CN111871164B - Intelligent regulation and control method for low-temperature plasma catalytic degradation of CVOCs - Google Patents
Intelligent regulation and control method for low-temperature plasma catalytic degradation of CVOCs Download PDFInfo
- Publication number
- CN111871164B CN111871164B CN202010542377.5A CN202010542377A CN111871164B CN 111871164 B CN111871164 B CN 111871164B CN 202010542377 A CN202010542377 A CN 202010542377A CN 111871164 B CN111871164 B CN 111871164B
- Authority
- CN
- China
- Prior art keywords
- cvocs
- low
- conversion rate
- temperature
- neural network
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B01—PHYSICAL OR CHEMICAL PROCESSES OR APPARATUS IN GENERAL
- B01D—SEPARATION
- B01D53/00—Separation of gases or vapours; Recovering vapours of volatile solvents from gases; Chemical or biological purification of waste gases, e.g. engine exhaust gases, smoke, fumes, flue gases, aerosols
- B01D53/32—Separation of gases or vapours; Recovering vapours of volatile solvents from gases; Chemical or biological purification of waste gases, e.g. engine exhaust gases, smoke, fumes, flue gases, aerosols by electrical effects other than those provided for in group B01D61/00
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B01—PHYSICAL OR CHEMICAL PROCESSES OR APPARATUS IN GENERAL
- B01D—SEPARATION
- B01D53/00—Separation of gases or vapours; Recovering vapours of volatile solvents from gases; Chemical or biological purification of waste gases, e.g. engine exhaust gases, smoke, fumes, flue gases, aerosols
- B01D53/34—Chemical or biological purification of waste gases
- B01D53/46—Removing components of defined structure
- B01D53/68—Halogens or halogen compounds
- B01D53/70—Organic halogen compounds
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B01—PHYSICAL OR CHEMICAL PROCESSES OR APPARATUS IN GENERAL
- B01D—SEPARATION
- B01D53/00—Separation of gases or vapours; Recovering vapours of volatile solvents from gases; Chemical or biological purification of waste gases, e.g. engine exhaust gases, smoke, fumes, flue gases, aerosols
- B01D53/34—Chemical or biological purification of waste gases
- B01D53/74—General processes for purification of waste gases; Apparatus or devices specially adapted therefor
- B01D53/86—Catalytic processes
- B01D53/8659—Removing halogens or halogen compounds
- B01D53/8662—Organic halogen compounds
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B01—PHYSICAL OR CHEMICAL PROCESSES OR APPARATUS IN GENERAL
- B01D—SEPARATION
- B01D2259/00—Type of treatment
- B01D2259/80—Employing electric, magnetic, electromagnetic or wave energy, or particle radiation
- B01D2259/818—Employing electrical discharges or the generation of a plasma
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A50/00—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE in human health protection, e.g. against extreme weather
- Y02A50/20—Air quality improvement or preservation, e.g. vehicle emission control or emission reduction by using catalytic converters
Landscapes
- Chemical & Material Sciences (AREA)
- Engineering & Computer Science (AREA)
- Analytical Chemistry (AREA)
- General Chemical & Material Sciences (AREA)
- Oil, Petroleum & Natural Gas (AREA)
- Chemical Kinetics & Catalysis (AREA)
- Environmental & Geological Engineering (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Exhaust Gas Treatment By Means Of Catalyst (AREA)
- Physical Or Chemical Processes And Apparatus (AREA)
Abstract
The invention discloses an intelligent regulation and control method for low-temperature plasma catalytic degradation of CVOCs. The invention utilizes an artificial neural network model to establish main process parameters of two-stage low-temperature plasma catalysis, CVOCs conversion rate and CO 2 The selective mapping relation is realized by optimizing the set values of the process parameters, accurately regulating and controlling the ozone concentration and the pollutant concentration of the plasma discharge unit, and synergistically optimizing the process conditions of the ozone catalytic oxidation unit, so that CO is converted efficiently while CVOCs are converted efficiently 2 The selectivity reaches a higher level. In addition, the invention solves the problem that the two-stage low-temperature plasma catalysis is difficult to simultaneously meet the requirements of high conversion rate and high CO 2 The selectivity problem can be flexibly adjusted and optimized according to different treatment processes and emission requirements, and the method is suitable for regulating and controlling the low-temperature plasma catalytic degradation process of various CVOCs (chemical vapor deposition) such as aromatic chlorides, non-aromatic chlorides and polymeric chlorides.
Description
Technical Field
The invention belongs to the technical field of catalysis, and particularly relates to an intelligent regulation and control method for low-temperature plasma catalytic degradation of CVOCs.
Background
Volatile Organic Compounds (VOCs) are Organic Compounds having a boiling point of 50 to 260 ℃ at normal pressure and a saturated vapor pressure of more than 133.32Pa at room temperature, and include aromatic hydrocarbons such as ethyl acetate, chlorobenzene, xylene and styrene, and pollutants such as aldehydes and esters. chlorine-Containing Volatile Organic Compounds (CVOCs) are typical pollutants discharged in industrial production such as petroleum industry, coating industry, chemical synthesis pharmaceutical projects and the like, are various in types, complex in components, easy to accumulate, and extremely high in toxicity and biological degradation resistance, and become a great problem in treatment of VOCs at present.
The Low-Temperature Plasma (LTP) technology is a processing method which is not affected by the toxicity of CVOCs, high-energy electrons generated by Plasma discharge at Low Temperature and normal pressure have inelastic collision with the CVOCs to enable the CVOCs to carry out cracking reaction, and active particles such as free radicals, excited state atoms, molecules and the like with high activity are generated through reactions such as ionization, dissociation and the like, so that the CVOCs are further degraded in an oxidation mode. Compared with CVOCs treatment technologies such as catalytic combustion, photocatalysis and biodegradation, the low-temperature plasma can be operated at normal temperature and normal pressure, is simple in device, small in occupied area and strong in particulate matter interference resistance, can be opened and closed at any time, and does not need pretreatment on oil mist, oil smoke and the like. However, LTP technology is associated with CVOCs degradation efficiency and deleterious byproduct reduction (O) 3 NOx and other carbonaceous by-products that are not fully degraded) still need to be further improved. Post-positioned low-temperature Plasma Catalysis (PPC) is used as a combination mode in Plasma Catalysis, so that CVOCs can pass through an LTP device first and undergo complex homogeneous reaction to complete the pre-degradation treatment of the CVOCs, and then high-energy active substances generated by the LTP device synergistically catalyze and degrade the CVOCs at low temperature (normal temperature or less than or equal to 150 ℃), the requirement of catalytic reaction on temperature is reduced, the degradation efficiency of the CVOCs is remarkably improved, and the method is a green technology suitable for CVOCs treatment.
In the field of plasma catalysis, pollutant conversion rate and CO 2 Selectivity is generally non-linearly dependent, and is primarily influenced by many factors, both plasma process parameters and reaction conditions, as well as catalyst type and activity, of the pre-LTP degradation treatment. Aiming at different practical treatment requirements, the conversion efficiency and CO of CVOCs are adjusted 2 The selective double-position regulation and control is carried out, and how to control the power supply parameters (discharge voltage and frequency) and the carrier gas background (relative humidity and O) of the plasma unit 2 /N 2 ) Gas flow rate, etc., and catalyst temperature, ozone concentration, intake air contaminant concentration, etc., of the low temperature catalytic unitConditions for efficient conversion of contaminants while CO is being produced 2 The selectivity reaches a higher level, and is one of the main problems facing the PPC technology to process CVOCs. At present, the power supply parameters of the PPC technology are often adjusted according to the air inlet volume and the pollutant concentration in the practical application process, the technological parameter values are generally determined by historical experience or debugging through condition experiments, and flexible adjustment is difficult to be made according to specific treatment requirements.
Disclosure of Invention
The invention aims to provide an intelligent regulation and control method for low-temperature plasma catalytic degradation of CVOCs (chemical vapor deposition) based on numerical simulation of a neural network, and a PPC (pentatricopeptide repeats) degradation test is carried out by taking chlorobenzene as a typical CVOCs to respectively obtain chlorobenzene conversion rate and CO 2 The optimized value of each process parameter when the selectivity reaches the maximum.
The method comprises the following specific steps:
step one, performing a multivariable experiment in a PPC system; the PPC system adopts a two-stage reaction device which is composed of a plasma discharge unit at the front end and a low-temperature catalytic oxidation unit at the rear end; each set of experiments varied the voltage, frequency, relative humidity, O, of the plasma discharge cell with the inlet CVOCs concentration known 2 /N 2 And n discharge process parameters in the gas flow rate, wherein n is more than or equal to 1 and less than or equal to 5. Respectively monitoring and recording the intermediate CVOCs conversion rate and ozone concentration in multivariable experiments and the CVOCs conversion rate and CO at the output end of the low-temperature catalytic oxidation unit 2 And (4) selectivity. After multivariable experiments, obtaining different n discharge process parameters and corresponding intermediate CVOCs conversion rate, ozone concentration, terminal CVOCs conversion rate and CO 2 And optionally, forming a training set. The intermediate CVOCs conversion rate is the CVOCs conversion rate of the gas outlet end of the plasma discharge unit.
Step two, constructing a first BP neural network which takes n discharge process parameters as input variables and takes the intermediate CVOCs conversion rate and the ozone concentration as output variables; the first BP neural network is trained using a training set.
Step three, constructing a model with the intermediate CVOCs conversion rate and the ozone concentration as input variables and the terminal CVOCs conversion rate and CO 2 Second BP neural selectively being output variableVia a network. The second BP neural network is trained using the training set.
Step four, setting the conversion rate and CO of terminal CVOCs 2 The desire for selectivity. Carrying out grid search on each input variable of the second BP neural network, and predicting the terminal CVOCs conversion rate and the terminal CO corresponding to different intermediate CVOCs conversion rates and ozone concentrations 2 Selectivity; selecting the terminal CVOCs conversion rate and CO which best meet the expectation 2 And (4) selecting the corresponding intermediate CVOCs conversion rate and ozone concentration as optimization parameter results.
Taking the intermediate CVOCs conversion rate and the ozone concentration obtained in the step four as expectations, carrying out grid search on n input variables of the first BP neural network, and predicting the corresponding intermediate CVOCs conversion rate and the ozone concentration under different n discharge process parameters; selecting the intermediate CVOCs conversion rate and the ozone concentration which most meet the expectation; with its corresponding n discharge process parameters.
And step six, performing CVOCs degradation in a PPC system according to the n discharge technological parameters obtained in the step five.
Preferably, the multivariate experiment in step one also changes the catalyst temperature such that the training set contains the catalyst temperature. The input variables of the second BP neural network constructed in step three also include catalyst temperature. The fourth step also obtains the conversion rate of terminal CVOCs and CO 2 The selectivity most closely corresponds to the desired catalyst temperature. And adjusting the temperature of the catalyst in the PPC system in the sixth step to the value obtained in the fourth step.
Preferably, the catalyst in the low-temperature catalytic oxidation unit is one or more of a metal catalyst, a metal oxide catalyst and a metal organic framework compound. In the first step, the temperature of the catalyst is adjusted by an electric heating wire in the low-temperature catalytic oxidation unit.
Preferably, the CVOCs to be degraded are chlorobenzene; the number of discharge process parameters n =5;
preferably, the voltage, frequency, relative humidity, and O are set in consideration of the maximum chlorobenzene conversion 2 /N 2 The gas flow rate and the catalyst temperature were 85kv, 10kHZ, 9%, 120L/min, 65 ℃, at which time the chlorobenzene conversion is 99.8%, CO 2 The selectivity was 43.5%.
In consideration of CO 2 Setting voltage, frequency, relative humidity, and O under the condition of maximum selectivity 2 /N 2 The gas flow rate and catalyst temperature were 74kv, 1.5kHZ, 13%, 1.4, 10L/min, 45 ℃ respectively, at which time CO 2 The selectivity was 69.6% and the chlorobenzene conversion was 93.5%.
In the comprehensive consideration of chlorobenzene conversion and CO 2 Under the selective condition, the voltage is set to be 78-82 kv, the frequency is set to be 2-3 kHZ, the relative humidity is set to be 10-12 percent, and O 2 /N 2 Is 1, 4.3-1, the gas flow rate is 12-15L/min, the catalyst temperature is 45-50 ℃, the chlorobenzene conversion rate under the optimized condition is 98-99%, and the CO content is as follows 2 The selectivity is 60-65%.
Preferably, the first BP neural network adopts a three-layer BP neural network model with a topological structure of n-11-2. The second BP neural network adopts a three-layer BP neural network model with a topological structure of 3-7-2.
Preferably, the CVOCs to be degraded are dichloromethane; the number of discharge process parameters n =3; the 3 discharge process parameters are the voltage, the frequency and the relative humidity of the plasma discharge unit respectively.
Preferably, the topology of the first BP neural network is 3. The topology of the second BP neural network is 3. The third output variable of the second BP neural network is the energy efficiency of dichloromethane degradation.
Preferably, in the second and third steps, the first BP neural network and the second BP neural network are trained by MATLAB software. The training function of the first BP neural network is a rainlm function, and the learning rate is 0.1. The training function of the second BP neural network is a slingda function, and the learning rate is 0.1.
Preferably, in step six, the conversion of CVOCs, CO if terminal 2 The selectivity does not meet the expectation set in the fourth step; and D, reselecting an optimized parameter result in the data set obtained in the step four, and re-executing the step five to obtain new process parameters.
The invention has the beneficial effects that:
1. the invention utilizes an artificial neural network model to establish main process parameters of two-stage low-temperature plasma catalysis, CVOCs conversion rate and CO 2 The selective mapping relation is realized by optimizing the set values of the process parameters, accurately regulating and controlling the ozone concentration and the pollutant concentration of the plasma discharge unit, and synergistically optimizing the process conditions of the ozone catalytic oxidation unit, so that CO is converted efficiently while CVOCs are converted efficiently 2 The selectivity reaches a higher level.
2. The invention can flexibly adjust and optimize expectation aiming at different treatment processes and emission requirements, and can meet the requirements of maximum conversion rate of CVOCs and CO 2 The selectivity is maximum, and the regulation and control targets are considered.
3. The method is suitable for regulating and controlling the low-temperature plasma catalytic degradation process of various CVOCs (chemical vapor deposition) such as aromatic hydrocarbon chlorides (such as chlorobenzene and polychlorinated chlorobenzene), non-aromatic hydrocarbon chlorides (such as dichloromethane and trichloroethylene), polymeric chlorides (such as polychlorinated biphenyl) and the like, and has strong adaptability to different industrial application scenes.
Drawings
FIG. 1 is a schematic diagram of the reaction process of CVOCs in a PPC system;
fig. 2 is a schematic diagram of a coupling hierarchy of a first BP and a second BP in embodiment 1 of the present invention;
FIG. 3 is a graph comparing predicted and measured values of intermediate chlorobenzene conversion in example 1 of the present invention;
FIG. 4 is a graph comparing the predicted and measured values of ozone concentration in example 1 of the present invention;
FIG. 5 is a graph comparing predicted and measured values of terminal chlorobenzene conversion in example 1 of the present invention;
FIG. 6 shows terminal CO in example 1 of the present invention 2 The prediction of selective ozone concentration is plotted against the test value.
Detailed Description
The invention is further described below with reference to the accompanying drawings.
Example 1
Intelligent regulation and control of low-temperature plasma catalytic degradation CVOCsThe method comprises the step of degrading the CVOCs by a Post-positioned low-temperature Plasma Catalysis system (Post-positioned Plasma Catalysis system) through a PPC system. The reaction process of CVOCs in the PPC system is shown in figure 1, and the two-stage reaction device is composed of a plasma discharge unit (LTP device) at the front end and a low-temperature catalytic oxidation unit at the rear end; in the plasma discharge unit, oxygen-containing carrier gas excites high-activity substances such as high-energy electrons, oxygen atoms, hydroxyl radicals, ozone and the like under the condition of dielectric barrier discharge of 40-100 kv, so that the CVOCs which are difficult to degrade are rapidly oxidized into intermediate products such as alcohols, phenols, carboxylic acids, chloralkanes and the like. The products enter a rear-end catalytic oxidation unit together with ozone and are further oxidized into CO through catalytic oxidation reaction of the ozone 2 、H 2 O, HCl and other small molecular substances.
The catalyst in the low-temperature catalytic oxidation unit adopts one or more of metal catalyst, metal oxide catalyst and metal organic framework compounds (MOFs) used at normal temperature. The plasma discharge unit adopts dielectric barrier discharge, and the working voltage is 220V.
In the process of catalyzing and degrading chlorobenzene by using low-temperature plasma, when only considering that the chlorobenzene conversion rate is maximum, the voltage, the frequency, the relative humidity and the O are set 2 /N 2 The gas flow rate and the catalyst temperature were 85kv, 10kHZ, 9%, 1.1, 20L/min, 65 ℃ respectively, at which time the chlorobenzene conversion was 99.8%, CO 2 The selectivity was 43.5%.
The low-temperature plasma catalytic degradation process of chlorobenzene only considers CO 2 When the selectivity is maximum, setting voltage, frequency, relative humidity, O 2 /N 2 The gas flow rate and catalyst temperature were 74kv, 1.5kHZ, 13%, 1.4, 10L/min, 45 ℃ respectively, at which time CO 2 The selectivity was 69.6% and the chlorobenzene conversion was 93.5%.
The low-temperature plasma catalytic degradation process of chlorobenzene comprehensively considers the chlorobenzene conversion rate and CO 2 In the selectivity, the voltage is set to be 78-82 kv, the frequency is set to be 2-3 kHZ, the relative humidity is set to be 10-12 percent, and O is set to be 2 /N 2 Is 1The chlorobenzene conversion rate can reach 98-99 percent, and CO can reach 2 The selectivity can reach 60-65%.
The intelligent regulation and control method for the low-temperature plasma catalytic degradation of CVOCs comprises the following steps:
step one, designing and implementing a multivariable experiment; each set of experiments simultaneously changes the voltage, the frequency, the relative humidity and the O of the plasma discharge unit under the premise that the concentration of the inlet CVOCs is known and basically stable 2 /N 2 And gas flow rate, and the temperature of the catalyst in the low-temperature catalytic oxidation unit is changed by the electric heating wire. In the test, the chlorobenzene conversion rate and the ozone concentration of the gas outlet end of the plasma discharge unit and the chlorobenzene conversion rate and CO of the output end of the low-temperature catalytic oxidation unit are respectively monitored and recorded 2 And (4) selectivity.
After multivariable experiments, the voltage, frequency, relative humidity and O of different plasma discharge units are obtained 2 /N 2 Gas flow rate, catalyst temperature in the low-temperature catalytic oxidation unit and chlorobenzene conversion rate and ozone concentration of the gas outlet end of the corresponding plasma discharge unit, chlorobenzene conversion rate and CO of the output end of the low-temperature catalytic oxidation unit 2 Selective, forming training sets of neural networks
Step two, constructing voltage, frequency, relative humidity and O of the plasma discharge unit 2 /N 2 The gas flow rate is five input variables, and the chlorobenzene conversion rate and the ozone concentration at the gas outlet end of the plasma discharge unit are two output variables; the first BP neural network was trained in MATLAB software using the corresponding data in the training set.
The first BP neural network adopts a three-layer BP neural network model with a topological structure of 5-11-2, and comprises an output layer, a hidden layer and an output layer; the training function is the rainlm function, and the learning rate is 0.1.
Step three, constructing three input variables of chlorobenzene conversion rate and ozone concentration at the gas outlet end of the plasma discharge unit and catalyst temperature in the low-temperature catalytic oxidation unit, and constructing chlorobenzene conversion rate and CO at the output end of the low-temperature catalytic oxidation unit 2 A second BP neural network that is selectively two output variables. Use ofThe corresponding data in the training set was used to train the second BP neural network in MATLAB software.
The second BP neural network adopts a three-layer BP neural network model with a topological structure of 3-7-2, and comprises an output layer, a hidden layer and an output layer; the training function is the slingda function, and the learning rate is 0.1.
Step four, setting the chlorobenzene conversion rate and CO of the tail end (low-temperature catalytic oxidation unit) according to the actual treatment requirement 2 A desired value or range of selectivity; for example: with maximum conversion of terminal chlorobenzene to desired or terminal CO 2 The selectivity is at most desirable.
Grid searching is carried out on three input variables of the second BP neural network, and chlorobenzene conversion rate and CO of the corresponding tail end (low-temperature catalytic oxidation unit) at the gas outlet end and ozone concentration of different plasma discharge units and the temperature of the catalyst in the low-temperature catalytic oxidation unit are predicted 2 Selectivity; conversion of chlorobenzene at each end obtained, CO 2 Selectivity values (each prediction yielded a set of chlorobenzene conversions, CO 2 Selectivity value), the most desirable terminal chlorobenzene conversion, CO, is selected 2 Selectivity; and the chlorobenzene conversion rate and the ozone concentration of the gas outlet end of the corresponding plasma discharge unit and the temperature of the catalyst in the low-temperature catalytic oxidation unit are used as optimization parameter results.
Step five, taking the chlorobenzene conversion rate and the ozone concentration of the gas outlet end of the plasma discharge unit obtained in the step four as expected values, carrying out grid search on five input variables of the first BP neural network, and predicting different voltages, frequencies, relative humidity and O 2 /N 2 Chlorobenzene conversion rate and ozone concentration at the gas outlet end of the corresponding plasma discharge unit at the gas flow rate; selecting chlorobenzene conversion rate and ozone concentration which best meet the expectation from the obtained chlorobenzene conversion rate and ozone concentration values (which are as close as possible to the expectation value in the step); at its corresponding voltage, frequency, relative humidity, O 2 /N 2 Gas flow rate as a result of the optimization parameters.
Step six, the low-temperature catalytic oxidation unit obtained in the step four is usedCatalyst temperature, voltage, frequency, relative humidity, O obtained in step five 2 /N 2 Taking the gas flow rate as a process parameter, and carrying out CVOCs degradation in a PPC system; if terminal chlorobenzene conversion, CO 2 The selectivity does not meet the expectation set in the fourth step; then the optimized parameter results are reselected from the data set obtained in step four and step five is re-executed to obtain new process parameters (catalyst temperature, voltage, frequency, relative humidity, O) 2 /N 2 Gas flow rate).
The method comprises the steps of constructing a neural network model with a first BP and a second BP coupled; when the expected requirements for CVOCs processing are changed, the process parameters meeting the requirements can be obtained only by executing the fourth step and the fifth step again, so that the processing effect of the PPC system is greatly improved, and the debugging process is simplified.
Chlorobenzene is taken as a target pollutant, and the initial chlorobenzene concentration is kept at 250 +/-10 ppm; the PPC system adopts dielectric barrier discharge, and the working voltage is 220V; the catalyst is copper manganese oxide (CuMnOx) prepared by an oxidation-reduction precipitation method, and is fixed in a dielectric barrier discharge device by quartz wool. Setting the investigation range of each process parameter, wherein the discharge voltage is 40-100 kv, the frequency is 1-10 kHZ, the relative humidity is 0-50%, and O 2 /N 2 The method is characterized by comprising the following steps of 1. Wherein the relative humidity is monitored and controlled by a dehumidifier, O 2 /N 2 The temperature of the catalyst is controlled by an electric heating wire.
Performing a grid search test within the set range of the variables to obtain multiple variables (voltage, frequency, relative humidity, etc.), O2 /N 2 And gas flow rate) under the synchronous change condition, constructing a three-layer BP neural network model (first BP) with 5 inputs and 2 outputs, wherein the number of neurons in an implicit layer under the optimal prediction precision is 11, a training function is a train lm function, and the learning rate is 0.1. The catalyst temperature and two output variables (ozone concentration and chlorobenzene concentration at the gas outlet end of the plasma unit) of the first BP are used as input, and PPC is used for loadingChlorobenzene conversion and CO at export 2 And selectively taking the selection as an output, constructing a three-layer BP neural network model (second BP) with a topological structure of 3-7-2, wherein the number of hidden layer neurons under the optimal prediction precision is 7, the training function is a thindda function, and the learning rate is 0.1.
The schematic coupling hierarchical structure of the first BP and the second BP is shown in FIG. 2, and two prediction models respectively establish the relation between the influence factors of the plasma unit and the ozone catalytic oxidation unit and the treatment efficiency.
Randomly selecting 20 different process conditions to verify the prediction performance of the first BP and the second BP, wherein the prediction results and the test results are shown in figure 3, the prediction accuracy of the total four output variables of the first BP and the second BP neural network is high, and the coefficient R is determined 2 All reach above 0.99. The highest chlorobenzene conversion rate and CO at the outlet of the PPC device respectively 2 The selectivity is at most desirable, and a grid search is performed by means of a second BP prediction model. The results showed that the chlorobenzene conversion at the outlet of the PPC unit reached a maximum of 99.8% when the ozone concentration, chlorobenzene concentration and catalyst temperature at the outlet of the plasma unit were 920ppm,35ppm and 65 ℃ respectively, and that the CO at this time was 2 The selectivity is 43.5%, and the corresponding voltage, frequency, relative humidity and O can be obtained by carrying out grid search through the first BP prediction model by taking the ozone concentration and the chlorobenzene concentration as expectation 2 /N 2 Gas flow rates of 85kv, 10kHZ, 9%, 1; when the ozone concentration, chlorobenzene concentration and catalyst temperature at the gas outlet end of the plasma unit are 650ppm,25ppm and 45 ℃ respectively, CO at the outlet of the PPC device 2 The selectivity reaches the maximum value of 69.6 percent, the chlorobenzene conversion rate is 93.5 percent, the ozone concentration and the chlorobenzene concentration are taken as expectations, and the grid search is carried out through the first BP prediction model to obtain corresponding voltage, frequency, relative humidity and O 2 /N 2 The gas flow rates are respectively 74kv, 1.5kHZ, 13%, 1; when the chlorobenzene conversion rate and CO are comprehensively considered 2 In the case of selectivity, the voltage is 78-82 kv, the frequency is 2-3 kHZ, the relative humidity is 10-12%, and O 2 /N 2 1, 4.3-1, gas flow rate of 12-15L/min, catalyst temperature of 45-50 DEG CThe chlorobenzene conversion rate under the optimized condition can reach 98-99 percent, and CO can reach 2 The selectivity can reach 60-65%. Respectively optimize the chlorobenzene conversion rate and CO 2 And carrying out experimental verification on the prediction scheme with the optimal selectivity, wherein the result is shown in the following table, and the verification result shows that the predicted optimized value of the intelligent regulation and control method is basically consistent with the actual test value.
table-Process parameter Regulation and control schemes under different optimization objectives and verification results
Example 2
The present example simulates the actual industrial waste gas discharge conditions, uses dichloromethane as the target pollutant, and sets the concentration of dichloromethane in the waste gas to be 1g/m 3 (about 260 ppm), the amount of air introduced into the exhaust gas treatment apparatus was maintained at 100m 3 H is used as the reference value. The process was carried out using a dielectric barrier discharge PPC unit, and the catalyst packed in the catalytic oxidation unit was a copper manganese oxide catalyst (CuMnOx) as described in example 1.
The intelligent control method for processing dichloromethane by the PPC technology in the embodiment is similar to that in embodiment 1, except that: because the background and the air volume of the carrier gas of the industrial waste gas are determined by the production process and are not regulated and controlled generally, the gas flow rate and the O 2 /N 2 Not as input variables to the neural network. Furthermore, considering the economic requirements of the treatment process from the actual engineering, the energy efficiency of the degradation of methylene chloride is one of the optimization objectives. After parameter tuning, the topology of the first BP neural network is 3. The topology of the second BP neural network is 3. The rest conditions are the same. By the intelligent regulation and control method for neural network modeling prediction, energy efficiency, dichloromethane conversion rate and CO are comprehensively considered 2 The optimized regulation scheme for selectively obtaining the catalyst has the voltage of 75kv, the frequency of 1.5kHZ, the relative humidity of 16 percent, the temperature of the catalyst of 50 ℃, and the energy efficiency measured under the condition of the process parameters of 0.713 g/kw.h -1 The conversion of methylene chloride was 99.0%, CO 2 The selectivity was 76%.
Claims (4)
1. An intelligent regulation and control method for low-temperature plasma catalytic degradation of CVOCs is characterized in that: step one, performing a multivariable experiment in a PPC system; the PPC system adopts a two-stage reaction device which is composed of a plasma discharge unit at the front end and a low-temperature catalytic oxidation unit at the rear end; each set of experiments varied the voltage, frequency, relative humidity, O, of the plasma discharge cell with the inlet CVOCs concentration known 2 /N 2 And n discharge process parameters in the gas flow rate, wherein n is more than or equal to 3 and less than or equal to 5; and changing the catalyst temperature of the low-temperature catalytic oxidation unit;
if the degraded CVOCs are chlorobenzene; the number of the discharge process parameters n =5; the first BP neural network adopts a three-layer BP neural network model with a topological structure of 5-11-2; the second BP neural network adopts a three-layer BP neural network model with a topological structure of 3-7-2;
if the degraded CVOCs are dichloromethane; the number of discharge process parameters n =3; the 3 discharge process parameters are respectively the voltage, the frequency and the relative humidity of the plasma discharge unit; the topology of the first BP neural network is 3; the topology of the second BP neural network is 3; the third output variable of the second BP neural network is the energy efficiency of dichloromethane degradation;
respectively monitoring and recording the intermediate CVOCs conversion rate and ozone concentration in multivariable experiments and the CVOCs conversion rate and CO at the output end of the low-temperature catalytic oxidation unit 2 Selectivity; after multivariable experiments, obtaining different n discharge process parameters, catalyst temperature in the low-temperature catalytic oxidation unit and corresponding intermediate CVOCs conversion rate, ozone concentration, CVOCs conversion rate at the output end of the low-temperature catalytic oxidation unit, and CO 2 Selectively, forming a training set; the intermediate CVOCs conversion rate is the CVOCs conversion rate of the gas outlet end of the plasma discharge unit;
step two, constructing a first BP neural network which takes n discharge process parameters as input variables and takes the intermediate CVOCs conversion rate and the ozone concentration as output variables; training a first BP neural network using a training set;
step three, constructing a CVOCs conversion rate and CO conversion rate of the output end of the low-temperature catalytic oxidation unit by taking the intermediate CVOCs conversion rate, the ozone concentration and the catalyst temperature in the low-temperature catalytic oxidation unit as input variables 2 A second BP neural network which is selectively an output variable; training a second BP neural network using the training set;
step four, setting the CVOCs conversion rate and CO at the output end of the low-temperature catalytic oxidation unit 2 The desire for selectivity; grid search is carried out on each input variable of the second BP neural network, and the CVOCs conversion rate and CO at the output end of the low-temperature catalytic oxidation unit corresponding to different intermediate CVOCs conversion rates, ozone concentrations and catalyst temperatures in the low-temperature catalytic oxidation unit are predicted 2 Selectivity; selecting the CVOCs conversion rate and CO at the output end of the low-temperature catalytic oxidation unit which most meets the expectation 2 The selectivity is taken as the result of the optimization parameters by the corresponding intermediate CVOCs conversion rate, the ozone concentration and the catalyst temperature in the low-temperature catalytic oxidation unit;
step five, taking the intermediate CVOCs conversion rate and the ozone concentration obtained in the step four as expectations, carrying out grid search on n input variables of the first BP neural network, and predicting the intermediate CVOCs conversion rate and the ozone concentration corresponding to n different discharge process parameters; selecting the intermediate CVOCs conversion rate and the ozone concentration which most meet the expectation; n discharge process parameters corresponding to the discharge process parameters;
step six, performing CVOCs degradation on the catalyst temperature in the low-temperature catalytic oxidation unit obtained in the step four and the n discharge technological parameters obtained in the step five in a PPC system; if the CVOCs conversion rate and CO at the output end of the low-temperature catalytic oxidation unit 2 The selectivity does not meet the expectation set in the fourth step; and D, reselecting an optimized parameter result in the data set obtained in the step four, and re-executing the step five to obtain new process parameters.
2. The intelligent regulation and control method for low-temperature plasma catalytic degradation (CVOCs) according to claim 1, wherein the method comprises the following steps: the catalyst in the low-temperature catalytic oxidation unit is one or more of a metal catalyst, a metal oxide catalyst and a metal organic framework compound; in the first step, the temperature of the catalyst is adjusted by an electric heating wire in the low-temperature catalytic oxidation unit.
3. The intelligent regulation and control method for low-temperature plasma catalytic degradation (CVOCs) according to claim 1, wherein the method comprises the following steps: setting voltage, frequency, relative humidity, O in consideration of maximum chlorobenzene conversion 2 /N 2 The gas flow rate and the catalyst temperature were 85kv, 10kHZ, 9%, 1.1, 20L/min, 65 ℃ respectively, at which time the chlorobenzene conversion was 99.8%, CO 2 The selectivity was 43.5%;
in consideration of CO 2 Setting voltage, frequency, relative humidity, and O under the condition of maximum selectivity 2 /N 2 The gas flow rate and catalyst temperature were 74kv, 1.5kHZ, 13%, 1.4, 10L/min, 45 ℃ respectively, at which time CO 2 The selectivity is 69.6 percent, and the chlorobenzene conversion rate is 93.5 percent;
in the comprehensive consideration of chlorobenzene conversion and CO 2 Under the selective condition, setting the voltage to be 78-82kv, the frequency to be 2-3kHZ, the relative humidity to be 10-12%, and O 2 /N 2 The ratio of the chlorobenzene to the chlorobenzene is 1:4.3 to 1:4.4, the gas flow rate is 12 to 15L/min, the catalyst temperature is 45 to 50 ℃, the chlorobenzene conversion rate under the optimized conditions is 98 to 99 percent, and the ratio of the chlorobenzene to the chlorobenzene is as follows 2 The selectivity is 60 to 65 percent.
4. The intelligent regulation and control method for low-temperature plasma catalytic degradation (CVOCs) according to claim 1, wherein the method comprises the following steps: in the second step and the third step, the first BP neural network and the second BP neural network are trained through MATLAB software; the training function of the first BP neural network is a trainlm function, and the learning rate is 0.1; the training function of the second BP neural network is a slingda function, and the learning rate is 0.1.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010542377.5A CN111871164B (en) | 2020-06-15 | 2020-06-15 | Intelligent regulation and control method for low-temperature plasma catalytic degradation of CVOCs |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010542377.5A CN111871164B (en) | 2020-06-15 | 2020-06-15 | Intelligent regulation and control method for low-temperature plasma catalytic degradation of CVOCs |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111871164A CN111871164A (en) | 2020-11-03 |
| CN111871164B true CN111871164B (en) | 2023-03-21 |
Family
ID=73157536
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010542377.5A Active CN111871164B (en) | 2020-06-15 | 2020-06-15 | Intelligent regulation and control method for low-temperature plasma catalytic degradation of CVOCs |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111871164B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114177746A (en) * | 2021-07-29 | 2022-03-15 | 浙江树人学院（浙江树人大学） | Plasma-biological drum coupling reactor and method for treating VOCs (volatile organic compounds) by using same |
| CN114405532B (en) * | 2022-03-04 | 2023-12-12 | 浙江工业大学 | Monoatomic catalyst for tar catalytic reforming and preparation parameter optimization method thereof |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108325362A (en) * | 2017-06-02 | 2018-07-27 | 中国石油化工股份有限公司 | A kind of method that low temperature plasma coupled biological method handles VOCs and foul gas |
| CN108479343A (en) * | 2018-06-08 | 2018-09-04 | 威海涌鑫环保工程有限公司 | A kind of Intelligentized control method based on three warehouse formula organic waste gas treatment equipments |
| CN110026068A (en) * | 2019-04-08 | 2019-07-19 | 东南大学 | A kind of large-scale coal fired power plant CO based on Neural network inverse control2Trapping system and feed forward control method |
| CN110173710A (en) * | 2019-05-07 | 2019-08-27 | 西安交通大学 | A kind of energy-saving climatic dynamics dynamic regulation device and method for eliminating plume |
| WO2020088485A1 (en) * | 2018-11-02 | 2020-05-07 | 浙江大学 | Intelligent multi-pollutant ultra-low emission system and global optimization method |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11048842B2 (en) * | 2018-09-19 | 2021-06-29 | Basf Se | Simulation of unit operations of a chemical plant for acid gas removal |
-
2020
- 2020-06-15 CN CN202010542377.5A patent/CN111871164B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108325362A (en) * | 2017-06-02 | 2018-07-27 | 中国石油化工股份有限公司 | A kind of method that low temperature plasma coupled biological method handles VOCs and foul gas |
| CN108479343A (en) * | 2018-06-08 | 2018-09-04 | 威海涌鑫环保工程有限公司 | A kind of Intelligentized control method based on three warehouse formula organic waste gas treatment equipments |
| WO2020088485A1 (en) * | 2018-11-02 | 2020-05-07 | 浙江大学 | Intelligent multi-pollutant ultra-low emission system and global optimization method |
| CN110026068A (en) * | 2019-04-08 | 2019-07-19 | 东南大学 | A kind of large-scale coal fired power plant CO based on Neural network inverse control2Trapping system and feed forward control method |
| CN110173710A (en) * | 2019-05-07 | 2019-08-27 | 西安交通大学 | A kind of energy-saving climatic dynamics dynamic regulation device and method for eliminating plume |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111871164A (en) | 2020-11-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111871164B (en) | Intelligent regulation and control method for low-temperature plasma catalytic degradation of CVOCs | |
| Saleem et al. | Role of CO2 in the conversion of toluene as a tar surrogate in a nonthermal plasma dielectric barrier discharge reactor | |
| CN105521705A (en) | Method for treating organic waste gas through low-temperature plasma concerted catalysis | |
| EP2581127B1 (en) | Method for air purification, preferably of volatile organic compounds | |
| CN108970348B (en) | Low-temperature plasma generator, method for treating pollutants by using low-temperature plasma and application of low-temperature plasma | |
| Kang et al. | Photolysis and photooxidation of typical gaseous VOCs by UV Irradiation: Removal performance and mechanisms | |
| CN107913590B (en) | Tail gas treatment device of chlorohydrination method propylene epoxide device | |
| CN108325362A (en) | A kind of method that low temperature plasma coupled biological method handles VOCs and foul gas | |
| CN211753814U (en) | High concentration hydrogen sulfide and organic waste gas treatment system | |
| Chen et al. | Environment pollutants removal with non-thermal plasma technology | |
| CN111467954A (en) | Device and method for catalyzing and degrading VOCs (volatile organic compounds) by low-temperature plasma and ultraviolet light | |
| CN101322916B (en) | Sulfur-containing compound gas cleaning filter medium device and method as well as used catalyst adsorption material thereof | |
| CN105664631B (en) | The fume purifier of fume of microwave plasma torch | |
| Wan et al. | Effect of gas components and particulate matter on the conversion of nitric oxide by dielectric barrier discharge | |
| CN207102266U (en) | A kind of low temperature plasma Coupling Adsorption method processing VOCs and foul gas device | |
| CN114682064B (en) | SF (sulfur hexafluoride) 6 Method for degrading waste gas by radio frequency discharge | |
| CN108970363B (en) | Interval type low-temperature plasma generator filled with ozone decomposer | |
| CN107081036B (en) | Ultraviolet activated ozone oxidation VOCs dust removal purification device and method | |
| CN108325360B (en) | Interval type low-temperature plasma generator | |
| CN107899402B (en) | Tail gas treatment method for chlorohydrination method propylene oxide device | |
| Ikaunieks et al. | Non-thermal plasma for VOC treatment in flue gases | |
| CN114741966A (en) | Grading optimization control method for organic pollutant production and treatment synergy of reactive distillation | |
| CN109185902B (en) | Gas-phase organic matter catalytic combustion method and device based on two-dimensional zeolite catalyst | |
| CN210752007U (en) | Injection type low-temperature plasma integrated system | |
| KR100339849B1 (en) | De-NOx or De-SOx and Simultaneous De-NOx and De-SOx Photocatalysis and Plasma Hybrid System |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |