A control method of a furnace pressure purging system based on DCS control

Abstract

The application discloses a kind of control method of furnace pressure purging system based on DCS control, comprising the following steps: S1, a group of measuring point sampling tubes are arranged on multiple furnace pressure measuring points on furnace respectively, and the furnace pressure prediction module and the sampling electromagnetic valve, furnace transmitter, purging electromagnetic valve, purging transmitter and pressure protection switch connected on each measuring point sampling tube are accessed intelligent purging program control in thermal power distributed control module;S2, furnace pressure soft measurement model is constructed by the furnace pressure prediction module;S3, in initial state, purging electromagnetic valve is in closed state, measurement electromagnetic valve is in open state, furnace pressure soft measurement model generates current furnace pressure prediction value, and the actual value of current furnace pressure is measured by the furnace transmitter;S4, the thermal power distributed control module judges the state of measuring point sampling tube according to the furnace pressure prediction value and furnace pressure actual value obtained in current S3.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of automation, in particular to a control method of a furnace pressure purging system based on DCS control. BACKGROUND

[0002] The thermal furnace pressure measuring point of a thermal power unit plays a crucial role in the operation of the furnace negative pressure regulation system. The furnace pressure high and low measuring point protection, as the main protection of the boiler, is clearly stipulated in the twenty-five key requirements for preventing major accidents in power production: the boiler furnace pressure protection device must not be withdrawn during unit operation, and the safety of the boiler cannot be effectively guaranteed during the protection withdrawal period. Therefore, the blockage of this measuring point will cause the protection to fail, which will bring disastrous consequences to the boiler.

[0003] However, the sampling device of this measuring point often fails to operate due to ash blockage during actual operation. In order to deal with the blockage, the boiler furnace pressure high and low protection needs to be withdrawn to blow the sampling pipe, which is not only dangerous but also cannot completely eliminate the defect of frequent ash blockage of the boiler furnace negative pressure sampling pipe. In most cases, the blockage is checked and cleared during the unit shutdown maintenance process, which is often more serious at this time, increasing the difficulty of handling and the maintenance workload.

[0004] Currently, there have been some researches on improving the reliability and preventing blockage of this measuring point: for example, in the improvement of the sampling device, a compensation type wind pressure measurement anti-blocking and blowing device is improved by using the dynamic pressure compensation method of fluid mechanics; a PLC control device is installed on site for back blowing, and a manual program control back blowing device is added to the DCS system. However, most of these devices use periodic blowing or manual blowing, which cannot timely predict abnormal negative pressure sampling and handle it in time. SUMMARY

[0005] The purpose of the present application is to provide a control method of a furnace pressure purging system based on DCS control to solve the above problems.

[0006] To achieve the above purpose, the present application adopts the following technical solutions:

[0007] A furnace pressure purging system based on DCS control includes a thermal distributed control module, a sampling tube, a sampling circuit, a sampling solenoid valve, a furnace transmitter, a furnace pressure prediction module, a purging circuit, a purging solenoid valve, and a purging transmitter. One end of the sampling tube is inserted into a furnace pressure measuring point, and the other end is connected to the sampling circuit and the purging circuit via a first tee connector. A pressure protection switch is installed at the other end of the sampling circuit, and the other end of the purging circuit is connected to a purging instrument. The sampling solenoid valve and the purging solenoid valve are respectively installed in the furnace pressure measurement module. The sampling circuit and purging circuit are used to control circuit switches. The furnace transmitter is connected to the sampling circuit via a second tee connector and is used to measure the actual value of the furnace pressure. The purging transmitter is connected to the purging circuit via a third tee connector and is used to measure the actual value of the sampling tube pressure. The furnace pressure prediction module is used to generate the predicted value of the furnace pressure. The sampling solenoid valve, furnace transmitter, furnace pressure prediction module, purging solenoid valve, purging transmitter, pressure protection switch and purging instrument are all communicatively connected to the thermal distributed control module.

