Water quality monitoring management platform

Abstract

The present application relates to the technical field of surface water supervision, and particularly relates to a water quality monitoring management platform, comprising a monitoring network and a management module; the monitoring network comprises shooting devices arranged at various water withdrawal path nodes, wherein the water withdrawal path nodes are provided with plankton which is sensitive to pollutants; the monitoring network is used to acquire the characteristic data of the plankton on the corresponding water withdrawal path nodes through the shooting devices; the management module is used to output a pollution diffusion path probability graph according to the characteristic data of the plankton, generate a water quality management scheme according to the pollution diffusion path probability graph, and realize real-time monitoring and accurate management of the surface water quality.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of surface water supervision, and particularly relates to a water quality monitoring management platform. BACKGROUND

[0002] At present, monitoring the water quality of surface water is an important measure to prevent and control surface water pollution and protect the ecological environment of a river basin. It can timely grasp the status of pollutants in the water, thereby providing a decision basis for taking precise management measures and avoiding irreversible ecological damage caused by the spread of pollution.

[0003] The conventional supervision method mainly relies on collecting water samples by artificial patrol inspection and then performing laboratory detection. However, this kind of method has a significant problem, that is, there are many surface water outfall points, pollutants are formed instantaneously and quickly spread along with rainfall runoff, and the conventional scheme can only provide static water quality data at an isolated time point and an isolated position, and cannot capture the dynamic migration process of pollutants in a complex water outflow path, so that it is difficult to continuously track the travel trajectory of pollutants in time and space, resulting in difficulties in tracing the pollution source and analyzing the diffusion path, and failing to meet the needs of real-time monitoring and precise management. SUMMARY

[0004] In order to meet the needs of real-time monitoring and precise management of the water quality of surface water, the present application provides a water quality monitoring management platform.

[0005] The water quality monitoring management platform provided by the present application adopts the following technical scheme:

[0006] A water quality monitoring management platform comprises a monitoring network and a management module.

[0007] The monitoring network comprises a shooting device arranged at each water outflow path node, wherein the water outflow path node is provided with plankton sensitive to pollutants.

[0008] The monitoring network is configured to acquire plankton feature data corresponding to the water outflow path node through each shooting device.

[0009] The management module is configured to output a pollution diffusion path probability graph according to the plankton feature data, and generate a water quality management scheme according to the pollution diffusion path probability graph.

[0010] Further, the step of acquiring plankton feature data corresponding to the water outflow path node through each shooting device comprises:

[0011] capturing an image of the plankton corresponding to the water outflow path node through the shooting device to obtain an initial image;

[0012] The initial image is preprocessed to output the plankton feature data.

[0013] Further, the step of preprocessing the initial image to output the plankton feature data includes:

[0014] After converting the initial image into a grayscale image, the grayscale image is denoised to obtain a denoised grayscale image.

[0015] Based on the denoised grayscale image, a binary image is generated, and the binary image is segmented according to the target contour of the plankton to obtain a segmented image.

[0016] After an initial region of interest image is cropped from the segmented image through a preset shape, the initial region of interest image is calibrated to obtain a calibrated region of interest image.

[0017] From the calibrated region of interest image, a preset morphological feature parameter is extracted to obtain the plankton feature data.

[0018] Further, the step of outputting a pollution diffusion path probability map according to the plankton feature data includes:

[0019] According to the plankton feature data and real-time environmental factor data, the concentration spatial distribution of the pollutant is inverted to generate a pollutant intensity heat map.

[0020] According to the pollutant intensity heat map and the preset hydrodynamic parameters, the pollution diffusion path probability map is output.

[0021] Further, the step of outputting the pollution diffusion path probability map according to the pollutant intensity heat map and the preset hydrodynamic parameters includes:

[0022] According to the pollutant intensity heat map and the preset hydrodynamic parameters, an initial diffusion trajectory set is obtained.

[0023] Based on the initial diffusion trajectory set, the directional transfer probability of the pollutant between adjacent water recession path nodes is calculated to generate a probability matrix of pollutant propagation between nodes.

[0024] According to the probability matrix and real-time hydrological disturbance factors, path diffusion simulation is performed to obtain a pollution diffusion path set.

[0025] On the basis of the pollution diffusion path set, a probability distribution surface is generated, and in combination with a preset risk threshold, the pollution diffusion path probability map is output.

[0026] Further, the step of generating a water quality management scheme according to the pollution diffusion path probability map includes:

[0027] In the pollution diffusion path probability map, target areas with risk diffusion probability values ​​exceeding preset risk diffusion probability values ​​are extracted, and a pollution load spatial allocation matrix is ​​generated by combining preset hydrodynamic parameters and real-time environmental factor data corresponding to the target areas.

[0028] The water quality management plan is generated based on the pollution load spatial allocation matrix, the pre-set hydraulic structure operating parameters, and the plankton characteristic data.

[0029] Furthermore, the step of generating the water quality management plan based on the pollution load spatial allocation matrix, pre-set hydraulic structure operating parameters, and the plankton characteristic data includes:

[0030] After constructing a multi-objective optimization function based on the pollution load spatial allocation matrix, the pre-set hydraulic structure operation parameters are coordinated to obtain candidate scheduling strategies.

[0031] Water quality management simulation is performed based on the candidate scheduling strategies to obtain an environmental benefit simulation report;

[0032] Based on the environmental benefit simulation report and the plankton characteristic data, a deviation analysis is performed to obtain the deviation results. The weights of the multi-objective optimization function are adjusted based on the deviation results, and the water quality management plan is output.

[0033] Beneficial effects achieved:

[0034] This application provides a water quality monitoring and management platform, including a monitoring network and a management module. The monitoring network includes imaging devices deployed at each drainage path node, wherein plankton sensitive to pollutants are present at each drainage path node. The monitoring network is used to acquire plankton characteristic data at the corresponding drainage path node through each imaging device. The management module is used to output a pollution diffusion path probability map based on the plankton characteristic data, and generate a water quality management plan based on the pollution diffusion path probability map.

[0035] In this application, by deploying imaging equipment and setting up plankton sensitive to pollutants at various drainage path nodes of surface water, the monitoring network can acquire plankton characteristic data at these drainage path nodes. This plankton characteristic data serves as an indirect but sensitive indicator of the impact of pollutants. Based on these plankton characteristic data distributed at different nodes, the management module can output a pollution diffusion path probability map by analyzing their variation patterns and spatial correlations, thereby revealing the migration direction of pollutants in complex drainage paths. Based on the dynamic diffusion trend presented by the pollution diffusion path probability map, a targeted water quality management plan can be generated, thereby achieving real-time monitoring and precise management of surface water quality. Attached Figure Description

[0036] Figure 1 This is a schematic diagram of the system modules of a water quality monitoring and management platform according to this application;

[0037] Figure 2 This is a schematic diagram illustrating the steps involved in achieving water quality monitoring through a monitoring network, as described in this application.

[0038] Figure 3 This is a flowchart illustrating the steps involved in implementing water quality monitoring and management through the management module in this application.

[0039] Explanation of icon numbers:

[0040] 10. Monitoring network; 20. Management module. Detailed Implementation

[0041] The following is in conjunction with the appendix Figures 1-3 This application will be described in further detail.

[0042] In the description of this application, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the present invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0043] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection between two components. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0044] This application discloses a water quality monitoring and management platform.

[0045] Please refer to Figure 1 The water quality monitoring and management platform proposed in this embodiment includes a monitoring network 10 and a management module 20.

