CN-122024893-A - Boiler water quality prediction analysis method and system based on big data

CN122024893ACN 122024893 ACN122024893 ACN 122024893ACN-122024893-A

Abstract

The invention relates to the technical field of boiler water quality analysis, in particular to a boiler water quality prediction analysis method and a system based on big data, which acquire pollution discharge and dosing action signals, and (3) constructing a behavior numbering sequence, extracting a corresponding water quality measurement fragment, calibrating a variation trend, classifying response behaviors, synchronizing heat load data to form a linkage track, and summarizing trend to output a prediction sequence. The invention continuously records various operations according to time sequence by identifying pollution discharge and dosing signals and constructing a control behavior numbering sequence, extracts continuous measurement data segments of conductivity and dissolved oxygen by combining the sequences, presents water quality change and control actions in a correlated way, distinguishes water quality response types based on the corresponding relation between trend and behavior category, introduces thermal load change records and performs time connection to form a water quality change track with a load background, and constructs a water quality trend prediction sequence related to control behavior and load change by summarizing trend directions of various stages.

Inventors

CHEN GUILAN
CHEN SHUANGYANG
Shen Sunwenjing
JIN WEI
SHU ZELI

Assignees

南京海邦环保科技有限公司

Dates

Publication Date: 20260512
Application Date: 20260210

Claims (9)

1. The boiler water quality prediction analysis method based on big data is characterized by comprising the following steps of: S1, acquiring action signals of a pollution discharge control device and a dosing control device in the operation of a boiler, extracting the starting change of a pollution discharge valve and the change moment of a dosing driving state, corresponding each signal to actual behaviors, and numbering the behaviors in a classified manner according to time sequence to obtain a boiler control behavior numbering sequence; s2, extracting measurement contents of conductivity and dissolved oxygen from furnace water monitoring records by using time of each behavior number in the boiler control behavior number sequence, sorting relevant measured values before and after behaviors into time segments, and correlating water quality change data according to the behavior numbers to obtain a water quality change splicing segment set; S3, checking the data trend of the conductivity and the dissolved oxygen in a concentrated manner according to the water quality change splicing fragments, butting the part with the change trend with the corresponding control behavior, and linking each data segment to the behavior number and the action type to obtain a behavior response water quality change mapping table; And S4, extracting corresponding heat load data from the steam output record according to the aligned data segments in the behavior response water quality change mapping table, sequentially connecting the heat load change and the water quality segments by taking time points as clues, and carrying out unified processing in combination with behavior identification to obtain a heat load guiding jump linkage track group.
2. The method for predicting and analyzing boiler water quality based on big data according to claim 1, wherein the boiler control behavior numbering sequence comprises a numbering identifier, a behavior type label and a behavior time sequence record, the water quality change splicing fragment set comprises a conductivity change curve, a dissolved oxygen fluctuation section and a water quality measurement time axis, the behavior response water quality change mapping table comprises a behavior response type, a water quality change mode and a water quality fluctuation intensity classification, and the thermal load guiding jump linkage track group comprises a thermal load change section, a control behavior corresponding relation and a time linkage index.
3. The method for predicting and analyzing boiler water quality based on big data as set forth in claim 1, wherein the step of obtaining the boiler control behavior number sequence is as follows: S101, acquiring a motion signal of a pollution discharge control device and a motion signal of a dosing control device in the running process of a boiler, correspondingly acquiring time and signal states one by one, distinguishing the state change process according to signal sources, and extracting a time point when the state of a pollution discharge valve is changed and a time point when the state of the dosing motion is switched to obtain a switching moment of the pollution discharge motion and the dosing motion; S102, according to the switching moment of the pollution discharge and dosing actions, calling the corresponding time period content in the boiler behavior data, extracting the pollution discharge and dosing action content which is close to the time, and confirming the corresponding relation by judging whether the time is close to the time, so as to obtain an action time corresponding relation sequence; S103, calling the corresponding relation sequence of the action time, numbering the corresponding relation sequence of the action time in sequence according to time sequence, sorting the numbers and the corresponding pollution discharge or dosing actions according to time sequence, outputting the numbering sequence, and obtaining a boiler control action sequence number set.
4. The method for predicting and analyzing the boiler water quality based on big data according to claim 1, wherein the step of obtaining the water quality change spliced fragment set is as follows: s201, calling the conductivity and dissolved oxygen sequence recorded in the boiler water monitoring device at intervals of minutes by using the action time corresponding to each number in the boiler control action sequence numbering set, respectively taking fixed intervals forwards and backwards in time, arranging conductivity and dissolved oxygen data in time sequence, and numbering the conductivity and dissolved oxygen data to obtain a water quality sequence fragment set; s202, according to the water quality sequence fragment set, calling a numbering sequence in a boiler control action sequence numbering set, sorting the conductivity and dissolved oxygen data of the corresponding actions, sequentially juxtaposing the conductivity and the dissolved oxygen data into time fragments under the same number, and screening out data missing items to obtain a water quality fragment set corresponding to the actions; And S203, aiming at each group of time slices in the water quality slice group corresponding to the action, uniformly arranging according to the sequence of numbers, splicing the conductivity data and the dissolved oxygen data in time sequence, and taking each action number as a connection index position to obtain a water quality change splicing slice set.
5. The method for predicting and analyzing boiler water quality based on big data as set forth in claim 1, wherein the step of obtaining the behavior response water quality change mapping table comprises the steps of: S301, checking time sequences in sequence according to the conductivity and dissolved oxygen measurement content corresponding to each group of codes in the water quality change splicing fragment set, comparing the change directions between adjacent measurement results, identifying the time position of the first switching of continuous trend, recording the time and the corresponding number, and obtaining a change trend starting time set; S302, calling the initial time set of the variation trend, searching the action category corresponding to each number in the boiler control action sequence number set for the number item corresponding to each time position, correspondingly pairing each time with the action category, and importing the time and the action category into a list arranged according to the number sequence to obtain a action number corresponding category list; S303, classifying the change trend starting time corresponding to each action category into a unified table together with the number according to the action number corresponding category list, arranging the data positions according to the number sequence, classifying and sorting the corresponding trend time points according to the action categories, and obtaining the action response water quality change mapping table.
6. The method for predicting and analyzing the boiler water quality based on big data according to claim 1, wherein the step of obtaining the thermal load guiding jump linkage track group is as follows: S401, calling boiler load data in steam output record content according to data segments corresponding to each number in the behavior response water quality change mapping table, arranging the heat load data of each fixed interval before and after the action time according to time sequence, and importing the heat load data into a list according to the number sequence to obtain a time segment set corresponding to the heat load; s402, invoking the time segment set corresponding to the heat load, performing time alignment treatment on the heat load time segment in each numbered item and the water quality data in the water quality change spliced segment set, wherein an alignment mode adopts the starting time coincidence as a reference standard, and sequentially arranging the two time segment data according to time to obtain a heat load water quality combined segment set; S403, according to the heat load water quality combined segment groups, carrying out position pairing on the continuous jump positions of the heat load data in each group and the change trend in the water quality data, carrying out combined index processing on time points of the two types of data, and sequentially summarizing according to numbers to obtain a heat load guiding jump linkage track group.
7. The big data based boiler water quality prediction analysis method according to claim 1, characterized in that the method further comprises the step of S5: S5, calling data and control behaviors in different boiler operation stages in the heat load guiding jump linkage track group to be processed one by one, collecting the continuation direction of the water quality measurement trend, outputting the water quality measurement trend in a linkage way with the control behaviors, and indicating the change of the water quality state of the boiler by using a continuous data description mode to obtain a boiler water quality trend prediction sequence; the boiler water quality trend prediction sequence comprises a water quality change trend line, a water quality state continuation characteristic and a control behavior association tag.
8. The method for predicting and analyzing boiler water quality based on big data according to claim 7, wherein the step of obtaining the predicted sequence of boiler water quality trend is: S501, according to the water quality measurement trend of each numbered item in the heat load guiding jump linkage track group, extracting the continuous change direction of the conductivity and the dissolved oxygen, recording the continuous range of the trend direction in the time sequence, and obtaining a water quality trend continuous section set by corresponding numbers to control actions one by one; s502, calling the water quality trend continuation section set, attaching the control action category corresponding to each number to the rear side position corresponding to the water quality continuation direction to form a water quality trend and action combination sequence divided according to the number, and rectifying according to the number sequence to obtain a control action water quality corresponding sequence; And S503, extracting continuous time periods of the continuous direction of the water quality in each sequence according to the corresponding sequence of the water quality of the control behavior, classifying according to the category of the control action, and connecting all action corresponding trend segments in series to form a complete change direction path to obtain a boiler water quality trend prediction sequence.
9. A big data based boiler water quality prediction analysis system for use in the big data based boiler water quality prediction analysis method of any of claims 1-8, the system comprising: The control behavior extraction module is used for acquiring a valve state signal of the boiler pollution discharge control device and a driving signal of the dosing control device, positioning a change time point of each operation by identifying the starting time and the ending time of the state change, numbering the operation behaviors according to the signal sequence relationship, and corresponding the numbers to the control actions to obtain a boiler control behavior number sequence; The water quality data splicing module is used for calling each behavior number recorded in the boiler control behavior number sequence and a corresponding time point, extracting continuous measurement records of conductivity data and dissolved oxygen data from a boiler water monitoring device, and pairing measurement contents with the control behavior numbers according to a time relation to obtain a water quality change splicing fragment set; The water quality trend attribution module extracts a conductivity continuous variation sequence and a dissolved oxygen continuous variation sequence according to the measurement content in the water quality variation splicing fragment set, compares data segments with direction conversion in the two types of measurement data, identifies a variation starting time point and a trend continuing direction, and corresponds a corresponding variation trend to a control behavior number to obtain a behavior response water quality variation mapping table; The thermal load linkage module calls the serial number information of the control behaviors in the behavior response water quality change mapping table, calls the steam pressure value and the steam flow velocity value from the steam output related record, matches the data time period of the behavior corresponding to the serial number, merges the pressure and flow velocity change direction and the corresponding water quality change trend according to the serial number, and obtains a thermal load guiding jump linkage track group; And the trend prediction generation module extracts the conductivity fluctuation trend and the dissolved oxygen fluctuation trend under each number according to the water quality change trend and the load change direction which are merged in the heat load guiding jump linkage track group, compares the conductivity fluctuation trend with the dissolved oxygen fluctuation trend under each number, and intensively outputs the trend types under each number action in a trend direction attribution mode to obtain a boiler water quality trend prediction sequence.

Description

Boiler water quality prediction analysis method and system based on big data Technical Field The invention relates to the technical field of boiler water quality analysis, in particular to a boiler water quality prediction analysis method and system based on big data. Background The technical field of boiler water quality analysis relates to sampling, detection and analysis of boiler operation water, and core matters comprise water sample collection, a detection method of ions and impurities in water, a data recording and processing mode and a water quality adjusting flow based on detection results. The method is widely applied to industries such as electric power, chemical industry and the like, and aims to prevent scaling, corrosion and other faults caused by water quality problems and ensure equipment operation safety and efficiency. The traditional boiler water quality prediction analysis method and system refer to a mode of predicting a water quality change trend by adopting empirical judgment, statistical analysis or regression calculation based on historical water quality data and operation parameters. The general flow comprises the steps of regularly collecting a water sample, detecting components in the water in a laboratory, summarizing historical and current data, setting an early warning threshold value, and comparing and judging water quality change by using a simple model. The correlation identification among the multiple variables is weak, and the method is difficult to adapt to the prediction requirement under the complex operation condition. The existing method relies on interval sampling and static detection, and is difficult to capture continuous change of water quality in a short period in boiler operation, so that state monitoring lacks key nodes. The trend is judged through static data review, an identification mechanism for water quality response caused by operation behaviors is lacking, and trigger factors of occurrence of change cannot be definitely determined. In a scene of frequent load fluctuation, fluctuation of parameters such as conductivity, dissolved oxygen and the like is easy to be misjudged, the system response is not timely, adjustment errors or early warning lag are easy to be caused, and the accuracy and timeliness of water quality regulation are reduced. Disclosure of Invention In order to solve the technical problems in the prior art, the embodiment of the invention provides a boiler water quality prediction analysis method and system based on big data. The technical scheme is as follows: a boiler water quality prediction analysis method based on big data comprises the following steps: S1, acquiring action signals of a pollution discharge control device and a dosing control device in the operation of a boiler, extracting the starting change of a pollution discharge valve and the change moment of a dosing driving state, corresponding each signal to actual behaviors, and numbering the behaviors in a classified manner according to time sequence to obtain a boiler control behavior numbering sequence; s2, extracting measurement contents of conductivity and dissolved oxygen from furnace water monitoring records by using time of each behavior number in the boiler control behavior number sequence, sorting relevant measured values before and after behaviors into time segments, and correlating water quality change data according to the behavior numbers to obtain a water quality change splicing segment set; S3, checking the data trend of the conductivity and the dissolved oxygen in a concentrated manner according to the water quality change splicing fragments, butting the part with the change trend with the corresponding control behavior, and linking each data segment to the behavior number and the action type to obtain a behavior response water quality change mapping table; s4, extracting corresponding heat load data from the steam output record according to the aligned data segments in the behavior response water quality change mapping table, sequentially connecting the heat load change and the water quality segments by taking time points as clues, and carrying out unified processing by combining with behavior identification to obtain a heat load guiding jump linkage track group; and S5, calling the data and control behaviors in different boiler operation stages in the heat load guiding jump linkage track group to process one by one, collecting the continuation direction of the water quality measurement trend, outputting in linkage with the control behavior, and indicating the change of the boiler water quality state by using a continuous data description mode to obtain a boiler water quality trend prediction sequence. As a further scheme of the invention, the boiler control behavior numbering sequence comprises a numbering identifier, a behavior type label and a behavior time sequence record, the water quality change splicing fragment set comprises a conductivity change curv