WO2022095534A1 - 一种预测火电厂氨逃逸的方法 - Google Patents
一种预测火电厂氨逃逸的方法 Download PDFInfo
- Publication number
- WO2022095534A1 WO2022095534A1 PCT/CN2021/112270 CN2021112270W WO2022095534A1 WO 2022095534 A1 WO2022095534 A1 WO 2022095534A1 CN 2021112270 W CN2021112270 W CN 2021112270W WO 2022095534 A1 WO2022095534 A1 WO 2022095534A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- data
- denitration reactor
- outlet
- nox concentration
- inlet
- Prior art date
Links
- QGZKDVFQNNGYKY-UHFFFAOYSA-N Ammonia Chemical compound N QGZKDVFQNNGYKY-UHFFFAOYSA-N 0.000 title claims abstract description 255
- 229910021529 ammonia Inorganic materials 0.000 title claims abstract description 127
- 238000000034 method Methods 0.000 title claims abstract description 28
- 238000012549 training Methods 0.000 claims abstract description 25
- 238000012795 verification Methods 0.000 claims abstract description 7
- UGFAIRIUMAVXCW-UHFFFAOYSA-N Carbon monoxide Chemical compound [O+]#[C-] UGFAIRIUMAVXCW-UHFFFAOYSA-N 0.000 claims description 31
- 239000003546 flue gas Substances 0.000 claims description 31
- 238000013528 artificial neural network Methods 0.000 claims description 27
- 238000006477 desulfuration reaction Methods 0.000 claims description 24
- 230000023556 desulfurization Effects 0.000 claims description 24
- 230000006870 function Effects 0.000 claims description 19
- 238000002347 injection Methods 0.000 claims description 19
- 239000007924 injection Substances 0.000 claims description 19
- 238000012545 processing Methods 0.000 claims description 12
- 238000012544 monitoring process Methods 0.000 claims description 11
- 238000005259 measurement Methods 0.000 claims description 8
- 238000010606 normalization Methods 0.000 claims description 6
- 230000002159 abnormal effect Effects 0.000 claims description 4
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 238000013527 convolutional neural network Methods 0.000 claims description 3
- 230000002068 genetic effect Effects 0.000 claims description 3
- 238000005457 optimization Methods 0.000 claims description 3
- 239000002245 particle Substances 0.000 claims description 3
- 230000000306 recurrent effect Effects 0.000 claims description 3
- 238000012706 support-vector machine Methods 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims description 2
- 238000007781 pre-processing Methods 0.000 abstract 1
- 238000003062 neural network model Methods 0.000 description 7
- 238000000738 capillary electrophoresis-mass spectrometry Methods 0.000 description 5
- 239000011159 matrix material Substances 0.000 description 5
- WWILHZQYNPQALT-UHFFFAOYSA-N 2-methyl-2-morpholin-4-ylpropanal Chemical compound O=CC(C)(C)N1CCOCC1 WWILHZQYNPQALT-UHFFFAOYSA-N 0.000 description 2
- 239000003054 catalyst Substances 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 239000000428 dust Substances 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 241000267854 Brustiarius nox Species 0.000 description 1
- 230000005856 abnormality Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000003638 chemical reducing agent Substances 0.000 description 1
- 238000002485 combustion reaction Methods 0.000 description 1
- 230000009849 deactivation Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000010438 heat treatment Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 230000000246 remedial effect Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000010200 validation analysis Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
- G01N33/0004—Gaseous mixtures, e.g. polluted air
- G01N33/0009—General constructional details of gas analysers, e.g. portable test equipment
- G01N33/0062—General constructional details of gas analysers, e.g. portable test equipment concerning the measuring method, e.g. intermittent, or the display, e.g. digital
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
- G01N33/0004—Gaseous mixtures, e.g. polluted air
- G01N33/0009—General constructional details of gas analysers, e.g. portable test equipment
- G01N33/0027—General constructional details of gas analysers, e.g. portable test equipment concerning the detector
- G01N33/0036—Specially adapted to detect a particular component
- G01N33/0037—Specially adapted to detect a particular component for NOx
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
- G01N33/0004—Gaseous mixtures, e.g. polluted air
- G01N33/0009—General constructional details of gas analysers, e.g. portable test equipment
- G01N33/0027—General constructional details of gas analysers, e.g. portable test equipment concerning the detector
- G01N33/0036—Specially adapted to detect a particular component
- G01N33/0042—Specially adapted to detect a particular component for SO2, SO3
-
- G01N33/0068—
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Definitions
- the invention belongs to the field of measurement of gaseous components in flue gas of thermal power plants, and particularly relates to a method for predicting ammonia escape in thermal power plants.
- the SCR process is generally used for NOx removal in thermal power plants; the unreacted ammonia gas at the outlet of the SCR denitration reactor is very important for the safe operation of downstream equipment, and some abnormal conditions will cause abnormal rise of ammonia escape, such as catalyst hole blockage, combustion System abnormality, AIG nozzle blockage, catalyst deactivation, etc. If the escaped ammonia is too high, it will react with SO3 in the flue gas to generate ammonium hydrogen sulfate, which will adhere to the heating surface of the downstream air preheater, resulting in an increase in the flow resistance, an increase in the current of the induced draft fan, an increase in power consumption, and even an increase in the The unit does not stop.
- ammonia escape online monitoring system of most power plants in China cannot accurately monitor ammonia escape data. show. This makes the problem of excessive ammonia escape not detected in time, and remedial measures are not taken until the performance of the denitration equipment is lower than the critical performance and the air preheater is seriously blocked.
- the purpose of the present invention is to provide a method for predicting the ammonia escape in thermal power plants.
- a method for predicting ammonia escape from a thermal power plant by using a grid method to measure the NOx concentration at the inlet and outlet of denitration reactor A, the NOx concentration at the inlet and outlet of denitration reactor B, the ammonia escape at the outlet of denitration reactor A, and the concentration of NOx at the outlet of denitration reactor B.
- Outlet ammonia escapes; when the relative standard deviation of the measured outlet NOx concentration distribution meets the requirements, then collect the DCS data of the online monitoring system in the same time period as the actual measurement, and preprocess the collected DCS data to obtain standard data.
- training data and validation data use the training data as input and ammonia slip data as output data to train ammonia slip prediction model A and ammonia slip prediction model B, and use the trained model to predict the real-time ammonia slip concentration.
- a further improvement of the present invention is that the DCS data includes NOx concentration data, flue gas temperature data, O2 concentration data, ammonia injection amount data, SO2 concentration data and air preheater differential pressure data.
