Distributed SCR (Selective Catalytic Reduction) denitration system and precise ammonia spraying control method thereof

Abstract

The invention discloses a distributed SCR (Selective Catalytic Reduction) denitration system and an accurate ammonia spraying control method thereof, and relates to the field of flue gas removal. According to the method, a first collection point, a second collection point and a third collection point are arranged in the distributed SCR denitration system; a thermocouple array, a differential pressure sensor and a flow velocity sensor are respectively arranged at a top airflow inlet, a middle catalyst channel area and a bottom smoke outlet of the catalyst module, so that a disturbance acquisition system covering multiple areas of the tower body is formed, and gas temperature original data, differential pressure original data and flow velocity fluctuation original data are acquired. Feature extraction and normalization preprocessing are carried out on original disturbance data in a control platform to form a standardized disturbance ammonia spraying vector set, so that layered disturbance field perception and ammonia spraying fine control support are realized on the structural level, ammonia spraying configuration redundancy or insufficiency caused by local information loss is avoided, the ammonia water utilization efficiency of each module is improved, and the system performance is improved. And the ammonia spraying accuracy of the whole tower is improved.

Description

Technical Field

[0001] This invention relates to the field of flue gas removal, specifically to a distributed SCR denitrification system and its precise ammonia injection control method. Background Technology

[0002] This invention relates to the field of flue gas removal, and more particularly to a clean production technology for controlling nitrogen oxide (NOx) pollution. More specifically, it relates to a distributed selective catalytic reduction (SCR) denitrification system and its precise ammonia injection control method, which is applicable to low NOx emission scenarios in coal-fired, oil-fired boilers and industrial furnaces. It is used to improve ammonia water utilization efficiency and reduce the risk of ammonia escape while ensuring denitrification efficiency.

[0003] Currently, most SCR denitrification systems widely used in engineering adopt a centralized ammonia injection structure, that is, a single ammonia injection unit is set up upstream of the denitrification reactor to uniformly supply ammonia. However, due to the non-uniformity of NOx concentration distribution between catalyst layers, this method is difficult to achieve stratified optimization and adjustment, often resulting in excessive or insufficient ammonia injection in certain areas, leading to high ammonia slip values ​​or excessive NOx concentrations. In addition, existing systems generally rely on single-point sensor sampling, lacking fine perception of internal disturbance behavior in the tower, and the control algorithm has a slow response and cannot adaptively adjust the ammonia injection intensity in real time, which seriously restricts the improvement of denitrification efficiency and operational stability.

[0004] The root cause of these problems lies in the lack of monitoring capabilities for lateral and longitudinal disturbance distribution in traditional denitrification systems. Flow field disturbances, pressure differential fluctuations, and flue gas velocity changes between catalyst modules cannot be dynamically captured, causing upstream abnormal disturbances to propagate downwards undetected, resulting in spatial imbalance. This imbalance not only reduces the utilization rate of ammonia water reaction, causing ammonia escape values ​​to remain above 3 ppm in the uneconomical operating range for extended periods, but also may lead to frequent fluctuations in denitrification efficiency, causing NOx emission concentrations to exceed ultra-low emission limits. This poses legal compliance risks, increases environmental protection operation and maintenance costs, and seriously threatens the long-term stable operation of the flue gas purification system. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a distributed SCR denitrification system and its precise ammonia injection control method, solving the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a distributed SCR denitrification system and its precise ammonia injection control method, comprising the following steps: S1. Set up collection points in the distributed SCR denitrification system and set up sensor groups in the collection points to collect raw disturbance data in real time. Then, transmit the collected raw disturbance data to the denitrification control platform through the industrial bus protocol. S2. In the control platform, feature extraction is performed on the original disturbance data to obtain the disturbance ammonia injection vector set. The disturbance ammonia injection vector set is then preprocessed to obtain the standardized disturbance ammonia injection vector set. S3. Calculate based on the standardized disturbance ammonia injection vector set, output the disturbance intensity quantification index Spr, and conduct a preliminary comparative evaluation based on the output results of the disturbance intensity quantification index Spr. S4. Based on the preliminary comparative evaluation results, the denitrification trend analysis mechanism is triggered. The actual ammonia injection value Freal of the catalyst module is corrected according to the disturbance intensity quantification index Spr, and the corrected deviation value ΔF is recorded. Then, based on the disturbance intensity quantification index Spr and the deviation value ΔF, the denitrification trend deviation prediction value Bia is calculated and output. S5. The predicted value of denitrification trend deviation Bia is superimposed with the disturbance intensity quantification index Spr to calculate the tower-level synergistic score Sco. A secondary comparative evaluation is then performed based on the output of the tower-level synergistic score Sco.

[0007] Preferably, S1 includes S11 and S12; S12. The collection points include a first collection point, a second collection point, and a third collection point; and a sensor group is set in each collection point to collect the original disturbance data in real time. The first sampling point is located at the top airflow inlet of each catalyst module in the distributed SCR denitrification system; the second sampling point is located in the middle catalyst channel area; and the third sampling point is located at the bottom flue gas outlet. The sensor group includes a thermocouple array, a differential pressure sensor, and a flow velocity sensor; The original disturbance data includes original gas temperature data, original pressure difference data, and original flow velocity fluctuation data; S12. The disturbance sensor group inside the acquisition point preprocesses and packages the data through the local embedded data acquisition controller. After formatting the original disturbance data into Modbus RTU protocol data frames, it is sent to the denitrification control platform through the RS485 bus network. The platform receives and decodes the original disturbance data and stores it in the real-time control buffer through the PLC communication interface module.