[0008] Preferably, both the sampling solenoid valve and the purging solenoid valve comprise two sets.

[0009] Preferably, it also includes a furnace pressure monitoring and purging display module, which is used to display the furnace pressure and purging status.

[0010] A control method for a furnace pressure purging system based on DCS control includes the following steps:

[0011] S1. Set up a set of sampling tubes at multiple furnace pressure measuring points on the furnace, and connect the furnace pressure prediction module and the sampling solenoid valve, furnace transmitter, purging solenoid valve, purging transmitter and pressure protection switch connected to each of the sampling tubes to the thermal distributed control module for intelligent purging program control.

[0012] S2. Construct a soft measurement model of furnace pressure using the furnace pressure prediction module;

[0013] S3. In the initial state, the purge solenoid valve is closed and the measurement solenoid valve is open. The furnace pressure soft measurement model generates the current furnace pressure prediction value and measures the current furnace pressure actual value through the furnace transmitter.

[0014] S4. The thermal distribution control module judges the status of the sampling tube at the measuring point based on the predicted value of the furnace pressure and the actual value of the furnace pressure obtained in S3. If the status of the sampling tube at the measuring point is abnormal, it outputs a fault alarm, judges the fault type, and triggers a purging command to enter step S5.

[0015] S5. The thermal distribution control module performs position holding processing on the furnace transmitter and pressure protection switch corresponding to the sampling tube of the fault alarm point, controls the measurement solenoid valve to close, and purges the solenoid valve to open. The thermal distribution control module sets the purging time according to the abnormal condition of the sampling tube of the measuring point, and the purging instrument purges the sampling tube of the measuring point through the purging circuit.

[0016] S6. During the purging process, the purging transmitter measures the actual pressure value of the sampling tube at the measuring point in real time, and determines whether the sampling tube is purged based on the actual pressure value. If the tube is purged ahead of schedule or the purging time ends, the purging solenoid valve is closed. Once the purging transmitter detects that there is no residual purging gas in the sampling tube at the measuring point, the sampling solenoid valve is opened, and the furnace transmitter resumes normal measurement of the actual furnace pressure value.

[0017] S7. The furnace pressure soft measurement model generates the current furnace pressure prediction value and measures the current furnace pressure actual value through the furnace transmitter. The thermal distributed control module compares the currently acquired furnace pressure prediction value with the actual furnace pressure value. If the two are close, the signal position protection processing of the furnace transmitter and pressure protection switch is released, and the adjustment and protection function of the furnace pressure related logic circuit is opened.

[0018] Preferably, the specific steps for constructing the furnace pressure soft measurement model by the furnace pressure prediction module in step S2 are as follows:

[0019] S21. Obtain the operating parameters of the furnace thermal power unit;

[0020] S22. Analyze and judge the operating parameters of the furnace power unit, and select the furnace pressure as a candidate relevant parameter;

[0021] S23. Perform correlation analysis on the candidate relevant parameters of furnace pressure to obtain the contribution rate of each candidate relevant parameter, and select the relevant parameters of furnace pressure according to the contribution rate.

[0022] S24. The furnace pressure-related parameters are normalized using the Mapminmax function, and the operating parameters of the furnace power unit for one day are selected as the original data for model training and learning.

[0023] S25. Sample the original data of the model training and learning, set the sampling interval to 20s, and obtain 4320 sets of effective data of furnace pressure related parameters.

[0024] S26. Select 300 sets of valid data related to furnace pressure parameters as training samples for the furnace pressure soft measurement model, and optimize the training samples.

[0025] S27. Select the radial basis function as the model kernel function, train the model using optimized training samples, and construct a soft measurement model for furnace pressure.

[0026] Preferably, the furnace pressure-related parameters include primary air fan current, primary air volume, secondary air volume, forced draft fan current, furnace flue gas temperature, induced draft fan current, and induced draft fan blade opening.

