CN112835950A - System and method for acquiring standard emission operation curve of wet desulphurization system based on DCS data mining - Google Patents

System and method for acquiring standard emission operation curve of wet desulphurization system based on DCS data mining Download PDF

Info

Publication number
CN112835950A
CN112835950A CN202011435762.6A CN202011435762A CN112835950A CN 112835950 A CN112835950 A CN 112835950A CN 202011435762 A CN202011435762 A CN 202011435762A CN 112835950 A CN112835950 A CN 112835950A
Authority
CN
China
Prior art keywords
data
data frame
effective
observation data
flue gas
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202011435762.6A
Other languages
Chinese (zh)
Other versions
CN112835950B (en
Inventor
文乐
曾德勇
高彦飞
张望宏
史章峰
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Huaneng Shaanxi Power Generation Co ltd
Original Assignee
Huaneng Shaanxi Power Generation Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Huaneng Shaanxi Power Generation Co ltd filed Critical Huaneng Shaanxi Power Generation Co ltd
Priority to CN202011435762.6A priority Critical patent/CN112835950B/en
Publication of CN112835950A publication Critical patent/CN112835950A/en
Application granted granted Critical
Publication of CN112835950B publication Critical patent/CN112835950B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/24—Querying
    • G06F16/245—Query processing
    • G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
    • G06F16/2465—Query processing support for facilitating data mining operations in structured databases
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/26—Visual data mining; Browsing structured data
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/28—Databases characterised by their database models, e.g. relational or object models
    • G06F16/284—Relational databases
    • G06F16/285—Clustering or classification
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]

Abstract

The invention provides a system and a method for acquiring a standard emission operating curve of a wet desulphurization system by DCS data mining, which comprises the following steps: acquiring a mining data frame from a DCS database of a coal unit; obtaining an effective data frame from the obtained mining data frame; the raw flue gas SO corresponding to each group of observation data in the obtained effective data frame2Respectively carrying out normalization processing on the concentration and the unit load to obtain an effective data frame after the normalization processing; obtaining different subclasses of each effective subdata frame; calculating the representative mass center of each sub-class of each effective sub-data frame; obtaining the load of the unit as the abscissa and the SO of the raw flue gas according to each representative mass center of the effective sub-data frames2The concentration is the coal fired boiler wet flue gas desulfurization system of the ordinate and reaches standardAn operating curve of emissions; the invention obtains the operation curve of the desulfurization equipment conforming to the actual working condition, and the operation curve is referred by operators to reasonably adjust the operation parameters and ensure SO2The pollutants reach the standard and are discharged.