[0046] The monitoring network 10 includes imaging equipment deployed at various drainage path nodes, where plankton sensitive to pollutants are present at the drainage path nodes;

[0047] Monitoring network 10 is used to acquire plankton characteristic data at corresponding drainage path nodes through various imaging devices;

[0048] The management module 20 is used to output a pollution diffusion path probability map based on plankton characteristic data, and generate a water quality management plan based on the pollution diffusion path probability map.

[0049] In this embodiment, a water quality monitoring and management platform comprising a monitoring network 10 and a management module 20 is constructed to achieve efficient monitoring and precise management of surface water pollution. Specifically, the monitoring network 10 includes imaging devices deployed at various drainage path nodes and utilizes plankton, which are sensitive to pollutants, as natural indicators. This non-invasive biomonitoring method captures early diffusion signals of pollutants, overcoming the drawbacks of traditional sensors such as high cost and susceptibility to interference in complex aquatic environments. The role of the monitoring network 10 in acquiring plankton characteristic data through imaging devices is to transform the physiological responses of plankton, such as changes in chlorophyll fluorescence and morphological abnormalities, into quantifiable pollution indicators, providing a biosensing basis for subsequent analysis.

[0050] The management module 20 outputs a pollution diffusion path probability map based on plankton characteristic data, which can dynamically simulate the migration trajectory and risk distribution of pollutants in the drainage path, overcoming the limitations of low efficiency and vague positioning of manual investigation. Finally, the water quality management plan generated based on the pollution diffusion path probability map can transform predictive data into executable instructions, realizing closed-loop control from pollution early warning to treatment response.

[0051] This water quality monitoring and management platform, through the integration of biological monitoring and intelligent decision-making, can improve the timeliness of pollution monitoring and the targeting of management measures while reducing hardware dependence, providing low-cost and high-precision technical support for the treatment of surface water pollution.

[0052] It should be noted that the plankton in this embodiment are hyperaccumulating plants that are sensitive to pollutants such as nitrogen and phosphorus, such as water hyacinth.

[0053] Reference Figure 2 As shown, the specific implementation method for real-time monitoring of surface water quality through a monitoring network is as follows:

[0054] Step S11: Use a camera to capture images of plankton at the corresponding drainage path nodes to obtain initial images.

[0055] By deploying dedicated imaging equipment at the nodes of the surface water receding path, the equipment adopts an orthogonal layered flash shadowless lighting design. It eliminates the interference of suspended particles in the water by synergistic illumination from multiple light sources at specific angles, while avoiding strong light stimulation that could cause abnormal behavior of plankton.

[0056] During filming, the camera lens is pointed vertically downwards at the plankton community. It can be set to shoot at a high speed of 25 frames per second within a distance of 0.5 meters, using linear scanning technology to capture the details of biological morphology line by line. The camera's built-in wide-angle lens covers a monitoring area with a diameter of 1 meter, ensuring a complete record of the spatial distribution of plankton. The imaging process uses backlighting to create high-contrast silhouettes of plankton against a dark background, ensuring sharp outlines.

[0057] This technology enables the acquisition of distortion-free, low-noise planktonic silhouette images, i.e., initial images, in complex aquatic environments. This provides a high-fidelity visual data foundation for subsequent feature extraction, overcoming the problems of image blurring and loss of detail caused by light scattering, biological movement, and occlusion by suspended objects in traditional underwater photography.

[0058] Step S12: Preprocess the initial image to output plankton feature data.

[0059] The purpose of this embodiment is to transform the physiological response characteristics of plankton in the initial image into quantitative indicators that can directly characterize the degree of water pollution, namely plankton characteristic data, so as to establish a mapping relationship between changes in plankton morphology and the degree of water pollution.