[0027] Preferably, the correlation analysis calculation formula in step S23 is:

[0028]

[0029] Where R is the contribution rate, x is the measured value of the candidate relevant parameter, y is the value of the target parameter corresponding to the candidate relevant parameter, Cov(x, y) is the covariance of the two, and D(x) and D(y) are the variances of the two.

[0030] Preferably, in step S26, the training samples are used to calculate the similarity between two sets of data using a similarity calculation function. For data with high similarity, only one set of data is retained and the other set of data is deleted, thereby reducing redundancy between training sample data and thus optimizing the data.

[0031] Preferably, the similarity calculation function is:

[0032] R ij =exp(-||x i -x j || 2 / δ)

[0033] Among them, X i For the i-th sample data, X j Let R represent the j-th sample data. ij This represents the similarity between two sets of data, where δ is the normalization parameter.

[0034] If R ij If R < ε, then retain both sets of data; if R < ε, then retain both sets of data. ij If the similarity threshold is greater than ε, then two sets of data are retained, where ε is the threshold of the similarity function.

[0035] By adopting the above technical solution, the present invention has the following beneficial effects compared with the prior art:

[0036] This invention provides a control method for a furnace pressure purging system based on DCS control. By constructing a soft-sensor model of furnace pressure, the method predicts the furnace pressure at measuring points and compares the predicted and actual furnace pressure values ​​in real time. Once severe blockage is detected in the sampling tube or anti-blockage device, an automatic purging program is triggered to isolate the measuring point; simultaneously, the measuring point signal is preserved. The purging solenoid valve is opened, and the purging instrument performs remote automatic purging with air. The purging duration is automatically adjusted according to the severity of the blockage, and the purging transmitter determines whether the blockage has been cleared. During equipment operation, the system can also be configured via a distributed thermal control module for periodic remote purging, alarm purging, and local manual purging, integrating multiple functions to effectively protect the boiler.

[0037] Drawings

[0038] Figure 1 This is a block diagram of the purging system of the present invention.

[0039] Figure 2 This is a flowchart of the purging system control method of the present invention;

[0040] Figure 3 This is a schematic diagram showing the state of the furnace pressure monitoring and purging display module of the present invention;

[0041] Figure 4 This is a logic block diagram for selecting the position of the furnace pressure measurement point signal in this invention. Detailed Implementation

[0042] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0043] Example

[0044] Cooperate Figures 1 to 4As shown, this invention discloses a furnace pressure purging system based on DCS control, including a thermal distributed control module, a sampling tube 1, a sampling circuit 2, a sampling solenoid valve 3, a furnace transmitter 4, a furnace pressure prediction module, a purging circuit 5, a purging solenoid valve 6, and a purging transmitter 7. One end of the sampling tube 1 is inserted into the furnace pressure measuring point, and the other end is connected to the sampling circuit 2 and the purging circuit 5 respectively through a first tee connector. The other end of the sampling circuit 2 is connected to a pressure protection switch 8, and the other end of the purging circuit 5 is connected to a purging instrument. The sampling solenoid valve 3 and... The purge solenoid valve 6 is installed on the sampling circuit 2 and the purging circuit respectively to control the circuit switch. The furnace transmitter 4 is connected to the sampling circuit 2 through the second tee connector and is used to measure the actual value of the furnace pressure. The purge transmitter 7 is connected to the purging circuit 5 through the third tee connector and is used to measure the actual value of the sampling tube pressure. The furnace pressure prediction module is used to generate the furnace pressure prediction value. The sampling solenoid valve 3, the furnace transmitter 4, the furnace pressure prediction module, the purge solenoid valve 6, the purge transmitter 7, the pressure protection switch and the purging instrument are all connected to the thermal distributed control module.

[0045] Both the sampling solenoid valve 3 and the purging solenoid valve 6 consist of two sets.

