Water environment monitoring system and method based on big data analysis

Abstract

The invention discloses a water environment monitoring system and method based on big data analysis, and the system comprises a data collection and processing module which is responsible for collecting and processing original data related to a water environment, and transmits the processed data to a water flow and pollutant diffusion modeling module; and the water flow and pollutant diffusion modeling module is responsible for constructing a hydrodynamic model by utilizing a CFD simulation method based on the processed data, and predicting a pollutant diffusion trend through a convection-diffusion equation. According to the invention, the data analysis and prediction module deeply excavates potential information in acquired data and simulation results through a K-means clustering algorithm and an LSTM algorithm, and realizes accurate prediction of pollutant diffusion and water quality change time series. And the early warning and emergency response module can discover potential water quality pollution events in advance based on the prediction results and the real-time monitoring data, and give out early warning in time, so that the quick response capability of coping with the water quality pollution events is enhanced.

Description

Technical Field

[0001] The present invention relates to the technical field of environmental monitoring, and particularly to a water environment monitoring system and method based on big data analysis. Background Art

[0002] Big data refers to the high-speed acquisition, transmission, storage, and analysis of a large number of complex data of various types and sources, extracting its value using relatively economical methods, and obtaining high-value products and services by processing massive data. With the continuous progress of information technology, big data technology has been widely applied in the field of water quality monitoring, providing strong technical support for the construction of water environment monitoring systems.

[0003] After retrieval, the invention patent with the Chinese patent number CN114387235A discloses a water environment monitoring method and system, which acquires an image of a target water area, preprocesses the image of the target water area as an image to be recognized, acquires the region of interest of the image to be recognized, and extracts test samples from the region of interest using a preset sliding window; classifies the extracted test samples, and classifies the region of interest of the image to be recognized into a water gauge type or a water flow type; determines the interface between the water gauge type and the water flow type according to the classification result and uses it as the pixel water level value of the image to be recognized; obtains the actual water level value according to the pixel water level value, and determines whether the actual water level value is greater than or equal to a preset water level value; if it is greater than or equal to, an alarm message is sent; if it is less than, the water quality of the target water area is monitored. The present invention can realize the monitoring of the water level height of the target water area through image recognition, and can send an alarm message when the water level value exceeds the preset water level value, thereby avoiding situations such as dam break or overtopping of the target water area.

[0004] However, in actual use of the above method, water environment monitoring is often interfered by the factor of water flow velocity, which may affect the accuracy of water quality monitoring. For example, the change of water flow velocity will directly affect the diffusion and degradation process of pollutants in water. Fast water flow can accelerate the diffusion of pollutants, covering a wider area in a short time, while slow water flow may cause pollutants to accumulate in local areas, increasing the risk of water quality deterioration. Therefore, a water environment monitoring system and method based on big data analysis are proposed. Summary of the Invention

[0005] The purpose of the present invention is to solve the drawback that water environment monitoring in the prior art is often interfered by the factor of water flow velocity, which may affect the accuracy of water quality monitoring, and to propose a water environment monitoring system and method based on big data analysis.

[0006] To achieve the above purpose, the present invention adopts the following technical solutions:

[0007] A water environment monitoring system based on big data analysis includes:

[0008] Data acquisition and processing module: responsible for collecting and processing raw data related to the water environment, and the data acquisition and processing module transfers the processed data to the water flow and pollutant diffusion modeling module;

[0009] Water flow and pollutant diffusion modeling module: responsible for constructing a hydrodynamic model based on the processed data using the CFD simulation method, predicting the pollutant diffusion trend through the convection-diffusion equation, and establishing an accumulation model of pollutants in a local area. The water flow and pollutant diffusion modeling module transfers the simulation results to the data analysis and prediction module;

[0010] Data analysis and prediction module: responsible for mining the collected data and simulation results using the K-means clustering algorithm, and then predicting the time series of pollutant diffusion and water quality changes using the LSTM algorithm. The data analysis and prediction module transfers the prediction results to the early warning and emergency response module;

[0011] Early warning and emergency response module: responsible for issuing early warnings for potential water quality pollution events based on the pollution trend prediction results and real-time monitoring data, and formulating emergency response strategies. The early warning and emergency response module transfers the early warning results to the model optimization and feedback module;

[0012] Model optimization and feedback module: responsible for comparing the prediction results with the actual monitoring data and optimizing the model parameters. The model optimization and feedback module feeds back the optimized parameters to the water flow and pollutant diffusion modeling module and the data analysis and prediction module.

[0013] The above technical solution further includes:

[0014] Preferably, the specific steps for collecting raw data related to the water environment are as follows:

[0015] According to the purpose and requirements of water environment monitoring, determine the water quality parameters to be monitored, such as dissolved oxygen, pH value, turbidity, temperature, heavy metal content, organic pollutant concentration, etc., and identify and determine the sources of obtaining the monitoring index data, such as online monitoring equipment (such as water quality sensors), laboratory analysis results, historical monitoring databases, remote sensing data, public reports, etc.;

[0016] Configure data acquisition equipment, such as water quality sensors, data recorders, etc., according to the monitoring indicators and data sources, install the equipment at the monitoring location, and calibrate the equipment to ensure its accuracy and reliability. The calibration process may involve using standard solutions for comparison tests, or making corrections with reference to other known accurate measurement results;

[0017] Collect data at set time intervals or trigger conditions (such as abnormal water quality changes).

[0018] Preferably, the specific steps for constructing a hydrodynamic model using the CFD simulation method are as follows:

[0019] Problem definition and geometric modeling: According to the monitored water body conditions, define the geometric model of the simulation area, including the watershed, rivers, lakes, or specific water environment areas of the water area. Use computational fluid dynamics software, such as ANSYS Fluent, etc., to generate the geometric grid of the watershed and perform grid division on the geometric domain. For example, use the finite volume method (FVM) for grid generation to ensure that the grid is fine enough to obtain accurate flow field and pollutant diffusion results;

[0020] Modeling of the water body kinematic equation: Use the Navier-Stokes equation to describe the motion behavior of the fluid where u is the fluid velocity vector, ρ is the fluid density, ν is the fluid viscosity coefficient, p is the fluid pressure, f is the external force term, such as gravity, and set boundary conditions to simulate the real situation of the water body, including the inlet boundary, outlet boundary, and wall boundary. The inlet boundary represents the initial values of the water flow velocity, temperature, or concentration. The outlet boundary represents the pressure boundary condition, and the wall boundary represents setting the wall no-slip condition or dealing with the adsorption and release of pollutants.

[0021] Preferably, the specific steps for predicting the pollutant diffusion trend through the convection-diffusion equation are as follows:

[0022] Pollutant diffusion is described by the convection-diffusion equation as: where C is the pollutant concentration, D is the pollutant diffusion coefficient, is the diffusion term, representing the spatial diffusion of the pollutant concentration, and R(C) is the reaction term, representing the possible degradation or transformation process of the pollutant;

[0023] Discretize and solve the convection-diffusion equation to obtain the variation of the pollutant concentration field with time. The discretization formula is: where, is the pollutant concentration at position i at time n, is the pollutant concentration at position j at time n, Δt is the time step, Δx is the spatial step, and u ij is the flow velocity.

[0024] Preferably, the specific process for establishing the accumulation model of pollutants in the local area is as follows:

[0025] The concentration calculation formula for local accumulation is: where C local (x,t) represents the accumulation of the pollutant concentration in the local area, t represents the time elapsed during the simulation process, and τ represents the time variable.

[0026] Preferably, the specific steps for mining the collected data and simulation results using the K-means clustering algorithm are as follows:

[0027] Data preprocessing: Preprocess the collected data and simulation results, such as denoising and cleaning, standardization or normalization, time series arrangement, etc., to ensure data quality, integrity, and consistency;

[0028] Data integration and feature engineering: Integrate data from different modules (such as water flow and pollutant diffusion modeling modules and real-time monitoring data) to generate a unified feature matrix;

[0029] Select the value of K: According to the theoretical knowledge of water pollution diffusion and in combination with the actual situation, preset the value of K, for example, how many water quality regions or pollution levels are expected to be distinguished;

[0030] Apply the K-means algorithm for clustering: Randomly select K initial cluster centers, assign each data point to the cluster center closest to it, that is, assign the data points to their respective clusters according to metrics such as Euclidean distance, and update the centroid of the cluster according to the mean of the data points in each cluster. Repeat the assignment and update steps until the cluster centers no longer change or reach the preset maximum number of iterations;

[0031] Analyze the clustering results: According to the data characteristics in each cluster, analyze the representativeness of the clusters, assign labels to each cluster, such as "low pollution area" and "high pollution area", classify the clusters according to pollutant concentration and flow velocity attributes, and visualize the clustering results using a three-dimensional scatter plot.

[0032] Preferably, the specific process for predicting the time series of pollutant diffusion and water quality changes using the LSTM algorithm is as follows:

[0033] Data preparation: Obtain the preprocessed water quality monitoring data, including time series data of key water quality indicators such as dissolved oxygen, pH value, ammonia nitrogen, etc., and divide the collected data into a training set, a validation set, and a test set. The training set accounts for 70% - 80%, and the validation set and the test set each account for 10% - 15%;

[0034] LSTM model construction: Define the LSTM network structure, including an input layer, LSTM layers (possibly including multiple layers), a fully connected layer, and an output layer, and set the number of units in the LSTM layer, learning rate, optimizer, and loss function parameters;

[0035] Model training: Input the data of the training set into the LSTM model, including the features of time series data and the corresponding target values (i.e., water quality indicators within a future period). Through the memory units and gating mechanisms (input gate, forget gate, output gate) of the LSTM layer, calculate the output of the model. Use the loss function to calculate the error between the model output and the actual target value. According to the gradient of the loss function, update the parameters of the LSTM model through the backpropagation algorithm to minimize the error. Use the optimizer to adjust the learning rate to accelerate the training process of the model;

[0036] Model prediction: Input the test set into the trained LSTM model, and calculate the output of the model through forward propagation, that is, obtain the predicted values of water quality indicators within a future period. Use the test set data to evaluate the prediction performance of the model, and optimize the model according to the evaluation results.

[0037] Preferably, the early warning and emergency response module includes a pollution trend assessment unit, an early warning determination unit, an emergency response decision-making unit, and an information release and notification unit. The pollution trend assessment unit is responsible for assessing the current situation and future change trends of water quality. The early warning determination unit is responsible for judging whether to trigger an early warning according to the output of the pollution trend assessment unit and the early warning threshold. The emergency response decision-making unit is responsible for formulating emergency response strategies. The information release and notification unit is responsible for transmitting the early warning information and emergency response measures to relevant personnel.

[0038] Preferably, a water environment monitoring method corresponding to a water environment monitoring system based on big data analysis includes:

[0039] S1: Collect the original data related to the water environment and perform preprocessing;

[0040] S2: Use the CFD simulation method to construct a hydrodynamic model for simulating the water flow movement state based on the processed data. By solving the convection-diffusion equation, predict the pollutant diffusion trend and distribution, and establish a pollutant accumulation model in a local area to evaluate the pollutant accumulation effect;

[0041] S3: Use the K-means clustering algorithm to mine the collected data and simulation results, and use the LSTM algorithm to predict the time series of pollutant diffusion and water quality changes;

[0042] S4: According to the pollution trend prediction results and real-time monitoring data, give early warnings for water quality pollution events, and formulate emergency response strategies according to the early warning results;

[0043] S5: Compare the prediction results with the actual data and optimize the model parameters.

[0044] The present invention has the following beneficial effects:

[0045] 1. In the present invention, the water flow and pollutant diffusion modeling module uses the CFD simulation method and the convection-diffusion equation to accurately simulate the dynamic relationship between the water flow velocity and the pollutant diffusion, thereby more accurately predicting the diffusion trend of pollutants and the accumulation situation in local areas.

[0046] 2. In the present invention, the data analysis and prediction module deeply mines the potential information in the collected data and simulation results through the K-means clustering algorithm and the LSTM algorithm, and realizes the accurate prediction of the time series of pollutant diffusion and water quality change. Based on these prediction results and real-time monitoring data, the early warning and emergency response module can detect potential water quality pollution events in advance and issue early warnings in a timely manner, thereby enhancing the rapid response ability to water quality pollution events. Description of the Drawings

[0047] Figure 1 It is a system architecture diagram of a water environment monitoring system based on big data analysis proposed by the present invention;

[0048] Figure 2 It is a flowchart of a water environment monitoring method based on big data analysis proposed by the present invention. Detailed Embodiments

[0049] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0050] As Figure 1 shown, a water environment monitoring system based on big data analysis includes:

[0051] The data acquisition and processing module: responsible for collecting the original data related to the water environment and performing processing, and the data acquisition and processing module transfers the processed data to the water flow and pollutant diffusion modeling module;

[0052] The water flow and pollutant diffusion modeling module: responsible for constructing a hydrodynamic model based on the processed data using the CFD simulation method, predicting the pollutant diffusion trend through the convection-diffusion equation, and establishing an accumulation model of pollutants in local areas. The water flow and pollutant diffusion modeling module transfers the simulation results to the data analysis and prediction module;

[0053] Data analysis and prediction module: responsible for mining the collected data and simulation results using the K-means clustering algorithm, and then predicting the time series of pollutant diffusion and water quality changes using the LSTM algorithm. The data analysis and prediction module transmits the prediction results to the early warning and emergency response module;

[0054] Early warning and emergency response module: responsible for issuing early warnings for potential water quality pollution events based on the pollution trend prediction results and real-time monitoring data, and formulating emergency response strategies. The early warning and emergency response module transmits the early warning results to the model optimization and feedback module;

[0055] Model optimization and feedback module: responsible for comparing the prediction results with the actual monitoring data and optimizing the model parameters. The model optimization and feedback module feeds back the optimized parameters to the water flow and pollutant diffusion modeling module and the data analysis and prediction module.

[0056] In the embodiment of the present invention, first, the data acquisition and processing module collects the original data related to the water environment, including water flow velocity, water quality parameters, etc. These data are the basis for subsequent analysis and modeling. The data processing process includes data cleaning, verification, and standardization to ensure the accuracy and consistency of the data. This helps to reduce analysis errors caused by data quality problems. The water flow and pollutant diffusion modeling module uses the CFD (Computational Fluid Dynamics) simulation method to construct a hydrodynamic model, which can accurately simulate the impact of water flow velocity on the pollutant diffusion and degradation processes. By solving the convection-diffusion equation, the system can predict the diffusion trend of pollutants and the accumulation situation in local areas under different water flow velocities. This prediction ability helps to more accurately evaluate the water quality status. The data analysis and prediction module uses the K-means clustering algorithm to mine the collected data and simulation results to discover the potential laws and patterns in the data. The LSTM (Long Short-Term Memory) algorithm is used to predict the time series of pollutant diffusion and water quality changes. The LSTM algorithm can capture the long-term dependencies in the time series data, so it can more accurately predict the water quality change trend. The early warning and emergency response module issues timely warnings for potential water quality pollution events based on the pollution trend prediction results and real-time monitoring data. When the system detects the risk of water quality deterioration, it can immediately formulate emergency response strategies, such as closing the pollution source, strengthening water quality monitoring, etc., to mitigate the impact of water quality deterioration.

[0057] In one embodiment, the specific steps for collecting the original data related to the water environment are as follows:

[0058] Determine the water quality parameters to be monitored, and identify and determine the sources for obtaining the monitoring index data;

[0059] Configure the data acquisition equipment according to the monitoring indexes and data sources, install the equipment at the monitoring location, and calibrate the equipment;

[0060] Collect data at set time intervals or trigger conditions.

[0061] In an embodiment of the present invention, water quality monitoring is carried out in the middle reaches of a certain river, and the monitoring parameters include pH value, dissolved oxygen, ammonia nitrogen, and total phosphorus. This section of the river is affected by the discharge of industrial sewage upstream, and special attention needs to be paid to the concentrations of ammonia nitrogen and total phosphorus. Then, a portable pH meter, a DO sensor, an ammonia nitrogen detector, and a total phosphorus monitor are selected. They are installed in the middle reaches of the river, and one sensor is installed at each of the bottom layer and the surface layer for water quality monitoring. pH, DO, ammonia nitrogen, and total phosphorus data are collected once an hour. When the ammonia nitrogen concentration exceeds the set threshold (such as 1.0 mg / L), the system automatically triggers an alarm to notify relevant personnel for subsequent processing.

[0062] In one embodiment, the specific steps for constructing a hydrodynamic model using the CFD simulation method are as follows:

[0063] Problem definition and geometric modeling: According to the monitored water body conditions, define the geometric model of the simulation area, use computational fluid dynamics software to generate the geometric grid of the watershed, and perform grid division on the geometric domain;

[0064] Modeling of the water kinematic equation: Use the Navier-Stokes equation to describe the motion behavior of the fluid Among them, u is the fluid velocity vector, ρ is the fluid density, ν is the fluid viscosity coefficient, p is the fluid pressure, f is the external force term, and boundary conditions are set to simulate the real situation of the water body, including the inlet boundary, the outlet boundary, and the wall boundary.

[0065] In an embodiment of the present invention, the inlet boundary sets the velocity, flow rate, or flow rate distribution of the water flow. For example, assume that the flow velocity of the river channel at the inlet is 1 m / s and the velocity distribution is set to be uniform. The outlet boundary sets the pressure to a constant value (such as atmospheric pressure) or sets the flow rate to a given value. For example, the pressure at the outlet is set to 0 Pa, indicating free water discharge. The wall boundary assumes that the tangential velocity of the fluid is zero, indicating the no-slip condition. In addition, the influence of bottom roughness may also need to be considered.

[0066] In one embodiment, the specific steps for predicting the pollutant diffusion trend through the convection-diffusion equation are as follows:

[0067] Pollutant diffusion is described by the convection-diffusion equation as: Among them, C is the pollutant concentration, D is the pollutant diffusion coefficient, is the diffusion term, and R(C) is the reaction term;

[0068] The convection-diffusion equation is discretized and solved to obtain the change of the pollutant concentration field over time. The discretization formula is: Wherein, is the pollutant concentration at position i at time n, is the pollutant concentration at position j at time n, Δt is the time step, Δx is the space step, and u ij is the flow velocity.

[0069] In one embodiment, the specific process of establishing the accumulation model of pollutants in the local area is as follows:

[0070] The calculation formula for the locally accumulated concentration is: Wherein, C local (x,t) represents the accumulation of pollutant concentration in the local area, t represents the time elapsed during the simulation process, and τ represents the time variable.

[0071] In one embodiment, the specific steps of using the K-means clustering algorithm to mine the collected data and simulation results are as follows:

[0072] Data preprocessing: Preprocess the collected data and simulation results;

[0073] Data integration and feature engineering: Integrate the data from different modules to generate a unified feature matrix;

[0074] Select the value of K: Preset the value of K according to the theoretical knowledge of water pollution diffusion and the actual situation;

[0075] Apply the K-means algorithm for clustering: Randomly select K initial clustering centers, assign each data point to the nearest clustering center, update the centroid of the cluster according to the mean of the data points in each cluster, and repeat the assignment and update steps until the clustering centers no longer change or reach the preset maximum number of iterations;

[0076] Analyze the clustering results: Analyze the representativeness of each cluster according to the data characteristics in each cluster, assign a label to each cluster, classify the clusters according to the pollutant concentration and flow velocity attributes, and visualize the clustering results using a three-dimensional scatter plot.

[0077] In the embodiment of the present invention, the purpose of data integration is to integrate the data from different modules (such as pollution sources, flow velocity fields, temperature changes, etc.) into a unified feature matrix. This process ensures the multi-dimensionality of the data and enables a more comprehensive analysis of the pollution diffusion pattern. Let the feature matrix be:

[0078] Concentration (C) Flow rate (u) Temperature (T) Time (t) Position (x) 0.1 0.5 20 0 0 0.15 0.4 21 10 10 0.2 0.45 19 20 20 0.25 0.6 22 30 30 0.3 0.55 20 40 40

[0079] Selecting the number of clusters K is a crucial step in K-means clustering. According to the theory of water pollution diffusion, the selection of the number of clusters should be based on the changing trend of pollutant concentration and the different states of the water body. Assuming that K = 3 is selected based on domain experience, it means dividing the data into 3 clusters. Then, the K-means algorithm will randomly select 3 initial cluster centers. Suppose the obtained clustering results are as follows:

[0080]

[0081]

[0082] In the clustering results, the cluster labels 1, 2, and 3 correspond to different water body states. Cluster 1: The pollutant concentration is low, the flow rate is medium, and the temperature is moderate, indicating that the pollutant has not spread to this area, or the flow rate is low and the diffusion is slow. Cluster 2: The pollutant concentration is medium, the flow rate is fast, and the temperature is slightly low, indicating that the pollutant is spreading in some areas with a fast flow rate and the concentration is increasing. Cluster 3: The pollutant concentration is high, the flow rate is fast, and the temperature is high, indicating the area near the pollution source where the pollutant concentration increases rapidly.

[0083] In one embodiment, the specific process of using the LSTM algorithm to predict the time series of pollutant diffusion and water quality change is as follows:

[0084] Data preparation: Obtain the preprocessed water quality monitoring data, and divide the collected data into a training set, a validation set, and a test set;

[0085] LSTM model construction: Define the LSTM network structure, including an input layer, an LSTM layer, a fully connected layer, and an output layer, and set the number of units in the LSTM layer, the learning rate, the optimizer, and the loss function parameters;

[0086] Model training: Input the data of the training set into the LSTM model. Through the memory units and gating mechanisms of the LSTM layer, calculate the output of the model. Use the loss function to calculate the error between the model output and the actual target value. According to the gradient of the loss function, update the parameters of the LSTM model through the backpropagation algorithm, and use the optimizer to adjust the learning rate;

[0087] Model prediction: Input the test set into the trained LSTM model, calculate the output of the model through forward propagation, use the test set data to evaluate the prediction performance of the model, and optimize the model according to the evaluation results.

[0088] In the embodiment of the present invention, for pollutant concentration prediction, the loss function uses the mean squared error (MSE): where y i is the true value of the i-th sample, is the value predicted by the model, and N is the number of samples. The backpropagation algorithm calculates the gradient of each weight through the chain rule, and the formula is: where is the loss function, T is the number of time steps, is the gradient of the output gate with respect to the weights, is the loss function is the partial derivative of the weight matrix W with respect to the loss function. W is the weight matrix in the LSTM model, representing the connection weights between the input layer, hidden layer, and output layer. The trained LSTM model calculates the output through forward propagation during prediction. The input test data is x t , and the corresponding predicted value is obtained. The formula is: where f LSTM represents the forward calculation process of the LSTM network.

[0089] In one embodiment, the early warning and emergency response module includes a pollution trend assessment unit, an early warning determination unit, an emergency response decision-making unit, and an information release and notification unit. The pollution trend assessment unit is responsible for evaluating the current situation and future change trends of water quality. The early warning determination unit is responsible for judging whether to trigger an early warning based on the output of the pollution trend assessment unit and the early warning threshold. The emergency response decision-making unit is responsible for formulating emergency response strategies. The information release and notification unit is responsible for transmitting the early warning information and emergency response measures to relevant personnel.

[0090] As Figure 2 shown, a water environment monitoring method based on big data analysis includes:

[0091] S1: Collect the original data related to the water environment and perform preprocessing;

[0092] S2: Use the CFD simulation method to construct a hydrodynamic model for simulating the water flow movement state based on the processed data. By solving the convection-diffusion equation, predict the pollutant diffusion trend and distribution, and establish a cumulative model of pollutants in a local area to evaluate the pollutant accumulation effect;

[0093] S3: Use the K-means clustering algorithm to mine the collected data and simulation results, and use the LSTM algorithm to predict the time series of pollutant diffusion and water quality changes;

[0094] S4: Based on the pollution trend prediction results and real-time monitoring data, issue early warnings for water quality pollution events, and formulate emergency response strategies according to the early warning results;

[0095] S5: Compare the prediction results with the actual data to optimize the model parameters.

[0096] Although embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A water environment monitoring system based on big data analysis, characterized in that, Including: Data acquisition and processing module: responsible for collecting and processing the original data related to the water environment, and the data acquisition and processing module transmits the processed data to the water flow and pollutant diffusion modeling module; Water flow and pollutant diffusion modeling module: responsible for constructing a hydrodynamic model based on the processed data using the CFD simulation method, predicting the pollutant diffusion trend through the convection-diffusion equation, and establishing a cumulative model of pollutants in the local area. The water flow and pollutant diffusion modeling module transmits the simulation results to the data analysis and prediction module; Data analysis and prediction module: responsible for mining the collected data and simulation results using the K-means clustering algorithm, and then predicting the time series of pollutant diffusion and water quality changes using the LSTM algorithm. The data analysis and prediction module transmits the prediction results to the early warning and emergency response module; Early warning and emergency response module: responsible for issuing early warnings for potential water quality pollution incidents based on the pollution trend prediction results and real-time monitoring data, and formulating emergency response strategies. The early warning and emergency response module transmits the early warning results to the model optimization and feedback module; Model optimization and feedback module: responsible for comparing the prediction results with the actual monitoring data and optimizing the model parameters. The model optimization and feedback module feeds back the optimized parameters to the water flow and pollutant diffusion modeling module and the data analysis and prediction module.

2. The water environment monitoring system based on big data analysis according to claim 1, characterized in that, The specific steps for collecting the original data related to the water environment are as follows: Determine the water quality parameters to be monitored, identify and determine the sources for obtaining the monitoring index data; Configure the data acquisition equipment according to the monitoring indexes and data sources, install the equipment at the monitoring location, and calibrate the equipment; Collect data at set time intervals or trigger conditions.

3. The water environment monitoring system based on big data analysis according to claim 1, characterized in that, The specific steps for constructing a hydrodynamic model using the CFD simulation method are as follows: Problem definition and geometric modeling: According to the monitored water body conditions, define the geometric model of the simulation area, generate the geometric grid of the basin using computational fluid dynamics software, and perform grid division on the geometric domain; Modeling of Hydrodynamics Equations: Using the Navier-Stokes equations to describe the motion behavior of fluids Among them, u is the fluid velocity vector, ρ is the fluid density, ν is the fluid viscosity coefficient, p is the fluid pressure, f is the external force term, and boundary conditions are set to simulate the real situation of water bodies, including inlet boundaries, outlet boundaries, and wall boundaries.

4. The water environment monitoring system based on big data analysis according to claim 1, characterized in that, The specific steps for predicting the pollutant diffusion trend through the convection-diffusion equation are as follows: The pollutant diffusion is described by the convection-diffusion equation, which is expressed as: where C is the pollutant concentration, D is the pollutant diffusion coefficient, is the diffusion term, and R(C) is the reaction term; The convective-diffusion equation is discretized and solved to obtain the variation of the pollutant concentration field with time. The discretization formula is as follows: where is the pollutant concentration at position i at time n, is the pollutant concentration at position j at time n, Δt is the time step, Δx is the space step, and u ij is the flow velocity.

5. A water environment monitoring system based on big data analysis according to claim 1, characterized in that, The specific process for establishing a cumulative model of pollutants in the local area is as follows: The calculation formula for the locally accumulated concentration is as follows: where C local (x, t) represents the accumulation of the pollutant concentration in the local area, t represents the time elapsed during the simulation process, and τ represents the time variable.

6. The water environment monitoring system based on big data analysis according to claim 1, characterized in that The specific steps for mining the collected data and simulation results using the K-means clustering algorithm are as follows: Data preprocessing: Preprocess the collected data and simulation results; Data integration and feature engineering: Integrate the data from different modules to generate a unified feature matrix; Select the value of K: Preset the value of K according to the theoretical knowledge of water pollution diffusion and the actual situation; Apply the K-means algorithm for clustering: Randomly select K initial clustering centers, assign each data point to the nearest clustering center, update the centroid of the cluster according to the mean value of the data points in each cluster, and repeat the assignment and update steps until the clustering centers no longer change or reach the preset maximum number of iterations; Analyze the clustering results: According to the data characteristics in each cluster, analyze the representativeness of the cluster, assign a label to each cluster, classify the clusters according to the pollutant concentration and flow velocity attributes, and visualize the clustering results using a three-dimensional scatter plot.

7. A water environment monitoring system based on big data analysis according to claim 1, characterized in that The specific process of predicting the time series of pollutant diffusion and water quality change using the LSTM algorithm is as follows: Data preparation: Obtain preprocessed water quality monitoring data, and divide the collected data into a training set, a validation set, and a test set; LSTM model construction: Define the LSTM network structure, including an input layer, an LSTM layer, a fully connected layer, and an output layer, and set the number of units in the LSTM layer, the learning rate, the optimizer, and the loss function parameters; Model training: Input the data of the training set into the LSTM model, calculate the output of the model through the memory units and gating mechanisms of the LSTM layer, calculate the error between the model output and the actual target value using the loss function, update the parameters of the LSTM model through the backpropagation algorithm according to the gradient of the loss function, and adjust the learning rate using the optimizer; Model prediction: Input the test set into the trained LSTM model, calculate the output of the model through forward propagation, evaluate the model prediction performance using the test set data, and optimize the model according to the evaluation results.

8. A water environment monitoring system based on big data analysis according to claim 1, characterized in that The early warning and emergency response module includes a pollution trend assessment unit, an early warning determination unit, an emergency response decision-making unit, and an information release and notification unit. The pollution trend assessment unit is responsible for assessing the current situation and future change trends of water quality. The early warning determination unit is responsible for judging whether to trigger an early warning based on the output of the pollution trend assessment unit and the early warning threshold. The emergency response decision-making unit is responsible for formulating emergency response strategies. The information release and notification unit is responsible for transmitting the early warning information and emergency response measures to relevant personnel.

9. A water environment monitoring method based on big data analysis corresponding to the water environment monitoring system based on big data analysis according to claims 1-8, characterized in that, It includes: S1: Collect the original data related to the water environment and perform preprocessing; S2: Use the CFD simulation method to construct a hydrodynamic model for simulating the water flow movement state based on the processed data, predict the pollutant diffusion trend and distribution by solving the convection-diffusion equation, and establish an accumulation model of pollutants in a local area to evaluate the pollutant accumulation effect; S3: Use the K-means clustering algorithm to mine the collected data and simulation results, and use the LSTM algorithm to predict the time series of pollutant diffusion and water quality change; S4: Warn of water quality pollution events based on the pollution trend prediction results and real-time monitoring data, and formulate emergency response strategies according to the warning results; S5: Compare the prediction results with the actual data to optimize the model parameters.

Images