- a further improvement of the present invention is that the concrete steps are as follows:
- the grid method is used to measure the NOx concentration at the inlet of denitration reactor A, the NOx concentration at the outlet of denitration reactor A, the NOx concentration at the inlet of denitration reactor B, the NOx concentration at the outlet of denitration reactor B, the ammonia escape at the outlet of denitration reactor A, and the NOx concentration at the outlet of denitration reactor A.
- step S2 Calculate the relative standard deviation of the NOx distribution at the outlet of the denitration reactor A and the outlet of the denitration reactor B according to the NOx concentration at the outlet of the denitration reactor A and the NOx concentration at the outlet of the denitration reactor B measured in step S1.
- the relative standard deviation is less than 30% Go to step S3; when the relative standard deviation is greater than 30%, repeat step S1 until the CV is less than 30%;
- the DCS data includes the temperature of the flue gas at the inlet of the denitration reactor A, the temperature of the flue gas at the inlet of the denitration reactor B, the O2 concentration at the inlet of the denitration reactor A, the denitration reactor A O2 concentration at the inlet of reactor B, NOx concentration at the inlet of denitration reactor A, NOx concentration at the outlet of denitration reactor A, NOx concentration at the inlet of denitration reactor B, NOx concentration at the outlet of denitration reactor B, ammonia injection amount of denitration reactor A, denitration reaction Ammonia injection amount of device B, NOx concentration of flue gas at the outlet of desulfurization tower, SO 2 concentration at the inlet of desulfurization tower, differential pressure of air preheater A and differential pressure of air preheater B;
- step S4 According to the NOx concentration at the inlet of the denitration reactor A, the NOx concentration at the outlet of the denitration reactor A, the NOx concentration at the inlet of the denitration reactor B, and the NOx concentration at the outlet of the denitration reactor B measured in step S1, the NOx concentration at the inlet of the denitration reactor A in step S3, The NOx concentration at the outlet of the denitration reactor A, the NOx concentration at the inlet of the denitration reactor B, and the NOx concentration at the outlet of the denitration reactor B are corrected, and when the deviation is less than 20%, step S5 is performed;
- step S5 Process the DCS data in step S3 to obtain standard data; divide the standard data into two parts, one part is used as model input data A or model input data B for training, and the other part is used as model input data A or Model input data B;
- ammonia slip prediction model A or ammonia slip prediction model B obtained by training is input into the model input data A or ammonia slip prediction model B used for verification for prediction, and the prediction data is obtained;
- step S7 collect the DCS data in step S3 in real time, perform outlier processing and normalization processing on the obtained DCS data, and obtain real-time model input data; adopt the trained ammonia slip prediction model A and ammonia slip prediction model obtained in step S6 B, Calculation of real-time model input data to predict real-time ammonia slip concentrations.
- a further improvement of the present invention is that, in step S4, when the deviation is greater than 20%, the online monitoring system is calibrated, repaired and maintained until the deviation between the NOx concentration collected by the online monitoring system and the measured NOx concentration distribution is less than 20%.
- a further improvement of the present invention is that, in step S5, abnormal value processing and normalization processing are performed on the DCS data in step S3 to obtain standard data.
- the model input data A includes the following parameters: flue gas temperature at the inlet of denitration reactor A , O concentration at the inlet of denitration reactor A, NOx concentration at the inlet of denitration reactor A, and outlet of denitration reactor A NOx concentration, ammonia injection amount of denitration reactor A, SO 2 concentration at the inlet of desulfurization tower, differential pressure of air preheater A, SO 2 concentration at the inlet of desulfurization tower and NOx concentration of flue gas at the outlet of desulfurization tower;
- the model input data B includes the following parameters: flue gas temperature at the inlet of denitration reactor B, O2 concentration at the inlet of denitration reactor B, NOx concentration at the inlet of denitration reactor B, NOx concentration at the outlet of denitration reactor B, ammonia injection amount of denitration reactor B, The SO 2 concentration at the inlet of the desulfurization tower, the differential pressure of the air preheater B and the NOx concentration of the flue gas at the outlet of the desulfurization tower.
- a further improvement of the present invention is that when the average relative error between the measured data and the predicted data in the same time period is less than 3%, the predicted data and the actual measured data are consistent.
- a further improvement of the present invention is that, in step S6, the ammonia slip prediction model A and the ammonia slip prediction model B adopt one or more of the following combinations: radial basis function neural network, BP neural network, genetic algorithm neural network, extreme learning machine, probabilistic neural network, generalized regression neural network, convolutional neural network, deep belief network, recurrent neural network, generalized regression neural network, particle swarm-based least squares support vector machine and ant colony optimization algorithm neural network.
- the present invention has the beneficial effects as follows: since the thermal power plant is equipped with an on - line monitoring system for testing NOx concentration, O2 concentration, air preheater differential pressure, ammonia injection amount, SO2 concentration, and accurate It is believed that the present invention uses these CEMS online measurement data as input variables to predict ammonia slip that is inaccurate in direct measurement, and overcomes the technical problem that the online measurement of ammonia slip is basically inaccurate in the prior art.
- the ammonia escape prediction model A and the ammonia escape prediction model B adopted in the present invention are stable and reliable, have short training time, strong generalization ability and accurate prediction data.
- the prediction model is continuously calibrated according to the measured ammonia slip data to ensure the accuracy of the ammonia slip prediction within a period of time, timely detect the problem of ammonia slip exceeding the standard, and avoid the blockage of the air preheater and the non-stop of the unit caused by ammonia slip.
- Figure 1 is a schematic diagram of the structure of the radial basis function neural network model.
- Each boiler in a thermal power plant is generally equipped with 2 denitration reactors and 2 air preheaters.
- One denitration reactor downstream corresponds to an air preheater. In order to distinguish the denitration reactors, they are generally named as denitration reactor A and denitration reaction.
- the corresponding downstream air preheaters are named as air preheater A and air preheater B.
- the NOx concentration at the inlet of denitration reactor A, the NOx concentration at the outlet of denitration reactor A, the NOx concentration at the inlet of denitration reactor B, the NOx concentration at the outlet of denitration reactor B, the ammonia escape at the outlet of denitration reactor A and the NOx concentration at the outlet of denitration reactor B are measured. Ammonia escape from the outlet; when all the measured NOx concentration distribution deviations meet the requirements, then collect DCS data.
