Desulfurization slurry blinding early warning control system and operation method thereof

Abstract

The application discloses a desulfurization slurry blinding early warning control system and an operation method thereof. ‑ concentration, Al 3+ concentration, Fe 2+ / Fe 3+ total concentration, SO3 2‑ concentration and pH value; a data processing and early warning module adopts a multi-factor risk index model, generates a graded early warning signal based on the physical index data collected by monitoring; a central control system sends different control instructions to an execution module according to different early warning signals sent by the data processing and early warning module; the execution module includes a raw material replacement module, a pollution control module, a limestone slurry adding module and an oxidation air volume adjusting module, and is used for executing the control instructions sent by the central control system respectively. The application realizes real-time early warning on the slurry blinding risk by monitoring key parameters and combining the multi-factor risk index model, and the system automatically executes the control instructions, reduces human intervention and improves the stability and safety of the desulfurization process.

Description

Technical Field

[0001] This invention belongs to the field of wet flue gas desulfurization (FGD) technology, specifically relating to a desulfurization slurry blindness early warning control system based on multi-parameter online monitoring and its operation method. Background Technology

[0002] In traditional wet flue gas desulfurization (FGD) systems, limestone or lime slurry reacts with SO2-containing flue gas to form calcium sulfite (CaSO3·0.5H2O), which is then converted into gypsum (CaSO4·2H2O) under the action of oxidizing air. However, coal combustion flue gas often contains fluorides and heavy metals (such as Mn). 2+ Fe 2+ Al 3+ Components such as limestone (e.g., limestone slurry) accumulate in the desulfurization slurry, causing precipitation blockage and even system "poisoning" failure. Traditional control methods rely solely on on-site experience to adjust the limestone slurry supply and periodic chemical analysis, which cannot achieve real-time early warning and rapid response. This easily leads to large system fluctuations, increased downtime, and reduced desulfurization efficiency and equipment reliability.

[0003] To address the aforementioned issues, this invention proposes a desulfurization slurry blindness early warning and control system based on multi-factor fusion for risk level assessment. The system includes an online monitoring unit, a data processing and early warning module, a control unit, and actuators. It can monitor key parameters in the slurry in real time, such as fluoride ions, heavy metal ions, and pH value, and determine the system's risk level through a data fusion algorithm. When the risk level reaches a preset threshold, the system automatically triggers measures such as wastewater discharge, pH adjustment, and oxidation air volume adjustment to achieve closed-loop rapid correction. This technology has been successfully applied in wet FGD systems at two power plants, significantly improving desulfurization efficiency and reducing equipment failure rates, demonstrating promising application prospects. Summary of the Invention

[0004] This invention proposes a desulfurization slurry blindness early warning and control system, including an online monitoring unit, a data processing and early warning module, a control unit, and an actuator; and designs a tiered collaborative operation method, which judges the risk level through multi-factor fusion, and sequentially triggers the adjustment of environmental protection system operating parameters → replacement of abnormal substances → replacement of blinding slurry, so as to achieve closed-loop rapid correction.

[0005] The technical solution adopted in this invention is as follows:

[0006] A desulfurization slurry blindness early warning and control system, comprising:

[0007] The online monitoring unit is used to collect the physical properties of the desulfurization slurry in real time. These physical properties include the following factors: F - Concentration, Al 3+ Concentration, Fe 2+ / Fe3+ Total concentration, SO3 2- Concentration and pH value;

[0008] The online monitoring unit transmits the collected physical property data to the data processing and early warning module for processing;

[0009] The data processing and early warning module adopts a multi-factor risk index model to generate graded early warning signals based on the physical property index data collected through monitoring.

[0010] The central control system coordinates the closed-loop automatic operation of each unit and achieves adaptive optimization. It sends different control commands to the execution module according to different early warning signals issued by the data processing and early warning module. The central control system also receives data collected from the online monitoring unit for adaptive adjustment.

[0011] The execution module includes a raw material replacement module, a sewage control module, a limestone slurry dosing module, and an oxidation air volume adjustment module. Each module of the execution module is used to execute control commands issued by the central control system.