Description

System and method for acquiring standard emission operation curve of wet desulphurization system based on DCS data mining
Technical Field
The invention relates to the field of wet desulphurization systems of coal-fired boilers, in particular to a system and a method for acquiring a standard emission operation curve of a wet desulphurization system for DCS data mining.
Background
At present, in order to enhance the profitability of enterprises, coal-fired power stations begin to intensively mix and burn low-calorific-value and high-sulfur coal types, and the high-sulfur coal types are easy to cause raw flue gas SO2The concentration is increased, if the ultimate output of the operation condition of the desulfurization system is reached, the clean flue gas SO is very easily caused2The hourly mean value of the concentration exceeds 35mg/Nm3Therefore, the method is examined by environmental protection departments. Generally, performance parameters and characteristic curves of a desulfurization system are obtained by a performance test method, but the test method needs to consume higher labor, material and financial costs, and a test result can only represent equipment characteristics under a test working condition, and the representativeness of the test result is weakened when the operation working condition and the boundary condition are changed.
Disclosure of Invention
The invention aims to provide a system and a method for acquiring a standard emission operation curve of a wet desulphurization system for DCS data mining, which solve the defects of high cost and low efficiency of the method for acquiring performance parameters and characteristic curves of the desulphurization system in the prior art.
In order to achieve the purpose, the invention adopts the technical scheme that:
the invention provides a method for acquiring a standard emission operating curve of a wet desulphurization system for DCS data mining, which comprises the following steps:
step 1, taking 50% rated load to 100% rated load of a coal-fired unit as a sampling range, and acquiring a mining data frame from a DCS (distributed control system) database of the coal-fired unit in a preset sampling period, wherein the mining data frame comprises a plurality of groups of observation data, and each group of observation data comprises raw flue gas SO2Concentration, clean flue gas SO2Concentration, flue gas flow, serial number of a slurry circulating pump, pH value of slurry in an absorption tower, liquid level of the absorption tower, desulfurization efficiency and unit load;
step 2, obtaining an effective data frame from the mining data frame obtained in the step 1, wherein a QQ diagram corresponding to the pH value of the absorption tower slurry in each group of observation data in the effective data frame and a QQ diagram corresponding to the liquid level of the absorption tower are consistent with the trend of a normal distribution straight line;
step 3, the raw flue gas SO corresponding to each group of observation data in the effective data frame obtained in the step 2 is processed2Respectively carrying out normalization processing on the concentration and the unit load to obtain an effective data frame after the normalization processing;
step 4, dividing all observation data in the effective data frame after normalization processing obtained in the step 3 according to the slurry circulating pumps obtained in the step 1 to obtain a plurality of sub data frames, wherein each sub data frame corresponds to the slurry circulating pump with the same number;
step 5, selecting clean flue gas SO from all observation data in each sub data frame obtained in step 42The concentration exceeds the observation data corresponding to the set threshold, and the observation data is used as effective observation data of the sub data frames, so that a plurality of effective sub data frames are obtained; clustering the obtained observation data in each effective sub data frame to obtain different subclasses of each effective sub data frame;
step 6, calculating the representative mass center of each subclass of each effective subdata box in the step 5;
step 7, obtaining the original flue gas SO by taking the unit load as the abscissa and taking the original flue gas load as the abscissa according to each representative mass center of the effective sub data frames obtained in the step 62And the concentration is the operation curve of the standard emission of the coal-fired boiler wet desulphurization system of the ordinate.
Preferably, in step 2, the effective data frame is obtained from the mining data frame obtained in step 1, and the specific method is as follows:
s201, deleting observation data corresponding to the PH value of the slurry of the maximum absorption tower, observation data corresponding to the PH value of the slurry of the minimum absorption tower, observation data corresponding to the liquid level value of the maximum absorption tower and observation data corresponding to the liquid level value of the minimum absorption tower from the plurality of groups of observation data obtained in the step 1; obtaining a data frame;
s202, respectively drawing a QQ diagram corresponding to the PH value of the absorption tower slurry in each group of observation data in the data frame and a QQ diagram corresponding to the liquid level of the absorption tower;
s203, judging whether the trend of the normal distribution straight line of the QQ diagrams corresponding to the PH values of all the obtained absorption tower serosity and the QQ diagrams corresponding to the liquid level of the absorption tower are consistent, wherein if the trend of the normal distribution straight line of the QQ diagrams corresponding to the PH values of the absorption tower serosity and the QQ diagrams corresponding to the liquid level of the absorption tower of all the observation data in the data frame is consistent, taking the data frame as an effective data frame; otherwise, entering S204;
s204, deleting the observation data corresponding to the PH value of the slurry of the maximum absorption tower, the observation data corresponding to the PH value of the slurry of the minimum absorption tower, the observation data corresponding to the liquid level value of the maximum absorption tower and the observation data corresponding to the liquid level value of the minimum absorption tower from the data frame obtained in S201 to obtain a primary data frame,
s205, taking the sequential data frames obtained in S204 as the data frames in S201, and iteratively executing S202 to S204; until the trend of the QQ graph corresponding to the PH value of the absorption tower slurry of each group of observation data in the obtained primary data frame and the trend of the QQ graph corresponding to the liquid level of the absorption tower are consistent with the trend of a normal distribution straight line; and then the obtained primary data frame is used as an effective data frame.
Preferably, in step 3, the raw flue gas SO corresponding to each set of observation data in the valid data frame obtained in step 2 is obtained by the following formula2Respectively carrying out normalization treatment on the concentration and the unit load:
Figure BDA0002827256480000031
in the formula: y isi、yiRespectively before and after normalization2Concentration; max (Y) represents all raw flue gases SO before normalization2The maximum value of the concentration; min (Y) represents all raw flue gases SO before normalization2The minimum value of concentration; fi、fiRespectively normalizing the unit loads before and after; max (f) represents the maximum value of the unit load before all normalization; min (F) represents the minimum value of the unit load before all normalization.