- DCS data includes NOx concentration data, flue gas temperature data, O 2 concentration data, ammonia injection amount data, SO 2 concentration data and pressure difference Data; preprocess the collected DCS data to obtain standard data, divide the standard data into training data and verification data, use the training data as input and ammonia escape data as output data to train ammonia escape prediction model A and ammonia escape prediction model B , using the trained model to predict the real-time ammonia slip concentration.
- the present invention comprises the following steps:
- the grid method is used to measure the NOx concentration at the inlet of denitration reactor A, the NOx concentration at the outlet of denitration reactor A, the O2 concentration at the inlet of denitration reactor A, the O2 concentration at the inlet of denitration reactor B, the NOx concentration at the inlet of denitration reactor B, and the NOx concentration at the inlet of denitration reactor B.
- step S3 When the CV of the NOx distribution of the denitration reactor A and the denitration reactor B is less than 30%, the DCS data in the same time period as the grid method in step S1 is collected by CEMS, and the DCS data includes the flue gas at the inlet of the denitration reactor A.
- step S4 According to the data measured in step S1 (namely, the NOx concentration at the inlet of denitration reactor A, the NOx concentration at the outlet of denitration reactor A, the NOx concentration at the inlet of denitration reactor B, and the NOx concentration at the outlet of denitration reactor B), all the data in the DCS data in step S3 are analyzed. NOx concentration (ie NOx concentration at the inlet of denitration reactor A, NOx concentration at the outlet of denitration reactor A, NOx concentration at the inlet of denitration reactor B, NOx concentration at the outlet of denitration reactor B) is corrected. When the deviation is greater than 20%, the relevant online monitoring system (CEMS) Carry out calibration, inspection and maintenance until the deviation of the NOx concentration collected by the CEMS and the measured NOx concentration distribution is less than 20%, then step S5 is performed.
- CEMS online monitoring system
- step S5 Perform outlier processing and normalization processing on the DCS data in step S3 to obtain standard data; take the first 50% to 90% of the standard data as the model input data A or model input data B for training, and the rest Standard data is used as model input data A or model input data B for verification;
- the model input data A includes the following parameters: flue gas temperature at the inlet of denitration reactor A, O2 concentration at the inlet of denitration reactor A, NOx concentration at the inlet of denitration reactor A, NOx concentration at the outlet of denitration reactor A, and ammonia injection of denitration reactor A volume, air preheater A differential pressure, SO 2 concentration at the inlet of the desulfurization tower and NOx concentration in the flue gas at the outlet of the desulfurization tower;
- the model input data B includes the following parameters: flue gas temperature at the inlet of denitration reactor B, O2 concentration at the inlet of denitration reactor B, NOx concentration at the inlet of denitration reactor B, NOx concentration at the outlet of denitration reactor B, ammonia injection amount of denitration reactor B, The SO 2 concentration at the inlet of the desulfurization tower, the differential pressure of the air preheater B and the NOx concentration of the flue gas at the outlet of the desulfurization tower.
- step S6 respectively establish ammonia slip prediction model A and ammonia slip prediction model B: respectively use the model input data A for training obtained in step S5 and the ammonia slip at the outlet of denitration reactor A measured in step S1 at the same time period as input data and output The data is used to train the ammonia slip prediction model A; the model input data B for training obtained in step S5 and the ammonia slip at the outlet of the denitrification reactor B in the same time period measured in step S1 are respectively used as input data and output data to the ammonia slip prediction model. B for training.
- the ammonia slip prediction model A or B obtained by training is input into the model input data A or B for verification to obtain the predicted data.
- the average relative error between the measured data and the predicted data in the same time period in step S1 is less than 3%, it can be considered that The predicted data and the measured data show good agreement, and the ammonia slip prediction model is applied in step S7.
- Ammonia escape prediction model A and ammonia escape prediction model B may adopt one or more of the following combinations: radial basis function neural network, BP neural network, genetic algorithm neural network, extreme learning machine, probabilistic neural network, generalized regression neural network , Convolutional Neural Network, Deep Belief Network, Recurrent Neural Network, Generalized Regression Neural Network, Particle Swarm-based Least Squares Support Vector Machine and Ant Colony Optimization Algorithm Neural Network.
- the ammonia slip prediction model can be implemented on the DCS, or an external optimal control system independent of the DCS can be used.
- S8 Carry out on-site measurement of ammonia slip and NOx distribution within a certain period of time (such as within three months), and compare it with the ammonia slip obtained from the ammonia slip prediction model, so as to achieve the purpose of calibrating the forecast model and ensure the accuracy of ammonia slip prediction within a period of time. Accuracy.
- the model for predicting ammonia escape using radial basis function neural network is as follows:
- the radial basis function neural network model includes an input layer, a hidden layer, and an output layer.
- the input layer includes 8 nodes
- the hidden layer includes N nodes
- the output layer includes 1 node.
- the matrix form output by the radial basis function neural network model is:
- the error sum of squares is used as the training target error function of the radial basis function neural network, namely
- the gradients of the objective function to the data center c j and the expansion constant ⁇ j are:
- model input data A includes the following parameters: flue gas temperature at the inlet of denitration reactor A, O2 concentration at the inlet of denitration reactor A, NOx concentration at the inlet of denitration reactor A, NOx concentration at the outlet of denitration reactor A, ammonia injection amount of denitration reactor A, Air preheater A differential pressure, desulfurization tower inlet SO 2 concentration, desulfurization tower outlet flue gas NOx concentration;
- model input data B includes the following parameters: denitration reactor B inlet flue gas temperature, denitration reactor B inlet O 2 concentration, NOx concentration at the inlet of denitration reactor B, NOx concentration at the outlet of denitration reactor B, ammonia injection amount of denitration reactor B, SO 2 concentration at the inlet of the desulfurization tower, differential pressure of air preheater B, NOx concentration of the flue gas at the outlet of the desulfurization tower.
Landscapes
- Engineering & Computer Science (AREA)
- Chemical & Material Sciences (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Immunology (AREA)
- Biochemistry (AREA)
- Analytical Chemistry (AREA)
- Medicinal Chemistry (AREA)
- Food Science & Technology (AREA)
- Combustion & Propulsion (AREA)
- Pathology (AREA)
- Biophysics (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Evolutionary Computation (AREA)
- Data Mining & Analysis (AREA)
- Computational Linguistics (AREA)
- Biomedical Technology (AREA)
- Artificial Intelligence (AREA)
- Treating Waste Gases (AREA)
Abstract
一种预测火电厂氨逃逸的方法，实测脱硝反应器A入口NOx浓度、脱硝反应器A出口NOx浓度、脱硝反应器B入口NOx浓度、脱硝反应器B出口NOx浓度、脱硝反应器A出口氨逃逸以及脱硝反应器B出口氨逃逸；当实测的所有NOx浓度满足要求时，再采集DCS数据；对采集的DCS数据进行预处理，得到标准数据，将标准数据分为训练数据与验证数据，利用训练数据作为输入，氨逃逸数据作为输出数据训练氨逃逸预测模型A以及氨逃逸预测模型B，利用训练好的模型预测实时的氨逃逸浓度。方法采用的氨逃逸预测模型A以及氨逃逸预测模型B稳定可靠，训练时间短，泛化能力强，预测数据准确。
Description
本发明属于火电厂烟气中气态组分的测量领域，具体涉及一种预测火电厂氨逃逸的方法。
火电厂的NOx脱除普遍采用了SCR工艺；SCR脱硝反应器出口未反应的氨气对下游设备的安全运行至关重要，一些异常情况都会引起氨逃逸的非正常上升，如催化剂孔堵塞、燃烧系统异常、AIG喷口堵塞、催化剂失活等。逃逸氨过高，会与烟气中的SO
3反应，生成硫酸氢铵，黏附在下游的空气预热器受热面，造成流通阻力升高，导致引风机电流增大、电耗上升，甚至导致机组非停。同时，过高的逃逸氨浓度会增加脱硝还原剂(NH
3)的耗量，增加脱硝运行成本。因此，SCR装置通常都要求控制在3μL/L以内，SNCR通常要求控制在10μL/L以内。为此，采用在线监测系统(CEMS)来检测烟气中的氨逃逸浓度。
受高浓度粉尘、振动、单点测量代表性低等因素的影响，国内大多数电厂的氨逃逸在线监测系统无法准确监测氨逃逸数据，即使存在氨逃逸超标问题，也无法被氨逃逸在线监测系统显示。这使得氨逃逸过大问题无法及时被发现，直至脱硝设备性能低于临界性能和空气预热器堵塞严重了，才采取补救措施。
准确、便捷、抗粉尘干扰能力强、抗SOx干扰能力强的NH
3浓度的测试或者预测方法对脱硝设备、空气预热器、机组的安全运行意义极大。
发明内容
针对SCR脱硝氨逃逸在线监测系统检测不准确而引发的空气预热器硫酸氢 铵严重堵塞问题，本发明的目的是提供了一种预测火电厂氨逃逸的方法。
为实现上述目的，本发明采用的技术方案如下：
一种预测火电厂氨逃逸的方法，通过采用网格法实测脱硝反应器A入口及出口的NOx浓度、脱硝反应器B入口及出口的NOx浓度、脱硝反应器A出口氨逃逸、脱硝反应器B出口氨逃逸；当实测的出口NOx浓度分布相对标准偏差满足要求时，再采集与实测同时间段的在线监测系统的DCS数据，对采集的DCS数据进行预处理，得到标准数据，将标准数据分为训练数据与验证数据，利用训练数据作为输入，氨逃逸数据作为输出数据训练氨逃逸预测模型A以及氨逃逸预测模型B，利用训练好的模型预测实时的氨逃逸浓度。
本发明进一步的改进在于，DCS数据包括NOx浓度数据、烟气温度数据、O
2浓度数据、喷氨量数据、SO
2浓度数据以及空气预热器差压数据。
本发明进一步的改进在于，具体步骤如下：
S1：采用网格法实测脱硝反应器A入口NOx浓度、脱硝反应器A出口NOx浓度、脱硝反应器B入口NOx浓度、脱硝反应器B出口NOx浓度、脱硝反应器A出口氨逃逸以及脱硝反应器B出口氨逃逸；
S2：根据步骤S1实测的脱硝反应器A出口NOx浓度与脱硝反应器B出口NOx浓度，分别计算脱硝反应器A出口、脱硝反应器B的出口NOx分布相对标准偏差，相对标准偏差小于30％时进行步骤S3；相对标准偏差大于30％时，重复步骤S1，直到CV小于30％；
S3：采集与步骤S1中网格法实测同时间段内的DCS数据，DCS数据包括脱硝反应器A入口烟气温度、脱硝反应器B入口烟气温度、脱硝反应器A入口O
2浓度、脱硝反应器B入口O
2浓度、脱硝反应器A入口NOx浓度、脱硝反应器A 出口NOx浓度、脱硝反应器B入口NOx浓度、脱硝反应器B出口NOx浓度、脱硝反应器A喷氨量、脱硝反应器B喷氨量、脱硫塔出口烟气的NOx浓度、脱硫塔入口SO
2浓度、空气预热器A差压以及空气预热器B差压；
S4：根据步骤S1实测的脱硝反应器A入口NOx浓度、脱硝反应器A出口NOx浓度、脱硝反应器B入口NOx浓度以及脱硝反应器B出口NOx浓度对步骤S3中脱硝反应器A入口NOx浓度、脱硝反应器A出口NOx浓度、脱硝反应器B入口NOx浓度以及脱硝反应器B出口NOx浓度进行校正，偏差小于20％时，进行步骤S5；
S5：对步骤S3的DCS数据进行处理，得到标准数据；将标准数据分为两部分，一部分作为得到训练用的模型输入数据A或模型输入数据B，另一部分作为验证用的模型输入数据A或模型输入数据B；
S6：分别建立氨逃逸预测模型A以及氨逃逸预测模型B，分别将步骤S5得到的训练用的模型输入数据A、步骤S1实测的同时间段的脱硝反应器A出口氨逃逸作为输入数据、输出数据对氨逃逸预测模型A进行训练；分别把步骤S5得到的训练用的模型输入数据B、步骤S1实测的同时间段的脱硝反应器B出口氨逃逸作为输入数据、输出数据对氨逃逸预测模型B进行训练；
训练得到的氨逃逸预测模型A或氨逃逸预测模型B输入验证用的模型输入数据A或氨逃逸预测模型B进行预测，得到预测数据；
S7：实时采集步骤S3中的DCS数据，对得到的DCS数据进行异常值处理和归一化处理，得到实时模型输入数据；采用步骤S6得到的训练后的氨逃逸预测模型A以及氨逃逸预测模型B，对实时模型输入数据进行计算，预测实时的氨逃逸浓度。
本发明进一步的改进在于，步骤S4中，偏差大于20％时，对在线监测系统进行标定、检修维护，直至在线监测系统采集的NOx浓度与实测的NOx浓度分布偏差小于20％。
本发明进一步的改进在于，步骤S5中，对步骤S3的DCS数据进行异常值处理和归一化处理，得到标准数据。
本发明进一步的改进在于，步骤S5中，模型输入数据A包括以下参数：脱硝反应器A入口烟气温度、脱硝反应器A入口O
2浓度、脱硝反应器A入口NOx浓度、脱硝反应器A出口NOx浓度、脱硝反应器A喷氨量、脱硫塔入口SO
2浓度、空气预热器A差压、脱硫塔入口SO
2浓度与脱硫塔出口烟气的NOx浓度；
模型输入数据B包括以下参数：脱硝反应器B入口烟气温度、脱硝反应器B入口O
2浓度、脱硝反应器B入口NOx浓度、脱硝反应器B出口NOx浓度、脱硝反应器B喷氨量、脱硫塔入口SO
2浓度、空气预热器B差压与脱硫塔出口烟气的NOx浓度。
本发明进一步的改进在于，同时间段的实测数据与预测数据的平均相对误差低于3％时，预测数据和实测数据具有一致性。
本发明进一步的改进在于，步骤S6中，氨逃逸预测模型A、氨逃逸预测模型B采用以下的一种或多种组合：径向基函数神经网络、BP神经网络、遗传算法神经网络、极限学习机、概率神经网络、广义回归神经网络、卷积神经网络、深度信念网络、循环神经网络、广义回归神经网络、基于粒子群的最小二乘支持向量机与蚁群优化算法神经网络。
与现有技术相比，本发明具有的有益效果为：由于火电厂装有测试NOx浓度、O
2浓度、空气预热器差压、喷氨量、SO
2浓度的在线监测系统，而且准确可 信，本发明根据这些CEMS在线测量数据作为输入变量，预测直接测量不准的氨逃逸，克服了现有技术中，对氨逃逸的在线测量基本不准确的技术问题。本发明采用的氨逃逸预测模型A以及氨逃逸预测模型B稳定可靠，训练时间短，泛化能力强，预测数据准确。
进一步的，根据实测的氨逃逸数据不断校准预测模型，保证一段时间内的氨逃逸预测的准确度，及时发现氨逃逸超标问题，避免氨逃逸导致的空气预热器堵塞、机组非停等问题。
图1为径向基函数神经网络模型结构示意图。
下面结合附图对本发明进行详细说明。
火电厂每台锅炉一般设置有2个脱硝反应器和2个空气预热器，一个脱硝反应器下游对应着一个空气预热器，为了区分脱硝反应器，一般命名为脱硝反应器A和脱硝反应器B；相对应的下游空气预热器命名为空气预热器A以及空气预热器B。
通过采用网格法实测脱硝反应器A入口NOx浓度、脱硝反应器A出口NOx浓度、脱硝反应器B入口NOx浓度、脱硝反应器B出口NOx浓度、脱硝反应器A出口氨逃逸以及脱硝反应器B出口氨逃逸；当实测的所有NOx浓度分布偏差满足要求时，再采集DCS数据，DCS数据包括NOx浓度数据、烟气温度数据、O
2浓度数据、喷氨量数据、SO
2浓度数据以及压差数据；对采集的DCS数据进行预处理，得到标准数据，将标准数据分为训练数据与验证数据，利用训练数据作为输入，氨逃逸数据作为输出数据训练氨逃逸预测模型A以及氨逃逸预测模型 B，利用训练好的模型预测实时的氨逃逸浓度。
具体的，本发明包括以下步骤：
S1：采用网格法实测脱硝反应器A入口NOx浓度、脱硝反应器A出口NOx浓度、脱硝反应器A入口O
2浓度、脱硝反应器B入口O
2浓度、脱硝反应器B入口NOx浓度、脱硝反应器B出口NOx浓度、脱硝反应器A出口氨逃逸以及脱硝反应器B出口氨逃逸。
S2：根据步骤S1实测的脱硝反应器A出口NOx浓度与脱硝反应器B出口NOx浓度，分别计算脱硝反应器A出口、脱硝反应器B的出口NOx分布相对标准偏差(CV)，CV小于30％时进行步骤S3，CV大于30％时，对相关的喷氨格栅的手动阀门开度进行调节并重复步骤S1，直到CV小于30％。
S3：脱硝反应器A、脱硝反应器B的NOx分布的CV小于30％时，通过CEMS采集与步骤S1中网格法实测同时间段内的DCS数据，DCS数据包括脱硝反应器A入口烟气温度、脱硝反应器B入口烟气温度、脱硝反应器A入口O
2浓度、脱硝反应器B入口O
2浓度、脱硝反应器A入口NOx浓度、脱硝反应器A出口NOx浓度、脱硝反应器B入口NOx浓度、脱硝反应器B出口NOx浓度、脱硝反应器A喷氨量、脱硝反应器B喷氨量、脱硫塔出口烟气的NOx浓度、脱硫塔入口SO
2浓度、空气预热器A差压以及空气预热器B差压。
S4：根据步骤S1实测的数据(即脱硝反应器A入口NOx浓度、脱硝反应器A出口NOx浓度、脱硝反应器B入口NOx浓度、脱硝反应器B出口NOx浓度)对步骤S3中DCS数据中所有NOx浓度(即脱硝反应器A入口NOx浓度、脱硝反应器A出口NOx浓度、脱硝反应器B入口NOx浓度、脱硝反应器B出口NOx浓度)进行校正，偏差大于20％时，对相关在线监测系统(CEMS)进行标定、 检修维护，直至CEMS采集的NOx浓度与实测的NOx浓度分布偏差小于20％时，进行步骤S5。
S5：对步骤S3的DCS数据进行异常值处理和归一化处理，得到标准数据；取标准数据前50％～90％的部分作为得到训练用的模型输入数据A或模型输入数据B，剩余的标准数据作为验证用的模型输入数据A或模型输入数据B；
其中，模型输入数据A包括以下参数：脱硝反应器A入口烟气温度、脱硝反应器A入口O
2浓度、脱硝反应器A入口NOx浓度、脱硝反应器A出口NOx浓度、脱硝反应器A喷氨量、空气预热器A差压、脱硫塔入口SO
2浓度与脱硫塔出口烟气的NOx浓度；
模型输入数据B包括以下参数：脱硝反应器B入口烟气温度、脱硝反应器B入口O
2浓度、脱硝反应器B入口NOx浓度、脱硝反应器B出口NOx浓度、脱硝反应器B喷氨量、脱硫塔入口SO
2浓度、空气预热器B差压与脱硫塔出口烟气的NOx浓度。
S6：分别建立氨逃逸预测模型A以及氨逃逸预测模型B：分别把步骤S5得到的训练用的模型输入数据A、步骤S1实测的同时间段的脱硝反应器A出口氨逃逸作为输入数据、输出数据对氨逃逸预测模型A进行训练；分别把步骤S5得到的训练用的模型输入数据B、步骤S1实测的同时间段的脱硝反应器B出口氨逃逸作为输入数据、输出数据对氨逃逸预测模型B进行训练。
训练得到的氨逃逸预测模型A或B输入验证用的模型输入数据A或B进行预测得到预测数据，步骤S1的同时间段的实测数据与预测数据的平均相对误差低于3％时，可认为预测数据和实测数据显示出了良好的一致性，氨逃逸预测模型在步骤S7中应用。
氨逃逸预测模型A、氨逃逸预测模型B可以采用以下的一种或多种组合：径向基函数神经网络、BP神经网络、遗传算法神经网络、极限学习机、概率神经网络、广义回归神经网络、卷积神经网络、深度信念网络、循环神经网络、广义回归神经网络、基于粒子群的最小二乘支持向量机与蚁群优化算法神经网络。
S7：依据OPC、TCP/IP、Modbus等各种工业通讯协议，实现与DCS的双向通讯，实时采集步骤S3中的DCS数据；对得到的实时数据进行异常值处理和归一化处理，得到实时模型输入数据；采用步骤S6得到的训练后的氨逃逸预测模型A、氨逃逸预测模型B，对实时模型输入数据进行计算，预测得到实时的氨逃逸浓度。
氨逃逸预测模型可以在DCS上实现，也可以采用独立于DCS的外挂式优化控制系统。
