A method and system for self-adaptive control of ammonia injection grid of SCR denitration system in thermal power plant

Abstract

The application discloses a thermal power plant SCR denitration system ammonia injection grid self-adaptive control method and system, including ammonia injection grid module, thermal impedance monitoring module, sound wave tomography module, differential pressure monitoring module, sound wave cleaning module and central control module, the central control module identifies the ash branch according to the temperature response curve collected by the thermal impedance monitoring module, according to the sound wave signal collected by the sound wave tomography module, the ammonia concentration distribution field and the temperature distribution field are reconstructed, the differential pressure signal collected by the differential pressure monitoring module is extracted spectrum characteristics, the influence of the ash branch on the injection uniformity or the catalyst layer stability is verified comprehensively, and the cleaning target branch is determined. The application realizes the on-line perception of the ammonia injection branch state, the visual monitoring of the injection uniformity, the early diagnosis of the catalyst layer flow instability, the thermal-acoustic resonance self-cleaning and the closed-loop intelligent control of the predictive maintenance, effectively improves the denitration efficiency, reduces the ammonia escape and prolongs the equipment life.

Description

Technical Field

[0001] This invention belongs to the field of flue gas denitrification technology, and particularly relates to an adaptive control method and system for ammonia injection grid in an SCR denitrification system of a thermal power plant. Background Technology

[0002] Selective catalytic reduction (SCR) technology is commonly used in flue gas denitrification in thermal power plants. Its core principle is to inject ammonia into the flue gas under the action of a catalyst, reducing nitrogen oxides to nitrogen and water. The ammonia injection grid, a key piece of equipment for ammonia injection, typically consists of several parallel branch pipes, each with a nozzle at its outlet. Uniform mixing of ammonia and flue gas is achieved by adjusting the flow rate of each branch pipe. Current ammonia injection control methods mainly rely on feedback regulation based on SCR outlet nitrogen oxide concentration measurements, adjusting the total ammonia injection rate through a PID controller. For the ammonia injection grid itself, operators periodically check the opening of the flow regulating valves in each branch pipe and manually correct them based on experience. Some advanced systems set flue gas composition measurement points at the SCR inlet, and, combined with computational fluid dynamics simulation results, preset the ammonia injection ratio for each zone to improve injection uniformity. The catalyst interlayer differential pressure, as a routine monitoring parameter, is mainly used to determine the degree of catalyst blockage and guide sootblower operation and catalyst replacement cycles.

[0003] However, existing technologies still have significant shortcomings. First, there is a lack of direct monitoring methods for the internal condition of ammonia injection branch pipes. Nozzle ash accumulation is a gradual process, difficult to detect in its early stages, and only becomes apparent when severe ash accumulation leads to a deterioration in the distribution of nitrogen oxides at the outlet. At this point, cleaning is extremely difficult and requires shutdown. The lack of thermal impedance monitoring technology makes online sensing of the flow status at the branch pipe level impossible. Second, traditional control relies only on a limited number of chemical sensor measurement points, failing to obtain panoramic information on the ammonia distribution across the flue cross-section, making it difficult to pinpoint specific areas of uneven injection. Furthermore, feedback control has inherent lag, making feedforward prediction impossible. Third, conventional differential pressure signals are only used to determine the average degree of catalyst layer blockage, and their high-frequency fluctuation information is not effectively utilized. These fluctuation characteristics precisely contain early signs of local flow instability. Fourth, existing systems lack online self-cleaning capabilities. Once branch pipes become blocked, manual inspection and cleaning can only be carried out after shutdown, resulting in high maintenance costs and low equipment availability. In addition, current control strategies are mostly passive responses, lacking the ability to predict the development trend of ash accumulation and proactive intervention methods, failing to actively optimize operating parameters to delay the deterioration process before a failure occurs. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, the present invention provides an adaptive control method and system for ammonia injection grid in SCR denitrification system of thermal power plant, which solves the problems of unpredictable ammonia injection branch pipe status, passive control, inability to self-clean, and lack of predictive maintenance in the prior art.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] An adaptive control method and system for ammonia injection grid in an SCR denitrification system of a thermal power plant, comprising:

[0007] The ammonia injection grid module includes several parallel ammonia injection branch pipes, and each ammonia injection branch pipe inlet is equipped with a flow regulating valve;

[0008] A thermal impedance monitoring module is installed on the outer wall of the ammonia injection branch pipe. It is used to apply thermal pulses to the corresponding ammonia injection branch pipe and collect the temperature response curve of the ammonia injection branch pipe.

[0009] The acoustic tomography module includes an acoustic generator and an acoustic receiver arranged on opposite sides of the cross-section of the SCR inlet flue, for transmitting and receiving multipath acoustic signals;

[0010] The differential pressure monitoring module includes a high-frequency differential pressure transmitter connected to the inlet and outlet sides of the SCR catalyst layer, used to collect differential pressure signals on both sides of the catalyst layer.

[0011] The acoustic cleaning module includes a piezoelectric transducer coupled to the outer wall of each ammonia injection branch pipe for applying acoustic excitation to the corresponding ammonia injection branch pipe.

[0012] The central control module is communicatively connected to the flow regulating valve, thermal impedance monitoring module, acoustic tomography module, differential pressure monitoring module, and acoustic cleaning module.

[0013] The central control module is configured to: identify ammonia injection branch pipes with ash accumulation based on the temperature response curve acquired by the thermal impedance monitoring module; reconstruct the ammonia concentration distribution field and flue gas temperature distribution field of the SCR inlet flue cross section based on the acoustic signal acquired by the acoustic tomography module; extract spectral features based on the differential pressure signal acquired by the differential pressure monitoring module; verify the impact of ammonia injection branch pipes with ash accumulation on injection uniformity or catalyst layer stability based on the ammonia concentration distribution field or spectral features, and determine the target branch pipe for cleaning; determine the acoustic resonance frequency based on the flue gas temperature distribution field; control the acoustic cleaning module to apply acoustic excitation to the target branch pipe for cleaning at the resonance frequency, and control the thermal impedance monitoring module to apply auxiliary thermal pulses to the branch pipe synchronously or at different times; determine the cleaning endpoint based on the temperature response curve of the branch pipe acquired in real time during the cleaning process, and control the acoustic cleaning module and the thermal impedance monitoring module to stop working.

[0014] Preferably, the central control module is also configured to: upon initial commissioning or after each cleaning, control the thermal impedance monitoring module to apply thermal pulses to each ammonia injection branch pipe, collect the temperature response curve of each ammonia injection branch pipe, and extract characteristic parameters as the reference characteristic parameters of the branch pipe in the clean state.

[0015] The acoustic tomography module is controlled to acquire acoustic signals, reconstruct the ammonia concentration distribution field of the cross section of the SCR inlet flue, and calculate the statistical value of ammonia concentration in each region of the cross section as the initial benchmark for ammonia concentration uniformity.

