Accurate dosing control system for water treatment

Abstract

The invention discloses an accurate dosing control system for water treatment, relates to the technical field of dosing control, and is used for solving the problem of poor dosing control accuracy in the water treatment process. Comprising a data acquisition module, a condition library construction module, a multi-target processing module and an optimization decision module. According to the method, a sensor is arranged in a water body to collect pollutant parameters, data are preprocessed, the types and the concentrations of pollutants are recognized through a support vector machine model, after recognition, an optimal processing parameter combination is called from a removal condition library, the condition library is dynamically updated by using an Apriori algorithm, and when multiple pollutants coexist, the optimal processing parameter combination is called from the removal condition library. Removal conditions are automatically adjusted according to interaction rules among pollutants, collaborative removal is realized, an optimal decision is solved by constructing a multi-objective optimization model with removal efficiency, medicament cost, processing time and environmental adaptability as targets and combining a genetic algorithm and a particle swarm optimization algorithm, intelligent dosing control is realized, and the system is suitable for large-scale popularization and application. And the accuracy and efficiency of water treatment are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of dosing control, and more particularly to a precise dosing control system for water treatment. Background Art

[0002] With the development of sensor technology, automatic control technology and data processing technology, precise dosing control system has gradually become an important innovation direction in the field of water treatment. The system monitors water quality parameters such as pH, turbidity, conductivity, dissolved oxygen, etc. in real time, combines advanced control algorithms and feedback adjustment mechanisms, and dynamically adjusts the dosage of chemicals to ensure the efficiency, stability and environmental protection of the water treatment process. This precise dosing control is not only suitable for traditional urban sewage treatment and industrial wastewater treatment, but can also be widely used in drinking water purification, agricultural water treatment and other fields.

[0003] Deficiencies in existing technologies: Wastewater treatment technology is deficient in the coordinated treatment of multiple pollutants, especially when the optimal dosage and reaction conditions of different pollutants (such as organic matter, ammonia nitrogen, phosphorus, etc.) are different, it is difficult to achieve accurate and efficient dosing control. It usually relies on a fixed dosing ratio or a simple control rule, and it is difficult to respond to the dynamic fluctuations of pollutant concentrations in wastewater in real time, resulting in insufficient treatment of certain pollutants or waste of reagents. At the same time, the lack of sufficient consideration of mutual interference between reagents may cause the weakening of drug efficacy or the generation of by-products, further affecting the treatment effect. The lack of adaptability makes it impossible to flexibly adjust the dosing strategy according to real-time environmental changes, resulting in large fluctuations in the effect of coordinated treatment. Due to the lack of multi-objective optimization capabilities, it is difficult to achieve a balance between the removal rate of each pollutant, reagent consumption and operating costs, which restricts the efficiency and economy of wastewater treatment. Summary of the invention

[0004] In order to overcome the above-mentioned defects of the prior art, there is a solution as follows to solve the problem of poor dosing control accuracy in the water treatment process in the above-mentioned background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A precise dosing control system for water treatment, comprising a data acquisition module, a condition library building module, a multi-objective processing module and an optimization decision module, wherein the modules are connected via signals;

[0007] The data acquisition module is used to set sensors in the water body to collect pollutant parameters, pre-process the collected data, use the support vector machine classification model to identify the type of pollutants, and output the type and concentration of pollutants;

[0008] The condition library construction module is used to construct a removal condition library after identifying the type and concentration of pollutants, call the condition combination for water quality treatment according to the removal conditions stored in the removal condition library, and use the Apriori algorithm to update the condition combination. In the case of coexistence of multiple pollutants, the removal conditions are automatically adjusted according to the interaction rules between pollutants;

[0009] The multi-objective processing module is used to determine the optimization objectives and constraints of the water treatment process, integrate multiple objectives to obtain a multi-objective function, and construct a multi-objective optimization model for water treatment decision-making;

[0010] The optimization decision module is used to combine the genetic algorithm and the particle swarm optimization algorithm to determine the global optimal solution of multi-objective requirements in the water treatment decision-making process, and optimize the water treatment decision according to the global optimal solution.

[0011] In a preferred embodiment, a sensor is set in a water body to collect pollutant parameters, and the collected data is preprocessed, a support vector machine classification model is used to identify the type of pollutants, and the type and concentration of pollutants are output. The specific steps are as follows:

[0012] Pollutant parameters include conductivity, fluorescence intensity, turbidity, and redox potential;

[0013] Remove noise from collected data, identify and correct abnormal data points, and convert preprocessed data into feature vectors;

[0014] Build a feature library to store the feature ranges and classification labels of pollutants;

[0015] Based on the feature library and collected data, the radial basis kernel function is used to map the feature vectors of the collected data. According to the classification labels of pollutants in the feature library and the feature vectors of pollutants, the support vector machine classification model is used to identify the types of pollutants, and the classification results, confidence levels and concentration values ​​of pollutants are obtained.