S8：在一段时间(如三个月内)再开展氨逃逸、NOx分布现场实测，与氨逃逸预测模型得到的氨逃逸进行对比，达到校准预测模型的目的，保证一段时间内的氨逃逸预测的准确度。
下面为具体实施例。
参见图1，采用径向基函数神经网络的预测氨逃逸的模型具体为：
径向基函数神经网络模型包括输入层、隐含层、输出层，所述输入层包括8个节点，所述隐含层包括N个节点，所述输出层包括1个节点。对8个采样点{(xi,ti)|xi∈Rn,ti∈R}，i＝1,2,^,8，结构为8—N—1的径向基函数神经网络模型输出为：
其中，i＝1,2,3,4,5,6,7,8，w
j(j＝1,2,3…)是第j个隐含层节点到输出层节点的权值；
其中cj为核函数的数据中心，σj为该核函数的扩展常数，N为节点数量，对于所有样本，所述隐含层节点输出矩阵为：
径向基函数神经网络模型输出的矩阵形式为：
其中
为连接隐含层与输出层的权值矩阵，
为径向基函数神经网络模型的实际输出矩阵；
对径向基函数神经网络模型参数进行调整，得到预测氨逃逸的径向基函数神经网络模型，将径向基函数神经网络输出值Y和真实值T＝(t
1t
2Λt
n)之间的误差平方和作为径向基函数神经网络的训练目标误差函数，即
为寻找最优输出权值W使网络输出值Y和真实值T＝(t
1t
2Λt
n)之间的误差平方和以及输出权值W范数最小，分两步对径向基函数神经网络参数进行优化：
首先通过矩阵
的广义逆求得W的最优值W；
然后以Y和T的误差平方和为目标函数，通过梯度下降算法优化隐节点数据中心：
c
j以及扩展常数σ
j，目标函数对数据中心c
j和扩展常数σ
j的梯度分别为：
数据中心cj以及扩展常数σ
j的更新公式为：
η为学习率，k＝1,2,^,n。
模型输入数据A包括以下参数：脱硝反应器A入口烟气温度、脱硝反应器A入口O
2浓度、脱硝反应器A入口NOx浓度、脱硝反应器A出口NOx浓度、脱硝反应器A喷氨量、空气预热器A差压、脱硫塔入口SO
2浓度、脱硫塔出口烟气的NOx浓度；模型输入数据B包括以下参数：脱硝反应器B入口烟气温度、脱硝反应器B入口O
2浓度、脱硝反应器B入口NOx浓度、脱硝反应器B出口NOx浓度、脱硝反应器B喷氨量、脱硫塔入口SO
2浓度、空气预热器B差压、脱硫塔出口烟气的NOx浓度。
Claims (7)
- 一种预测火电厂氨逃逸的方法，其特征在于，通过采用网格法实测脱硝反应器A入口及出口的NOx浓度、脱硝反应器B入口及出口的NOx浓度、脱硝反应器A出口氨逃逸、脱硝反应器B出口氨逃逸；当实测的出口NOx浓度分布相对标准偏差满足要求时，再采集与实测同时间段的在线监测系统的DCS数据，对采集的DCS数据进行预处理，得到标准数据，将标准数据分为训练数据与验证数据，利用训练数据作为输入，氨逃逸数据作为输出数据训练氨逃逸预测模型A以及氨逃逸预测模型B，利用训练好的模型预测实时的氨逃逸浓度；其中，DCS数据包括NOx浓度数据、烟气温度数据、O 2浓度数据、喷氨量数据、SO 2浓度数据以及空气预热器差压数据。
- 根据权利要求1所述的一种预测火电厂氨逃逸的方法，其特征在于，具体步骤如下：S1：采用网格法实测脱硝反应器A入口NOx浓度、脱硝反应器A出口NOx浓度、脱硝反应器B入口NOx浓度、脱硝反应器B出口NOx浓度、脱硝反应器A出口氨逃逸以及脱硝反应器B出口氨逃逸；S2：根据步骤S1实测的脱硝反应器A出口NOx浓度与脱硝反应器B出口NOx浓度，分别计算脱硝反应器A出口、脱硝反应器B的出口NOx分布相对标准偏差，相对标准偏差小于30％时进行步骤S3；相对标准偏差大于30％时，重复步骤S1，直到CV小于30％；S3：采集与步骤S1中网格法实测同时间段内的DCS数据，DCS数据包括脱硝反应器A入口烟气温度、脱硝反应器B入口烟气温度、脱硝反应器A入口O 2浓度、脱硝反应器B入口O 2浓度、脱硝反应器A入口NOx浓度、脱硝反应器A出口NOx浓度、脱硝反应器B入口NOx浓度、脱硝反应器B出口NOx浓度、 脱硝反应器A喷氨量、脱硝反应器B喷氨量、脱硫塔出口烟气的NOx浓度、脱硫塔入口SO 2浓度、空气预热器A差压以及空气预热器B差压；S4：根据步骤S1实测的脱硝反应器A入口NOx浓度、脱硝反应器A出口NOx浓度、脱硝反应器B入口NOx浓度以及脱硝反应器B出口NOx浓度对步骤S3中脱硝反应器A入口NOx浓度、脱硝反应器A出口NOx浓度、脱硝反应器B入口NOx浓度以及脱硝反应器B出口NOx浓度进行校正，偏差小于20％时，进行步骤S5；S5：对步骤S3的DCS数据进行处理，得到标准数据；将标准数据分为两部分，一部分作为得到训练用的模型输入数据A或模型输入数据B，另一部分作为验证用的模型输入数据A或模型输入数据B；S6：分别建立氨逃逸预测模型A以及氨逃逸预测模型B，分别将步骤S5得到的训练用的模型输入数据A、步骤S1实测的同时间段的脱硝反应器A出口氨逃逸作为输入数据、输出数据对氨逃逸预测模型A进行训练；分别把步骤S5得到的训练用的模型输入数据B、步骤S1实测的同时间段的脱硝反应器B出口氨逃逸作为输入数据、输出数据对氨逃逸预测模型B进行训练；训练得到的氨逃逸预测模型A或氨逃逸预测模型B输入验证用的模型输入数据A或氨逃逸预测模型B进行预测，得到预测数据；S7：实时采集步骤S3中的DCS数据，对得到的DCS数据进行异常值处理和归一化处理，得到实时模型输入数据；采用步骤S6得到的训练后的氨逃逸预测模型A以及氨逃逸预测模型B，对实时模型输入数据进行计算，预测实时的氨逃逸浓度。
- 根据权利要求2所述的一种预测火电厂氨逃逸的方法，其特征在于，步 骤S4中，偏差大于20％时，对在线监测系统进行标定、检修维护，直至在线监测系统采集的NOx浓度与实测的NOx浓度分布偏差小于20％。
- 根据权利要求2所述的一种预测火电厂氨逃逸的方法，其特征在于，步骤S5中，对步骤S3的DCS数据进行异常值处理和归一化处理，得到标准数据。
- 根据权利要求2所述的一种预测火电厂氨逃逸的方法，其特征在于，步骤S5中，模型输入数据A包括以下参数：脱硝反应器A入口烟气温度、脱硝反应器A入口O 2浓度、脱硝反应器A入口NOx浓度、脱硝反应器A出口NOx浓度、脱硝反应器A喷氨量、脱硫塔入口SO 2浓度、空气预热器A差压、脱硫塔入口SO 2浓度与脱硫塔出口烟气的NOx浓度；模型输入数据B包括以下参数：脱硝反应器B入口烟气温度、脱硝反应器B入口O 2浓度、脱硝反应器B入口NOx浓度、脱硝反应器B出口NOx浓度、脱硝反应器B喷氨量、脱硫塔入口SO 2浓度、空气预热器B差压与脱硫塔出口烟气的NOx浓度。
- 根据权利要求2所述的一种预测火电厂氨逃逸的方法，其特征在于，同时间段的实测数据与预测数据的平均相对误差低于3％时，预测数据和实测数据具有一致性。
- 根据权利要求2所述的一种预测火电厂氨逃逸的方法，其特征在于，步骤S6中，氨逃逸预测模型A、氨逃逸预测模型B采用以下的一种或多种组合：径向基函数神经网络、BP神经网络、遗传算法神经网络、极限学习机、概率神经网络、广义回归神经网络、卷积神经网络、深度信念网络、循环神经网络、广义回归神经网络、基于粒子群的最小二乘支持向量机与蚁群优化算法神经网络。
