A recycling treatment process for photovoltaic fluorine-containing wastewater

Abstract

The application discloses a kind of photovoltaic fluorine-containing wastewater recycling treatment process, specifically related to wastewater treatment field, including the following steps: by multi-stage adjustment mode to fluorine-containing wastewater pH adjustment, by activated carbon or diatomite as filter medium, remove particulate impurities and suspended solids, so that the solid content in wastewater is reduced to set threshold value.The application realizes the efficiency and automation of wastewater treatment by multi-stage pH adjustment, optimized filtration, electrolytic treatment and ion exchange resin defluorination process, and combines sensor real-time monitoring and decision tree model judgment.In the process of wastewater treatment, through various optimization methods, it is ensured that fluorine ions, heavy metals and organic matter are fully removed at each stage, while the dynamic adjustment of the proportion of filter medium is ensured, so that the overall process is stable and efficient, and the recycling standard is reached.

Description

Technical Field

[0001] This invention relates to the field of wastewater treatment technology, and more specifically, to a recycling process for photovoltaic fluoride-containing wastewater. Background Technology

[0002] The recycling and reuse treatment of photovoltaic fluoride-containing wastewater refers to the treatment of fluoride-containing wastewater generated during the photovoltaic industry through a series of technical means, so that it meets the environmental protection standards for reuse or discharge, thereby reducing wastewater discharge, lowering treatment costs, and achieving effective recycling of water resources. This type of treatment usually involves a combination of physical, chemical, and biological methods.

[0003] To address the challenges of treating fluoride-containing wastewater and the incomplete removal of pollutants in photovoltaic production, existing technologies may experience inconsistencies in efficiency during the filtration and removal of suspended solids, heavy metals, and organic matter under varying concentrations or load conditions. This affects the overall wastewater treatment effect. Furthermore, incomplete fluoride ion removal is a significant concern, as fluoride ions are not easily removed. Existing technologies may result in substandard fluoride ion concentrations during treatment, hindering the achievement of recycling standards. Additionally, there are shortcomings in real-time monitoring and control of wastewater treatment parameters, preventing timely dynamic adjustments and reducing treatment effectiveness and system stability. Summary of the Invention

[0004] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a recycling treatment process for photovoltaic fluoride-containing wastewater. This process involves adjusting the pH value of the fluoride-containing wastewater through a multi-stage adjustment method and using activated carbon or diatomaceous earth as a filter medium to remove particulate impurities and suspended solids, thereby reducing the solid content in the wastewater to a set threshold and solving the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution, comprising the following steps:

[0006] pH value of fluoride-containing wastewater is adjusted through a multi-stage regulation method;

[0007] Activated carbon or diatomaceous earth is used as a filter medium to remove particulate impurities and suspended solids, reducing the solid content in the wastewater to a set threshold. After the content of suspended solids in the wastewater is controlled below the set suspended solids content threshold by the filter medium, alumina and zeolite are used to remove residual organic matter and trace heavy metal ions in the wastewater, resulting in primary filtered wastewater.

[0008] The primary filtration wastewater is decomposed by electrolysis to oxidize or reduce the residual organic matter and heavy metal ions in the wastewater. The fluoride ion concentration of the wastewater after electrochemical treatment is monitored. When the concentration drops to the set fluoride ion concentration threshold, the defluorination stage begins.

[0009] In the defluorination stage, selective ion exchange resin is added to the wastewater after electrochemical treatment to remove fluoride ions in a targeted manner. After the wastewater is treated by ion exchange, when the fluoride ion content drops to the set fluoride ion content threshold, it enters the recycling mode.

[0010] The fluoride ion concentration, pH value, and heavy metal content of wastewater are monitored using a fluoride ion selective electrode, a heavy metal sensor, and a pH sensor, generating a monitoring dataset. This dataset is then used as input to a decision tree model. The threshold values ​​corresponding to the monitoring dataset form a tree structure, with each node representing a feature for judgment and each branch representing a decision based on the feature threshold. Cross-validation is used to verify the model's accuracy. The constructed decision tree is then applied to monitor fluoride ion concentration, pH value, and heavy metal content data. The rules set by the decision tree determine whether the data meets the recycling standards, and the system switches to recycling mode when the fluoride ion concentration drops to the system's set concentration threshold.

