A coal-fired flue gas SCR denitrification control system and method based on zoned ammonia injection

Abstract

The present invention relates to a coal-fired flue gas SCR denitrification control system and method based on zoned ammonia injection, the system comprising an SCR zoned measurement system and an SCR multi-zoned hybrid prediction control system; the SCR zoned measurement system comprises an SCR inlet flue, an SCR injection device, an SCR reactor and an SCR outlet flue connected in sequence; a flow measurement instrument and a temperature measurement instrument are arranged in the SCR inlet flue, and a NO x Measuring instrument; SCR multi-partition hybrid prediction control system includes a multi-partition hybrid prediction model and a multi-partition control strategy module. The multi-partition hybrid prediction model is used to predict the NO x Concentration; real-time measurement of NO in each partition x The concentration is used to modify the prediction model; the multi-zone control strategy module is used to determine the total amount of NH3 and the amount of NH3 in each zone. The present invention reduces the operating cost.

Description

Technical Field

[0001] The present invention relates to the technical field of SCR denitration, and in particular to a control system and method for SCR denitration of coal-fired flue gas based on zoned ammonia injection. Background Art

[0002] At present, typical coal-fired power plants use SCR (selective catalytic reduction) denitrification systems. The ammonia injection point is set in the vertical flue at the entrance of the SCR reactor. The flue cross-section is large, resulting in the cross-sectional velocity field, temperature field and NO x The concentration distribution is uneven; the ammonia injection pipeline uses a manual valve, which cannot realize online adjustment of the ammonia injection ratio and amount; the existing DCS (Distributed Control System) system has a time lag in responding to operating conditions; in order to meet the ultra-low emission requirements, an excessive ammonia injection control method is manually adopted; ultimately, the SCR denitrification system outlet NO x Frequent fluctuations, high system operation material consumption, excessive ammonia injection causes subsequent air preheater blockage and corrosion, and increases the operating cost of the denitrification system. Summary of the invention

[0003] The purpose of the present invention is to provide a coal-fired flue gas SCR denitration control system and method based on zoned ammonia injection to reduce the NO x Fluctuations, to reduce system operating costs.

[0004] To achieve the above object, the present invention provides the following solutions:

[0005] A coal-fired flue gas SCR denitrification control system based on zoned ammonia injection, comprising: an SCR zoned measurement system and an SCR multi-zoned hybrid prediction control system;

[0006] The SCR partition measurement system comprises an SCR inlet flue, an SCR injection device, an SCR reactor and an SCR outlet flue connected in sequence; a flow measuring instrument and a temperature measuring instrument are arranged in the SCR inlet flue, wherein the flow measuring instrument is used to measure the flue gas flow of each partition in real time, and the temperature measuring instrument is used to measure the temperature of each partition in real time; and a NO x Measuring instrument, the NO x The measuring instrument is used to measure the NO in each partition in real time. x concentration;

[0007] The SCR multi-zone hybrid prediction control system includes a multi-zone hybrid prediction model and a multi-zone control strategy module. The multi-zone hybrid prediction model is used to predict the NO in each zone according to the boiler operation data, the temperature of each zone and the flue gas flow of each zone. x concentration; the NO xThe measuring instrument measures the NO of each partition in real time x The concentration is used to modify the multi-partition hybrid prediction model; the multi-partition control strategy module is used to adjust the NO concentration of each partition according to the output of the multi-partition hybrid prediction model. x Concentration and set outlet NO x concentration, determining the total amount of NH3 and the amount of NH3 in each partition, and determining the total amount of reducing agent input to the SCR injection device and the amount of reducing agent in each partition according to the total amount of NH3 and the amount of NH3 in each partition.

[0008] Optionally, the boiler operation data includes load, coal feed rate, oxygen amount, temperature, primary air volume and secondary air volume.

[0009] Optionally, the multi-partition hybrid prediction model is a recurrent neural network model or a long short-term memory network model.