[0008] Preferably, S2 includes S21; S21. In the denitrification control platform, feature extraction is performed on the original disturbance data to obtain the disturbance ammonia injection vector set; The disturbance ammonia injection vector set includes the radial rotational temperature difference Trot of the i-th catalyst module, the inlet-outlet shear pressure drop difference Pshear of the i-th catalyst module, and the standard deviation of the flue gas velocity Dgas of the i-th catalyst module. The radial rotational temperature difference Trot of the i-th catalyst module is obtained by calculating the radial distribution difference of the original gas temperature data. The inlet and outlet shear pressure drop difference Pshear of the i-th catalyst module is obtained by fitting the upstream and downstream pressure drop slopes to the raw pressure difference data. The standard deviation Dgas of the flue gas velocity of the i-th catalyst module is obtained by performing a joint analysis of the time-domain standard deviation and frequency-domain energy density on the raw data of the velocity fluctuation.

[0009] Preferably, S2 further includes S22; S22. The perturbation ammonia injection vector set is preprocessed. The preprocessing includes normalization. The normalization is performed by using a maximum and minimum value normalization algorithm to map each feature value to the [0,1] interval, and the normalization result is subjected to mean filtering to smooth short-term perturbation fluctuations, thus forming a standardized perturbation ammonia injection vector set.

[0010] Preferably, S3 includes S31; S31. Based on the standardized perturbation ammonia injection vector set, the perturbation intensity quantification index Spr is output. The perturbation intensity quantification index Spr is used to reflect the comprehensive quantitative value of the flue gas perturbation degree of each catalyst module. Its derivation is based on three perturbation feature quantities in the standardized perturbation ammonia injection vector set. The derivation process includes squaring the three perturbation feature quantities in the standardized perturbation ammonia injection vector set of each catalyst module, and then arithmetically averaging the square values ​​of the three perturbation feature quantities to balance the contribution of the perturbation dimension. Then, the square root of the square values ​​of the three perturbation feature quantities is processed to normalize the amplitude, and the dimensionless perturbation intensity quantification index Spr of each catalyst module is obtained.

[0011] Preferably, S3 further includes S32; S32. A preliminary comparative evaluation is performed on the disturbance intensity quantification index Spr by comparing Spr with a set disturbance classification threshold, wherein the disturbance classification threshold includes a first disturbance threshold and a second disturbance threshold; wherein When the disturbance intensity quantification index Spr is greater than the first disturbance threshold, it indicates an abnormal disturbance and triggers the denitrification trend analysis mechanism. When the disturbance intensity quantification index Spr is between the second disturbance threshold and the first disturbance threshold, it indicates a disturbance criticality. The ammonia injection control value is reduced by 10%, and the ammonia injection operation is delayed by 5 seconds. When the disturbance intensity quantification index Spr is less than the second disturbance threshold, it indicates that the disturbance is normal, and standard PID ammonia injection control is performed according to the NOx flue gas concentration. The denitrification trend analysis mechanism is triggered only when the disturbance intensity quantification index Spr is greater than the first disturbance threshold.

[0012] Preferably, S4 includes S41; S41. After the initial comparative evaluation triggers the destocking trend analysis mechanism, the ammonia injection flow rate is reduced by 15%, and ammonia injection is prohibited in the current control cycle. The corrected actual ammonia injection value Freal is obtained. The difference between the actual ammonia injection value Freal and the theoretical ammonia injection value Fbase is calculated to obtain the corrected deviation value ΔF. Among them, the theoretical ammonia injection value Fbase is the basic ammonia injection flow rate value output by the PID algorithm.

[0013] Preferably, S4 further includes S42; S42. The calculation of the denitrification trend deviation prediction value Biaj is derived based on the coupling relationship between the disturbance intensity quantification index Spri and the corrected ammonia injection deviation value ΔFi. The specific derivation process includes: for each target catalyst module, obtaining the deviation value ΔF and the disturbance intensity quantification index Spr of all upper catalyst modules, and using the product of the deviation value ΔF and the disturbance intensity quantification index Spr as the intensity factor of disturbance migration. Using the vertical physical distance d between catalyst modules as the square term of the attenuation factor, the influence intensity factors of all upper catalyst modules are weighted and superimposed to obtain the denitrification trend deviation prediction value Bia of the target catalyst module.

[0014] Preferably, S5 includes S51; S51. The derivation is based on the superposition and fusion of the denitrification trend deviation prediction value Bia and the disturbance intensity quantification index Spr. The derivation process includes: for all catalyst modules, the corresponding denitrification trend deviation prediction value Bia is divided by 1 and the disturbance intensity quantification index Spr is added to suppress the risk of the disturbed catalyst module. Then, the calculation results of all catalyst modules are arithmetically averaged to obtain the tower-level synergistic score Sco.

[0015] Preferably, S5 includes S52; S52. Compare the tower-level collaborative score Sco with the preset multi-level risk control thresholds to classify the system operating status levels and execute the corresponding control adjustment mechanisms. The control levels and corresponding mechanisms are as follows: When Sco is less than the first collaborative threshold, it is judged as the stable level of the distributed SCR denitrification system, and the current ammonia injection control strategy remains unchanged. When Sco is between the first and second collaborative thresholds, it is judged as the warning level of the distributed SCR denitrification system. The proportional gain Kp of the ammonia injection control PID of each catalyst module is increased by 10%, and the ammonia injection sampling period is shortened from 5 seconds to 1 second. When Sco is greater than the second collaborative threshold, the distributed SCR denitrification system is judged to be at a risk level, and the ammonia injection controller parameter reset strategy is executed. The ammonia injection controller parameter reset strategy specifically resets three parameters of the PID controller, namely proportional gain Kp, integral gain Ki, and derivative gain Kd. Specifically, the proportional gain Kp is reset to 0.45, the integral gain Ki is reset to 0.01, and the derivative gain Kd is reset to 0.20. At the same time, the tower airflow redistribution mechanism and the bypass diversion channel are triggered to achieve overall system stability control.