[0046] It also includes a furnace pressure monitoring and purging display module, which displays the furnace pressure and purging status. The module also allows manual switching to a manual purging mode, with the thermal distribution control module controlling the manual purging of the sampling tubes at the measuring points.

[0047] This invention also discloses a control method for a furnace pressure purging system based on DCS control, comprising the following steps:

[0048] S1. Set up a set of sampling tubes at multiple furnace pressure measuring points on the furnace, and connect the furnace pressure prediction module and the sampling solenoid valve, furnace transmitter, purging solenoid valve, purging transmitter and pressure protection switch connected to each sampling tube to the thermal distributed control module for intelligent purging program control.

[0049] S2. Construct a soft measurement model of furnace pressure using the furnace pressure prediction module;

[0050] S3. In the initial state, the purge solenoid valve is closed and the measurement solenoid valve is open. The furnace pressure soft measurement model generates the current furnace pressure prediction value and measures the current furnace pressure actual value through the furnace transmitter.

[0051] S4. The thermal distribution control module judges the status of the sampling tube at the measuring point based on the predicted value of the furnace pressure and the actual value of the furnace pressure obtained in S3. If there is an abnormality in the status of the sampling tube at the measuring point, a fault alarm is output, the fault type is judged, and a purging command is triggered to enter step S5.

[0052] S5. The thermal distribution control module performs position holding processing on the furnace transmitter and pressure protection switch corresponding to the sampling tube of the fault alarm point, controls the measurement solenoid valve to close, and purges the solenoid valve to open. The thermal distribution control module sets the purging time according to the abnormal condition of the sampling tube of the measuring point, and the purging instrument purges the sampling tube of the measuring point through the purging circuit.

[0053] S6. During the purging process, the purging transmitter measures the actual pressure value of the sampling tube at the measuring point in real time and determines whether the sampling tube is purged based on the actual pressure value. If the tube is purged in advance or the purging time ends, the purging solenoid valve is closed. Once the purging transmitter detects that there is no residual purging gas in the sampling tube at the measuring point, the sampling solenoid valve is opened, and the furnace transmitter resumes normal measurement of the actual furnace pressure value.

[0054] S7. The furnace pressure soft measurement model generates the current furnace pressure prediction value and measures the current furnace pressure actual value through the furnace transmitter. The thermal distributed control module compares the currently acquired furnace pressure prediction value with the actual furnace pressure value. If the two are close, the signal position protection processing of the furnace transmitter and pressure protection switch is released, and the adjustment and protection function of the furnace pressure related logic circuit is opened.

[0055] Both furnace transmitter 4 and pressure protection switch 8 are furnace pressure measuring points. Furnace transmitter 4 is an analog quantity, and pressure protection switch 8 is a digital quantity.

[0056] In this implementation, the signal position retention processing of the furnace transmitter and pressure protection switch, i.e., the furnace pressure measuring point position retention logic, is explained as follows: Figure 4 As shown: The three furnace pressure measuring points in the diagram are judged by a three-selection function block, which outputs the selected pressure value. The three-selection function block provides four signal selection modes: 1. Select high value; 2. Select low value; 3. Select median value; 4. Select average value. It also judges and alarms the deviation and quality of the three input signals.

[0057] Under normal measurement conditions, the signal during purging at the measuring point is Boolean 0. At this time, the signal switching selection function block outputs a Y value that tracks the input value X2. When the intelligent purging detection system detects and determines that the measuring point is blocked and requires purging, it triggers the purging logic and simultaneously flips the purging signal at the measuring point to Boolean 1. At this point, the signal switching selection function block outputs a Y value that tracks the input value X1. Because X1 always tracks the output value Y, the Y output value remains unchanged from its current value at the time of purging switching. When the purging of the measuring point is completed and the purging signal returns to Boolean 0, the switching block output value Y returns to the same level as the measured value channel X2.