[0010] Optionally, it also includes an SCR multi-zone control system, which includes a zoned ammonia injection control box, a main pneumatic control valve, and multiple zoned pneumatic control valves. The input of the zoned ammonia injection control box is connected to the multi-zone control strategy module, and the output of the zoned ammonia injection control box is connected to the main pneumatic control valve and each zoned pneumatic control valve. Each zoned pneumatic control valve is connected to the SCR reactor. The zoned ammonia injection control box is used to convert the total amount of reducing agent input by the multi-zone control strategy module and the amount of reducing agent in each zone into a response electrical signal and transmit it to the main pneumatic control valve and each zoned pneumatic control valve to control the opening of the main pneumatic control valve and each zoned pneumatic control valve.

[0011] Optionally, the temperature measuring instrument is a matrix temperature measuring instrument.

[0012] The present invention also discloses a coal-fired flue gas SCR denitration control method based on zoned ammonia injection. The coal-fired flue gas SCR denitration intelligent control method applies the coal-fired flue gas SCR denitration intelligent control system, including:

[0013] Obtain boiler operation data in real time;

[0014] Obtain the flow rate and temperature of each partition of the SCR inlet flue in real time;

[0015] Real-time acquisition of NO in each zone of the SCR outlet flue x concentration;

[0016] The boiler operation data, the flow rate of each partition and the temperature of each partition are input into the multi-partition hybrid prediction model, and the NO x The concentration prediction value is based on the NO concentration of each partition obtained from the SCR outlet flue. x The multi-partition hybrid prediction model is modified according to the concentration;

[0017] According to the NO of each partition output by the multi-partition hybrid prediction model x Concentration and set outlet NO x Concentration, determine the total amount of NH3 and the amount of NH3 in each zone;

[0018] The total amount of reducing agent input to the SCR injection device and the amount of reducing agent in each partition are determined according to the total amount of NH3 and the amount of NH3 in each partition.

[0019] Optionally, the boiler operation data includes load, coal feed rate, oxygen amount, temperature, primary air volume and secondary air volume.

[0020] Optionally, the multi-partition hybrid prediction model is a recurrent neural network model or a long short-term memory network model.

[0021] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0022] The present invention provides a coal-fired flue gas SCR denitrification control system and method based on zoned ammonia injection, which uses a multi-zone hybrid prediction model to predict the NO in each zone of the SCR outlet flue. x The concentration is predicted by using the NO concentration of each partition collected in real time from the SCR outlet flue. x The multi-zone mixed prediction model is modified according to the NO concentration of each zone. x The predicted value of the concentration controls the total amount of reducing agent input into the SCR injection device and the amount of reducing agent in each partition, thereby improving the accuracy of the partition control, thereby reducing the problem of excessive ammonia injection and reducing the operating cost of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0024] Figure 1 This is a schematic diagram of the structure of a coal-fired flue gas SCR denitration control system based on zoned ammonia injection according to the present invention;

[0025] Figure 2 The present invention is a flow chart of a method for SCR denitrification of coal-fired flue gas based on zoned ammonia injection. DETAILED DESCRIPTION

[0026] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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 creative work are within the scope of protection of the present invention.

[0027] The purpose of the present invention is to provide a coal-fired flue gas SCR denitration control system and method based on zoned ammonia injection, which reduces the operating cost of the system.

[0028] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0029] Figure 1 The schematic diagram of the structure of the SCR denitrification control system of coal-fired flue gas based on zoned ammonia injection of the present invention is as follows: Figure 1 As shown, a coal-fired flue gas SCR denitrification control system based on zoned ammonia injection includes: an SCR zoned measurement system, an SCR multi-zone hybrid prediction control system and an SCR multi-zone control system.

[0030] The SCR partition measurement system includes a boiler, an SCR inlet flue, an SCR injection device, an SCR reactor and an SCR outlet flue connected in sequence; a flow measuring instrument and a temperature measuring instrument are set in the SCR inlet flue. The flow measuring instrument is used to measure the flue gas flow of each partition in real time, and the temperature measuring instrument is used to measure the temperature of each partition in real time. The SCR outlet flue is provided with a NO x Measuring instrument, NO x (Nitrogen oxides) measuring instrument is used to measure NO in each partition in real time x concentration.

[0031] The SCR multi-partition hybrid prediction control system includes a multi-partition hybrid prediction model and a multi-partition control strategy module. The multi-partition hybrid prediction model is used to predict the NOx of each partition based on the boiler operation data, the temperature of each partition and the flue gas flow of each partition. x Concentration; NO x The measuring instrument measures the NO of each partition in real time x The concentration is used to correct the multi-partition hybrid prediction model; the multi-partition control strategy module is used to adjust the NO concentration of each partition according to the output of the multi-partition hybrid prediction model. x Concentration and set outlet NO x The total amount of NH3 and the amount of NH3 in each partition are determined, and the total amount of reducing agent input to the SCR injection device and the amount of reducing agent in each partition are determined according to the total amount of NH3 and the amount of NH3 in each partition.

[0032] Boiler operation data include load, coal feed, oxygen content, temperature, primary air volume and secondary air volume.

[0033] The multi-partition hybrid prediction model is a recurrent neural network model or a long short-term memory network model.

[0034] Reducing agents mainly include liquid ammonia, urea and ammonia water.

[0035] The SCR multi-zone control system includes a zoned ammonia injection control box, a main pneumatic control valve, and multiple zoned pneumatic control valves. The input of the zoned ammonia injection control box is connected to the multi-zone control strategy module, and the output of the zoned ammonia injection control box is connected to the main pneumatic control valve and each zoned pneumatic control valve. Each zoned pneumatic control valve is connected to the SCR reactor. The zoned ammonia injection control box is used to convert the total amount of reducing agent input by the multi-zone control strategy module and the amount of reducing agent in each zone into a response electrical signal and transmit it to the main pneumatic control valve and each zoned pneumatic control valve to control the opening of the main pneumatic control valve and each zoned pneumatic control valve.

[0036] The temperature measuring instrument is a matrix temperature measuring instrument.

[0037] The following is a detailed description of the SCR denitrification control system for coal-fired flue gas based on zoned ammonia injection of the present invention. The SCR denitrification control system for coal-fired flue gas based on zoned ammonia injection includes an SCR zoned measurement system, an SCR multi-zone hybrid predictive control system and an SCR multi-zone intelligent control system (SCR multi-zone control system).

[0038] The SCR partition measurement system includes boiler, SCR inlet flue, SCR injection device, SCR reactor and SCR outlet flue in sequence. The boiler measurement data (boiler operation data) are load, coal feed, oxygen, temperature and primary / secondary air volume; the SCR injection device inlet flue (SCR inlet flue) is equipped with partition flow and partition temperature measuring instruments to measure the flue gas flow and temperature of different partitions in the flue in real time; the SCR outlet flue is equipped with partition NO x Measuring instrument, real-time measurement of NO in each area of ​​the flue cross section at the denitrification outlet x concentration.

[0039] The SCR multi-partition hybrid prediction control system includes a multi-partition hybrid prediction model and a multi-partition intelligent control strategy (multi-partition control strategy module). The multi-partition hybrid prediction model is based on the RNN (Recurrent Neural Network) and LSTM (Long Short-Term Memory) algorithms to establish a hybrid model that couples the SCR denitrification process mechanism with the denitrification operating parameters (real-time parameters provided by the above SCR partition measurements). It can achieve NO in flue gas. x Total concentration and partition NO xPrediction of the content of exported NO x The concentration is used as a calibration value to correct the model and realize the online optimization of the multi-partition hybrid prediction model. The online optimization of the multi-partition hybrid prediction model specifically includes the following steps:

[0040] (1) Using the historical operation data of the denitrification system (SCR system), set the SCR outlet flue NO x The concentration is the target value, and the real-time parameters provided by the SCR partition measurement system are the input values. Based on the RNN and LSTM neural network algorithms, the NO concentration in the SCR outlet flue is predicted. x The historical operation data includes boiler operation data, flow rate of each zone of SCR inlet flue, temperature of each zone of SCR inlet flue, and NO concentration of each zone of SCR outlet flue. x concentration.

[0041] (2) Adjust the weights of the operating parameters of the historical data in the hybrid model, and use the neural network algorithm to continuously perform the prediction-comparison-adjustment process to establish an accurate multi-partition hybrid prediction model.

[0042] (3) Input the real-time operation data into the multi-partition hybrid prediction model to obtain the NO in flue gas x Total concentration and partition NO x The real-time prediction value of the content will export NO x The real-time concentration data is used as the calibration value, and the multi-partition hybrid prediction model fine-tunes the operating parameter weights according to the calibration value to achieve continuous optimization of the model during operation.

[0043] The multi-zone intelligent control strategy is based on the predicted NO x Total and partition NO x content, based on the set outlet NO x The value is taken as the target, and the total amount of NH3 and the amount of NH3 in each partition are calculated, so as to calculate the amount of reducing agent. The calculation formula for the total amount of NH3 and the amount of NH3 in each partition is as follows:

[0044] m NH3 =δ×(m NOx -m out )×Q g / M NOx / 1000

[0045] m NH3,n =δ n ×(m NOx，n -m out )×Q g，n / M NOx / 1000

[0046] Where: mNH3 is the total amount of NH3, m NH3,n is the amount of NH3 in the nth partition, m out The set SCR outlet flue NO x value; δ is the ammonia nitrogen molar ratio, δ n is the molar ratio of ammonia nitrogen in the nth partition; Q g is the total flue gas volume, Q g，n is the smoke volume of the nth partition; M NOx NO x Molar mass; m NOx For the predicted NO x Total amount, m NOx，n The predicted NO of the nth partition x quantity.

[0047] The SCR multi-zone intelligent control system includes a zoned ammonia injection control box, a main pneumatic regulating valve and a zoned pneumatic regulating valve. The staged ammonia injection control box converts the total amount of reducing agent and the zoned usage obtained by the above control strategy into a corresponding electrical signal and transmits it to the main pneumatic regulating valve and the zoned pneumatic regulating valve to control the valve opening and realize intelligent regulation of the denitration system zones.

[0048] The present invention adds a matrix temperature and flow meter in front of the ammonia injection grid to obtain accurate temperature and flue gas flow data of each partition in front of the ammonia injection grid, and installs a partition NO x Measuring instruments provide data support for the algorithm model; the ammonia injection regulating valves in each zone are upgraded from manual valves to pneumatic regulating valves. Real-time data collection is carried out on the pollutant concentration in each zone after the SCR reactor, the flue gas temperature and flow in each zone before the ammonia injection grid, and the combustion of the front-end boiler.

[0049] Using boiler combustion data and data such as the temperature and flue gas flow rate of each zone before the ammonia injection grid, a multi-zone hybrid prediction model established by a neural network algorithm is used to predict NO x The concentration is predicted and analyzed, and the total ammonia amount is accurately controlled through the total amount control module in the partition intelligent control strategy.

[0050] By utilizing data such as the flue gas temperature and flow rate of each zone in front of the ammonia injection grid and the pollutant concentration of each zone after the SCR reactor, the zone control module in the zone intelligent control strategy is used to accurately control the ammonia amount in each zone, changing the previous dilemma of lagging measurement and control, and realizing functions such as advanced prediction, precise control and online adjustment.

[0051] The optimized opening parameters of each regulating valve are transmitted to the DCS through communication. After disturbance-free switching, the ammonia injection amount of each partition is adjusted online in real time, and edge computing is used to optimize the model parameters in real time.

[0052] Figure 2The present invention is a flow chart of a method for controlling SCR denitrification of coal-fired flue gas based on zoned ammonia injection, as shown in FIG. Figure 2 As shown, a method for controlling SCR denitration of coal-fired flue gas based on zoned ammonia injection is applied. A method for controlling SCR denitration of coal-fired flue gas based on zoned ammonia injection is applied. A method for controlling SCR denitration of coal-fired flue gas based on zoned ammonia injection comprises the following steps:

[0053] Step 201: Obtain boiler operation data in real time.

[0054] Step 202: Obtain the flow rate and temperature of each partition of the SCR inlet flue in real time.

[0055] Step 203: Obtain NO in each zone of the SCR outlet flue in real time x concentration.

[0056] Step 204: Input the boiler operation data, the flow rate of each partition and the temperature of each partition into the multi-partition hybrid prediction model, and output the NO x Concentration prediction value, using the NO concentration of each zone obtained from the SCR outlet flue x The multi-partition mixed prediction model was corrected based on the concentration.

[0057] Step 205: The NO of each partition output by the multi-partition hybrid prediction model x Concentration and set outlet NO x Concentration, determine the total amount of NH3 and the amount of NH3 in each zone.

[0058] Step 206: Determine the total amount of reducing agent input to the SCR injection device and the amount of reducing agent in each partition according to the total amount of NH3 and the amount of NH3 in each partition.

[0059] Boiler operation data include load, coal feed, oxygen content, temperature, primary air volume and secondary air volume.

[0060] The multi-partition hybrid prediction model is a recurrent neural network model or a long short-term memory network model.

[0061] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0062] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A coal-fired flue gas SCR denitrification control system based on zoned ammonia injection, characterized in that: include: SCR partition measurement system and SCR multi-partition hybrid predictive control system; The SCR partition measurement system comprises an SCR inlet flue, an SCR injection device, an SCR reactor and an SCR outlet flue connected in sequence; a flow measuring instrument and a temperature measuring instrument are arranged in the SCR inlet flue, wherein the flow measuring instrument is used to measure the flue gas flow of each partition in real time, and the temperature measuring instrument is used to measure the temperature of each partition in real time; and a NO x Measuring instrument, the NO x The measuring instrument is used to measure the NO in each partition in real time. x concentration; The SCR multi-zone hybrid prediction control system includes a multi-zone hybrid prediction model and a multi-zone control strategy module. The multi-zone hybrid prediction model is used to predict the NO in each zone according to the boiler operation data, the temperature of each zone and the flue gas flow of each zone. x concentration; the NO x The measuring instrument measures the NO of each partition in real time x The concentration is used to modify the multi-partition hybrid prediction model; the multi-partition control strategy module is used to adjust the NO concentration of each partition according to the output of the multi-partition hybrid prediction model. x Concentration and set outlet NO x concentration, determine the total amount of NH3 and the amount of NH3 in each partition, and determine the total amount of reducing agent input to the SCR injection device and the amount of reducing agent in each partition according to the total amount of NH3 and the amount of NH3 in each partition; the boiler operation data includes load, coal feed rate, oxygen amount, temperature, primary air volume and secondary air volume; The multi-partition hybrid prediction model is a recurrent neural network model or a long short-term memory network model; The calculation formulas for the total amount of NH3 and the amount of NH3 in each zone are as follows: m NH3 =δ×(m NOx -m out )×Q g / M NOx / 1000 m NH3，n =δ n ×(m NOx，n -m out )×Q g，n / M NOx / 1000 Where: m NH3 is the total amount of NH3, m NH3，n is the amount of NH3 in the nth partition, m out The SCR outlet flue NO is set x value; δ is the ammonia nitrogen molar ratio, δ n is the molar ratio of ammonia nitrogen in the nth partition; Q g is the total flue gas volume, Q g，n is the smoke volume of the nth partition; M NOx NO x Molar mass; m NOx For the predicted NO x Total amount, m NOx，n The predicted NO of the nth partition x ; The online optimization of the multi-partition hybrid prediction model specifically includes the following steps: (1) Using the historical operation data of the denitrification system, set the SCR outlet flue NO x The concentration is the target value, and the real-time parameters provided by the SCR partition measurement system are the input values. Based on the RNN and LSTM neural network algorithms, the NO concentration in the SCR outlet flue is predicted. x The predicted value is compared with the target value to obtain the prediction accuracy; the historical operation data includes boiler operation data, flow rate of each partition of the SCR inlet flue, temperature of each partition of the SCR inlet flue, and NO concentration of each partition of the SCR outlet flue. x concentration; (2) Adjust the weights of the operating parameters of the historical data in the hybrid model, and use the neural network algorithm to continuously perform the process of prediction-comparison-adjustment to establish an accurate multi-partition hybrid prediction model; (3) Input the real-time operation data into the multi-partition hybrid prediction model to obtain the NO in flue gas x Total concentration and partition NO x The real-time prediction value of the content will export NO x The real-time concentration data is used as the calibration value, and the multi-partition hybrid prediction model fine-tunes the operating parameter weights according to the calibration value to achieve continuous optimization of the model during operation.

2. The SCR denitrification control system for coal-fired flue gas based on zoned ammonia injection according to claim 1 is characterized in that: It also includes an SCR multi-zone control system, which includes a zoned ammonia injection control box, a main pneumatic control valve, and multiple zoned pneumatic control valves. The input of the zoned ammonia injection control box is connected to the multi-zone control strategy module, and the output of the zoned ammonia injection control box is connected to the main pneumatic control valve and each zoned pneumatic control valve. Each zoned pneumatic control valve is connected to the SCR reactor. The zoned ammonia injection control box is used to convert the total amount of reducing agent input by the multi-zone control strategy module and the amount of reducing agent in each zone into a response electrical signal and transmit it to the main pneumatic control valve and each zoned pneumatic control valve to control the opening of the main pneumatic control valve and each zoned pneumatic control valve.

3. The SCR denitrification control system for coal-fired flue gas based on zoned ammonia injection according to claim 1 is characterized in that: The temperature measuring instrument is a matrix temperature measuring instrument.

4. A method for controlling SCR denitrification of coal-fired flue gas based on zoned ammonia injection, characterized in that: The SCR denitration control method for coal-fired flue gas based on zoned ammonia injection applies the SCR denitration control system for coal-fired flue gas based on zoned ammonia injection according to any one of claims 1 to 3, comprising: Obtain boiler operation data in real time; Obtain the flow rate and temperature of each partition of the SCR inlet flue in real time; Obtain NO in each zone of the SCR outlet flue in real time x concentration; The boiler operation data, the flow rate of each partition and the temperature of each partition are input into the multi-partition hybrid prediction model, and the NO x The concentration prediction value is based on the NO concentration of each partition obtained from the SCR outlet flue. x The multi-zone hybrid prediction model is corrected by the concentration; the boiler operation data includes load, coal feed, oxygen content, temperature, primary air volume and secondary air volume; According to the NO of each partition output by the multi-partition hybrid prediction model x Concentration and set outlet NO x Concentration, determine the total amount of NH3 and the amount of NH3 in each zone; Determine the total amount of reducing agent input to the SCR injection device and the amount of reducing agent in each partition according to the total amount of NH3 and the amount of NH3 in each partition; The multi-partition hybrid prediction model is a recurrent neural network model or a long short-term memory network model; The calculation formulas for the total amount of NH3 and the amount of NH3 in each zone are as follows: m NH3 =δ×(m NOx -m out )×Q g / M NOx / 1000 m NH3,n =δ n ×(m NOx,n -m out )×Q g,n / M NOx / 1000 Where: m NH3 is the total amount of NH3, m NH3,n is the amount of NH3 in the nth partition, m out The SCR outlet flue NO is set x value; δ is the molar ratio of ammonia nitrogen, δ n is the molar ratio of ammonia nitrogen in the nth partition; Q g is the total flue gas volume, Q g,n is the smoke volume of the nth partition; M NOx NO x The molar mass, m NOx For the predicted NO x Total amount, m NOx,n The predicted NO of the nth partition x ; The online optimization of the multi-partition hybrid prediction model specifically includes the following steps: (1) Using the historical operation data of the denitrification system, set the SCR outlet flue NO x The concentration is the target value, and the real-time parameters provided by the SCR partition measurement system are the input values. Based on the RNN and LSTM neural network algorithms, the NO concentration in the SCR outlet flue is predicted. x The predicted value is compared with the target value to obtain the prediction accuracy; the historical operation data includes boiler operation data, flow rate of each partition of the SCR inlet flue, temperature of each partition of the SCR inlet flue, and NO concentration of each partition of the SCR outlet flue. x concentration; (2) Adjust the weights of the operating parameters of the historical data in the hybrid model, and use the neural network algorithm to continuously perform the process of prediction-comparison-adjustment to establish an accurate multi-partition hybrid prediction model; (3) Input the real-time operation data into the multi-partition hybrid prediction model to obtain the NO in flue gas x Total concentration and partition NO x The real-time prediction value of the content will export NO x The real-time concentration data is used as the calibration value, and the multi-partition hybrid prediction model fine-tunes the operating parameter weights according to the calibration value to achieve continuous optimization of the model during operation.

Images