Sewage treatment data management system based on artificial intelligence

Abstract

The present invention discloses a sewage treatment data management system based on artificial intelligence, including a data acquisition module, a data preprocessing module, a sewage treatment process water quality parameter management module, a sewage treatment equipment management module, a hyperparameter optimization module and a sewage treatment intelligent auxiliary module. The present invention relates to the technical field of sewage data management, and specifically refers to a sewage treatment data management system based on artificial intelligence. This scheme proposes a mode combining sewage treatment process water quality parameter management with sewage treatment equipment management, real-time monitoring of water quality data and equipment status, and accurate positioning of sewage treatment process problems; innovatively introduces non-parametric kernel density estimation and dynamic window width parameter selection strategy to solve the problem of inaccurate selection of cluster centers and cluster shapes; uses reversible chaotic mapping to initialize populations, population event selection factors and individual migration step adjustment strategies to enhance global search capabilities and obtain the optimal parameter combination.

Description

Technical Field

[0001] The present invention relates to the technical field of sewage data management, and in particular to a sewage treatment data management system based on artificial intelligence. Background Art

[0002] With the acceleration of urbanization and the growth of population, the demand for sewage treatment is increasing, and traditional sewage treatment technology is facing many challenges. In order to effectively respond to these challenges and improve the efficiency and quality of sewage treatment, intelligent sewage treatment data management systems have emerged. Modern sewage treatment systems require real-time and accurate data monitoring and analysis to optimize the treatment process and ensure water quality and environmental protection. However, traditional sewage treatment data management systems usually have the technical problem of only monitoring the water quality parameters of the sewage treatment process and the status of sewage treatment equipment, resulting in the inability to accurately locate problems when they occur in the sewage treatment process, thereby affecting the efficiency of sewage treatment data management. General clustering methods have technical problems such as incorrect selection of cluster centers and inaccurate cluster shapes. Traditional clustering methods have improper parameter settings and weak global search capabilities in parameter optimization algorithms, resulting in inaccurate final clustering results. Summary of the invention

[0003] In view of the above situation, in order to overcome the defects of the prior art, the present invention provides a sewage treatment data management system based on artificial intelligence. In view of the technical problem that traditional sewage treatment data management systems usually only monitor the water quality parameters of the sewage treatment process and the status of the sewage treatment equipment, resulting in the inability to accurately locate problems when they occur in the sewage treatment process, thereby affecting the efficiency of sewage treatment data management, this solution creatively introduces a mode of combining sewage treatment process water quality parameter management with sewage treatment equipment management. By real-time monitoring of water quality data and equipment status, it achieves accurate positioning of sewage treatment process problems and gives relevant treatment suggestions, thereby solving the problem of low efficiency in the sewage treatment data management process; in view of the clustering problem in general clustering methods In order to solve the technical problems of incorrect center selection and inaccurate cluster shape, which affect the flexibility and stability of the clustering method, this scheme introduces non-parametric kernel density estimation and dynamic window width parameter selection strategy to more accurately define the local density, solve the problems of inaccurate selection of cluster centers and cluster shapes, and improve the flexibility and stability of clustering; in order to solve the technical problems of improper parameter setting and weak global search ability of parameter optimization algorithm in traditional clustering methods, which lead to inaccurate final clustering results, this scheme uses reversible chaotic mapping to initialize the population, population event selection factors and individual migration step adjustment strategy to enhance the global search ability and adaptive balance parameter ability, thereby obtaining the optimal parameter value and solving the problem of inaccurate output results of the final monitoring model.

[0004] The technical solution adopted by the present invention is as follows: the artificial intelligence-based sewage treatment data management system provided by the present invention includes a data acquisition module, a data preprocessing module, a sewage treatment process water quality parameter management module, a sewage treatment equipment management module, a hyperparameter optimization module and a sewage treatment intelligent auxiliary module;

[0005] The data acquisition module collects historical sewage treatment water quality data, historical sewage treatment equipment operation data, real-time sewage treatment water quality data, real-time sewage treatment equipment operation data and external environment data to obtain sewage treatment raw data;

[0006] The data preprocessing module performs data cleaning, standardization, feature selection and data aggregation on the original sewage treatment data to obtain preliminary sewage treatment data;

[0007] The sewage treatment process water quality parameter management module establishes a water quality parameter abnormality monitoring model to monitor the water quality data in the sewage treatment process in real time, obtains the water quality monitoring results of the current sewage treatment process, and determines the current water quality status;

[0008] The sewage treatment equipment management module monitors the operating status of the sewage treatment equipment during the sewage treatment process by establishing an equipment operating status monitoring model to obtain equipment operating status monitoring results;

[0009] The hyperparameter optimization module obtains the optimal parameter combination of the equipment operation status monitoring model through a search algorithm;

[0010] The sewage treatment intelligent auxiliary module locates problems in the sewage treatment process through the output results of the sewage treatment process water quality parameter management module and the output results of the sewage treatment equipment management module, and gives relevant treatment suggestions based on specific problems.

[0011] Furthermore, the data acquisition module specifically obtains original sewage treatment data from the sewage treatment plant by collecting data; the original sewage treatment data includes historical sewage treatment water quality data, historical sewage treatment equipment operation data, real-time sewage treatment water quality data, real-time sewage treatment equipment operation data and external environment data; the historical sewage treatment water quality data and the real-time sewage treatment water quality data both include chemical oxygen demand, biological oxygen demand, total nitrogen, total phosphorus, dissolved oxygen, pH value and suspended solids data; the historical sewage treatment equipment operation data and the real-time sewage treatment equipment operation data both include load, vibration frequency, energy consumption, speed, pressure, temperature and equipment operation status of the aeration equipment; the external environment data includes air temperature, humidity, precipitation, wind speed, air pressure, flow rate of the sewage source and initial water quality data.

[0012] Furthermore, the data preprocessing module specifically performs data cleaning, standardization, feature selection and data aggregation on the original sewage treatment data to obtain preliminary sewage treatment data; the data cleaning is to process missing values, outliers and duplicate values; the standardization is to standardize the data based on the maximum and minimum normalization method; the feature selection is to use the correlation analysis method to calculate the correlation between the features and the water quality indicators and the equipment status, select the features with strong correlation with the sewage treatment effect, and eliminate redundant features; the data aggregation includes time dimension aggregation and space dimension aggregation, the time dimension aggregation is to perform statistical processing on the data at the time granularity of hours and days, and the space dimension aggregation is to integrate the data of different sewage treatment tanks, equipment and sensors to generate comprehensive features.

[0013] Furthermore, the sewage treatment process water quality parameter management module specifically establishes a water quality parameter abnormality monitoring model to monitor the water quality data in the sewage treatment process in real time, obtains the water quality monitoring results in the current sewage treatment process, and determines whether the current water quality state is abnormal; including the following steps:

[0014] A water quality parameter anomaly monitoring model is established, specifically by using a nonparametric kernel density estimation method and a dynamic window width parameter selection method to effectively identify the core points in the cluster and generate more accurate clustering results; the following steps are included:

[0015] The distance between sample data points is calculated, specifically, the distance between sample data points is measured values ​​of various water quality parameters in historical sewage treatment water quality data and real-time sewage treatment water quality data. The formula used is expressed as follows:

[0016] ;

[0017] In the formula, Represents sample data points and sample data points The distance between

[0018] Calculate the local density of the sample data points, specifically perform local density estimation by a nonparametric kernel density estimation method, including the following steps:

[0019] Kernel density estimation is performed, and the formula used is expressed as follows:

[0020] ;

[0021] ;

[0022] In the formula, represents the density estimate of the sample data point x, n represents the number of samples, h represents the window width, represents the Gaussian kernel function, represents the exponential function, q represents the qth sample data point in the data set;

[0023] Perform local density estimation, specifically apply the kernel density estimation formula to the local density calculation of the data points, and extract the constant term from the formula. The formula used is expressed as follows:

[0024] ;

[0025] In the formula, Represents data points Local density value;

[0026] Dynamic window width parameter selection, specifically selecting the optimal window width by minimizing the mean integrated square error, includes the following steps:

[0027] The mean integrated square error is calculated using the following formula:

[0028] ;

[0029] ;

[0030] In the formula, represents the mean integrated square error, f(·) represents the actual probability density function, which represents the true distribution of the data, represents the asymptotic mean of the mean integrated squared error, x represents the sample data point, represents the mathematical expectation, represents the first-order moment of the Gaussian kernel function K, represents the second-order moment of the Gaussian kernel function K, Indicates a trend of gradual growth, It represents the integral of the square of the second derivative of f(x), which measures the curvature of the density function;

[0031] Select the optimal window width; the formula used is as follows:

[0032] ;

[0033] ;

[0034] In the formula, Express Partial derivative with respect to window width h;

[0035] Calculate the relative distance; the formula used is expressed as follows:

[0036] ;

[0037] In the formula, Represents sample data points Local density value; Represents sample data points The minimum distance to the nearest point with higher density, j means that the condition is met The sample data point index of ;

[0038] Identify the cluster center, specifically by calculating the decision value of each sample data point to determine the cluster center, the decision value is obtained by multiplying the local density and the relative distance; the formula used is expressed as follows:

[0039] ;

[0040] In the formula, represents the decision value, represents the local density of data sample points, Represents the minimum distance between a sample data point and the nearest point with higher density;

[0041] Cluster assignment, specifically, for each non-cluster center point, find the point with the closest distance and higher local density, and assign the point to the corresponding cluster center;

[0042] Training of a water quality parameter anomaly monitoring model, specifically using the historical sewage treatment water quality data and external environment data in the preliminary sewage treatment data as input data to train the water quality parameter anomaly monitoring model, and analyzing the clustering results, marking the water quality state corresponding to each cluster, the water quality state including normal and abnormal, and obtaining a trained water quality parameter anomaly monitoring model;

[0043] The water quality parameter abnormality monitoring specifically uses the real-time sewage treatment water quality data and external environment data in the preliminary sewage treatment data as input data of the trained water quality parameter abnormality monitoring model to obtain the water quality monitoring results in the current sewage treatment process.

[0044] Furthermore, the sewage treatment equipment management module specifically monitors the operation status of the sewage treatment equipment during the sewage treatment process by establishing an equipment operation status monitoring model to obtain equipment operation status monitoring results; specifically includes the following steps:

[0045] Establishing the equipment operation status monitoring model includes the following steps:

[0046] Initialize the internal structure of the model LSTM, which includes the following steps:

[0047] The calculation process of the forget gate uses the following formula:

[0048] ;

[0049] In the formula, f trepresents the output of the forget gate, X t It indicates input data. Indicates the hidden state at the previous moment, W f represents the weight matrix corresponding to the forget gate, represents the bias parameter of the forget gate, Represents the sigmoid function;

[0050] Input gate calculation process, the formula used is as follows:

[0051] ;

[0052] ;

[0053] ;

[0054] In the formula, i t represents the output of the input gate, W i represents the corresponding input gate weight matrix, represents the bias parameter of the input gate, c t Indicates the current cell state. represents the candidate value of the new cell state, c t-1 Indicates the cell state at the last moment, W c Represents the corresponding cell state weight matrix, B c The bias parameter representing the cell state, represents element-wise multiplication, It is the hyperbolic tangent function;

[0055] The output gate calculation process uses the following formula:

[0056] ;

[0057] ;

[0058] In the formula, o t represents the output of the output gate, W o It represents the corresponding output gate weight matrix, B o Represents the bias parameter of the output gate, h t Indicates the hidden state at the current moment;

[0059] To calculate the bidirectional LSTM, the formula used is as follows:

[0060] ;

[0061] ;

[0062] ;

[0063] Where h represents the hidden state of the combined forward and backward LSTM. Indicates the hidden state in the positive direction at the current moment, Indicates the hidden state in the positive direction at the previous moment, Indicates the reverse hidden state at the current moment. Indicates the hidden state in the reverse direction at the next moment;

[0064] The final output value is obtained using the following formula:

[0065] ;

[0066] In the formula, Represents the output value of the monitoring model, represents the weight matrix corresponding to the output, represents the Softmax activation function, B y Represents the bias parameter of the output layer;

[0067] Equipment operation status monitoring model training, specifically using the historical sewage treatment equipment operation data in the preliminary sewage treatment data to train the equipment operation status monitoring model to obtain a trained equipment operation status monitoring model;

[0068] The real-time monitoring of the equipment operation status specifically involves using the real-time sewage treatment equipment operation data in the preliminary sewage treatment data as input data of the trained equipment operation status monitoring model to obtain equipment operation status monitoring results.

[0069] Furthermore, the hyperparameter optimization module specifically obtains the optimal parameter combination of the equipment operation status monitoring model through a search algorithm; comprising the following steps:

[0070] Initialization parameters, specifically by constructing the initial parameters of the algorithm; the initial parameters of the algorithm include the population size N, the population dimension D and the maximum number of iterations ;

[0071] Initializing the population, specifically, initializing the population by generating chaotic sequence values ​​through a reversible chaotic mapping, includes the following steps:

[0072] Generate N chaotic sequence values ​​and initialize the random value z0; the formula used is as follows:

[0073] ;

[0074] ;

[0075] In the formula, represents the i-th chaotic sequence value of the d-th dimension, represents the i+1th chaotic sequence value of the dth dimension, round represents a random number between 0 and 1, and A represents The mapping matrix, Indicates the current value of each dimension. represents the i-th chaotic sequence value of the first dimension, represents the i-th chaotic sequence value of the second dimension, and T represents the transposition operation;

[0076] The chaotic sequence value conversion search individual is expressed as follows:

[0077] ;

[0078] In the formula, represents the initialization position of the dth dimension of the i-th individual, lb d andub d Respectively represent the search lower and upper limits of dimension d;

[0079] Calculate the fitness value, specifically calculate the individual fitness value f in the population i ; The performance of the equipment operation status monitoring model established based on the individual position is used as the individual fitness value;

[0080] The population event selection factor is calculated using the following formula:

[0081] ;

[0082] In the formula, represents the event selection factor of the t-th generation population, and t represents the current number of iterations;

[0083] Individual exploration event stage, specifically if the current iteration population event selection factor , the individual enters the exploration event stage, and the formula used is as follows:

[0084] ;

[0085] In the formula, represents the d-th dimension position of the i-th individual in the t+1-th generation population, p represents a random integer uniformly distributed in the range [1, D], and r represents a random integer uniformly distributed in the range [1, N]. represents a random number uniformly distributed in the range [0,1]. represents the random dimensional position of a random individual in the t-th generation population, represents the random dimensional position of the i-th individual in the t-th generation population, It represents the remainder when d is divided by 2, where d represents the index of the dimension. represents the randomness parameter used to control the search space, representing a random number uniformly distributed in the range [0,2π];

[0086] Individual predation event stage, specifically if the current iteration population event selection factor , then the individual enters the predation event stage, and the formula used is as follows:

[0087] ;

[0088] In the formula, represents the position of the individual with the best fitness value in the population in the tth iteration on dimension d, represents the d-th dimension position of the i-th individual in the t-th generation population, represents the d-th dimension position of a random individual in the t-th generation population, Represents a random number in the range [-1,1], and represents a random number uniformly distributed in the range [0,1]. and represents random numbers that follow a normal distribution;

[0089] In the individual migration stage, the position is updated by adjusting the individual migration step size; the formula used is as follows:

[0090] ;

[0091] ;

[0092] In the formula, Indicates the probability of encountering danger during the migration phase, represents the individual migration step length, Up and Lw represent the maximum and minimum values ​​of the solution space respectively, represents the fitness value of the best individual, represents the fitness value of the worst individual, , and Represents a random number uniformly distributed in the range [0,1];

[0093] Search determination, specifically, by constructing a search termination condition, searching and determining the optimal individual position, and obtaining the optimal individual position data setting;

[0094] The search termination conditions include threshold termination and iteration termination;

[0095] The threshold termination is specifically to set the fitness threshold, when the individual fitness value f i When it is above the fitness threshold, the hyperparameter search is completed;

[0096] The iteration termination specifically refers to terminating the iteration and obtaining the optimal individual position when the maximum number of iterations is reached;

[0097] The optimal individual position includes equipment operation status monitoring model parameters.

[0098] Furthermore, the sewage treatment intelligent auxiliary module can locate the problematic links in the sewage treatment process based on the output results of the sewage treatment process water quality parameter management module and the output results of the sewage treatment equipment management module; if the water quality is abnormal, the specific parameters of the sewage treatment process are adjusted according to the output results of the sewage treatment process water quality parameter management module; if the equipment is abnormal, suggestions for sewage treatment equipment and maintenance are given according to the output results of the sewage treatment equipment management module.

[0099] The beneficial effects achieved by the present invention using the above scheme are as follows:

[0100] (1) Traditional sewage treatment data management systems usually have the technical problem of only monitoring the water quality parameters of the sewage treatment process and the status of sewage treatment equipment. As a result, problems cannot be accurately located when they occur in the sewage treatment process, thus affecting the efficiency of sewage treatment data management. This solution creatively introduces a model that combines sewage treatment process water quality parameter management with sewage treatment equipment management. By real-time monitoring of water quality data and equipment status, it can accurately locate problems in the sewage treatment process and provide relevant treatment suggestions, thus solving the problem of inefficiency in the sewage treatment data management process.

[0101] (2) In order to solve the technical problems of incorrect cluster center selection and inaccurate cluster shape in general clustering methods, which affect the flexibility and stability of clustering methods, this scheme introduces non-parametric kernel density estimation and dynamic window width parameter selection strategy to more accurately define local density and distance, solve the problems of inaccurate selection of cluster centers and cluster shapes, and improve the flexibility and stability of clustering.

[0102] (3) In order to solve the technical problems of improper parameter settings in traditional clustering methods and weak global search capabilities of parameter optimization algorithms, which lead to inaccurate final clustering results, this scheme uses reversible chaotic mapping to initialize the population, population event selection factors and individual migration step adjustment strategies to enhance the global search capability and adaptive balance parameter capabilities, thereby obtaining the optimal parameter values ​​and solving the problem of inaccurate output results of the final monitoring model. BRIEF DESCRIPTION OF THE DRAWINGS

[0103] Figure 1 A schematic diagram of the modules of the artificial intelligence-based sewage treatment data management system provided by the present invention;

[0104] Figure 2It is a flow chart of the water quality parameter management module of the sewage treatment process;

[0105] Figure 3 A flow chart of establishing a water quality parameter abnormality monitoring model in the water quality parameter management module of the sewage treatment process;

[0106] Figure 4 This is a flow chart of the sewage treatment equipment management module;

[0107] Figure 5 This is a flowchart of the hyperparameter optimization module;

[0108] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention. DETAILED DESCRIPTION

[0109] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments; based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0110] In the description of the present invention, it is necessary to understand that terms such as “upper”, “lower”, “front”, “back”, “left”, “right”, “top”, “bottom”, “inside” and “outside” indicating directions or positional relationships are based on the directions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the referred system or element must have a specific direction, be constructed and operated in a specific direction, and therefore cannot be understood as a limitation on the present invention.

[0111] Example 1, see Figure 1 , the artificial intelligence-based sewage treatment data management system provided by the present invention includes a data acquisition module, a data preprocessing module, a sewage treatment process water quality parameter management module, a sewage treatment equipment management module, a hyperparameter optimization module and a sewage treatment intelligent auxiliary module;

[0112] The data acquisition module collects historical sewage treatment water quality data, historical sewage treatment equipment operation data, real-time sewage treatment water quality data, real-time sewage treatment equipment operation data and external environment data to obtain sewage treatment raw data; and sends the data to the data preprocessing module;

[0113] The data preprocessing module receives the data sent by the data acquisition module, performs data cleaning, standardization, feature selection and data aggregation on the original sewage treatment data to obtain preliminary sewage treatment data; and sends the data to the sewage treatment process water quality parameter management module and the sewage treatment equipment management module;

[0114] The sewage treatment process water quality parameter management module receives the data sent by the data preprocessing module, establishes a water quality parameter abnormality monitoring model, monitors the water quality data in the sewage treatment process in real time, obtains the water quality monitoring results of the current sewage treatment process, and judges the current water quality status; and sends the data to the sewage treatment intelligent auxiliary module;

[0115] The sewage treatment equipment management module receives the data sent by the data preprocessing module and the hyperparameter optimization module, monitors the operation status of the sewage treatment equipment during the sewage treatment process by establishing an equipment operation status monitoring model, and obtains the equipment operation status monitoring results; and sends the data to the sewage treatment intelligent auxiliary module;

[0116] The hyperparameter optimization module obtains the optimal parameter combination of the equipment operation status monitoring model through a search algorithm; and sends the data to the sewage treatment equipment management module;

[0117] The sewage treatment intelligent auxiliary module receives data sent by the sewage treatment process water quality parameter management module and the sewage treatment equipment management module, locates problems in the sewage treatment process based on the output results of the sewage treatment process water quality parameter management module and the output results of the sewage treatment equipment management module, and gives relevant treatment suggestions based on specific problems.

[0118] Example 2, see Figure 1 This embodiment is based on the above embodiment, and the data acquisition module specifically obtains the original sewage treatment data from the sewage treatment plant by collecting data; the original sewage treatment data includes historical sewage treatment water quality data, historical sewage treatment equipment operation data, real-time sewage treatment water quality data, real-time sewage treatment equipment operation data and external environment data; the historical sewage treatment water quality data and the real-time sewage treatment water quality data both include chemical oxygen demand, biological oxygen demand, total nitrogen, total phosphorus, dissolved oxygen, pH value and suspended solids data; the historical sewage treatment equipment operation data and the real-time sewage treatment equipment operation data both include load, vibration frequency, energy consumption, speed, pressure, temperature and equipment operation status of the aeration equipment; the external environment data includes air temperature, humidity, precipitation, wind speed, air pressure, flow rate of sewage source and initial water quality data.

[0119] Example 3, see Figure 1This embodiment is based on the above embodiment, and the data preprocessing module specifically performs data cleaning, standardization, feature selection and data aggregation on the original sewage treatment data to obtain preliminary sewage treatment data; the data cleaning is to process missing values, abnormal values ​​and duplicate values; the standardization is to standardize the data based on the maximum and minimum normalization method; the feature selection is to use the correlation analysis method to calculate the correlation between the features and the water quality indicators and the equipment status, select the features with strong correlation with the sewage treatment effect, and eliminate redundant features; the data aggregation includes time dimension aggregation and space dimension aggregation, the time dimension aggregation is to perform statistical processing on the data at the time granularity of hours and days, and the space dimension aggregation is to integrate the data of different sewage treatment tanks, equipment and sensors to generate comprehensive features.

[0120] Example 4, see Figure 1 , Figure 2 and Figure 3 This embodiment is based on the above embodiment. The sewage treatment process water quality parameter management module specifically establishes a water quality parameter abnormality monitoring model to monitor the water quality data in the sewage treatment process in real time, obtain the water quality monitoring results of the current sewage treatment process, and determine whether the current water quality state is abnormal; it includes the following steps:

[0121] A water quality parameter anomaly monitoring model is established, specifically by using a nonparametric kernel density estimation method and a dynamic window width parameter selection method to effectively identify the core points in the cluster and generate more accurate clustering results; the following steps are included:

[0122] The distance between sample data points is calculated, specifically, the distance between sample data points is measured values ​​of various water quality parameters in historical sewage treatment water quality data and real-time sewage treatment water quality data. The formula used is expressed as follows:

[0123] ;

[0124] In the formula, Represents sample data points and sample data points The distance between

[0125] Calculate the local density of the sample data points, specifically perform local density estimation by a nonparametric kernel density estimation method, including the following steps:

[0126] Kernel density estimation is performed, and the formula used is expressed as follows:

[0127] ;

[0128] ;

[0129] In the formula, represents the density estimate of the sample data point x, n represents the number of samples, h represents the window width, represents the Gaussian kernel function, represents the exponential function, q represents the qth sample data point in the data set;

[0130] Perform local density estimation, specifically apply the kernel density estimation formula to the local density calculation of the data points, and extract the constant term from the formula. The formula used is expressed as follows:

[0131] ;

[0132] In the formula, Represents data points Local density value;

[0133] Dynamic window width parameter selection, specifically selecting the optimal window width by minimizing the mean integrated square error, includes the following steps:

[0134] The mean integrated square error is calculated using the following formula:

[0135] ;

[0136] ;

[0137] In the formula, represents the mean integrated square error, f(·) represents the actual probability density function, which represents the true distribution of the data, represents the asymptotic mean of the mean integrated squared error, x represents the sample data point, represents the mathematical expectation, represents the first-order moment of the Gaussian kernel function K, represents the second-order moment of the Gaussian kernel function K, Indicates a trend of gradual growth, It represents the integral of the square of the second derivative of f(x), which measures the curvature of the density function;

[0138] Select the optimal window width; the formula used is as follows:

[0139] ;

[0140] ;

[0141] In the formula, Express Partial derivative with respect to window width h;

[0142] Calculate the relative distance; the formula used is expressed as follows:

[0143] ;

[0144] In the formula, Represents sample data points Local density value; Represents sample data points The minimum distance to the nearest point with higher density, j means that the condition is met The sample data point index of ;

[0145] Identify the cluster center, specifically by calculating the decision value of each sample data point to determine the cluster center, the decision value is obtained by multiplying the local density and the relative distance; the formula used is expressed as follows:

[0146] ;

[0147] In the formula, represents the decision value, represents the local density of data sample points, Represents the minimum distance between a sample data point and the nearest point with higher density;

[0148] Cluster assignment, specifically, for each non-cluster center point, find the point with the closest distance and higher local density, and assign the point to the corresponding cluster center;

[0149] Training of a water quality parameter anomaly monitoring model, specifically using the historical sewage treatment water quality data and external environment data in the preliminary sewage treatment data as input data to train the water quality parameter anomaly monitoring model, and analyzing the clustering results, marking the water quality state corresponding to each cluster, the water quality state including normal and abnormal, and obtaining a trained water quality parameter anomaly monitoring model;

[0150] The water quality parameter abnormality monitoring specifically uses the real-time sewage treatment water quality data and external environment data in the preliminary sewage treatment data as input data of the trained water quality parameter abnormality monitoring model to obtain the water quality monitoring results in the current sewage treatment process.

[0151] By performing the above operations, this scheme introduces non-parametric kernel density estimation and dynamic window width parameter selection strategy to more accurately define local density and distance, solve the problems of inaccurate selection of cluster centers and cluster shapes, and improve the flexibility and stability of clustering, in order to solve the technical problems of incorrect cluster center selection and inaccurate cluster shapes in general clustering methods, which affect the flexibility and stability of clustering methods.

[0152] Example 5, see Figure 1 and Figure 4The sewage treatment equipment management module specifically monitors the operating status of the sewage treatment equipment during the sewage treatment process by establishing an equipment operating status monitoring model to obtain the equipment operating status monitoring results; specifically includes the following steps:

[0153] Establishing the equipment operation status monitoring model includes the following steps:

[0154] Initialize the internal structure of the model LSTM, which includes the following steps:

[0155] The calculation process of the forget gate uses the following formula:

[0156] ;

[0157] In the formula, f t represents the output of the forget gate, X t It indicates input data. Indicates the hidden state at the previous moment, W f represents the weight matrix corresponding to the forget gate, represents the bias parameter of the forget gate, Represents the sigmoid function;

[0158] Input gate calculation process, the formula used is as follows:

[0159] ;

[0160] ;

[0161] ;

[0162] In the formula, i t represents the output of the input gate, W i represents the corresponding input gate weight matrix, represents the bias parameter of the input gate, c t Indicates the current cell state. represents the candidate value of the new cell state, c t-1 Indicates the cell state at the last moment, W c Represents the corresponding cell state weight matrix, B c The bias parameter representing the cell state, represents element-wise multiplication, It is the hyperbolic tangent function;

[0163] The output gate calculation process uses the following formula:

[0164] ;

[0165] ;

[0166] In the formula, o t represents the output of the output gate, W o It represents the corresponding output gate weight matrix, B o Represents the bias parameter of the output gate, h t Indicates the hidden state at the current moment;

[0167] To calculate the bidirectional LSTM, the formula used is as follows:

[0168] ;

[0169] ;

[0170] ;

[0171] Where h represents the hidden state of the combined forward and backward LM, Indicates the hidden state in the positive direction at the current moment, Indicates the hidden state in the positive direction at the previous moment, Indicates the reverse hidden state at the current moment. Indicates the hidden state in the reverse direction at the next moment;

[0172] The final output value is obtained using the following formula:

[0173] ;

[0174] In the formula, Represents the output value of the monitoring model, represents the weight matrix corresponding to the output, represents the activation function, B y Represents the bias parameter of the output layer;

[0175] Equipment operation status monitoring model training, specifically using the historical sewage treatment equipment operation data in the preliminary sewage treatment data to train the equipment operation status monitoring model to obtain a trained equipment operation status monitoring model;

[0176] The real-time monitoring of the equipment operation status specifically involves using the real-time sewage treatment equipment operation data in the preliminary sewage treatment data as input data of the trained equipment operation status monitoring model to obtain equipment operation status monitoring results.

[0177] Example 6, see Figure 1 and Figure 5 This embodiment is based on the above embodiment, and the hyperparameter optimization module specifically obtains the optimal parameter combination of the equipment operation status monitoring model through a search algorithm; it includes the following steps:

[0178] Initialization parameters, specifically by constructing the initial parameters of the algorithm; the initial parameters of the algorithm include the population size N, the population dimension D and the maximum number of iterations ;

[0179] Initializing the population, specifically, initializing the population by generating chaotic sequence values ​​through a reversible chaotic mapping, includes the following steps:

[0180] Generate N chaotic sequence values ​​and initialize the random value z0; the formula used is as follows:

[0181] ;

[0182] ;

[0183] In the formula, represents the i-th chaotic sequence value of the d-th dimension, represents the i+1th chaotic sequence value of the dth dimension, round represents a random number between 0 and 1, and A represents The mapping matrix, Indicates the current value of each dimension. represents the i-th chaotic sequence value of the first dimension, represents the i-th chaotic sequence value of the second dimension, and T represents the transposition operation;

[0184] The chaotic sequence value conversion search individual is expressed as follows:

[0185] ;

[0186] In the formula, represents the initialization position of the dth dimension of the i-th individual, lb d andub d Respectively represent the search lower and upper limits of dimension d;

[0187] Calculate the fitness value, specifically calculate the individual fitness value f in the population i ; The performance of the equipment operation status monitoring model established based on the individual position is used as the individual fitness value;

[0188] The population event selection factor is calculated using the following formula:

[0189] ;

[0190] In the formula, represents the event selection factor of the t-th generation population, and t represents the current number of iterations;

[0191] Individual exploration event stage, specifically if the current iteration population event selection factor , the individual enters the exploration event stage, and the formula used is as follows:

[0192] ;

[0193] In the formula, represents the d-th dimension position of the i-th individual in the t+1-th generation population, p represents a random integer uniformly distributed in the range [1, D], and r represents a random integer uniformly distributed in the range [1, N]. represents a random number uniformly distributed in the range [0,1]. represents the random dimensional position of a random individual in the t-th generation population, represents the random dimensional position of the i-th individual in the t-th generation population, It represents the remainder when d is divided by 2, where d represents the index of the dimension. represents the randomness parameter used to control the search space, representing a random number uniformly distributed in the range [0,2π];

[0194] Individual predation event stage, specifically if the current iteration population event selection factor , then the individual enters the predation event stage, and the formula used is as follows:

[0195] ;

[0196] In the formula, represents the position of the individual with the best fitness value in the population in the tth iteration on dimension d, represents the d-th dimension position of the i-th individual in the t-th generation population, represents the d-th dimension position of a random individual in the t-th generation population, Represents a random number in the range [-1,1], and represents a random number uniformly distributed in the range [0,1]. and represents random numbers that follow a normal distribution;

[0197] In the individual migration stage, the position is updated by adjusting the individual migration step size; the formula used is as follows:

[0198] ;

[0199] ;

[0200] In the formula, Indicates the probability of encountering danger during the migration phase, represents the individual migration step length, Up and Lw represent the maximum and minimum values ​​of the solution space respectively, represents the fitness value of the best individual, represents the fitness value of the worst individual, , and Represents a random number uniformly distributed in the range [0,1];

[0201] Search determination, specifically, by constructing a search termination condition, searching and determining the optimal individual position, and obtaining the optimal individual position data setting;

[0202] The search termination conditions include threshold termination and iteration termination;

[0203] The threshold termination is specifically to set the fitness threshold, when the individual fitness value f i When it is above the fitness threshold, the hyperparameter search is completed;

[0204] The iteration termination specifically refers to terminating the iteration and obtaining the optimal individual position when the maximum number of iterations is reached;

[0205] The optimal individual position includes equipment operation status monitoring model parameters.

[0206] By performing the above operations, in order to solve the technical problems of improper parameter settings in traditional clustering methods and weak global search capabilities of parameter optimization algorithms, which lead to inaccurate final clustering results, this scheme uses reversible chaotic mapping to initialize the population, population event selection factors and individual migration step adjustment strategies to enhance the global search capability and adaptive balance parameter capabilities, thereby obtaining the optimal parameter values ​​and solving the problem of inaccurate output results of the final monitoring model.

[0207] Embodiment 7, see Figure 1 This embodiment is based on the above embodiment. The sewage treatment intelligent auxiliary module can locate the link where problems occur in the sewage treatment process based on the output results of the sewage treatment process water quality parameter management module and the output results of the sewage treatment equipment management module; if the water quality is abnormal, the specific parameters of the sewage treatment process are adjusted according to the output results of the sewage treatment process water quality parameter management module; if the equipment is abnormal, suggestions for sewage treatment equipment and maintenance are given according to the output results of the sewage treatment equipment management module.

[0208] By performing the above operations, the traditional sewage treatment data management system usually has the technical problem of only monitoring the water quality parameters of the sewage treatment process and the status of the sewage treatment equipment, which makes it impossible to accurately locate the problems when they occur in the sewage treatment process, thus affecting the efficiency of sewage treatment data management. This solution creatively introduces a model that combines the management of water quality parameters in the sewage treatment process with the management of sewage treatment equipment. By real-time monitoring of water quality data and equipment status, it can accurately locate the problems in the sewage treatment process and give relevant treatment suggestions, thus solving the problem of inefficiency in the sewage treatment data management process.

[0209] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0210] While the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that many changes, modifications, substitutions and variations can be made to the embodiments without departing from the principles and spirit of the invention.

[0211] The present invention and its embodiments are described above, and such description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if ordinary technicians in the field are inspired by it, without departing from the purpose of the invention, they can design a structure and embodiment similar to the technical solution without creativity, which should belong to the protection scope of the present invention.

Claims

1. The artificial intelligence-based sewage treatment data management system is characterized by: The system includes data acquisition module, data preprocessing module, sewage treatment process water quality parameter management module, sewage treatment equipment management module, hyperparameter optimization module and sewage treatment intelligent auxiliary module; The data acquisition module collects historical sewage treatment water quality data, historical sewage treatment equipment operation data, real-time sewage treatment water quality data, real-time sewage treatment equipment operation data and external environment data to obtain sewage treatment raw data; The data preprocessing module performs data cleaning, standardization, feature selection and data aggregation on the original sewage treatment data to obtain preliminary sewage treatment data; The sewage treatment process water quality parameter management module establishes a water quality parameter abnormality monitoring model through a clustering algorithm of a non-parametric kernel density estimation method and a dynamic window width parameter selection strategy, monitors the water quality data in the sewage treatment process in real time, obtains the water quality monitoring results of the current sewage treatment process, and judges the current water quality status; The sewage treatment equipment management module monitors the operating status of the sewage treatment equipment during the sewage treatment process by establishing an equipment operating status monitoring model to obtain equipment operating status monitoring results; The hyperparameter optimization module obtains the optimal parameter combination of the equipment operation status monitoring model in the sewage treatment equipment management module by using a search algorithm that initializes the population with a reversible chaotic map, a population event selection factor, and an individual migration step adjustment strategy; The sewage treatment intelligent auxiliary module locates the problems in the sewage treatment process through the output results of the sewage treatment process water quality parameter management module and the output results of the sewage treatment equipment management module, and gives relevant treatment suggestions according to the specific problems; The sewage treatment process water quality parameter management module specifically includes the following steps: A water quality parameter anomaly monitoring model is established, specifically by using a nonparametric kernel density estimation method and a dynamic window width parameter selection method to effectively identify the core points in the cluster and generate more accurate clustering results; the following steps are included: The distance between sample data points is calculated, specifically, the distance between sample data points is measured values ​​of various water quality parameters in historical sewage treatment water quality data and real-time sewage treatment water quality data. The formula used is expressed as follows: ； In the formula, Represents sample data points and sample data points The distance between Calculate the local density of the sample data points, specifically perform local density estimation by a nonparametric kernel density estimation method, including the following steps: Kernel density estimation is performed, and the formula used is expressed as follows: ； ； In the formula, represents the density estimate of the sample data point x, n represents the number of samples, h represents the window width, represents the Gaussian kernel function, represents the exponential function, q represents the qth sample data point in the data set; Perform local density estimation, specifically apply the kernel density estimation formula to the local density calculation of the data points, and extract the constant term from the formula. The formula used is expressed as follows: ； In the formula, Represents data points Local density value; Dynamic window width parameter selection, specifically selecting the optimal window width by minimizing the mean integrated square error, includes the following steps: The mean integrated square error is calculated using the following formula: ； ； In the formula, represents the mean integrated square error, f(·) represents the actual probability density function, which represents the true distribution of the data, represents the asymptotic mean of the mean integrated squared error, x represents the sample data point, represents the mathematical expectation, represents the first-order moment of the Gaussian kernel function K, represents the second-order moment of the Gaussian kernel function K, Indicates a trend of gradual growth, It represents the integral of the square of the second derivative of f(x), which measures the curvature of the density function; Select the optimal window width; the formula used is as follows: ； ； In the formula, Express Partial derivative with respect to window width h; Calculate the relative distance; the formula used is expressed as follows: ； In the formula, Represents sample data points Local density value; Represents sample data points The minimum distance to the nearest point with higher density, j means that the condition is met The sample data point index of ; Identify the cluster center, specifically by calculating the decision value of each sample data point to determine the cluster center, the decision value is obtained by multiplying the local density and the relative distance; the formula used is expressed as follows: ； In the formula, represents the decision value, represents the local density of data sample points, Represents the minimum distance between a sample data point and the nearest point with higher density; Cluster assignment, specifically, for each non-cluster center point, find the point with the closest distance and higher local density, and assign the point to the corresponding cluster center; Training of a water quality parameter anomaly monitoring model, specifically using the historical sewage treatment water quality data and external environment data in the preliminary sewage treatment data as input data to train the water quality parameter anomaly monitoring model, and analyzing the clustering results, marking the water quality state corresponding to each cluster, the water quality state including normal and abnormal, and obtaining a trained water quality parameter anomaly monitoring model; Abnormal water quality parameter monitoring, specifically, using the real-time sewage treatment water quality data and external environment data in the preliminary sewage treatment data as input data of the trained water quality parameter abnormality monitoring model to obtain the water quality monitoring results in the current sewage treatment process; The sewage treatment equipment management module specifically includes the following steps: Establishing the equipment operation status monitoring model includes the following steps: Initialize the internal structure of the model LSTM, which includes the following steps: The calculation process of the forget gate uses the following formula: ； In the formula, f t represents the output of the forget gate, X t It indicates input data. Indicates the hidden state at the previous moment, W f represents the weight matrix corresponding to the forget gate, represents the bias parameter of the forget gate, Represents the sigmoid function; Input gate calculation process, the formula used is as follows: ； ； ； In the formula, i t represents the output of the input gate, W i represents the corresponding input gate weight matrix, represents the bias parameter of the input gate, c t Indicates the current cell state. represents the candidate value of the new cell state, c t-1 Indicates the cell state at the last moment, W c Represents the corresponding cell state weight matrix, B c The bias parameter representing the cell state, represents element-wise multiplication, It is the hyperbolic tangent function; The output gate calculation process uses the following formula: ； ； In the formula, o t represents the output of the output gate, W o It represents the corresponding output gate weight matrix, B o Represents the bias parameter of the output gate, h t Indicates the hidden state at the current moment; To calculate the bidirectional LSTM, the formula used is as follows: ； ； ； Where h represents the hidden state of the combined forward and backward LSTM. Indicates the hidden state in the positive direction at the current moment, Indicates the hidden state in the positive direction at the previous moment, Indicates the reverse hidden state at the current moment. Indicates the hidden state in the reverse direction at the next moment; The final output value is obtained using the following formula: ； In the formula, Represents the output value of the monitoring model, represents the weight matrix corresponding to the output, represents the Softmax activation function, B y Represents the bias parameter of the output layer; Equipment operation status monitoring model training, specifically using the historical sewage treatment equipment operation data in the preliminary sewage treatment data to train the equipment operation status monitoring model to obtain a trained equipment operation status monitoring model; Real-time monitoring of equipment operation status, specifically, using the real-time sewage treatment equipment operation data in the preliminary sewage treatment data as input data of the trained equipment operation status monitoring model to obtain equipment operation status monitoring results; The hyperparameter optimization module specifically includes the following steps: Initialization parameters, specifically by constructing the initial parameters of the algorithm; the initial parameters of the algorithm include the population size N, the population dimension D and the maximum number of iterations ; Initializing the population, specifically, initializing the population by generating chaotic sequence values ​​through a reversible chaotic mapping, includes the following steps: Generate N chaotic sequence values ​​and initialize the random value z0; the formula used is as follows: ； ； In the formula, represents the i-th chaotic sequence value of the d-th dimension, represents the i+1th chaotic sequence value of the dth dimension, round represents a random number between 0 and 1, and A represents The mapping matrix, Indicates the current value of each dimension. represents the i-th chaotic sequence value of the first dimension, represents the i-th chaotic sequence value of the second dimension, and T represents the transposition operation; The chaotic sequence value conversion search individual is expressed as follows: ； In the formula, represents the initialization position of the dth dimension of the i-th individual, lb d andub d Respectively represent the search lower and upper limits of dimension d; Calculate the fitness value, specifically calculate the individual fitness value f in the population i ; The performance of the equipment operation status monitoring model established based on the individual position is used as the individual fitness value; The population event selection factor is calculated using the following formula: ； In the formula, represents the event selection factor of the t-th generation population, and t represents the current number of iterations; Individual exploration event stage, specifically if the current iteration population event selection factor , the individual enters the exploration event stage, and the formula used is as follows: ； In the formula, represents the d-th dimension position of the i-th individual in the t+1-th generation population, p represents a random integer uniformly distributed in the range [1, D], and r represents a random integer uniformly distributed in the range [1, N]. represents a random number uniformly distributed in the range [0,1]. represents the random dimensional position of a random individual in the t-th generation population, represents the random dimensional position of the i-th individual in the t-th generation population, It represents the remainder when d is divided by 2, where d represents the index of the dimension. represents the randomness parameter used to control the search space, representing a random number uniformly distributed in the range [0,2π]; Individual predation event stage, specifically if the current iteration population event selection factor , then the individual enters the predation event stage, and the formula used is as follows: ； In the formula, represents the position of the individual with the best fitness value in the population in the tth iteration on dimension d, represents the d-th dimension position of the i-th individual in the t-th generation population, represents the d-th dimension position of a random individual in the t-th generation population, Represents a random number in the range [-1,1], and represents a random number uniformly distributed in the range [0,1]. and represents random numbers that follow a normal distribution; In the individual migration stage, the position is updated by adjusting the individual migration step size; the formula used is as follows: ； ； In the formula, Indicates the probability of encountering danger during the migration phase, represents the individual migration step length, Up and Lw represent the maximum and minimum values ​​of the solution space respectively, represents the fitness value of the best individual, represents the fitness value of the worst individual, , and Represents a random number uniformly distributed in the range [0,1]; Search determination, specifically, by constructing a search termination condition, searching and determining the optimal individual position, and obtaining the optimal individual position data setting; The search termination conditions include threshold termination and iteration termination; The threshold termination is specifically to set the fitness threshold, when the individual fitness value f i When it is above the fitness threshold, the hyperparameter search is completed; The iteration termination specifically refers to terminating the iteration and obtaining the optimal individual position when the maximum number of iterations is reached; The optimal individual position includes equipment operation status monitoring model parameters.

2. The artificial intelligence-based sewage treatment data management system according to claim 1, characterized in that: The intelligent auxiliary module for sewage treatment is specifically based on the output results of the sewage treatment process water quality parameter management module and the output results of the sewage treatment equipment management module, and can locate the links where problems occur in the sewage treatment process; if the water quality is abnormal, the specific parameters of the sewage treatment process are adjusted according to the output results of the sewage treatment process water quality parameter management module; if the equipment is abnormal, suggestions for sewage treatment equipment and maintenance are given according to the output results of the sewage treatment equipment management module.

3. The artificial intelligence-based sewage treatment data management system according to claim 1, characterized in that: The data acquisition module specifically obtains the original sewage treatment data from the sewage treatment plant by collecting data; the original sewage treatment data includes historical sewage treatment water quality data, historical sewage treatment equipment operation data, real-time sewage treatment water quality data, real-time sewage treatment equipment operation data and external environment data; the historical sewage treatment water quality data and the real-time sewage treatment water quality data both include chemical oxygen demand, biological oxygen demand, total nitrogen, total phosphorus, dissolved oxygen, pH value and suspended solids data; the historical sewage treatment equipment operation data and the real-time sewage treatment equipment operation data both include load, vibration frequency, energy consumption, speed, pressure, temperature and equipment operation status of the aeration equipment; the external environment data includes air temperature, humidity, precipitation, wind speed, air pressure, flow rate of sewage source and initial water quality data.

4. The artificial intelligence-based sewage treatment data management system according to claim 1, characterized in that: The data preprocessing module specifically performs data cleaning, standardization, feature selection and data aggregation on the original sewage treatment data to obtain preliminary sewage treatment data; the data cleaning is to process missing values, outliers and duplicate values; the standardization is to standardize the data based on the maximum and minimum normalization method; the feature selection is to use the correlation analysis method to calculate the correlation between the features and the water quality indicators and the equipment status, select features according to the correlation, and eliminate redundant features; the data aggregation includes time dimension aggregation and space dimension aggregation, the time dimension aggregation is to perform statistical processing on the data at the time granularity of hours and days, and the space dimension aggregation is to integrate the data of different sewage treatment tanks, equipment and sensors to generate comprehensive features.

Images