[0016] In a preferred embodiment, after identifying the type and concentration of pollutants, a removal condition library is constructed, a condition combination is called for water quality treatment according to the removal conditions stored in the removal condition library, and the condition combination is updated using the Apriori algorithm. The specific steps include:

[0017] Initial construction of the removal condition library: According to the basic characteristics of pollutants and treatment experience, a pollutant removal condition library is established. For each pollutant, the removal parameters required for removal are set. The removal parameters include the optimal pH value, suitable temperature, recommended reagent type, initial dosage, and reaction time.

[0018] After identifying the pollutants and their concentrations, the corresponding removal parameters in the condition combination in the removal condition library are automatically called;

[0019] The Apriori algorithm is embedded in the removal condition library to learn the correlation between different pollutant removal conditions, determine the best treatment condition combination, and update the generated condition combination to the condition library.

[0020] In a preferred embodiment, for the coexistence of multiple pollutants, the removal conditions are automatically adjusted according to the interaction rules between the pollutants. The specific steps are as follows:

[0021] Analyze the historical removal parameters, including removal effect, reagent dosage, temperature, pH value, and treatment time, to determine the actual removal efficiency of each pollutant and the average removal effect between the existing removal conditions;

[0022] Based on the relationship between conditions and removal effects, pollutant removal parameters are adjusted. Based on removal parameter analysis and removal effect optimization, the removal conditions for each pollutant are dynamically updated in the condition library.

[0023] In the case of coordinated treatment of multiple pollutants, interaction rules are constructed to determine the interactions between pollutants, and the selection and dosage of reagents are optimized based on the interaction factors.

[0024] In a preferred embodiment, the method for determining the optimization objectives and constraints of a water treatment process, integrating multiple objectives to obtain a multi-objective function, and constructing a multi-objective optimization model for water treatment decision-making includes the following steps:

[0025] The optimization objectives include maximizing removal efficiency, minimizing reagent costs, minimizing reaction time, and maximizing adaptability to environmental conditions;

[0026] The constraints include minimum removal rate constraint, reagent dosage constraint, reaction time constraint and environmental condition constraint;

[0027] Integrate the calculations according to the optimization objectives to form a multi-objective function;

[0028] The multi-objective function results, optimization objectives and constraints are combined to form a multi-objective optimization model, and water treatment decisions are made.

[0029] In a preferred embodiment, a genetic algorithm and a particle swarm optimization algorithm are combined to determine the global optimal solution of multi-objective requirements in the water treatment decision-making process, and the water treatment decision optimization is performed according to the global optimal solution, including the following steps:

[0030] Use the initialized genetic algorithm to perform global search, randomly generate an initial population containing multiple sets of candidate solutions, and use the candidate solutions as the initial individuals of the genetic algorithm;

[0031] Calculate the fitness value of each initial individual according to the multi-objective function;

[0032] Use roulette or tournament selection methods to select individuals from the population based on their fitness values ​​to serve as parents for the next generation;

[0033] Perform a crossover operation on the selected parent individuals to generate new offspring individuals, and perform a mutation operation on the genes of the selected offspring individuals;

[0034] Repeat the fitness calculation, selection, crossover and mutation operations for multiple generations of iterative evolution. The genetic algorithm will converge to the initial optimal solution as the excellent solution set after multiple iterations.

[0035] The excellent solution set obtained by the genetic algorithm is used as the initial particle swarm of particle swarm optimization, and the particle swarm optimization performs local fine search. The specific steps are as follows:

[0036] Initialize the position and speed of the particle swarm, use the excellent solution set output by the genetic algorithm as the initial position of the particle swarm, each particle represents a candidate solution, and set the particle speed;

[0037] Calculate the fitness value of each particle to measure the quality of the current solution, and record the individual optimal position of each particle and the global optimal position of the particle swarm;

[0038] According to the speed update formula of the particle swarm optimization algorithm, the particle position is dynamically adjusted;

[0039] Repeat the fitness value calculation, velocity update and position update process to update the individual optimal position and global optimal position of the particle;

[0040] When the particle swarm reaches the maximum number of iterations or the global optimal position no longer changes in consecutive iterations, the algorithm terminates and the global optimal position in the particle swarm is taken as the optimal solution of the multi-objective optimization model;

[0041] Output the final optimal solution, which includes the optimal dosage of reagent, reaction time, temperature and pH value.

[0042] The technical effects and advantages of the precise dosing control system for water treatment of the present invention are as follows:

[0043] The present invention sets sensors in the water body to collect pollutant parameters (such as conductivity, fluorescence intensity, turbidity and redox potential), pre-processes the data, and uses the support vector machine classification model to identify the type and concentration of pollutants. After identifying the pollutants, the system calls the corresponding treatment parameter combination from the removal condition library, and updates the condition combination through the Apriori algorithm to dynamically optimize the condition library. At the same time, when multiple pollutants coexist, the removal conditions are automatically adjusted according to the interaction rules between the pollutants to ensure the synergistic removal effect of different pollutants.