[0012] Furthermore, the online monitoring unit uses an ion-selective electrode to detect F in the desulfurization slurry. - Concentration and SO3 2- The concentration of Al in the desulfurization slurry was detected using atomic absorption spectrometry or an electrochemical sensor. 3+ Concentration, Fe 2+ / Fe 3+ The total concentration was determined by measuring the pH value of the desulfurization slurry using a pH meter.

[0013] Furthermore, the multi-factor risk index model comprehensively considers the monitored and collected physical property data and the weighting coefficients of each physical property to calculate the risk index. RI Calculate using the following formula:

[0014] ;

[0015] , , , Representing F respectively - Concentration, Al 3+ Concentration, Fe 2+ / Fe 3+ Total concentration and SO3 2- The threshold values ​​for the four factors of concentration, a1, a2, a3, a4, and a5, are respectively F - Concentration, Al 3+ Concentration, Fe 2+ / Fe 3+ Total concentration, SO32- Weighting coefficients for the five factors: concentration, pH value, etc.

[0016] , , , These represent the real-time collected F values. - Concentration, Al 3+ Concentration, Fe 2+ / Fe 3+ Total concentration and SO3 2- Concentration; p H This represents the pH value of the desulfurization slurry collected in real time. H opt ΔpH represents the average pH value of the desulfurization slurry during normal operation, while ΔpH represents the maximum pH value of the desulfurization slurry during normal operation.

[0017] The thresholds for the above factors are set based on the historical normal operation data of the desulfurization tower, that is, based on the historical data of the physical properties of the desulfurization slurry being within the normal range during normal operation of the desulfurization tower. The thresholds set for each factor are the average values ​​of their respective ranges.

[0018] The data processing and early warning module triggers different levels of early warning signals based on the magnitude of the risk index RI value generated by the multi-factor risk index model.

[0019] Furthermore, when RI>0.7, RI>0.85, and RI>1.0, the data processing and early warning module triggers Level I, II, and III early warnings, respectively.

[0020] Based on the triggered Level I warning signal, the central control system sends control commands to the limestone slurry dosing module and / or oxidation air volume regulating module to adjust the system operating parameters, specifically the pH value and / or SO3 concentration. 2- concentration;

[0021] The central control system sends an instruction to the raw material replacement module to replace abnormal substances based on the triggered Level II warning signal. This is used to replace the limestone slurry, coal, or water raw materials involved in the operation of the desulfurization tower.

[0022] The central control system sends an instruction to the sewage control module to replace the blinding desulfurization slurry based on the triggered Level III warning signal.

[0023] Furthermore, the data processing and early warning unit further employs an improved K-means clustering algorithm to perform mutation detection and threshold fine-tuning on the historical RI time series generated by the multi-factor risk index model under the historical operating data of the desulfurization tower: First, the historical RI time series data are grouped, and the improved K-means algorithm is used to perform cluster analysis on each group of data to identify mutation points in the data, that is, periods when the RI value changes drastically; when a mutation is detected, the improved K-means algorithm dynamically adjusts the weighting coefficients and thresholds of each factor according to the trend of the historical RI time series data.

[0024] Furthermore, the improved K-means clustering algorithm of this invention calculates the mean of the data points belonging to each cluster based on historical data of desulfurization slurry properties within the normal range, using this mean as its initial centroid; performs multi-factor feature extraction on all historical data of desulfurization slurry properties; dynamically determines the optimal number of clusters using the silhouette coefficient method; then, calculates the distance from each data point to each cluster center using weighted Euclidean distance, assigns the data points to the nearest cluster center, recalculates the cluster centers of each cluster, constructs an objective function with the goal of minimizing the sum of squared errors, and iterates repeatedly until the objective function converges.

[0025] Furthermore, the central control system uses a PID controller to control the execution module, and the central control system uses the online learning algorithm Recursive Least Squares (RLS) to dynamically update the weighting coefficients of each factor and the PID parameters of the PID controller, thereby achieving adaptive optimization of system performance.

[0026] The steps for applying the RLS algorithm to PID parameter tuning are as follows:

[0027] System modeling: Establish a linear mathematical model for the system, representing the proportional relationship between the system's input and output values;

[0028] Parameter initialization: Initialize the PID controller parameters Kp, Ki, Kd and covariance matrix P;

[0029] Data acquisition: Real-time acquisition of data from the online monitoring unit, including set values ​​and feedback values ​​for each factor;

[0030] Error calculation: The error of the system is calculated as e(t) = r(t) - y(t), where r(t) is the set value of each factor at time t, i.e. the threshold of each factor; y(t) is the feedback value of each factor at time t.