[0011] In a preferred embodiment, the multi-stage adjustment method for fluoride-containing wastewater includes:

[0012] The pH of the wastewater is adjusted to the set range. First, an alkaline or acidic regulator is added to the fluoride-containing wastewater to adjust the pH value to the neutral to slightly alkaline range of pH 7-8, reducing the initial acidity or alkalinity of the wastewater. Then, a second pH adjustment is performed on the adjusted wastewater by adding an alkaline or acidic regulator to control the pH value within the set range of 6.5 to 7.5, ensuring the removal of fluoride ions from the wastewater during the filtration process.

[0013] In a preferred embodiment, when wastewater enters the filter medium, if the suspended solids concentration exceeds the set filtration threshold, the filtration efficiency will decrease, affecting the normal operation of the electrolysis unit. A particle swarm optimization model is used to adjust the ratio of activated carbon to diatomaceous earth, ensuring that the filtration efficiency is close to the set filtration threshold and reducing the impact of residual micro-impurities on subsequent treatment steps. The suspended solids filtration formula is as follows:

[0014]

[0015] in, This indicates the filtration efficiency, ranging from 0 to 1. It represents the filtration effect; the closer the value is to 1, the closer the filtration effect is to the set filtration threshold. This indicates the concentration of suspended solids in the wastewater; 50 represents the set threshold for suspended solids concentration, and represents the allowable concentration of suspended solids. When, in the formula The value of 1;5 represents the curve slope control coefficient in the formula, which affects the filtration efficiency. For suspended solids concentration The greater the slope, the more drastic the response of filtration efficiency to concentration changes; The base of the natural logarithm is used to calculate exponential decay; this formula is derived by relating the concentration of suspended solids to the base of the natural logarithm. The change in efficiency is processed exponentially to obtain the filtration efficiency. The value.

[0016] In a preferred embodiment, the proportion of filter media is dynamically adjusted based on a particle swarm optimization model to ensure filtration efficiency. By maintaining the set filtration threshold, the particle swarm optimization model can effectively find the media ratio at different concentrations, thereby increasing filtration efficiency. The formula for judgment is as follows:

[0017]

[0018] in, Represents particles The speed, the rate of proportional change; The inertia coefficient represents the degree to which a particle retains its current velocity; The learning factor determines the weights by which particles move toward individual and global extrema. Represents a random number, increasing the randomness of the search; Represents particles The new position; both the new particle velocity and the new position need to be evaluated to see if they meet a target value set by the process, that is, within a specified proportion range. In the velocity evaluation, the current and initial positions are combined, and the weights of random distribution are set according to a specified proportion. Each particle records its position corresponding to the specified velocity in the generation process, that is, its individual local position. Meanwhile, the global position is recorded as the range reached by all particles in the swarm that conforms to the process settings. When the fitness of the particles And the global position meets the process settings. If the optimal combination is found, the iteration stops.

[0019] In a preferred embodiment, the electrolysis and defluorination stage of the primary filtration wastewater treatment removes heavy metals and organic matter from the primary filtration wastewater through a combination of current and voltage. The electrolysis reaction rate is expressed by the following formula:

[0020]

[0021] in, This indicates the decomposition rate, or the rate at which pollutants are removed. Represents current and voltage, where This indicates the current intensity during the electrolysis process. This indicates the voltage applied during electrolysis; and The concentrations of heavy metals and organic matter in wastewater are indicated by a combination of current and voltage. Represented as a constant coefficient, it is used to adjust the units or scale of the calculation results. The exponent represents the nonlinear relationship in the formula, indicating the rate of change of the parameter, and is the rate of removal of organic matter and heavy metals from wastewater under current and voltage conditions; the exponent This indicates that the removal effect is not a simple linear relationship, but rather amplifies the influence of current and voltage to ensure pollutant removal. After electrolysis, the wastewater enters the defluorination stage, where fluoride ions are removed through ion exchange resin. The formula for the change in fluoride ion concentration during treatment by the ion exchange resin is:

[0022]

[0023] in, Indicates time The concentration of fluoride ions in the wastewater at that time; This indicates the initial fluoride ion concentration, meaning that the fluoride ion concentration in the wastewater is 10 mg / L at the start of the defluorination treatment. The base of the natural logarithm is approximately 2.718, used to calculate exponential decay. -0.05 represents the exponential decay coefficient, indicating the rate of natural decrease in fluoride ion concentration. The negative sign indicates that the concentration decreases over time. (Integral sign) Indicates from time 0 to The cumulative change, 0.02 represents the adsorption rate coefficient, which indicates the efficiency of fluoride ion removal. The larger the value, the faster the removal rate. This represents the current concentration difference of remaining fluoride ions, and indicates the amount of fluoride ions that need to be removed. t This represents a function used to smooth the adsorption process, where -0.1 controls the effect of time variation; This indicates that the adsorption process takes place within approximately 20 minutes. This indicates the natural decay of fluoride ion concentration; This indicates how fluoride ions are dynamically adsorbed during the treatment process, over time. As the fluoride ion concentration gradually decreases at a certain adsorption rate, this item represents the amount by which the fluoride ion concentration is gradually reduced. By considering the adsorption rate of ion exchange resin for removing fluoride ions, the difference in the current fluoride ion concentration, and the adjustment of the fluoride removal rate by time, the trend of fluoride ion concentration change over this period is calculated. When the fluoride ion concentration threshold is reached, and the fluoride ion content drops to the set threshold of 1 mg / L, the system automatically switches to the recycling mode.

[0024] In a preferred embodiment, a monitoring dataset containing the following data is generated by real-time monitoring of key indicators of wastewater using a fluoride ion selective electrode, a heavy metal sensor, and a pH sensor:

[0025] Fluoride ion concentration pH value indicates the acidity or alkalinity of wastewater; heavy metal concentration... The monitoring dataset is input into a decision tree model for judgment. Each node in the decision tree model represents a feature, such as fluoride ion concentration. Each branch makes a decision based on whether the feature value reaches a threshold, until a leaf node is reached to output the judgment result. In order to correct and monitor the data in real time, we can judge the fluoride ion concentration, pH value and heavy metal content monitored by the sensor in real time by its electrolysis reaction rate and fluoride ion concentration changes. The monitoring dataset is used as a decision tree model to judge whether the wastewater meets the recycling standards. Secondly, when the monitoring data fluctuates or differs greatly from historical data, the above formula is used for correction to ensure the accuracy of the judgment. The fluoride ion concentration formula is used to predict the expected change of fluoride ion at different time points, correct the sensor data and avoid error accumulation.

[0026] In a preferred embodiment, fluoride ion concentration, pH value, and heavy metal content are input into a decision tree model, and the fluoride ion concentration is... pH value, heavy metal concentration This forms the training set of monitoring data. Assuming the dataset contains 100 data points, 5-fold cross-validation is used, dividing the dataset into 5 subsets, each containing 20 data points. The subsets can be randomly split to ensure a uniform distribution of different data types within each subset. The first validation is performed as follows:

[0027] Using subset 1 as the validation set and subsets 2, 3, 4, and 5 as the training set, a decision tree model was trained on these 80 data points to learn how to determine whether the concentration of fluoride ions, pH value, and heavy metal concentration meet the recycling standards. The accuracy of the model was verified using 20 data points from subset 1, and the first verification results were obtained.

[0028] Second verification:

[0029] Using subset 2 as the validation set and subsets 1, 3, 4, and 5 as the training set, we validate the model on 20 data points from subset 2 to obtain the second validation result. This second validation result is then used to train the model. We repeat this process, using each subset in turn as the validation set and the remaining subsets as the training set, for a total of 5 iterations. In each iteration, we record the model's performance on the validation set, including accuracy and error rate, and calculate the average performance. We then average the results of the 5 validations to obtain the model's metrics. For example, if the accuracy rates of the 5 validations are 85%, 88%, 86%, 87%, and 89%, then the model's average accuracy is: The monitoring dataset is divided into multiple subsets for processing. Cross-validation uses different subsets to train and validate the model each time. In each iteration, the model is trained on the training set and evaluated on the validation set. The decision tree model judges the accuracy of fluoride ion concentration, pH value, and heavy metal concentration. The stability of the model is judged based on the average result of cross-validation. When the performance of a certain value is significantly lower than the average value, the model parameters need to be adjusted or a more complex model needs to be used. Cross-validation improves the judgment accuracy of the model and reduces the false positive rate.

[0030] The technical effects and advantages of this invention are as follows:

[0031] A fluoride ion selective electrode, a heavy metal sensor, and a pH sensor are introduced for real-time monitoring to generate a dataset. A decision tree model is used for judgment and correction to ensure the accuracy of the judgment. Cross-validation improves the stability and generalization ability of the model, thereby reducing false judgments and ensuring that the system can automatically switch to recycling mode when the fluoride ion concentration threshold is reached.

[0032] By removing organic matter, heavy metals and fluoride ions from wastewater through electrolysis and ion exchange resins, the wastewater treatment is ensured to meet the set standards, thereby achieving efficient reuse of wastewater, reducing environmental pollution and improving resource sustainability.

[0033] By adjusting the pH of the wastewater in multiple stages, fluoride ions and other pollutants are effectively removed under acidic and alkaline conditions, which improves the efficiency of subsequent filtration and electrolysis. Activated carbon or diatomaceous earth is used as the filter medium and combined with a particle swarm optimization model to make the filtration process more efficient and reduce the impact of suspended solids on the electrolysis unit and subsequent processes. Attached Figure Description

[0034] Figure 1 This is a process flow diagram of the present invention. Detailed Implementation

[0035] 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.

[0036] Refer to the instruction manual appendix Figure 1 An embodiment of the present invention provides a recycling process for photovoltaic fluoride-containing wastewater, comprising the following steps:

[0037] pH value of fluoride-containing wastewater is adjusted through a multi-stage regulation method;

[0038] Activated carbon or diatomaceous earth is used as a filter medium to remove particulate impurities and suspended solids, reducing the solid content in the wastewater to a set threshold. After the content of suspended solids in the wastewater is controlled below the set suspended solids content threshold by the filter medium, alumina and zeolite are used to remove residual organic matter and trace heavy metal ions in the wastewater, resulting in primary filtered wastewater.

[0039] The primary filtration wastewater is decomposed by electrolysis to oxidize or reduce the residual organic matter and heavy metal ions in the wastewater. The fluoride ion concentration of the wastewater after electrochemical treatment is monitored. When the concentration drops to the set fluoride ion concentration threshold, the defluorination stage begins.

[0040] In the defluorination stage, selective ion exchange resin is added to the wastewater after electrochemical treatment to remove fluoride ions in a targeted manner. After the wastewater is treated by ion exchange, when the fluoride ion content drops to the set fluoride ion content threshold, it enters the recycling mode.

[0041] The fluoride ion concentration, pH value, and heavy metal content of wastewater are monitored using a fluoride ion selective electrode, a heavy metal sensor, and a pH sensor, generating a monitoring dataset. This dataset is then used as input to a decision tree model. The threshold values ​​corresponding to the monitoring dataset form a tree structure, with each node representing a feature for judgment and each branch representing a decision based on the feature threshold. Cross-validation is used to verify the model's accuracy. The constructed decision tree is then applied to monitor fluoride ion concentration, pH value, and heavy metal content data. The rules set by the decision tree determine whether the data meets the recycling standards, and the system switches to recycling mode when the fluoride ion concentration drops to the system's set concentration threshold.

[0042] Multi-stage regulation methods for fluoride-containing wastewater include:

[0043] The pH of the wastewater is adjusted to the set range. Initially, an alkaline or acidic regulator is added to the fluoride-containing wastewater to adjust the pH to the neutral to slightly alkaline range of pH 7-8, reducing the initial acidity or alkalinity of the wastewater. A second pH adjustment is then performed on the adjusted wastewater, adding another alkaline or acidic regulator to control the pH within the set range of 6.5 to 7.5. This ensures the removal of fluoride ions from the wastewater during filtration. The initial adjustment to pH 7-8, followed by a second adjustment to pH 6.5-7.5, ensures that the chemical environment of the wastewater is suitable for the removal of fluoride ions and other pollutants in subsequent filtration and treatment steps. An appropriate pH value helps improve the removal efficiency of the filter media and ion exchange resin, preventing fluoride ions from forming compounds that are difficult to remove under excessively high or low pH conditions, thus ensuring a stable and efficient wastewater treatment process.

[0044] When wastewater enters the filter media, if the suspended solids concentration exceeds the set filtration threshold, the filtration efficiency will decrease, affecting the normal operation of the electrolysis unit. A particle swarm optimization model is used to adjust the ratio of activated carbon to diatomaceous earth, ensuring the filtration efficiency is close to the set filtration threshold and reducing the impact of residual micro-impurities on subsequent treatment steps. The suspended solids filtration formula is as follows:

[0045]

[0046] in, This indicates the filtration efficiency, ranging from 0 to 1. It represents the filtration effect; the closer the value is to 1, the closer the filtration effect is to the set filtration threshold. This indicates the concentration of suspended solids in the wastewater; 50 represents the set threshold for suspended solids concentration, and represents the allowable concentration of suspended solids. When, in the formula The value of 1;5 represents the curve slope control coefficient in the formula, which affects the filtration efficiency. For suspended solids concentration The greater the slope, the more drastic the response of filtration efficiency to concentration changes; The base of the natural logarithm is used to calculate exponential decay; this formula is derived by relating the concentration of suspended solids to the base of the natural logarithm. The change in efficiency is processed exponentially to obtain the filtration efficiency. The value;

[0047] It should be noted that the filtration efficiency The condition is to ensure the concentration of suspended solids in the wastewater. Within the set allowable threshold, such as The following measures are taken to ensure that the filtration effect is close to the ideal state, when the filtration efficiency is... A value close to 1 indicates that the wastewater filtration process meets the set standards, effectively removing suspended solids and avoiding adverse effects on subsequent treatment steps. With exponential decay treatment, the system can monitor and optimize the filtration process to ensure that the wastewater achieves the expected purity and quality after treatment.

[0048] The proportion of filter media is dynamically adjusted based on a particle swarm optimization model to ensure filtration efficiency. By maintaining the set filtration threshold, the particle swarm optimization model can effectively find the media ratio at different concentrations, thereby increasing filtration efficiency. The formula for judgment is as follows:

[0049]

[0050] in, Represents particles The speed, the rate of proportional change; The inertia coefficient represents the degree to which a particle retains its current velocity; The learning factor determines the weights by which particles move toward individual and global extrema. Represents a random number, increasing the randomness of the search; Represents particles The new position; both the new particle velocity and the new position need to be evaluated to see if they meet a target value set by the process, that is, within a specified proportion range. In the velocity evaluation, the current and initial positions are combined, and the weights of random distribution are set according to a specified proportion. Each particle records its position corresponding to the specified velocity in the generation process, that is, its individual local position. Meanwhile, the global position is recorded as the range reached by all particles in the swarm that conforms to the process settings. When the fitness of the particles And the global position meets the process settings. If the optimal combination is found, the iteration stops.

[0051] The process efficiency condition determined by the particle swarm optimization model is used to dynamically adjust the proportion of filter media. This keeps the filtration efficiency close to a set process threshold. By monitoring and adjusting the velocity and position of particles, the model can find a suitable combination, ensuring the process speed is appropriate and preventing abnormal substances from being affected by excessive levels and continuing processing. When the process efficiency at the globally suitable positions reaches the set value, such as... When the time is reached, it indicates that a combination that meets the quality requirements has been found, satisfying the process requirements, and the iteration stops;

[0052] In the electrolysis and defluorination stages of primary filtration wastewater treatment, heavy metals and organic matter in the primary filtration wastewater are removed through a combination of current and voltage. The electrolysis reaction rate formula is as follows:

[0053]

[0054] in, This indicates the decomposition rate, or the rate at which pollutants are removed. Represents current and voltage, where This indicates the current intensity during the electrolysis process. This indicates the voltage applied during electrolysis; and The concentrations of heavy metals and organic matter in wastewater are indicated by a combination of current and voltage. Represented as a constant coefficient, it is used to adjust the units or scale of the calculation results. The exponent represents the nonlinear relationship in the formula, indicating the rate of change of the parameter, and is the rate of removal of organic matter and heavy metals from wastewater under current and voltage conditions; the exponent This indicates that the removal effect is not a simple linear relationship, but rather amplifies the influence of current and voltage to ensure pollutant removal. After electrolysis, the wastewater enters the defluorination stage, where fluoride ions are removed through ion exchange resin. The formula for the change in fluoride ion concentration during treatment by the ion exchange resin is:

[0055]

[0056] in, Indicates time The concentration of fluoride ions in the wastewater at that time; This indicates the initial fluoride ion concentration, meaning that the fluoride ion concentration in the wastewater is 10 mg / L at the start of the defluorination treatment. The base of the natural logarithm is approximately 2.718, used to calculate exponential decay. -0.05 represents the exponential decay coefficient, indicating the rate of natural decrease in fluoride ion concentration. The negative sign indicates that the concentration decreases over time. (Integral sign) Indicates from time 0 to The cumulative change, 0.02 represents the adsorption rate coefficient, which indicates the efficiency of fluoride ion removal. The larger the value, the faster the removal rate. This represents the current concentration difference of remaining fluoride ions, and indicates the amount of fluoride ions that need to be removed. t This represents a function used to smooth the adsorption process, where -0.1 controls the effect of time variation; This indicates that the adsorption process takes place within approximately 20 minutes. This indicates the natural decay of fluoride ion concentration; This indicates how fluoride ions are dynamically adsorbed during the treatment process, over time. As the fluoride ion concentration gradually decreases at a certain adsorption rate, this item represents the amount by which the fluoride ion concentration is gradually reduced. By considering the adsorption rate of ion exchange resin to remove fluoride ions, the difference in the current fluoride ion concentration, and the adjustment of the fluoride removal rate by time, the trend of fluoride ion concentration change over this period is calculated. When the fluoride ion concentration threshold is reached, and the fluoride ion content drops to the set threshold of 1 mg / L, the system automatically switches to the recycling mode.

[0057] The condition for judging the change in fluoride ion concentration is to monitor the removal process of fluoride ions in wastewater, ensuring that the concentration gradually decreases during the treatment process and reaches the set threshold of 1 mg / L. Through this formula, the natural decay of fluoride ions and the resin adsorption removal effect can be analyzed to optimize the treatment process. When the fluoride ion concentration reaches the set threshold, the system will automatically switch to the recycling mode to ensure the high efficiency and stability of wastewater treatment.

[0058] By using fluoride ion selective electrodes, heavy metal sensors, and pH sensors to monitor key indicators of wastewater in real time, a monitoring dataset containing the following data was generated:

[0059] Fluoride ion concentration pH value indicates the acidity or alkalinity of wastewater; heavy metal concentration... The monitoring dataset is input into a decision tree model for judgment. Each node in the decision tree model represents a feature, such as fluoride ion concentration. Each branch makes a decision based on whether the feature value reaches a threshold, until the leaf node is reached to output the judgment result. In order to correct and monitor the data in real time, we can judge the fluoride ion concentration, pH value and heavy metal content monitored by the sensor in real time by its electrolysis reaction rate and fluoride ion concentration changes. The monitoring dataset is used as a decision tree model to judge whether the wastewater meets the recycling standards. Secondly, when the monitoring data fluctuates or differs greatly from historical data, the above formula is used for correction to ensure the accuracy of the judgment. The fluoride ion concentration formula is used to predict the expected change of fluoride ion at different time points, correct the sensor data and avoid error accumulation.

[0060] Furthermore, by monitoring in real time whether the fluoride ion concentration, pH value, and heavy metal content meet the wastewater recycling standards, and by using a decision tree model, electrolysis reaction rate formula, and fluoride ion concentration change formula for correction, the wastewater treatment effect can be judged in real time, ensuring the accuracy of the data. When the monitoring data fluctuates or differs significantly from historical data, the system can use these formulas to correct the sensor data, avoid error accumulation, and ensure the continuous, stable, and efficient operation of wastewater treatment.

[0061] Fluoride ion concentration, pH value, and heavy metal content are input into the decision tree model, with fluoride ion concentration... pH value, heavy metal concentration This forms the training set of monitoring data. Assuming the dataset contains 100 data points, 5-fold cross-validation is used, dividing the dataset into 5 subsets, each containing 20 data points. The subsets can be randomly split to ensure a uniform distribution of different data types within each subset. The first validation is performed as follows:

[0062] Using subset 1 as the validation set and subsets 2, 3, 4, and 5 as the training set, a decision tree model was trained on these 80 data points to learn how to determine whether the concentration of fluoride ions, pH value, and heavy metal concentration meet the recycling standards. The accuracy of the model was verified using 20 data points from subset 1, and the first verification results were obtained.

[0063] Second verification:

[0064] Using subset 2 as the validation set and subsets 1, 3, 4, and 5 as the training set, we validate the model on 20 data points from subset 2 to obtain the second validation result. This second validation result is then used to train the model. We repeat this process, using each subset in turn as the validation set and the remaining subsets as the training set, for a total of 5 iterations. In each iteration, we record the model's performance on the validation set, including accuracy and error rate, and calculate the average performance. We then average the results of the 5 validations to obtain the model's metrics. For example, if the accuracy rates of the 5 validations are 85%, 88%, 86%, 87%, and 89%, then the model's average accuracy is: The monitoring dataset is divided into multiple subsets for processing. Cross-validation uses different subsets to train and validate the model each time. In each iteration, the model is trained on the training set and evaluated on the validation set. The decision tree model judges the accuracy of fluoride ion concentration, pH value, and heavy metal concentration. The stability of the model is judged based on the average result of cross-validation. When the performance of a certain value is significantly lower than the average value, the model parameters need to be adjusted or a more complex model needs to be used. Cross-validation improves the judgment accuracy of the model and reduces the false positive rate.

[0065] Furthermore, determining the concentrations of fluoride ions, pH value, and heavy metals is to ensure that these key indicators meet the standards for wastewater recycling. By inputting these data into the decision tree model and performing cross-validation, the accuracy and stability of the model can be verified. This ensures that the system can accurately determine whether the wastewater meets the reuse standards in real-time monitoring, thereby maintaining an efficient and stable wastewater treatment process.

[0066] Furthermore, to verify the accuracy and stability of the decision tree model on data of fluoride ion concentration, pH value, and heavy metal concentration, the dataset was divided into different subsets for multiple training and validations using 5-fold cross-validation. The average accuracy of the model was calculated. If the model performs consistently in each validation, it indicates that the model has good generalization ability and stability, thus enabling reliable determination of whether wastewater treatment meets recycling standards.

[0067] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A recycling and reuse treatment process for photovoltaic fluoride-containing wastewater, characterized in that, Includes the following steps: pH value of fluoride-containing wastewater is adjusted through a multi-stage regulation method; Activated carbon or diatomaceous earth is used as a filter medium to remove particulate impurities and suspended solids, reducing the solid content in the wastewater to a set threshold. After the content of suspended solids in the wastewater is controlled below the set suspended solids content threshold by the filter medium, alumina and zeolite are used to remove residual organic matter and trace heavy metal ions in the wastewater, resulting in primary filtered wastewater. The primary filtration wastewater is decomposed by electrolysis to oxidize or reduce the residual organic matter and heavy metal ions in the wastewater. The fluoride ion concentration of the wastewater after electrochemical treatment is monitored. When the concentration drops to the set fluoride ion concentration threshold, the defluorination stage begins. In the defluorination stage, selective ion exchange resin is added to the wastewater after electrochemical treatment to remove fluoride ions in a targeted manner. After the wastewater is treated by ion exchange, when the fluoride ion content drops to the set fluoride ion content threshold, it enters the recycling mode. The fluoride ion concentration, pH value, and heavy metal content of wastewater are monitored using a fluoride ion selective electrode, a heavy metal sensor, and a pH sensor, generating a monitoring dataset. This dataset is then used as input to a decision tree model. The threshold values ​​corresponding to the monitoring dataset form a tree structure, with each node representing a feature for judgment and each branch representing a decision based on the feature threshold. Cross-validation is used to verify the model's accuracy. The constructed decision tree is then applied to monitor fluoride ion concentration, pH value, and heavy metal content data. The rules set by the decision tree determine whether the data meets the recycling standards, and the system switches to recycling mode when the fluoride ion concentration drops to the system's set concentration threshold.

2. The recycling and reuse treatment process for photovoltaic fluoride-containing wastewater according to claim 1, characterized in that, Multi-stage regulation methods for fluoride-containing wastewater include: The pH of the wastewater is adjusted to the set range. An alkaline or acidic regulator is added to the fluoride-containing wastewater initially. The pH is then adjusted a second time by adding an alkaline or acidic regulator to control the pH within the set range of 6.5 to 7.5, ensuring the removal of fluoride ions from the wastewater during the filtration process.

3. The recycling and reuse treatment process for photovoltaic fluoride-containing wastewater according to claim 2, characterized in that, When wastewater enters the filter media, if the suspended solids concentration exceeds the set filtration threshold, the filtration efficiency will decrease, affecting the normal operation of the electrolysis unit. A particle swarm optimization model is used to adjust the ratio of activated carbon to diatomaceous earth to ensure that the filtration efficiency is close to the set filtration threshold, reducing the impact of residual impurities on subsequent treatment steps. The suspended solids filtration formula is as follows: in, This indicates the filtration efficiency, ranging from 0 to 1. Indicates the concentration of suspended solids in wastewater; when When, in the formula It equals 1; The base of the natural logarithm is used to calculate exponential decay; this formula is derived by relating the concentration of suspended solids to the base of the natural logarithm. The change in efficiency is processed exponentially to obtain the filtration efficiency. The value.

4. The recycling and reuse treatment process for photovoltaic fluoride-containing wastewater according to claim 3, characterized in that, The proportion of filter media is dynamically adjusted based on a particle swarm optimization model to ensure filtration efficiency. Maintaining within the set filtration threshold, the particle swarm optimization model is used to find the media ratio at different concentrations, thereby increasing filtration efficiency. The formula for judgment is as follows: in, Represents particles The speed, the rate of proportional change; The inertia coefficient represents the degree to which a particle retains its current velocity; The learning factor determines the weights by which particles move toward individual and global extrema. Represents a random number, increasing the randomness of the search; Represents particles The new position; both the new particle velocity and the new position need to be evaluated to see if they meet a target value set by the process, that is, within a specified proportion range. In the velocity evaluation, the current and initial positions are combined, and the weights of random distribution are set according to a specified proportion. Each particle records its position corresponding to the specified velocity in the generation process, that is, its individual local position. Meanwhile, the global position is recorded as the range reached by all particles in the swarm that conforms to the process settings. When the fitness of the particles And the global position meets the process settings. If the optimal combination is found, the iteration stops.

5. The recycling and reuse treatment process for photovoltaic fluoride-containing wastewater according to claim 4, characterized in that, In the electrolysis and defluorination stages of primary filtration wastewater treatment, heavy metals and organic matter in the primary filtration wastewater are removed through a combination of current and voltage. The electrolysis reaction rate formula is as follows: in, This indicates the decomposition rate, or the rate at which pollutants are removed. Represents current and voltage, where This indicates the current intensity during the electrolysis process. This indicates the voltage applied during electrolysis; and The concentrations of heavy metals and organic matter in wastewater are represented by a combination of current and voltage. After electrolysis, the wastewater enters the defluorination stage, where fluoride ions are removed through ion exchange resins. The formula for the change in fluoride ion concentration during treatment by the ion exchange resins is: in, Indicates time The concentration of fluoride ions in the wastewater at that time; The base of the natural logarithm. Indicates from time 0 to Cumulative changes; This represents the current concentration difference of remaining fluoride ions, and indicates the amount of fluoride ions that need to be removed. t This represents a function used to smooth the adsorption process, where -0.1 controls the effect of time variation; This indicates how fluoride ions are dynamically adsorbed during the treatment process, over time. The term represents the amount by which the fluoride ion concentration is gradually reduced at a certain adsorption rate, taking into account the adsorption rate of the ion exchange resin for removing fluoride ions, the difference in the current fluoride ion concentration, and the effect of time on the fluoride removal rate. in The initial fluoride ion concentration is 10 mg / L, which is the concentration of fluoride ions in the wastewater at the start of the defluorination treatment; -0.05 represents the exponential decay coefficient, which indicates the rate at which the fluoride ion concentration decreases naturally; and 0.02 represents the adsorption rate coefficient, which indicates the efficiency of fluoride ion removal. This indicates that the adsorption process takes place around 20 minutes.

6. The recycling and reuse treatment process for photovoltaic fluoride-containing wastewater according to claim 5, characterized in that, By using fluoride ion selective electrodes, heavy metal sensors, and pH sensors to monitor key indicators of wastewater in real time, a monitoring dataset containing the following data was generated: Fluoride ion concentration pH value indicates the acidity or alkalinity of wastewater; heavy metal concentration... The monitoring dataset is input into the decision tree model for judgment; each node of the decision tree model represents a feature, and each branch makes a decision based on whether the feature value reaches the threshold, until the leaf node is reached and the judgment result is output; Among them, the changes in electrolysis reaction rate and fluoride ion concentration are used to determine the fluoride ion concentration, pH value, and heavy metal content monitored by the sensor.

7. The recycling and reuse treatment process for photovoltaic fluoride-containing wastewater according to claim 6, characterized in that, Fluoride ion concentration, pH value, and heavy metal content are input into the decision tree model, with fluoride ion concentration... pH value, heavy metal concentration This forms the training set of monitoring data. Assuming the dataset contains 100 data points, 5-fold cross-validation is used, dividing the dataset into 5 subsets, each containing 20 data points. The subsets are randomly split to ensure a uniform distribution of different data types within each subset. The first validation is performed as follows: Using subset 1 as the validation set and subsets 2, 3, 4, and 5 as the training set, a decision tree model was trained on these 80 data points to learn how to determine whether the concentration of fluoride ions, pH value, and heavy metal concentration meet the recycling standards. The accuracy of the model was verified using 20 data points from subset 1, and the first verification results were obtained. Second verification: Using subset 2 as the validation set and subsets 1, 3, 4, and 5 as the training set, the model is validated on 20 data points in subset 2 to obtain a second validation result, which is then used to train the model. This process is repeated, using each subset in turn as the validation set and the remaining subsets as the training set, for a total of 5 iterations. In each iteration, the model's performance on the validation set is recorded, including accuracy and error rate. The average performance is calculated, and the results of the 5 validations are averaged to obtain the model's metrics. The accuracy of the decision tree model in judging fluoride ion concentration, pH value, and heavy metal concentration is assessed. The stability of the model is determined based on the average result of cross-validation. If the performance for a certain value is lower than the average, the model parameters need to be adjusted or a more complex model needs to be used. Cross-validation improves the model's judgment accuracy and reduces the false positive rate.

Images