[0060] By employing a systematic image processing workflow to eliminate the impact of environmental disturbances on biological phenotypes, typical biological characteristics of polluted phytoplankton are accurately extracted, making these characteristics reliable indicators of pollution levels. For example, pre-processed biological characteristics can clearly show a positive correlation between leaf color changes and total phosphorus concentration in water, or a quantitative correspondence between population density growth and ammonia nitrogen load. This transforms phytoplankton into natural biosensors, providing a highly sensitive and low-cost pollution monitoring method for water quality management. By inverting pollution levels through biological response characteristics, it overcomes the problems of traditional physicochemical sensors being prone to failure and having high maintenance costs in complex aquatic environments, laying a reliable data foundation for subsequent pollution diffusion prediction and management decisions.

[0061] In one feasible implementation, step S12 includes the following steps:

[0062] Step S121: After converting the initial image to a grayscale image, the grayscale image is denoised to obtain a denoised grayscale image.

[0063] First, a single-channel grayscale image containing only brightness information is generated by weighted averaging of the red, green, and blue channel values ​​of each pixel in the initial image, in order to eliminate color interference and reduce data dimensionality.

[0064] Next, taking each pixel in the grayscale image as the center, a 3×3 or 3×5 neighborhood window is selected. The grayscale values ​​of all pixels in the window are sorted by size, and the median value is used to replace the original pixel value. For example, if the window pixel value sequence is [15,20,18,100,16,17,19,21,14], after sorting by size it becomes [14,15,16,17,18,19,20,21,100]. The median value 18 is used to replace each original pixel value. This eliminates noise caused by suspended particles in the water or the imaging equipment, while effectively preserving the edge features of plankton, resulting in a denoised grayscale image.

[0065] It should be noted that the weighted average calculation formula in this embodiment is as follows:

[0066]

[0067] Where Gray is the grayscale value corresponding to the pixel, 0.299 is the weight of the red channel, R is the brightness value of the red channel, 0.587 is the weight of the green channel, G is the brightness value of the green channel, 0.144 is the weight of the blue channel, and B is the brightness value of the blue channel.

[0068] Step S122: Based on the denoised grayscale image, a binary image is generated, and then the binary image is segmented according to the target contour of the plankton to obtain a segmented image.

[0069] The grayscale image is converted into a binary image containing only pure black and pure white pixels by setting a global or adaptive threshold. Specifically, the optimal threshold is automatically calculated using the maximum inter-class variance method. This algorithm traverses all possible thresholds and calculates the inter-class variance between plankton and water. The threshold corresponding to the maximum inter-class variance is the optimal segmentation threshold. For example, if the calculated threshold T=125, pixels with a grayscale value ≥125 are set to white, and pixels with a grayscale value <125 are set to black, thereby generating a binary image.

[0070] Next, the binary image is segmented based on the target contour of the plankton. The Canny edge detection algorithm is used to extract the boundary image of the continuous closed plankton contour. Then, morphological closing operation is used to fill the internal holes of the contour and smooth the edges. Finally, each connected region is marked by the boundary tracking method to separate the plankton from the water body and obtain a segmented image that only highlights the morphological features of the plankton. This accurately extracts the complete distribution area of ​​plankton in the image, eliminates the interference of water turbidity, uneven lighting and suspended particles, and provides a highly reliable target area for subsequent biological morphological feature analysis, ensuring the accuracy of pollution inversion.

[0071] It should be noted that the Canny edge detection algorithm first performs Gaussian filtering on the binary image to remove noise, resulting in a smooth image. Then, it calculates the gradient magnitude and direction of the smooth image, performs non-maximum suppression to refine the edges, and finally performs double thresholding to connect the broken contours, achieving continuous connection of the broken edges and ultimately outputting a complete and closed plankton contour boundary image.

[0072] Step S123: After extracting the initial image of interest from the segmented image using a preset shape, the initial image of interest is calibrated to obtain the calibrated image of interest.

[0073] Based on a preset geometric contour, such as a rectangle, circle, or polygon selection area, the target region of plankton is located and selected in the segmented image. Specifically, the smallest bounding rectangle of the plankton contour in the segmented image is first identified. Using the geometric center point as a reference, an initial image of interest containing the complete plankton is cropped according to a preset size, such as a rectangle with a side length of 100 pixels, to ensure that the plankton is not cropped or missing.

[0074] Next, by comparing the time-series images, such as the change in sharpness between the current frame and the previous frame, if the blurriness of the initial image of interest (IOI) corresponding to the current frame is higher than that of the IIO corresponding to the previous frame, a dynamic calibration mechanism is triggered to perform sharpening processing. This involves superimposing the Gaussian-blurred IIO with the original IIO to improve edge contrast, or recalculating the center point coordinates of the IIO based on the centroid offset of the contour. Finally, the calibrated IIO is output to eliminate image blurring and target offset caused by fluctuations in water turbidity or displacement of the imaging equipment. This ensures the spatial consistency and sharpness stability of the planktonic feature extraction area in the time dimension, providing a high-fidelity data foundation for subsequent measurement of biological morphological characteristic parameters and ensuring a low error rate in pollution response feature analysis.

[0075] Step S124: Extract preset morphological feature parameters from the calibrated image of interest to obtain planktonic feature data.

[0076] First, the binarized contours of plankton in the calibrated image of interest are identified. Basic morphological parameters of the plankton, such as perimeter, area, and aspect ratio of the minimum bounding rectangle, are calculated. The perimeter is accurately measured using the Freeman chain code algorithm, and the area is calculated by counting pixels through connected component labeling. Then, high-level morphological features are extracted, such as roundness = 4π × area / perimeter. 2 The body proportions are calculated as follows: body ratio = major axis / minor axis; leaf shape = number of contour concave points; eccentricity = focal length / major axis length. Among these, leaf shape is identified by the K-Curvature algorithm to identify the wavy undulation features of the leaf edge, and eccentricity is calculated based on the least circumscribed ellipse fitting.

[0077] The above parameters are mapped to plankton characteristic data through a preset pollution response model, such as a negative correlation between decreased roundness and total phosphorus concentration, and a positive correlation between increased foliage and ammonia nitrogen load. For example, a foliage coefficient of 0.35 corresponds to an ammonia nitrogen concentration of 2.1 mg / L. This establishes a quantitative correlation between plankton and the degree of water pollution, transforming changes in biological phenotypic characteristics into calculable pollution indicators. This overcomes the shortcomings of traditional physicochemical monitoring, such as delayed response and high cost in sudden pollution events, and provides a highly timely and low-cost biological sensing data source for water quality management.

[0078] Reference Figure 3 As shown, the specific implementation method for precise management of surface water quality through the management module is as follows:

[0079] Step S21: Based on plankton characteristic data and real-time environmental factor data, the spatial distribution of pollutant concentration is inverted to generate a pollutant intensity heat map.

[0080] Phytoplankton characteristic data, along with real-time environmental factors such as total nitrogen, dissolved oxygen, and water temperature, are input into a pre-trained multivariate regression model. This model dynamically calculates the pollution level at each drainage path node by analyzing the phytoplankton response to pollution and the correction effect of real-time environmental factors. Based on a spatial interpolation algorithm, the pollution level at each drainage path node is converted into a continuous spatial distribution surface. Through gradient color mapping, such as dark red → light yellow corresponding to high → low concentration, a pollutant intensity heatmap is generated. This breaks through the spatiotemporal limitations of traditional point source monitoring, enabling a visualized and dynamic presentation of pollution intensity across the entire area. It accurately locates the core pollution area in the drainage path, providing spatial decision-making basis for targeted governance and effectively improving the efficiency of pollution source tracing.

[0081] The formula for the multiple regression model is as follows:

[0082]

[0083] in, The degree of pollution; For constant terms; The regression coefficient for total nitrogen quantifies the impact of total nitrogen concentration on the degree of water pollution. Total nitrogen; The regression coefficient for dissolved oxygen quantifies the impact of dissolved oxygen concentration on the degree of water pollution. For dissolved oxygen, The regression coefficient for water temperature quantifies the impact of water temperature changes on the degree of water pollution. Water temperature; This is the error term.

[0084] Step S22: Output a probability map of pollution diffusion paths based on the pollutant intensity thermogram and preset hydrodynamic parameters.

[0085] Based on the pollutant intensity heat map and the pre-set hydrodynamic parameters, a pollution diffusion path probability map is output, which can predict and visualize the migration trajectory and impact range of pollutants in water bodies, and transform static pollution distribution data into a dynamic risk warning tool.

[0086] This pollution diffusion path probability map, by coupling the pollutant intensity heat map with pre-set hydrodynamic parameters, simulates the spatiotemporal evolution of pollutants spreading with water flow, generating a pollution diffusion path probability map that characterizes the likelihood of different areas being affected by pollution with probability values. This provides a basis for decision-making on pollution trend prediction, high-risk area identification, and emergency measure priority for water quality management, significantly improving the response speed and management accuracy of water pollution incidents.

[0087] Among them, the preset hydrodynamic parameters are the flow velocity, longitudinal diffusion coefficient, etc. of the corresponding water body.

[0088] Step S23: Extract target areas whose risk diffusion probability values ​​exceed preset risk diffusion probability values ​​from the pollution diffusion path probability map, and generate a pollution load spatial allocation matrix by combining preset hydrodynamic parameters and real-time environmental factor data corresponding to the target areas.

[0089] The preset risk diffusion probability value is a pre-set critical threshold for risk classification.

[0090] Spatial queries were performed on the pollution diffusion path probability map to filter out the set of grid cells whose risk diffusion probability values ​​exceeded a preset threshold. Polygonal boundaries of continuous high-risk areas were extracted using a boundary tracing algorithm and marked as target areas. Preset hydrodynamic parameters corresponding to the target areas were obtained, and combined with real-time environmental factor data, a dynamic load allocation model was used to calculate the spatial allocation matrix of the pollution load. Specifically:

[0091] Using the target area as the basic unit, the pollutant migration capacity index is calculated based on the flow velocity in the hydrodynamic parameters. The normalized pollution load value of each unit is calculated by superimposing real-time environmental factor data, and finally a pollution load spatial distribution matrix is ​​generated. This realizes the transformation of the probability prediction of high-risk areas into a quantifiable spatial distribution of pollution loads, breaking through the traditional model of uniform distribution and providing data support for differentiated processes such as gate control and purification facility deployment.

[0092] Step S24: Based on the pollution load spatial allocation matrix, pre-set hydraulic structure operating parameters, and plankton characteristic data, a water quality management plan is generated.

[0093] This embodiment uses a pollution load spatial allocation matrix to identify high-risk areas, combines pre-set hydraulic structure operation parameters to generate engineering control instructions, and introduces plankton characteristic data to verify the treatment effect in real time and trigger dynamic correction. Finally, it outputs a water quality management scheme that includes engineering control instructions, biological monitoring feedback mechanisms, and emergency response plans. This breaks through the limitations of traditional single-point treatment and realizes closed-loop management of "spatial targeted control - precise engineering scheduling - real-time biological feedback", effectively improving the accuracy of water quality management.

[0094] Among them, the pre-set operating parameters for hydraulic structures include gate opening and closing degree, pump station start and stop sequence, etc.

[0095] In one feasible implementation, step S22 includes the following steps:

[0096] Step S221: Based on the pollutant intensity thermogram and preset hydrodynamic parameters, obtain the initial diffusion trajectory set.

[0097] Using the concentration values ​​of grid nodes in the pollutant intensity heatmap as initial input, and combining them with pre-set hydrodynamic parameters, a water quality migration model based on the convection-diffusion equation is constructed:

[0098]

[0099] in, This indicates the rate of change of pollution levels over time. The velocity of the water body, The longitudinal diffusion coefficient is... This indicates the contribution of diffusion behavior to the change in concentration over time. This indicates the loss caused by reaction decay.