[0044] A multi-objective optimization model is further constructed, with removal efficiency, chemical cost, treatment time and environmental adaptability as optimization targets, integrating multi-objective requirements to meet the overall needs of water treatment. Through the combination of genetic algorithm and particle swarm optimization algorithm, the optimal solution is accurately searched in the global and local range, and finally the global optimal solution of multi-objective optimization is determined to achieve precise dosing, cost control and efficient treatment, thereby significantly improving the intelligence level of water treatment and the dynamic adaptability of the treatment process, and achieving more accurate dosing control in the water treatment process. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 The present invention is a schematic structural diagram of a precise dosing control system for water treatment. DETAILED DESCRIPTION

[0046] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0047] In order to achieve the above objectives, Figure 1 The present invention provides a structural schematic diagram of a precise dosing control system for water treatment, which specifically includes a data acquisition module, a condition library construction module, a multi-objective processing module and an optimization decision module, and each module is connected by a signal;

[0048] The data acquisition module is used to set sensors in the water body to collect pollutant parameters, pre-process the collected data, use the support vector machine classification model to identify the type of pollutants, and output the type and concentration of pollutants;

[0049] The condition library construction module is used to construct a removal condition library after identifying the type and concentration of pollutants, call the condition combination for water quality treatment according to the removal conditions stored in the removal condition library, and use the Apriori algorithm to update the condition combination. In the case of coexistence of multiple pollutants, the removal conditions are automatically adjusted according to the interaction rules between pollutants;

[0050] The multi-objective processing module is used to determine the optimization objectives and constraints of the water treatment process, integrate multiple objectives to obtain a multi-objective function, and construct a multi-objective optimization model for water treatment decision-making;

[0051] The optimization decision module is used to combine the genetic algorithm and the particle swarm optimization algorithm to determine the global optimal solution of multi-objective requirements in the water treatment decision-making process, and optimize the water treatment decision according to the global optimal solution.

[0052] Step 1: Identify pollutants and detect their concentration in real time. Identify pollutants in water and detect their concentration to obtain real-time information on pollutant types and concentrations. The specific steps are as follows:

[0053] Data collection and preprocessing are carried out. The basic parameters of pollutants are collected in real time through sensors installed in the water body, including conductivity (reflecting the ion concentration in the water), fluorescence intensity (used to identify specific organic matter), turbidity (indicating the content of suspended particles in the water), redox potential (reflecting the redox property of the water), pH, fluoride ion concentration, suspended matter concentration, total nitrogen, nitrate concentration, etc. The basic parameters of each pollutant will form a data set for subsequent processing;

[0054] After data collection, the system preprocesses the data, including removing noise from the data (for example, through filtering algorithms) and identifying and correcting abnormal data points. Through data preprocessing, it ensures that the parameters collected by the sensor can reflect the actual pollutant situation in the water body and reduce errors in subsequent identification;

[0055] Construct a pollutant identification model, identify the main pollutant types in wastewater based on real-time collected water parameters, and classify pollutants in combination with the pollutant feature library. The specific steps are as follows:

[0056] The collected pre-processed data (such as conductivity C, fluorescence intensity F, turbidity T, redox potential R, etc.) are converted into multi-dimensional feature vectors. Assuming that n types of pollutants are to be identified, the feature vector of each pollutant is defined as: You can use cross terms (such as C·F, T·R) as cross terms and nonlinear transformations (such as log(1+C), sin(F)) to enhance the expressiveness of features;

[0057] Perform feature library matching and build a feature library to store the feature ranges and classification labels of common pollutants. For example, for a certain pollutant P i , its characteristic value range in conductivity is defined as: C i ∈[C i，min , C i，max ], where C i，minrepresents the minimum characteristic value of the conductivity of pollutant i, C i，min is the maximum eigenvalue of conductivity. The eigenvalue ranges of fluorescence intensity, turbidity and redox potential are similar. When the eigenvector collected in real time is When the parameter values ​​in are within the corresponding pollutant characteristic range, it can be preliminarily judged that it may belong to a certain type of pollutant;

[0058] Based on the feature library and the collected data, the support vector machine (SVM) classification model is used to identify the types of pollutants. The goal of SVM is to find a classification hyperplane to separate the data points of different pollutant types. The optimization goal of SVM is expressed as: in, is the weight vector, b is the bias term, is the classification label of the i-th pollutant, is the feature vector of the ith pollutant; by adjusting the weight vector and the bias term, the model finds the optimal hyperplane so that each pollutant is divided in the feature space;

[0059] In order to improve the classification effect, the kernel function can be used to map the feature vector to a high-dimensional space. Commonly used kernel functions include polynomial kernels, radial basis kernels (RBF kernels), etc. For example, the radial basis kernel function is defined as: Among them, γ is the adjustment parameter, is the characteristic vector of the jth pollutant, which is used to control the range of the kernel function. Through the mapping of the kernel function, the model is classified in a high-dimensional space to improve the ability to identify complex pollutants;

[0060] The feature vector of the processed real-time collected data is input into the model to obtain the classification result and confidence of the pollutant. Assume that for a given feature vector, the pollutant type output by the classifier is P k , confidence level η i Calculated by the following formula: Among them, η i Indicates the confidence of the model in the classification result. A value close to 1 indicates a higher confidence, and a value close to 0 indicates a lower confidence.

[0061] After the above steps, the recognition model can classify the pollutants in the water and calculate the confidence level. The classification results and confidence level of the model will be passed to the next step of concentration calculation to provide basic data for subsequent dosing control.

[0062] According to the pollutant parameters of each pollutant (conductivity, fluorescence intensity, turbidity, redox potential, etc.), weighted calculation is directly performed to quickly obtain the concentration value.

[0063] It should be noted that the confidence value is used to evaluate the reliability of identification. If the confidence is lower than the set threshold, the system will mark the identification result as uncertain and reclassify the sample or conduct manual review. The weight coefficient is determined in advance through experimental data, and each coefficient represents the influence of the corresponding parameter on the pollutant concentration. The size of the weight is directly related to the sensitivity of the parameter to the pollutant concentration. It is usually determined based on historical data and experiments in the initial stage of the system. For example, organic pollutants are more sensitive to turbidity, so the weight coefficient of turbidity will be higher, while heavy metal pollutants may be more sensitive to redox potential, so the weight coefficient of redox potential will be higher.

[0064] Step 2: Construct and update the pollutant removal condition library. That is, after identifying the type and concentration of pollutants, the system needs to call the best removal conditions for each pollutant, and update and optimize the condition library in real time according to the historical data and dynamic changes in the water treatment process, thereby improving the dynamic adaptability of the treatment parameters and making the system more efficient in responding to different water quality fluctuations. The specific steps are as follows:

[0065] Initial construction of the removal condition library. In the initial stage of the system, a pollutant removal condition library is established based on the basic characteristics of the pollutants and treatment experience. For each pollutant P i , set the key removal parameters required for its removal, including: optimal pH value pH i , suitable temperature T i , recommended drug type D i , initial dosage Q i , reaction time t i ;

[0066] After identifying the pollutant P i Its concentration C i After that, the system automatically calls the condition combination in the removal condition library to remove the corresponding removal parameter {pH i , T i , D i , Q i , t i} is the initial treatment condition. If multiple pollutants are identified to coexist, the system will make preliminary adjustments to the treatment parameters based on their characteristics and treatment interactions. The specific expression is as follows: Among them, Q i，adj represents the adjusted pollutant P i The dosage, ρ ij The pollutant P i and P j The mutual influence coefficient between them is obtained by training historical data and represents the interaction between multiple pollutants;

[0067] An association rule mining algorithm (such as the Apriori algorithm) is embedded in the removal condition library to learn the association between different pollutant removal conditions. For example, the Apriori algorithm is used to mine the optimal treatment condition combination for the removal effect of coexisting pollutants in the multi-pollutant removal process. The specific steps are as follows:

[0068] Transactionalize historical processing data to form a data set containing pollutant types, removal conditions (pH value, temperature, dosage) and removal effects;

[0069] Apriori algorithm was applied to extract high-frequency condition combinations (such as {pH = 7, T = 25 °C, dosage = 50 mg / L}) and association rules with removal effects from transactional data.

[0070] The generated condition combination (rule) will be automatically updated to the condition library and become a new recommended parameter set.

[0071] Automatically update the condition library and construct interactive rules. Based on the historical data of wastewater treatment and the effect of pollutant treatment, dynamically adjust the optimal removal conditions of pollutants, and construct interactive rules based on the interactions between different pollutants. The specific steps are as follows:

[0072] The system regularly analyzes historical treatment data (removal parameters), covering parameters such as removal effect, reagent dosage, temperature, pH value, and treatment time. By analyzing historical data, the relationship between the actual removal efficiency of each pollutant and the existing removal conditions is identified. i，k represents the pollutant P in the kth treatment i For each pollutant, the system calculates the average removal effect of its condition library. Where N is the number of treatments in the historical data. If the removal effect of a pollutant is significantly lower than the set threshold, the system will mark its removal conditions and automatically adjust the removal parameters of the pollutant (such as pH value, temperature, dosage, etc.) based on the relationship between the conditions and the removal effect to improve the removal efficiency;

[0073] Based on historical data (removal parameters) analysis and removal effect optimization, the removal conditions of each pollutant in DIO are dynamically updated in the condition library. For example, if the optimal removal conditions (pH value, temperature, etc.) of a certain treatment are significantly different from the set conditions in the original condition library, the system will update this new optimal condition to the condition library and mark it as a priority call condition. This update process ensures the adaptability of the condition library in actual wastewater treatment, allowing the system to optimize the removal parameters in real time according to changes in water quality.

[0074] In the case of multi-pollutant synergistic treatment, the removal effects of different pollutants may affect each other. For example, some agents are effective against different pollutants at the same time, while other agents may weaken the effect due to chemical reactions between pollutants. The system constructs interaction rules to reflect the interaction between pollutants. The construction formula of the interaction rules can be: Among them, γ ij Indicates the pollutant P i and P j The interaction factors between Indicates P i and P j The removal effect when coexisting, P i and P j The removal effect when treated alone; when γ ij >0, indicating that there is a synergistic effect between the two pollutants. ij When <0, it indicates that there is an inhibitory effect between the two. The interaction factors are stored in the interaction rule table in the condition library, and these rules are applied in the actual treatment process to optimize the selection and dosage of the reagents.

[0075] It should be noted that the removal effect during single treatment refers to the removal rate measured after drug treatment when there is only one target pollutant in the wastewater. The measurement of this effect is used as the benchmark removal rate of pollutants under ideal conditions, which can be achieved through the design of single treatment experiments of pollutants; the removal effect during coexistence refers to the pollutant removal rate measured by the system during the actual treatment process under the condition of the simultaneous presence of multiple pollutants. This measurement is mainly used to evaluate the effect of coordinated treatment of multiple pollutants, which can be measured through coexistence treatment experiments of multiple pollutants.

[0076] The formation and real-time update of the dynamic response mechanism ensure that the system can adjust the treatment conditions in real time according to the changes in pollutant concentrations in the wastewater, fluctuations in environmental conditions, and the latest parameters in the condition library to improve the effect of synergistic removal of multiple pollutants. The specific steps are as follows:

[0077] After identifying the type and concentration of the pollutant, the system activates the dynamic response mechanism to call the latest removal conditions in the condition library, including the optimal pH value, temperature, agent type and dosage. The dynamic response mechanism matches the removal conditions in the condition library with the current pollutant concentration and decides whether to adjust the removal conditions according to the interactive rule table of the condition library. For example, the current pollutant P in the wastewater can be set. i The concentration of C i , pollutant P j The concentration of C j , when the interaction factor γ ij <0, the system adjusts the removal conditions through the following formula: Among them, Qi，adj is the adjusted dosage of the reagent, used to suppress interference, Q i The pollutant P i The initial dosage is adjusted to improve the adaptability of different pollutant removal conditions, avoid mutual interference, and improve the effect of synergistic removal;

[0078] The system continuously monitors the pollutant concentration, removal effect and environmental changes (such as temperature, pH, etc.) during the wastewater treatment process. When a significant fluctuation in concentration or environmental parameters is detected, the dynamic response mechanism automatically updates the current treatment parameters, calls the latest removal condition combination in the condition library, and adjusts the dosage of the reagent and the treatment conditions. For example, when the temperature drops beyond the preset threshold, the system automatically adjusts the reaction time and dosage to ensure the removal effect. The specific update formula is as follows: Among them, Q i，new is the updated dosage of the agent, T opt is the optimal treatment temperature for pollutants, T curent is the current temperature, δ is the adjustment coefficient, which is used to control the effect of temperature on the dosage. When the removal parameters or environmental conditions in the condition library change, the system immediately updates the current treatment conditions to ensure that the treatment effect remains optimal in the changing water environment.

[0079] Step 3: Establish a multi-objective optimization model. After completing pollutant identification, concentration calculation and removal condition library construction, the system establishes a multi-objective optimization model to achieve the best balance between pollutant removal efficiency, reagent cost, reaction time and environmental adaptability. This multi-objective optimization model enables the system to dynamically select the optimal dosing strategy. The specific steps for establishing the multi-objective optimization model are:

[0080] In the multi-objective optimization model to be established, the optimization objectives and constraints are determined. The optimization objectives can be set as multiple independent sub-objectives according to the actual operation requirements. The optimization objectives clarify the multi-dimensional objectives that the system needs to achieve in the pollutant removal process, such as removal efficiency, cost, processing time, etc.; the constraints ensure the safety and stability of the optimization process. Common optimization objectives include:

[0081] Maximize removal efficiency: Improve the removal efficiency of pollutants to ensure that the effluent quality meets the discharge standards. This optimization goal is mainly measured by the removal rate before and after treatment and is set as the goal to be maximized, that is, the higher the removal efficiency, the better;

[0082] Minimize reagent costs: Reduce the amount of reagents used to reduce the operating costs of wastewater treatment. The reagent cost target is mainly expressed by the amount of reagents added, that is, the lower the reagent cost, the better;

[0083] Minimize reaction time: shorten the treatment time to improve the treatment efficiency of the system. The reaction time refers to the time required for the reagent to be added to the effluent to meet the standard. The shorter the reaction time, the better.

[0084] Maximize environmental adaptability: Improve the system's ability to adapt to fluctuations in environmental parameters such as temperature and pH. This optimization goal is usually expressed by the fluctuation range of environmental parameters. The higher the environmental adaptability, the better.

[0085] In the process of water treatment optimization, the following constraints also need to be set:

[0086] Minimum removal rate constraint: the removal rate of each pollutant must meet the minimum emission standard; agent dosage constraint: the agent dosage must be controlled below the maximum dosage; reaction time constraint: the reaction time must be within the specified time; environmental condition constraint: pH value and temperature must be kept within the appropriate range;

[0087] According to the above optimization objectives, multiple optimization objectives are integrated into a multi-objective function, weight coefficients are set for each sub-objective, and each objective is converted into an objective function term. The multi-objective function form is as follows: Z = w1·f1(E i )-w2·f2(Q i )-w3·f3(t i )+w4·f4(ΔpH, ΔT);

[0088] in, represents the removal efficiency target term, C in，i , C out，i are the concentrations of pollutants before and after treatment, respectively;

[0089] Q is the target item of the reagent cost, which is defined as the sum of the reagent dosage multiplied by its cost. The lower the better. i is the dosage of the agent, Price Di is the price of the unit medicine;

[0090] is the reaction time target item, defined as the sum of the reaction times of each pollutant, the shorter the better, t i Indicates reaction time;

[0091] It is the environmental adaptability target item, which is defined as the sum of pH and temperature fluctuations. The lower the absolute value, the better. opt，i , T opt，i are the optimum pH value and temperature of the pollutant, pH current and T current is the pH and temperature of the current treatment conditions;

[0092] The weight coefficients w1, w2, w3, and w4 are used to adjust the priority of different optimization objectives. The weight values ​​are reasonably determined through actual operation conditions, experimental data, and demand analysis. The weight coefficients of different treatment stages are dynamically adjusted according to the actual treatment conditions. For example, when the removal efficiency requirement is higher, the weight of the removal efficiency objective item can be increased; when cost is prioritized, the weight of the reagent cost objective item can be increased.

[0093] The multi-objective optimization model can analyze the optimization objectives, constraints and results of the multi-objective function, comprehensively consider multiple key indicators (such as removal efficiency, reagent cost, reaction time, environmental adaptability, etc.), and achieve a balance between conflicting objectives. For example, under the premise of ensuring the removal efficiency, the multi-objective optimization model can reduce the operating cost by finding the amount of reagent used;

[0094] It should be noted that the multi-objective optimization model includes a multi-objective function, and on its basis, optimization objectives and constraints are added. The multi-objective optimization model defines the maximization or minimization method of the multi-objective function, and stipulates various constraints in the optimization process (such as minimum removal efficiency, maximum drug dosage, etc.) to improve the safety, practical feasibility and stability of the optimization process.

[0095] After constructing the multi-objective function, the optimization algorithm is selected to solve the multi-objective optimization model to obtain the parameter combination that makes the multi-objective function Z optimal. The genetic algorithm (GA) and particle swarm optimization algorithm (PSO) are combined to optimize the strategy and determine the global optimal solution of the multi-objective requirements in the wastewater treatment process. The following are the specific steps:

[0096] First, the genetic algorithm is used for preliminary global search to find potential good solutions, and the particle swarm optimization algorithm performs local fine search based on the genetic algorithm to accelerate convergence to the global optimal solution. The specific steps are as follows:

[0097] Use the initialized genetic algorithm to perform a global search and generate an initial population. Randomly generate an initial population containing multiple groups of candidate solutions. Each candidate solution represents a possible combination of treatment parameters (such as dosage of the agent, reaction time, pH value, temperature, etc.). These solutions will serve as the initial individuals of the genetic algorithm.

[0098] Calculate the fitness value. Calculate the fitness value of each individual (i.e., the value of the multi-objective function Z) according to the multi-objective function. The higher the fitness value, the closer the removal efficiency, reagent cost, reaction time, etc. of the combination scheme are to the optimal.

[0099] Use methods such as roulette or tournament selection to select individuals with higher fitness from the current population as the parents of the next generation. Individuals with higher fitness have a higher probability of being selected and entering the next generation.

[0100] Perform crossover operations on the selected parent individuals to generate new offspring individuals. The crossover rate is generally set to 0.6 to 0.8 to ensure the diversity of the population. Perform mutation operations on the genes of some offspring individuals (such as adjusting the value of a processing parameter). The mutation rate is generally set to 0.01 to 0.05 to prevent the algorithm from falling into a local optimum.

[0101] Repeat the fitness calculation, selection, crossover and mutation operations, and perform multiple generations of iterative evolution. The genetic algorithm will converge to the initial optimal solution after multiple iterations, that is, obtain a set of candidate solutions with higher fitness (i.e., excellent solution set), which have higher removal efficiency and lower reagent cost and processing time;

[0102] The excellent solution set obtained by the genetic algorithm is used as the initial particle swarm of particle swarm optimization, and the particle swarm optimization performs local fine search. The specific steps are as follows:

[0103] Initialize the position and speed of the particle swarm, use the excellent solution set output by the genetic algorithm as the initial position of the particle swarm, each particle represents a candidate solution, and the particle speed is initialized to a small value so that fine adjustments can be made near the initial solution;

[0104] Calculate the fitness value of each particle (i.e., the value of the multi-objective function Z) to measure the quality of the current solution and record the optimal position of each particle (the individual optimal position p best and the global optimal position of the particle swarm (global optimal position g best );

[0105] According to the velocity update formula of PSO, the particle position is dynamically adjusted to gradually approach the global optimal solution. The velocity update formula is: i =ω·v i +c1·r1·(p best -x i )+c2·r2·(g best -x i ), where ω is the inertia weight, which controls the influence of the current speed and usually decreases with the number of iterations; c1 and c2 are acceleration factors, which are used to adjust the speed of particles moving towards the individual optimal and global optimal directions; r1 and r2 are random numbers, which ensure the randomness of particles in the search process and avoid falling into the local optimal state; x i Indicates the current location;

[0106] Perform iterative updates and approach the global optimum, repeat the fitness calculation, speed update and position update process, continuously update the individual optimal position and global optimal position of the particles, and as the number of iterations increases, the particle swarm gradually converges to the global optimal solution;

[0107] When the particle swarm reaches the maximum number of iterations, or the global optimal position does not change in several consecutive iterations, the algorithm terminates. At this time, the global optimal position g in the particle swarm is best That is the optimal solution of the multi-objective optimization model;

[0108] After PSO is completed, the final optimal solution is output, including the optimal dosage of reagents, reaction time, temperature and pH value, etc. This solution meets all the requirements of the multi-objective optimization model (such as high removal efficiency, low cost, short processing time, strong environmental adaptability, etc.);

[0109] The optimal solution obtained by hybrid optimization is applied to the actual wastewater treatment process to verify whether the key indicators such as removal efficiency, cost and time in actual operation meet the expectations; if the actual operation results deviate from the expected values ​​of the optimization model, the inertia weight of PSO and the crossover and mutation rate of GA can be further adjusted to optimize the algorithm's solution process in order to improve the convergence and accuracy of the hybrid algorithm.

[0110] By combining genetic algorithms and particle swarm optimization, the system can perform global and local fine searches in multi-objective optimization to achieve fast and accurate optimal solutions. The genetic algorithm is responsible for the initial global search to find a good solution set, and the particle swarm optimization algorithm further optimizes through local search on this basis, and finally obtains the global optimal solution for multi-objective optimization, providing accurate and economical treatment parameters for the water treatment process.

[0111] The optimal solution obtained by hybrid optimization is applied to actual wastewater treatment, and key indicators such as removal rate, reagent cost and treatment time are measured to ensure that they meet the expected requirements. If the actual effect deviates greatly, the parameters of GA and PSO can be adjusted. Through this optimization method, the system realizes efficient, economical and adaptable dosing control of wastewater treatment, and improves the accuracy and stability of water treatment.

[0112] It should be noted that the threshold information in this embodiment is pre-set by professionals and will not be explained in detail here. Some parameter English letters in the embodiments have the same situation, but different meanings are explained when used, which will not be explained one by one here.

[0113] The present invention sets sensors in the water body to collect pollutant parameters (such as conductivity, fluorescence intensity, turbidity and redox potential), pre-processes the data, and uses the support vector machine classification model to identify the type and concentration of pollutants. After identifying the pollutants, the system calls the corresponding treatment parameter combination from the removal condition library, and updates the condition combination through the Apriori algorithm to dynamically optimize the condition library. At the same time, when multiple pollutants coexist, the removal conditions are automatically adjusted according to the interaction rules between the pollutants to ensure the synergistic removal effect of different pollutants.

[0114] A multi-objective optimization model is further constructed, with removal efficiency, chemical cost, treatment time and environmental adaptability as optimization targets, integrating multi-objective requirements to meet the overall needs of water treatment. Through the combination of genetic algorithm and particle swarm optimization algorithm, the optimal solution is accurately searched in the global and local range, and finally the global optimal solution of multi-objective optimization is determined to achieve precise dosing, cost control and efficient treatment, thereby significantly improving the intelligence level of water treatment and the dynamic adaptability of the treatment process, and achieving more accurate dosing control in the water treatment process.

[0115] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0116] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product.

[0117] Those of ordinary skill in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0118] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0119] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0120] Finally: The above description is only 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 in the protection scope of the present invention.

Claims

1. A precise dosing control system for water treatment, characterized in that: It includes data acquisition module, condition library construction module, multi-objective processing module and optimization decision module, and each module is connected through signals; The data acquisition module is used to set sensors in the water body to collect pollutant parameters, pre-process the collected data, use the support vector machine classification model to identify the type of pollutants, and output the type and concentration of pollutants; The condition library construction module is used to construct a removal condition library after identifying the type and concentration of pollutants, call the condition combination for water quality treatment according to the removal conditions stored in the removal condition library, and use the Apriori algorithm to update the condition combination. In the case of coexistence of multiple pollutants, the removal conditions are automatically adjusted according to the interaction rules between pollutants; The multi-objective processing module is used to determine the optimization objectives and constraints of the water treatment process, integrate multiple objectives to obtain a multi-objective function, and construct a multi-objective optimization model for water treatment decision-making; The optimization decision module is used to combine the genetic algorithm and the particle swarm optimization algorithm to determine the global optimal solution of multi-objective requirements in the water treatment decision-making process, and optimize the water treatment decision according to the global optimal solution.

2. The precise dosing control system for water treatment according to claim 1, characterized in that: It is used to set the sensor to collect pollutant parameters in the water body, pre-process the collected data, use the support vector machine classification model to identify the pollutant types, and output the pollutant types and concentrations. The specific steps are as follows: Pollutant parameters include conductivity, fluorescence intensity, turbidity, and redox potential; Remove noise from collected data, identify and correct abnormal data points, and convert preprocessed data into feature vectors; Build a feature library to store the feature ranges and classification labels of pollutants; Based on the feature library and collected data, the radial basis kernel function is used to map the feature vectors of the collected data. According to the classification labels of pollutants in the feature library and the feature vectors of pollutants, the support vector machine classification model is used to identify the types of pollutants, and the classification results, confidence levels and concentration values ​​of pollutants are obtained.

3. The precise dosing control system for water treatment according to claim 2, characterized in that: After identifying the type and concentration of pollutants, a removal condition library is constructed, and a condition combination is called for water quality treatment according to the removal conditions stored in the removal condition library, and the condition combination is updated using the Apriori algorithm. The specific steps include: Initial construction of the removal condition library: According to the basic characteristics of pollutants and treatment experience, a pollutant removal condition library is established. For each pollutant, the removal parameters required for removal are set. The removal parameters include the optimal pH value, suitable temperature, recommended reagent type, initial dosage, and reaction time. After identifying the pollutants and their concentrations, the corresponding removal parameters in the condition combination in the removal condition library are automatically called; The Apriori algorithm is embedded in the removal condition library to learn the correlation between different pollutant removal conditions, determine the best treatment condition combination, and update the generated condition combination to the condition library.

4. The precise dosing control system for water treatment according to claim 3, characterized in that: For the coexistence of multiple pollutants, the removal conditions are automatically adjusted according to the interaction rules between pollutants. The specific steps are as follows: Analyze the historical removal parameters, including removal effect, reagent dosage, temperature, pH value, and treatment time, to determine the actual removal efficiency of each pollutant and the average removal effect between the existing removal conditions; Based on the relationship between conditions and removal effects, pollutant removal parameters are adjusted. Based on removal parameter analysis and removal effect optimization, the removal conditions for each pollutant are dynamically updated in the condition library. In the case of coordinated treatment of multiple pollutants, interaction rules are constructed to determine the interactions between pollutants, and the selection and dosage of reagents are optimized based on the interaction factors.

5. The precise dosing control system for water treatment according to claim 4, characterized in that: It is used to determine the optimization objectives and constraints of the water treatment process, integrate multiple objectives to obtain a multi-objective function, and construct a multi-objective optimization model for water treatment decision-making, including the following steps: The optimization objectives include maximizing removal efficiency, minimizing reagent costs, minimizing reaction time, and maximizing adaptability to environmental conditions; The constraints include minimum removal rate constraint, reagent dosage constraint, reaction time constraint and environmental condition constraint; Integrate the calculations according to the optimization objectives to form a multi-objective function; The multi-objective function results, optimization objectives and constraints are combined to form a multi-objective optimization model, and water treatment decisions are made.

6. The precise dosing control system for water treatment according to claim 5, characterized in that: It is used to combine the genetic algorithm and the particle swarm optimization algorithm to determine the global optimal solution of multi-objective requirements in the water treatment decision-making process, and optimize the water treatment decision according to the global optimal solution, including the following steps: Use the initialized genetic algorithm to perform global search, randomly generate an initial population containing multiple sets of candidate solutions, and use the candidate solutions as the initial individuals of the genetic algorithm; Calculate the fitness value of each initial individual according to the multi-objective function; Use roulette or tournament selection methods to select individuals from the population based on their fitness values ​​to serve as parents for the next generation; Perform a crossover operation on the selected parent individuals to generate new offspring individuals, and perform a mutation operation on the genes of the selected offspring individuals; Repeat the fitness calculation, selection, crossover and mutation operations for multiple generations of iterative evolution. The genetic algorithm will converge to the initial optimal solution as the excellent solution set after multiple iterations. The excellent solution set obtained by the genetic algorithm is used as the initial particle swarm of particle swarm optimization, and the particle swarm optimization performs local fine search. The specific steps are as follows: Initialize the position and speed of the particle swarm, use the excellent solution set output by the genetic algorithm as the initial position of the particle swarm, each particle represents a candidate solution, and set the particle speed; Calculate the fitness value of each particle to measure the quality of the current solution, and record the individual optimal position of each particle and the global optimal position of the particle swarm; According to the speed update formula of the particle swarm optimization algorithm, the particle position is dynamically adjusted; Repeat the fitness value calculation, velocity update and position update process to update the individual optimal position and global optimal position of the particle; When the particle swarm reaches the maximum number of iterations or the global optimal position no longer changes in consecutive iterations, the algorithm terminates and the global optimal position in the particle swarm is taken as the optimal solution of the multi-objective optimization model; Output the final optimal solution, which includes the optimal dosage of reagent, reaction time, temperature and pH value.

Images