Sewage treatment control system based on data analysis

Abstract

The invention relates to the technical field of sewage treatment control, in particular to a sewage treatment control system based on data analysis. The system comprises a data acquisition module, a water quality prediction module, a dynamic adjustment module and an energy-saving feedback module. The data acquisition module acquires water quality historical data, environmental factor data and sewage flow data, the water quality prediction module is used for establishing a long-short term memory network model for predicting the water quality change trend, and the dynamic adjustment module is used for analyzing the relationship between water quality indexes and equipment operation parameters by using a Pearson's correlation coefficient analysis method. An energy-saving feedback module is adopted; a genetic algorithm is adopted to find an optimal operation combination among the devices; feedback adjustment is performed on the aeration rate, the water pump flow rate and the stirring speed parameters of the dynamic adjustment module, so that when the water quality changes, the operation parameters of the devices are adjusted in time, and the sewage treatment effect is ensured; and energy consumption is reduced.

Description

Technical Field

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

[0002] The sewage treatment control system is an automated system used to monitor, control and optimize the sewage treatment process. It uses various sensors, controllers, actuators and other equipment to achieve precise control of each process link in the sewage treatment plant to ensure that the sewage treatment process runs efficiently and stably and achieves the expected water quality treatment goals.

[0003] At present, the traditional sewage treatment control system simply adjusts the parameters of each device through experience to treat sewage, and lacks consideration of environmental factors. For example, high temperature weather may accelerate the metabolism of microorganisms in the water, thereby affecting the dissolved oxygen and COD content. When facing changes in water quality, the parameters of each device cannot be adjusted in time, which affects the sewage treatment effect. In order to be able to predict water quality index data based on historical water quality data, surrounding environment data and flow data, dynamically adjust the operating parameters of each device to ensure the sewage treatment effect, and at the same time, with energy saving as the goal, find the best equipment operation combination to ensure that the sewage treatment effect is not affected and achieve energy saving and emission reduction. Therefore, we propose a sewage treatment control system based on data analysis. Summary of the invention

[0004] The purpose of the present invention is to solve the problem that the traditional sewage treatment control system cannot adjust the parameters of each device in time when the water quality changes, which affects the sewage treatment effect. In order to be able to predict the water quality index data based on the historical water quality data, the surrounding environment data and the flow data, the operating parameters of each device are dynamically adjusted to ensure the sewage treatment effect. At the same time, with energy saving as the goal, the best equipment operation combination is found to ensure that the sewage treatment effect is not affected and to achieve energy saving and emission reduction.

[0005] To achieve the above-mentioned purpose, the present invention provides a sewage treatment control system based on data analysis, including a data acquisition module, a water quality prediction module, a dynamic adjustment module and an energy-saving feedback module; The data acquisition module collects historical water quality data, environmental factor data and sewage flow data, and inputs them into the water quality prediction module to establish a long short-term memory network model for predicting water quality change trends, and uses the Adam optimizer to update the model parameters; The dynamic adjustment module uses the Pearson correlation coefficient analysis method to analyze the water quality index and the equipment operation parameters, determines the strength and direction of the relationship between the water quality index and the equipment operation parameters, and numerically quantifies them. According to the water quality index value predicted by the water quality prediction module, based on the relationship between the water quality index and the equipment operation parameters, a dynamic adjustment strategy for the equipment parameters is formulated; The energy-saving feedback module aims at energy saving, uses genetic algorithm to find the best operation combination among aeration equipment, water pump and agitator, and performs feedback adjustment on the equipment parameters of the dynamic adjustment module with water quality compliance and equipment operation as constraints.

[0006] As a further improvement of the technical solution, the water quality prediction module includes a model building unit and a model optimization unit; The model building unit determines the input layer dimension based on the dimension of the input data, and adds multiple LSTM layers thereafter, maps the output of the LSTM layer to the final output dimension using a fully connected layer, selects a linear activation function in the output layer, and predicts the water quality index value; The model optimization unit uses a validation set or a test set to evaluate the trained model and adjusts the model parameters according to the evaluation results.

[0007] As a further improvement of the present technical solution, the model building unit uses a mean square error loss function to measure the difference between the predicted value and the true value. The formula of the mean square error loss function is: ; in, is the mean square error loss value, is the sample size, is the predicted value of water quality index, is the true value of water quality index.