Preferably, in step 4, according to the serial number of the slurry circulation pump obtained in step 1, all the observation data in the effective data frame after the normalization processing obtained in step 3 are divided to obtain a plurality of sub data frames, and the specific method is as follows:
and dividing all the observation data in the effective data frame after the normalization processing according to the serial number of the slurry circulating pump in each group of observation data to obtain a plurality of subdata frames.
Preferably, in step 5, the threshold value is the clean flue gas SO2Concentration multiplied by a factor of safety CsWherein, the safety factor CsThe value is 0.95;
clustering the obtained observation data in each effective subdata box, wherein the specific method comprises the following steps:
and clustering the obtained observation data in each effective sub data frame by using a density-based clustering algorithm DBSCAN to obtain different subclasses of each effective sub data frame.
Preferably, in step 6, the representative centroid of each sub-class of each valid sub-data box in step 5 is calculated by the following specific method:
respectively calculating the raw flue gas SO corresponding to each group of observation data in each sub-class in each effective sub-data frame2Mean value of concentration
Figure BDA0002827256480000041
And median number
Figure BDA0002827256480000042
And mean value of unit load
Figure BDA0002827256480000043
And median
Figure BDA0002827256480000044
The obtained raw flue gas SO2Mean value of concentration
Figure BDA0002827256480000045
Raw flue gas SO2Median of concentration
Figure BDA0002827256480000046
Mean value of unit load
Figure BDA0002827256480000047
Median of sum unit load
Figure BDA0002827256480000048
As two coordinate points, respectively
Figure BDA0002827256480000049
Will be provided with
Figure BDA00028272564800000410
And
Figure BDA00028272564800000411
the line segment between the two points is divided into 10 equal parts by using 9 dividing points, and the sum L of Euclidean distances of each dividing point is calculated by the following formulasumObtaining the sum of 9 Euclidean distances;
Figure BDA00028272564800000412
and selecting the segmentation point corresponding to the minimum sum of Euclidean distances from the 9 segmentation points as the representative centroid.
Preferably, in step 7, the unit load is taken as the abscissa and the raw flue gas SO is obtained according to each representative centroid of the effective sub data frames obtained in step 62The method is characterized in that the concentration is a running curve of standard discharge of the coal-fired boiler wet desulphurization system with a vertical coordinate, and the specific method comprises the following steps:
performing reverse normalization processing on each representative mass center in each effective sub data frame obtained in the step 6 to obtain a coordinate value of each representative mass center after the reverse normalization processing;
performing linear regression on the obtained coordinate values after the inverse normalization processing of all the representative centroids to fit a straight line, namely obtaining the original flue gas SO with the unit load as the abscissa2At a concentration ofAnd the vertical coordinate is the operation curve of the coal-fired boiler wet desulphurization system reaching the standard for emission.
A system for acquiring the emission curve of the wet desulphurization system by DCS data mining, which can be used for realizing the method for acquiring the emission curve of the wet desulphurization system by DCS data mining, comprises an acquisition unit, a data mining module, a data normalization processing module, a data division module, a data clustering processing module, a data analysis module and a data statistics module, wherein,
the acquisition unit is used for acquiring a mining data frame from a DCS database of the coal-fired unit in a preset sampling period by taking 50% rated load to 100% rated load of the coal-fired unit as a sampling range, wherein the mining data frame comprises a plurality of groups of observation data, and each group of observation data comprises raw flue gas SO2Concentration, clean flue gas SO2Concentration, flue gas flow, serial number of a slurry circulating pump, pH value of slurry in an absorption tower, liquid level of the absorption tower, desulfurization efficiency and unit load;
the data mining module is used for acquiring an effective data frame from the obtained mining data frame, wherein a QQ diagram corresponding to the PH value of the absorption tower slurry in each group of observation data in the effective data frame and a QQ diagram corresponding to the liquid level of the absorption tower are consistent with the trend of a normal distribution straight line;
the data normalization processing module is used for performing normalization processing on the raw flue gas SO corresponding to each group of observation data in the obtained effective data frame2Respectively carrying out normalization processing on the concentration and the unit load to obtain an effective data frame after the normalization processing;
the data dividing module is used for dividing all the observation data in the obtained effective data frame after the normalization processing according to the obtained slurry circulating pump to obtain a plurality of sub data frames, wherein each sub data frame corresponds to the slurry circulating pump with the same number;
the data clustering processing module is used for selecting clean flue gas SO from all the obtained observation data in each subdata frame2The concentration exceeds the observation data corresponding to the set threshold, and the observation data is used as effective observation data of the sub data frames, so that a plurality of effective sub data frames are obtained;clustering the obtained observation data in each effective sub data frame to obtain different subclasses of each effective sub data frame;
the data analysis module is used for calculating the representative centroid of each subclass of each effective subdata box;
the data statistical module is used for obtaining and obtaining the raw flue gas SO by taking the unit load as the abscissa according to each representative mass center of each effective subdata frame2And the concentration is the operation curve of the standard emission of the coal-fired boiler wet desulphurization system of the ordinate.
Compared with the prior art, the invention has the beneficial effects that:
according to the system and the method for acquiring the standard emission operating curve of the wet desulphurization system based on DCS data mining, the standard emission operating curve of the wet desulphurization system is acquired by using the actual operating parameters of a coal-fired unit, so that the defect of high cost of the existing performance test method can be effectively overcome; meanwhile, the operation curve of the desulfurization equipment conforming to the actual working condition is obtained for operators to refer to SO as to reasonably adjust the operation parameters and ensure SO2The pollutants reach the standard and are discharged.
Drawings
FIG. 1 is a schematic flow diagram of the present invention;