[0016] The differential pressure monitoring module collects differential pressure signals, performs spectral analysis on the continuously collected differential pressure signals, and extracts spectral feature values ​​as spectral background features under stable operating conditions.

[0017] Preferably, the ammonia injection branch pipe with ash accumulation is identified based on the temperature response curve acquired by the thermal impedance monitoring module as follows:

[0018] The characteristic parameters of the real-time temperature response curve are compared with the baseline characteristic parameters to calculate the deviation value. When the deviation value exceeds the first preset threshold, it is determined that the branch pipe has a risk of dust accumulation. When the deviation value exceeds the second preset threshold, it is determined that the branch pipe is severely blocked. The second preset threshold is greater than the first preset threshold.

[0019] The characteristic parameters of the temperature response curve include at least one of the following: heating rate, steady-state temperature difference, or time constant.

[0020] Preferably, the specific characteristics of the ammonia injection branch pipe with accumulated ash are verified based on the ammonia concentration distribution field or spectrum to determine the impact on injection uniformity or catalyst layer stability, and the target cleaning branch pipe is determined as follows:

[0021] Based on the ammonia concentration distribution field reconstructed by the acoustic tomography module, local areas where the ammonia concentration is lower than the preset uniformity index are identified.

[0022] Based on the spectral characteristics extracted by the differential pressure monitoring module, it is determined whether the spectral characteristics exceed the second preset threshold. If so, it is determined that there is flow instability in the catalyst layer.

[0023] When it is determined that there is dust accumulation in the ammonia injection branch pipe, and the ammonia concentration in the corresponding injection area is locally low, or the catalyst layer is unstable, the branch pipe is identified as the target cleaning branch pipe.

[0024] Preferably, the ammonia concentration distribution field and flue gas temperature distribution field of the SCR inlet flue cross section are reconstructed based on the acoustic signals acquired by the acoustic tomography module as follows:

[0025] The sound velocity distribution of the flue cross section is obtained by inverting the flight time of the sound waves along each propagation path, and then converted into the flue gas temperature distribution field based on the correspondence between sound velocity and temperature.

[0026] The attenuation coefficient distribution of the flue cross section is obtained by inverting the amplitude attenuation of sound waves along each propagation path, and then converted into an ammonia concentration distribution field based on the correspondence between the attenuation coefficient and the ammonia concentration.

[0027] Preferably, extracting spectral features from the differential pressure signal acquired by the differential pressure monitoring module specifically includes:

[0028] Perform a Fourier transform on the differential pressure signal to extract the energy integral value within a preset high-frequency band or the amplitude at at least one preset characteristic frequency.

[0029] Specifically, determining whether the spectral characteristics exceed the second preset threshold means that when the energy integral value of the high-frequency band exceeds a preset multiple of the corresponding value in the spectral background characteristics, or when the characteristic frequency amplitude exceeds a preset multiple of the corresponding value in the spectral background characteristics, it is determined that there is flow instability in the catalyst layer.

[0030] Preferably, the acoustic resonant frequency is determined based on the flue gas temperature distribution field as follows:

[0031] The gas velocity in the branch pipe is calculated based on the flue gas temperature value of the area where the clean target branch pipe is located, extracted from the flue gas temperature distribution field.

[0032] A preliminary estimate of the resonant frequency is determined based on the gas sound velocity, the geometry of the branch pipe, and the acoustic boundary conditions.

[0033] The control acoustic cleaning module sweeps the frequency around the initial estimated value. At the same time, the temperature sensor corresponding to the branch pipe in the thermal impedance monitoring module detects the vibration signal of the pipe wall. When the amplitude of the vibration signal reaches a local maximum, the current frequency is locked as the resonance frequency.

[0034] Preferably, the cleaning endpoint is determined based on the temperature response curve of the branch pipe collected in real time during the cleaning process.

[0035] The temperature response curve of the branch pipe is received in real time by the thermal impedance monitoring module. The deviation between its characteristic parameters and the reference characteristic parameters is calculated. When the deviation is lower than the third preset threshold, the cleaning is determined to be completed.

[0036] Preferably, a digital twin model is also included, and the central control module is further configured as follows:

[0037] The data received in real time from the thermal impedance monitoring module, the acoustic tomography module, and the differential pressure monitoring module are input into the digital twin model, which drives the digital twin model to output the predicted ammonia concentration distribution in the area of ​​the SCR inlet flue where no acoustic receiver is set or the predicted ash growth rate of each ammonia injection branch.

[0038] When the predicted rate of ash accumulation exceeds the fourth preset threshold, the opening of the flow regulating valve of the branch pipe is adjusted before the degree of ash accumulation reaches the first preset threshold, thus delaying the development of ash accumulation.

[0039] Preferably, an adaptive control method for ammonia injection grid in an SCR denitrification system of a thermal power plant includes the following steps:

[0040] Step 1: Establish the baseline temperature response characteristics of each ammonia injection branch pipe under clean conditions, the baseline initial ammonia concentration distribution of the SCR inlet flue section, and the spectral background characteristics of the catalyst layer under stable operation conditions;

[0041] Step 2: Real-time acquisition of temperature response curves, multipath acoustic signals of flue gas cross-section, and differential pressure signals of catalyst layer for each ammonia injection branch pipe;

[0042] Step 3: Based on the comparison between the real-time temperature response curve and the reference temperature response characteristics, identify candidate ammonia injection branch pipes with ash accumulation;

[0043] Step 4: Reconstruct the ammonia concentration distribution field and flue gas temperature distribution field of the current flue cross section based on the real-time collected acoustic signals, and identify local areas where the ammonia concentration is lower than the preset uniformity index.

[0044] Step 5: Perform spectrum analysis on the real-time acquired differential pressure signal, extract spectrum features, and determine whether there is catalyst layer flow instability;

[0045] Step 6: When the candidate ammonia injection branch identified in Step 3 corresponds to the local area space identified in Step 4, or when the candidate ammonia injection branch identified in Step 3 exists and Step 5 determines that there is flow instability, the candidate ammonia injection branch is identified as the cleaning target branch.

[0046] Step 7: Extract the flue gas temperature value of the area where the clean target branch pipe is located from the flue gas temperature distribution field reconstructed in Step 4, and determine the acoustic resonant frequency based on the flue gas temperature value.

[0047] Step 8: Control the acoustic cleaning module to apply acoustic excitation to the target branch pipe at the resonant frequency, and control the thermal impedance monitoring module to apply auxiliary thermal pulses to the branch pipe synchronously or at different times.

[0048] Step 9: During the cleaning process, the temperature response curve of the branch pipe is collected in real time. When the deviation between the branch pipe and the reference temperature response characteristic is lower than the third preset threshold, the application of acoustic excitation and auxiliary heat pulse is stopped.

[0049] Step 10: After cleaning, verify whether the temperature response characteristics of the branch pipe, the ammonia concentration in the corresponding area, and the differential pressure spectrum characteristics of the catalyst layer have returned to the preset range. If they have not returned, repeat steps 7 to 9 or output a maintenance alarm.