[0031] The setpoint r(t) is the target value that the system hopes to maintain for each monitoring factor. In this invention, r(t) corresponds to the "threshold" of each factor in the multi-factor risk model. The feedback value y(t) is the actual operating condition value measured in real time by the online monitoring unit, such as the real-time collected F... - Concentration, Al 3+ Concentration, Fe 2+ / Fe 3+ Total concentration, SO3 2- Concentration and pH value, etc.

[0032] Gain calculation: Calculate the gain vector g(t) according to the RLS algorithm:

[0033] ;

[0034] Where x(t) is the input vector at time t, and the input vector includes a set of physical property parameters of the desulfurization slurry and PID parameters, wherein the physical property parameters include F - Concentration, Al 3+ Concentration, Fe 2+ / Fe 3+ Total concentration, SO3 2- Concentration and pH value; λ is the forgetting factor, and P(t-1) is the covariance matrix of the previous time step (i.e., time step t-1);

[0035] Parameter update: Update the parameters of the PID controller:

[0036] θ(t) = θ(t-1) + g(t) * e(t)

[0037] Where θ(t) is the vector of parameters Kp, Ki, and Kd of the PID controller at time t, and e(t) is the error of the system at time t;

[0038] Covariance Update: Update the covariance matrix:

[0039] ;

[0040] Through the above steps, the RLS algorithm can adjust the parameters of the PID controller in real time to adapt to the dynamic changes of the system, thereby optimizing the control performance.

[0041] The operation method of the desulfurization slurry blindness early warning control system of the present invention includes the following steps:

[0042] Step 1: Use an online monitoring unit to collect real-time physical property indicators of the desulfurization slurry, including the F content of the desulfurization slurry. - Concentration, Al 3+ Concentration, Fe 2+ / Fe 3+ Total concentration, SO3 2-Concentration and pH value;

[0043] Step 2: The data processing and early warning module uses a multi-factor risk index model to process the data collected in Step 1, generate graded early warning signals, and determine the risk level as Level I, Level II, or Level III according to the severity of the risk.

[0044] Step 3: The central control system receives early warning signals from the data processing and early warning module, and sends different control commands to the execution module according to different early warning signals. When the risk level of the early warning is Level I, the limestone slurry addition module and / or oxidation air volume adjustment module are used to adjust the limestone slurry addition and / or oxidation air volume of the desulfurization tower. When the risk level of the early warning is Level II, the limestone slurry, coal or water raw materials involved in the operation of the desulfurization tower are replaced through the raw material replacement module. When the risk level of the early warning is Level III, the desulfurization slurry in the desulfurization tower is discharged through the sewage control module.

[0045] Compared with the prior art, the present invention has achieved the following beneficial effects:

[0046] 1) Real-time monitoring and early warning: Real-time early warning of the risk of blindness caused by slurry is achieved by monitoring key parameters online and combining them with a multi-factor risk index model.

[0047] 2) Automated control: The system automatically executes control commands, reducing human intervention and improving the stability and safety of the desulfurization process.

[0048] 3) Adaptive optimization: The system adopts an online learning algorithm to dynamically adjust system parameters to adapt to different working conditions and improve the system's adaptability and efficiency.

[0049] 4) Energy conservation and emission reduction: By optimizing control strategies, unnecessary energy consumption and chemical reagent addition are reduced, thus lowering operating costs. Attached Figure Description

[0050] Figure 1 This is a schematic diagram of the operation process of the desulfurization slurry blindness early warning control system of the present invention. Detailed Implementation

[0051] This invention provides a desulfurization slurry blindness early warning and control system, comprising:

[0052] The online monitoring unit is used to collect the physical properties of the desulfurization slurry in real time, including F. - Concentration, Al 3+ Concentration, Fe 2+ / Fe 3+ Total concentration, SO3 2- Concentration and pH value;

[0053] The online monitoring unit transmits the collected physical property data to the data processing and early warning module for processing;

[0054] The data processing and early warning module adopts a multi-factor risk index model to generate graded early warning signals based on the physical property index data collected through monitoring.