FIG. 2 is a QQ chart for judging whether the pH value measurement point data of the slurry in the absorption tower is approximately normally distributed;
FIG. 3 shows 3 subclasses and their representative centroids obtained after clustering by DBSCAN algorithm;
FIG. 4 is a calculation example of the standard discharge operating curve of the wet desulfurization system calculated by the method of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings.
The method for acquiring the standard emission operation curve of the wet desulphurization system based on DCS data mining obtains the operation curve of the desulphurization equipment according with the actual working condition, and the operation curve is referred by operators to reasonably adjust the operation parameters SO as to ensure that SO2The pollutants reach the standard and are discharged.
As shown in fig. 1 to 4, the method for obtaining the emission standard operation curve of the wet desulphurization system for DCS data mining provided by the present invention comprises the following steps:
step 1, taking 50% rated load to 100% rated load of a coal-fired unit as a sampling range, and acquiring a mining data frame from a DCS (distributed control system) database of the coal-fired unit in a preset sampling period, wherein the mining data frame comprises a plurality of groups of observation data, and each group of observation data comprises raw flue gas SO2Concentration, clean flue gas SO2Concentration, flue gas flow, serial number of a slurry circulating pump, pH value of slurry in an absorption tower, liquid level of the absorption tower, desulfurization efficiency and unit load;
step 2, obtaining an effective data frame from the mining data frame obtained in the step 1, wherein a QQ diagram corresponding to the pH value of the absorption tower slurry in each group of observation data in the effective data frame and a QQ diagram corresponding to the liquid level of the absorption tower are consistent with the trend of a normal distribution straight line;
step 3, the raw flue gas SO corresponding to each group of observation data in the effective data frame obtained in the step 2 is processed2Respectively carrying out normalization processing on the concentration and the unit load to obtain an effective data frame after the normalization processing;
step 4, dividing all observation data in the effective data frame after normalization processing obtained in the step 3 according to the slurry circulating pumps obtained in the step 1 to obtain a plurality of sub data frames, wherein each sub data frame corresponds to the slurry circulating pump with the same number;
step 5, selecting clean flue gas SO from all observation data in each sub data frame obtained in step 42The concentration exceeds the observation data corresponding to the set threshold, and the observation data is used as effective observation data of the sub data frames, so that a plurality of effective sub data frames are obtained;
clustering the obtained observation data in each effective sub data frame by using a density-based clustering algorithm DBSCAN to obtain different subclasses of each effective sub data frame;
step 6, calculating the representative mass center of each subclass of each effective subdata box in the step 5;
step 7, performing inverse normalization processing on each representative mass center of the effective sub data frames obtained in the step 6; obtaining coordinate values of each representative centroid after reverse normalization processing;
step 8, performing linear regression fitting on all the coordinate values after the representative centroid inverse normalization processing in the step 7 to obtain the raw flue gas SO with the unit load as the abscissa2And the concentration is the operation curve of the standard emission of the coal-fired boiler wet desulphurization system of the ordinate.
Further, in step 1, the sampling period is 60 seconds.
Further, in step 2, obtaining an effective data frame from the mining data frame obtained in step 1, specifically, the method includes:
s201, deleting observation data corresponding to the PH value of the slurry of the maximum absorption tower, observation data corresponding to the PH value of the slurry of the minimum absorption tower, observation data corresponding to the liquid level value of the maximum absorption tower and observation data corresponding to the liquid level value of the minimum absorption tower from the plurality of groups of observation data obtained in the step 1; obtaining a data frame;
s202, respectively drawing a QQ diagram corresponding to the PH value of the absorption tower slurry in each group of observation data in the data frame and a QQ diagram corresponding to the liquid level of the absorption tower;
s203, judging whether the trend of the normal distribution straight line of the QQ diagrams corresponding to the PH values of all the obtained absorption tower serosity and the QQ diagrams corresponding to the liquid level of the absorption tower are consistent, wherein if the trend of the normal distribution straight line of the QQ diagrams corresponding to the PH values of the absorption tower serosity and the QQ diagrams corresponding to the liquid level of the absorption tower of all the observation data in the data frame is consistent, taking the data frame as an effective data frame; otherwise, entering S204;
s204, deleting the observation data corresponding to the PH value of the slurry of the maximum absorption tower, the observation data corresponding to the PH value of the slurry of the minimum absorption tower, the observation data corresponding to the liquid level value of the maximum absorption tower and the observation data corresponding to the liquid level value of the minimum absorption tower from the data frame obtained in S201 to obtain a primary data frame,
s205, taking the sequential data frames obtained in S204 as the data frames in S201, and iteratively executing S202 to S204; until the trend of the QQ graph corresponding to the PH value of the absorption tower slurry of each group of observation data in the obtained primary data frame and the trend of the QQ graph corresponding to the liquid level of the absorption tower are consistent with the trend of a normal distribution straight line; and then the obtained primary data frame is used as an effective data frame.
Further, in step 3, the raw flue gas SO corresponding to each set of observation data in the effective data frame obtained in step 2 is calculated according to the following formula2Respectively carrying out normalization treatment on the concentration and the unit load:
Figure BDA0002827256480000081
in the formula: y isi、yiRespectively before and after normalization2Concentration measuring point data; max (Y) represents all raw flue gases SO before normalization2The maximum value of the concentration measuring point data; min (Y) represents all raw flue gases SO before normalization2The minimum value of the data of the concentration measuring points; fi、fiRespectively normalizing the unit load measuring point data before and after; max (F) represents the maximum value of the unit load measuring point data before normalization; and min (F) represents the minimum value of all the unit load measuring point data before normalization.
Further, in step 4, according to the serial number of the slurry circulation pump obtained in step 1, dividing all observation data in the effective data frame after normalization processing obtained in step 3 to obtain a plurality of sub data frames, and the specific method is as follows:
and dividing all the observation data in the effective data frame after the normalization processing according to the serial numbers of the slurry circulating pumps in each group of observation data to obtain a plurality of sub data frames, wherein the serial numbers of the slurry circulating pumps in each sub data frame are the same.
Further, in step 5, selecting clean flue gas SO from all observation data in each sub data frame obtained in step 42The method comprises the following specific steps that observation data corresponding to the concentration exceeding the set threshold are used as a plurality of effective observation data of each sub data frame:
setting a threshold value which is the clean flue gas SO2Concentration multiplied by a factor of safety CsWherein, the safety factor CsThe value is 0.95;
selecting clean flue gas SO from all observation data in each sub data frame obtained in step 42The concentration exceeds the observation data corresponding to the set threshold, and the observation data is used as effective observation data of the sub data frames, so that a plurality of effective sub data frames are obtained;
and clustering the obtained observation data in each effective sub data frame by using a density-based clustering algorithm DBSCAN to obtain different subclasses of each effective sub data frame.
Further, in step 6, calculating a representative centroid of each sub-class of each valid sub-data box in step 5, specifically: respectively calculating the raw flue gas SO corresponding to each group of observation data in each sub-class in each effective sub-data frame2Mean value of concentration
Figure BDA0002827256480000091
And median number
Figure BDA0002827256480000092
And mean value of unit load
Figure BDA0002827256480000093
And median
Figure BDA0002827256480000094
The obtained raw flue gas SO2Mean value of concentration
Figure BDA0002827256480000095
Raw flue gas SO2Median of concentration
Figure BDA0002827256480000096
Mean value of unit load
Figure BDA0002827256480000097
Median of sum unit load
Figure BDA0002827256480000098
As twoCoordinate points are respectively
Figure BDA0002827256480000099
Will be provided with
Figure BDA00028272564800000910
And
Figure BDA00028272564800000911
the line segment between the two points is divided into 10 equal parts by using 9 dividing points, and the sum L of Euclidean distances of each dividing point is calculated by the following formulasumObtaining the sum of 9 Euclidean distances;
Figure BDA00028272564800000912
selecting a segmentation point corresponding to the minimum Euclidean distance sum from the 9 segmentation points as a representative centroid;
fig. 3 shows 3 subcategories and their representative centroids obtained after clustering by the DBSCAN algorithm.
Further, in step 7, performing inverse normalization processing on each representative centroid in each effective sub data frame obtained in step 6 through the following formula to obtain a coordinate value of each representative centroid after inverse normalization processing;
Figure BDA00028272564800000913
in the formula: y isc、YcA certain subclass before and after reverse normalization represents the ordinate data of the centroid, fc、FcRespectively representing the abscissa data of the centroid for a certain subclass before and after inverse normalization;
further, in step 8, linear regression fitting is carried out on all the coordinate values obtained in step 7 after the reverse normalization processing of the representative centroid, namely, the unit load is taken as the abscissa, and the raw flue gas SO is obtained2Coal-fired boiler wet desulphurization system with concentration as ordinate and operation for up-to-standard dischargeThe row curve, as shown in fig. 4.
The invention provides a system for acquiring a standard emission operating curve of a wet desulphurization system for DCS data mining, which can be used for realizing the method for acquiring the standard emission operating curve of the wet desulphurization system for DCS data mining, and comprises an acquisition unit, a data mining module, a data normalization processing module, a data division module, a data clustering processing module, a data analysis module and a data statistics module, wherein,
the acquisition unit is used for acquiring a mining data frame from a DCS database of the coal-fired unit in a preset sampling period by taking 50% rated load to 100% rated load of the coal-fired unit as a sampling range, wherein the mining data frame comprises a plurality of groups of observation data, and each group of observation data comprises raw flue gas SO2Concentration, clean flue gas SO2Concentration, flue gas flow, serial number of a slurry circulating pump, pH value of slurry in an absorption tower, liquid level of the absorption tower, desulfurization efficiency and unit load;
the data mining module is used for acquiring an effective data frame from the obtained mining data frame, wherein a QQ diagram corresponding to the PH value of the absorption tower slurry in each group of observation data in the effective data frame and a QQ diagram corresponding to the liquid level of the absorption tower are consistent with the trend of a normal distribution straight line;
the data normalization processing module is used for performing normalization processing on the raw flue gas SO corresponding to each group of observation data in the obtained effective data frame2Respectively carrying out normalization processing on the concentration and the unit load to obtain an effective data frame after the normalization processing;
the data dividing module is used for dividing all the observation data in the obtained effective data frame after the normalization processing according to the obtained slurry circulating pump to obtain a plurality of sub data frames, wherein each sub data frame corresponds to the slurry circulating pump with the same number;
the data clustering processing module is used for selecting clean flue gas SO from all the obtained observation data in each subdata frame2The concentration exceeds the observation data corresponding to the set threshold, and the observation data is used as effective observation data of the sub data frames, so that a plurality of effective sub data frames are obtained; and is effective for each obtainedClustering the observation data in the sub data frames to obtain different subclasses of each effective sub data frame;
the data analysis module is used for calculating the representative centroid of each subclass of each effective subdata box;
the data statistical module is used for obtaining and obtaining the raw flue gas SO by taking the unit load as the abscissa according to each representative mass center of each effective subdata frame2And the concentration is the operation curve of the standard emission of the coal-fired boiler wet desulphurization system of the ordinate.
FIG. 4 shows the standard discharge operation curve of the wet desulfurization system calculated by the method of the present invention under the operation conditions of 4 or 5 slurry circulating pumps, and the unit SO can be obtained by the unit operation attendant by looking up the curve according to the current operation load of the unit2Raw flue gas SO required by emission not exceeding standard2The upper concentration limit, which is a pilot signal that the output of the desulfurization system is about to reach or exceed the limit during normal operation.

