Automatic ammonia dosing method for boiler water of thermal power plant

Abstract

This invention relates to the field of automatic control technology, specifically to an automatic ammonia dosing method for boiler water in thermal power plants. The method includes the following steps: obtaining boiler water pH and volumetric flow rate to construct a basic ammonia flow rate; collecting fluid temperature and pressure data for model calculation to generate dosing volume requirements; calculating flow velocity based on inner diameter to extract Reynolds parameters and generate turbulent flow rate; mapping based on scattering coefficients to generate the target dosing volumetric flow rate; and analyzing operating frequency control parameters to generate automatic ammonia dosing commands. In this invention, by extracting multidimensional fluid state analysis kinematic viscosity to calculate local over-limit disturbances, it accurately compensates for flow rate dynamic deviations caused by pipeline resistance and turbulent dissipation. It introduces physical scattering coefficients to perform spatial diffusion mapping to generate feedforward gain, and performs pulse bandwidth conversion based on target requirements to establish underlying control commands. This effectively eliminates the supply distortion caused by multidimensional flow field attenuation and the drawbacks of traditional manual lag, ensuring continuous and constant feedwater quality and stable automatic dosing process.

Description

Technical Field

[0001] This invention relates to the field of automatic control technology, and in particular to an automatic ammonia preparation method for boiler water dosing in thermal power plants. Background Technology

[0002] The field of automatic control technology mainly studies how to make dynamic systems operate according to predetermined laws without direct human intervention. Its core issues include feedback control mechanisms, system state monitoring, parameter adjustment and actuator driving. By constructing differential equations to analyze the response characteristics of the system, formulating proportional-integral-derivative control strategies, and using thermal sensors to obtain real-time parameters, which are then handed over to programmable logic controllers to perform arithmetic operations. The output current commands drive electric actuators to change the physical state of the controlled object.

[0003] The traditional automatic ammonia dosing method for boiler water in thermal power plants refers to the adjustment of fluctuations in the hydrogen ion concentration of boiler water in the boiler water system of thermal power plants. Operators periodically open the valve of the sampling pipeline to extract boiler water liquid samples, use a glass electrode pH meter to test the pH of the liquid, and refer to the water quality standard reference table based on the obtained hydrogen ion concentration value to calculate the required mass of pure ammonia water. Then, the metering pump start button is manually pressed, and the eccentricity of the diaphragm metering pump connecting rod is changed by manually rotating the stroke adjustment handwheel, thereby setting the volumetric flow rate of ammonia solution to be continuously injected into the chemical storage tank and feedwater pipeline.

[0004] Traditional methods rely on manual periodic opening of sampling pipeline valves and detection using glass electrodes, resulting in significant time lag in obtaining hydrogen ion concentration data. Consulting static water quality standard comparison tables cannot respond to dynamic changes in the thermodynamics of the fluid inside the pipeline. Manual calculation of ammonia water quality lacks adaptation to real-time flow fluctuations. Mechanical adjustment of the metering pump stroke by rotating the handwheel has execution lag and low accuracy. It is difficult to couple fluid viscosity and turbulent scattering characteristics, resulting in poor uniformity of reagent mixing. The discretization of sampling and execution links makes it impossible to quickly track sudden load changes, resulting in insufficient boiler water quality stability and high reagent consumption. Summary of the Invention

[0005] To address the technical problems of traditional methods that rely on manual periodic opening of sampling pipeline valves and detection using glass electrodes, resulting in significant time lag in obtaining hydrogen ion concentration data, inability to respond to dynamic changes in the thermodynamics of the fluid inside the pipeline when consulting static water quality standard comparison tables, lack of adaptation to real-time flow fluctuations when manually calculating ammonia water quality, low accuracy and lag in mechanical adjustment of metering pump stroke using handwheels, poor uniformity of chemical mixing due to difficulty in coupling fluid viscosity and turbulent scattering characteristics, and inability to quickly track load changes due to discretization of sampling and execution links, leading to insufficient boiler water quality stability and high chemical consumption, this invention provides an automatic ammonia dosing method for boiler water in thermal power plants.

[0006] To achieve the above objectives, the present invention employs an automatic ammonia preparation method for boiler water dosing in thermal power plants, comprising the following steps: S1: Obtain the acidity and alkalinity and volumetric flow rate of the boiler water in the main feedwater pipeline of the thermal power plant. Based on the target acidity and alkalinity, perform water quality deviation parameter analysis on the boiler water acidity and alkalinity to extract the basic water quality deviation. Perform weighted flow allocation on the volumetric flow rate to construct the basic ammonia water flow rate. S2: Call the basic ammonia flow rate, collect the fluid temperature and pressure inside the pipeline, input them into the thermo-fluid dynamics model for kinematic viscosity coupling calculation, and generate the dosing volume requirement; S3: Call the required dosing volume, calculate the Reynolds equivalent parameters of the flow field based on the main pipeline inner diameter parameters, and perform over-limit disturbance deviation measurement and convection mixing compensation to generate turbulence disturbance compensation flow rate. S4: Call the turbulence disturbance compensation flow rate, obtain the physical scattering coefficient of the dosing nozzle, perform nonlinear spatial diffusion response mapping to extract the scattering compensation gain parameter, perform nozzle resistance feedforward gain modulation on the turbulence disturbance compensation flow rate, and generate the target dosing volumetric flow rate. S5: Call the target dosing volumetric flow rate to perform digital pulse bandwidth conversion to extract the duty cycle digital features, analyze the operating frequency control parameters of the ammonia dosing metering pump, and generate automatic ammonia dosing control instructions.

[0007] As a further aspect of the present invention, the basic ammonia flow rate includes the baseline dosage, theoretical ammonia consumption, and basic mass flow rate; the dosage volume requirement includes the apparent volume, corrected volume, and dosing pump discharge rate; the turbulence disturbance compensation flow rate includes the local head loss converted flow rate, the friction head loss converted flow rate, and the dissipation compensation flow rate component; the target dosing volume flow rate includes the rated flow rate, the actual volumetric flow rate, and the effective discharge rate; and the automatic ammonia distribution control command includes the motor speed, drive current, and start / stop status.

[0008] As a further aspect of the present invention, the specific steps of S1 are as follows: S101: Collect the acidity and alkalinity of boiler water in the main feedwater pipeline of the thermal power plant and the target acidity and alkalinity. Subtract the target acidity and alkalinity from the boiler water acidity and alkalinity to extract the acidity and alkalinity difference. Read the deviation zone. If it is determined that the acidity and alkalinity difference falls within the deviation zone, map the scale parameter. Multiply the acidity and alkalinity difference with the scale parameter to generate the basic water quality deviation. S102: Call the basic water quality deviation and collect the volumetric flow rate, perform the conversion based on the inverse ratio rule to obtain the deviation weight factor, establish the correlation between the deviation weight factor and the volumetric flow rate, configure the weighted frequency for the volumetric flow rate, extract the frequency feature components and perform a dot product operation with the volumetric flow rate to obtain the weighted flow rate allocation matrix. S103: Extract the main allocation element according to the weighted flow allocation matrix, read the reference flow rate constant and multiply it with the main allocation element to obtain the feedforward ammonia flux, detect the pipeline resistance constant, reconstruct the ammonia flow rate vector by dividing the feedforward ammonia flux by the resistance constant, extract the ammonia flow rate vector scalar, and establish the basic ammonia flow rate.

[0009] As a further aspect of the present invention, the specific steps of S2 are as follows: S201: Call the basic ammonia flow rate, collect the fluid temperature and fluid pressure, subtract the fluid temperature from the preset baseline point to obtain the temperature deviation, subtract the fluid pressure from the pressure reference threshold to extract the pressure deviation, and perform vector splicing to establish a parameter feature map. S202: Extract the scale sequence, perform inner product processing on the parameter feature map and the scale sequence to obtain the thermodynamic state distribution matrix, extract the state feature tensor based on the thermodynamic state distribution matrix and the fluid constant, and retrieve the scalar in the viscosity lookup table according to the state feature tensor to generate the kinematic viscosity coupling factor. S203: Call the kinematic viscosity coupling factor and the basic ammonia flow rate to perform convolution correlation to extract the volume reference value, detect the pipeline coefficient, perform linear mapping between the volume reference value and the pipeline coefficient to extract the corrected residual, and accumulate it with the volume reference value to generate the dosing volume requirement.

[0010] As a further aspect of the present invention, the execution vector splicing to establish a parameter feature map refers to obtaining the data dimension corresponding to the temperature deviation and the data dimension corresponding to the pressure gradient, performing filling and alignment processing on the temperature deviation and the pressure gradient based on the data dimension, splicing the filled and aligned temperature deviation and the filled and aligned pressure gradient along the channel direction to extract the initial feature matrix, and performing normalization mapping on the matrix element values ​​contained in the initial feature matrix to construct the parameter feature map. The pressure benchmark threshold refers to extracting historical fluid pressure records corresponding to the base ammonia flow rate, performing interval filtering and noise reduction on the historical fluid pressure records to extract a benchmark pressure sample set, arranging the pressure sample values ​​in the benchmark pressure sample set in ascending order to construct a pressure distribution vector, extracting the middle position value of the pressure distribution vector and the average scalar value of the pressure sample values ​​contained in the pressure distribution vector, and performing weighted calculation based on the middle position value and the average scalar value to generate the pressure benchmark threshold.

[0011] As a further aspect of the present invention, the specific steps of S3 are as follows: S301: Call the dosing volume requirement, collect the inner diameter parameters and cross-sectional spectrum of the main pipeline, align the dosing volume requirement with the cross-sectional spectrum according to the area to extract the profile matrix, project the profile matrix and the inner diameter parameters of the main pipeline to extract the flow field vector, calculate the ratio of the flow field vector with the viscosity, and generate the Reynolds equivalent parameters of the flow field. S302: Based on the Reynolds equivalent parameters of the flow field, detect the extreme values, compare the Reynolds equivalent parameters of the flow field with the extreme values ​​to extract the disturbance amount, obtain the roughness, map the disturbance amount and roughness through interpolation to extract the drag scalar, obtain the dissipation rate, fuse the drag scalar and the dissipation rate, and establish an over-limit disturbance deviation metric. S303: For the aforementioned over-limit disturbance deviation measurement and dosing volume requirement, obtain a mass transfer map, aggregate the over-limit disturbance deviation measurement and the mass transfer map by node to extract the gain factor, multiply the dosing volume requirement by the gain factor to extract the demand vector, and decouple the demand vector periodically according to the time sequence to obtain the turbulence disturbance compensation flow rate.

[0012] As a further aspect of the present invention, the specific steps of S4 are as follows: S401: Call the turbulence disturbance compensation flow rate, obtain the nozzle scattering coefficient, collect spatial grid nodes, perform a dot product operation between the turbulence disturbance compensation flow rate and the spatial grid nodes to extract the initial vector, use the nozzle scattering coefficient to perform a logarithmic mapping to extract the response gradient, and establish a spatial diffusion response tensor; S402: Based on the spatial diffusion response tensor, detect the attenuation factor and reduce the dimension of the spatial diffusion response tensor to extract the field scalar, collect the nozzle flow rate, correlate the field scalar with the nozzle flow rate to extract the dynamic component, obtain the gain constant, compensate the dynamic component based on the gain constant, and obtain the scattering compensation gain parameter. S403: For the scattering compensation gain parameter and the turbulence disturbance compensation flow rate, monitor the structural resistance, extract the feedforward coefficient based on the damping modulation of the scattering compensation gain parameter according to the structural resistance, obtain the correction base, calibrate and extract the feedforward vector in combination with the correction base, and use the feedforward vector to feedforward modulate the turbulence disturbance compensation flow rate to generate the target dosing volumetric flow rate.

[0013] As a further aspect of the present invention, the step of using the nozzle scattering coefficient to perform logarithmic mapping to extract the response gradient refers to obtaining the natural logarithmic value of the nozzle scattering coefficient, performing partial derivative processing on the natural logarithmic value of the nozzle scattering coefficient and the initial vector to generate a partial derivative vector, and determining the partial derivative vector as the response gradient.

[0014] As a further aspect of the present invention, the specific steps of S5 are as follows: S501: The target dosing volumetric flow rate is called and the preset digital reference period is used to calculate the ratio to obtain the bandwidth time domain scalar. The bandwidth time domain scalar is mapped to the digital duty cycle constant to extract the digital feature factor. The digital feature factor is multiplied by the time domain proportional parameter to unify the dimensions and generate the duty cycle digital feature. S502: Based on the duty cycle digital characteristics and the rated speed of the pump body, the base output frequency is extracted by multiplying the base output frequency with the frequency offset coefficient. The dynamic offset frequency is extracted by weighting the base output frequency with the frequency offset coefficient. The dynamic offset frequency is compared with the system frequency limit value to perform a limiting operation and obtain the operating frequency control parameters. S503: For the operation frequency regulation parameters, the data frame is extracted from the load field of the header of the control message. The data frame and the security check mask are XORed to extract the tail of the transmission message. The device addressing identifier is concatenated with the data frame and the tail of the transmission message to establish an automatic ammonia distribution control command.

[0015] As a further aspect of the present invention, the step of performing an XOR operation on the data frame and the security check mask to extract the tail of the transmission message refers to splitting the data frame to obtain the binary sequence of the data frame and splitting the security check mask to obtain the binary sequence of the mask, performing a bit-by-bit alignment operation on the binary sequence of the data frame and the binary sequence of the mask and performing an XOR operation to compare and extract the comparison result, combining the comparison result according to the alignment position to generate a check sequence, and using the check sequence as the tail of the transmission message.

[0016] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In this invention, the acid-base and flow deviation of the fluid in the main water supply pipeline are extracted to establish the basic ammonia flow rate requirement. Temperature and pressure variables are collected simultaneously and analyzed by a thermodynamic model to determine the fluid motion viscosity coupling characteristics. The Reynolds equivalent parameters of the flow field are calculated by combining the inner diameter scale and cross-sectional projection, and local over-limit disturbances are measured. This accurately compensates for the flow rate deviation caused by pipeline resistance and turbulent dissipation. The physical scattering coefficient of the nozzle is introduced to perform spatial diffusion response mapping to generate feedforward gain. According to the target requirements, pulse bandwidth conversion is performed to establish a digital control command for ammonia distribution. This effectively eliminates the supply distortion and artificial lag caused by the dynamic attenuation of the multi-dimensional flow field, ensuring the continuous and constant water quality of the water supply medium and the stability of the dosing process. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the accompanying drawings without creative effort.