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011210798.4A CN112461995A (zh) | 2020-11-03 | 2020-11-03 | 一种预测火电厂氨逃逸的方法 |
| CN202011210798.4 | 2020-11-03 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2022095534A1 true WO2022095534A1 (zh) | 2022-05-12 |
Family
ID=74835041
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2021/112270 WO2022095534A1 (zh) | 2020-11-03 | 2021-08-12 | 一种预测火电厂氨逃逸的方法 |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN112461995A (zh) |
| WO (1) | WO2022095534A1 (zh) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112461995A (zh) * | 2020-11-03 | 2021-03-09 | 西安热工研究院有限公司 | 一种预测火电厂氨逃逸的方法 |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2006000877A2 (en) * | 2004-06-21 | 2006-01-05 | Eaton Corporation | Strategy for controlling nox emissions and ammonia slip in an scr system using a nonselective nox/nh3 sensor |
| CN106599586A (zh) * | 2016-12-19 | 2017-04-26 | 北京国能中电节能环保技术股份有限公司 | 基于神经网络的scr智能喷氨优化方法及装置 |
| CN107158946A (zh) * | 2017-05-27 | 2017-09-15 | 苏州西热节能环保技术有限公司 | 一种氨逃逸浓度实时在线预测与控制方法 |
| US20180024509A1 (en) * | 2016-07-25 | 2018-01-25 | General Electric Company | System modeling, control and optimization |
| CN108837698A (zh) * | 2018-07-02 | 2018-11-20 | 大唐环境产业集团股份有限公司 | 基于先进测量仪表和先进控制算法的scr脱硝喷氨优化方法及系统 |
| CN109766596A (zh) * | 2018-12-25 | 2019-05-17 | 国网新疆电力有限公司电力科学研究院 | 一种脱硝经济运行的专家系统构建方法 |
| US10436739B1 (en) * | 2014-10-06 | 2019-10-08 | Bjr Sensors Llc | Low cost, fast and sensitive NOx and NH3 sensor |
| CN111460726A (zh) * | 2020-01-22 | 2020-07-28 | 杭州电子科技大学 | 一种煤泥流化床锅炉脱硝系统氨逃逸的优化方法 |
| CN112461995A (zh) * | 2020-11-03 | 2021-03-09 | 西安热工研究院有限公司 | 一种预测火电厂氨逃逸的方法 |
Family Cites Families (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP5404320B2 (ja) * | 2009-10-30 | 2014-01-29 | 三菱重工業株式会社 | 内燃機関のＮＯｘ浄化装置 |
| CN104008427A (zh) * | 2014-05-16 | 2014-08-27 | 华南理工大学 | 基于bp神经网络的中央空调冷负荷的预测方法 |
| CN104190254B (zh) * | 2014-09-09 | 2016-08-24 | 国家电网公司 | 一种优化scr喷氨的方法 |
| CN104297008A (zh) * | 2014-10-11 | 2015-01-21 | 苏州华瑞能泰发电技术有限公司 | 基于现场性能测试的脱硝装置潜能评估与预测方法 |
| CN107694300B (zh) * | 2016-08-08 | 2020-01-17 | 中冶长天国际工程有限责任公司 | 活性炭脱硫脱硝系统的喷氨量控制方法和装置 |
| CN107694337A (zh) * | 2017-11-03 | 2018-02-16 | 吉林省电力科学研究院有限公司 | 基于神经网络预测控制的燃煤机组scr烟气脱硝控制方法 |
| CN108490132A (zh) * | 2018-04-02 | 2018-09-04 | 华能国际电力股份有限公司 | 一种基于比对差值法氨逃逸检测装置及方法 |
| CN208340474U (zh) * | 2018-05-03 | 2019-01-08 | 中国大唐集团科学技术研究院有限公司华东分公司 | 一种脱硝氮氧化物分级控制系统 |
| CN109493250B (zh) * | 2018-11-06 | 2021-09-07 | 大唐南京环保科技有限责任公司 | 一种scr反应器的脱硝能力的评估方法 |
| CN109499364B (zh) * | 2018-11-29 | 2021-05-28 | 东南大学 | 一种基于数字镜像的催化剂辅助设计方法 |
| CN109766577B (zh) * | 2018-12-07 | 2023-07-14 | 大唐南京环保科技有限责任公司 | 基于pls局部拟合的scr反应器信息快速获取方法 |
| CN110094251B (zh) * | 2019-05-05 | 2020-06-02 | 东南大学 | 基于分时段多模型建模的scr催化剂性能劣化分析方法 |
-
2020
- 2020-11-03 CN CN202011210798.4A patent/CN112461995A/zh active Pending
-
2021