Claims (8)

1. A method for acquiring a standard emission operation curve of a wet desulphurization system mined by DCS data is characterized by comprising the following steps:
step 1, taking 50% rated load to 100% rated load of a coal-fired unit as a sampling range, and acquiring a mining data frame from a DCS (distributed control system) database of the coal-fired unit in a preset sampling period, wherein the mining data frame comprises a plurality of groups of observation data, and each group of observation data comprises raw flue gas SO2Concentration, clean flue gas SO2Concentration, flue gas flow, serial number of a slurry circulating pump, pH value of slurry in an absorption tower, liquid level of the absorption tower, desulfurization efficiency and unit load;
step 2, obtaining an effective data frame from the mining data frame obtained in the step 1, wherein a QQ diagram corresponding to the pH value of the absorption tower slurry in each group of observation data in the effective data frame and a QQ diagram corresponding to the liquid level of the absorption tower are consistent with the trend of a normal distribution straight line;
step 3, the raw flue gas SO corresponding to each group of observation data in the effective data frame obtained in the step 2 is processed2Respectively carrying out normalization processing on concentration and unit loadObtaining an effective data frame after normalization processing;
step 4, dividing all observation data in the effective data frame after normalization processing obtained in the step 3 according to the slurry circulating pumps obtained in the step 1 to obtain a plurality of sub data frames, wherein each sub data frame corresponds to the slurry circulating pump with the same number;
step 5, selecting clean flue gas SO from all observation data in each sub data frame obtained in step 42The concentration exceeds the observation data corresponding to the set threshold, and the observation data is used as effective observation data of the sub data frames, so that a plurality of effective sub data frames are obtained; clustering the obtained observation data in each effective sub data frame to obtain different subclasses of each effective sub data frame;
step 6, calculating the representative mass center of each subclass of each effective subdata box in the step 5;
step 7, obtaining the original flue gas SO by taking the unit load as the abscissa and taking the original flue gas load as the abscissa according to each representative mass center of the effective sub data frames obtained in the step 62And the concentration is the operation curve of the standard emission of the coal-fired boiler wet desulphurization system of the ordinate.
2. The method of claim 1, wherein in step 2, the valid data frame is obtained from the mining data frame obtained in step 1, and the specific method is as follows:
s201, deleting observation data corresponding to the PH value of the slurry of the maximum absorption tower, observation data corresponding to the PH value of the slurry of the minimum absorption tower, observation data corresponding to the liquid level value of the maximum absorption tower and observation data corresponding to the liquid level value of the minimum absorption tower from the plurality of groups of observation data obtained in the step 1; obtaining a data frame;
s202, respectively drawing a QQ diagram corresponding to the PH value of the absorption tower slurry in each group of observation data in the data frame and a QQ diagram corresponding to the liquid level of the absorption tower;
s203, judging whether the trend of the normal distribution straight line of the QQ diagrams corresponding to the PH values of all the obtained absorption tower serosity and the QQ diagrams corresponding to the liquid level of the absorption tower are consistent, wherein if the trend of the normal distribution straight line of the QQ diagrams corresponding to the PH values of the absorption tower serosity and the QQ diagrams corresponding to the liquid level of the absorption tower of all the observation data in the data frame is consistent, taking the data frame as an effective data frame; otherwise, entering S204;
s204, deleting the observation data corresponding to the PH value of the slurry of the maximum absorption tower, the observation data corresponding to the PH value of the slurry of the minimum absorption tower, the observation data corresponding to the liquid level value of the maximum absorption tower and the observation data corresponding to the liquid level value of the minimum absorption tower from the data frame obtained in S201 to obtain a primary data frame,
s205, taking the sequential data frames obtained in S204 as the data frames in S201, and iteratively executing S202 to S204; until the trend of the QQ graph corresponding to the PH value of the absorption tower slurry of each group of observation data in the obtained primary data frame and the trend of the QQ graph corresponding to the liquid level of the absorption tower are consistent with the trend of a normal distribution straight line; and then the obtained primary data frame is used as an effective data frame.
3. The method of claim 1, wherein in step 3, the raw flue gas SO corresponding to each set of observation data in the valid data frame obtained in step 2 is processed by the following formula2Respectively carrying out normalization treatment on the concentration and the unit load:
Figure FDA0002827256470000021
in the formula: y isi、yiRespectively before and after normalization2Concentration; max (Y) represents all raw flue gases SO before normalization2The maximum value of the concentration; min (Y) represents all raw flue gases SO before normalization2The minimum value of concentration; fi、fiRespectively normalizing the unit loads before and after; max (f) represents the maximum value of the unit load before all normalization; min (F) represents the minimum value of the unit load before all normalization.
4. The method of claim 1, wherein in step 4, all observation data in the normalized valid data frame obtained in step 3 are divided according to the serial number of the slurry circulating pump obtained in step 1 to obtain a plurality of sub data frames, and the specific method is as follows:
and dividing all the observation data in the effective data frame after the normalization processing according to the serial number of the slurry circulating pump in each group of observation data to obtain a plurality of subdata frames.
5. The method of claim 1, wherein in step 5, the threshold is clean flue gas SO2Concentration multiplied by a factor of safety CsWherein, the safety factor CsThe value is 0.95;
clustering the obtained observation data in each effective subdata box, wherein the specific method comprises the following steps:
and clustering the obtained observation data in each effective sub data frame by using a density-based clustering algorithm DBSCAN to obtain different subclasses of each effective sub data frame.
6. The method of claim 1, wherein in step 6, the representative centroid of each sub-class of each active sub-data box in step 5 is calculated by:
respectively calculating the raw flue gas SO corresponding to each group of observation data in each sub-class in each effective sub-data frame2Mean value of concentration
Figure FDA0002827256470000031
And median number
Figure FDA0002827256470000032
And mean value of unit load
Figure FDA0002827256470000033
And median
Figure FDA0002827256470000034
The obtained raw flue gas SO2Mean value of concentration
Figure FDA0002827256470000035
Raw flue gas SO2Median of concentration
Figure FDA0002827256470000036
Mean value of unit load
Figure FDA0002827256470000037
Median of sum unit load
Figure FDA0002827256470000038
As two coordinate points, respectively
Figure FDA0002827256470000039
Will be provided with
Figure FDA00028272564700000310
And
Figure FDA00028272564700000311
the line segment between the two points is divided into 10 equal parts by using 9 dividing points, and the sum L of Euclidean distances of each dividing point is calculated by the following formulasumObtaining the sum of 9 Euclidean distances;
Figure FDA0002827256470000041
and selecting the segmentation point corresponding to the minimum sum of Euclidean distances from the 9 segmentation points as the representative centroid.
7. The method of claim 1, wherein in step 7, the raw flue gas SO is obtained according to each representative centroid of each valid subframe obtained in step 6, with the unit load as abscissa, and the raw flue gas load as abscissa2The method is characterized in that the concentration is a running curve of standard discharge of the coal-fired boiler wet desulphurization system with a vertical coordinate, and the specific method comprises the following steps:
performing reverse normalization processing on each representative mass center in each effective sub data frame obtained in the step 6 to obtain a coordinate value of each representative mass center after the reverse normalization processing;
performing linear regression fitting on the obtained coordinate values after the inverse normalization processing of all the representative mass centers to obtain the coal-fired boiler with the unit load as the abscissa and the raw flue gas SO2And the concentration is an operation curve of the standard emission of the wet desulphurization system with the ordinate.
8. A system for acquiring the emission standard operation curve of a DCS data mining wet desulphurization system, which is characterized in that the system can be used for realizing the method for acquiring the emission standard operation curve of the DCS data mining wet desulphurization system of any one of claims 1-7, and comprises an acquisition unit, a data mining module, a data normalization processing module, a data division module, a data clustering processing module, a data analysis module and a data statistics module, wherein,
the acquisition unit is used for acquiring a mining data frame from a DCS database of the coal-fired unit in a preset sampling period by taking 50% rated load to 100% rated load of the coal-fired unit as a sampling range, wherein the mining data frame comprises a plurality of groups of observation data, and each group of observation data comprises raw flue gas SO2Concentration, clean flue gas SO2Concentration, flue gas flow, serial number of a slurry circulating pump, pH value of slurry in an absorption tower, liquid level of the absorption tower, desulfurization efficiency and unit load;
the data mining module is used for acquiring an effective data frame from the obtained mining data frame, wherein a QQ diagram corresponding to the PH value of the absorption tower slurry in each group of observation data in the effective data frame and a QQ diagram corresponding to the liquid level of the absorption tower are consistent with the trend of a normal distribution straight line;
the data normalization processing module is used for performing normalization processing on the raw flue gas SO corresponding to each group of observation data in the obtained effective data frame2Respectively carrying out normalization processing on the concentration and the unit load to obtain an effective data frame after the normalization processing;
the data dividing module is used for dividing all the observation data in the obtained effective data frame after the normalization processing according to the obtained slurry circulating pump to obtain a plurality of sub data frames, wherein each sub data frame corresponds to the slurry circulating pump with the same number;
the data clustering processing module is used for selecting clean flue gas SO from all the obtained observation data in each subdata frame2The concentration exceeds the observation data corresponding to the set threshold, and the observation data is used as effective observation data of the sub data frames, so that a plurality of effective sub data frames are obtained; clustering the obtained observation data in each effective sub data frame to obtain different subclasses of each effective sub data frame;
the data analysis module is used for calculating the representative centroid of each subclass of each effective subdata box;
the data statistical module is used for obtaining and obtaining the raw flue gas SO by taking the unit load as the abscissa according to each representative mass center of each effective subdata frame2And the concentration is the operation curve of the standard emission of the coal-fired boiler wet desulphurization system of the ordinate.
CN202011435762.6A 2020-12-09 2020-12-09 System and method for acquiring standard emission operation curve of wet desulphurization system based on DCS data mining Active CN112835950B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011435762.6A CN112835950B (en) 2020-12-09 2020-12-09 System and method for acquiring standard emission operation curve of wet desulphurization system based on DCS data mining