[0058] The specific steps for constructing the furnace pressure soft measurement model in step S2 of the furnace pressure prediction module are as follows:

[0059] S21. Obtain the operating parameters of the furnace thermal power unit;

[0060] S22. Based on a comprehensive consideration of the formation mechanism and correlation analysis of furnace pressure, the operating parameters of the furnace thermal power unit are analyzed and judged, and alternative relevant parameters of furnace pressure are selected.

[0061] From the perspective of the formation mechanism of furnace pressure, the calculation of furnace pressure is very close to the following formula: PV=MRT, where P is the absolute pressure, V is the furnace volume, M is the mass of flue gas in the furnace, R is the flue gas constant, and T is the absolute temperature.

[0062] Since the volume is fixed and R is a constant, the pressure P is proportional to MT. Therefore, the flue gas temperature in the furnace is a relevant parameter that affects the pressure.

[0063] Considering the correlation with the equipment structure, 10 alternative relevant parameters can be obtained, namely, primary fan blade opening, primary fan current, primary fan outlet pressure, primary air volume, secondary air volume, forced draft fan current, forced draft fan blade opening, furnace air box differential pressure, induced draft fan current, and induced draft fan blade opening.

[0064] S23. Perform correlation analysis on the candidate relevant parameters of furnace pressure, obtain the contribution rate of each candidate relevant parameter, and select the relevant parameters of furnace pressure based on the contribution rate.

[0065] The correlation analysis calculation formula in step S23 is as follows:

[0066]

[0067] Where R is the contribution rate, x is the measured value of the candidate relevant parameter, y is the value of the target parameter corresponding to the candidate relevant parameter, Cov(x, y) is the covariance of the two, and D(x) and D(y) are the variances of the two.

[0068] By conducting correlation analysis on 10 candidate parameters, the primary air volume and secondary air volume, which are more representative and have a more direct impact on furnace pressure, were selected as the relevant parameters based on their respective contribution rates.

[0069] The table below shows the values ​​of multiple furnace pressure-related parameters and the furnace pressure values ​​at different time periods. The induced draft fan A current and induced draft fan B current represent the currents of the induced draft fans on both sides of the furnace, the primary air fan A current and primary air fan B current represent the currents of the primary air fans on both sides of the furnace, and the forced draft fan A current and forced draft fan B current represent the currents of the forced draft fans on both sides of the furnace.

[0070] Time Furnace pressure Induced draft fan A current Induced draft fan B current Primary air fan A current Primary air fan B current Primary air fan A current Primary air fan B current Total air volume Secondary air volume Primary air volume 1 -0.044 102 103 63 65 31 30 222.5 132 90.5 2 -0.031 108 109 63 67 33 32 229.4 136 93.4 3 -0.039 106 108 61 65 33 32 228.5 136 92.5 4 -0.036 107 109 62 66 32 32 231.1 136 95.1 5 -0.032 107 109 62 65 33 32 225.2 136 89.2 6 -0.035 107 109 64 65 32 31 231.8 136 95.8 7 -0.03 / 106 107 62 64 33 32 231.8 135 96.8 8 -0.035 118 118 59 63 34 33 253.3 158 95.3 9 -0.058 123 121 61 64 34 34 255.8 163 92.8 10 -0.025 128 123 60 65 33 33 259.3 163 96.3 11 -0.048 125 123 59 62 34 34 255.2 163 92.2 12 -0.032 108 110 59 63 32 31 229 140 89 13 -0.038 122 122 59 64 33 31 247.8 156 91.8 14 -0.015 105 105 59 62 32 31 223.8 133 90.8 15 -0.029 108 111 59 63 32 31 227.7 138 89.7 16 -0.059 96 98 60 63 31 30 210.8 122 88.8 17 -0.031 96 96 60 63 30 29 206.8 121 85.8 18 -0.017 96 97 59 61 31 30 207 123 84 19 -0.035 95 97 59 59 31 30 209.1 121 88.1 20 -0.033 96 99 63 67 31 30 218.2 121 97.2 21 -0.024 94 97 62 62 32 30 220.2 121 99.2 22 -0.011 125 125 66 67 33 32 255.8 156 99.8 23 -0.019 99 100 67 70 31 30 225.5 120 105.5 24 -0.053 86 85 61 63 30 30 189.2 98 91.2 25 -0.046 81 80 62 63 30 29 172.1 85 87.1 26 -0.043 82 84 60 64 30 28 177.8 92 85.8 27 -0.1 83 85 63 65 30 28 180.6 92 88.6 28 -0.032 80 81 63 65 30 29 173.7 86 87.7 29 -0.011 81 83 63 66 30 29 173.9 86 87.9 30 -0.048 83 85 62 65 30 28 174.6 90 84.6