[0016] This invention provides a distributed SCR denitrification system and its precise ammonia injection control method. It has the following beneficial effects: (1) This method involves setting up a first, second, and third collection point in a distributed SCR denitrification system, and installing thermocouple arrays, differential pressure sensors, and flow rate sensors at the top gas inlet, middle catalyst channel area, and bottom flue gas outlet of the catalyst module, respectively, to form a disturbance acquisition system covering multiple areas of the tower, collecting raw data on gas temperature, differential pressure, and flow rate fluctuations; combining Modbus RTU protocol format data encapsulation and RS485 network high-reliability transmission to achieve real-time high-fidelity acquisition of the raw disturbance data. After feature extraction and normalization preprocessing of the raw disturbance data in the control platform, a standardized disturbance ammonia injection vector set is formed, thereby achieving layered disturbance field perception and fine ammonia injection control support at the structural level, avoiding redundancy or insufficiency in ammonia injection configuration due to missing local information, improving the ammonia water utilization efficiency of each module, and enhancing the overall tower ammonia injection accuracy.

[0017] (2) Based on the standardized ammonia injection vector set, this method constructs a dimensionless perturbation intensity quantification index Spr by squaring, merging the mean and normalizing the amplitude of the three perturbation features Trot, Pshear and Dgas, which quantitatively reflects the perturbation level of each catalyst module. When Spr exceeds the set perturbation classification threshold range, the denitrification trend analysis mechanism is triggered. The ammonia injection deviation value ΔF is calculated by combining the actual ammonia injection value Freal and the theoretical ammonia injection value Fbase. Based on the product of ΔF and Spr, the downstream migration process of reaction residue caused by perturbation between modules is modeled. By considering the vertical physical distance square term dij² as the attenuation factor, the denitrification trend deviation prediction value Bia is derived, thereby identifying the denitrification trend risk path in the system in advance, effectively suppressing the NOx penetration anomaly caused by the migration of perturbation to the lower layer, and improving the reaction controllability of the downstream area.

[0018] (3) The method constructs a tower-level collaborative score Sco (Sco) based on the fusion relationship between the predicted value of denitrification trend deviation (Bia) and the quantitative index of disturbance intensity (Spr), and conducts a unified assessment of the disturbance risk coupling degree of the entire catalyst module. By setting multi-level collaborative control thresholds, Sco is evaluated in segments, and the operating level of the SCR denitrification system is divided accordingly. Under different levels, control strategies of stability maintenance, sensitivity enhancement, or PID resetting are implemented respectively. These include increasing Kp gain by 10%, shortening the ammonia injection sampling cycle from 5 seconds to 1 second, and resetting Kp / Ki / Kd to 0.45 / 0.01 / 0.20 respectively, accompanied by the linkage triggering of the airflow redistribution mechanism and the bypass diversion channel. This mechanism can effectively reduce the system's dependence on manual intervention and construct an intelligent ammonia injection control system with fast response speed, strong disturbance tolerance, and high dynamic correction capability. Attached Figure Description

[0019] Figure 1 This is a schematic diagram illustrating the steps of a distributed SCR denitrification system and its precise ammonia injection control method according to the present invention. Figure 2 A schematic diagram showing the installation arrangement of each catalyst module above and below; Figure 3 This is a circuit diagram for data transmission. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Example 1 This invention provides a distributed SCR denitrification system and its precise ammonia injection control method. Please refer to [link / reference]. Figure 1 , Figure 2 and Figure 3 This includes the following steps: S1. Set up collection points in the distributed SCR denitrification system and set up sensor groups in the collection points to collect raw disturbance data in real time. Then, transmit the collected raw disturbance data to the denitrification control platform through the industrial bus protocol. S2. In the control platform, feature extraction is performed on the original disturbance data to obtain the disturbance ammonia injection vector set. The disturbance ammonia injection vector set is then preprocessed to obtain the standardized disturbance ammonia injection vector set. S3. Calculate based on the standardized disturbance ammonia injection vector set, output the disturbance intensity quantification index Spr, and conduct a preliminary comparative evaluation based on the output results of the disturbance intensity quantification index Spr. S4. Based on the preliminary comparative evaluation results, the denitrification trend analysis mechanism is triggered. The actual ammonia injection value Freal of the catalyst module is corrected according to the disturbance intensity quantification index Spr, and the corrected deviation value ΔF is recorded. Then, based on the disturbance intensity quantification index Spr and the deviation value ΔF, the denitrification trend deviation prediction value Bia is calculated and output. S5. The predicted value of denitrification trend deviation Bia is superimposed with the disturbance intensity quantification index Spr to calculate the tower-level synergistic score Sco. A secondary comparative evaluation is then performed based on the output of the tower-level synergistic score Sco.

