A pollution source detection method and system for water environment management

Abstract

The present invention belongs to the technical field of pollution source detection. The present invention discloses a pollution source detection method and system for water environment management; comprising: collecting acoustic disturbance data through underwater acoustic sensors, and performing real-time edge preprocessing, synchronously collecting water environment data at different depths, calculating water environment gradient data, and then realizing data synchronization through a unified timestamp. Fourier transform is used to extract the frequency domain characteristics of the acoustic disturbance signal, calculate the main frequency, energy difference and spectrum skewness, and generate disturbance sound wave indicators accordingly. Combining the disturbance sound wave indicators with the water environment gradient data, a sliding window is used to construct a time series sample set, and a recurrent neural network model for predicting future pollution release risks is trained. Differentiated pollution release thresholds are set according to different water body types. When the model output exceeds the threshold, a corresponding level of sediment pollution warning instruction is automatically generated to achieve efficient and graded intelligent identification and response of pollution events.

Description

Technical Field

[0001] The present invention relates to the technical field of pollution source detection, and more specifically, to a pollution source detection method and system for water environment management. Background Art

[0002] With the acceleration of industrialization and urbanization, water pollution is becoming increasingly serious. Although various water environment improvement projects have achieved some success in surface water purification in recent years, sediments, as a significant "sink" of pollutants, can still release accumulated heavy metals, organic pollutants, and nutrients under certain conditions, becoming a key source of a "secondary rebound" of water pollution.

[0003] In actual applications, factors such as ship navigation, hydraulic construction, and water flow disturbance may trigger the transfer and release of pollutants in the sediment to the water body, resulting in sudden changes in water quality and even ecological problems such as "returning to black and smelly".

[0004] The Chinese patent with authorization announcement number CN109060414B discloses a river sediment sampling device, which designs a sampling method that has minimal disturbance to the sediment, maintains the integrity of the sediment sample, and has good pressure and heat preservation performance, facilitating subsequent experimental testing.

[0005] Chinese patent application number CN115128229B discloses a big data-based water pollution source detection and management system. By evenly distributing water quality monitoring points throughout a river, a first detection module regularly monitors water quality, while a second detection module locates the monitoring points and determines their standard water quality. By comparing actual water quality with the standard, the system dynamically adjusts the detection cycle, thereby improving detection efficiency.

[0006] However, current research on sediment disturbance release processes has largely focused on laboratory conditions or offline testing, lacking real-time, automated monitoring methods suitable for actual aquatic environments. Furthermore, due to the nonlinear, sudden, and complex coupling characteristics of pollution release processes, existing methods struggle to accurately predict their development trends, failing to provide timely and effective data support for water quality risk warnings and adjustments to governance strategies.

[0007] Therefore, there is an urgent need for an intelligent detection method and system that can perceive sediment disturbance and its pollution release behavior in real time and has dynamic prediction capabilities to improve the accuracy and response efficiency of water environment monitoring.

[0008] In view of this, the present invention proposes a pollution source detection method and system for water environment management to solve the above problems. Summary of the Invention

[0009] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a pollution source detection method and system for water environment treatment.

[0010] To achieve the above object, the present invention provides the following technical solutions:

[0011] A method for detecting pollution sources for water environment management, the method comprising:

[0012] Collect acoustic disturbance data and perform edge preprocessing in real time; synchronously collect water environment data while collecting acoustic disturbance data, and the water environment data includes four groups of indicators at the same time; each group of indicators includes water temperature, pH value, dissolved oxygen concentration and flow rate;

[0013] Calculate water environment gradient data based on water environment data; suppose water environment data is ; is the water environment data group number, =1, 2, 3, 4 are the first, second, third and fourth groups of water environment data respectively; is the water environment data category identifier, =1, 2, 3, 4 represent water temperature, pH value, dissolved oxygen concentration and flow rate respectively; the water environment gradient data is: Where, =1;

[0014] Introduce water environment gradient data and acoustic disturbance data into the time synchronization mechanism and give them a unified time stamp;

[0015] Obtain frequency domain information using Fourier transform based on acoustic disturbance data; obtain main frequency, energy difference and spectrum skewness based on spectrum information; calculate disturbance sound wave index based on spectrum skewness, main frequency and energy difference;

[0016] A data set was established using a sliding window method based on disturbance acoustic wave indicators and water environment gradient data. A machine learning model was trained based on the data set to predict whether sediment disturbance would exceed pollution standards at the next k moments.

[0017] Set different levels of pollution release thresholds for different water body types; compare the output of the machine learning model with the preset pollution release thresholds of different levels. When the output value of the machine learning model is greater than the pollution release threshold, generate different levels of sediment pollution warning instructions.