[0055] The central control system is used to coordinate the closed-loop automatic operation of each unit and achieve adaptive optimization. It sends different control commands to the execution module according to different early warning signals issued by the data processing and early warning module.

[0056] The execution module includes a raw material replacement module, a sewage control module, a limestone slurry dosing module, and an oxidation air volume adjustment module. Each module of the execution module is used to execute control commands issued by the central control system.

[0057] The multi-factor risk index model comprehensively considers the monitored and collected physical property data and the weighting coefficients of each physical property to calculate the risk index. RI When RI>0.7, RI>0.85, and RI>1.0, the data processing and early warning module triggers Level I, II, and III early warnings, respectively.

[0058] When the data processing and early warning module triggers different levels of early warnings, the central control system will send different control commands according to the early warning level. The specific control measures for each level are as follows:

[0059] Level I Risk (RI>0.7): This level of warning indicates a slight deviation in the system, requiring initial adjustments. At this stage, the system primarily adjusts SO3 levels. 2- Key physical properties such as concentration and pH value are monitored. Limestone slurry is added to adjust the pH value, maintaining the slurry within a suitable acid-base range, while adjusting the oxidation air volume helps promote the oxidation of sulfite ions, ensuring their conversion to gypsum, thereby indirectly controlling SO3. 2- concentration.

[0060] Level II Risk (RI>0.85): This level indicates increased risk, requiring more stringent system adjustments. At this point, the control system will issue instructions to replace abnormal substances, such as limestone slurry, coal quality, and water quality. These substances may affect the normal operation of the desulfurization reaction; replacing abnormal substances indirectly controls F... - Concentration, Al 3+ Concentration, Fe 2+ / Fe 3+ Total concentration, SO3 2- Concentration and pH value. By changing these raw materials, the stability of the desulfurization process can be ensured and the risk of system failure can be reduced.

[0061] Level III Risk (RI>1.0): This level indicates a high system risk, requiring the most stringent emergency measures. In this case, the control system will instruct the replacement of the slurry, i.e., a blowdown operation. By discharging the problematic slurry, the system can clean accumulated impurities and sediment, restoring normal operation and preventing equipment damage or "blinding" phenomena.

[0062] Example 1: A power plant 1 applied the desulfurization slurry blindness early warning control system of the present invention (operation process as follows) Figure 1 As shown), the system has collected online data for one year, removing abnormal data pairs. Take 210 mg / L, Take 890 mg / L, Take 650 mg / L, Take 99860 mg / L, p H opt Take 5.4. Obtain F through an online learning algorithm. - Concentration, Al 3+ Concentration, Fe 2+ / Fe 3+ Total concentration, SO3 2- The weighting coefficients for the five factors—concentration, pH value, etc.—are a1=0.261, a2=0.294, a3=0.132, a4=0.221, and a5=0.092, respectively. The desulfurization slurry blindness early warning control system of this invention is used to automatically regulate the operation of the desulfurization tower. The operational performance of the desulfurization tower within one year is shown in Table 1 under "After Modification".

[0063] The results in Table 1, "Before Modification," correspond to the operating performance of the desulfurization tower in a power plant 1 before the implementation of the desulfurization slurry blindness early warning control system of this invention.

[0064] Table 1

[0065] parameter Before renovation After renovation Annual average desulfurization efficiency (%) 95% 98% Average annual equipment failure rate (%) 1.5% 1% Human intervention (number of times / year) 5 times 2 times

[0066] Example 2: A power plant applied a desulfurization slurry blindness early warning and control system. The system collected online data for two years, eliminating abnormal data pairs. Take 190 mg / L, Take 1100 mg / L, Take 550 mg / L, Take 99970 mg / L, p H opt Take 5.3. Obtain F through an online learning algorithm. - Concentration, Al 3+ Concentration, Fe 2+ / Fe 3+ Total concentration, SO3 2-The weighting coefficients for the five factors—concentration, pH value, etc.—are a1=0.293, a2=0.287, a3=0.130, a4=0.180, and a5=0.110, respectively. The desulfurization slurry blindness early warning control system of this invention is used to automatically regulate the operation of the desulfurization tower. The operational performance of the desulfurization tower within one year is shown in Table 2 under the "After Modification" section.

[0067] The results in Table 2, "Before Modification," correspond to the operating performance of the desulfurization tower in a certain power plant 2 before the implementation of the desulfurization slurry blindness early warning control system of this invention.

[0068] Table 2

[0069] parameter Before renovation After renovation Annual average desulfurization efficiency (%) 96% 98% Average annual equipment failure rate (%) 1.5% 1.2% Human intervention (number of times / year) 4 times 2 times

Claims

1. A desulfurization slurry blindness early warning control system, characterized in that, include: The online monitoring unit is used to collect the physical properties of the desulfurization slurry in real time. Including the following factors: F - Concentration, Al 3+ Concentration, Fe 2+ / Fe 3+ Total concentration, SO3 2- Concentration and pH value; The online monitoring unit transmits the collected physical property data to the data processing and early warning module for processing; The data processing and early warning module adopts a multi-factor risk index model to generate graded early warning signals based on the physical property index data collected through monitoring. The central control system is used to coordinate the closed-loop automatic operation of each unit and achieve adaptive optimization. It sends different control commands to the execution module according to different early warning signals issued by the data processing and early warning module. The execution module includes a raw material replacement module, a sewage control module, a limestone slurry dosing module, and an oxidation air volume adjustment module. Each module of the execution module is used to execute control commands issued by the central control system. The multi-factor risk index model comprehensively considers the monitored and collected physical property data and the weighting coefficients of each physical property to calculate the risk index RI, which is calculated according to the following formula. C Felim , Representing F respectively - Concentration, Al 3+ Concentration, Fe 2+ / Fe 3+ Total concentration and SO3 2- The threshold values ​​for the four factors of concentration, a1, a2, a3, a4, and a5, are respectively F - Concentration, Al 3+ Concentration, Fe 2+ / Fe 3+ Total concentration, SO3 2- Weighting coefficients for the five factors: concentration, pH value, etc. C Fe , These represent the real-time collected F values. - Concentration, Al 3+ Concentration, Fe 2+ / Fe 3+ Total concentration and SO3 2- Concentration; pH represents the real-time pH value of the desulfurization slurry. opt ΔpH represents the average pH value of the desulfurization slurry during normal operation, while ΔpH represents the maximum pH value of the desulfurization slurry during normal operation. The thresholds for the above factors are set based on the historical normal operation data of the desulfurization tower, that is, based on the historical data of the physical properties of the desulfurization slurry being within the normal range during normal operation of the desulfurization tower. The thresholds set for each factor are the average values ​​of their respective ranges. The data processing and early warning module triggers different levels of early warning signals based on the magnitude of the risk index RI value generated by the multi-factor risk index model. The data processing and early warning unit further employs an improved K-means clustering algorithm to perform mutation detection and threshold fine-tuning on the historical RI time series generated by the multi-factor risk index model under the historical operating data of the desulfurization tower: First, the historical RI time series data are grouped, and the improved K-means algorithm is used to perform cluster analysis on each group of data to identify mutation points in the data, that is, periods when the RI value changes drastically; when a mutation is detected, the improved K-means algorithm dynamically adjusts the weighting coefficients and thresholds of each factor according to the trend of the historical RI time series data; The central control system uses the online learning algorithm, Recursive Least Squares (RLS), to dynamically update the weighting coefficients of each factor.

2. The desulfurization slurry blindness early warning control system as described in claim 1, characterized in that, The online monitoring unit uses an ion-selective electrode to detect F in the desulfurization slurry. - Concentration and SO3 2- The concentration of Al in the desulfurization slurry was detected using atomic absorption spectrometry or an electrochemical sensor. 3+ Concentration, Fe 2+ / Fe 3+ The total concentration was determined by measuring the pH value of the desulfurization slurry using a pH meter.

3. The desulfurization slurry blindness early warning control system as described in claim 1, characterized in that, When RI>0.7, RI>0.85, and RI>1.0, the data processing and early warning module triggers Level I, II, and III early warnings, respectively. Based on the triggered Level I warning signal, the central control system sends control commands to the limestone slurry dosing module and / or oxidation air volume regulating module to adjust the system operating parameters, specifically the pH value and / or SO3 concentration. 2- concentration; The central control system sends an instruction to the raw material replacement module to replace abnormal substances based on the triggered Level II warning signal. This is used to replace the limestone slurry, coal, or water raw materials involved in the operation of the desulfurization tower. The central control system sends an instruction to the sewage control module to replace the blinding desulfurization slurry based on the triggered Level III warning signal.

4. The desulfurization slurry blindness early warning control system as described in claim 1, characterized in that, The improved K-means clustering algorithm calculates the mean of the data points belonging to each cluster based on historical data showing that the physical properties of the desulfurization slurry are within the normal range, and uses this as its initial centroid. Multi-factor feature extraction was performed on all historical data of the physical properties of the desulfurization slurry; The optimal number of clusters is dynamically determined using the silhouette coefficient method. Then, the distance from each data point to each cluster center is calculated using weighted Euclidean distance. The data points are assigned to the nearest cluster center, and the cluster centers of each cluster are recalculated. An objective function is constructed with the goal of minimizing the sum of squared errors. The iteration is repeated until the objective function converges.

5. The desulfurization slurry blindness early warning control system as described in claim 1, characterized in that, The central control system uses a PID controller to control the execution module. The central control system also receives data collected from the online monitoring unit and uses the online learning algorithm, recursive least squares (RLS), to dynamically update the PID parameters of the PID controller, thereby achieving adaptive optimization of system performance. The steps for applying the RLS algorithm to PID parameter tuning are as follows: System modeling: Establish a linear mathematical model for the system, representing the proportional relationship between the system's input and output values; Parameter initialization: Initialize the PID controller parameters Kp, Ki, Kd and covariance matrix P; Data acquisition: Real-time acquisition of data from the online monitoring unit, including set values ​​and feedback values ​​for each factor; Error calculation: Calculate the system error e(t) = r(t) - y(t), where r(t) is the set value of each factor, i.e. the threshold; y(t) is the feedback value of each factor; Gain calculation: Calculate the gain vector g(t) according to the RLS algorithm: g(t)=P(t-1)*x(t) / [λ+x(t) T *P(t-1)*x(t)] Where x(t) is the input vector, which includes a set of physical property parameters of the desulfurization slurry and PID parameters. The physical property parameters include F... - Concentration, Al 3+ Concentration, Fe 2+ / Fe 3+ Total concentration, SO3 2- Concentration and pH value; λ is the forgetting factor, and P(t-1) is the covariance matrix of the previous time step; Parameter update: Update the parameters of the PID controller: θ(t) = θ(t-1) + g(t) * e(t) Where θ(t) is the vector of parameters Kp, Ki, and Kd of the PID controller, and e(t) is the system error; Covariance Update: Update the covariance matrix: P(t)=(1 / λ)*[P(t-1)-g(t)*x(t) T *P(t-1)] Through the above steps, the RLS algorithm can adjust the parameters of the PID controller in real time to adapt to the dynamic changes of the system, thereby optimizing the control performance.

6. The method of operating the system as described in any one of claims 1-5, characterized in that, Includes the following steps: Step 1: Use an online monitoring unit to collect real-time physical property indicators of the desulfurization slurry, including the F content of the desulfurization slurry. - Concentration, Al 3+ Concentration, Fe 2+ / Fe 3+ Total concentration, SO3 2- Concentration and pH value; Step 2: The data processing and early warning module uses a multi-factor risk index model to process the data collected in Step 1, generate graded early warning signals, and determine the risk level as Level I, Level II, or Level III according to the severity of the risk. Step 3: The central control system receives early warning signals from the data processing and early warning module, and sends different control commands to the execution module according to different early warning signals. When the risk level of the early warning is Level I, the limestone slurry addition module and / or oxidation air volume adjustment module are used to adjust the limestone slurry addition and / or oxidation air volume of the desulfurization tower. When the risk level of the early warning is Level II, the limestone slurry, coal or water raw materials involved in the operation of the desulfurization tower are replaced through the raw material replacement module. When the risk level of the early warning is Level III, the desulfurization slurry in the desulfurization tower is discharged through the sewage control module.

Images