[0018] Figure 1 This is a schematic diagram of the steps of the present invention; Figure 2 This is a detailed schematic diagram of S1 of the present invention; Figure 3 This is a detailed schematic diagram of S2 of the present invention; Figure 4 This is a detailed schematic diagram of S3 of the present invention; Figure 5 This is a detailed schematic diagram of S4 of the present invention; Figure 6 This is a detailed schematic diagram of S5 of the present invention. Detailed Implementation

[0019] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0020] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0021] Please see Figure 1 This invention provides an automatic ammonia preparation method for boiler water dosing in thermal power plants, comprising the following steps: S1: Obtain the acidity and alkalinity and volumetric flow rate of the boiler water in the main feedwater pipeline of the thermal power plant. Based on the target acidity and alkalinity, perform water quality deviation parameter analysis on the boiler water acidity and alkalinity to extract the basic water quality deviation. Perform weighted flow allocation on the volumetric flow rate to construct the basic ammonia water flow rate. S2: Call the basic ammonia flow rate, collect the fluid temperature and pressure inside the pipeline, input them into the thermo-fluid dynamics model for kinematic viscosity coupling calculation, and generate the dosing volume requirement; S3: Call the dosing volume requirement, extract the Reynolds equivalent parameters of the flow field by spatial scale velocity conversion based on the inner diameter parameters of the main pipeline, and perform over-limit disturbance deviation measurement and convection mixing compensation to generate turbulence disturbance compensation flow rate. S4: Call the turbulence disturbance compensation flow rate, obtain the physical scattering coefficient of the dosing nozzle, perform nonlinear spatial diffusion response mapping to extract the scattering compensation gain parameter, perform nozzle resistance feedforward gain modulation on the turbulence disturbance compensation flow rate, and generate the target dosing volumetric flow rate. S5: Call the target dosing volumetric flow rate to perform digital pulse bandwidth conversion to extract the duty cycle digital features, analyze the operating frequency control parameters of the ammonia dosing metering pump, and generate automatic ammonia dosing control instructions.

[0022] The basic ammonia flow rate includes the baseline dosage, theoretical ammonia consumption, and basic mass flow rate. The dosing volume requirement includes the apparent volume, corrected volume, and dosing pump discharge. The turbulence disturbance compensation flow rate includes the local head loss equivalent flow rate, the friction head loss equivalent flow rate, and the dissipation compensation flow rate component. The target dosing volume flow rate includes the rated flow rate, the actual volumetric flow rate, and the effective discharge rate. The automatic ammonia distribution control commands include the motor speed, drive current, and start / stop status.

[0023] Please see Figure 2 The specific steps of S1 are as follows: S101: Collect the acidity and alkalinity of boiler water in the main feedwater pipeline of the thermal power plant and the target acidity and alkalinity. Subtract the target acidity and alkalinity from the boiler water acidity and alkalinity to extract the acidity and alkalinity difference. Read the deviation zone. If it is determined that the acidity and alkalinity difference falls within the deviation zone, map the scale parameter. Multiply the acidity and alkalinity difference with the scale parameter to generate the basic water quality deviation. Acidity and alkalinity sensors deployed on the inner wall of the main feedwater pipeline acquire the boiler water's acidity and alkalinity in real time at a sampling frequency of 10 Hz. The acquired data undergoes median filtering to remove high-frequency noise. Simultaneously, the target acidity and alkalinity for the corresponding operating condition is retrieved from the central control database. This target acidity and alkalinity is set by reading the current boiler operating load parameters and combining them with historical best water quality records. The acquired boiler water acidity and alkalinity are subtracted from the target acidity and alkalinity to obtain the acidity and alkalinity difference representing the current water quality deviation. For this acidity and alkalinity difference, a preset deviation range is read from the configuration list. The value range of this deviation range is set with reference to the continuous fluctuation data of water quality compliance over the past 30 days. The absolute extreme values ​​extracted from the fluctuation data are summed and divided by the total number of values ​​(20) to calculate the upper and lower limit absolute values ​​of the range with an absolute value of 0.5. A direct numerical comparison operation is then performed to determine whether the absolute value of the acidity and alkalinity difference falls within the 0.5 range. If the deviation falls within the specified boundary, a scale parameter mapping operation is triggered. The dynamic scale parameter corresponding to the pH difference is retrieved from a pre-built parameter lookup table. A multiplication operation is then performed, multiplying the pH difference by the mapped scale parameter to generate the basic water quality deviation. In the actual operation scenario, the pre-processed current boiler water pH is collected as 9.2, while the target pH retrieved from the database is 9.0. Subtracting 9.0 from 9.2 yields a pH difference of 0.2. The deviation boundary is read from a range of values ​​with upper and lower limits of 0.5. After comparison, the absolute value of 0.2 is less than 0.5, indicating the deviation falls within the boundary. Based on this 0.2 difference, the corresponding scale parameter 1.5 is extracted from the parameter lookup table using linear interpolation. A multiplication operation is then performed, multiplying the pH difference of 0.2 by the scale parameter 1.5 to derive the basic water quality deviation of 0.3. The numerical result of 0.3 represents the baseline adjustment required under the current minor fluctuations in water quality, and is used as core data to proceed to the next processing step. The advantage of this calculation logic is that by establishing the product relationship between the acid-base difference and dynamic scale parameters, it quantifies the true degree of deviation under minor fluctuations. The calculated baseline water quality deviation is then applied to the physical chemical dosing network for verification.

[0024] S102: Call the basic water quality deviation and collect the volumetric flow rate, perform the transformation based on the inverse ratio rule to obtain the deviation weight factor, establish the correlation between the deviation weight factor and the volumetric flow rate, configure the weighted frequency for the volumetric flow rate, extract the frequency feature components and perform a dot product operation with the volumetric flow rate to obtain the weighted flow rate allocation matrix. Upon receiving the baseline water quality deviation from the aforementioned steps, the volumetric flow rate for the current period is simultaneously collected using an ultrasonic flow meter installed on the main pipeline. Abnormal peak values ​​are removed from the collected flow data. A conversion operation is performed based on an inverse proportionality rule to obtain the deviation weighting factor. This conversion operation specifically involves dividing a preset static weighting constant by the absolute value of the baseline water quality deviation to obtain a dynamic parameter inversely proportional to the degree of deviation. A data association is established between the deviation weighting factor and the volumetric flow rate, combining them into a two-dimensional data pair. For this data pair, a weighted frequency is configured for the volumetric flow rate. Specifically, based on the flow range of the volumetric flow rate, the corresponding discrete numerical frequency is matched from the frequency mapping library. For every 50 cubic meters per hour increase beyond the baseline flow rate, the weighted frequency increases by a fixed 2. The frequency characteristic components of this weighted frequency are extracted, and the extracted frequency characteristic components, the deviation weighting factor, and the volumetric flow rate are multiplied together to generate a weighted flow allocation matrix. Substituting the actual monitoring values, assuming the received baseline water quality deviation is 0.3 and the volumetric flow rate after cleaning is 200 cubic meters per hour. A static weighting constant of 0.6 is set. Dividing 0.6 by the basic water quality deviation of 0.3 yields a deviation weighting factor of 2.0. A combined association is established between the deviation weighting factor 2.0 and the volumetric flow rate of 200. For the volumetric flow rate of 200, according to the matching rules, the initial frequency corresponding to the basic flow rate of 100 is 3. The additional 100 cubic meters per hour increases the weighted frequency by 4, resulting in a weighted frequency of 7. The weighted frequency of 7 is extracted as a feature component. It, the deviation weighting factor 2.0, and the volumetric flow rate of 200 are multiplied together. For a 1D scalar scenario, this multiplication degenerates into a simple multiplication. Multiplying 7, 2.0, and 200 yields a single-element baseline value of 2800 for the weighted flow allocation matrix. This result, 2800, is the core element of the initially generated weighted flow allocation matrix, representing the global dosing demand scale after dual modulation by flow rate and deviation. The advantage of this operational logic is that by introducing an inverse proportional rule and a dot product modulation of weighted frequencies, static water quality deviation is converted into a distribution matrix that is strongly correlated with real-time flow.

[0025] S103: Extract the main allocation element based on the weighted flow allocation matrix, read the reference flow rate constant and multiply it with the main allocation element to obtain the feedforward ammonia flux, detect the pipeline resistance constant, reconstruct the ammonia flow rate vector by dividing the feedforward ammonia flux by the resistance constant, extract the ammonia flow rate vector scalar, and establish the basic ammonia flow rate. For the generated weighted flow allocation matrix, the internal data elements are traversed row by row, and the diagonal component with the largest value is extracted as the main allocation primitive. The pre-calibrated reference flow rate constant is read from the configuration register; this constant is obtained by averaging the historical flow rates of the best-performing section during long-term operation. A multiplication operation is performed, multiplying the main allocation primitive by the reference flow rate constant to obtain the feedforward ammonia flux. Subsequently, the pipeline resistance constant is detected by the differential pressure transmitter at the end of the pipeline network. The logic for obtaining this constant is to read the real-time inlet and outlet pressure difference and divide it by a fixed pipe length coefficient. A division operation is performed, dividing the calculated feedforward ammonia flux by the detected pipeline resistance constant to reconstruct the ammonia flow rate vector containing velocity and direction information. The absolute value of this ammonia flow rate vector is extracted as scalar data to establish the base ammonia flow rate. In the example calculation, the extracted value of the main allocation primitive in the weighted flow allocation matrix is ​​set to 2800. The set reference flow rate constant is read from the register as 0.05. A multiplication operation is performed, multiplying the main distribution unit 2800 by the reference flow rate constant 0.05 to calculate the feedforward ammonia flux as 140. The differential pressure transmitter detects an inlet / outlet pressure difference of 20. Assuming a fixed pipe length coefficient of 10, a division operation is performed to obtain a pipeline resistance constant of 2.0. A division operation is then performed, dividing the feedforward ammonia flux 140 by the pipeline resistance constant 2.0, reconstructing the ammonia flow rate vector to a value of 70. The pure numerical scalar value of 70 is extracted from this vector to establish a base ammonia flow rate value of 70. This value of 70 serves as the baseline dosing flow rate, unaffected by subsequent random disturbances, and is used as the downstream logic for the lower-level control core data flow. The advantage of this operational logic is that, through the product of the main distribution unit and the reference flow rate, and the reconstruction by division using the pipeline resistance, the weakening effect of physical pipeline obstacles on the dosing flux is accurately eliminated.

[0026] Please see Figure 3 The specific steps of S2 are as follows: S201: Call the basic ammonia flow rate, collect the fluid temperature and fluid pressure, subtract the fluid temperature from the preset baseline point to obtain the temperature deviation, subtract the fluid pressure from the pressure reference threshold to extract the pressure deviation, and perform vector splicing to establish a parameter feature map. The system reads the generated baseline ammonia flow rate and simultaneously collects current fluid temperature and pressure data via thermocouples and pressure transmitters. The collected data undergoes smoothing filtering. The collected fluid temperature is subtracted from the preset baseline value in memory to obtain the temperature deviation. The fluid pressure is subtracted from a pressure benchmark threshold to extract the pressure deviation. The specific process for determining this pressure benchmark threshold is as follows: First, based on the baseline ammonia flow rate, historical fluid pressure records for 60 consecutive sampling periods under the corresponding operating conditions are retrieved from the database. This data is then subjected to a moving average filter with a range of 5 for noise reduction, and abrupt changes are removed to extract a benchmark pressure sample set. Next, the pressure sample values ​​within the set are sorted in ascending order to construct a 1D pressure distribution vector. The value at the middle position of this distribution vector is extracted, and then the average scalar value obtained by dividing the sum of the sample values ​​by the total number is extracted. The arithmetic mean of these two values ​​is used as the final pressure benchmark threshold. The process involves vector concatenation to obtain one data dimension occupied by the temperature deviation and another by the pressure gradient. Zero-padding is then applied to align these dimensions. The aligned temperature deviation and pressure gradient are then concatenated end-to-end along the channel direction to extract an initial feature matrix containing both pressure and temperature attributes. Finally, minimax normalization is applied to the elements of the matrix to construct a parametric feature map.

[0027] Table 1: Historical Fluid Pressure Record Data Table 1 lists historical fluid pressure data for five cycles. Assuming the current fluid temperature is 150°C and the preset baseline is 120°C, the temperature deviation is 30. Assuming the calculated pressure baseline threshold is 104 and the current fluid pressure is 110°C, the pressure deviation is 6. The values ​​30 and 6 are padded with zeros and concatenated to form the initial feature matrix. The maximum extreme value in the feature matrix is ​​30, and the minimum is 0. Normalization is applied to the value 6 to obtain 0.2, and the value 30 is mapped to 1.0. Based on this, a parametric feature map containing the elements 0.2 and 1.0 is constructed.

[0028] S202: Extract the scale sequence, perform inner product processing on the parameter feature map and the scale sequence to obtain the thermodynamic state distribution matrix, extract the state feature tensor based on the thermodynamic state distribution matrix and fluid constant, and retrieve the scalar in the viscosity lookup table according to the state feature tensor to generate the kinematic viscosity coupling factor. The system receives the constructed parameter feature map and extracts a preset scale sequence from a built-in coefficient library. This scale sequence consists of a set of fixed values ​​representing decreasing environmental sensitivity. For each element value in the parameter feature map, an inner product operation is performed with the corresponding value in the scale sequence. This involves multiplying each element and summing the results to obtain a thermodynamic state distribution matrix representing macroscopic thermodynamic characteristics. Based on the obtained thermodynamic state distribution matrix and the fluid constants read from the fluid physical property table, a state feature tensor is extracted. Specifically, the result of the thermodynamic state distribution matrix is ​​multiplied by the fluid constants to generate a state feature tensor with a defined physical dimension. According to the specific values ​​of the generated state feature tensor, a comparison is performed item by item in a pre-imported viscosity lookup table to find the scalar with the smallest difference from the state feature tensor value, generating a kinematic viscosity coupling factor representing the current fluid viscous drag characteristics. In the specific numerical extrapolation, the input parameter feature map contains normalized values ​​of 0.2 and 1.0. The corresponding scale sequence values ​​extracted from the coefficient library are 0.5 and 0.8, respectively. The inner product processing is applied to the parameter feature map and scale sequence. First, 0.2 is multiplied by 0.5 to obtain 0.1, and 1.0 is multiplied by 0.8 to obtain 0.8. Then, 0.1 and 0.8 are summed to obtain the core scalar result of the thermodynamic state distribution matrix, which is 0.9. The fluid constant under the current operating condition is set to 10. The distribution matrix result 0.9 is multiplied by the fluid constant 10 to extract the state feature tensor value of 9.0. Based on the value 9.0, a nearest-neighbor matching search is performed in the viscosity lookup table to find the scalar value with the smallest difference, which is 1.2. Finally, the kinematic viscosity coupling factor is generated as 1.2. This calculated result of 1.2 directly reflects the fluid viscous drag amplification factor under the combined effect of current temperature and pressure, and serves as the core multiplier for subsequent flow rate correction.

[0029] S203: Call the kinematic viscosity coupling factor and the basic ammonia flow rate to perform convolution correlation to extract the volume reference value, detect the pipeline coefficient, perform linear mapping between the volume reference value and the pipeline coefficient to extract the corrected residual, and accumulate it with the volume reference value to generate the dosing volume requirement. The generated kinematic viscosity coupling factor and the base ammonia flow rate generated in the early stage are retrieved and convolved to extract the volumetric baseline value. This convolutional correlation, at the current single-value discrete data level, is a direct multiplication of the two values ​​with an added unit smoothing coefficient. The network coefficient of the current physical network is detected by a network feature scanning sensor; this coefficient is directly determined by the geometric ratio of the pipe wall average roughness parameter to the pipe inner diameter parameter. A linear mapping operation is performed between the extracted volumetric baseline value and the detected network coefficient. Specifically, the volumetric baseline value is multiplied by the network coefficient, and then a preset fixed offset for pipe noise floor is subtracted to extract the correction residual used to compensate for system wear errors. Finally, the calculated correction residual is summed with the original volumetric baseline value to generate the dosing volume requirement, encompassing the combined effects of environmental and fluid physical properties. A calculation example is performed using actual monitoring data and preset values, assuming the retrieved kinematic viscosity coupling factor is 1.2 and the base ammonia flow rate is 70. Multiplying the two values ​​(1.2 multiplied by 70) yields 84. With a smoothing coefficient of 1, the volume reference value is extracted as 84. The average pipe wall roughness parameter is then measured to be 0.02, and the pipe inner diameter parameter is 100. The ratio is calculated to obtain the actual pipe network coefficient of 0.0002. For ease of calculation, an equivalent pipe network coefficient of 0.1 is used as the amplified reference. A linear mapping operation is performed, multiplying the volume reference value 84 by the equivalent pipe network coefficient 0.1 to obtain 8.4. A fixed offset for pipe noise floor is set to 2.4. Subtracting 2.4 from 8.4 yields a corrected residual of 6.0. The corrected residual of 6.0 is then added to the volume reference value 84, resulting in 90. This result, 90 cubic centimeters per second, represents the final generated dosing volume requirement, forming the basic action target line for subsequent control nodes.

[0030] Please see Figure 4 The specific steps of S3 are as follows: S301: Call the dosing volume requirement, collect the main pipeline inner diameter parameters and cross-sectional spectrum, align the dosing volume requirement with the cross-sectional spectrum according to the area to extract the profile matrix, project the profile matrix and the main pipeline inner diameter parameters to extract the flow field vector, calculate the ratio of the flow field vector with the viscosity, and generate the flow field Reynolds equivalent parameter. The chemical dosing volume requirement is invoked, and simultaneously, laser rangefinders and acoustic analyzers are used to collect the main pipe's inner diameter parameters and real-time cross-sectional spectrum data. The collected cross-sectional spectrum data records the distribution ratio of velocity layers in different radial coordinate systems within the pipe cross-section. Following the rule of equal area division of the pipe cross-section, the chemical dosing volume requirement is aligned with the cross-sectional spectrum data, i.e., the total chemical dosing requirement is allocated according to the area ratio data of different velocity layers, extracting a profile matrix containing multi-dimensional values. This profile matrix data is then projected onto the collected main pipe inner diameter parameters, i.e., the values ​​of the elements in the profile matrix are divided by the cross-sectional area values ​​calculated corresponding to the main pipe inner diameter parameters, extracting flow field vectors characterizing the independent velocity intensity of different regions. Furthermore, the values ​​of the elements in this flow field vector are compared with the dynamic viscosity values ​​of the fluid, i.e., a sequential division operation is performed to generate the Reynolds equivalent parameters of the flow field reflecting the degree of laminar or turbulent flow. In the numerical verification example, the acquired chemical dosing volume requirement is set to 90, and the cross-sectional area value converted from the collected main pipe inner diameter parameters is set to 100. The cross-sectional spectrum is divided into two calculation regions, a central region and an edge region, based on area. Data shows that the central region accounts for 0.6% of the area, while the edge region accounts for 0.4%. Based on this area ratio, a dosing requirement of 90 is allocated, yielding values ​​of 54 and 36 respectively. These are used to construct two elements of a 1D profile matrix. The profile matrix element is then divided by the area of ​​100. Dividing the central region's 54 by 100 yields a flow field vector value of 0.54, while dividing the edge region's 36 by 100 yields a flow field vector value of 0.36. The dynamic viscosity of the fluid under the current environment is set to 0.02. A ratio calculation is performed: dividing the central flow field vector (0.54) by 0.02 yields a Reynolds equivalent parameter of 27, and dividing the edge flow field vector (0.36) by 0.02 yields an equivalent parameter of 18. The central region value of 27 is extracted as the representative core equivalent parameter. This value of 27 directly reflects the intensity of local turbulence in the core fluid of the current pipeline network.

[0031] S302: Based on the Reynolds equivalent parameters of the flow field, detect the extreme values, compare the Reynolds equivalent parameters of the flow field with the extreme values ​​to extract the disturbance quantity, obtain the roughness, map the disturbance quantity and roughness through interpolation to extract the drag scalar, obtain the dissipation rate, fuse the drag scalar and dissipation rate, and establish an over-limit disturbance deviation metric. Based on the calculated Reynolds equivalent parameters of the flow field, the upper limit of the Reynolds extreme values ​​recorded by the system within the current control cycle is detected using a global traversal sorting algorithm. A difference comparison operation is performed between the flow field Reynolds equivalent parameters and the detected extreme values; that is, the difference between the extreme values ​​and the equivalent parameters is extracted as the disturbance quantity representing the flow field fluctuation intensity. The current roughness parameters of the pipeline are obtained by consulting the internal wall wear database. The extracted disturbance quantity is mapped to the obtained roughness parameters using linear interpolation; specifically, the disturbance quantity is directly multiplied by the roughness parameters and then the system's inherent reference friction coefficient is added to extract a resistance scalar characterizing the degree of impediment in the pipeline network. The kinetic energy dissipation rate of the fluid in the pipe wall boundary layer is read by calling the energy consumption model. The extracted resistance scalar and the obtained dissipation rate are fused, i.e., a product operation is performed to establish an over-limit disturbance deviation metric for quantifying abnormal states within the flow field. The actual operating parameters are then substituted into the calculation stage, assuming the received flow field Reynolds equivalent parameters are 27 and the upper limit of the Reynolds extreme values ​​obtained from the global scan detection is 35. A difference comparison was performed, subtracting the equivalent parameter 27 from the extreme value 35 to extract a disturbance of 8. The roughness parameter corresponding to the current pipe inner wall was retrieved and set to 0.05. A linear mapping calculation was performed, multiplying the disturbance of 8 by the roughness parameter 0.05 to obtain 0.4. The baseline friction coefficient was retrieved and preset to 1.0. 0.4 was added to 1.0 to extract a resistance scalar of 1.4. The current kinetic energy dissipation rate of the fluid was then obtained and set to 2.0. A fusion product operation was performed, multiplying the resistance scalar of 1.4 by the dissipation rate of 2.0 to obtain a result of 2.8. The calculated numerical result of 2.8 is the over-limit disturbance deviation measure, quantifying the implicit flow damping caused by the combination of turbulence and pipe wall friction.

[0032] S303: For the measurement of the deviation of the over-limit disturbance and the demand for the dosage volume, obtain the mass transfer map, and extract the gain factor by converging the measurement of the deviation of the over-limit disturbance and the mass transfer map according to the node. Multiply the demand for the dosage volume by the gain factor to extract the demand vector. Decouple the demand vector periodically according to the time sequence to obtain the turbulence disturbance compensation flow rate. For the established over-limit disturbance deviation metric and the requested dosing volume, a standard mass transfer map is extracted by reading the pre-stored 3D spatial diffusion model in the central controller. This mass transfer map consists of baseline mass transfer efficiency data recording multiple key physical nodes in the flow field cross-section. According to the planned grid nodes within the mass transfer map, the over-limit disturbance deviation metric and the baseline mass transfer efficiencies extracted from multiple nodes are converged. Specifically, this operation involves adding the deviation metric and the sum of the node mass transfer efficiencies, then dividing by the total number of nodes to extract the global average gain factor. The dosing volume requirement is multiplied by the extracted gain factor to extract the overall demand vector after mass transfer correction. Following the time series rhythm, this demand vector is decoupled by division according to a preset control cycle span parameter, i.e., the overall demand vector is divided by the cycle span value to calculate the turbulence disturbance compensation flow rate to be output within a single operating cycle.

[0033] Table 2: Mass Transfer Efficiency Matrix Node Distribution Table As shown in Table 2, the baseline mass transfer efficiency values ​​of five representative nodes within the mass transfer diagram were extracted. Assuming the input out-of-limit disturbance deviation metric is 2.8 and the required dosage volume is 90, a convergence operation was performed on each node, and the sum of the mass transfer efficiencies of the five nodes was calculated as: 0.95 + 0.85 + 0.82 + 0.80 + 0.80 = 4.22. Adding the deviation metric 2.8 to the sum 4.22 yields 7.02, which is then divided by the total number of nodes (5), resulting in an average gain factor of 1.4. Multiplying the required dosage volume of 90 by the gain factor 1.4 yields a corrected overall demand vector of 126. Assuming the preset control cycle span parameter is 2, the demand vector 126 was decoupled by division over a cycle span of 2, resulting in a value of 63. This value 63 represents the absolute amount of the compensation flow rate that actually needs to be dynamically output under the control cycle.

[0034] Please see Figure 5 The specific steps of S4 are as follows: S401: Call the turbulence disturbance compensation flow rate, obtain the nozzle scattering coefficient, collect the spatial grid nodes, perform a dot product operation between the turbulence disturbance compensation flow rate and the spatial grid nodes to extract the initial vector, use the nozzle scattering coefficient to perform a logarithmic mapping to extract the response gradient, and establish the spatial diffusion response tensor. The turbulence disturbance compensation flow rate is invoked, and the solidified nozzle scattering coefficient is obtained by consulting the actuator configuration parameter table. Simultaneously, spatial grid node coordinate data arranged within the nozzle diffusion area are collected using a spatial scanner. The turbulence disturbance compensation flow rate is multiplied one by one with the collected spatial grid node coordinate data to generate an initial vector set with intensity attributes. A logarithmic mapping operation is performed using the acquired nozzle scattering coefficient to extract the response gradient. This operation specifically involves obtaining the natural logarithm of the nozzle scattering coefficient, then performing proportional multiplication on the natural logarithm and the element values ​​in the initial vector set to generate a partial derivative vector, which is then used as the response gradient. A calculation example is performed using the operating parameters, assuming the received turbulence disturbance compensation flow rate is 63. The coordinate data of two core grid nodes, 2 and 3, are collected. A dot multiplication operation is performed, multiplying 63 by 2 and 3 respectively, extracting elements 126 and 189 from the initial vector set. The nozzle scattering coefficient is obtained as 2.718, and its natural logarithm is calculated to be approximately 1.0. The natural logarithm value 1.0 is multiplied by each element of the initial vector. Multiplying 1.0 by 126 yields the first element 126 of the partial derivative vector, and multiplying it by 189 yields the second element 189, generating a partial derivative vector containing elements 126 and 189. This partial derivative vector is designated as the response gradient, and a spatial diffusion response tensor is established.

[0035] S402: Based on the spatial diffusion response tensor, the attenuation factor is detected and the field scalar is extracted by dimensionality reduction of the spatial diffusion response tensor. The nozzle flow rate is collected, and the dynamic component is extracted by correlating the field scalar with the nozzle flow rate. The gain constant is obtained, and the dynamic component is compensated based on the gain constant to obtain the scattering compensation gain parameter. Based on the constructed spatial diffusion response tensor, the time difference between the injection command and the physical feedback is detected by a global timer and converted into an attenuation factor. A dimensionality reduction operation is then performed on the spatial diffusion response tensor to extract the field scalar. Specifically, this operation involves multiplying the elements of the tensor by the attenuation factor and then calculating the arithmetic mean. Nozzle velocity parameters are acquired using a flow probe installed in the main nozzle section. The extracted field scalar is then multiplied and correlated with the acquired nozzle velocity parameters to extract the dynamic component; that is, the two are directly multiplied. The set gain constant is read from the system compensation calibration register, and based on this gain constant, additive compensation calculations are performed on the calculated dynamic component to derive the scattering compensation gain parameter. In the specific data calculation example, the elements contained in the constructed spatial diffusion response tensor are assumed to be 126 and 189. The current attenuation factor value is detected and calculated to be 0.5. Dimensionality reduction is performed on the tensor. First, 126 and 189 are multiplied by 0.5 respectively to obtain intermediate values ​​of 63 and 94.5. The sum of 63 and 94.5 is then divided by the number of elements (2) to obtain the average value, resulting in a field scalar of 78.75. The current nozzle velocity parameter is acquired as 10. The field scalar 78.75 is correlated and multiplied with the nozzle velocity 10, extracting the dynamic component as 787.5. Then, a pre-set system gain constant of 12.5 is retrieved. Additive compensation calculation is performed, adding the dynamic component 787.5 to the gain constant 12.5, finally yielding a result of 800. This result, 800, is the generated scattering compensation gain parameter, constituting the total compensation amplification factor required to be supplied to the actuator under physical conditions.

[0036] S403: For the scattering compensation gain parameter and the turbulence disturbance compensation flow rate, monitor the structural resistance, extract the feedforward coefficient based on the damping modulation of the scattering compensation gain parameter according to the structural resistance, obtain the correction base, combine the correction base to calibrate and extract the feedforward vector, and use the feedforward vector to feedforward modulate the turbulence disturbance compensation flow rate to generate the target dosing volumetric flow rate. Based on the generated scattering compensation gain parameter and the previously saved turbulence disturbance compensation flow rate, the structural resistance values ​​of the nozzles and pipelines are monitored using a network of physical strain gauges distributed along the pipe section. A damping modulation operation is performed on the scattering compensation gain parameter according to the monitored structural resistance to extract the feedforward coefficient. This operation is represented by subtracting the equivalent resistance loss value calculated from the structural resistance from the scattering compensation gain parameter. A correction base value for correcting long-period drift of the system is obtained from the control bus. A calibration feedforward vector extraction operation is performed using this correction base value, that is, multiplying the feedforward coefficient with the correction base value to extract a stable feedforward vector parameter. The extracted feedforward vector is used to perform feedforward modulation on the turbulence disturbance compensation flow rate, that is, multiplying the value of the feedforward vector with the turbulence disturbance compensation flow rate to generate the target dosing volumetric flow rate. The collected data is substituted into the verification calculation, assuming the received scattering compensation gain parameter is 800. The monitored pipeline structural resistance value is 50, and the corresponding resistance loss equivalent is set to 50. A damping modulation subtraction operation is performed, subtracting the loss equivalent of 50 from the scattering compensation gain parameter of 800 to obtain a feedforward coefficient of 750. The preset system correction base of 0.002 is obtained from the control bus. A multiplication calibration operation is performed, multiplying the feedforward coefficient 750 by the correction base of 0.002 to obtain a feedforward vector scalar value of 1.5. The previously calculated and locked turbulence disturbance compensation flow rate of 63 is retrieved. A feedforward modulation multiplication operation is performed, multiplying the turbulence disturbance compensation flow rate of 63 by the feedforward vector scalar value of 1.5, ultimately deriving a calculation result of 94.5. This result of 94.5 is the generated target dosing volumetric flow rate, prepared as a hard numerical command for packaging and execution.

[0037] Please see Figure 6 The specific steps of S5 are as follows: S501: Calculate the ratio between the target dosing volumetric flow rate and the preset digital reference cycle to obtain the bandwidth time-domain scalar. Map the bandwidth time-domain scalar to the digital duty cycle constant to extract the digital feature factor. Multiply the digital feature factor with the time-domain scaling parameter to unify the dimensions and generate the duty cycle digital feature. The system calls the generated target dosing volumetric flow rate and simultaneously reads the preset digital reference period parameter from the control board's onboard clock chip. It then performs a division ratio calculation between the target dosing volumetric flow rate and the digital reference period parameter to obtain a bandwidth time-domain scalar representing the pulse trigger density per unit time. Next, it retrieves the underlying control duty cycle mapping rule parameter table and performs a division mapping calculation between the obtained bandwidth time-domain scalar and the fixed digital duty cycle constant in the table to extract a digital feature factor representing the proportion of high-level operation of the control waveform. This digital feature factor is then multiplied by the time-domain proportion parameter required by the underlying hardware port to unify the dimensions, generating a duty cycle digital feature that can be directly written by the pulse width modulator. Substituting the specific value into the logic operation, assuming the received target dosing volumetric flow rate is 94.5 and the preset digital reference period parameter is 3.5, a division ratio calculation is performed, dividing 94.5 by 3.5 to obtain a bandwidth time-domain scalar of 27. Finally, a fixed digital duty cycle constant value is retrieved and set to 45. A division mapping calculation is performed, dividing the bandwidth time-domain scalar 27 by the digital duty cycle constant 45, extracting a digital feature factor of 0.6. The time-domain scaling parameter required by the underlying hardware firmware is set to 100. A dimensionless multiplication operation is performed, multiplying the digital feature factor 0.6 by the time-domain scaling parameter 100, deriving a result of 60. This calculated value of 60 represents a 60% duty cycle for the generated control signal, used as a digital feature instruction for dividing the time action window of the actuator.

[0038] S502: The base output frequency is extracted by multiplying the duty cycle digital characteristics with the pump body rated speed. The dynamic bias frequency is extracted by weighting the base output frequency with the frequency bias coefficient. The dynamic bias frequency is compared with the system frequency limit value to perform a limiting operation and obtain the operating frequency control parameters. The calculated duty cycle digital characteristic is converted into a decimal factor. Combined with the pump's rated speed parameter read from the drive library, the duty cycle digital characteristic factor is multiplied by the pump's rated speed to extract the corresponding base output frequency for that state. The voltage sampling circuit obtains the current grid frequency variation amplitude, which is converted into a frequency offset coefficient via a lookup table. The extracted base output frequency is then multiplied by this frequency offset coefficient using a weighted multiplication operation to extract the dynamic offset frequency after correcting for external interference. The calculated dynamic offset frequency is compared with the safety-set frequency limit value, and a limiting operation is performed. If the dynamic offset frequency is greater than or equal to the limit value, the control parameter is set to the limit value; otherwise, the original calculated value is retained, thus obtaining the operating frequency control parameter.

[0039] Table 3: Comparison Table of Power Grid Fluctuation and Frequency Offset Coefficient Table 3 details the bias coefficients corresponding to the frequency fluctuation amplitude. In the example, the input duty cycle characteristic is converted to a decimal factor of 0.6, and the output reference corresponding to the pump's rated speed is 50. A multiplication operation is performed, multiplying 0.6 by 50 to extract the base output frequency of 30. Monitoring the grid frequency fluctuation of 0.2 Hz, the corresponding frequency bias coefficient is found to be 0.98. A weighted calculation is performed, multiplying the base output frequency of 30 by the bias coefficient 0.98 to extract the dynamic bias frequency of 29.4. The frequency limit value is found to be 45. An amplitude limit comparison operation is performed; 29.4 is less than 45, so the safety interception mechanism is not triggered, and 29.4 is retained. This value of 29.4 is the safe operating frequency control parameter input to the inverter.

[0040] S503: Extract data frames from the load field of the header of the control message for the operation frequency regulation parameters, perform an XOR operation between the data frames and the security check mask to extract the tail of the transmission message, and concatenate the device addressing identifier with the data frames and the tail of the transmission message to establish an automatic ammonia distribution control command. For the integer portion of the acquired operating frequency control parameter, it is used as core data and filled into the load field of the control message header as specified in the communication protocol to extract the data frame containing the execution command. The data frame is then XORed with a randomly generated security check mask to extract the tail of the transmission message. This extraction process specifically involves segmenting the data frame to obtain the corresponding binary sequence, segmenting the security check mask to obtain the mask binary sequence, performing bit-by-bit alignment on the data frame binary sequence and the mask binary sequence, comparing and extracting each bit's result, combining the comparison results according to the alignment position to generate a check sequence, and using this check sequence as the tail of the transmission message. The pre-assigned device address identifier is concatenated with the data frame and the tail of the transmission message to establish the automatic ammonia distribution control command. In the actual data encapsulation and packaging example, the acquired integer value of the operating frequency control parameter is assumed to be 29. This is converted to binary and filled into the load field to generate a data frame, with the binary sequence of this data frame being 11101. The corresponding binary sequence of the security check mask is obtained as 10110. Perform a bit-by-bit XOR operation on 11101 and 10110. The first bit (1) XORed with 1 results in 0, the second bit (1) XORed with 0 results in 1, the third bit (1) XORed with 1 results in 0, the fourth bit (0) XORed with 1 results in 1, and the fifth bit (1) XORed with 0 results in 1. Combine the comparison results according to the alignment positions to generate a check sequence of 01011. This sequence 01011 is directly designated as the tail of the transmitted message. The device addressing identifier sequence from the control module is read as 100. Perform a concatenation operation, connecting the identifier sequence 100, the data frame sequence 11101, and the tail sequence 01011 in the original order to finally establish the automatic ammonia distribution control command binary code string as 1001110101011. This final data string will be pushed to the gateway for physical layer transmission.

[0041] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of protection of the described technical solutions.

Claims

1. A method for automatically dosing ammonia in a boiler water of a thermal power plant, characterized in that, Includes the following steps: S1: Obtain the acidity and alkalinity and volumetric flow rate of the boiler water in the main feedwater pipeline of the thermal power plant. Based on the target acidity and alkalinity, perform water quality deviation parameter analysis on the boiler water acidity and alkalinity to extract the basic water quality deviation. Perform weighted flow allocation on the volumetric flow rate to construct the basic ammonia water flow rate. S2: Call the basic ammonia flow rate, collect the fluid temperature and fluid pressure inside the pipeline, input them into the thermo-fluid dynamics model for kinematic viscosity coupling calculation, and generate the dosing volume requirement; S3: Call the required dosing volume, calculate the Reynolds equivalent parameters of the flow field based on the main pipeline inner diameter parameters, and perform over-limit disturbance deviation measurement and convection mixing compensation to generate turbulence disturbance compensation flow rate. S4: Call the turbulence disturbance compensation flow rate, obtain the physical scattering coefficient of the dosing nozzle, perform nonlinear spatial diffusion response mapping to extract the scattering compensation gain parameter, perform nozzle resistance feedforward gain modulation on the turbulence disturbance compensation flow rate, and generate the target dosing volumetric flow rate. S5: Call the target dosing volumetric flow rate to perform digital pulse bandwidth conversion to extract the duty cycle digital features, analyze the operating frequency control parameters of the ammonia dosing metering pump, and generate automatic ammonia dosing control instructions.

2. The method for automatic dosing of ammonia in a boiler water of a thermal power plant as claimed in claim 1, wherein, The basic ammonia flow rate includes the baseline dosage, theoretical ammonia consumption, and basic mass flow rate. The dosing volume requirement includes the apparent volume, corrected volume, and dosing pump discharge. The turbulence disturbance compensation flow rate includes the local head loss converted flow rate, the friction head loss converted flow rate, and the dissipation compensation flow rate component. The target dosing volume flow rate includes the rated flow rate, the actual volumetric flow rate, and the effective discharge. The automatic ammonia distribution control commands include motor speed, drive current, and start / stop status.

3. The method for automatic dosing of ammonia in a boiler water of a thermal power plant as claimed in claim 1, wherein, The specific steps of S1 are as follows: S101: Collect the acidity and alkalinity of boiler water in the main feedwater pipeline of the thermal power plant and the target acidity and alkalinity. Subtract the target acidity and alkalinity from the boiler water acidity and alkalinity to extract the acidity and alkalinity difference. Read the deviation zone. If it is determined that the acidity and alkalinity difference falls within the deviation zone, map the scale parameter. Multiply the acidity and alkalinity difference with the scale parameter to generate the basic water quality deviation. S102: Call the basic water quality deviation and collect the volumetric flow rate, perform the conversion based on the inverse ratio rule to obtain the deviation weight factor, establish the correlation between the deviation weight factor and the volumetric flow rate, configure the weighted frequency for the volumetric flow rate, extract the frequency feature components and perform a dot product operation with the volumetric flow rate to obtain the weighted flow rate allocation matrix. S103: Extract the main allocation element according to the weighted flow allocation matrix, read the reference flow rate constant and multiply it with the main allocation element to obtain the feedforward ammonia flux, detect the pipeline resistance constant, reconstruct the ammonia flow rate vector by dividing the feedforward ammonia flux by the resistance constant, extract the ammonia flow rate vector scalar, and establish the basic ammonia flow rate.

4. The method for automatic dosing of ammonia in a boiler water of a thermal power plant as claimed in claim 3, wherein, The specific steps of S2 are as follows: S201: Call the basic ammonia flow rate, collect the fluid temperature and fluid pressure, subtract the fluid temperature from the preset baseline point to obtain the temperature deviation, subtract the fluid pressure from the pressure reference threshold to extract the pressure deviation, and perform vector splicing to establish a parameter feature map. S202: Extract the scale sequence, perform inner product processing on the parameter feature map and the scale sequence to obtain the thermodynamic state distribution matrix, extract the state feature tensor based on the thermodynamic state distribution matrix and the fluid constant, and retrieve the scalar in the viscosity lookup table according to the state feature tensor to generate the kinematic viscosity coupling factor. S203: Call the kinematic viscosity coupling factor and the basic ammonia flow rate to perform convolution correlation to extract the volume reference value, detect the pipeline coefficient, perform linear mapping between the volume reference value and the pipeline coefficient to extract the corrected residual, and accumulate it with the volume reference value to generate the dosing volume requirement.

5. The automatic ammonia dosing method for boiler water in thermal power plants according to claim 4, characterized in that, The execution vector splicing to establish the parameter feature map refers to obtaining the data dimension corresponding to the temperature deviation and the data dimension corresponding to the pressure gradient, performing filling and alignment processing on the temperature deviation and pressure gradient based on the data dimension, splicing the filled and aligned temperature deviation and pressure gradient along the channel direction to extract the initial feature matrix, and performing normalization mapping on the matrix element values ​​contained in the initial feature matrix to construct the parameter feature map. The pressure benchmark threshold refers to extracting historical fluid pressure records corresponding to the base ammonia flow rate, performing interval filtering and noise reduction on the historical fluid pressure records to extract a benchmark pressure sample set, arranging the pressure sample values ​​in the benchmark pressure sample set in ascending order to construct a pressure distribution vector, extracting the middle position value of the pressure distribution vector and the average scalar value of the pressure sample values ​​contained in the pressure distribution vector, and performing weighted calculation based on the middle position value and the average scalar value to generate the pressure benchmark threshold.

6. The automatic ammonia dosing method for boiler water in thermal power plants according to claim 4, characterized in that, The specific steps for S3 are as follows: S301: Call the dosing volume requirement, collect the inner diameter parameters and cross-sectional spectrum of the main pipeline, align the dosing volume requirement with the cross-sectional spectrum according to the area to extract the profile matrix, project the profile matrix and the inner diameter parameters of the main pipeline to extract the flow field vector, calculate the ratio of the flow field vector with the viscosity, and generate the Reynolds equivalent parameters of the flow field. S302: Based on the Reynolds equivalent parameters of the flow field, detect the extreme values, compare the Reynolds equivalent parameters of the flow field with the extreme values ​​to extract the disturbance amount, obtain the roughness, map the disturbance amount and roughness through interpolation to extract the drag scalar, obtain the dissipation rate, fuse the drag scalar and the dissipation rate, and establish an over-limit disturbance deviation metric. S303: For the aforementioned over-limit disturbance deviation measurement and dosing volume requirement, obtain a mass transfer map, aggregate the over-limit disturbance deviation measurement and the mass transfer map by node to extract the gain factor, multiply the dosing volume requirement by the gain factor to extract the demand vector, and decouple the demand vector periodically according to the time sequence to obtain the turbulence disturbance compensation flow rate.

7. The automatic ammonia dosing method for boiler water in thermal power plants according to claim 6, characterized in that, The specific steps of S4 are as follows: S401: Call the turbulence disturbance compensation flow rate, obtain the nozzle scattering coefficient, collect spatial grid nodes, perform a dot product operation between the turbulence disturbance compensation flow rate and the spatial grid nodes to extract the initial vector, use the nozzle scattering coefficient to perform a logarithmic mapping to extract the response gradient, and establish a spatial diffusion response tensor; S402: Based on the spatial diffusion response tensor, detect the attenuation factor and reduce the dimension of the spatial diffusion response tensor to extract the field scalar, collect the nozzle flow rate, correlate the field scalar with the nozzle flow rate to extract the dynamic component, obtain the gain constant, compensate the dynamic component based on the gain constant, and obtain the scattering compensation gain parameter. S403: For the scattering compensation gain parameter and the turbulence disturbance compensation flow rate, monitor the structural resistance, extract the feedforward coefficient based on the damping modulation of the scattering compensation gain parameter according to the structural resistance, obtain the correction base, calibrate and extract the feedforward vector in combination with the correction base, and use the feedforward vector to feedforward modulate the turbulence disturbance compensation flow rate to generate the target dosing volumetric flow rate.

8. The automatic ammonia dosing method for boiler water in thermal power plants according to claim 7, characterized in that, The step of extracting the response gradient by performing a logarithmic mapping using the nozzle scattering coefficient refers to obtaining the natural logarithmic value of the nozzle scattering coefficient, performing partial derivative processing on the natural logarithmic value of the nozzle scattering coefficient and the initial vector to generate a partial derivative vector, and determining the partial derivative vector as the response gradient.

9. The automatic ammonia dosing method for boiler water in thermal power plants according to claim 7, characterized in that, The specific steps of S5 are as follows: S501: The target dosing volumetric flow rate is called and the preset digital reference period is used to calculate the ratio to obtain the bandwidth time domain scalar. The bandwidth time domain scalar is mapped to the digital duty cycle constant to extract the digital feature factor. The digital feature factor is multiplied by the time domain proportional parameter to unify the dimensions and generate the duty cycle digital feature. S502: Based on the duty cycle digital characteristics and the rated speed of the pump body, the base output frequency is extracted by multiplying the base output frequency with the frequency offset coefficient. The dynamic offset frequency is extracted by weighting the base output frequency with the frequency offset coefficient. The dynamic offset frequency is compared with the system frequency limit value to perform a limiting operation and obtain the operating frequency control parameters. S503: For the operation frequency regulation parameters, the data frame is extracted from the load field of the header of the control message. The data frame and the security check mask are XORed to extract the tail of the transmission message. The device addressing identifier is concatenated with the data frame and the tail of the transmission message to establish an automatic ammonia distribution control command.

10. The automatic ammonia dosing method for boiler water in thermal power plants according to claim 9, characterized in that, The step of extracting the tail of the transmission message by performing an XOR operation on the data frame and the security check mask refers to splitting the data frame to obtain the binary sequence of the data frame and splitting the security check mask to obtain the binary sequence of the mask, performing a bit-by-bit alignment operation on the binary sequence of the data frame and the binary sequence of the mask and performing an XOR operation to compare and extract the comparison result, combining the comparison result according to the alignment position to generate a check sequence, and using the check sequence as the tail of the transmission message.

Images