[0018] Preferably, the acoustic disturbance data refers to the echo spectrum signal generated by the sediment particles after being disturbed; the acquisition frequency of the acoustic disturbance data is 10 times per second, and the edge preprocessing includes: signal conversion and invalid signal elimination;

[0019] The first group of water environment data consists of four indicators at the bottom sediment (water temperature, pH value, dissolved oxygen concentration and flow rate); the second group of water environment data consists of four indicators 5 cm vertically above the bottom sediment; the third group of water environment data consists of four indicators 15 cm vertically above the bottom sediment; the fourth group of water environment data consists of four indicators 50 cm vertically above the bottom sediment; there are four groups of water environment gradient data; each group is a three-dimensional vector; the timestamp information includes the collection date, hour, minute, second and quarter identifier.

[0020] Preferably, the Fourier transform is defined as:

[0021] Where: is the Fourier transform frequency distribution; is the time domain signal; is the frequency after Fourier transform; is the starting time of Fourier transform; For time; is an imaginary unit; is the time length of the window;

[0022] The dominant frequency is the frequency with the largest amplitude in the spectrum; the energy difference refers to the difference in frequency domain energy between the two time windows before and after the disturbance; the spectrum skewness indicates the symmetry of the spectrum distribution;

[0023] Spectral skewness ;

[0024] Where: For the frequencies; for The corresponding amplitude on is the mean frequency of the spectrum; is the spectrum standard deviation; is the total number of frequency points.

[0025] Preferably, disturbance sound wave indicator Where, is the main frequency; is the energy difference; =10 -6 ; =1.2;

[0026] The process of establishing a data set includes: using the disturbance acoustic wave index and water environment gradient data with the same timestamp as feature data, and the label corresponding to the feature data is the pollution release label at time k after the timestamp; when the pollutant concentration is detected at any position 5cm, 15cm, or 50cm above the bottom mud, the pollution release label is set to 1; otherwise, it is set to 0;

[0027] A set of feature data and labels is constructed as a sample, and multiple groups of samples are collected to establish a data set; the data set is divided into a training set, a validation set, and a test set, where the training set accounts for 60% of the data set, and the validation set and the test set each account for 20% of the data set.

[0028] Preferably, the training process of the machine learning model includes:

[0029] Using the training set as input to a machine learning model, the model outputs a predicted value of a pollution release label at k moments in the future, uses the characteristic data of real-time disturbance acoustic wave indicators and water environment gradient data as prediction targets, and minimizes the machine learning model loss function value as a training objective; training stops when the machine learning model loss function value is less than or equal to a preset target loss value;

[0030] The machine learning model loss function is the mean square error; the mean square error is calculated by Minimize to train the model; in the loss function is the loss function value, i is the sample group number; is the number of sample groups; is the label of the i-th group of samples, The pollution release label predicted for the i-th group of samples;

[0031] The machine learning model is a recurrent neural network model, which can be a long short-term memory network or a gated recurrent unit.

[0032] Preferably, the urban river threshold sets the third-level pollution release threshold, the artificial water bodies such as storage ponds set the second-level pollution release threshold; the eutrophic lakes set the first-level pollution release threshold; the third-level pollution release threshold > the second-level pollution release threshold > the first-level pollution release threshold.

[0033] Preferably, the first-level sediment pollution warning instruction includes the system immediately sending a high-level warning, notifying the environmental emergency team to respond quickly, and launching the emergency pollution control plan, including quickly deploying pollution interception measures, carrying out pollution source blocking and rapid water quality restoration, etc., to prevent the spread of pollution;

[0034] Level 2 sediment pollution warning instructions include: the system automatically generates a medium-level warning, notifying on-site management personnel to conduct focused inspections and intensive water quality monitoring in relevant waters, and to deploy pollution control resources in advance if necessary for further response;

[0035] The third-level sediment pollution warning instructions include: the system issues a low-level reminder, recommending that staff conduct regular inspections of the water area, make preparations for pollution prevention and control, pay attention to changes in pollution trends, and record relevant data for analysis and evaluation.

[0036] A pollution source detection system for water environment management, the system comprising:

[0037] The data collection module collects acoustic disturbance data and performs edge preprocessing in real time. It also collects water environment data simultaneously with the acoustic disturbance data. The water environment data includes four groups of indicators at the same time. Each group of indicators includes water temperature, pH value, dissolved oxygen concentration, and flow rate.

[0038] The data processing module calculates the water environment gradient data based on the water environment data; assuming that the water environment data is ; is the water environment data group number, =1, 2, 3, 4 are the first, second, third and fourth groups of water environment data respectively; is the water environment data category identifier, =1, 2, 3, 4 represent water temperature, pH value, dissolved oxygen concentration and flow rate respectively; the water environment gradient data is: Where, =1;

[0039] The data synchronization module introduces the water environment gradient data and acoustic disturbance data into the time synchronization mechanism and adds a unified time stamp;

[0040] The data enhancement module uses Fourier transform to obtain frequency domain information based on the acoustic disturbance data; obtains the main frequency, energy difference and spectrum skewness based on the spectrum information; and calculates the disturbance sound wave index based on the spectrum skewness, main frequency and energy difference;

[0041] The model training module uses a sliding window method to establish a data set based on disturbance acoustic wave indicators and water environment gradient data. Based on the data set, a machine learning model is trained to predict whether the sediment disturbance will exceed the pollution standard at the next k moments.

[0042] The post-processing module sets different levels of pollution release thresholds for different water body types; based on the comparison of the output of the machine learning model with the preset pollution release thresholds of different levels, when the output value of the machine learning model is greater than the pollution release threshold, different levels of sediment pollution warning instructions are generated.

[0043] An electronic device comprises: a processor and a memory, wherein the memory stores a computer program that can be called by the processor; the processor executes the above-mentioned pollution source detection method for water environment treatment by calling the computer program stored in the memory.

[0044] A computer-readable storage medium stores instructions. When the instructions are executed on a computer, the computer is caused to execute the above-mentioned pollution source detection method for water environment treatment.

[0045] The technical effects and advantages of the pollution source detection method and system for water environment management presented in this invention are as follows: This application utilizes underwater acoustic sensors to collect acoustic disturbance data, extracts the main frequency, energy difference, and spectral skewness through Fourier transform, and calculates disturbance acoustic wave indicators. Simultaneously, water environment data (water temperature, pH, dissolved oxygen, and flow rate) at different heights above the sediment are collected to construct a water environment gradient signature. This signature is aligned with the disturbance acoustic wave indicators by timestamp to form a sliding window sample. Finally, a recurrent neural network model is trained to predict the risk of sediment pollution release in the future. The model output is compared with differentiated pollution thresholds to trigger graded warning instructions.

[0046] This application integrates multi-source data fusion, time series modeling, and a hierarchical response mechanism to achieve highly sensitive, real-time, and predictable identification of pollution releases caused by sediment disturbance. Gradient water data and acoustic wave characteristics work together to enhance the model's comprehensive judgment of disturbance paths and pollution intensity. Differentiated threshold settings improve the system's adaptability and practicality across different water types, reducing false alarms and missed alerts. The hierarchical warning instruction design provides practical operational guidance, enabling refined, intelligent, and efficient water environment management. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 Schematic diagram of a pollution source detection method for water environment management according to the present invention;

[0048] Figure 2 Schematic diagram of a pollution source detection system for water environment management according to the present invention;

[0049] Figure 3 This is a schematic diagram of the structure of an electronic device provided by an embodiment of the present application;

[0050] Figure 4 This is a schematic diagram of the computer-readable storage medium structure provided by an embodiment of the present application. DETAILED DESCRIPTION

[0051] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 making creative efforts are within the scope of protection of the present invention. Example 1

[0052] See also Figure 1 As shown, the pollution source detection method and system for water environment management described in this embodiment include:

[0053] Acoustic disturbance data is collected using highly sensitive underwater acoustic sensors deployed at the bottom of the water. These acoustic disturbance data refer to the echo spectrum signals generated by disturbed sediment particles. The disturbance is artificially simulated, such as from a boat rowing or other water disturbances.

[0054] Acoustic disturbance data is collected 10 times per second and uploaded to the edge computing device in real time via the RS485 communication protocol. The edge computing device performs real-time edge preprocessing, which includes signal conversion and invalid signal rejection. Signal conversion converts the collected acoustic disturbance data into a high-precision digital waveform on a unified timeline. Invalid signal rejection identifies and removes useless and interfering information from the received signal. The methods for invalid signal rejection are well-known in the art and will not be detailed here.

[0055] Water environment data is collected simultaneously with the acoustic disturbance data, including four indicators: water temperature, pH value, dissolved oxygen concentration, and flow rate. It should be noted that the water environment data includes four groups of indicators at the same time; each group of indicators includes four categories: water temperature, pH value, dissolved oxygen concentration, and flow rate;

[0056] The first group of water environment data is four types of indicators at the bottom mud (water temperature, pH value, dissolved oxygen concentration and flow rate); the second group of water environment data is four types of indicators 5 cm vertically above the bottom mud; the third group of water environment data is four types of indicators 15 cm vertically above the bottom mud; the fourth group of water environment data is four types of indicators 50 cm vertically above the bottom mud.

[0057] This setting is because when the bottom sediment is disturbed, only the composition of the bottom sediment is often considered, which cannot timely reflect the diffusion of the bottom sediment and the actual nonlinear pollution behavior. By setting the water environment data of the gradient height, the changes in the close range after the bottom sediment disturbance can be reflected first, and as the diffusion range increases, the set gradient gradually increases, effectively preventing the collection of invalid data or unclear data, and facilitating the accuracy of subsequent analysis.

[0058] Water temperature was measured using a high-precision digital thermometer in degrees Celsius (°C); pH was measured using a water quality ion electrode; dissolved oxygen concentration was measured using a fluorescence sensor with a range of 0–20 mg / L; and flow velocity was measured using a Doppler flowmeter.

[0059] Environmental data, reflecting changes in hydrodynamic and biochemical conditions, are key drivers of sediment release. Higher flow rates increase the likelihood of sediment disturbance and resuspension. Water temperature changes influence microbial activity, indirectly regulating pollutant release rates. pH and dissolved oxygen levels determine the release forms and migration pathways of heavy metals and nutrients.

[0060] Calculate water environment gradient data based on water environment data; suppose water environment data is ; is the water environment data group number, =1, 2, 3, 4 are the first, second, third and fourth groups of water environment data respectively; is the water environment data category identifier, =1, 2, 3, 4 represent water temperature, pH value, dissolved oxygen concentration and flow rate respectively; the water environment gradient data is:

[0061] Where, =1;

[0062] The water environment gradient data also consists of four groups; each group is a three-dimensional vector; the collection of water environment gradient data is to improve the sensitivity to abnormal pollution release events; traditional sediment collection and analysis only exist in the laboratory, which is far from enough if the timely response capability of the environmental protection department needs to be improved. By obtaining water environment gradient data and amplifying the gradient impact of sediment disturbance, it is conducive to timely and real-time detection and early warning of water pollution. The vector method can better reflect the gradient trend than single data.

[0063] Water environment gradient data and acoustic disturbance data are integrated into a time synchronization mechanism, which assigns unified timestamps to both data, accurate to the second. The timestamps include the acquisition date, hour, minute, second, and quarterly identifier, which is used to construct the time series input feature matrix. The quarterly information is automatically determined by the system's built-in calendar module. The inclusion of temporal features enhances the model's ability to identify periodic disturbance events and seasonal pollution release trends.

[0064] Calculating the disturbance sound wave index based on the acoustic disturbance data; the calculation process includes: obtaining frequency domain information using Fourier transform based on the acoustic disturbance data; Fourier transform is defined as:

[0065] Where: is the Fourier transform frequency distribution; is the time domain signal; is the frequency after Fourier transform; is the starting time of Fourier transform; For time; is an imaginary unit; is the time length of the window.

[0066] Based on the spectrum information, the main frequency, energy difference and spectrum skewness are obtained. The main frequency is the frequency with the largest amplitude in the spectrum, indicating that the strongest energy in the current disturbance signal is concentrated on a specific frequency, which helps to identify the key frequency of the pollution disturbance; the energy difference refers to the difference in frequency domain energy between the two time windows before and after the disturbance; the spectrum skewness indicates the symmetry of the spectrum distribution; if the skewness is large, it means that the spectrum is asymmetric and may be interfered with or abnormal; if the skewness is small, it means that the energy distribution of this disturbance is uniform or concentrated, and the structure is more stable and reliable. Spectrum skewness ;

[0067] Where: For the frequencies; for The corresponding amplitude on is the mean frequency of the spectrum; is the spectrum standard deviation; is the total number of frequency points;

[0068] Calculate the disturbance sound wave index based on spectrum skewness, main frequency, and energy difference:

[0069] Where, is the main frequency; is the energy difference; =10 -6 ; =1.2;

[0070] The higher the main frequency, the more subtle and intense the disturbance may be. The greater the energy difference, the greater the natural disturbance and the greater the disturbance sound wave index. The greater the spectrum skewness, the more asymmetric and dispersed the spectrum, and the lower the corresponding signal quality. Therefore, the greater the spectrum skewness, the smaller the disturbance sound wave index. The disturbance sound wave index is an important indicator for judging whether the disturbance is significant and stable, and avoiding irrelevant disturbances. The more concentrated the disturbance, the more intense the energy change. The smaller the skewness, the more significant the disturbance is. At this time, the sediment disturbance has a clear directionality, which is convenient for identifying disturbances.

[0071] A data set was established using a sliding window method based on disturbance acoustic wave indicators and water environment gradient data. The disturbance acoustic wave indicators and water environment gradient data with the same timestamp were used as feature data. The label corresponding to the feature data was the pollution release label at time k after the timestamp. When the pollutant concentration exceeded the standard at any position 5 cm, 15 cm, or 50 cm above the bottom mud, the pollution release label was set to 1; otherwise, it was set to 0.

[0072] The pollutant concentration is obtained in real time by online water quality monitoring instruments, including but not limited to online ammonia nitrogen analyzers and online heavy metal monitors. The standards for pollutant concentration exceeding the standard are formulated by relevant departments.

[0073] A set of feature data and labels is constructed as a sample, and multiple sets of samples are collected to establish a dataset; the dataset is divided into a training set, a validation set, and a test set, where the training set accounts for 60% of the dataset, and the validation set and the test set each account for 20% of the dataset;

[0074] The present invention provides the following example of the sliding window method: Assume that there are 5 minutes of characteristic data of disturbance sound wave index and water environment gradient data, and mark the 5 minutes of characteristic data as a characteristic data set , { , , , , }, ={ }, Indicates the characteristic data of the first minute, and so on; Release the label for the pollution in the 1+k minute; and Construct a set of samples; the next set of samples is { , , , , }and{ }; Use sliding windows to build multiple samples and labels;

[0075] It should be noted that constructing a data set in this way can effectively extract the potential impact of disturbance signals on pollution release at different time stages, and the multi-layer water environment gradient data can characterize the path after disturbance; introducing acoustic disturbance indicators as input variables can improve the model's sensitivity and discrimination for instantaneous disturbance identification.

[0076] A machine learning model is trained based on the dataset to predict whether the sediment disturbance will exceed the pollution standard at the next k moments.

[0077] Using the training set as input to a machine learning model, the model outputs a predicted value of a pollution release label at k moments in the future, uses the characteristic data of real-time disturbance acoustic wave indicators and water environment gradient data as prediction targets, and minimizes the machine learning model loss function value as a training objective; training stops when the machine learning model loss function value is less than or equal to a preset target loss value;

[0078] The machine learning model loss function is the mean square error; the mean square error is calculated by Minimize to train the model; in the loss function is the loss function value, i is the sample group number; is the number of sample groups; is the label of the i-th group of samples, The pollution release label predicted for the i-th group of samples;

[0079] Preferably, the machine learning model is a recurrent neural network model, which can be a long short-term memory network (LSTM) or a gated recurrent unit (GRU).

[0080] Other model parameters of the machine learning model, such as the depth of the network model, the number of neurons in each layer, the activation function used by the network model, the convergence condition, the ratio of the training set, the test set, and the validation set, and the loss function, are all obtained through actual engineering implementation and continuous experimental tuning.

[0081] Different levels of pollution release thresholds are set for different water body types. These thresholds are set by qualified water environment management personnel based on factors such as the functional positioning, pollution carrying capacity and management standards of the specific water body;

[0082] Based on the comparison between the output of the machine learning model and the preset pollution release thresholds of different levels, when the output value of the machine learning model is greater than the pollution release threshold, different levels of sediment pollution warning instructions are generated;

[0083] Since different water bodies have significantly different sensitivities to pollutant concentrations, targeted threshold designs are needed. Taking urban rivers as an example, they mostly undertake drainage, landscaping and certain ecological functions, and have a certain tolerance for pollution. In order to avoid frequent false alarms that affect urban management efficiency, the pollution release threshold can be appropriately raised as the third-level pollution release threshold to ensure that the early warning system is more stable and practical. For artificial water bodies such as storage ponds, their original design intention is to receive rainwater and sewage and reduce non-point source pollution. A certain degree of pollution accumulation is allowed during operation, but strict requirements are placed on outflow control. Therefore, a second-level pollution release threshold is set; eutrophic lakes are extremely sensitive to pollution inputs and are easily endogenous polluted due to sediment disturbance, leading to ecological problems such as algal blooms or black and odorous water. Such water bodies should have lower threshold standards to identify potential risks caused by minor disturbances in advance and set a first-level pollution release threshold. The third-level pollution release threshold > the second-level pollution release threshold > the first-level pollution release threshold;

[0084] The advantage of differentiated threshold design is that it not only enhances the practical guiding significance of the model prediction results and makes the output results closer to the response needs of actual water body management, but also avoids the false alarm or missed alarm problems caused by unified thresholds, thereby improving the stability and reliability of the early warning system.

[0085] The first-level sediment pollution warning instructions include the system immediately sending a high-level warning, notifying the environmental emergency team to respond quickly, and launching an emergency pollution control plan, including quickly deploying pollution interception measures, sealing pollution sources, and quickly repairing water quality to prevent the spread of pollution.

[0086] The second-level sediment pollution warning instructions include: the system automatically generates a medium-level warning, notifies on-site management personnel to conduct key inspections and intensive water quality monitoring of relevant waters, and deploys pollution control resources in advance when necessary to prepare for further response.

[0087] The third-level sediment pollution warning instructions include: the system issues a low-level reminder, recommending that staff conduct regular inspections of the water area, make preparations for pollution prevention and control, pay attention to changes in pollution trends, and record relevant data for analysis and evaluation.

[0088] This application uses underwater acoustic sensors to collect acoustic disturbance data. Fourier transforms are used to extract the dominant frequency, energy difference, and spectral skewness, and calculate disturbance acoustic wave indicators. Simultaneously, water environment data (water temperature, pH, dissolved oxygen, and flow rate) at different heights above the sediment are collected to construct a water environment gradient signature. This is then aligned with the disturbance acoustic wave indicators by timestamp to form a sliding window sample. Finally, a recurrent neural network model is trained to predict the risk of sediment pollution release at future times. The model output is then compared with differentiated pollution thresholds to trigger graded warning instructions.

[0089] This application integrates multi-source data fusion, time series modeling, and a hierarchical response mechanism to achieve highly sensitive, real-time, and predictable identification of pollution releases caused by sediment disturbance. Gradient water data and acoustic wave characteristics work together to enhance the model's comprehensive judgment of disturbance paths and pollution intensity. Differentiated threshold settings improve the system's adaptability and practicality across different water types, reducing false alarms and missed alerts. The hierarchical warning instruction design provides practical operational guidance, enabling refined, intelligent, and efficient water environment management. Example 2

[0090] See also Figure 2 As shown, for the parts not described in detail in this embodiment, please refer to the description of embodiment 1. A pollution source detection system for water environment treatment is provided, the system comprising:

[0091] The data collection module collects acoustic disturbance data and performs edge preprocessing in real time. It also collects water environment data simultaneously with the acoustic disturbance data. The water environment data includes four groups of indicators at the same time. Each group of indicators includes water temperature, pH value, dissolved oxygen concentration, and flow rate.

[0092] The data processing module calculates the water environment gradient data based on the water environment data; assuming that the water environment data is ; is the water environment data group number, =1, 2, 3, 4 are the first, second, third and fourth groups of water environment data respectively; is the water environment data category identifier, =1, 2, 3, 4 represent water temperature, pH value, dissolved oxygen concentration and flow rate respectively; the water environment gradient data is: Where, =1;

[0093] The data synchronization module introduces the water environment gradient data and acoustic disturbance data into the time synchronization mechanism and adds a unified time stamp;

[0094] The data enhancement module uses Fourier transform to obtain frequency domain information based on the acoustic disturbance data; obtains the main frequency, energy difference and spectrum skewness based on the spectrum information; and calculates the disturbance sound wave index based on the spectrum skewness, main frequency and energy difference;

[0095] The model training module uses a sliding window method to establish a data set based on disturbance acoustic wave indicators and water environment gradient data. Based on the data set, a machine learning model is trained to predict whether the sediment disturbance will exceed the pollution standard at the next k moments.

[0096] The post-processing module sets different levels of pollution release thresholds for different water body types; based on the comparison of the output of the machine learning model with the preset pollution release thresholds of different levels, when the output value of the machine learning model is greater than the pollution release threshold, different levels of sediment pollution warning instructions are generated.

[0097] The modules are connected via wired and / or wireless networks. Example 3

[0098] See also Figure 3 According to another aspect of the present application, an electronic device 500 is provided. The electronic device 500 may include one or more processors and one or more memories. The memories may store computer-readable code that, when executed by the one or more processors, may execute the above-described method and system for detecting pollution sources for water environment management.

[0099] The method or system according to the embodiment of the present application can also be used by Figure 3 The electronic device architecture shown in FIG. Figure 3 As shown, the electronic device 500 may include a bus 501, one or more CPUs 502, a read-only memory (ROM) 503, a random access memory (RAM) 504, a communication port 505 connected to a network, an input / output component 506, a hard disk 507, etc. The storage device in the electronic device 500, such as the ROM 503 or the hard disk 507, may store a pollution source detection method for water environment treatment provided by the present application. Furthermore, the electronic device 500 may also include a user interface 508. Of course, Figure 3 The architecture shown is only exemplary and can be omitted according to actual needs when implementing different devices. Figure 3One or more components of an electronic device are shown. Example 4

[0100] See also Figure 4 As shown, a computer-readable storage medium 600 according to an embodiment of the present application is shown. Computer-readable instructions are stored on the computer-readable storage medium 600. When the computer-readable instructions are executed by the processor, a pollution source detection method for water environment treatment according to an embodiment of the present application described with reference to the above figures can be executed. The storage medium 600 includes, but is not limited to, for example, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and cache memory. Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc.

[0101] In addition, according to the embodiments of the present application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, the present application provides a non-transitory machine-readable storage medium that stores machine-readable instructions that can be executed by a processor to execute instructions corresponding to the method steps provided in the present application. When the computer program is executed by a central processing unit (CPU), the above-mentioned functions defined in the method of the present application are performed.

[0102] The above embodiments can be implemented in whole or in part via software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product comprises one or more computer instructions or computer programs. When loaded or executed on a computer, the processes or functions described in accordance with the embodiments of the present invention are fully or partially performed. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transferred from one computer-readable storage medium to another. For example, the computer instructions can be transferred from one website, computer, server, or data center to another website, computer, server, or data center via a wired or wireless network. The computer-readable storage medium can be any available medium accessible by a computer, or a data storage device such as a server or data center that contains a collection of one or more available media. The available medium can be magnetic media (e.g., floppy disks, hard disks, or magnetic tapes), optical media (e.g., DVDs), or semiconductor media. The semiconductor media can be a solid-state drive.

[0103] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in the present invention can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present invention.

[0104] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0105] In the several embodiments provided by the present invention, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a method and system for detecting pollution sources for water environment treatment. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0106] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0107] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0108] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included in the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

[0109] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A pollution source detection method for water environment management, characterized in that: include: Collect acoustic disturbance data and perform edge pre-processing in real time; Water environment data is collected simultaneously with the acoustic disturbance data. The water environment data includes four groups of indicators at the same time; each group of indicators includes four categories: water temperature, pH value, dissolved oxygen concentration and flow rate; Calculate water environment gradient data based on water environment data; suppose water environment data is ; is the water environment data group number, =1, 2, 3, 4 are the first, second, third and fourth groups of water environment data respectively; is the water environment data category identifier, =1, 2, 3, 4 represent water temperature, pH value, dissolved oxygen concentration and flow rate respectively; the water environment gradient data is: Where, =1; Introduce water environment gradient data and acoustic disturbance data into the time synchronization mechanism and give them a unified time stamp; Obtain frequency domain information using Fourier transform based on acoustic disturbance data; obtain main frequency, energy difference and spectrum skewness based on spectrum information; calculate disturbance sound wave index based on spectrum skewness, main frequency and energy difference; A data set was established using a sliding window method based on disturbance acoustic wave indicators and water environment gradient data. A machine learning model was trained based on the data set to predict whether sediment disturbance would exceed pollution standards at the next k moments. Setting different levels of pollution release thresholds for different water body types; Based on the comparison between the output of the machine learning model and the preset pollution release thresholds of different levels, when the output value of the machine learning model is greater than the pollution release threshold, different levels of sediment pollution warning instructions are generated.

2. A method for detecting pollution sources for water environment management according to claim 1, characterized in that: The acoustic disturbance data refers to the echo spectrum signal generated by the sediment particles after being disturbed; The frequency of collecting acoustic disturbance data is 10 times per second, and the edge preprocessing includes: signal conversion and invalid signal elimination; The first group of water environment data is four types of indicators at the bottom sediment; the second group of water environment data is four types of indicators 5 cm vertically above the bottom sediment; the third group of water environment data is four types of indicators 15 cm vertically above the bottom sediment; the fourth group of water environment data is four types of indicators 50 cm vertically above the bottom sediment; there are four groups of water environment gradient data; each group is a spliced ​​vector; the timestamp information includes the collection date, hour, minute, second and quarter identifier.

3. A pollution source detection method for water environment management according to claim 2, characterized in that: The Fourier transform is defined as: Where: is the Fourier transform frequency distribution; is the time domain signal; is the frequency after Fourier transform; is the starting time of Fourier transform; For time; is an imaginary unit; is the time length of the window; The dominant frequency is the frequency with the largest amplitude in the spectrum; the energy difference refers to the difference in frequency domain energy between the two time windows before and after the disturbance; the spectrum skewness indicates the symmetry of the spectrum distribution; Spectral skewness ; Where: For the frequencies; for The corresponding amplitude on is the mean frequency of the spectrum; is the spectrum standard deviation; is the total number of frequency points.

4. A method for detecting pollution sources for water environment management according to claim 3, characterized in that: Disturbance Sonic Wave Indicator Where, is the main frequency; is the energy difference; =10 -6 ; =1.2; The process of establishing a data set includes: using the disturbance acoustic wave index and water environment gradient data with the same timestamp as feature data, and the label corresponding to the feature data is the pollution release label at time k after the timestamp; when the pollutant concentration is detected at any position 5cm, 15cm, or 50cm above the bottom mud, the pollution release label is set to 1; otherwise, it is set to 0; A set of feature data and labels is constructed as a sample, and multiple groups of samples are collected to establish a data set; the data set is divided into a training set, a validation set, and a test set, where the training set accounts for 60% of the data set, and the validation set and the test set each account for 20% of the data set.

5. A pollution source detection method for water environment management according to claim 4, characterized in that: The training process of the machine learning model includes: Using the training set as input to a machine learning model, the model outputs a predicted value of a pollution release label at k moments in the future, uses the characteristic data of real-time disturbance acoustic wave indicators and water environment gradient data as prediction targets, and minimizes the machine learning model loss function value as a training objective; training stops when the machine learning model loss function value is less than or equal to a preset target loss value; The machine learning model loss function is the mean square error; the mean square error is calculated by Minimize to train the model; in the loss function is the loss function value, i is the sample group number; is the number of sample groups; is the label of the i-th group of samples, The pollution release label predicted for the i-th group of samples; The machine learning model is a recurrent neural network model, which can be a long short-term memory network or a gated recurrent unit.

6. A method for detecting pollution sources for water environment management according to claim 5, characterized in that: The urban river threshold sets a third-level pollution release threshold, the artificial water body of the regulating reservoir sets a second-level pollution release threshold; the eutrophic lake sets a first-level pollution release threshold; the third-level pollution release threshold > the second-level pollution release threshold > the first-level pollution release threshold.

7. A method for detecting pollution sources for water environment management according to claim 6, characterized in that: Level 1 sediment pollution warning instructions include the system immediately sending a high-level warning, notifying the environmental emergency response team to quickly respond, and initiating the emergency pollution control plan, including the rapid deployment of pollution interception measures, the implementation of pollution source blocking and rapid water quality restoration to prevent the spread of pollution; Level 2 sediment pollution warning instructions include: the system automatically generates a medium-level warning, notifying on-site management personnel to conduct focused inspections and intensive water quality monitoring in relevant waters, and to deploy pollution control resources in advance if necessary for further response; The third-level sediment pollution warning instructions include: the system issues a low-level reminder, recommending that staff conduct regular inspections of the water area, make preparations for pollution prevention and control, pay attention to changes in pollution trends, and record relevant data for analysis and evaluation.

8. A pollution source detection system for water environment management, the system comprising: The data collection module collects acoustic disturbance data and performs edge preprocessing in real time. It also collects water environment data simultaneously with the acoustic disturbance data. The water environment data includes four groups of indicators at the same time. Each group of indicators includes water temperature, pH value, dissolved oxygen concentration, and flow rate. The data processing module calculates the water environment gradient data based on the water environment data; assuming that the water environment data is ; is the water environment data group number, =1, 2, 3, 4 are the first, second, third and fourth groups of water environment data respectively; is the water environment data category identifier, =1, 2, 3, 4 represent water temperature, pH value, dissolved oxygen concentration and flow rate respectively; the water environment gradient data is: Where, =1; The data synchronization module introduces the water environment gradient data and acoustic disturbance data into the time synchronization mechanism and adds a unified time stamp; The data enhancement module uses Fourier transform to obtain frequency domain information based on the acoustic disturbance data; obtains the main frequency, energy difference and spectrum skewness based on the spectrum information; and calculates the disturbance sound wave index based on the spectrum skewness, main frequency and energy difference; The model training module uses a sliding window method to establish a data set based on disturbance acoustic wave indicators and water environment gradient data. Based on the data set, a machine learning model is trained to predict whether the sediment disturbance will exceed the pollution standard at the next k moments. Post-processing module, setting different levels of pollution release thresholds for different water types; Based on the comparison between the output of the machine learning model and the preset pollution release thresholds of different levels, when the output value of the machine learning model is greater than the pollution release threshold, different levels of sediment pollution warning instructions are generated.

9. An electronic device, characterized in that: include: A processor and a memory, wherein the memory stores a computer program that can be called by the processor; The processor executes the pollution source detection method for water environment treatment according to any one of claims 1 to 7 by calling the computer program stored in the memory.

10. A computer-readable storage medium, characterized in that: Instructions are stored, and when the instructions are run on a computer, the computer is caused to execute a pollution source detection method for water environment treatment according to any one of claims 1 to 7.

Images