[0008] As a further improvement of the technical solution, the model optimization unit uses grid search to preliminarily traverse the model parameters, performs random search in the excellent area of ​​the grid search, and selects the optimal hyperparameter combination.

[0009] As a further improvement of the technical solution, the dynamic adjustment module includes a contact determination unit and a strategy formulation unit; The Pearson correlation coefficient analysis method for determining the connection unit analyzes the relationship between the water quality index and the equipment operation parameter, determines the strength and direction of the relationship between the water quality index and the equipment operation parameter, and quantifies it numerically; The strategy formulation unit sets a clear target value or target range for each water quality indicator based on the water quality discharge standard and the operation requirements of the sewage treatment plant, and formulates a dynamic adjustment strategy for the equipment parameters based on the relationship between the water quality indicator and the equipment operation parameters.

[0010] As a further improvement of the technical solution, the determining connection unit uses a t-test to perform a significance test on the calculated correlation coefficient, and the calculation formula of the t-test statistic is: ; in, is a statistic, To measure the correlation coefficient of the linear relationship between water quality indicators and equipment parameters, is the sample size.

[0011] As a further improvement of the technical solution, the strategy formulation unit sets positive relationship adjustment rules and negative relationship adjustment rules as basic adjustment rules, and determines the adjustment direction by calculating the deviation.

[0012] As a further improvement of the present technical solution, the strategy formulation unit divides the deviation intervals into mild deviation, moderate deviation and severe deviation, and sets an adjustment gradient for each interval.

[0013] As a further improvement of the present technical solution, the energy-saving feedback module takes energy saving as the goal, sets the objective function as minimizing the total energy consumption, defines the inverse of the total energy consumption as the fitness function value, and uses the penalty function method to process the constraints.

[0014] As a further improvement of the technical solution, the energy-saving feedback module uses selection, crossover and mutation to perform genetic operations, with the termination condition being that the fitness value no longer increases.

[0015] Compared with the prior art, the present invention has the following beneficial effects: In the sewage treatment control system based on data analysis, the data acquisition module collects historical water quality data, environmental factor data and sewage flow data, arranges the cleaned data in time series, and uses the water quality prediction module to establish a long short-term memory network model for predicting water quality change trends, and trains and optimizes it. The dynamic adjustment module uses the Pearson correlation coefficient analysis method to analyze the relationship between water quality indicators and equipment operating parameters, determines the strength and direction of the relationship between water quality indicators and equipment operating parameters, and quantifies them numerically. According to the water quality indicator values ​​predicted by the water quality prediction module, based on the relationship between water quality indicators and equipment operating parameters, a dynamic adjustment strategy for equipment parameters is formulated, so as to achieve timely adjustment of the operating parameters of each device when water quality changes, ensure the sewage treatment effect, and improve sewage treatment efficiency. The energy-saving feedback module aims at energy saving and uses genetic algorithms to find the best operating combination among aeration equipment, water pumps and agitators. It also uses water quality compliance and equipment operation as constraints to feedback and adjust the aeration volume, water pump flow and stirring speed parameters of the dynamic adjustment module. It adjusts the operating combination of various equipment while effectively treating sewage, thereby improving energy utilization and reducing energy consumption.

[0016] In addition to the above-described purposes, features and advantages, the present invention has other purposes, features and advantages. The present invention will be further described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a schematic diagram of the overall process of the present invention; Figure 2 It is a schematic diagram of the overall detailed process of the present invention.

[0018] The meaning of each number in the figure is: 100. Data collection module; 200. Water quality prediction module; 210. Model building unit; 220. Model optimization unit; 300. Dynamic adjustment module; 310. Contact determination unit; 320. Strategy formulation unit; 400. Energy-saving feedback module. DETAILED DESCRIPTION

[0019] The following will be combined with the accompanying drawings in the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0020] At present, traditional sewage treatment control systems cannot adjust the parameters of various equipment in time when facing changes in water quality, which affects the sewage treatment effect. In order to be able to predict water quality index data based on historical water quality data, surrounding environment data and flow data, dynamically adjust the operating parameters of various equipment to ensure the sewage treatment effect. At the same time, with energy saving as the goal, find the best equipment operation combination to ensure that the sewage treatment effect is not affected and achieve energy saving and emission reduction.

[0021] Therefore, the present invention proposes to collect water quality historical data, environmental factor data and sewage flow data through a data acquisition module, arrange the cleaned data in time series, use a water quality prediction module to establish a long short-term memory network model for predicting water quality change trends, and train and optimize the model, a dynamic adjustment module uses the Pearson correlation coefficient analysis method to analyze the relationship between water quality indicators and equipment operating parameters, and formulates a dynamic adjustment strategy for equipment parameters based on the relationship between water quality indicators and equipment operating parameters according to the water quality indicator values ​​predicted by the water quality prediction module. The energy-saving feedback module takes energy saving as the goal, uses a genetic algorithm to find the best operating combination between aeration equipment, a water pump and agitator, and performs feedback adjustment on the aeration volume, water pump flow and stirring speed parameters of the dynamic adjustment module.

[0022] The details are as follows: See also Figure 1 As shown, the present invention provides a sewage treatment control system based on data analysis, including a data acquisition module 100, a water quality prediction module 200, a dynamic adjustment module 300 and an energy-saving feedback module 400; The data acquisition module 100 collects historical water quality data, environmental factor data and sewage flow data, and inputs them into the water quality prediction module 200 to establish a long short-term memory network model for predicting water quality change trends, and uses the Adam optimizer to update the model parameters; Obtain historical records of dissolved oxygen, chemical oxygen demand, biochemical oxygen demand, ammonia nitrogen, and pH indicators from the sewage treatment plant's monitoring system and water quality test reports. The data time span should be as long as possible to cover water quality conditions in different seasons and under different operating conditions. Collect data on environmental factors such as temperature, air pressure, rainfall, and sunshine duration through meteorological departments and environmental monitoring stations to ensure that the time of these data corresponds to the water quality data. Obtain sewage flow data from the flow meter equipment of the sewage treatment plant and record the flow size at different time points.

[0023] like Figure 2 As shown, the water quality prediction module 200 includes a model building unit 210 and a model optimization unit 220; The model building unit 210 determines the input layer dimension according to the dimension of the input data, and adds multiple LSTM layers thereafter, maps the output of the LSTM layer to the final output dimension using a fully connected layer, selects a linear activation function in the output layer, and predicts the water quality index value; The model optimization unit 220 uses the validation set or the test set to evaluate the trained model and adjusts the model parameters according to the evaluation results; The preprocessed data is divided into training set, validation set and test set according to a certain ratio (70% training set, 15% validation set, 15% test set). The training set is used for model training, the validation set is used to adjust the model's hyperparameters, and the test set is used to evaluate the final performance of the model.

[0024] In order to better measure the difference between the predicted value and the true value, the model building unit 210 uses a mean square error loss function to measure the difference between the predicted value and the true value. The formula of the mean square error loss function is: ; in, is the mean square error loss value, is the sample size, is the predicted value of water quality index, is the true value of water quality index.

[0025] During the training process, the model will continuously adjust its own parameters according to the feedback of the loss function to make the predicted value as close to the true value as possible. When the model achieves satisfactory performance on the validation set, the test set is used to conduct a final evaluation of the model. If the results on the test set also show that the model has good predictive ability, then the model can be used to predict new and unknown water quality index values. When the new input feature data is input into the trained model, the model will output the corresponding water quality index prediction value.

[0026] In order to better select the optimal hyperparameter combination, the model optimization unit 220 uses a grid search to preliminarily traverse the model parameters, performs a random search in the excellent area of ​​the grid search, and selects the optimal hyperparameter combination; Design a relatively coarse-grained grid, that is, the number of values ​​for each hyperparameter is relatively small. For example, for the learning rate hyperparameter, select a value set such as [0.001, 0.01, 0.1], use this coarse-grained grid to perform grid search, traverse all possible hyperparameter combinations, evaluate the model performance on the validation set, and record the hyperparameter combinations with better performance and their corresponding performance indicators; Analyze the results of the grid search in the above steps and determine the area where the hyperparameter combination with better performance is located. For example, if it is found that the model performance is better when the learning rate is around 0.01 and the number of hidden layer neurons is between 100-200, then this area will be used as the key area for the next random search. Random search will be performed in the determined key area. During the random search, the values ​​of the hyperparameters can be randomly sampled more carefully according to the range of the key area. For example, the learning rate can be randomly sampled between 0.005-0.015, and the number of hidden layer neurons can be randomly sampled between 100-200. Perform multiple random sampling and model training evaluations, record the results of each time, select the hyperparameter combination with the best performance as the final result, and update the model parameters.

[0027] The dynamic adjustment module 300 further analyzes the relationship between the water quality index and the equipment operating parameters using the Pearson correlation coefficient analysis method, determines the strength and direction of the relationship between the water quality index and the equipment operating parameters, and numerically quantifies it, and formulates a dynamic adjustment strategy for the equipment parameters based on the relationship between the water quality index and the equipment operating parameters according to the water quality index value predicted by the water quality prediction module 200; The dynamic adjustment module 300 includes a contact determination unit 310 and a strategy formulation unit 320; Determine the relationship unit 310 Pearson correlation coefficient analysis method to analyze the relationship between the water quality index and the equipment operation parameter, determine the strength and direction of the relationship between the water quality index and the equipment operation parameter, and quantify it numerically; The strategy formulation unit 320 sets a clear target value or target range for each water quality indicator based on the water quality discharge standard and the operation requirements of the sewage treatment plant, and formulates a dynamic adjustment strategy for the equipment parameters based on the relationship between the water quality indicator and the equipment operation parameters; The Pearson correlation coefficient is used to measure the degree of linear correlation between two variables, which can help us explore the relationship between water quality index values ​​and equipment operating parameters. The Pearson correlation coefficient range is [-1,1]. If it is close to 1, it means that the two variables are positively linearly correlated, that is, when one variable increases, the other also increases. If it is close to -1, it means negative linear correlation, when one variable increases, the other decreases. If it is close to 0, it means weak linear correlation. The p value is used to test the significance of correlation, usually with a threshold of 0.05. If the p value is less than 0.05, it indicates that the correlation between the two variables is significant at this significance level. If it is greater than 0.05, the correlation is not significant, which may be caused by random factors.

[0028] In order to calculate the p value more accurately, the determining association unit 310 uses a t-test to perform a significance test on the calculated correlation coefficient, and the calculation formula of the t-test statistic is: ; in, is a statistic, To measure the correlation coefficient of the linear relationship between water quality indicators and equipment parameters, is the sample size.

[0029] According to the above formula, the sample size and the calculated correlation coefficients for the linear relationship between water quality indicators and equipment parameters Substitute in and calculate Value, determine the degrees of freedom: ; According to the formula: ; Calculate the p-value. Usually, when the P-value is less than the pre-set significance level, the null hypothesis is rejected and the correlation between the two variables is considered significant. When the P-value is greater than or equal to the significance level, the null hypothesis cannot be rejected, which means that there is insufficient evidence to show that there is a significant correlation between the two variables.

[0030] In order to better determine the adjustment direction, the strategy formulation unit 320 sets the positive relationship adjustment rule and the negative relationship adjustment rule as the basic adjustment rule, and determines the adjustment direction by calculating the deviation; Positive adjustment rule setting: Through data analysis and actual experience, determine the water quality indicators and equipment operating parameters that have a positive relationship. For example, aeration volume and dissolved oxygen are usually positively correlated, that is, as aeration volume increases, the dissolved oxygen content in the water will also increase. The stirring speed and mixing uniformity may be positively correlated. The faster the stirring speed, the more uniform the sewage mixing, and the decomposition efficiency of certain pollutants may be improved; Monitor water quality indicators in real time, compare them with target values, and calculate deviation values. The deviation value is the actual value minus the target value. When the actual value of the water quality indicator is lower than the target value, increase the corresponding equipment operating parameters. When the actual value is higher than the target value, reduce the equipment operating parameters. Negative adjustment rule setting: Common negative relationships include dosage and certain pollutant concentrations. Increasing the dosage of the agent will reduce the pollutant concentration. Sludge return ratio and effluent suspended solids content. Increasing the sludge return ratio may reduce the effluent suspended solids content. Similarly, the water quality index is monitored in real time, and its deviation from the target value is calculated. When the actual value of the water quality index is higher than the target value, the corresponding equipment operating parameters are increased; when the actual value is lower than the target value, the equipment operating parameters are reduced; When increasing the equipment operating parameters, be careful to avoid side effects caused by over-adjustment. For example, too much reagent may lead to increased subsequent treatment costs, secondary pollution and other problems. Therefore, various factors should be considered comprehensively during the adjustment process to ensure the rationality and effectiveness of the adjustment measures.

[0031] In order to better set the adjustment gradient of each interval, the strategy formulation unit 320 divides the deviation intervals into mild deviation, moderate deviation and severe deviation, and sets the adjustment gradient of each interval; According to the target values ​​of water quality indicators and actual operation experience, the deviation range of water quality indicators is divided into several intervals. For example, for the chemical oxygen demand (COD) indicator, it can be divided into mild deviation (deviation within ±10%), moderate deviation (deviation between 10%-30%) and severe deviation (deviation exceeds 30%) intervals; Mild deviation range: Small adjustment: In the range of slight deviation, a smaller adjustment gradient should be used to avoid excessive adjustment leading to system instability. For example, when the dissolved oxygen index deviation is within ±5%, the aeration volume of the aeration equipment is adjusted to 1%-3% of the rated aeration volume each time; Progressive adjustment: gradually adjust equipment parameters in smaller steps, closely monitor water quality changes, and make timely adjustments based on feedback results; Moderate deviation interval: Moderate adjustment: In the moderate deviation range, the adjustment gradient can be appropriately increased. For example, when the ammonia nitrogen index deviation is between 10% and 20%, the dosage of the dosing equipment is adjusted to 5% to 10% of the current dosage each time; Combined with forecast adjustment: refer to the water quality forecast model to estimate the water quality changes after adjustment and set the adjustment range more targetedly; Severe deviation range: Large-scale adjustment: In the severe deviation range, a large adjustment gradient is required to restore the water quality to the target range as quickly as possible. For example, when the COD index deviation exceeds 30%, the aeration volume of the aeration equipment can be increased by 10%-20% at one time, and the operating intensity of other auxiliary equipment can be increased at the same time; Emergency measures coordination: In addition to adjusting equipment parameters, some emergency measures can also be taken, such as adding pretreatment steps and adjusting water inlet flow.

[0032] In addition, the energy-saving feedback module 400 takes energy saving as the goal, uses genetic algorithms to find the best operation combination between the aeration equipment, water pump and agitator, and takes water quality compliance and equipment operation as constraints to feedback and adjust the equipment parameters of the dynamic adjustment module 300; In order to better set the fitness function, the energy-saving feedback module 400 takes energy saving as the goal, sets the objective function as the minimum total energy consumption, defines the inverse of the total energy consumption as the fitness function value, uses the penalty function method to handle the constraints, and uses the genetic algorithm to find the best operation combination between the aeration equipment, the water pump and the agitator. The steps are as follows: Aeration volume, water pump flow, and stirring speed are used as decision variables. Each variable corresponds to an operating parameter of the equipment. With energy saving as the goal, the objective function can be set to minimize the total energy consumption. The total energy consumption is the sum of the energy consumption of each device. The energy consumption of the equipment is related to the equipment power and operating time. Set constraints: Water quality compliance constraints: various water quality indicators after sewage treatment (such as chemical oxygen demand, ammonia nitrogen, etc.) must meet relevant standards; Equipment operation constraints: The operating parameters of each equipment must be within its safe and feasible range, for example, the aeration volume has minimum and maximum limits; Real number coding is used to directly represent the operating parameters of each device with a real number. A chromosome is a vector composed of parameters such as aeration volume, water pump flow, stirring speed, etc. For example, chromosome [A, Q, S], where A represents aeration volume, Q represents water pump flow, and S represents stirring speed; A certain number of chromosomes are randomly generated to form the initial population. The population size is set according to the complexity of the problem, generally ranging from dozens to hundreds. The fitness function is related to the objective function. Since the goal is to minimize the total energy consumption, the fitness value can be defined as the inverse of the total energy consumption. The lower the energy consumption, the higher the fitness value. Use the penalty function method to handle constraints. If the operation combination corresponding to the chromosome does not meet the water quality standard or equipment operation constraints, a penalty term is added to the fitness function to reduce its fitness value until the optimal operation combination between the aeration equipment, water pump and agitator is generated. In order to better perform genetic operations, the energy-saving feedback module 400 uses selection, crossover and mutation to perform genetic operations, and the termination condition is that the fitness value no longer increases; Selection operation: commonly used roulette selection or tournament selection. Taking roulette selection as an example, the probability of each chromosome being selected is proportional to the proportion of its fitness value to the total fitness value; Crossover operation: single-point crossover or multi-point crossover. For example, single-point crossover randomly selects a crossover point and exchanges the parts of the two chromosomes after the point to generate two new chromosomes. Mutation operation: randomly change a gene value in a chromosome with a certain mutation probability. During mutation, the new gene value must be within the parameter value range; In each iteration, selection, crossover, and mutation operations are performed in sequence to generate a new population, and the fitness value of each chromosome in the new population is calculated. This process is repeated until the termination condition is met.

[0033] In summary, the working principle of this solution is as follows: In the sewage treatment control system based on data analysis, the data acquisition module 100 collects water quality historical data, environmental factor data and sewage flow data, arranges the cleaned data in time series, and uses the water quality prediction module 200 to establish a long short-term memory network model for predicting water quality change trends, and trains and optimizes it. The dynamic adjustment module 300 uses the Pearson correlation coefficient analysis method to analyze the relationship between water quality indicators and equipment operating parameters, determines the strength and direction of the relationship between water quality indicators and equipment operating parameters, and quantifies it numerically. According to the water quality indicator value predicted by the water quality prediction module 200, the water quality indicator and the equipment operating parameters are used to calculate the relationship between the water quality indicators and the equipment operating parameters. Based on the relationship between the equipment operating parameters, a dynamic adjustment strategy for the equipment parameters is formulated to achieve timely adjustment of the operating parameters of each device when the water quality changes, to ensure the sewage treatment effect and improve the sewage treatment efficiency. The energy-saving feedback module 400 takes energy saving as the goal, uses genetic algorithms to find the best operating combination among the aeration equipment, water pump and agitator, and takes water quality compliance and equipment operation as constraints to feedback and adjust the aeration volume, water pump flow and stirring speed parameters of the dynamic adjustment module 300, so as to adjust the operating combination between each device on the premise of effectively treating sewage, improve energy utilization and reduce energy consumption.

[0034] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and descriptions are only preferred examples of the present invention and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A sewage treatment control system based on data analysis, characterized in that: It comprises a data acquisition module (100), a water quality prediction module (200), a dynamic adjustment module (300) and an energy-saving feedback module (400); The data collection module (100) collects historical water quality data, environmental factor data and sewage flow data, and inputs the data into the water quality prediction module (200) to establish a long short-term memory network model for predicting water quality change trends, and uses an Adam optimizer to update model parameters; The dynamic adjustment module (300) uses the Pearson correlation coefficient analysis method to analyze the water quality index and the equipment operation parameters, determines the strength and direction of the relationship between the water quality index and the equipment operation parameters, and formulates a dynamic adjustment strategy for the equipment parameters based on the relationship between the water quality index and the equipment operation parameters according to the water quality index value predicted by the water quality prediction module (200); The energy-saving feedback module (400) aims at energy saving, uses a genetic algorithm to find the best operation combination among the aeration equipment, the water pump and the agitator, and performs feedback adjustment on the equipment parameters of the dynamic adjustment module (300) with water quality compliance and equipment operation as constraints.

2. The sewage treatment control system based on data analysis according to claim 1 is characterized in that: The water quality prediction module (200) comprises a model building unit (210) and a model optimization unit (220); The model building unit (210) receives data from the data acquisition module (100) as input data, including obtaining historical water quality data of dissolved oxygen, chemical oxygen demand, biochemical oxygen demand, ammonia nitrogen, and pH index from the monitoring system of the sewage treatment plant and the water quality test report, collecting environmental factor data of temperature, air pressure, rainfall, and light duration through the meteorological department and the environmental monitoring station, ensuring that the time of the environmental factor data corresponds to the water quality data, determining the input layer dimension based on the dimension of the input data, and adding multiple LSTM layers thereafter, mapping the output of the LSTM layer to the final output dimension using a fully connected layer, selecting a linear activation function in the output layer, dividing the input data into a training set, a validation set, and a test set, and the training set is used for training the long short-term memory network model to predict the water quality index value; The model optimization unit (220) uses the Adam optimizer to evaluate the trained model using a validation set or a test set, and adjusts the model parameters according to the evaluation results.

3. The sewage treatment control system based on data analysis according to claim 2 is characterized in that: The model building unit (210) uses a mean square error loss function to measure the difference between the predicted value and the true value. The formula of the mean square error loss function is: ; in, is the mean square error loss value, is the sample size, is the predicted value of water quality index, is the true value of water quality index.

4. The sewage treatment control system based on data analysis according to claim 2 is characterized in that: The model optimization unit (220) uses grid search to perform a preliminary traversal of model parameters, performs a random search in an excellent area of ​​the grid search, and selects the optimal hyperparameter combination to update the model parameters.

5. The sewage treatment control system based on data analysis according to claim 1 is characterized in that: The dynamic adjustment module (300) comprises a contact determination unit (310) and a strategy formulation unit (320); The determining connection unit (310) measures the linear correlation between two variables by using the Pearson correlation coefficient in the Pearson correlation coefficient analysis method, explores the relationship between the water quality index value and the equipment operation parameter, determines the strength and direction of the relationship between the water quality index value and the equipment operation parameter, and numerically quantifies the relationship; The strategy formulation unit (320) sets a clear target value or target range for each water quality indicator based on the water quality discharge standard and the operation requirements of the sewage treatment plant, and formulates a dynamic adjustment strategy for the equipment parameters based on the relationship between the water quality indicator and the equipment operation parameters.

6. The sewage treatment control system based on data analysis according to claim 5 is characterized in that: The determining association unit (310) uses a t-test to perform a significance test on the calculated Pearson correlation coefficient, and the calculation formula of the t-test statistic is: ; in, is a statistic, To measure the correlation coefficient of the linear relationship between water quality indicators and equipment parameters, is the sample size, and the sample size and the calculated correlation coefficients for the linear relationship between water quality indicators and equipment parameters Substitute and calculate Value, determine the degrees of freedom: ; According to the formula: ; Calculate the p-value used to test the significance of the correlation. Usually, when the p-value is less than the pre-set significance level, reject the null hypothesis and consider that the correlation between the two water quality indicators and the equipment parameters is significant. When the p-value is greater than or equal to the significance level, fail to reject the null hypothesis, which means that there is insufficient evidence to show that there is a significant correlation between the two water quality indicators and the equipment parameters.

7. The sewage treatment control system based on data analysis according to claim 5 is characterized in that: The strategy formulation unit (320) sets a positive relationship adjustment rule and a negative relationship adjustment rule as basic adjustment rules, and determines the adjustment direction by calculating the deviation.

8. The sewage treatment control system based on data analysis according to claim 7 is characterized in that: The strategy formulation unit (320) divides the deviation into mild deviation, moderate deviation and severe deviation, and sets an adjustment gradient for each interval.

9. The sewage treatment control system based on data analysis according to claim 1 is characterized in that: The energy-saving feedback module (400) uses a genetic algorithm to find the best operation combination between the aeration equipment, the water pump and the agitator in the following steps: the aeration volume, the water pump flow rate and the agitation speed are used as decision variables, each variable corresponds to an operation parameter of the equipment, energy saving is taken as the goal, and the objective function is set to minimize the total energy consumption; Using real number coding, the operating parameters of each device are directly represented by a real number, and a certain number of chromosomes are randomly generated to form the initial population. A chromosome is a vector composed of aeration volume, water pump flow, and stirring speed parameters. The inverse of the total energy consumption is defined as the fitness function value, and the penalty function method is used to deal with the constraints. If the operating combination corresponding to the chromosome does not meet the water quality standards or equipment operation constraints, a penalty term is added to the fitness function to reduce its fitness value until the optimal operating combination of the aeration equipment, water pump and agitator is generated.

10. The sewage treatment control system based on data analysis according to claim 1, characterized in that: The energy-saving feedback module (400) uses selection, crossover and mutation to perform genetic algorithm operations, with the termination condition being that the fitness value no longer increases.

Images