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011435762.6A CN112835950B (en) 2020-12-09 2020-12-09 System and method for acquiring standard emission operation curve of wet desulphurization system based on DCS data mining

Publications (2)

Publication Number Publication Date
CN112835950A true CN112835950A (en) 2021-05-25
CN112835950B CN112835950B (en) 2023-03-28

Family

ID=75923488

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011435762.6A Active CN112835950B (en) 2020-12-09 2020-12-09 System and method for acquiring standard emission operation curve of wet desulphurization system based on DCS data mining

Country Status (1)

Country Link
CN (1) CN112835950B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115155269A (en) * 2022-09-09 2022-10-11 启东凯顺机械制造有限公司 Automatic control method of gas fine desulfurization system

Citations (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2000051055A2 (en) * 1999-02-22 2000-08-31 Vialogy Corporation Method and apparatus for monitoring therapy effectiveness
US20060042525A1 (en) * 2004-08-27 2006-03-02 Neuco, Inc. Method and system for SCR Optimization
US20060052902A1 (en) * 2004-08-27 2006-03-09 Neuco, Inc. Method and system for SNCR optimization
CN104123593A (en) * 2014-07-16 2014-10-29 上海交通大学 Coal consumption characteristic curve on-line rolling update based multi-mode load scheduling method
CN105223036A (en) * 2015-09-28 2016-01-06 广东电网有限责任公司电力科学研究院 MgO flue gas desulfurization performance on-site verification method and system
CN105808902A (en) * 2014-12-27 2016-07-27 上海麦杰环境科技有限公司 Qualitative method used for analyzing operational condition of wet desulphurization system
CN106089328A (en) * 2016-08-10 2016-11-09 西安热工研究院有限公司 Steam turbine pitch rating curve discrimination method based on DCS data mining
CN106703904A (en) * 2016-11-18 2017-05-24 华能国际电力开发公司铜川照金电厂 Method for optimizing steam distribution curves of steam turbines on basis of data mining technologies
CN107940501A (en) * 2017-11-30 2018-04-20 国网辽宁省电力有限公司电力科学研究院 Air and flue system control optimization method after the transformation of fired power generating unit desulphurization denitration
CN109523180A (en) * 2018-11-27 2019-03-26 国投北部湾发电有限公司 A kind of thermal power plant's coal consumption and heat supply online monitoring system
CN110033141A (en) * 2019-04-22 2019-07-19 大唐环境产业集团股份有限公司 A kind of method for building up of desulphurization system operating condition database
CN110496507A (en) * 2019-08-12 2019-11-26 厦门邑通软件科技有限公司 The method of calcium sulfite concentration is fitted in a kind of wet desulfurizing process
CN111881554A (en) * 2020-06-29 2020-11-03 东北电力大学 Optimization control method for boiler changing along with air temperature