[0022] In this embodiment, the method relies on a first, second, and third sampling point deployed in multiple areas within the SCR reactor to collect raw data on gas temperature, pressure difference, and flow rate fluctuations. This data is then uniformly transmitted to the denitrification control platform via the RS485 industrial bus protocol. A disturbance ammonia injection vector set is constructed through disturbance feature extraction, and normalization and mean filtering are performed to form a standardized disturbance ammonia injection vector set. Furthermore, a dimensionless disturbance intensity quantification index, Spr, is derived, and based on its comparison with the disturbance classification threshold, a denitrification trend analysis mechanism is triggered. By calculating the coupling effect between the corrected ammonia injection deviation value ΔF and the disturbance intensity quantification index Spr, and combining the vertical positional relationship of the modules, a denitrification trend deviation prediction value, Bia, is derived. Based on this, a tower-level collaborative score, Sco, is formed, achieving a unified characterization and classification of disturbance distribution among catalyst modules, ammonia injection control intensity, and risk level. This method not only enables three-dimensional dynamic perception of multi-point disturbances, precise response adjustment, and distributed feedback correction, but also promotes adaptive adjustment of PID parameters and optimization of ammonia injection rhythm through a Sco-driven collaborative scoring and hierarchical control mechanism. Simultaneously, it supports the coordinated execution of tower airflow redistribution mechanisms and bypass control, fundamentally enhancing the system's self-sensing, autonomous adjustment, and self-recovery capabilities in response to operational disturbances. Therefore, this invention significantly improves the ammonia utilization efficiency and ammonia injection distribution uniformity of the denitrification system, stably controlling ammonia slip concentration below 3 ppm, and achieving an overall ammonia nitrogen reaction utilization rate exceeding 92%. While meeting ultra-low emission standards, it reduces unit operating energy consumption and commissioning costs, demonstrating significant environmental, economic, and engineering promotion value.

[0023] Example 2 Please see Figure 1 , Figure 2 and Figure 3 Specifically: S1 includes S11 and S12; S12. The collection points include a first collection point, a second collection point, and a third collection point; and a sensor group is set up in each collection point to collect raw disturbance data in real time. The first sampling point is located at the top airflow inlet of each catalyst module in the distributed SCR denitrification system; the second sampling point is located in the middle catalyst channel area; and the third sampling point is located at the bottom flue gas outlet. The sensor array includes a thermocouple array, a differential pressure sensor, and a flow rate sensor; The raw disturbance data includes raw gas temperature data, raw pressure difference data, and raw flow velocity fluctuation data; S12. The disturbance sensor group inside the acquisition point preprocesses and packages the data through the local embedded data acquisition controller. After formatting the original disturbance data into Modbus RTU protocol data frames, it is sent to the denitrification control platform through the RS485 bus network. The platform receives and decodes the original disturbance data and stores it in the real-time control buffer through the PLC communication interface module.

[0024] In this embodiment, the method utilizes a distributed SCR denitrification system. Three key airflow regions within the denitrification tower are designated as the first, second, and third sampling points, corresponding to the top airflow inlet of the catalyst module, the middle catalyst channel region, and the bottom flue gas outlet, respectively. Each sampling point deploys a sensor array consisting of a thermocouple array, a differential pressure sensor, and a flow velocity sensor, enabling real-time, multi-dimensional acquisition of raw gas temperature, differential pressure, and flow velocity fluctuation data. The disturbance data output from these sensor arrays is preprocessed and packaged by a local embedded data acquisition controller, then uniformly formatted into Modbus RTU protocol data frames and transmitted to the denitrification control platform via an RS485 bus network. On the platform side, a PLC communication interface module is used for data decoding and real-time control caching, achieving high-efficiency and high-reliability data interaction throughout the entire process. Through the combined use of the aforementioned structural deployment and data transmission mechanism, a disturbance sensing network covering the entire elevation, cross-section, and time period of the distributed SCR denitrification system tower was constructed. This solves the problems of delayed response and inability to reflect the non-uniformity of airflow between catalyst layers in traditional single-point sampling, enabling rapid identification and dynamic quantification of disturbance sources. Furthermore, local preprocessing and bus communication significantly reduce signal transmission delay and communication failure rate, improving data processing efficiency and providing a stable and accurate data foundation for subsequent core processes such as disturbance vector construction, ammonia injection control adjustment, and trend analysis.

[0025] Example 3 Please see Figure 1 and Figure 3 Specifically: S2 includes S21; S21. In the denitrification control platform, feature extraction is performed on the original disturbance data to obtain the disturbance ammonia injection vector set; The perturbation ammonia injection vector set includes the radial rotational temperature difference Trot of the i-th catalyst module, the inlet-outlet shear pressure drop difference Pshear of the i-th catalyst module, and the standard deviation of the flue gas velocity Dgas of the i-th catalyst module. The radial rotational temperature difference Trot of the i-th catalyst module is obtained by calculating the radial distribution difference of the raw gas temperature data. The inlet and outlet shear pressure drop difference Pshear of the i-th catalyst module is obtained by fitting the upstream and downstream pressure drop slopes to the raw pressure difference data. The standard deviation Dgas of the flue gas velocity of the i-th catalyst module was obtained by performing a joint analysis of the time-domain standard deviation and frequency-domain energy density of the raw velocity fluctuation data.

[0026] S2 also includes S22; S22. The ammonia injection vector set of disturbance is preprocessed. The preprocessing includes normalization. The normalization process maps each feature value to the [0,1] interval by using the maximum and minimum value normalization algorithm. The normalization result is then subjected to mean filtering to smooth short-term disturbance fluctuations, thus forming a standardized ammonia injection vector set of disturbance.

[0027] In this embodiment, the method constructs a perturbation ammonia injection vector set based on raw perturbation data received from multiple acquisition points through a denitrification control platform and structured analysis of different types of data. Specifically, the radial distribution difference along the catalyst module cross-section is calculated from the raw gas temperature data to extract the radial rotational temperature difference Trot for each i-th catalyst module; the pressure difference trend analysis is performed on the raw pressure difference data at the upstream and downstream positions of the module, and the pressure difference slope is fitted using the least squares method to obtain the shear pressure drop difference Pshear; time-domain standard deviation analysis and frequency-domain energy density spectrum analysis are performed on the raw flow velocity fluctuation data, and the results of the two characteristic quantity calculations are fused to construct the flue gas velocity standard deviation Dgas. These three perturbation characteristic parameters together constitute the perturbation ammonia injection vector set describing the perturbation state of the catalyst module operation. To avoid information offset in scale, dimension, and fluctuation characteristics of the raw perturbation features, a standardization preprocessing mechanism is introduced in this embodiment. First, a maximum and minimum value normalization algorithm is used to normalize the perturbation ammonia injection vector set to uniformly map the perturbation ammonia injection vector set to the [0,1] interval, ensuring the consistency of subsequent processing in the numerical space dimension. Then, a moving average filtering algorithm is applied to each normalized perturbation feature quantity to smooth short-term high-frequency fluctuations, reduce noise interference, and improve feature stability. Finally, a standardized perturbation ammonia injection vector set is constructed as a unified input parameter source.

[0028] Example 4 Please see Figure 1 Specifically: S3 includes S31; S31. Based on the standardized perturbation ammonia injection vector set, the perturbation intensity quantification index Spr is output. The perturbation intensity quantification index Spr is used to reflect the comprehensive quantitative value of the flue gas perturbation degree of each catalyst module. Its derivation is based on three perturbation feature quantities in the standardized perturbation ammonia injection vector set. The derivation process includes squaring the three perturbation feature quantities in the standardized perturbation ammonia injection vector set of each catalyst module to enhance the sensitivity to high values, and then arithmetically averaging the squared values ​​of the three perturbation feature quantities to balance the contribution of the perturbation dimension. Finally, the square root of the squared values ​​of the three perturbation feature quantities is processed to normalize the amplitude, and the dimensionless perturbation intensity quantification index Spr of each catalyst module is obtained.

[0029] S3 also includes S32; S32. A preliminary comparative evaluation is performed on the disturbance intensity quantification index Spr by comparing Spr with a set disturbance classification threshold, which includes a first disturbance threshold and a second disturbance threshold; wherein... When the disturbance intensity quantification index Spr is greater than the first disturbance threshold, it indicates an abnormal disturbance and triggers the denitrification trend analysis mechanism. When the disturbance intensity quantification index Spr is between the second disturbance threshold and the first disturbance threshold, it indicates a disturbance criticality. The ammonia injection control value is reduced by 10%, and the ammonia injection operation is delayed by 5 seconds. When the disturbance intensity quantification index Spr is less than the second disturbance threshold, it indicates that the disturbance is normal, and standard PID ammonia injection control is performed according to the NOx flue gas concentration. The denitrification trend analysis mechanism is triggered only when the disturbance intensity quantification index Spr is greater than the first disturbance threshold.

[0030] In this embodiment, to accurately characterize the comprehensive intensity of disturbances experienced by the catalyst module, a disturbance intensity quantification index Spr is introduced as an aggregated expression of multi-dimensional disturbance information vectors. Specifically, firstly, three disturbance feature quantities for each catalyst module are extracted based on a standardized disturbance ammonia injection vector set: radial rotation temperature difference Trot, inlet-outlet shear pressure drop difference Pshear, and flue gas velocity standard deviation Dgas. To enhance the response sensitivity of each feature under high disturbance conditions, each feature value is squared and its arithmetic mean is calculated to balance the contribution of different disturbance dimensions to the overall disturbance intensity. Subsequently, the square root of the result is taken to complete the amplitude normalization process, finally obtaining the dimensionless disturbance intensity quantification index Spr. This formula realizes a unified measurement of different types of disturbance features and can stably and sensitively reflect the current disturbance complexity of the catalyst module. Based on this index, the system further designs a disturbance classification evaluation mechanism. By setting a first disturbance threshold and a second disturbance threshold, the disturbance intensity quantification index Spr is classified into intervals to drive different levels of ammonia injection control strategies. When the disturbance intensity quantification index Spr is less than the second disturbance threshold, the current disturbance level is determined to be stable, and the system adopts the standard NOx concentration closed-loop PID control method to perform conventional ammonia injection; when the disturbance intensity quantification index Spr is between the second disturbance threshold and the first disturbance threshold, it is characterized as a disturbance critical state, and the ammonia injection control value is actively reduced by 10%, and the ammonia injection action is delayed by 5 seconds to avoid excessive ammonia injection due to instantaneous fluctuations; when the disturbance intensity quantification index Spr exceeds the first disturbance threshold, it is determined to be a disturbance abnormality, and the denitrification trend analysis mechanism is immediately triggered to enter the trend prediction and flow correction control process.

[0031] Example 5 Please see Figure 1 Specifically: S4 includes S41; S41. After the initial comparative evaluation triggers the destocking trend analysis mechanism, the ammonia injection flow rate is reduced by 15%, and ammonia injection is prohibited in the current control cycle. The corrected actual ammonia injection value Freal is obtained. The difference between the actual ammonia injection value Freal and the theoretical ammonia injection value Fbase is calculated to obtain the corrected deviation value ΔF. Among them, the theoretical ammonia injection value Fbase is the basic ammonia injection flow rate value output by the PID algorithm.

[0032] S4 also includes S42; S42. The calculation of the denitrification trend deviation prediction value Biaj is derived based on the coupling relationship between the disturbance intensity quantification index Spri and the corrected ammonia injection deviation value ΔFi. The derivation is based on the following: the ammonia injection correction behavior caused by the disturbance in the upper catalyst module will cause unreacted ammonia and NOx to migrate to the lower layer, and the degree of influence will decrease rapidly with the increase of the vertical distance between layers. The specific derivation process includes: for each target catalyst module, obtaining the deviation value ΔF and the disturbance intensity quantification index Spr of all the catalyst modules above it, and using the product of the deviation value ΔF and the disturbance intensity quantification index Spr as the intensity factor of the disturbance migration, and using the vertical physical distance d between the catalyst modules as the square term of the attenuation factor, the influence intensity factors of all the upper catalyst modules are weighted and superimposed to obtain the denitrification trend deviation prediction value Bia of the target catalyst module; The denitrification trend deviation prediction value Bia is used to assess the downward trend in denitrification efficiency that may be transmitted downstream due to insufficient ammonia injection caused by upper-layer disturbances, and serves as the trigger for whether to implement lower-layer early ammonia injection compensation or activate bypass control.

[0033] In this embodiment, the method first evaluates the results of a preliminary comparison between the disturbance intensity quantification index Spr and the classification threshold. If it is determined that Spr has exceeded the first disturbance threshold, the denitrification trend analysis mechanism is triggered. The system immediately executes the ammonia injection flow rate reduction correction strategy, reducing the theoretical ammonia injection value Fbase output by the original PID by 15%, and suspending ammonia injection execution within the current control cycle to avoid overcompensation of denitrification control caused by transient disturbance propagation. At this time, the actual ammonia injection value Freal will deviate from Fbase, and the difference ΔF between the two constitutes the dynamic correction amount of ammonia injection control. For a specific numerical example, assuming the ammonia injection deviation ΔF of the first, second, and third catalyst modules are 0.08, 0.12, and 0.05 respectively, corresponding to Spr values ​​of 0.68, 0.75, and 0.62, and dij values ​​of 3m, 2m, and 1m respectively, the predicted denitrification trend deviation value Bia of the fourth target module is calculated as follows: Bia4 = 0.08 × 0.68 / 3 2 +0.12×0.75 / 2 2 +0.05×0.62 / 1 2 =0.05954; that is, the fourth-layer catalyst module faces a NOx treatment capacity decline trend of approximately 0.059 in the next cycle. This value exceeds the system's set deviation response threshold. For example, when it reaches 0.05, it will trigger early ammonia injection compensation or open the bypass control mechanism. In summary, by introducing the construction method of the denitrification trend deviation prediction value Bia, not only can the disturbance itself be perceived in real time, but the disturbance propagation trend can also be assessed, and the downstream control strategy can be adjusted in advance to effectively avoid NOx emission exceedance caused by disturbance accumulation. At the same time, it improves the utilization efficiency of ammonia injection resources, enhances the system's response robustness to complex flue gas disturbances, and significantly improves the overall denitrification stability, accuracy, and adaptive control capability of the SCR system.

[0034] Example 6 Please see Figure 1 Specifically: S5 includes S51; S51. The derivation is based on the superposition and fusion of the predicted value of denitrification trend deviation Bia and the quantitative index of disturbance intensity Spr. The derivation is based on the following: the predicted value of denitrification trend deviation Bia is the potential NOx deviation risk to the target module caused by insufficient ammonia injection due to upper disturbance, and the quantitative index of disturbance intensity Spr is used to reflect the current operational instability of the module itself. The combination of the two can construct a composite index reflecting the degree of risk synergy of the entire tower system. The derivation process includes: for all catalyst modules, the corresponding predicted value of denitrification trend deviation Bia is divided by 1 and the value of the quantitative index of disturbance intensity Spr is added to suppress the risk of the disturbed catalyst module. Then, the calculation results of all catalyst modules are arithmetically averaged to obtain the tower-level synergy score Sco.

[0035] S5 includes S52; S52. Compare the tower-level collaborative score Sco with the preset multi-level risk control thresholds to classify the system operating status levels and execute the corresponding control adjustment mechanisms. The control levels and corresponding mechanisms are as follows: When Sco is less than the first collaborative threshold, it is judged as the stable level of the distributed SCR denitrification system, and the current ammonia injection control strategy remains unchanged. When Sco is between the first and second collaborative thresholds, it is judged as the early warning level of the distributed SCR denitrification system. The proportional gain Kp of the ammonia injection control PID of each catalyst module is increased by 10%, and the ammonia injection sampling period is shortened from 5 seconds to 1 second to improve the system response rate and ammonia injection control accuracy. When Sco exceeds the second collaborative threshold, the distributed SCR denitrification system is classified as risky, and an ammonia injection controller parameter reset strategy is executed. Specifically, the ammonia injection controller parameter reset strategy resets three parameters of the PID controller, namely proportional gain Kp, integral gain Ki, and derivative gain Kd. The specific reset values ​​are: proportional gain Kp is reset to 0.45, integral gain Ki is reset to 0.01, and derivative gain Kd is reset to 0.20. At the same time, the tower airflow redistribution mechanism and bypass diversion channel are triggered to achieve overall system stability control.

[0036] In this embodiment, to achieve coordinated optimization control at the tower scale of the distributed SCR denitrification system, a tower-level coordinated score (Sco) is introduced as a comprehensive criterion for evaluating the degree of risk coupling of the entire tower. First, in step S51, the system uses the predicted denitrification trend deviation value (Bia) of each catalyst module and the disturbance intensity quantification index (Spri) as core inputs. The design logic of this formula is as follows: the predicted denitrification trend deviation value (Bia) measures the transmission impact of upper-level disturbances on the NOx risk of the current module, while the disturbance intensity quantification index (Spri) reflects the local disturbance sensitivity of the current module. By dividing Bia by (1+Spri), while amplifying the impact of high Bia values, the suppression effect of Spr is introduced, enabling the assignment of risk constraint factors to "regions with high disturbance risk but large module volatility," thereby avoiding control misguidance. Finally, the arithmetic mean of the weighted risk values ​​of all modules is calculated to obtain the tower-level coordinated score (Sco), which serves as an evaluation index of the system's denitrification coordinated situation.

[0037] For example, if the predicted deviation values ​​(Bia) of the denitrification trend for four catalyst modules in a system are 0.08, 0.06, 0.05, and 0.07, respectively, and the corresponding disturbance intensity quantification index (Spri) values ​​are 0.5, 0.3, 0.6, and 0.4, then the calculation process of the tower-level synergistic score (Sco) is as follows: Sco=1 / 4(0.08 / 1+0.5+0.06 / 1+0.3+0.05 / 1+0.6+0.07 / 1+0.4)=0.0452 After completing the Sco calculation, step S52 is executed. By comparing Sco with the system's preset multi-level risk control thresholds, the system can intelligently classify the system into three operating levels and implement corresponding control strategies. When Sco is lower than the first coordination threshold, such as 0.03, the system is judged to be in a stable state and no adjustment is made. When Sco is between the first and second synergistic thresholds, such as 0.03–0.06, the system enters the warning level. At this time, the PID proportional gain Kp of each catalyst module is increased by 10%, and the sampling period is shortened from 5 seconds to 1 second to enhance the adjustment response speed. If Sco exceeds the second coordination threshold, such as 0.06, the system determines it to be at a risk level and immediately executes the PID parameter reset strategy: Kp=0.45, Ki=0.01, Kd=0.20, triggering the tower airflow redistribution mechanism and bypass diversion channel to ensure the stability of the denitrification system. Through the above implementation method, this invention constructs a tower-scale closed-loop control logic with multi-source disturbance perception, trend risk fusion assessment, and system-level response linkage on the basis of the original PID local adjustment, realizing a technological leap from "module response" to "system-level coordinated response". Its beneficial effects include: effectively avoiding system instability caused by the downstream propagation of local disturbances, enhancing the adaptive capability of the SCR system, reducing the unstable fluctuation rate of NOx emissions, improving predictive controllability, and improving the accuracy of ammonia water dosing. Ammonia escape control is significantly optimized, and the overall safety margin and operating economy of the denitrification system are improved.

[0038] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A distributed SCR denitrification system and its precise ammonia injection control method, characterized in that: Includes the following steps: S1. Set up collection points in the distributed SCR denitrification system and set up sensor groups in the collection points to collect raw disturbance data in real time. Then, transmit the collected raw disturbance data to the denitrification control platform through the industrial bus protocol. S2. In the control platform, feature extraction is performed on the original disturbance data to obtain the disturbance ammonia injection vector set. The disturbance ammonia injection vector set is then preprocessed to obtain the standardized disturbance ammonia injection vector set. S3. Calculate based on the standardized disturbance ammonia injection vector set, output the disturbance intensity quantification index Spr, and conduct a preliminary comparative evaluation based on the output results of the disturbance intensity quantification index Spr. S4. Based on the preliminary comparative evaluation results, the denitrification trend analysis mechanism is triggered. The actual ammonia injection value Freal of the catalyst module is corrected according to the disturbance intensity quantification index Spr, and the corrected deviation value ΔF is recorded. Then, based on the disturbance intensity quantification index Spr and the deviation value ΔF, the denitrification trend deviation prediction value Bia is calculated and output. S5. The predicted value of denitrification trend deviation Bia is superimposed with the disturbance intensity quantification index Spr to calculate the tower-level synergistic score Sco. A secondary comparative evaluation is then performed based on the output of the tower-level synergistic score Sco.

2. The distributed SCR denitrification system and its precise ammonia injection control method according to claim 1, characterized in that: S1 includes S11 and S12; S12. The collection points include a first collection point, a second collection point, and a third collection point; and a sensor group is set in each collection point to collect the original disturbance data in real time. The first sampling point is located at the top airflow inlet of each catalyst module in the distributed SCR denitrification system; the second sampling point is located in the middle catalyst channel area; and the third sampling point is located at the bottom flue gas outlet. The sensor group includes a thermocouple array, a differential pressure sensor, and a flow velocity sensor; The original disturbance data includes original gas temperature data, original pressure difference data, and original flow velocity fluctuation data; S12. The disturbance sensor group inside the acquisition point preprocesses and packages the data through the local embedded data acquisition controller. After formatting the original disturbance data into Modbus RTU protocol data frames, it is sent to the denitrification control platform through the RS485 bus network. The platform receives and decodes the original disturbance data and stores it in the real-time control buffer through the PLC communication interface module.

3. The distributed SCR denitrification system and its precise ammonia injection control method according to claim 1, characterized in that: S2 includes S21; S21. In the denitrification control platform, feature extraction is performed on the original disturbance data to obtain the disturbance ammonia injection vector set; The disturbance ammonia injection vector set includes the radial rotational temperature difference Trot of the i-th catalyst module, the inlet-outlet shear pressure drop difference Pshear of the i-th catalyst module, and the standard deviation of the flue gas velocity Dgas of the i-th catalyst module. The radial rotational temperature difference Trot of the i-th catalyst module is obtained by calculating the radial distribution difference of the original gas temperature data. The inlet and outlet shear pressure drop difference Pshear of the i-th catalyst module is obtained by fitting the upstream and downstream pressure drop slopes to the raw pressure difference data. The standard deviation Dgas of the flue gas velocity of the i-th catalyst module is obtained by performing a joint analysis of the time-domain standard deviation and frequency-domain energy density on the raw data of the velocity fluctuation.

4. The distributed SCR denitrification system and its precise ammonia injection control method according to claim 3, characterized in that: S2 further includes S22; S22. The perturbation ammonia injection vector set is preprocessed. The preprocessing includes normalization. The normalization is performed by using a maximum and minimum value normalization algorithm to map each feature value to the [0,1] interval, and the normalization result is subjected to mean filtering to smooth short-term perturbation fluctuations, thus forming a standardized perturbation ammonia injection vector set.

5. The distributed SCR denitrification system and its precise ammonia injection control method according to claim 4, characterized in that: S3 includes S31; S31. Based on the standardized perturbation ammonia injection vector set, the perturbation intensity quantification index Spr is output. The perturbation intensity quantification index Spr is used to reflect the comprehensive quantitative value of the flue gas perturbation degree of each catalyst module. Its derivation is based on three perturbation feature quantities in the standardized perturbation ammonia injection vector set. The derivation process includes squaring the three perturbation feature quantities in the standardized perturbation ammonia injection vector set of each catalyst module, and then arithmetically averaging the square values ​​of the three perturbation feature quantities to balance the contribution of the perturbation dimension. Then, the square root of the square values ​​of the three perturbation feature quantities is processed to normalize the amplitude, and the dimensionless perturbation intensity quantification index Spr of each catalyst module is obtained.

6. The distributed SCR denitrification system and its precise ammonia injection control method according to claim 5, characterized in that: S3 further includes S32; S32. A preliminary comparative evaluation is performed on the disturbance intensity quantification index Spr by comparing the disturbance intensity quantification index Spr with the set disturbance classification threshold, wherein the disturbance classification threshold includes a first disturbance threshold and a second disturbance threshold. in When the disturbance intensity quantification index Spr is greater than the first disturbance threshold, it indicates an abnormal disturbance and triggers the denitrification trend analysis mechanism. When the disturbance intensity quantification index Spr is between the second disturbance threshold and the first disturbance threshold, it indicates a disturbance criticality. The ammonia injection control value is reduced by 10%, and the ammonia injection operation is delayed by 5 seconds. When the disturbance intensity quantification index Spr is less than the second disturbance threshold, it indicates that the disturbance is normal, and standard PID ammonia injection control is performed according to the NOx flue gas concentration. The denitrification trend analysis mechanism is triggered only when the disturbance intensity quantification index Spr is greater than the first disturbance threshold.

7. The distributed SCR denitrification system and its precise ammonia injection control method according to claim 6, characterized in that: S4 includes S41; S41. After the initial comparative evaluation triggers the destocking trend analysis mechanism, the ammonia injection flow rate is reduced by 15%, and ammonia injection is prohibited in the current control cycle. The corrected actual ammonia injection value Freal is obtained. The difference between the actual ammonia injection value Freal and the theoretical ammonia injection value Fbase is calculated to obtain the corrected deviation value ΔF. Among them, the theoretical ammonia injection value Fbase is the basic ammonia injection flow rate value output by the PID algorithm.

8. The distributed SCR denitrification system and its precise ammonia injection control method according to claim 7, characterized in that: S4 further includes S42; S42. The calculation of the denitrification trend deviation prediction value Biaj is derived based on the coupling relationship between the disturbance intensity quantification index Spri and the corrected ammonia injection deviation value ΔFi. The specific derivation process includes: for each target catalyst module, obtaining the deviation value ΔF and the disturbance intensity quantification index Spr of all upper catalyst modules, and using the product of the deviation value ΔF and the disturbance intensity quantification index Spr as the intensity factor of disturbance migration. Using the vertical physical distance d between catalyst modules as the square term of the attenuation factor, the influence intensity factors of all upper catalyst modules are weighted and superimposed to obtain the denitrification trend deviation prediction value Bia of the target catalyst module.

9. A distributed SCR denitrification system and its precise ammonia injection control method according to claim 8, characterized in that: S5 includes S51; S51. The derivation is based on the superposition and fusion of the denitrification trend deviation prediction value Bia and the disturbance intensity quantification index Spr. The derivation process includes: for all catalyst modules, the corresponding denitrification trend deviation prediction value Bia is divided by 1 and the disturbance intensity quantification index Spr is added to suppress the risk of the disturbed catalyst module. Then, the calculation results of all catalyst modules are arithmetically averaged to obtain the tower-level synergistic score Sco.

10. A distributed SCR denitrification system and its precise ammonia injection control method according to claim 9, characterized in that: S5 includes S52; S52. Compare the tower-level collaborative score Sco with the preset multi-level risk control thresholds to classify the system operating status levels and execute the corresponding control adjustment mechanisms. The control levels and corresponding mechanisms are as follows: When Sco is less than the first collaborative threshold, it is judged as the stable level of the distributed SCR denitrification system, and the current ammonia injection control strategy remains unchanged. When Sco is between the first and second collaborative thresholds, it is judged as the warning level of the distributed SCR denitrification system. The proportional gain Kp of the ammonia injection control PID of each catalyst module is increased by 10%, and the ammonia injection sampling period is shortened from 5 seconds to 1 second. When Sco is greater than the second collaborative threshold, the distributed SCR denitrification system is judged to be at a risk level, and the ammonia injection controller parameter reset strategy is executed. The ammonia injection controller parameter reset strategy specifically resets three parameters of the PID controller, namely proportional gain Kp, integral gain Ki, and derivative gain Kd. Specifically, the proportional gain Kp is reset to 0.45, the integral gain Ki is reset to 0.01, and the derivative gain Kd is reset to 0.

20. At the same time, the tower airflow redistribution mechanism and the bypass diversion channel are triggered to achieve overall system stability control.

Images