material, and λ is the thermal conductivity of the insulation material.

3. The method according to claim 1, characterized in that The error compensation model is used to analyze the intake air flow rate, the intake air concentration and the switching period to obtain a predicted deviation value of a dynamic parameter, including: Normalizing the intake air flow rate, the intake air concentration and the switching period respectively to obtain normalized intake air flow rate, intake air concentration and switching period; Inputting the normalized intake air flow rate, intake air concentration and switching period into the error compensation model to obtain a normalized value of the predicted deviation of the dynamic parameter; The predicted deviation normalized value is denormalized to obtain the predicted deviation value.

4. The method according to claim 1, characterized in that: The process of constructing the error compensation model includes: Obtain the original data set, and normalize the original data set to obtain the training data set; Iteratively training the radial basis function neural network using the training data set to obtain a loss function value of the radial basis function neural network for each training; The parameters of the radial basis function neural network are updated according to the loss function value obtained in each training until the radial basis function neural network converges to obtain the error compensation model.

5. The method according to claim 4, characterized in that The radial basis function neural network includes an input layer, a hidden layer and an output layer; the radial basis function neural network is iteratively trained using the training data set to obtain the loss function value of the radial basis function neural network in each training, including: For each training, the mapping from the input layer to the hidden layer is represented Determining the output of the training data set at the hidden layer; According to the mapping representation from the hidden layer to the output layer Determine sample prediction results; Determine the loss function value according to the sample prediction result and the corresponding true result; Where X is the n-dimensional input vector (x1, x2, ..., x n ), α is the output of the m-dimensional hidden layer node (α1, α2, …, α m );C i is the center of the ith basis function, a vector of the same dimension as X; σ is the Gaussian kernel width, the number of hidden layer nodes m is determined by changing the number of nodes from small to large using the statistical F test method, and the position of the node center is determined by the K-means clustering method; Y is the r-dimensional output vector (y1, y2, …, y n );||XC i || is the vector XC i The norm of ω ij is the weight from the i-th hidden layer node to the j-th output.

6. The method according to claim 1, characterized in that The method of correcting the initial prediction value by using the prediction deviation value to obtain the prediction value of the dynamic parameter of the volatile organic compound during the combustion process includes: Obtaining the sum of the prediction deviation value and the initial prediction value; The sum of the prediction deviation value and the initial prediction value is determined as the prediction value of the dynamic parameter.

7. The method according to claim 1, characterized in that The dynamic parameters include bed temperature distribution and exhaust gas conversion rate; the predicted value of the dynamic parameter does not meet the preset conditions including: the predicted value of the bed temperature distribution is greater than the preset temperature threshold, or the predicted value of the exhaust gas conversion rate is less than the preset conversion threshold.

8. A combustion control system for volatile organic compounds, characterized in that: The system includes a catalytic combustion reactor and a computer device; The computer device is used to obtain the intake air flow rate, intake air concentration and reversing period of the catalytic combustion reactor when the volatile organic compounds are burned in the catalytic combustion reactor; and input the intake air flow rate, the intake air concentration and the reversing period into the mechanism model and the error compensation model respectively to obtain the initial predicted value and predicted deviation value of the dynamic parameter of the volatile organic compounds during the combustion process; then correct the initial predicted value by the predicted deviation value to obtain the predicted value of the dynamic parameter of the volatile organic compounds during the combustion process, and finally generate a control signal when the predicted value of the dynamic parameter does not meet the preset condition; the control signal is used to adjust the state of the catalytic combustion reactor; The mechanism model is a group of partial differential control equations based on mass conservation and energy conservation, which is established according to the dynamic nonlinear interactive process of interphase flow, heat transfer, mass transfer and catalytic reaction in the catalytic combustion reactor, and is used to describe the combustion process of the volatile organic matter in the catalytic combustion reactor; the error compensation model is an online real-time corrected machine learning algorithm constructed by learning historical data through a radial basis function neural network.

9. The system according to claim 8, characterized in that The system also includes a reversing control valve, a gas concentration online measuring device, a gas flow meter and a temperature measuring device; the gas concentration online measuring device is respectively installed on the inlet and outlet pipes of the catalytic combustion reactor, the gas flow meter is installed on the inlet and outlet pipe positions of the catalytic combustion reactor, and the temperature measuring device is installed on the key temperature measuring point position of the catalytic combustion reactor; the reversing control valve is installed at the corresponding position of the system; The gas concentration online measuring device is used to measure the gas concentration at the inlet and outlet of the catalytic combustion reactor and transmit the concentration signal to the computer device; The gas flow meter is used to measure the inlet and outlet gas concentrations of the catalytic combustion reactor and transmit the gas flow signal to the computer device; The temperature measuring device is used to measure the temperature distribution signal in the bed layer and transmit the temperature distribution signal in the bed layer to the computer device; The reversing control valve is used to change the gas flow direction in the catalytic combustion reactor by adjusting the flow direction of the volatile organic matter gas, and transmit the reversing signal to the computer device.

10. A combustion control device for volatile organic compounds, characterized in that: The device comprises: An acquisition module, used to acquire the intake air flow rate, intake air concentration and reversing period of the catalytic combustion reactor when volatile organic matter is burned in the catalytic combustion reactor; An input module is used to input the intake air flow rate, the intake air concentration and the reversing period into a mechanism model to obtain an initial prediction value of the dynamic parameters of the volatile organic matter during the combustion process; the mechanism model is a group of partial differential control equations based on mass conservation and energy conservation established according to the dynamic nonlinear interactive process of interphase flow, heat transfer, mass transfer and catalytic reaction in a catalytic combustion reactor, so as to describe the combustion process of the volatile organic matter in the catalytic combustion reactor; An analysis module, used to analyze the intake air flow rate, the intake air concentration and the commutation period through an error compensation model to obtain a predicted deviation value of a dynamic parameter; the error compensation model is an online real-time corrected machine learning algorithm constructed by learning historical data through a radial basis function neural network; A correction module, used to correct the initial prediction value by using the prediction deviation value to obtain a prediction value of a dynamic parameter of the volatile organic compound during the combustion process; A control module is used to generate a control signal when the predicted value of the dynamic parameter does not meet a preset condition; the control signal is used to adjust the state of the catalytic combustion reactor.

Images