Patent Citations (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2000051055A2 (en) * 1999-02-22 2000-08-31 Vialogy Corporation Method and apparatus for monitoring therapy effectiveness
US20060042525A1 (en) * 2004-08-27 2006-03-02 Neuco, Inc. Method and system for SCR Optimization
US20060052902A1 (en) * 2004-08-27 2006-03-09 Neuco, Inc. Method and system for SNCR optimization
CN104123593A (en) * 2014-07-16 2014-10-29 上海交通大学 Coal consumption characteristic curve on-line rolling update based multi-mode load scheduling method
CN105808902A (en) * 2014-12-27 2016-07-27 上海麦杰环境科技有限公司 Qualitative method used for analyzing operational condition of wet desulphurization system
CN105223036A (en) * 2015-09-28 2016-01-06 广东电网有限责任公司电力科学研究院 MgO flue gas desulfurization performance on-site verification method and system
CN106089328A (en) * 2016-08-10 2016-11-09 西安热工研究院有限公司 Steam turbine pitch rating curve discrimination method based on DCS data mining
CN106703904A (en) * 2016-11-18 2017-05-24 华能国际电力开发公司铜川照金电厂 Method for optimizing steam distribution curves of steam turbines on basis of data mining technologies
CN107940501A (en) * 2017-11-30 2018-04-20 国网辽宁省电力有限公司电力科学研究院 Air and flue system control optimization method after the transformation of fired power generating unit desulphurization denitration
CN109523180A (en) * 2018-11-27 2019-03-26 国投北部湾发电有限公司 A kind of thermal power plant's coal consumption and heat supply online monitoring system
CN110033141A (en) * 2019-04-22 2019-07-19 大唐环境产业集团股份有限公司 A kind of method for building up of desulphurization system operating condition database
CN110496507A (en) * 2019-08-12 2019-11-26 厦门邑通软件科技有限公司 The method of calcium sulfite concentration is fitted in a kind of wet desulfurizing process
CN111881554A (en) * 2020-06-29 2020-11-03 东北电力大学 Optimization control method for boiler changing along with air temperature

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
QIONG HUANG,ET AL.: "Efficient Designated Confirmer Signature and DCS-Based Ambiguous Optimistic Fair Exchange", 《 IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY》 *
叶秋生: "湿式石灰石—石膏法烟气脱硫专家系统的开发研究", 《中国优秀硕士学位论文全文数据库》 *
范常浩等: "基于SOPSO算法的CFB机组联合脱硫系统经济性优化研究", 《计算机与应用化学》 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115155269A (en) * 2022-09-09 2022-10-11 启东凯顺机械制造有限公司 Automatic control method of gas fine desulfurization system

Also Published As

Publication number Publication date
CN112835950B (en) 2023-03-28

Similar Documents

Publication Publication Date Title
CN114911209B (en) Garlic processing wastewater treatment management system based on data analysis
CN109472040B (en) Method and device for dynamically correcting yaw wind alignment deviation of wind turbine generator
CN114749006B (en) Optimization method for wet desulfurization slurry circulating pump
CN113806675B (en) NOx emission and oil consumption characteristic analysis method
CN112835950B (en) System and method for acquiring standard emission operation curve of wet desulphurization system based on DCS data mining
CN115738622B (en) Tail gas emission detection system of desulfurization equipment
WO2024041591A1 (en) Coordinated method for optimizing wind power-photovoltaic energy storage ratio
CN113792988A (en) Online monitoring data anomaly identification method for enterprise
CN116228017A (en) Intelligent overall process control method based on enterprise environment-friendly big data aggregation
CN109753720B (en) Air quality monitoring data anomaly investigation method, system, equipment and medium
CN114781757A (en) Multi-unit operation and maintenance strategy optimization method for offshore wind farm
CN114295778A (en) Thermal power plant-based flue gas emission quality evaluation method and evaluation system
CN107730070B (en) Intelligent substation relay protection system maintenance method based on state monitoring
CN105808902B (en) Qualitative method for analyzing operation condition of wet desulphurization system
CN113030633B (en) GA-BP neural network-based power distribution network fault big data analysis method and system
CN116797649A (en) Incineration treatment performance analysis method and system based on industrial big data
CN112580948A (en) Pollution treatment facility emission reduction potential assessment method
CN112001511A (en) Equipment reliability and dynamic risk evaluation method, system and equipment based on data mining
CN111523770A (en) Comprehensive evaluation method for improving water environment quality through water diversion and drainage
CN211575141U (en) Tank wagon loading VOC recovery processing system
CN108110283B (en) Flow battery energy storage system abnormal data identification method
CN113468813A (en) Desulfurization system inlet SO2Concentration prediction method and device and electronic equipment
CN112101821A (en) Method for evaluating cost index of ultralow emission system of thermal generator set
CN103559394A (en) Method for calculating desulfurization efficiency and commissioning rate of desulfurization facility of coal-fired unit
CN113869782B (en) Method for identifying environmental protection management and control abnormity based on time sequence decomposition of enterprise power consumption data

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant