Intelligent backwashing method and device for water treatment equipment based on ultrafiltration membrane

Abstract

The application discloses an intelligent backwashing method and device for a water treatment equipment based on an ultrafiltration membrane, relates to the technical field of washing methods, and comprises the following steps: S1, trend analysis is conducted on a historical membrane pollution index sequence, and a rapid water production stage and a stable water production stage of each filtration water production process are obtained through division; S2, trend analysis is conducted on an online membrane pollution index sequence, and a trend slope of an online stable water production stage is obtained; and S3, the trend slope of the online stable water production stage is compared with an online backwashing slope threshold value, and current water production process execution or maintenance is controlled according to the execution label. Through analysis of each water production stage of offline historical data, the application provides a parameterized description of a trend division basis for a rapid water production stage and a stable water production stage from a data driving angle, and provides a theoretical basis for an online backwashing autonomous decision-making mechanism of a water production process of an ultrafiltration membrane source water filtration system.

Description

Technical Field

[0001] This invention relates to the field of rinsing methods, and in particular to an intelligent backwashing method and apparatus for water treatment equipment based on ultrafiltration membranes. Background Technology

[0002] Process industries are the cornerstone of my country's real economy, boasting large-scale product output and high industrial linkages, occupying a pivotal position in the national economy. Among them, ultrafiltration membrane filtration-based raw water treatment equipment, as a typical process-type municipal infrastructure, has broader social demand and application scenarios compared to other process industry facilities. The informatization and networking technologies of the production process provide an effective implementation path for the intelligent transformation of process-type municipal infrastructure. Smart water systems based on edge computing and cloud computing collect pre-treatment data from the raw water filtration process of various water plants, upload it to the cloud for aggregation, and establish a big data real-time monitoring platform containing operational information of equipment in each water plant. This enables centralized real-time monitoring and maintenance information dissemination for the operation of equipment in different water plants, thereby ensuring water production and extending the service life of ultrafiltration membranes. However, existing ultrafiltration membrane-based raw water filtration systems have the following limitations in equipment maintenance:

[0003] (1) Equipment maintenance relies on alarms for abnormal information and the issuance of maintenance orders. Existing smart water systems primarily rely on retrieving real-time or historical data from abnormal subsystems of each water plant to monitor its operational status. Engineers then determine and intervene in water production and backwashing timing, and issue equipment maintenance orders. The system's autonomous decision-making capability needs improvement.

[0004] (2) The backwashing maintenance strategy for ultrafiltration membrane filtration water production is relatively simple. Existing ultrafiltration membrane filtration water production requires backwashing to reduce the fouling degree of the ultrafiltration membrane. However, most of these backwashing strategies adopt timed backwashing mechanisms, ignoring the differences in ultrafiltration membrane fouling degree caused by factors such as source water quality, seasonal factors, and environmental differences. Since water cannot be produced during backwashing, the timed backwashing strategy cannot achieve an optimal balance between water production and membrane lifespan;

[0005] (3) The criteria for judging the degree of ultrafiltration membrane fouling are relatively singular. Due to the influence of data measurement conditions and sensor accuracy, the membrane fouling index of existing ultrafiltration membranes is mainly approximated by the magnitude of the permeate pressure under the siphon principle. Since the pressure data fluctuates greatly, it is not easy to statistically summarize and generalize the data characteristics and trends of a single normal permeate stage, which poses a challenge to the intelligent backwashing maintenance strategy of ultrafiltration membranes.

[0006] In summary, existing smart water systems based on ultrafiltration membranes, relying solely on timed backwashing strategies based on permeate pressure data, cannot guarantee the lifespan of ultrafiltration membranes while maintaining maximum permeate flow. Therefore, to further meet the intelligent transformation needs of process-oriented municipal infrastructure, exemplified by ultrafiltration source water filtration systems, it is crucial to mine permeate process data, capture trends in ultrafiltration membrane fouling, and develop backwashing maintenance strategies with autonomous decision-making mechanisms. Summary of the Invention

[0007] The purpose of this invention is to provide an intelligent backwashing method and apparatus for water treatment equipment based on ultrafiltration membranes, so as to solve the problems mentioned in the background art.

[0008] To achieve the above objectives, the present invention provides the following technical solution: an intelligent backwashing method for water treatment equipment based on ultrafiltration membranes, comprising the following steps:

[0009] S1. Obtain offline historical data of the ultrafiltration membrane source water filtration system and preprocess it to obtain a historical membrane fouling index sequence; perform trend analysis on the historical membrane fouling index sequence to divide each filtration water production process into a rapid water production stage and a stable water production stage; based on the trend slope analysis of the stable water production stage, formulate the online backwashing slope threshold of the ultrafiltration membrane.

[0010] S2. Real-time acquisition of online operating data of the ultrafiltration membrane source water filtration system, and preprocessing to obtain the online membrane fouling index sequence within the cumulative window; trend analysis of the online membrane fouling index sequence to obtain the trend slope of the online stable water production stage;

[0011] S3. Compare the trend slope of the online stable water production stage with the online backwash slope threshold, generate an ultrafiltration membrane autonomous backwash execution tag based on the comparison result, and control the execution or maintenance of the current water production process based on the execution tag.

[0012] Preferably, the step S1 of acquiring offline historical data and performing preprocessing specifically includes:

[0013] Collect multiple sets of historical data, including process status, cumulative water production, and deep water pressure of the produced water, to form a historical analysis dataset;

[0014] Data from the normal water production stage were filtered out, and the dataset was divided into multiple subsets of normal water production process data based on the sequence numbers of the start and end sample points of each normal water production process.

[0015] For each subset, the corresponding historical membrane fouling index is calculated based on the current permeate pressure, the initial permeate pressure, and the cumulative increase in permeate volume, and a historical membrane fouling index sequence is formed.

[0016] Preferably, step S1, which involves performing trend analysis on the historical membrane fouling index sequence to divide it into a rapid water production stage and a stable water production stage, specifically includes:

[0017] The historical membrane fouling index sequence is subjected to noise immunity processing to obtain an immunity membrane fouling index sequence;

[0018] The anti-interference membrane fouling index sequence is fitted by constructing input and output window data of cumulative sequence number and establishing a quadratic regression model using the least squares method.

[0019] Based on the changing trends of the fitting results, the data segments with rapid downward trends corresponding to the rapid water production stage and the data segments with slow downward trends corresponding to the stable water production stage are identified and divided.

[0020] Preferably, the step S1, which involves determining the online backwash slope threshold based on the trend slope analysis of the stable water production stage, specifically includes:

[0021] For the slow downward trend data segment of the defined stable water production stage, construct the input and output window data for first-order linear regression;

[0022] A first-order regression model was established using the least squares method, and the trend slope of the stable water production stage was calculated.

[0023] Based on process requirements or expert knowledge, the online backwashing slope threshold of the ultrafiltration membrane is set according to the calculated trend slope.

[0024] Preferably, the step S2 of acquiring and preprocessing online operational data specifically includes:

[0025] Real-time data collection of online water production volume, water production pressure, and cumulative increase in water production during normal water production stages;

[0026] The online membrane fouling index is calculated in real time based on the current production water pressure, initial production water pressure, and cumulative increase in production water volume.

[0027] The online membrane fouling indices calculated within the set time window are used to construct an online membrane fouling index sequence.

[0028] Preferably, the trend analysis of the online membrane fouling index sequence in step S2 specifically includes:

[0029] The online membrane fouling index sequence is subjected to noise immunity processing to obtain an online immunity membrane fouling index sequence.

[0030] Using the same method as offline analysis, a quadratic regression fitting was performed on the online anti-fouling membrane fouling index sequence to distinguish between the online rapid water production stage and the online stable water production stage;

[0031] First-order linear regression analysis was performed on the data of the defined online stable water production stage to calculate the real-time trend slope of the online stable water production stage.

[0032] Preferably, the step S3, which involves generating an ultrafiltration membrane self-backwashing execution label based on the comparison results, specifically includes:

[0033] When the trend slope of the online stable water production stage is less than the online backwash slope threshold, it is determined that the backwash condition has not been met, a backwash execution tag in the off state is generated, and the water production process continues.

[0034] When the trend slope of the online stable water production stage is greater than or equal to the online backwash slope threshold, it is determined that the backwash condition has been met, and a backwash execution tag in the open state is generated.

[0035] Preferably, after performing the backwashing operation based on the execution tag in the open state, the method further includes:

[0036] Save the relevant runtime data in the current cumulative window to the database or cloud;

[0037] Clear the online membrane fouling index data in the current window;

[0038] The backwash execution label is turned off, and preparations are made for monitoring and determining the next water production process.

[0039] Preferably, the noise immunity processing involves superimposing a Gaussian white noise sequence with the same dimension onto the original membrane fouling index sequence.

[0040] A smart backwashing device for water treatment equipment based on ultrafiltration membrane, the device comprising:

[0041] The offline data acquisition and preprocessing module is used to acquire and preprocess offline historical data of the ultrafiltration membrane source water filtration system to obtain a historical membrane fouling index sequence.

[0042] The offline membrane fouling index analysis module is used to perform trend analysis on the historical membrane fouling index sequence and divide it into rapid water production stage and stable water production stage.

[0043] The offline stable water production analysis module is used to perform trend slope analysis based on the defined stable water production stages and to determine the online backwashing slope threshold for the ultrafiltration membrane.

[0044] The online data acquisition and preprocessing module is used to acquire and preprocess the online operating data of the ultrafiltration membrane source water filtration system in real time to obtain the online membrane fouling index sequence within the cumulative window.

[0045] The online membrane fouling index analysis module is used to perform trend analysis on the online membrane fouling index sequence to obtain the trend slope of the online stable water production stage;

[0046] The online backwashing condition determination and execution module is used to compare the trend slope of the online stable water production stage with the online backwashing slope threshold, and generate and execute the ultrafiltration membrane autonomous backwashing command.

[0047] The technical effects and advantages of this invention are as follows:

[0048] (1) By analyzing the water production stages of offline historical data, from the perspective of data-driven analysis, we can explore the parameterized description of the trend division of the rapid water production stage and the stable water production stage in the water production process, which provides a theoretical basis for the online backwashing autonomous decision-making mechanism of the water production process of the ultrafiltration membrane source water filtration system.

[0049] (2) By calculating the membrane fouling index, the single-parameter judgment mode of using only the product water pressure as the membrane fouling process under actual operation is changed. Moreover, the anti-disturbance membrane fouling index under Gaussian white noise avoids the interference of measurement data fluctuations caused by sensor reasons on the slope of the stable product water stage.

[0050] (3) By uploading the cumulative window data under the autonomous decision-making mechanism, a data source is provided for the membrane information and intelligent database of the ultrafiltration water production process. This is not only beneficial to the subsequent maintenance of the ultrafiltration membrane water production equipment, but also provides an effective reference for the intelligent collaborative transformation of source water pretreatment and other processes. Attached Figure Description

[0051] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention, but do not constitute a limitation thereof. In the drawings:

[0052] Figure 1 This is a flowchart illustrating the judgment logic of the backwashing method of the present invention;

[0053] Figure 2 This is a schematic diagram illustrating the trend analysis of the rapid water production stage and the stable water production stage in the offline water production process of the present invention;

[0054] Figure 3 This is a schematic diagram of the backwashing device of the present invention. Detailed Implementation

[0055] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0056] This invention provides, for example Figures 1-3The method shown is an intelligent backwashing method for water treatment equipment based on ultrafiltration membrane, the method comprising: collecting... Historical data of the ultrafiltration membrane source water filtration process for each group, with variables for each group including: 1. 1. Process status; 2. 3. Cumulative water production; : Production water pressure, obtained Each dimension is Dimensional historical analysis dataset;

[0057] Then, the historical analysis dataset is preprocessed, including: filtering process state variable information and retaining variables under normal water production stages. , Specifically: record the sequence numbers of the starting and ending sample points for each normal water production stage in the historical analysis dataset. This constitutes a sequence numbering matrix for the normal water production stage intervals: ,in, The first The sequence numbers of the starting and ending sample points in the historical analysis data for each normal water production process. The total number of normal water production processes in the historical analysis dataset; based on the numbering matrix. The historical analysis dataset is divided into For each subset of normal water production process data, the membrane fouling index is calculated.

[0058]

[0059] in, For subset Chinese serial number Membrane fouling index, For subset Chinese serial number Water production pressure, For subset The initial production water pressure in the middle, For subset The cumulative increase in water production;

[0060] Furthermore, by The sequence of membrane fouling indices that constitutes each stage of filtration permeate: ;

[0061] Subsequently, a trend analysis was conducted on the membrane fouling index at each stage, including: To conduct noise immunity bombardment, specifically: constructing and Gaussian white noise sequences of the same dimension Calculate the anti-interference membrane fouling index

[0062]

[0063] Furthermore, the rapid water production phase of each filtration process is obtained, specifically by capturing the phase using a quadratic baseline trend curve. The rapid downward trend is further illustrated by constructing the cumulative sequence number input and output window data: ,

[0064]

[0065]

[0066] in, To calculate the cumulative number of sequence numbers, use , Establish a least squares regression model and obtain the fitted values ​​of the output window. ,in Based on The least squares regression parameters,

[0067] Furthermore, utilizing determination The rapid downward trend, specifically: when In hour( This is the last estimate of the cumulative window quadratic fitting result. The reciprocal of the cumulative window quadratic fitting result If the fitted result shows an upward trend (with one estimated value), then it is determined that... The rapid decline (corresponding to the rapid water production stage) trend data is as follows: , The slow decline (corresponding to the stable water production stage) trend data is as follows Otherwise To proceed to the next round , The cumulative window size and the determination of a rapid downward trend;

[0068] Then, a trend slope analysis was performed on the stable water production stage, including: To capture the slope of the first-order baseline trend curve, specifically: construct input and output window data. ,

[0069]

[0070]

[0071] Furthermore, utilizing , Establish a least squares regression model to obtain Zhongyu Corresponding slope Based on the duration / trend of each stage of the stable water production process and the process requirements / expert system, an online backwashing slope threshold for the ultrafiltration membrane was determined. ;

[0072] Next, online data from the ultrafiltration membrane source water filtration system was acquired and preprocessed, including: screening process state variable information and utilizing data from the normal water production stage. Real-time online water production Water production pressure Cumulative increase in water production and initial pressure Calculate the online membrane fouling index

[0073]

[0074] Furthermore, a sequence of online membrane fouling indices is constructed within the cumulative window of the filtration permeate stage: ,right To conduct noise immunity bombardment, specifically: constructing and Gaussian white noise sequences of the same dimension Calculate the cumulative window of online anti-interference membrane pollution index.

[0075]

[0076] Then, a trend analysis was performed on the online anti-fouling membrane fouling index under the cumulative window, including: trend slope analysis during the stable water production stage; specifically: obtaining the rapid water production stage of the filtration process under the cumulative window; more specifically: constructing cumulative sequence number input and output window data. ,

[0077]

[0078]

[0079] Among them, utilizing , Establish a least squares regression model and obtain the fitted values ​​of the output window. ,in These are the least squares regression parameters;

[0080] Furthermore, utilizing determination The rapid downward trend, specifically: when In If the fitting result shows an upward trend, then it is determined that... The rapid decline (corresponding to the online rapid water production stage) trend data is as follows: , The slow decline trend data (corresponding to the online stable water production stage) is as follows: ;

[0081] Going a step further, to To capture the slope of the first-order baseline trend curve, specifically: construct input and output window data. ,

[0082]

[0083]

[0084] use , Establish a least squares regression model to obtain Zhongyu Corresponding slope ;

[0085] Next, the trends during the online stable water production stage were compared with the established online backwashing slope threshold for the ultrafiltration membrane, including: comparison , The value, specifically: when If the current online stable water production stage does not meet the backwashing conditions, the ultrafiltration membrane backwashing execution tag remains closed, the water production process continues, and the accumulated window parameters are set to... Simultaneously, perform the next round of slope determination;

[0086] Finally, the online backwashing operation of the ultrafiltration membrane is performed according to the ultrafiltration membrane self-backwashing execution label, including: comparison. , The value, when When the current online stable water production stage is determined to be ready for backwashing, the ultrafiltration membrane backwashing execution tag is turned on, and the control system automatically executes the ultrafiltration membrane backwashing operation.

[0087] Furthermore, the online membrane fouling index data under the cumulative window is processed, including: [processing the data within the cumulative window]. , Once the data is saved to the database / cloud, clear the online anti-interference membrane fouling index in the current window. After the backwashing is completed, the backwashing execution label is turned off, and the trend slope of the stable water production stage of the next water production process is determined.

[0088] like Figure 2 The curve changes shown fully replicate the core judgment process of offline membrane fouling index trend analysis, transforming the abstract mathematical fitting and stage division steps into intuitive image changes. In the image presentation, the original anti-fouling membrane fouling index curve initially shows a rapid downward trend, corresponding to the rapid water production stage in the early stage of ultrafiltration membrane water production when the membrane pores are not heavily clogged and the water production efficiency is relatively high. Subsequently, the curve turns to a gentle downward trend, corresponding to the stable water production stage where the membrane pores are gradually clogged and the water production efficiency tends to stabilize. By constructing a quadratic least squares regression model based on the cumulative sequence number and the anti-fouling membrane fouling index, the technicians obtain a fitting curve that shows an inflection point of decline followed by an increase. When the last estimated value of the fitting curve is greater than the second to last estimated value, this inflection point is the stage division critical point. The image clearly distinguishes the rapid water production stage and the stable water production stage on the left and right sides using this critical point as the boundary.

[0089] A smart backwashing device for water treatment equipment based on ultrafiltration membrane, the device comprising:

[0090] The offline data acquisition and preprocessing module is used to acquire and preprocess offline historical data of the ultrafiltration membrane source water filtration system to obtain a historical membrane fouling index sequence.

[0091] The offline membrane fouling index analysis module is used to perform trend analysis on the historical membrane fouling index sequence and divide it into rapid water production stage and stable water production stage.

[0092] The offline stable water production analysis module is used to perform trend slope analysis based on the defined stable water production stages and to determine the online backwashing slope threshold for the ultrafiltration membrane.

[0093] The online data acquisition and preprocessing module is used to acquire and preprocess the online operating data of the ultrafiltration membrane source water filtration system in real time to obtain the online membrane fouling index sequence within the cumulative window.

[0094] The online membrane fouling index analysis module is used to perform trend analysis on the online membrane fouling index sequence to obtain the trend slope of the online stable water production stage;

[0095] The online backwashing condition determination and execution module is used to compare the trend slope of the online stable water production stage with the online backwashing slope threshold, and generate and execute the ultrafiltration membrane autonomous backwashing command.

[0096] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A smart backwashing method for water treatment equipment based on ultrafiltration membrane, characterized in that, Includes the following steps: S1. Obtain offline historical data of the ultrafiltration membrane source water filtration system and preprocess it to obtain a historical membrane fouling index sequence; perform trend analysis on the historical membrane fouling index sequence to divide the rapid water production stage and stable water production stage of each filtration water production process. Based on the trend slope analysis of the stable water production stage, the online backwashing slope threshold of the ultrafiltration membrane is determined. S2. Real-time acquisition of online operating data of the ultrafiltration membrane source water filtration system, and preprocessing to obtain the online membrane fouling index sequence within the cumulative window; trend analysis of the online membrane fouling index sequence to obtain the trend slope of the online stable water production stage; S3. Compare the trend slope of the online stable water production stage with the online backwash slope threshold, generate an ultrafiltration membrane autonomous backwash execution tag based on the comparison result, and control the execution or maintenance of the current water production process based on the execution tag. Step S2 involves acquiring and preprocessing online operational data, specifically including: Real-time data collection of online water production volume, water production pressure, and cumulative increase in water production during normal water production stages; The online membrane fouling index is calculated in real time based on the current production water pressure, initial production water pressure, and cumulative increase in production water volume. The online membrane fouling indices calculated within a set time window are used to construct an online membrane fouling index sequence. Step S2 involves trend analysis of the online membrane fouling index sequence, specifically including: The online membrane fouling index sequence is subjected to noise immunity processing to obtain an online immunity membrane fouling index sequence. Using the same method as offline analysis, a quadratic regression fitting was performed on the online anti-fouling membrane fouling index sequence to distinguish between the online rapid water production stage and the online stable water production stage; First-order linear regression analysis was performed on the data of the defined online stable water production stage to calculate the real-time trend slope of the online stable water production stage; Step S3, which involves generating an ultrafiltration membrane self-backwashing execution tag based on the comparison results, specifically includes: When the trend slope of the online stable water production stage is less than the online backwash slope threshold, it is determined that the backwash condition has not been met, a backwash execution tag in the off state is generated, and the water production process continues. When the trend slope of the online stable water production stage is greater than or equal to the online backwash slope threshold, it is determined that the backwash condition has been met, and a backwash execution tag in the open state is generated. After performing the backwash operation based on the execution tag indicating the open state, the process also includes: Save the relevant runtime data in the current cumulative window to the database or cloud; Clear the online membrane fouling index data in the current window; The backwash execution label is turned off, and preparations are made for monitoring and determining the next water production process.

2. The intelligent backwashing method for water treatment equipment based on ultrafiltration membrane according to claim 1, characterized in that, Step S1 involves acquiring offline historical data and preprocessing it, specifically including: Collect multiple sets of historical data, including process status, cumulative water production, and deep water pressure of the produced water, to form a historical analysis dataset; Data from the normal water production stage were filtered out, and the dataset was divided into multiple subsets of normal water production process data based on the sequence numbers of the start and end sample points of each normal water production process. For each subset, the corresponding historical membrane fouling index is calculated based on the current permeate pressure, the initial permeate pressure, and the cumulative increase in permeate volume, and a historical membrane fouling index sequence is formed.

3. The intelligent backwashing method for water treatment equipment based on ultrafiltration membrane according to claim 2, characterized in that, Step S1 involves trend analysis of the historical membrane fouling index sequence to divide it into rapid water production phases and stable water production phases, specifically including: The historical membrane fouling index sequence is subjected to noise immunity processing to obtain an immunity membrane fouling index sequence; The anti-interference membrane fouling index sequence is fitted by constructing input and output window data of cumulative sequence number and establishing a quadratic regression model using the least squares method. Based on the changing trends of the fitting results, the data segments with rapid downward trends corresponding to the rapid water production stage and the data segments with slow downward trends corresponding to the stable water production stage are identified and divided.

4. The intelligent backwashing method for water treatment equipment based on ultrafiltration membrane according to claim 3, characterized in that, Step S1 involves determining the online backwash slope threshold based on trend slope analysis during the stable water production phase, specifically including: For the slow downward trend data segment of the defined stable water production stage, construct the input and output window data for first-order linear regression; A first-order regression model was established using the least squares method, and the trend slope of the stable water production stage was calculated. Based on process requirements or expert knowledge, the online backwashing slope threshold of the ultrafiltration membrane is set according to the calculated trend slope.

5. The intelligent backwashing method for water treatment equipment based on ultrafiltration membrane according to claim 4, characterized in that, The noise immunity process involves superimposing a Gaussian white noise sequence of the same dimension onto the original membrane fouling index sequence.

6. An intelligent backwashing device for water treatment equipment based on ultrafiltration membrane, characterized in that, The apparatus for implementing the intelligent backwashing method as described in any one of claims 1-5 comprises: The offline data acquisition and preprocessing module is used to acquire and preprocess offline historical data of the ultrafiltration membrane source water filtration system to obtain a historical membrane fouling index sequence. The offline membrane fouling index analysis module is used to perform trend analysis on the historical membrane fouling index sequence and divide it into rapid water production stage and stable water production stage. The offline stable water production analysis module is used to perform trend slope analysis based on the defined stable water production stages and to determine the online backwashing slope threshold for the ultrafiltration membrane. The online data acquisition and preprocessing module is used to acquire and preprocess the online operating data of the ultrafiltration membrane source water filtration system in real time to obtain the online membrane fouling index sequence within the cumulative window. The online membrane fouling index analysis module is used to perform trend analysis on the online membrane fouling index sequence to obtain the trend slope of the online stable water production stage; The online backwashing condition determination and execution module is used to compare the trend slope of the online stable water production stage with the online backwashing slope threshold, and generate and execute the ultrafiltration membrane autonomous backwashing command.

Images