[0071] By analyzing the formation mechanism of the furnace pressure system and combining the measurement principle of the relevant parameters in the table above with the correlation coefficient method, seven furnace pressure-related parameters were identified, including primary air fan current, primary air volume, secondary air volume, forced draft fan current, furnace flue gas temperature, induced draft fan current, and induced draft fan blade opening. These seven furnace pressure-related parameters work together to affect the furnace pressure value.

[0072] S24. The furnace pressure-related parameters are normalized using the Mapminmax function, and the operating parameters of the furnace power unit for one day are selected as the original data for model training and learning.

[0073] S25. Sample the original data for model training and learning, set the sampling interval to 20s, and obtain 4320 sets of valid data related to furnace pressure parameters.

[0074] S26. Select 300 sets of valid data related to furnace pressure parameters as training samples for the furnace pressure soft measurement model, and optimize the training samples.

[0075] S27. Select the radial basis function as the model kernel function, train the model using optimized training samples, and construct a soft measurement model for furnace pressure.

[0076] In step S26, the similarity between two sets of data is calculated using a similarity calculation function. For data with high similarity, only one set of data is kept and the other set of data is deleted, thereby reducing redundancy between training sample data and thus optimizing the data.

[0077] The similarity calculation function is:

[0078] R ij =exp(-||x i -x j || 2 / δ)

[0079] Among them, X i For the i-th sample data, X j Let R represent the j-th sample data. ij This represents the similarity between two sets of data, where δ is the normalization parameter.

[0080] If R ij If R < ε, then retain both sets of data; if R < ε, then retain both sets of data. ij If the similarity threshold is greater than ε, then two sets of data are retained, where ε is the similarity function threshold. In this embodiment, the similarity function threshold is 0.97.

[0081] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A control method for a furnace pressure purging system based on DCS control, the method being implemented based on a thermal distributed control module, a sampling tube at a measuring point, a sampling circuit, a sampling solenoid valve, a furnace transmitter, a furnace pressure prediction module, a purging circuit, a purging solenoid valve, a purging transmitter, and a furnace pressure monitoring and purging display module, characterized in that, Includes the following steps: S1. Set up a set of sampling tubes at multiple furnace pressure measuring points on the furnace, and connect the furnace pressure prediction module and the sampling solenoid valve, furnace transmitter, purging solenoid valve, purging transmitter and pressure protection switch connected to each of the sampling tubes to the thermal distributed control module for intelligent purging program control. S2. Construct a soft measurement model of furnace pressure using the furnace pressure prediction module; S3. In the initial state, the purge solenoid valve is closed and the measurement solenoid valve is open. The furnace pressure soft measurement model generates the current furnace pressure prediction value and measures the current furnace pressure actual value through the furnace transmitter. S4. The thermal distribution control module judges the status of the sampling tube at the measuring point based on the predicted value of the furnace pressure and the actual value of the furnace pressure obtained in S3. If the status of the sampling tube at the measuring point is abnormal, it outputs a fault alarm, judges the fault type, and triggers a purging command to enter step S5. S5. The thermal distribution control module performs position holding processing on the furnace transmitter and pressure protection switch corresponding to the sampling tube of the fault alarm point, controls the measurement solenoid valve to close, and purges the solenoid valve to open. The thermal distribution control module sets the purging time according to the abnormal condition of the sampling tube of the measuring point, and the purging instrument purges the sampling tube of the measuring point through the purging circuit. S6. During the purging process, the purging transmitter measures the actual pressure value of the sampling tube at the measuring point in real time, and determines whether the sampling tube is purged based on the actual pressure value. If the tube is purged ahead of schedule or the purging time ends, the purging solenoid valve is closed. Once the purging transmitter detects that there is no residual purging gas in the sampling tube at the measuring point, the sampling solenoid valve is opened, and the furnace transmitter resumes normal measurement of the actual furnace pressure value. S7. The furnace pressure soft measurement model generates the current furnace pressure prediction value and measures the current furnace pressure actual value through the furnace transmitter. The thermal distributed control module compares the currently acquired furnace pressure prediction value with the actual furnace pressure value. If the two are close, the signal position protection processing of the furnace transmitter and pressure protection switch is released, and the adjustment and protection function of the furnace pressure related logic circuit is opened.

2. The control method for a furnace pressure purging system based on DCS control as described in claim 1, characterized in that: The specific steps for constructing the furnace pressure soft measurement model by the furnace pressure prediction module in step S2 are as follows: S21. Obtain the operating parameters of the furnace thermal power unit; S22. Analyze and judge the operating parameters of the furnace thermal power unit, and select the furnace pressure as a candidate relevant parameter; S23. Perform correlation analysis on the candidate relevant parameters of furnace pressure to obtain the contribution rate of each candidate relevant parameter, and select the relevant parameters of furnace pressure according to the contribution rate. S24. The furnace pressure-related parameters are normalized using the Mapminmax function, and the operating parameters of the furnace power unit for one day are selected as the original data for model training and learning. S25. Sample the original data of the model training and learning, set the sampling interval to 20s, and obtain 4320 sets of effective data of furnace pressure related parameters. S26. Select 300 sets of valid data related to furnace pressure parameters as training samples for the furnace pressure soft measurement model, and optimize the training samples. S27. Select the radial basis function as the model kernel function, train the model using optimized training samples, and construct a soft measurement model for furnace pressure.

3. The control method for a furnace pressure purging system based on DCS control as described in claim 2, characterized in that: The furnace pressure-related parameters include primary air fan current, primary air volume, secondary air volume, forced draft fan current, furnace flue gas temperature, induced draft fan current, and induced draft fan blade opening.

4. The control method for a furnace pressure purging system based on DCS control as described in claim 2, characterized in that: The correlation analysis calculation formula in step S23 is as follows: Where R is the contribution rate, x is the measured value of the candidate relevant parameter, y is the value of the target parameter corresponding to the candidate relevant parameter, Cov(x, y) is the covariance of the two, and D(x) and D(y) are the variances of the two.

5. The control method for a furnace pressure purging system based on DCS control as described in claim 2, characterized in that: In step S26, the training samples are used to calculate the similarity between two sets of data using a similarity calculation function. For data with high similarity, only one set of data is kept and the other set of data is deleted, thereby reducing redundancy between training sample data and thus optimizing the data.

6. The control method for a furnace pressure purging system based on DCS control as described in claim 5, characterized in that: The similarity calculation function is: R ij =exp(-||x i -x j ||2 / δ) Among them, X i For the i-th sample data, X j Let R represent the j-th sample data. ij This represents the similarity between two sets of data, where δ is the normalization parameter. If R ij If R < ε, then retain both sets of data; if R < ε, then retain both sets of data. ij If the similarity threshold is greater than ε, then two sets of data are retained, where ε is the threshold of the similarity function.

Images