A Multi-Objective Optimization Control Method for Wastewater Treatment Process Based on Knowledge Selection

Abstract

This invention relates to the field of chemical or physical variable control technology, and discloses a multi-objective optimization control method for wastewater treatment based on knowledge selection. The invention monitors the substances in the wastewater within a wastewater storage tank by installing a series of online monitoring devices and coordinating these devices. Simultaneously, it reduces data redundancy by collecting data from different wastewater samples from each online monitoring device and preprocessing the collected data. Furthermore, it extracts features from the wastewater data by inputting the processed data into an Elman neural network. The invention also improves the processing accuracy of the Elman neural network by optimizing it with a non-dominated genetic algorithm. Finally, it adjusts the wastewater treatment process based on the results of the Elman neural network processing, thereby improving the accuracy and efficiency of the wastewater treatment process.

Description

Technical Field

[0001] This invention relates to the field of chemical or physical variable control technology, specifically a multi-objective optimization control method for wastewater treatment processes based on knowledge selection. Background Technology

[0002] Urban wastewater treatment is a typical energy-intensive industry. With the increasing scale of urban wastewater treatment and the continuous improvement of wastewater treatment efficiency in my country, its operating energy consumption is also getting higher and higher.

[0003] The existing technology CN103597418B monitors the pH of the material in the process to be controlled by using a nonlinear model of the pH of the material, standard orthogonal basis functions and ordinal spline basis functions, and controls the pH of the material by adding substances to neutralize it. However, this technology has great limitations and is insufficient to meet practical needs.

[0004] Furthermore, urban wastewater treatment processes passively accept factors such as influent water quality and quantity, and the operating environment, while also being affected by weather changes, temperature, precipitation, and pH levels, resulting in dynamic nonlinearity. Therefore, urban wastewater treatment is an operation characterized by strong nonlinearity, dynamic uncertainty, and strong coupling. Moreover, due to the lack of effective optimization and control strategies, satisfactory operational results are difficult to achieve in urban wastewater treatment processes. Summary of the Invention

[0005] (a) Technical problems to be solved

[0006] To address the shortcomings of existing technologies, this invention provides a multi-objective optimization control method for wastewater treatment processes based on knowledge selection. This method has advantages such as high efficiency and stability, and solves the problem that urban wastewater treatment processes are difficult to achieve satisfactory operating results due to the lack of effective optimization control strategies.

[0007] (II) Technical Solution

[0008] To address the aforementioned technical problem that the lack of effective optimization control strategies makes it difficult to achieve satisfactory operational results in urban wastewater treatment processes, this invention provides the following technical solution:

[0009] This embodiment discloses a multi-objective optimization control method for wastewater treatment processes based on knowledge selection, specifically including the following steps:

[0010] S1. Data on the main substances in wastewater are collected in real time by installing a series of detection devices in the wastewater storage tank.

[0011] S2. Based on the installation of a series of detection devices in the wastewater storage tank to collect classified wastewater data in real time and perform data preprocessing, the preprocessed wastewater data is obtained.

[0012] S3. Optimize the initial weights and thresholds of the Elman neural network using a non-dominated genetic algorithm to obtain an optimized Elman neural network; process the pre-treated wastewater data using the optimized Elman neural network to obtain wastewater data processed by the Elman neural network, the wastewater data including the content data of each component in the wastewater;

[0013] S4. Analyze the wastewater data processed by the Elman neural network, combining the weights and the set thresholds.

[0014] S5. Based on the comparison results, the wastewater is directed to the corresponding wastewater treatment center for treatment.

[0015] Preferably, the step of collecting data on the main substances in the wastewater in real time by installing a series of detection devices in the wastewater storage tank includes:

[0016] The installed series of testing equipment includes: an online pH meter, an online dissolved oxygen meter, and an online component analyzer;

[0017] The online pH meter is used to monitor the pH value of wastewater in real time;

[0018] The online component analyzer is used to monitor the concentration changes of various components in wastewater in real time.

[0019] The online dissolved oxygen meter is used to monitor the concentration of dissolved oxygen in wastewater in real time;

[0020] Based on the concentration of dissolved oxygen in wastewater measured by an online dissolved oxygen meter, the collected wastewater is distinguished by setting a concentration threshold for dissolved oxygen in the wastewater;

[0021] The dissolved oxygen concentration was set at >2.0 mg / L under aerobic conditions, <0.2 mg / L under anaerobic conditions, 0.2-0.5 mg / L under hypoxic conditions, and 0.5-2.0 mg / L under normal conditions.

[0022] Wastewater is initially classified into anaerobic, anoxic, and aerobic categories.

[0023] Preferably, the step of collecting classified wastewater data in real time by installing a series of detection devices in the wastewater storage tank and performing data preprocessing to obtain preprocessed wastewater data includes:

[0024] The wastewater data collected in real time and classified is stored in matrix form;

[0025] The auxiliary variable data sample matrix X is obtained according to Formula 1. m×n Mean and variance:

[0026]

[0027] The sample matrix is ​​standardized to zero mean according to Formula 2, and the standardized matrix is ​​calculated.

[0028]

[0029] Where m is the number of samples, n is the number of sample components, and x ij X represents the j-th component of the i-th sample; j S is the mean of the i-th sample component. j This represents the standard deviation of the i-th sample;

[0030] The standardized matrix Z is calculated based on formulas 3 and 4. m×n covariance matrix R m×n ;

[0031]

[0032] Where r represents the covariance and T represents the matrix transpose;

[0033] The different eigenvalues ​​λ of R can be obtained by solving formula 5. j (j=1,2,...,n), and arrange the n eigenvalues ​​of R in descending order;

[0034]

[0035] Where λJ represents the sum of different eigenvalues ​​λ j The matrix formed by these components, where R represents the covariance matrix;

[0036] The result is used to calculate the unit eigenvector b corresponding to the corresponding eigenvalue using Formula 6. j (j=1,2,...,n), b j =(b 1j b 2j , ..., b nj );

[0037] Rb=λ J b (6)

[0038] The cumulative variance contribution rate of the principal components is calculated based on the calculated eigenvalues, and the number of principal components k is determined based on the cumulative variance contribution rate.

[0039]

[0040] According to Formula 8, the standardized matrix Z is... m×n Projected onto k-dimensional coordinates:

[0041]

[0042] Where U1 is the first principal component, U2 is the second principal component, and U... k It is the k-th principal component.

[0043] Preferably, the optimization of the initial weights and thresholds of the Elman neural network using a non-dominated genetic algorithm to obtain the optimized Elman neural network specifically includes the following steps:

[0044] The objective function for the weights is expressed by the following formula:

[0045]

[0046] Among them, w max w min These represent the maximum and minimum values ​​of the weights, w. max =0.9, w min =0.4; k is the number of iterations, k max This represents the maximum number of iterations.

[0047] By setting the initial basic parameters of the Elman neural network and encoding the weights and thresholds of the Elman neural network;

[0048] S31. Initialize the population, set basic parameters such as maximum number of iterations, population size, crossover probability, and mutation probability, and set appropriate constraints on input and output variables.

[0049] S32. Perform fast non-dominated sorting on the population and calculate the crowding degree, then proceed to the algorithm iteration process;

[0050] S33. Perform a selection operation based on the non-dominated hierarchy ranking and crowding degree of individuals. For individuals with the same non-dominated hierarchy ranking, select those with lower crowding degree; for individuals with different non-dominated hierarchy rankings, select those with higher crowding degree.

[0051] S34, Crossover and Mutation Operations;

[0052] S35. Merge populations, iterate and filter: Merge the parent and child populations from step S34 to generate a new merged population. Use the method described in step S33 to filter and generate a new population. Stop when the maximum number of iterations is reached, and the algorithm terminates.

[0053] Preferably, the fast non-dominated sorting specifically includes the following steps:

[0054] S321. Traverse all solutions in the set and calculate the corresponding sets n and S for each solution. p S represents the number of dominant solutions p, and is a single numerical value; p Let p represent the solutions dominated by solution p, which is a set.

[0055] S322. Find the solution where n=0, assign it to set F1, and rank its corresponding non-dominated layer by 1.

[0056] S323. Iterate through each individual j in set F1 and find the solution set S dominated by individual j. j and S j n corresponding to each individual h h Subtract 1;

[0057] S324, Find S j The corresponding n h Find the solution that equals 0, assign it to set F2, and assign its corresponding non-dominated layer sort value to 2. Repeat steps S322-S324 to complete the non-dominated sorting operation for all individuals in the set.

[0058] Preferably, the congestion calculation method is as follows:

[0059] Set the population size to N, initialize the crowding distance for each individual in the population, and set n... d =0, n∈1,2,...,N;

[0060] Define each objective function f m Perform a quick non-dominated sort, setting f m max For the individual objective function f m The maximum value of f m min For the individual objective function f m The minimum value;

[0061] The crowding of solutions on the boundary of each objective function is set to infinity, as follows:

[0062] I(d1)=∞,I(d n )=∞ (10)

[0063] The crowding level of the remaining individuals is:

[0064]

[0065] Where I(d1) represents the solution of the first individual in the objective function, I represents the crowding degree, and k represents the kth individual.

[0066] Preferably, the binary crossover process is as follows:

[0067]

[0068] Where, p i，k For its parent individual, c i，kThe offspring individuals generated by its crossover, i, k represent the k-th individual of the i-th generation, β k ≥0 represents an individual that undergoes crossover, and its probability density function is:

[0069]

[0070] Where, η c Let c represent the c-th offspring individual, and β represent the crossover operation;

[0071] The process of generating offspring through polynomial mutation is achieved using the polynomial mutation operator, as shown in the following formula:

[0072]

[0073] Where u∈[0,1], c k p represents the offspring individuals resulting from the mutation. k Represents its parent individual, δ k It can be expressed by the following formula:

[0074]

[0075]

[0076] Where, r k η represents the non-dominated order of an individual. m This is the variation distribution index.

[0077] Preferably, the pretreated wastewater data is processed using an optimized Elman neural network to obtain...

[0078] The wastewater data processed by the Elman neural network includes:

[0079] The structure of an Elman neural network includes an input layer, hidden layers, a supporting layer, and an output layer.

[0080] Furthermore, the nonlinear state expression of the Elman neural network structure is as follows:

[0081] x(t)=f(w1x c (t)+w2u(t-1)) (19)

[0082] x c (t)=x(t-1) (20)

[0083] y t =g(w3x(t)) (21)

[0084] Where r, n, and m represent the number of nodes in the input layer, hidden layer, and output layer, respectively, and the number of nodes in the connecting layer is the same as the number of nodes in the hidden layer, denoted as n, x(t), x c y(t) and y(t) represent the outputs of the hidden layer, the receiving layer, and the output layer at time t, respectively; w1 represents the connection weight from the receiving layer to the hidden layer; w2 represents the connection weight from the input layer to the hidden layer; w3 represents the connection weight from the hidden layer to the output layer; g() represents the transfer function of the output layer; f() represents the activation function of the hidden layer; and u represents the input vector of the input layer.

[0085] Furthermore, the activation function of the hidden layer is the Sigmoid function:

[0086]

[0087] g() takes a linear function:

[0088] y t =w3x(t) (23)

[0089] Furthermore, the Elman neural network's learning algorithm employs gradient descent, and its error index function is expressed as follows:

[0090]

[0091] Among them, y d y(q) represents the actual output of the Elman neural network at step q, and y(q) represents the input of the Elman neural network at step q.

[0092] Preferably, the analysis based on the wastewater data processed by the Elman neural network, combined with weights and a set threshold, includes:

[0093] Based on the comparison results after combining weights and set thresholds, the top n principal components with a cumulative variance contribution rate greater than 85% are selected, and the top n principal components are sorted according to their contribution rates. The principal components in the wastewater are then treated sequentially based on the sorting.

[0094] After treatment, components with a cumulative variance contribution rate of not less than 15% are sorted according to their contribution rate, and wastewater with the same component ranking is mixed.

[0095] Preferably, directing the wastewater to the corresponding wastewater treatment center for treatment based on the comparison results includes:

[0096] After the first n principal components with a cumulative variance contribution rate greater than 85% in the wastewater are treated, the treated wastewater is subjected to the S2-S5 operations again until it meets the wastewater treatment standards.

[0097] (III) Beneficial Effects

[0098] Compared with existing technologies, this invention provides a multi-objective optimization control method for wastewater treatment processes based on knowledge selection, which has the following beneficial effects:

[0099] 1. This invention monitors wastewater components in real time using a series of online detection devices. First, based on an online dissolved oxygen meter, the collected wastewater is differentiated by setting a dissolved oxygen concentration threshold to determine whether it is aerobic or anaerobic. Simultaneously, according to the online dissolved oxygen meter's criteria, the wastewater is transferred to the corresponding environment. The online detection devices then monitor changes in various components within the wastewater in real time. This differentiation and transfer of collected wastewater reduces the influence of environmental parameters on wastewater composition and improves the accuracy of the real-time monitoring equipment. This invention designs an optimized control strategy for the wastewater treatment process, addressing the challenging problem of efficient and stable operation currently faced by urban wastewater treatment plants. It aims to improve the operational efficiency of the treatment process and enhance the competitiveness of the wastewater treatment industry by researching optimized control theories and technologies for urban wastewater treatment processes. This invention differentiates wastewater from multiple perspectives, measuring the content of various components and taking corresponding treatments, thereby improving the safety and effectiveness of wastewater treatment.

[0100] 2. This invention processes the data collected by online detection equipment by constructing a matrix for storage and processing. It then calculates the eigenvalues ​​of the matrix by solving for the variance of the constructed matrix, calculates the cumulative variance contribution rate of the principal components based on the eigenvalues, determines the number of principal components based on the cumulative variance contribution rate, and determines the order of wastewater treatment based on the contribution of the principal components, thus ensuring the effectiveness and completeness of the wastewater treatment process.

[0101] 3. This invention transmits the processed wastewater matrix data to an Elman neural network. The Elman neural network's input layer, hidden layer, continuation layer, and output layer, along with weights and thresholds, further process the wastewater data, thereby reducing wastewater treatment errors.

[0102] 4. This invention optimizes the initial weights and thresholds of the Elman neural network by using a non-dominated genetic algorithm. Through crossover mutation and iteration of the weights and thresholds encoded in the input Elman neural network, the multi-objective optimization problem is transformed into the problem of optimizing a fitness function, which reduces the complexity of the algorithm and improves the running speed. Attached Figure Description

[0103] Figure 1 This is a schematic diagram of the multi-objective optimization process structure for wastewater treatment based on knowledge selection according to the present invention;

[0104] Figure 2This is a schematic diagram of the process structure for optimizing the initial weights and thresholds of an Elman neural network based on a non-dominated genetic algorithm, as described in this invention. Detailed Implementation

[0105] 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.

[0106] This embodiment discloses a multi-objective optimization control method for wastewater treatment processes based on knowledge selection, specifically including the following steps:

[0107] S1. By installing a series of detection devices in the wastewater storage tank, data on the main substances in the wastewater are collected in real time, and the wastewater is initially classified.

[0108] The installed series of testing equipment includes: online pH meter, online dissolved oxygen meter, and online component analyzer;

[0109] The online pH meter is used to monitor the pH value of wastewater in real time;

[0110] The online component analyzer is used to monitor the concentration changes of various components in wastewater in real time.

[0111] The online dissolved oxygen meter is used to monitor the concentration of dissolved oxygen in wastewater in real time;

[0112] Based on the concentration of dissolved oxygen in wastewater measured by an online dissolved oxygen meter, the collected wastewater is distinguished by setting a concentration threshold for dissolved oxygen in the wastewater;

[0113] The dissolved oxygen concentration was set at >2.0 mg / L under aerobic conditions, <0.2 mg / L under anaerobic conditions, 0.2-0.5 mg / L under hypoxic conditions, and 0.5-2.0 mg / L under normal conditions.

[0114] Wastewater is initially classified based on anaerobic, anoxic, and aerobic conditions.

[0115] S2. Based on the installation of a series of detection devices in the wastewater storage tank to collect classified wastewater data in real time and perform data preprocessing, the preprocessed wastewater data is obtained, including the content data of each component in the wastewater.

[0116] The pretreated wastewater data is stored in matrix form;

[0117] The auxiliary variable data sample matrix X is obtained according to Formula 1.m×n Mean and variance:

[0118]

[0119] Furthermore, the sample matrix is ​​standardized to zero mean according to Formula 2 to obtain the standardized matrix;

[0120]

[0121] Where m is the number of samples, n is the number of sample components, and x ij X represents the j-th component of the i-th sample; j S is the mean of the i-th sample component. j This represents the standard deviation of the i-th sample;

[0122] Furthermore, the standardized matrix Z is calculated according to Formulas 3 and 4. m×n covariance matrix R m×n ;

[0123]

[0124] Where r represents the covariance and T represents the matrix transpose;

[0125] Furthermore, the different eigenvalues ​​λ of R are solved according to Formula 5. j (j=1,2,...,n), and arrange the n eigenvalues ​​of R in descending order;

[0126] R-λ J E|=0 (5)

[0127] Where λJ represents the sum of different eigenvalues ​​λ j The matrix formed by these components, where R represents the covariance matrix;

[0128] Furthermore, the results are used to calculate the unit eigenvector b corresponding to the respective eigenvalues ​​using Formula 6. j (j=1,2,...,n), b j =(b 1j b 2j , ..., b nj );

[0129] Rb=λ J b (6)

[0130] Furthermore, the cumulative variance contribution rate of the principal components is calculated based on the calculated eigenvalues, and the number of principal components k is determined according to the cumulative variance contribution rate.

[0131]

[0132] Furthermore, according to Formula 8, the standardized matrix Z is... m×n Projected onto k-dimensional coordinates:

[0133] U i,j =Z i T b j ,i=1,2...,m,j=1,2...,k (8)

[0134] Where U1 is the first principal component, U2 is the second principal component, and U... k It is the k-th principal component;

[0135] Furthermore, the principal components in the wastewater are sorted in descending order according to their cumulative variance contribution rate, and the component with the highest cumulative variance contribution rate is treated first.

[0136] S3. Optimize the initial weights and thresholds of the Elman neural network using a non-dominated genetic algorithm to obtain an optimized Elman neural network; process the pre-treated wastewater data using the optimized Elman neural network to obtain wastewater data processed by the Elman neural network.

[0137] Furthermore, the initial weights and thresholds of the Elman neural network are optimized using a genetic algorithm, specifically including the following steps:

[0138] The objective function for the weights is expressed by the following formula:

[0139]

[0140] Among them, w max w min These represent the maximum and minimum values ​​of the weights, w. max =0.9, w min =0.4; k is the number of iterations, k max This represents the maximum number of iterations.

[0141] Furthermore, the initial basic parameters of the Elman neural network are set, and the weights and thresholds of the Elman neural network are encoded.

[0142] S31. Initialize the population, set basic parameters such as maximum number of iterations, population size, crossover probability, and mutation probability, and set appropriate constraints on input and output variables.

[0143] S32. Perform fast non-dominated sorting on the population and calculate the crowding degree, then proceed to the algorithm iteration process;

[0144] The fast nondominated sort specifically includes the following steps:

[0145] S321. Traverse all solutions in the set and calculate the corresponding sets n and S for each solution. p S represents the number of dominant solutions p, and is a single numerical value; p Let p represent the solutions dominated by solution p, which is a set.

[0146] S322. Find the solution where n=0, assign it to set F1, and rank its corresponding non-dominated layer by 1.

[0147] S323. Iterate through each individual j in set F1 and find the solution set S dominated by individual j. j and S j n corresponding to each individual h h Subtract 1;

[0148] S324, Find S j The corresponding n h Find the solutions that equal to 0, assign them to set F2, and assign the corresponding non-dominated layer sort value to 2. Repeat steps S422-S424 to complete the non-dominated sorting operation for all individuals in the set.

[0149] S33. Perform a selection operation based on the non-dominated hierarchy ranking and crowding degree of individuals. For individuals with the same non-dominated hierarchy ranking, select those with lower crowding degree; for individuals with different non-dominated hierarchy rankings, select those with higher crowding degree.

[0150] The congestion level is calculated as follows:

[0151] Set the population size to N, initialize the crowding distance for each individual in the population, and set n... d =0, n∈1,2,...,N;

[0152] Define each objective function f m Perform a quick non-dominated sort, setting f m max For the individual objective function f m The maximum value of f m min For the individual objective function f m The minimum value;

[0153] The crowding of solutions on the boundary of each objective function is set to infinity, as follows:

[0154] I(d1)=∞,I(d n )=∞ (10)

[0155] The crowding level of the remaining individuals is:

[0156]

[0157] Where I(d1) represents the solution of the first individual in the objective function, I represents the crowding degree, and k represents the kth individual;

[0158] S34, Crossover and Mutation Operations;

[0159] Data is recombined and mutated by simulating binary crossover and polynomial mutation methods;

[0160] Furthermore, the binary crossover process is as follows:

[0161]

[0162] Where, p i，k For its parent individuals, c i，k The offspring individuals generated by its crossover, i, k represent the k-th individual of the i-th generation, β k ≥0 represents an individual that undergoes crossover, and its probability density function is:

[0163]

[0164] Where, η c Let c represent the c-th offspring individual, and β represent the crossover operation;

[0165] Furthermore, the process of generating offspring through polynomial mutation is achieved using the polynomial mutation operator, as shown in the following formula:

[0166]

[0167] Where u∈[0,1], c k p represents the offspring individuals resulting from the mutation. k Represents its parent individual, δ k It can be expressed by the following formula:

[0168]

[0169] Where, r k η represents the non-dominated order of an individual. m It is the distribution index of variation;

[0170] S35. Merge populations, iterate and filter: Merge the parent and child populations from step S34 to generate a new merged population. Use the method described in step S33 to filter and generate a new population. Stop when the maximum number of iterations is reached and the algorithm terminates.

[0171] Furthermore, the wastewater data is processed using an optimized Elman neural network;

[0172] The structure of an Elman neural network includes an input layer, hidden layers, a supporting layer, and an output layer.

[0173] Furthermore, the nonlinear state expression of the Elman neural network structure is as follows:

[0174] x(t)=f(w1x c (t)+w2u(t-1)) (19)

[0175] x c (t)=x(t-1) (20)

[0176] y t =g(w3x(t)) (21)

[0177] Where r, n, and m represent the number of nodes in the input layer, hidden layer, and output layer, respectively, and the number of nodes in the connecting layer is the same as the number of nodes in the hidden layer, denoted as n, x(t), x c (t), y t Let w1 represent the outputs of the hidden layer, the receiving layer, and the output layer at time t, respectively; w2 represent the connection weights from the receiving layer to the hidden layer; w3 represent the connection weights from the hidden layer to the output layer; g() represents the transfer function of the output layer; f() represents the activation function of the hidden layer; and u represents the input vector of the input layer.

[0178] Furthermore, the activation function of the hidden layer is the Sigmoid function:

[0179]

[0180] g() takes a linear function:

[0181] y t =w3x(t) (23)

[0182] Furthermore, the Elman neural network's learning algorithm employs gradient descent, and its error index function is expressed as follows:

[0183]

[0184] Among them, y d y(q) represents the actual output of the Elman neural network at step q, and y(q) represents the input of the Elman neural network at step q.

[0185] S4. Analyze the wastewater data processed by the Elman neural network, combining the weights and the set thresholds.

[0186] Based on the comparison results after combining weights and set thresholds, the top n principal components with a cumulative variance contribution rate greater than 85% are selected, and the top n principal components are sorted according to their contribution rates. The principal components in the wastewater are then treated sequentially based on the sorting.

[0187] After treatment, components with a cumulative variance contribution rate of not less than 15% are sorted according to their contribution rate, and wastewater with the same component ranking is mixed.

[0188] S5. Based on the comparison results, the wastewater is directed to the corresponding wastewater treatment center for treatment.

[0189] Furthermore, after the first n principal components with a cumulative variance contribution rate greater than 85% in the wastewater are treated, the treated wastewater is subjected to the S2-S5 operations again until the wastewater treatment standard is met.

[0190] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A multi-objective optimization control method for wastewater treatment processes based on knowledge selection, characterized in that, Includes the following steps: S1. Data on the main substances in wastewater are collected in real time by installing a series of detection devices in the wastewater storage tank. Specifically, this includes: using an online dissolved oxygen meter to differentiate collected wastewater by setting a concentration threshold for dissolved oxygen in the wastewater, determining whether the collected wastewater is aerobic or anaerobic; and transferring the wastewater to the corresponding environment according to the online dissolved oxygen meter's judgment criteria, and then using online detection equipment to monitor the changes in various components in the wastewater in real time. S2. Based on the installation of a series of detection devices in the wastewater storage tank to collect classified wastewater data in real time and perform data preprocessing, the preprocessed wastewater data is obtained, including the content data of each component in the wastewater. S3. Optimize the initial weights and thresholds of the Elman neural network using a non-dominated genetic algorithm to obtain an optimized Elman neural network; process the pre-treated wastewater data using the optimized Elman neural network to obtain wastewater data processed by the Elman neural network. S4. Analyze the wastewater data processed by the Elman neural network, combining the weights and the set thresholds. Specifically, this includes: selecting the top n principal components with a cumulative variance contribution rate greater than 85% based on the comparison results after combining weights and setting thresholds, sorting the top n principal components according to their contribution rates, and processing the principal components in the wastewater according to the sorting; after processing, sorting the components with a cumulative variance contribution rate of not less than 15% according to their contribution rates and mixing the wastewater with the same component sorting. S5. Based on the comparison results, the wastewater is directed to the corresponding wastewater treatment center for treatment. Specifically, this includes: after the treatment of the first n principal components with a cumulative variance contribution rate greater than 85% in the wastewater, the treated wastewater is subjected to repeated operations S2-S5 until the wastewater treatment standard is met.

2. The multi-objective optimization control method for wastewater treatment process based on knowledge selection according to claim 1, characterized in that, The method of collecting real-time data on the main substances in wastewater by installing a series of detection devices in the wastewater storage tank includes: The installed series of testing equipment includes: an online pH meter, an online dissolved oxygen meter, and an online component analyzer; The online pH meter is used to monitor the pH value of wastewater in real time; The online component analyzer is used to monitor the concentration changes of various components in wastewater in real time. The online dissolved oxygen meter is used to monitor the concentration of dissolved oxygen in wastewater in real time; Based on the concentration of dissolved oxygen in wastewater measured by an online dissolved oxygen meter, the collected wastewater is distinguished by setting a concentration threshold for dissolved oxygen in the wastewater; The dissolved oxygen concentration was set at >2.0 mg / L under aerobic conditions, <0.2 mg / L under anaerobic conditions, 0.2-0.5 mg / L under hypoxic conditions, and 0.5-2.0 mg / L under normal conditions. Wastewater is initially classified into anaerobic, anoxic, and aerobic categories.

3. The multi-objective optimization control method for wastewater treatment process based on knowledge selection according to claim 1, characterized in that, The method involves installing a series of detection devices in a wastewater storage tank to collect classified wastewater data in real time and performing data preprocessing to obtain preprocessed wastewater data, including: The wastewater data collected in real time and classified is stored in matrix form; The auxiliary variable data sample matrix X is obtained according to Formula 1. m×n Mean and variance: The sample matrix is ​​standardized to zero mean according to Formula 2, and the standardized matrix is ​​calculated. Where m is the number of samples, n is the number of sample components, and x ij This represents the j-th component of the i-th sample; S is the mean of the i-th sample component. j This represents the standard deviation of the i-th sample; The standardized matrix Z is calculated based on formulas 3 and 4. m×n covariance matrix R n×n ; Where r represents the covariance and T represents the matrix transpose; The different eigenvalues ​​λ of R can be obtained by solving formula 5. j (j=1,2,...,n), and arrange the n eigenvalues ​​of R in descending order; |R-λ J E|=0 (5) Where, λ J This indicates that the eigenvalues ​​λ are different. j The matrix formed by these components, where R represents the covariance matrix; The result is used to calculate the unit eigenvector b corresponding to the corresponding eigenvalue using Formula 6. j (j=1,2,...,n), b j =(b 1j b 2j , ..., b nj ); Rb=λ J b (6) The cumulative variance contribution rate of the principal components is calculated based on the calculated eigenvalues, and the number of principal components k is determined based on the cumulative variance contribution rate. According to Formula 8, the standardized matrix Z is... m×n Projected onto k-dimensional coordinates: Where U1 is the first principal component, U2 is the second principal component, and U... k It is the k-th principal component.

4. The multi-objective optimization control method for wastewater treatment process based on knowledge selection according to claim 1, characterized in that, The process of optimizing the initial weights and thresholds of the Elman neural network using a non-dominated genetic algorithm to obtain the optimized Elman neural network specifically includes the following steps: The objective function for the weights is expressed by the following formula: Among them, w max w min These represent the maximum and minimum values ​​of the weights, w. max =0.9, w min =0.4; k is the number of iterations, k max This represents the maximum number of iterations. By setting the initial basic parameters of the Elman neural network and encoding the weights and thresholds of the Elman neural network; S31. Initialize the population, set basic parameters such as maximum number of iterations, population size, crossover probability, and mutation probability, and set appropriate constraints on input and output variables. S32. Perform fast non-dominated sorting on the population and calculate the crowding degree, then proceed to the algorithm iteration process; S33. Perform a selection operation based on the non-dominated hierarchy ranking and crowding degree of individuals. For individuals with the same non-dominated hierarchy ranking, select those with lower crowding degree; for individuals with different non-dominated hierarchy rankings, select those with higher crowding degree. S34. Crossover and mutation operations; S35. Merge populations, iterate and filter: Merge the parent and child populations from step S34 to generate a new merged population. Use the method described in step S33 to filter and generate a new population. Stop when the maximum number of iterations is reached, and the algorithm terminates.

5. The multi-objective optimization control method for wastewater treatment process based on knowledge selection according to claim 4, characterized in that, The fast non-dominated sorting specifically includes the following steps: S321. Traverse all solutions in the set and calculate the corresponding sets n and S for each solution. p S represents the number of dominant solutions p, and is a single numerical value; p Let p represent the solutions dominated by solution p, which is a set. S322. Find the solution where n=0, assign it to set F1, and rank its corresponding non-dominated layer by 1. S323. Iterate through each individual j in set F1 and find the solution set S dominated by individual j. j and S j n corresponding to each individual h h Subtract 1; S324, Find S j The corresponding n h Find the solution that equals 0, assign it to set F2, and assign its corresponding non-dominated layer sort value to 2. Repeat steps S322-S324 to complete the non-dominated sorting operation for all individuals in the set.

6. The multi-objective optimization control method for wastewater treatment process based on knowledge selection according to claim 4, characterized in that, The binary crossover process is as follows: Where, p i，k For its parent individual, c i，k The offspring individuals generated by its crossover, i, k represent the k-th individual of the i-th generation, β k ≥0 represents an individual that undergoes crossover, and its probability density function is: Where, η c Let c represent the c-th offspring individual, and β represent the crossover operation; The process of generating offspring through polynomial mutation is achieved using the polynomial mutation operator, as shown in the following formula: Where u∈[0,1], c k p represents the offspring individuals resulting from the mutation. k Represents its parent individual, δ k It can be expressed by the following formula: Where, r k η represents the non-dominated order of an individual. m This is the variation distribution index.

7. The multi-objective optimization control method for wastewater treatment process based on knowledge selection according to claim 1, characterized in that, The process of processing the pretreated wastewater data using an optimized Elman neural network to obtain Elman neural network-processed wastewater data includes: The structure of an Elman neural network includes an input layer, hidden layers, a supporting layer, and an output layer. The nonlinear state expression of the Elman neural network structure is as follows: x(t)=f(w1x c (t)+w2u(t-1)) (19) x c (t)=x(t-1) (20) y t =g(w3x(t)) (21) Where r, n, and m represent the number of nodes in the input layer, hidden layer, and output layer, respectively, and the number of nodes in the connecting layer is the same as the number of nodes in the hidden layer, denoted as n, x(t), x c (t), y t Let w1 represent the outputs of the hidden layer, the receiving layer, and the output layer at time t, respectively; w2 represent the connection weights from the receiving layer to the hidden layer; w3 represent the connection weights from the hidden layer to the output layer; g() represents the transfer function of the output layer; f() represents the activation function of the hidden layer; and u represents the input vector of the input layer. The activation function of the hidden layer is the Sigmoid function: g() takes a linear function: y t =w3x(t) (23) The Elman neural network learning algorithm uses gradient descent, and its error index function is expressed as follows: Among them, y d y(q) represents the actual output of the Elman neural network at step q, and y(q) represents the input of the Elman neural network at step q.

Images