- 2021-08-12 WO PCT/CN2021/112270 patent/WO2022095534A1/zh active Application Filing
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2006000877A2 (en) * | 2004-06-21 | 2006-01-05 | Eaton Corporation | Strategy for controlling nox emissions and ammonia slip in an scr system using a nonselective nox/nh3 sensor |
| US10436739B1 (en) * | 2014-10-06 | 2019-10-08 | Bjr Sensors Llc | Low cost, fast and sensitive NOx and NH3 sensor |
| US20180024509A1 (en) * | 2016-07-25 | 2018-01-25 | General Electric Company | System modeling, control and optimization |
| CN106599586A (zh) * | 2016-12-19 | 2017-04-26 | 北京国能中电节能环保技术股份有限公司 | 基于神经网络的scr智能喷氨优化方法及装置 |
| CN107158946A (zh) * | 2017-05-27 | 2017-09-15 | 苏州西热节能环保技术有限公司 | 一种氨逃逸浓度实时在线预测与控制方法 |
| CN108837698A (zh) * | 2018-07-02 | 2018-11-20 | 大唐环境产业集团股份有限公司 | 基于先进测量仪表和先进控制算法的scr脱硝喷氨优化方法及系统 |
| CN109766596A (zh) * | 2018-12-25 | 2019-05-17 | 国网新疆电力有限公司电力科学研究院 | 一种脱硝经济运行的专家系统构建方法 |
| CN111460726A (zh) * | 2020-01-22 | 2020-07-28 | 杭州电子科技大学 | 一种煤泥流化床锅炉脱硝系统氨逃逸的优化方法 |
| CN112461995A (zh) * | 2020-11-03 | 2021-03-09 | 西安热工研究院有限公司 | 一种预测火电厂氨逃逸的方法 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112461995A (zh) | 2021-03-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108956876B (zh) | 一种烟气在线连续监测系统的测量时延修正方法 | |
| CN108380043B (zh) | 一种scr脱硝装置分区喷氨调节控制方法 | |
| CN102565274B (zh) | 一种电厂scr脱硝系统中催化剂活性的修正方法 | |
| CN113433911B (zh) | 基于浓度精准预测的脱硝装置喷氨精准控制系统与方法 | |
| CN102759931B (zh) | 烟气脱硝的控制方法和装置 | |
| CN109493250A (zh) | 一种scr反应器的脱硝能力的评估方法 | |
| CN106596198A (zh) | 一种在线so3测量系统及方法 | |
| WO2022095534A1 (zh) | 一种预测火电厂氨逃逸的方法 | |
| CN111753389B (zh) | 一种燃煤锅炉入炉原煤热值在线软测量方法和装置 | |
| CN113175678B (zh) | 垃圾焚烧的监控方法及装置 | |
| CN104504238A (zh) | 一种基于神经网络的燃煤机组氮氧化物排放浓度预测方法 | |
| CN112098461B (zh) | 一种烟气引入法在线水露点酸露点测量装置 | |
| CN113740090A (zh) | 一种用于火电厂的空预器防堵方法及系统 | |
| CN114062098A (zh) | 基于稀释法同步标定的NOx浓度分布监测系统及监测方法 | |
| CN113970937A (zh) | 一种scr脱硝精准喷氨控制方法 | |
| CN113283052A (zh) | 飞灰含碳量软测量方法及燃煤锅炉燃烧优化方法和系统 | |
| CN109260948B (zh) | 波动负荷下脱硝均匀性测量优化系统及方法 | |
| CN116519555A (zh) | 一种稀释法烟气监测系统 | |
| CN113488111B (zh) | 一种scr脱硝系统喷氨量优化建模方法 | |
| CN111540412A (zh) | 一种基于最小二乘法的scr反应器入口烟气软测量方法 | |
| CN113836794B (zh) | 一种软硬结合的飞灰含碳量在线监测方法 | |
| Zain et al. | Development of a neural network Predictive Emission Monitoring System for flue gas measurement | |
| CN113262629B (zh) | 一种基于脱硝装置分区测量的锅炉烟气流量计算方法 | |
| CN112379035B (zh) | 一种在线评估scr脱硝装置催化剂性能的系统及方法 | |
| CN214151507U (zh) | 基于引风机出口NOx浓度快速测量反馈控制SCR喷氨量的结构 |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 21888233 Country of ref document: EP Kind code of ref document: A1 |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| 122 | Ep: pct application non-entry in european phase |
Ref document number: 21888233 Country of ref document: EP Kind code of ref document: A1 |
|
| 32PN | Ep: public notification in the ep bulletin as address of the adressee cannot be established |
Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 17-10-2023) |