Intelligent decision method and system for cleaning of RO membrane for reuse of printing and dyeing wastewater

Abstract

The application provides a printing and dyeing wastewater recycling RO membrane cleaning intelligent decision method and system, which comprises the following steps: obtaining membrane performance parameters and various pollutant concentration parameters in real time, combining pollution characteristic factors and historical data models, calculating pollution dominant indexes of various pollutants, identifying dominant pollutants and evaluating the synergistic pollution effect of the pollutants; based on the expected removal efficiency of the candidate cleaning scheme on the dominant pollutants, the pollution dominant index and the synergistic pollution relief ability, the cleaning decision score of each scheme is calculated; when the highest score exceeds the preset cleaning starting threshold, the corresponding cleaning scheme is selected, and the process parameters are determined according to the dominant pollutants and the synergistic pollution index; after cleaning, the pollution characteristic factors and the cleaning threshold are adjusted according to the feedback, and the subsequent decision is continuously optimized.

Description

Technical Field

[0001] This application belongs to the field of intelligent decision-making, and in particular relates to an intelligent decision-making method and system for cleaning RO membranes used in the reuse of dyeing and printing wastewater. Background Technology

[0002] Dyeing and printing wastewater contains a large number of pollutants, which can be filtered out using RO membranes. The state of the RO membrane directly affects the filtration efficiency. Existing technologies mostly employ timed, pre-set cleaning programs, which are inherently delayed, intervening only when membrane fouling has become severe. Furthermore, these methods cannot identify the specific types of pollutants causing the fouling or their relative importance, potentially leading to poor cleaning results, excessive use of chemicals, unnecessary downtime, and potential damage to the membrane material. While methods combining influent water quality or establishing experience-based expert systems often remain at a qualitative or semi-quantitative level, they struggle to calculate the contribution of different pollutants to the current state of membrane performance deterioration. Membrane fouling is not the result of a single pollutant, but rather a process of interaction and mutual promotion among multiple pollutants—a synergistic fouling effect. For example, organic matter can provide nutrients for microorganisms, while microbial metabolites can promote the crystallization of inorganic salts into scale. Existing decision-making methods ignore this synergistic effect, resulting in cleaning strategies that are often ineffective in addressing the root cause of the fouling. Furthermore, existing decision-making models and parameters, such as the fouling coefficient or cleaning start threshold, lack the ability to be adjusted and optimized based on feedback from actual cleaning results. They cannot cope with fluctuations in influent water quality and changes in the membrane system's state, making it difficult to achieve optimal long-term operation. Summary of the Invention

[0003] This invention proposes an intelligent decision-making method for cleaning RO membranes used in the reuse of dyeing and printing wastewater. This method addresses the problem that existing decision models and parameters lack the ability to adjust and optimize based on feedback from actual cleaning results, and are unable to cope with fluctuations in influent water quality and long-term changes in the membrane system's state. The method includes: Obtain membrane performance parameters representing the current operating status of the reverse osmosis membrane system, as well as various pollutant concentration parameters representing the feed water quality; based on the pollutant concentration parameters, and combined with pollution characteristic factors representing the pollution tendency of each pollutant and a correlation model derived from historical data, calculate the contribution of each pollutant to the current deterioration of membrane performance parameters, and obtain the pollution dominance index of each pollutant. A set of dominant pollutants is identified based on the pollution dominance index, and a synergistic pollution index representing the synergistic pollution effect of a specific combination of pollutants in the set of dominant pollutants is calculated; for multiple candidate cleaning schemes, a cleaning decision score is calculated for each candidate cleaning scheme based on the expected removal efficiency of each scheme for the dominant pollutants, the pollution dominance index, the expected mitigation efficiency for synergistic pollution, and the synergistic pollution index. When the maximum value of the cleaning decision score among the candidate cleaning schemes exceeds the preset cleaning start threshold, the cleaning scheme corresponding to the maximum value is selected, and the specific process parameters of the scheme are determined based on the dominant pollutant and co-contamination index that contribute to the maximum score. Based on the dominant pollutants identified during this cleaning process and feedback on the cleaning effect, the pollution characteristic factors used to calculate the pollution dominance index and the cleaning initiation threshold are adjusted to optimize subsequent decision-making.

[0004] Furthermore, this invention also relates to an intelligent decision-making system for cleaning RO membranes used in the reuse of dyeing and printing wastewater, comprising the following modules: The module is used to acquire membrane performance parameters representing the current operating status of the reverse osmosis membrane system, as well as various pollutant concentration parameters representing the feed water quality. Based on the pollutant concentration parameters, and combined with pollution characteristic factors representing the pollution tendency of each pollutant and a correlation model derived from historical data, the contribution of each pollutant to the current deterioration of membrane performance parameters is calculated, and the pollution dominance index of each pollutant is obtained. The first calculation module is used to identify a group of dominant pollutants based on the pollution dominance index, and to calculate a synergistic pollution index representing the synergistic pollution effect of a specific combination of pollutants in the group of dominant pollutants; and to calculate a cleaning decision score for each candidate cleaning scheme based on the expected removal efficiency of each scheme for the dominant pollutants, the pollution dominance index, the expected mitigation efficiency for synergistic pollution, and the synergistic pollution index. The second calculation module is used to select the cleaning scheme corresponding to the maximum value when the maximum value of the cleaning decision score of each candidate cleaning scheme exceeds the preset cleaning start threshold, and to determine the specific process parameters of the scheme based on the dominant pollutant and co-contamination index that contribute to the maximum score. The adjustment module is used to adjust the pollution characteristic factors used to calculate the pollution dominance index and the cleaning initiation threshold based on the dominant pollutants identified in this cleaning and the feedback on the cleaning effect, thereby optimizing subsequent decisions.

[0005] This invention identifies the dominant contaminants causing the current fouling state by quantitatively calculating the contribution of each contaminant to membrane performance degradation. It also demonstrates the synergistic fouling effect among multiple contaminants, thus revealing the mechanism of membrane fouling. Based on the identification of the dominant and synergistic contaminants, a scoring mechanism selects a cleaning scheme from multiple candidate schemes and determines optimized process parameters for the selected scheme. Furthermore, by utilizing a feedback correction mechanism for cleaning effectiveness, the key parameters of the decision-making model can be adjusted according to actual operating results, ensuring the accuracy of the decision-making system during long-term operation. Therefore, it improves the scientific nature of cleaning decisions, achieves better maintenance of reverse osmosis membranes, and reduces operating costs. Attached Figure Description

[0006] Figure 1 A flowchart of the first embodiment; Figure 2 This is a schematic diagram of data acquisition and processing; Figure 3 A schematic diagram of the dominant pollution indices for each pollutant; Figure 4 A diagram illustrating the decision scores for each candidate cleaning scheme; Figure 5 To adjust the closed-loop diagram. Detailed Implementation

[0007] The features and exemplary embodiments of various aspects of this application will be described in detail below. To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain this application and not to limit it. For those skilled in the art, this application can be implemented without some of these specific details. The following description of the embodiments is merely to provide a better understanding of this application by illustrating examples.

[0008] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes said element.

[0009] See the first embodiment. Figure 1 This invention proposes an intelligent decision-making method for cleaning RO membranes used in the reuse of dyeing and printing wastewater, comprising: S1. Obtain membrane performance parameters representing the current operating status of the reverse osmosis membrane system, and various pollutant concentration parameters representing the feed water quality; based on the pollutant concentration parameters, and combined with pollution characteristic factors representing the pollution tendency of each pollutant and a correlation model derived from historical data, calculate the contribution of each pollutant to the current deterioration of membrane performance parameters, and obtain the pollution dominance index of each pollutant. The reverse osmosis system utilizes an online monitoring and control system, namely the SCADA system, to collect real-time data on feed water pressure, permeate flow rate, concentrate flow rate, feed water conductivity, permeate conductivity, and water temperature. Based on this data, standardized permeate flux, desalination rate, and standardized pressure difference are calculated as membrane performance parameters. Simultaneously, online water quality analyzers acquire data on feed water turbidity, total organic carbon (TOC), and pH. Combined with regular offline water sample analysis, this yields concentration parameters for various pollutants in the feed water, including calcium ions, magnesium ions, silica, iron ions, and adenosine triphosphate (ATP), which indicates microbial content.

[0010] The pollution characteristic factors are preset weighting coefficients for each pollutant; for example, 0.8 is set for silica, which is prone to fouling, and 0.4 is set for organic matter. The correlation model is a machine learning model trained based on historical operating data, used to establish a quantitative relationship between pollutant concentration and membrane performance degradation rate, and outputs the dynamic weight of each pollutant to the current degradation state. For example, a gradient boosting decision tree model is used, which takes multiple pollutant concentrations and key operating parameters as input features and the membrane performance degradation rate over a future period as the prediction target for training. When a decision needs to be made, the current pollutant concentration parameter is input into the trained model. The model not only outputs a predicted value for the performance degradation rate but also calculates the contribution of each input feature to the prediction result through its internal mechanism. Subsequently, the contribution of a single pollutant is normalized to obtain the pollution dominance index of that pollutant.

[0011] In an optional embodiment, obtaining membrane performance parameters representing the current operating status of the reverse osmosis membrane system, and various pollutant concentration parameters representing the feed water quality, includes: The normalized permeate flux, transmembrane pressure difference, and desalination rate are collected in real time by online sensors and used as the membrane performance parameters. Suspended solids (SS), chemical oxygen demand (COD), and silica (Silica) in the influent are measured periodically using water quality analyzers. calcium ions The concentrations of total organic carbon (TOC) are used as the pollutant concentration parameters.

[0012] Specifically, the first step is to install and calibrate a series of monitoring devices at key locations in the reverse osmosis membrane system. For example, electromagnetic flow meters are installed in the permeate pipeline, pressure transmitters in the feed and concentrate pipelines, and conductivity meters in the feed and permeate pipelines. These online sensors continuously collect data at a frequency of once per minute. The controller calculates the normalized permeate flux based on real-time permeate flow rate, operating pressure, and temperature, and calculates the desalination rate based on the feed permeate conductivity, while simultaneously recording the transmembrane pressure difference. For example, if the normalized permeate flux decreases from an initial value of 1.0 to 0.85 within 30 days, the transmembrane pressure difference increases from 1.0 MPa to 1.2 MPa, and the desalination rate slightly decreases from 99.5% to 99.2%, this data reflects the deterioration trend of membrane performance.

[0013] The second step is to deploy water quality analysis instruments to periodically monitor the influent water quality. For example, an online water quality analyzer is installed at the pretreatment effluent outlet. This analyzer integrates a turbidity meter, a COD analysis module, a silica meter, a calcium ion selective electrode, and a TOC analysis module. The instrument is set to sample and analyze once per hour, outputting the concentration values ​​of each pollutant. For example, in a certain hour's measurement, the data obtained are: suspended solids (SS) concentration 12 mg / L, chemical oxygen demand (COD) concentration 25 mg / L, silica concentration 20 mg / L, calcium ion concentration 75 mg / L, and total organic carbon concentration 8 mg / L. Figure 2 .

[0014] In an optional embodiment, the calculation of the contribution of each contaminant to the current deterioration of membrane performance parameters, to obtain the pollution dominance index of each contaminant, includes: For the i-th pollutant, the pollution dominance index of the pollutant Calculated using the following formula: in, Let i be the current concentration of the i-th pollutant. Let i be the pollution characteristic factor of the i-th pollutant. is the weighting coefficient of the i-th pollutant in the correlation model output on the decrease in normalized permeable flux, and j is the index of all pollutant types.

[0015] Current pollutant concentration Water quality monitoring from the previous stage, such as It is 75 mg / L. Pollution characteristic factors. These are parameters pre-set based on expert knowledge or long-term statistical data, reflecting the fouling tendency of pollutants, such as the fouling characteristic factor of calcium ions, which are prone to scaling. The value was set at 0.8, while organic matter... Set to 0.4. Weighting coefficient. The output is from a pre-trained correlation model, which is a gradient boosting decision tree model that learns the impact of different pollutant combinations on the rate of membrane flux decline based on historical operating data. Under the current operating conditions, after analyzing the input data, the model outputs the characteristic importance of each pollutant as a weight coefficient; for example, the model outputs the weight coefficient for calcium ions. The total organic carbon is 0.5. It is 0.1.

[0016] More specifically, the gradient boosting decision tree model is an additive model consisting of 150 base learners, i.e., decision trees. The maximum depth of each decision tree is limited to 6 layers to prevent overfitting. The model is built iteratively, with each subsequent tree built to fit the prediction residuals of the previous tree. The training set for this model comes from the historical operating data of the reverse osmosis membrane system over the past three years, containing more than 20,000 sample points. Each sample point consists of a feature vector and a label. The feature vector contains five contaminant concentration parameters: SS, COD, ... , The data is analyzed across seven dimensions: TOC, influent temperature, and operating pressure. The label represents the actual rate of decline of the normalized permeable flux within 24 hours following the given time point. The training process employs a gradient descent algorithm to minimize the loss function, which is the mean squared error. , where y is the actual rate of descent. The values ​​are the model's predicted values. During training, the learning rate is set to 0.05, and the optimal number of trees is selected using 10-fold cross-validation. The real-time input to the model is a seven-dimensional vector containing the current concentrations of five pollutants and two operating parameters; for example, [12, 25, 20, 75, 8, 28, 1.1] represents the current measured values ​​of each parameter. The model's output is the contribution of each input feature to the prediction result, i.e., the feature importance score. This score, after normalization, is used as the weight coefficient. use.

[0017] The numerator is calculated, with calcium ions contributing 30 as an example. This calculation is repeated for all pollutant types, such as SS, COD, silica, calcium ions, and TOC, and the results are summed to obtain the denominator. Assuming the sum of the contribution measures of all pollutants, i.e., the denominator, is 65, then the pollution dominance index of calcium ions is... Approximately 0.46. The pollution dominance index for each pollutant was calculated sequentially, resulting in a normalized set of indices. This set represents the relative contribution of each pollutant under the current membrane fouling state, such as... Figure 3 .

[0018] S2, Identify a group of dominant pollutants based on the pollution dominance index, and calculate a synergistic pollution index representing the synergistic pollution effect of a specific combination of pollutants in the group of dominant pollutants; For multiple candidate cleaning schemes, calculate the cleaning decision score of each candidate cleaning scheme based on the expected removal efficiency of each scheme for the dominant pollutants, the pollution dominance index, the expected mitigation efficiency for synergistic pollution, and the synergistic pollution index. All pollutants whose pollution dominance index exceeds a threshold are identified as dominant pollutants. Based on a pre-defined co-contamination rule base, a search is conducted within this group of dominant pollutants to determine if any known synergistic combinations exist, such as a combination of organic matter and microorganisms, or a combination of iron ions and silica. If such combinations exist, a co-contamination index is calculated for that combination. In one embodiment, the calculation formula is the membrane performance degradation rate corresponding to the combination acting as the dominant pollutant, divided by the sum of the membrane performance degradation rates when each pollutant acts as the dominant pollutant individually. In another embodiment, the co-contamination index is calculated using a weighted integral of pollutant concentrations over the past operating cycle to reflect the cumulative effect.

[0019] A candidate library of cleaning solutions is predefined, such as solution 1 (citric acid cleaning), solution 2 (high pH alkaline cleaning), and solution 3 (oxidative cleaning). An expected removal efficiency coefficient is preset for each solution and each contaminant, with a value between 0 and 1. For example, citric acid has a removal efficiency of 0.9 for calcium scale and 0.2 for organic matter. Similarly, an expected mitigation efficiency coefficient is preset for each solution and each co-contamination combination. The cleaning decision score for each candidate solution is calculated by weighted summation using the formula: ∑(Pollution dominance index of a dominant contaminant × Expected removal efficiency of the cleaning solution for that contaminant) + ∑(Co-contamination index × Expected mitigation efficiency of the cleaning solution for that co-contamination).

[0020] In an optional embodiment, the step of identifying a group of dominant pollutants based on the pollution dominance index and calculating a synergistic pollution index representing the synergistic pollution effect of a specific combination of pollutants in the group of dominant pollutants includes: All pollutants are sorted in descending order of their dominant pollution indices, and those with dominant pollution indices greater than a preset dominant pollution threshold are selected. The pollutants are the dominant pollutants; For a combination of two pollutants i and j as the dominant pollutants, the membrane performance degradation rate corresponding to the combination being the dominant pollutant is retrieved from the historical database. This rate is then divided by the sum of the membrane performance degradation rates when each pollutant is the dominant pollutant. This ratio is used as the co-pollution index of the pollutant combination (i,j). .

[0021] Sort the dominant pollution indices of each pollutant calculated in the previous step in descending order, for example, to obtain the sequence as follows: , , Set a dominant threshold. The threshold is set to 0.18. All pollutants with a pollution dominance index greater than 0.18 are identified as dominant pollutants. In this example, calcium ions and suspended solids (SS) are identified as dominant pollutants because their pollution dominance indices both exceed the threshold.

[0022] For the combination of calcium ions and suspended solids (SS), the historical operating database was queried. The database was retrieved for all historical periods in which both calcium ions and SS were the dominant pollutants. The average rate of decline in normalized permeable flux during these periods was calculated, yielding a synergistic deterioration rate of 0.009 per day. Then, periods in which only calcium ions were the sole dominant pollutant were retrieved, with an average deterioration rate of 0.005 per day; and periods in which only SS were the sole dominant pollutant were retrieved, with an average deterioration rate of 0.003 per day. The synergistic pollution index was calculated according to the definition. The value greater than 1 indicates a synergistic fouling effect between calcium ions and suspended solids (SS), meaning that membrane fouling caused by their co-existence is more severe than when they act alone.

[0023] In an optional embodiment, calculating the cleaning decision score for each candidate cleaning scheme based on the expected removal efficiency of each scheme for the dominant pollutant, the pollution dominance index, the expected mitigation efficiency for co-polluting pollutants, and the co-polluting index includes: For the k-th candidate cleaning scheme, the cleaning decision score of the scheme is... Calculated using the following weighted summation formula: Wherein, D is the dominant pollutant set, and S is the pollutant combination set with synergistic effects; To determine the expected removal efficiency of scheme k for the dominant pollutant i, The pollution dominance index for pollutant i; For scheme k, the combination of pollutants The expected mitigation efficacy of synergistic pollution control. The co-contamination index; and These are preset weighting factors.

[0024] Specifically, several candidate cleaning schemes were preset, such as scheme 1 being citric acid pickling, scheme 2 being sodium hydroxide alkaline pickling, and scheme 3 being a sequential cleaning process of alkaline pickling followed by acid pickling. The weighting factors were set empirically. and Based on the calculations in the previous step, the dominant pollutant set D includes calcium ions and suspended solids (SS), with pollution dominance indices of [missing information]. and The pollutant combination S exhibiting a synergistic effect is calcium ions and suspended solids (SS), with a synergistic pollution index. The efficiency parameters of each scheme are retrieved from the knowledge base, such as the calcium ion removal efficiency of scheme 3. Removal efficiency for suspended solids (SS) Furthermore, Scheme 3 demonstrates efficacy in mitigating the synergistic pollution of calcium ions and suspended solids (SS). .

[0025] Taking Scheme 3 as an example, the score calculation process is as follows: The weighted effectiveness for the dominant pollutant is calculated to be 0.4179. The weighted effectiveness for co-pollutants is calculated to be 0.2869. The two parts are added together to obtain the total score for Scheme 3. Repeat this calculation for all candidate solutions.

[0026] S3, when the maximum value of the cleaning decision score among the candidate cleaning schemes exceeds the preset cleaning start threshold, the cleaning scheme corresponding to the maximum value is selected, and the specific process parameters of the scheme are determined according to the dominant pollutant and synergistic pollution index that contribute to the maximum score. A pre-set cleaning initiation threshold, such as 0.7, is used. When the maximum value among all calculated candidate solution decision scores (e.g., Solution 1's cleaning decision score is 0.85), exceeding 0.7, cleaning is initiated. The cleaning solution with the highest decision score is selected for execution. Specific process parameters for the cleaning solution are determined by querying a pre-set process parameter rule base. This rule base stores a series of rules mapping diagnostic results to operational parameters. Each rule's condition section includes: cleaning solution type, dominant contaminant type, specific synergistic contaminant combination, and their PDI and SCI value ranges; the result section contains the corresponding detailed process parameters, such as cleaning agent type, concentration, pH, temperature, and cycle time. Figure 4 To prevent the reverse osmosis system from being newly put into operation when the feed water quality is poor, resulting in high calculated PDI and CDS, thus triggering the cleaning start-up threshold, the system may optionally combine the transmembrane pressure difference or the decrease in permeate flux to determine whether to trigger the cleaning. For example, if the pressure difference is greater than a threshold, or the decrease in permeate flux is greater than another threshold, then the cleaning is triggered.

[0027] In an optional embodiment, determining the specific process parameters of the scheme based on the dominant pollutant and co-polluting index that contribute to the maximum score includes: Establish a process parameter rule base, which maps the dominant pollutant type, synergistic pollutant combination and the corresponding pollutant dominance index and synergistic pollutant index value range to the specific process parameters of the cleaning scheme, including cleaning agent type, concentration, pH value, temperature and circulation time. When a specific dominant pollutant or a combination of pollutants with a high synergistic pollution index is identified, the corresponding combination of process parameters is matched and determined from the rule base.

[0028] Specifically, the process parameter rule base is a structured database that stores a series of IF-THEN rules. For example, one rule is: IF selection scheme is sequential cleaning AND the dominant contaminants are calcium ions and suspended solids SSAND AND in the range of 0.4 to 0.6 For pH values ​​greater than 1.1, the alkaline washing step uses a 0.5% sodium hydroxide solution at pH 12, a temperature of 35°C, and a circulation time of 60 minutes; the acid washing step uses a 2.0% citric acid solution at pH 2, a temperature of 30°C, and a circulation time of 90 minutes. Another rule may apply... When the concentration is below 0.4, the concentration of citric acid in the corresponding pickling step will decrease to 1.0%.

[0029] If the optimal solution has been determined to be sequential cleaning, and the dominant pollutants are calcium ions and suspended solids (SS), then... , Using the diagnostic results as query criteria, a match is performed in the process parameter rule base. The condition of the first rule mentioned above is found to fully satisfy the current situation, i.e. The value 0.46 is in the range of 0.4 to 0.6 and The value 1.125 is greater than 1.1. Therefore, detailed process parameters are extracted from the THEN section of this rule and output to the control system. The generated cleaning instructions are: use 0.5% sodium hydroxide solution to circulate and clean for 60 minutes at pH 12 and 35°C; use 2.0% citric acid solution to circulate and clean for 90 minutes at pH 2 and 30°C.

[0030] S4. Based on the dominant pollutants identified in this cleaning and the feedback on the cleaning effect, adjust the pollution characteristic factors used to calculate the pollution dominance index and the cleaning initiation threshold to optimize subsequent decisions.

[0031] After cleaning, the membrane performance recovery rate is calculated by comparing the standardized permeate flux before and after cleaning. If the dominant contaminant identified in this cleaning is organic matter, and alkaline cleaning is implemented, but the performance recovery rate is lower than expected (e.g., below 90%), it indicates that the current model may have overestimated the contamination tendency of organic matter. Based on the preset adjustment step size, the organic matter contamination characteristic factor is lowered by a small value, for example, from 0.5 to 0.48. Regarding the cleaning initiation threshold, if it is found that the actual deterioration of membrane performance exceeds the ideal range when several cleaning decisions are initiated consecutively, leading to increased cleaning difficulty, the cleaning initiation threshold is adjusted from, for example, 0.70 to 0.68 to achieve earlier warning and intervention.

[0032] In an optional embodiment, the adjustment of the pollution characteristic factors used to calculate the pollution dominance index and the cleaning initiation threshold, and the optimization of subsequent decisions, includes: After the cleaning process is completed, assess the actual normalized permeate flux recovery rate. If the recovery rate is lower than the preset lower threshold Then, the pollution characteristic factors of the dominant pollutants targeted in this cleaning will be increased by a first adjustment amount based on the original values. And reduce the cleaning initiation threshold; If the recovery rate is higher than the preset upper limit threshold If so, the cleaning initiation threshold is increased.

[0033] Specifically, after implementing the cleaning protocol determined in the previous step, the performance of the reverse osmosis membrane is re-monitored. Assume the normalized permeate flux before cleaning is 0.82, and after cleaning it is 0.93, while the initial ideal flux of the membrane is 1.0. The cleaning recovery rate is calculated to be 61.1%. This recovery rate is compared with a preset threshold, assuming a lower limit threshold. 70%, upper limit threshold It is 90%.

[0034] Since the calculated recovery rate of 61.1% is lower than the lower threshold... The cleaning efficiency was 70%, indicating a poor assessment of the pollution severity. Therefore, an adjustment mechanism was triggered. It identified calcium ions and suspended solids (SS) as the dominant pollutants in this cleaning operation and assigned corresponding pollution characteristic factors. Adjust upwards. For example, adjust upwards... Increase by a preset adjustment amount from 0.8. That is, adjust 0.05 to 0.85; The threshold was increased from 0.6 to 0.65. This makes the system more sensitive to the two contaminants in future calculations. Simultaneously, the cleaning initiation threshold was lowered; for example, if the original threshold was 0.75, it was adjusted to 0.73 to allow for earlier intervention, thereby completing the closed-loop optimization of the entire decision-making system. Figure 5 .

[0035] In the second embodiment, the present invention also proposes an intelligent decision-making system for cleaning RO membranes used for dyeing and printing wastewater reuse, comprising the following modules: Obtain membrane performance parameters representing the current operating status of the reverse osmosis membrane system, as well as various pollutant concentration parameters representing the feed water quality; based on the pollutant concentration parameters, and combined with pollution characteristic factors representing the pollution tendency of each pollutant and a correlation model derived from historical data, calculate the contribution of each pollutant to the current deterioration of membrane performance parameters, and obtain the pollution dominance index of each pollutant. A set of dominant pollutants is identified based on the pollution dominance index, and a synergistic pollution index representing the synergistic pollution effect of a specific combination of pollutants in the set of dominant pollutants is calculated; for multiple candidate cleaning schemes, a cleaning decision score is calculated for each candidate cleaning scheme based on the expected removal efficiency of each scheme for the dominant pollutants, the pollution dominance index, the expected mitigation efficiency for synergistic pollution, and the synergistic pollution index. When the maximum value of the cleaning decision score among the candidate cleaning schemes exceeds the preset cleaning start threshold, the cleaning scheme corresponding to the maximum value is selected, and the specific process parameters of the scheme are determined based on the dominant pollutant and co-contamination index that contribute to the maximum score. Based on the dominant pollutants identified during this cleaning process and feedback on the cleaning effect, the pollution characteristic factors used to calculate the pollution dominance index and the cleaning initiation threshold are adjusted to optimize subsequent decision-making.

[0036] In an optional embodiment, obtaining membrane performance parameters representing the current operating status of the reverse osmosis membrane system, and various pollutant concentration parameters representing the feed water quality, includes: The normalized permeate flux, transmembrane pressure difference, and desalination rate are collected in real time by online sensors and used as the membrane performance parameters. Suspended solids (SS), chemical oxygen demand (COD), and silica (Silica) in the influent are measured periodically using water quality analyzers. calcium ions The concentrations of total organic carbon (TOC) are used as the pollutant concentration parameters.

[0037] In an optional embodiment, the calculation of the contribution of each contaminant to the current deterioration of membrane performance parameters, to obtain the pollution dominance index of each contaminant, includes: For the i-th pollutant, the pollution dominance index of the pollutant Calculated using the following formula: in, Let i be the current concentration of the i-th pollutant. Let i be the pollution characteristic factor of the i-th pollutant. is the weighting coefficient of the i-th pollutant in the correlation model output on the decrease in normalized permeable flux, and j is the index of all pollutant types.

[0038] In an optional embodiment, the step of identifying a group of dominant pollutants based on the pollution dominance index, and calculating a synergistic pollution index representing the synergistic pollution effect of a specific combination of pollutants in the group of dominant pollutants, includes: All pollutants are sorted in descending order of their dominant pollution indices, and those with dominant pollution indices greater than a preset dominant pollution threshold are selected. The pollutants are the dominant pollutants; For a combination of two pollutants i and j as the dominant pollutants, the membrane performance degradation rate corresponding to the combination being the dominant pollutant is retrieved from the historical database. This rate is then divided by the sum of the membrane performance degradation rates when each pollutant is the dominant pollutant. This ratio is used as the co-pollution index of the pollutant combination (i,j). .

[0039] In an optional embodiment, calculating the cleaning decision score for each candidate cleaning scheme based on the expected removal efficiency of each scheme for the dominant pollutant, the pollution dominance index, the expected mitigation efficiency for co-polluting pollutants, and the co-polluting index includes: For the k-th candidate cleaning scheme, the cleaning decision score of the scheme is... Calculated using the following weighted summation formula: Wherein, D is the dominant pollutant set, and S is the pollutant combination set with synergistic effects; To determine the expected removal efficiency of scheme k for the dominant pollutant i, The pollution dominance index for pollutant i; For scheme k, the combination of pollutants The expected mitigation efficacy of synergistic pollution control. The co-contamination index; and These are the preset weighting factors.

[0040] In an optional embodiment, determining the specific process parameters of the scheme based on the dominant pollutant and co-polluting index that contribute to the maximum score includes: Establish a process parameter rule base, which maps the dominant pollutant type, synergistic pollutant combination and the corresponding pollutant dominance index and synergistic pollutant index value range to the specific process parameters of the cleaning scheme, including cleaning agent type, concentration, pH value, temperature and circulation time. When a specific dominant pollutant or a combination of pollutants with a high synergistic pollution index is identified, the corresponding combination of process parameters is matched and determined from the rule base.

[0041] In an optional embodiment, the adjustment of the pollution characteristic factors used to calculate the pollution dominance index and the cleaning initiation threshold, and the optimization of subsequent decisions, includes: After the cleaning process is completed, assess the actual normalized permeate flux recovery rate. If the recovery rate is lower than the preset lower threshold Then, the pollution characteristic factors of the dominant pollutants targeted in this cleaning will be increased by a first adjustment amount based on the original values. And reduce the cleaning initiation threshold; If the recovery rate is higher than the preset upper limit threshold If so, the cleaning initiation threshold is increased.

[0042] It should be clarified that this application is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of this application is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of this application.

[0043] The functional modules shown in the above-described block diagram can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, they can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this application are programs or code segments used to perform the required tasks. Programs or code segments can be stored on a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried on a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, radio frequency (RF) links, etc. Code segments can be downloaded via computer networks such as the Internet, intranets, etc.

[0044] It should also be noted that the exemplary embodiments mentioned in this application describe methods or systems based on a series of steps or apparatus. However, this application is not limited to the order of the above steps; that is, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0045] The aspects of this application have been described above with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It should be understood that each block in the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that these instructions, executable via the processor of the computer or other programmable data processing apparatus, enable the implementation of the functions / actions specified in one or more blocks of the flowchart illustrations and / or block diagrams. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field-programmable logic circuit. It is also understood that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can also be implemented by dedicated hardware performing the specified functions or actions, or can be implemented by a combination of dedicated hardware and computer instructions.

[0046] The above description is merely a specific implementation of this application. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, modules, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. It should be understood that the protection scope of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the protection scope of this application.

Claims

1. A smart decision-making method for cleaning RO membranes used in the reuse of dyeing and printing wastewater, characterized in that, Includes the following steps: Obtain membrane performance parameters representing the current operating status of the reverse osmosis membrane system, as well as various pollutant concentration parameters representing the feed water quality; based on the pollutant concentration parameters, and combined with pollution characteristic factors representing the pollution tendency of each pollutant and a correlation model derived from historical data, calculate the contribution of each pollutant to the current deterioration state of the membrane performance parameters, and obtain the pollution dominance index of each pollutant. The correlation model is a machine learning model trained based on historical operating data, used to establish a quantitative relationship between pollutant concentration and membrane performance deterioration rate and output the weight coefficient of each pollutant on the deterioration state of membrane performance parameters. A set of dominant pollutants is identified based on the pollution dominance index, and a synergistic pollution index representing the synergistic pollution effect of a specific combination of pollutants in the set of dominant pollutants is calculated; for multiple candidate cleaning schemes, a cleaning decision score is calculated for each candidate cleaning scheme based on the expected removal efficiency of each scheme for the dominant pollutants, the pollution dominance index, the expected mitigation efficiency for synergistic pollution, and the synergistic pollution index. When the maximum value of the cleaning decision score among the candidate cleaning schemes exceeds the preset cleaning start threshold, the cleaning scheme corresponding to the maximum value is selected, and the specific process parameters of the scheme are determined based on the dominant pollutant and co-contamination index that contribute to the maximum value. Based on the dominant pollutants identified during this cleaning process and feedback on the cleaning effect, the pollution characteristic factors used to calculate the pollution dominance index and the cleaning initiation threshold are adjusted to optimize subsequent decisions. The calculation of the contribution of each pollutant to the current deterioration of membrane performance parameters yields the pollution dominance index for each pollutant, including: For the i-th pollutant, the pollution dominance index of the pollutant Calculated using the following formula: in, Let i be the current concentration of the i-th pollutant. Let i be the pollution characteristic factor of the i-th pollutant. is the weighting coefficient of the i-th pollutant in the correlation model output on the decrease in normalized permeable flux, and j is the index of all pollutant types.

2. The method according to claim 1, characterized in that, The acquisition of membrane performance parameters representing the current operating status of the reverse osmosis membrane system, and various pollutant concentration parameters representing the feed water quality, includes: The normalized permeate flux, transmembrane pressure difference, and desalination rate are collected in real time by online sensors and used as the membrane performance parameters. Suspended solids (SS), chemical oxygen demand (COD), and silica (Silica) in the influent are measured periodically using water quality analyzers. calcium ions The concentrations of total organic carbon (TOC) are used as the pollutant concentration parameters.

3. The method according to claim 1, characterized in that, The step of identifying a group of dominant pollutants based on the pollution dominance index, and calculating a synergistic pollution index representing the synergistic pollution effect of a specific combination of pollutants in the group of dominant pollutants, includes: All pollutants are sorted in descending order of their dominant pollution indices, and those with dominant pollution indices greater than a preset dominant pollution threshold are selected. The pollutants are the dominant pollutants; For a combination of two dominant pollutants i and j, the membrane performance degradation rate corresponding to the combination as the dominant pollutant is retrieved from the historical database. This rate is then divided by the sum of the membrane performance degradation rates when each pollutant is the dominant pollutant, and the ratio is used as the co-pollution index of the pollutant i and j combination. .

4. The method according to claim 2, characterized in that, The step of calculating the cleaning decision score for each candidate cleaning scheme based on the expected removal efficiency of each scheme for the dominant pollutant, the pollution dominance index, the expected mitigation efficiency for co-polluting pollutants, and the co-polluting index includes: For the k-th candidate cleaning scheme, the cleaning decision score of the scheme is... Calculated using the following weighted summation formula: Wherein, D is the dominant pollutant set, and S is the pollutant combination set with synergistic effects; To determine the expected removal efficiency of scheme k for the dominant pollutant i, The pollution dominance index for pollutant i; To determine the expected mitigation efficacy of scheme k for the synergistic pollution reduction of pollutant combination i and j, The co-contamination index; and These are the preset weighting factors.

5. The method according to claim 1, characterized in that, The step of determining the specific process parameters of the scheme based on the dominant pollutant and co-polluting index that contribute to the maximum value includes: Establish a process parameter rule base, which maps the dominant pollutant type, synergistic pollutant combination and the corresponding pollutant dominance index and synergistic pollutant index value range to the specific process parameters of the cleaning scheme, including cleaning agent type, concentration, pH value, temperature and circulation time. When a specific dominant pollutant or a combination of pollutants with a high synergistic pollution index is identified, the corresponding combination of process parameters is matched and determined from the rule base.

6. The method according to claim 1, characterized in that, The adjustment is used to calculate the pollution characteristic factors of the pollution-dominant index and the cleaning initiation threshold, and to optimize subsequent decisions, including: After the cleaning process is completed, assess the actual normalized permeate flux recovery rate. If the recovery rate is lower than the preset lower threshold Then, the pollution characteristic factors of the dominant pollutants targeted in this cleaning will be increased by a first adjustment amount based on the original values. And reduce the cleaning initiation threshold; If the recovery rate is higher than the preset upper limit threshold If so, the cleaning initiation threshold is increased.

7. A smart decision-making system for cleaning RO membranes used in dyeing and printing wastewater reuse, characterized in that, Includes the following modules: The module is used to acquire membrane performance parameters representing the current operating status of the reverse osmosis membrane system, as well as various pollutant concentration parameters representing the feed water quality. Based on the pollutant concentration parameters, and combined with pollution characteristic factors representing the pollution tendency of each pollutant and a correlation model derived from historical data, the contribution of each pollutant to the current deterioration of membrane performance parameters is calculated, and the pollution dominance index of each pollutant is obtained. The correlation model is a machine learning model trained based on historical operating data, used to establish a quantitative relationship between pollutant concentration and membrane performance deterioration rate and output the weight coefficient of each pollutant on the deterioration of membrane performance parameters. The first calculation module is used to identify a group of dominant pollutants based on the pollution dominance index, and to calculate a synergistic pollution index representing the synergistic pollution effect of a specific combination of pollutants in the group of dominant pollutants; and to calculate the cleaning decision score of each candidate cleaning scheme based on the expected removal efficiency of each scheme for the dominant pollutants, the pollution dominance index, the expected mitigation efficiency for synergistic pollution, and the synergistic pollution index. The second calculation module is used to select the cleaning scheme corresponding to the maximum value when the maximum value of the cleaning decision score of each candidate cleaning scheme exceeds the preset cleaning start threshold, and to determine the specific process parameters of the scheme based on the dominant pollutant and co-contamination index that contribute to the maximum value. The adjustment module is used to adjust the pollution characteristic factors used to calculate the pollution dominance index and the cleaning start threshold based on the dominant pollutants identified in this cleaning and the feedback on the cleaning effect, thereby optimizing subsequent decisions. The calculation of the contribution of each pollutant to the current deterioration of membrane performance parameters yields the pollution dominance index for each pollutant, including: For the i-th pollutant, the pollution dominance index of the pollutant Calculated using the following formula: in, Let i be the current concentration of the i-th pollutant. Let i be the pollution characteristic factor of the i-th pollutant. is the weighting coefficient of the i-th pollutant in the correlation model output on the decrease in normalized permeable flux, and j is the index of all pollutant types.

8. The system according to claim 7, characterized in that, The acquisition of membrane performance parameters representing the current operating status of the reverse osmosis membrane system, and various pollutant concentration parameters representing the feed water quality, includes: The normalized permeate flux, transmembrane pressure difference, and desalination rate are collected in real time by online sensors and used as the membrane performance parameters. Suspended solids (SS), chemical oxygen demand (COD), and silica (Silica) in the influent are measured periodically using water quality analyzers. calcium ions The concentrations of total organic carbon (TOC) are used as the pollutant concentration parameters.

Images