[0050] The present invention discloses an adaptive control method and system for ammonia injection grid in an SCR denitrification system of a thermal power plant, and its technical effects and advantages are as follows:

[0051] 1. This invention acquires the temperature response curves of each ammonia injection branch pipe through a thermal impedance monitoring module, and quantitatively identifies the degree of ash accumulation in the branch pipe by utilizing changes in thermal response characteristic parameters, transforming the traditionally unmonitorable internal state of the branch pipe into a quantifiable indicator; it transmits and receives multipath acoustic signals through an acoustic tomography module, and reconstructs the ammonia concentration distribution field and flue gas temperature distribution field of the SCR inlet flue cross section, realizing visualized monitoring of the spatial distribution of ammonia injection; it acquires the catalyst layer differential pressure signal at high frequency through a differential pressure monitoring module and extracts spectral features, mining the micro-flow information contained in the conventional differential pressure signal to determine the flow instability state of the catalyst layer. The coordinated monitoring of these three physical fields forms a comprehensive sensing network from the branch pipe level, cross section level to the catalyst layer level, significantly improving diagnostic accuracy and robustness.

[0052] 2. This invention comprehensively determines the target branch to be cleaned based on thermal impedance identification of ash-accumulated branch pipes, acoustic tomography identification of abnormal concentration areas, and pressure difference spectrum determination of flow instability, avoiding misjudgment based on a single indicator. It calculates the gas velocity within the target branch pipe based on the flue gas temperature distribution field reconstructed by acoustic tomography, and uses a temperature sensor to detect vibration signals and lock the actual resonant frequency through frequency sweeping, ensuring that the acoustic excitation is always in the optimal resonant state. During the cleaning process, it monitors the changes in the temperature response curve of the target branch pipe in real time, automatically stopping excitation when the characteristic parameter deviation is below a preset threshold, achieving on-demand and precise cleaning. After cleaning, the effect is verified through multi-source information, forming a complete closed-loop control system of "perception-diagnosis-decision-execution-verification".

[0053] 3. In this invention, the heating element of the thermal impedance monitoring module serves both to apply thermal pulses to monitor the ash accumulation state and to provide thermal stress as a thermal pulse source during the cleaning process; the temperature sensor is used both to collect temperature response curves and to detect pipe wall vibration signals when the resonant frequency is locked; the synergistic effect of acoustic excitation and thermal pulses allows the ash layer to simultaneously withstand vibration fatigue and thermal stress peeling, resulting in significantly higher cleaning efficiency than a single method. This hardware reuse and energy coupling design achieves the integration of diagnosis and treatment without increasing hardware costs.

[0054] 4. This invention integrates real-time monitoring data using a digital twin model to extrapolate the ammonia concentration distribution in unmeasurable areas and predict the ash accumulation trend in each branch pipe. When it is predicted that a branch pipe will reach the ash accumulation threshold within a preset time, the opening of the flow regulating valve in that branch pipe is adjusted in advance to slow down the ash accumulation, transforming passive cleaning into proactive prevention. Combined with the accumulation of historical data and the establishment of correlation models, the system possesses continuous learning and optimization capabilities, enabling it to adaptively adjust control strategies for different operating conditions.

[0055] 5. This invention fundamentally ensures the consistency of flow rate in each branch of the ammonia injection grid by accurately identifying and promptly removing ash accumulation in the ammonia injection branch pipes. This allows ammonia and flue gas to be uniformly mixed at the catalyst inlet section, thereby minimizing ammonia escape and reducing the risk of ammonium bisulfate formation while ensuring denitrification efficiency, and delaying catalyst deactivation and air preheater blockage. The online self-cleaning function avoids manual cleaning during downtime, significantly improving equipment availability and operating economy. The early diagnosis function of catalyst layer flow instability can adjust operating parameters in time before blockage occurs, extending catalyst life and reducing replacement frequency and maintenance costs. Attached Figure Description

[0056] Figure 1 This invention presents an adaptive control method for ammonia injection grid in an SCR denitrification system of a thermal power plant, and a system block diagram thereof.

[0057] Figure 2 This invention presents a flowchart of an adaptive control method and system for ammonia injection grid in an SCR denitrification system of a thermal power plant. Detailed Implementation

[0058] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0059] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include," "contain," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "includes..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0060] refer to Figure 1-2This invention provides an adaptive control method and system for ammonia injection grids in a thermal power plant's SCR denitrification system. The system includes an ammonia injection grid module, a thermal impedance monitoring module, an acoustic tomography module, a differential pressure monitoring module, an acoustic cleaning module, and a central control module. The central control module identifies ammonia injection branch pipes with ash accumulation based on the temperature response curve acquired by the thermal impedance monitoring module. It reconstructs the ammonia concentration distribution field and flue gas temperature distribution field of the SCR inlet flue cross-section based on the acoustic signal acquired by the acoustic tomography module. It extracts spectral features from the differential pressure signal acquired by the differential pressure monitoring module and verifies the impact of the ash-accumulated branch pipes on injection uniformity or catalyst layer stability by comprehensively considering the ammonia concentration distribution field or spectral features to determine the target branch pipe for cleaning. Furthermore, it determines the acoustic resonance frequency based on the flue gas temperature distribution field, controls the acoustic cleaning module to apply acoustic excitation at the resonance frequency, and controls the thermal impedance monitoring module to synchronously or sequentially apply auxiliary thermal pulses. Simultaneously, it determines the cleaning endpoint based on the temperature response curve acquired in real-time during the cleaning process. This invention achieves intelligent control of the entire process, including online sensing of ammonia injection branch pipe status, visual monitoring of injection uniformity, early diagnosis of catalyst layer flow instability, and online self-cleaning via thermoacoustic resonance.

[0061] An adaptive control system for ammonia injection grid in an SCR denitrification system of a thermal power plant includes:

[0062] Ammonia injection grid module: includes several parallel ammonia injection branch pipes, each with a flow regulating valve at the inlet and a nozzle at the outlet for injecting ammonia gas into the SCR inlet flue.

[0063] Thermal impedance monitoring module: Located on the outer wall of each ammonia injection branch pipe, including a thin-film heating element and a temperature sensor, used to apply thermal pulses to the corresponding ammonia injection branch pipe and acquire the temperature response curve of the ammonia injection branch pipe.

[0064] Acoustic tomography module: includes acoustic generators and acoustic receivers arranged on opposite sides of the cross-section of the SCR inlet flue, for transmitting and receiving multipath acoustic signals.

[0065] Differential pressure monitoring module: includes a high-frequency differential pressure transmitter connected to the inlet and outlet sides of the SCR catalyst layer, used to collect differential pressure signals on both sides of the catalyst layer.

[0066] Acoustic cleaning module: includes a piezoelectric transducer coupled to the outer wall of each ammonia injection branch pipe, used to apply acoustic excitation to the corresponding ammonia injection branch pipe.

[0067] The central control module communicates with the flow regulating valve, thermal impedance monitoring module, acoustic tomography module, differential pressure monitoring module, and acoustic cleaning module. It is used to identify the status of the ammonia injection branch pipe, make cleaning decisions, and execute control based on the data collected by each module.

[0068] Benchmark establishment:

[0069] After the system is initially put into operation or after each major overhaul, the central control module executes the following baseline establishment procedure:

[0070] Step 1: Establish the baseline temperature response characteristics of each ammonia injection branch pipe under clean conditions, the baseline initial ammonia concentration distribution of the SCR inlet flue section, and the spectral background characteristics of the catalyst layer under stable operation conditions.

[0071] The specific implementation is as follows:

[0072] The thermal impedance monitoring module is controlled to apply a 5W heat pulse with a duration of 10s to each ammonia injection branch, and the temperature response curve of each branch is collected. The heating rate (°C / s) is extracted as the reference heating rate of the branch.

[0073] The acoustic tomography module was controlled to emit 2kHz acoustic waves. After receiving multipath signals, the initial ammonia concentration distribution field of the SCR inlet flue cross section was reconstructed. The cross-sectional average concentration and coefficient of variation were calculated as the initial reference for ammonia concentration uniformity. The initial average concentration was 300 mg / m³, and the initial coefficient of variation was 4.5%.

[0074] The differential pressure monitoring module continuously acquires the differential pressure signal of the catalyst layer at a sampling frequency of 10Hz for 10 minutes, performs fast Fourier transform, extracts the energy integral value of the 1~10Hz frequency band as the background energy integral value, and the background energy integral value is 0.25V²; extracts the characteristic frequency amplitude of 1Hz as the background characteristic frequency amplitude, and the background characteristic frequency amplitude is 0.08V.

[0075] Example 1

[0076] This embodiment provides an adaptive control method and system for the ammonia injection grid in an SCR denitrification system of a thermal power plant, used for adaptive control under normal operating conditions. Specific implementation details include:

[0077] Purpose of implementation: To verify the stable operation capability of the system of the present invention under normal operating conditions, and to demonstrate how the system maintains the existing ammonia injection distribution without triggering unnecessary intervention.

[0078] System Implementation: The system employs the ammonia injection grid module, thermal impedance monitoring module, acoustic tomography module, differential pressure monitoring module, acoustic cleaning module, and central control module of this invention. Each module is connected and configured as described above. Reference Characteristic Parameters: The reference heating rate of each branch pipe is between 0.8 and 1.2℃ / s, the average concentration is 300 mg / m³, the coefficient of variation is 4.5%, the background energy integral value is 0.25V², and the background characteristic frequency amplitude is 0.08V.

[0079] Implementation steps:

[0080] Step 1: The baseline has been established and the system is now in automatic operation mode.

[0081] Step 2: Real-time acquisition of temperature response curves, multipath acoustic signals of flue gas cross-sections, and differential pressure signals of catalyst layers for each ammonia injection branch. The sampling period is set to 1 minute.

[0082] Step 3: Based on the comparison between the real-time acquired temperature response curve and the baseline temperature response characteristics, candidate ammonia injection branch pipes with ash accumulation are identified. The current heating rate of each branch pipe is calculated and compared with the baseline heating rate to obtain the deviation rate. The monitoring results show that the deviation rates of all branch pipes are between 0.95 and 1.10, all less than the first preset threshold of 1.5, and no candidate ammonia injection branch pipes were identified.

[0083] Step 4: Reconstruct the ammonia concentration distribution field and flue gas temperature distribution field of the current flue cross-section based on the real-time acquired acoustic signals, and identify local areas where the ammonia concentration is lower than the preset uniformity index. The reconstruction results show that the ammonia concentration distribution of the cross-section is uniform, with a coefficient of variation of 4.8%, which is 10% lower than the preset uniformity index, and no local areas were identified.

[0084] Step 5: Perform spectral analysis on the real-time acquired differential pressure signal, extract spectral features, and determine whether catalyst layer flow instability exists. Perform a Fast Fourier Transform on the current 1-minute differential pressure signal, calculate the high-frequency energy as 0.27V², the ratio of which to the background energy integral value is 1.08, less than the second preset threshold of 2.0; the characteristic frequency amplitude at 1Hz is 0.09V, the ratio of which to the background characteristic frequency amplitude is 1.13, less than the preset multiple of 3 corresponding to the second preset threshold. Therefore, it is determined that there is no catalyst layer flow instability.

[0085] Step 6: Since no candidate ammonia injection branch was identified in Step 3, no target branch for cleaning was determined, so the subsequent steps are skipped.

[0086] Steps 7 through 10: Not executed.

[0087] Implementation results: The system maintained the existing ammonia injection distribution without any intervention. The outlet NOx was stable at 45 mg / m³, the ammonia slip was less than 2 ppm, and the catalyst bed differential pressure was stable, which verified the stability and non-intervention of the system under normal operating conditions.

[0088] Example 2

[0089] This embodiment provides an adaptive control method and system for ammonia injection grid in an SCR denitrification system of a thermal power plant, used for early warning and prevention of slight ash accumulation in branch pipes. Specific implementation details include:

[0090] Purpose of implementation: To verify the early identification capability of the system of the present invention for slight dust accumulation, and the preventive control effect of delaying the development of dust accumulation through flow regulation.

[0091] Implementation system: Same as in Example 1. Baseline characteristic parameters are the same as above.

[0092] Implementation steps:

[0093] Step 1: Baseline establishment is complete.

[0094] Step 2: Collect data in real time. Sampling period is 1 minute.

[0095] Step 3: Based on the comparison between the real-time temperature response curve and the baseline temperature response characteristics, candidate ammonia injection branch pipes with ash accumulation are identified. Calculations show that the current heating rate of branch pipe No. 12 is 1.92℃ / s, the baseline heating rate is 1.20℃ / s, and the deviation rate is 1.60, which exceeds the first preset threshold of 1.5. Therefore, branch pipe No. 12 is recorded as a candidate ammonia injection branch pipe.

[0096] Step 4: Reconstruct the ammonia concentration distribution field and flue gas temperature distribution field of the current flue cross-section based on the real-time acquired acoustic signals, and identify local areas where the ammonia concentration is lower than the preset uniformity index. The reconstruction results show that the ammonia concentration in the area corresponding to branch pipe No. 12 is 276 mg / m³, which is 92% of the average concentration of 300 mg / m³. It is lower than the average value but still above the preset uniformity index of 80%, and is not recorded as a local area.

[0097] Step 5: Perform spectral analysis on the real-time acquired differential pressure signal, extract spectral features, and determine whether catalyst layer flow instability exists. The high-frequency energy is calculated to be 0.28V², with a ratio of 1.12 to the background energy integral value, which is less than 2.0; the characteristic frequency amplitude at 1Hz is 0.09V, with a ratio of 1.13 to the background characteristic frequency amplitude, which is less than 3 times. Therefore, it is determined that there is no catalyst layer flow instability.

[0098] Step 6: Step 3 identified a candidate ammonia injection branch pipe, but Step 4 did not identify a local area and Step 5 did not determine the existence of flow instability. The conditions of "the candidate ammonia injection branch pipe corresponds to the local area" or "the candidate ammonia injection branch pipe exists and there is flow instability" are not met. Therefore, the candidate ammonia injection branch pipe is not identified as the cleaning target branch pipe.

[0099] Steps 7 through 10: Cleaning was not performed. However, based on the predictive capabilities of the digital twin model, the central control module initiated preventative measures: opening the flow regulating valve of branch pipe 12 by 8% to increase the ammonia injection pressure and delay ash accumulation.

[0100] Implementation results: Subsequent continuous monitoring showed that the deviation rate of branch pipe No. 12 remained stable between 1.55 and 1.62, and did not deteriorate further to the second preset threshold of 2.0. The cleaning trigger was effectively delayed, verifying the effectiveness of early warning and prevention and control.

[0101] Example 3

[0102] This embodiment provides an adaptive control method and system for ammonia injection grid in an SCR denitrification system of a thermal power plant, used for thermoacoustic resonance cleaning under severe ash accumulation. Specific implementation details include:

[0103] Purpose of implementation: To verify the effectiveness of the system of the present invention in accurately identifying severely dusty branch pipes, locking the resonant frequency, and performing thermoacoustic resonance cleaning throughout the entire process.

[0104] Implementation system: Same as in Example 1. Baseline characteristic parameters are the same as above.

[0105] Implementation steps:

[0106] Step 1: Baseline establishment is complete.

[0107] Step 2: Collect data in real time. Sampling period is 1 minute.

[0108] Step 3: Based on the comparison between the real-time temperature response curve and the baseline temperature response characteristics, candidate ammonia injection branch pipes with ash accumulation are identified. Calculations show that the current heating rate of branch pipe No. 5 is 2.64℃ / s, the baseline heating rate is 1.20℃ / s, and the deviation rate is 2.20, which exceeds the second preset threshold of 2.0. Therefore, branch pipe No. 5 is recorded as a candidate ammonia injection branch pipe.

[0109] Step 4: Reconstruct the ammonia concentration distribution field and flue gas temperature distribution field of the current flue cross-section based on the real-time acquired acoustic signals, and identify local areas where the ammonia concentration is lower than the preset uniformity index. The reconstruction results show that the ammonia concentration in the area corresponding to branch pipe No. 5 is 195 mg / m³, which is 65% of the average concentration of 300 mg / m³, and lower than the preset uniformity index of 80%. Record the coordinates of this local area.

[0110] Step 5: Perform spectral analysis on the real-time acquired differential pressure signal, extract spectral features, and determine whether there is catalyst layer flow instability. The high-frequency energy is calculated to be 0.63V², with a ratio of 2.52 to the background energy integral value, exceeding the second preset threshold of 2.0; the characteristic frequency amplitude at 1Hz is 0.24V, with a ratio of 3.0 to the background characteristic frequency amplitude, not exceeding 3 times. Therefore, it is determined that flow instability exists in the catalyst layer.

[0111] Step 6: The candidate ammonia injection branch pipe (No. 5) identified in Step 3 corresponds to the local area space identified in Step 4 (after coordinate matching, the injection area of ​​branch pipe No. 5 coincides with the area with low ammonia concentration), which meets the condition. Therefore, the candidate ammonia injection branch pipe is determined as the cleaning target branch pipe.

[0112] Step 7: Extract the flue gas temperature value of the area where the clean target branch pipe is located from the flue gas temperature distribution field reconstructed in Step 4, and determine the acoustic resonant frequency based on the flue gas temperature value.

[0113] The flue gas temperature in the No. 5 branch pipe area was 380℃ (653K).

[0114] The calculated sound velocity of the gas inside the branch pipe is approximately 512 m / s;

[0115] The geometric dimensions of branch pipe No. 5 are: length 1.2m, inner diameter 20mm, acoustic boundary condition is a closed structure at one end, and the preliminary estimated value of the resonance frequency is about 107Hz.

[0116] The control acoustic cleaning module sweeps the frequency within a range of ±10Hz with 107Hz as the center. At the same time, the temperature sensor corresponding to the branch pipe in the thermal impedance monitoring module detects the vibration signal of the pipe wall. When the amplitude of the vibration signal reaches a local maximum at 108Hz, 108Hz is locked as the resonance frequency.

[0117] Step 8: Control the acoustic cleaning module to apply acoustic excitation (power 40W) to the target branch pipe at 108Hz, and control the thermal impedance monitoring module to simultaneously apply auxiliary thermal pulses (cycle 1s, duty cycle 50%) to the branch pipe.

[0118] Step 9: During the cleaning process, the temperature response curve of the branch pipe is acquired in real time, and the deviation rate is calculated. When the deviation rate drops from 2.20 to 1.18 (below the third preset threshold of 1.2), the application of acoustic excitation and auxiliary heat pulse is stopped. The cleaning duration is 25 seconds.

[0119] Step 10: After cleaning, verify whether the temperature response characteristics of the branch pipe, the ammonia concentration in the corresponding area, and the differential pressure spectrum characteristics of the catalyst layer have returned to the preset range.

[0120] The retested heating rate of branch pipe No. 5 was 1.32℃ / s, with a deviation rate of 1.10, which recovered to below the third preset threshold of 1.2.

[0121] The ammonia concentration distribution field was reconstructed. The ammonia concentration in the area corresponding to branch pipe No. 5 was 288 mg / m³, which is 96% of the average concentration of 300 mg / m³, and was restored to within ±10% of the initial baseline.

[0122] The differential pressure spectrum was remeasured, and the high-frequency energy was 0.33V², with a ratio of 1.32 to the integral value of the background energy, which recovered to below the second preset threshold of 2.0.

[0123] All requirements met; cleaning complete.

[0124] Results: The unit was not shut down, the denitrification efficiency remained unchanged, and the severely clogged branch pipe was successfully cleaned online, verifying the effectiveness of thermoacoustic resonance cleaning.

[0125] Example 4

[0126] This embodiment provides an adaptive control method and system for the ammonia injection grid in a thermal power plant's SCR denitrification system, used for the diagnosis and rapid response to catalyst bed flow instability. Specific implementation details include:

[0127] Objective: To verify the ability of the system of the present invention to identify catalyst layer flow instability based solely on the spectral characteristics of differential pressure fluctuations in the absence of branch pipe blockage, and to achieve rapid response through ammonia injection distribution adjustment.

[0128] Implementation system: Same as in Example 1. Baseline characteristic parameters are the same as above.

[0129] Implementation steps:

[0130] Step 1: Baseline establishment is complete.

[0131] Step 2: Collect data in real time. Sampling period is 1 minute.

[0132] Step 3: Based on the comparison between the real-time acquired temperature response curve and the baseline temperature response characteristics, identify candidate ammonia injection branch pipes with ash accumulation. Calculate the deviation rate of all branch pipes; if it is less than 1.4, and no branch pipe exceeds the first preset threshold of 1.5, no candidate ammonia injection branch pipe is identified.

[0133] Step 4: Reconstruct the ammonia concentration distribution field and flue gas temperature distribution field of the current flue cross-section based on the real-time acquired acoustic signals, and identify local areas where the ammonia concentration is lower than the preset uniformity index. The reconstruction results show that the ammonia concentration in each area is between 285 and 315 mg / m³, all within 95% to 105% of the average concentration of 300 mg / m³, and not lower than the preset uniformity index of 80%, so no local areas were identified.

[0134] Step 5: Perform spectral analysis on the real-time acquired differential pressure signal, extract spectral features, and determine whether there is catalyst layer flow instability. The high-frequency energy is calculated to be 0.80V², with a ratio of 3.20 to the background energy integral value, exceeding the second preset threshold of 2.0; the characteristic frequency amplitude at 1Hz is 0.32V, with a ratio of 4.00 to the background characteristic frequency amplitude, exceeding the preset multiple of 3 times. Therefore, it is determined that flow instability exists in the catalyst layer.

[0135] Step 6: No candidate ammonia injection branch was identified in Step 3, but Step 5 determined that flow instability existed. According to the control logic, the condition "a candidate ammonia injection branch exists and Step 5 determined that flow instability exists" is not met (because the candidate branch does not exist), therefore cleaning is not triggered. However, the central control module initiates ammonia injection distribution adjustment based on the flow instability diagnosis.

[0136] Steps 7 through 9: Cleaning was not performed.

[0137] Step 10: Cleaning was not performed, but after the system adjusted the ammonia injection distribution, steps 2 to 5 were repeated for verification. After 30 minutes, the differential pressure spectrum was remeasured. The high-frequency energy was 0.35V², with a ratio of 1.40 to the background energy integral value. The characteristic frequency amplitude at 1Hz was 0.10V, with a ratio of 1.25 to the background characteristic frequency amplitude. All values ​​returned to the preset range.

[0138] Implementation results: The system can promptly identify catalyst layer flow instability by using the spectral characteristics of differential pressure fluctuations, and quickly eliminate the anomaly by adjusting the ammonia injection distribution, thus avoiding the potential risk of local catalyst blockage. This verifies the diagnostic and response capabilities of the system under conditions without branch pipe blockage.

[0139] Example 5

[0140] This embodiment provides an adaptive control method and system for the ammonia injection grid in an SCR denitrification system of a thermal power plant, used for predictive maintenance via digital twins to avoid clogging. Specific implementation details include:

[0141] Objective: To verify the predictive maintenance function of the built-in digital twin model in the system of this invention, and to demonstrate its ability to intervene in advance to delay the development of dust accumulation.

[0142] Implementation system: Same as in Example 1, with a built-in digital twin model in the central control module. Baseline characteristic parameters are the same as above.

[0143] Implementation steps:

[0144] Step 1: Baseline establishment is complete.

[0145] Step 2: Collect data in real time. Sampling period is 1 minute.

[0146] Step 3: Based on the comparison between the real-time acquired temperature response curve and the reference temperature response characteristics, identify candidate ammonia injection branch pipes with ash accumulation. Currently, the deviation rate of each branch pipe is less than 1.3, and no candidate ammonia injection branch pipes have been identified.

[0147] Step 4: Reconstruct the ammonia concentration distribution field and flue gas temperature distribution field of the current flue cross section based on the real-time collected acoustic signals. No local areas were identified.

[0148] Step 5: Spectral analysis of the real-time acquired differential pressure signal revealed no flow instability.

[0149] Step 6: Identify the branch pipe without a clean target.

[0150] Steps 7 through 10: Not executed.

[0151] Digital twin prediction and intervention:

[0152] The central control module continuously inputs real-time data from the thermal impedance monitoring module, acoustic tomography module, and differential pressure monitoring module into the digital twin model;

[0153] The digital twin model outputs the predicted ash accumulation growth rate of each ammonia injection branch pipe, showing that the ash accumulation growth rate of branch pipe No. 7 is relatively fast, and the deviation rate is expected to reach the first preset threshold of 1.5 after 6 hours.

[0154] Based on this predicted value, the central control module adjusts the opening of the flow regulating valve of branch pipe No. 7 by 12% before the degree of dust accumulation reaches the first preset threshold.

[0155] Continuously monitor the changes in the deviation rate of branch pipe No. 7.

[0156] Implementation Results: After 4 hours, the deviation rate of branch pipe No. 7 only rose to 1.30, which did not reach the first preset threshold of 1.5, effectively avoiding cleaning triggering and extending the branch pipe maintenance cycle. This verified the predictive accuracy of the digital twin model and the effectiveness of early intervention.

[0157] Comparative Example 1

[0158] Purpose of implementation: To verify the technical superiority of the system of the present invention by comparing it with conventional control schemes.

[0159] System Implementation: A conventional PID control system is used. The ammonia injection grid has a traditional uniform partition structure, lacking thermal impedance monitoring, acoustic chromatography, differential pressure fluctuation analysis, and acoustic cleaning modules. Total volume regulation relies solely on NOx concentration feedback from the outlet CEMS, and the ammonia injection valves in each partition are manually adjustable.

[0160] Implementation steps:

[0161] Once the system is put into operation, set the initial opening degree of the ammonia injection valves in each zone according to the design values.

[0162] During operation, operators periodically adjust the total ammonia injection valve manually based on the outlet CEMS data;

[0163] When NOx fluctuations occur at the outlet or ammonia escape increases, operators try to adjust the valves in each zone based on their experience.

[0164] There are no branch-level monitoring methods, making it impossible to locate the blocked branch.

[0165] The furnace was shut down for maintenance after three months of operation.

[0166] Implementation results:

[0167] During operation, NOx fluctuations at the outlet were ±15 mg / m³, and ammonia slip was 5-8 ppm.

[0168] During the shutdown and maintenance, it was found that the nozzles of the ammonia injection branch pipes were blocked in many places, especially in the edge area of ​​the flue.

[0169] The pressure differential of the SCR catalyst layer increased by 30% compared to the initial value. Upon opening the cap, local signs of ammonium bisulfate blockage were found.

[0170] The blockage in each branch pipe needs to be checked and cleared manually, which will take 3 days.

[0171] Compared with the embodiments of the present invention:

[0172] In this embodiment of the invention, the NOx output of the unit during simultaneous operation is stable at 45±3 mg / m³, and the ammonia slip is <3 ppm.

[0173] Only three automatic cleaning events occurred, and no serious blockages were found;

[0174] The catalyst bed pressure differential increased by only 5%, and no local blockage occurred;

[0175] No manual cleaning of branch pipes is required; the system performs automatic maintenance.

[0176] The comparison shows that the system of the present invention can significantly improve the uniformity of ammonia injection, extend equipment life, and reduce maintenance costs, demonstrating outstanding technological progress.

[0177] Compared to Examples 1-5 and Comparative Example 1, this invention demonstrates its comprehensive technical effectiveness under different operating conditions through five examples. Example 1 verifies the stability of the system during normal operation, maintaining outlet NOx at a stable level of 45 mg / m³ and ammonia slip below 2 ppm without any intervention, proving that multi-module collaborative monitoring does not falsely trigger cleaning actions. Example 2 addresses a scenario with slight dust accumulation. The system proactively identified a deviation rate of 1.60 in branch pipe 12 through thermal impedance monitoring. Combining this with acoustic tomography and differential pressure analysis, it confirmed that immediate cleaning was unnecessary. Instead, a preventative measure of opening the regulating valve by 8% was taken, successfully controlling the deviation rate between 1.55 and 1.62, avoiding cleaning triggering, and verifying the effectiveness of early warning and preventative control. Example 3 addresses severe dust accumulation, with a deviation rate of 2.20 in branch pipe 5. The ammonia concentration in the corresponding area drops to 65% of the average, and the pressure differential high-frequency energy ratio reaches 2.52. The system accurately locates the cleaning target and removes the dust within 25 seconds using thermoacoustic resonance cleaning. After cleaning, the deviation rate recovers to 1.10, the ammonia concentration rises to 96%, and the pressure differential energy ratio drops to 1.32, achieving online self-healing without shutdown. Example 4 addresses catalyst layer flow instability (without branch pipe blockage). The system promptly diagnoses the issue based solely on pressure differential spectrum characteristics (energy ratio 3.20, characteristic frequency amplitude ratio 4.00) and restores the pressure differential energy ratio to 1.40 within 30 minutes through ammonia injection distribution adjustment, avoiding the risk of local catalyst blockage. Example 5 demonstrates the digital twin predictive maintenance function, predicting the risk of dust accumulation in branch pipe 7 6 hours in advance. By intervening with a 12% increase in valve opening, the actual deviation rate only rises to 1.30, effectively delaying the cleaning trigger. The five implementation examples fully cover the entire chain of intelligent capabilities, from routine monitoring, early prevention, precise cleaning, rapid response to advanced prediction.

[0178] Compared to conventional PID control, this invention offers significant advantages. The conventional method employs traditional uniform zone ammonia injection and total volume regulation, lacking any branch-level monitoring or automatic cleaning mechanisms. After three months of operation, outlet NOx fluctuations reached ±15 mg / m³, and ammonia slip increased to 5-8 ppm. A shutdown for maintenance revealed multiple branch blockages, a 30% increase in catalyst bed differential pressure, and signs of ammonium bisulfate blockage, requiring manual inspection of each branch, which took three days. In contrast, this invention quantifies branch-level ash accumulation through thermal impedance monitoring, visualizes cross-sectional ammonia concentration using acoustic tomography, senses catalyst bed health status through differential pressure spectrum analysis, and achieves online cleaning through thermoacoustic resonance. Combined with digital twin predictive maintenance, this allows the unit to maintain stable outlet NOx at 45±3 mg / m³, ammonia slip below 3 ppm, catalyst differential pressure increase of only 5%, and triggers automatic cleaning only three times without manual intervention within the same operating cycle. This significantly improves denitrification efficiency and equipment reliability, reduces operating costs, and demonstrates outstanding technological advancement.

[0179] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of protection of the claims.

[0180] In conclusion, the above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An adaptive control system for an ammonia injection grid in an SCR denitrification system of a thermal power plant, characterized in that, include: The ammonia injection grid module includes several parallel ammonia injection branch pipes, and each ammonia injection branch pipe inlet is equipped with a flow regulating valve; A thermal impedance monitoring module is installed on the outer wall of the ammonia injection branch pipe. It is used to apply thermal pulses to the corresponding ammonia injection branch pipe and collect the temperature response curve of the ammonia injection branch pipe. The acoustic tomography module includes an acoustic generator and an acoustic receiver arranged on opposite sides of the cross-section of the SCR inlet flue, for transmitting and receiving multipath acoustic signals; The differential pressure monitoring module includes a high-frequency differential pressure transmitter connected to the inlet and outlet sides of the SCR catalyst layer, used to collect differential pressure signals on both sides of the catalyst layer. The acoustic cleaning module includes a piezoelectric transducer coupled to the outer wall of each ammonia injection branch pipe for applying acoustic excitation to the corresponding ammonia injection branch pipe. The central control module is communicatively connected to the flow regulating valve, thermal impedance monitoring module, acoustic tomography module, differential pressure monitoring module, and acoustic cleaning module. The central control module is configured to: identify ammonia injection branch pipes with ash accumulation based on the temperature response curve acquired by the thermal impedance monitoring module; reconstruct the ammonia concentration distribution field and flue gas temperature distribution field of the SCR inlet flue cross section based on the acoustic signal acquired by the acoustic tomography module; extract spectral features based on the differential pressure signal acquired by the differential pressure monitoring module; verify the impact of ammonia injection branch pipes with ash accumulation on injection uniformity or catalyst layer stability based on the ammonia concentration distribution field or spectral features, and determine the target branch pipe for cleaning; determine the acoustic resonance frequency based on the flue gas temperature distribution field; control the acoustic cleaning module to apply acoustic excitation to the target branch pipe for cleaning at the resonance frequency, and control the thermal impedance monitoring module to apply auxiliary thermal pulses to the branch pipe synchronously or at different times; determine the cleaning endpoint based on the temperature response curve of the branch pipe acquired in real time during the cleaning process, and control the acoustic cleaning module and the thermal impedance monitoring module to stop working. Among them, the impact of ammonia injection branch pipes with ash accumulation on injection uniformity or catalyst layer stability was verified based on the ammonia concentration distribution field or spectral characteristics, and the specific clean target branch pipes were determined as follows: Based on the ammonia concentration distribution field reconstructed by the acoustic tomography module, local areas where the ammonia concentration is lower than the preset uniformity index are identified. Based on the spectral characteristics extracted by the differential pressure monitoring module, it is determined whether the spectral characteristics exceed the second preset threshold. If so, it is determined that there is flow instability in the catalyst layer. When it is determined that there is dust accumulation in the ammonia injection branch pipe, and the ammonia concentration in the corresponding injection area is locally low, or the catalyst layer is unstable, the branch pipe is identified as the cleaning target branch pipe. Specifically, the acoustic resonant frequency is determined based on the flue gas temperature distribution field as follows: The gas velocity in the branch pipe is calculated based on the flue gas temperature value of the area where the clean target branch pipe is located, extracted from the flue gas temperature distribution field. A preliminary estimate of the resonant frequency is determined based on the gas sound velocity, the geometry of the branch pipe, and the acoustic boundary conditions. The control acoustic cleaning module sweeps the frequency around the initial estimated value. At the same time, the temperature sensor corresponding to the branch pipe in the thermal impedance monitoring module detects the vibration signal of the pipe wall. When the amplitude of the vibration signal reaches a local maximum, the current frequency is locked as the resonance frequency.

2. The adaptive control system for ammonia injection grid in an SCR denitrification system of a thermal power plant as described in claim 1, characterized in that, The central control module is also configured to: upon initial commissioning or after each cleaning, control the thermal impedance monitoring module to apply thermal pulses to each ammonia injection branch pipe, collect the temperature response curve of each ammonia injection branch pipe, and extract characteristic parameters as the reference characteristic parameters of the branch pipe in the clean state. The acoustic tomography module is controlled to acquire acoustic signals, reconstruct the ammonia concentration distribution field of the cross section of the SCR inlet flue, and calculate the statistical value of ammonia concentration in each region of the cross section as the initial benchmark for ammonia concentration uniformity. The differential pressure monitoring module collects differential pressure signals, performs spectral analysis on the continuously collected differential pressure signals, and extracts spectral feature values ​​as spectral background features under stable operating conditions.

3. The adaptive control system for ammonia injection grid in an SCR denitrification system of a thermal power plant as described in claim 2, characterized in that, Based on the temperature response curve acquired by the thermal impedance monitoring module, the specific ammonia injection branch pipe with ash accumulation was identified as follows: The characteristic parameters of the real-time temperature response curve are compared with the baseline characteristic parameters to calculate the deviation value. When the deviation value exceeds the first preset threshold, it is determined that the branch pipe has a risk of dust accumulation. When the deviation value exceeds the second preset threshold, it is determined that the branch pipe is severely blocked. The second preset threshold is greater than the first preset threshold. The characteristic parameters of the temperature response curve include at least one of the following: heating rate, steady-state temperature difference, or time constant.

4. The adaptive control system for ammonia injection grid in an SCR denitrification system of a thermal power plant as described in claim 1, characterized in that, The ammonia concentration distribution field and flue gas temperature distribution field of the SCR inlet flue cross section are reconstructed based on the acoustic signals collected by the acoustic tomography module as follows: The sound velocity distribution of the flue cross section is obtained by inverting the flight time of the sound waves along each propagation path, and then converted into the flue gas temperature distribution field based on the correspondence between sound velocity and temperature. The attenuation coefficient distribution of the flue cross section is obtained by inverting the amplitude attenuation of sound waves along each propagation path, and then converted into an ammonia concentration distribution field based on the correspondence between the attenuation coefficient and the ammonia concentration.

5. The adaptive control system for ammonia injection grid in an SCR denitrification system of a thermal power plant as described in claim 1, characterized in that, The extraction of spectral features from the differential pressure signal acquired by the differential pressure monitoring module specifically includes: Perform a Fourier transform on the differential pressure signal to extract the energy integral value within a preset high-frequency band or the amplitude at at least one preset characteristic frequency. The determination of whether the spectral characteristics exceed the second preset threshold is as follows: when the energy integral value of the high frequency band exceeds the preset multiple of the corresponding value in the spectral background characteristics, or when the characteristic frequency amplitude exceeds the preset multiple of the corresponding value in the spectral background characteristics, it is determined that there is flow instability in the catalyst layer.

6. The adaptive control system for ammonia injection grid in an SCR denitrification system of a thermal power plant as described in claim 2, characterized in that, The cleaning endpoint is determined based on the temperature response curve of the branch pipe collected in real time during the cleaning process. The temperature response curve of the branch pipe is received in real time by the thermal impedance monitoring module. The deviation between its characteristic parameters and the reference characteristic parameters is calculated. When the deviation is lower than the third preset threshold, the cleaning is determined to be completed.

7. The adaptive control system for ammonia injection grid in an SCR denitrification system of a thermal power plant as described in claim 1, characterized in that, It also includes a digital twin model, and the central control module is configured as follows: The data received in real time from the thermal impedance monitoring module, the acoustic tomography module, and the differential pressure monitoring module are input into the digital twin model, which drives the digital twin model to output the predicted ammonia concentration distribution in the area of ​​the SCR inlet flue where no acoustic receiver is set or the predicted ash growth rate of each ammonia injection branch. When the predicted rate of ash accumulation exceeds the fourth preset threshold, the opening of the flow regulating valve of the branch pipe is adjusted before the degree of ash accumulation reaches the first preset threshold, thus delaying the development of ash accumulation.

8. An adaptive control method for an ammonia injection grid in an SCR denitrification system of a thermal power plant, applied to the system described in any one of claims 1-7, characterized in that, Includes the following steps: Step 1: Establish the baseline temperature response characteristics of each ammonia injection branch pipe under clean conditions, the baseline initial ammonia concentration distribution of the SCR inlet flue section, and the spectral background characteristics of the catalyst layer under stable operation conditions; Step 2: Real-time acquisition of temperature response curves, multipath acoustic signals of flue gas cross-section, and differential pressure signals of catalyst layer for each ammonia injection branch pipe; Step 3: Based on the comparison between the real-time temperature response curve and the reference temperature response characteristics, identify candidate ammonia injection branch pipes with ash accumulation; Step 4: Reconstruct the ammonia concentration distribution field and flue gas temperature distribution field of the current flue cross section based on the real-time collected acoustic signals, and identify local areas where the ammonia concentration is lower than the preset uniformity index. Step 5: Perform spectrum analysis on the real-time acquired differential pressure signal, extract spectrum features, and determine whether there is catalyst layer flow instability; Step 6: When the candidate ammonia injection branch identified in Step 3 corresponds to the local area space identified in Step 4, or when the candidate ammonia injection branch identified in Step 3 exists and Step 5 determines that there is flow instability, the candidate ammonia injection branch is identified as the cleaning target branch. Step 7: Extract the flue gas temperature value of the area where the clean target branch pipe is located from the flue gas temperature distribution field reconstructed in Step 4, and determine the acoustic resonant frequency based on the flue gas temperature value. Step 8: Control the acoustic cleaning module to apply acoustic excitation to the target branch pipe at the resonant frequency, and control the thermal impedance monitoring module to apply auxiliary thermal pulses to the branch pipe synchronously or at different times. Step 9: During the cleaning process, the temperature response curve of the branch pipe is collected in real time. When the deviation between the branch pipe and the reference temperature response characteristic is lower than the third preset threshold, the application of acoustic excitation and auxiliary heat pulse is stopped. Step 10: After cleaning, verify whether the temperature response characteristics of the branch pipe, the ammonia concentration in the corresponding area, and the differential pressure spectrum characteristics of the catalyst layer have returned to the preset range. If they have not returned, repeat steps 7 to 9 or output a maintenance alarm.

Images