[0100] This water quality migration model is solved discretically using the finite difference method or the finite volume method. At each time step, it calculates the comprehensive migration process of pollutants with water flow, including advection, turbulent diffusion, and biochemical decay, and generates an initial diffusion trajectory set that evolves in time series. This enables accurate simulation of the spatiotemporal migration path of pollutants driven by hydrodynamics, breaking through the limitations of static pollution analysis and providing a dynamic trajectory prediction basis for pollution early warning and emergency interception.

[0101] Step S222: Based on the initial diffusion trajectory set, calculate the probability of directional transfer of pollutants between adjacent drainage path nodes, and generate a probability matrix of pollutant propagation between nodes.

[0102] Analyze the initial diffusion trajectory set, count the frequency of pollutant transfer between adjacent nodes within a specific time window, and calculate the transfer probability from node i to j using conditional probability:

[0103]

[0104] in, This represents the transition probability from node i to j. This represents the number of transitions from i to j. This represents the total number of times node i has been left.

[0105] Next, the interference of the node's own historical state on the transfer direction is eliminated by the transfer entropy algorithm, the intensity of the directional information flow from i to j is quantified, and an n×n-dimensional probability matrix M is constructed. This breaks through the limitations of traditional symmetric correlation analysis, accurately identifies the asymmetric migration pattern of pollutants in the discharge path, and provides spatial causal basis for management decisions such as gate control and emergency interception.

[0106] Step S223: Based on the probability matrix and real-time hydrological disturbance factors, perform path diffusion simulation to obtain a set of pollution diffusion paths.

[0107] Based on the probability matrix, the transfer probability is dynamically corrected by combining real-time hydrological disturbance factors. When the rainfall intensity increases, the flow velocity enhancement coefficient amplifies the downstream transfer probability; when the gate is closed, the target node transfer probability decay coefficient reduces the probability of the blocked path; and the tidal phase modulates the estuary node probability weights through a periodic function.

[0108] Next, the Markov chain Monte Carlo method is used for path simulation. Specifically, starting from the pollution source node, a single diffusion path is generated node by node based on the corrected transfer probability. This process is repeated until all nodes have been calculated, generating a set of paths containing high, medium and low probabilities of coverage, i.e., a set of pollution diffusion paths. This is used to quantify the disturbance pattern of extreme hydrological events, such as rainstorms combined with ebb tides, on the migration path of pollutants, and to generate multi-scenario diffusion trajectories.

[0109] Step S224: Based on the set of pollution diffusion paths, generate a probability distribution surface and, in conjunction with a preset risk threshold, output a probability map of pollution diffusion paths.

[0110] The set of pollution diffusion paths is spatially rasterized, and the frequency of each raster cell being covered by the path is counted. A Gaussian kernel function is used to calculate the spatially continuous probability density value, generating a smooth probability distribution surface. Then, a preset risk threshold (e.g., high risk ≥ 0.7, medium risk 0.4-0.7, low risk < 0.4) is used to segment the surface. Gradient color mapping (e.g., dark red → orange → light yellow corresponding to high → medium → low risk) is employed to generate a pollution diffusion path probability map. High-risk areas (probability ≥ 0.7) can be automatically marked as emergency response core areas (e.g., with flashing boundary warnings). This transforms the abstract set of pollution diffusion paths into an intuitive pollution diffusion path probability map, accurately quantifying sudden hydrological events, such as the probability of pollution exposure in different areas under heavy rain and gate closure, effectively improving the efficiency of emergency response to pollution events.

[0111] In one feasible implementation, step S24 includes the following steps:

[0112] Step S241: After constructing a multi-objective optimization function based on the pollution load spatial allocation matrix, coordinate the pre-set hydraulic structure operating parameters to obtain candidate scheduling strategies.

[0113] Using the proportion of pollution load in each target area reflected in the pollution load spatial allocation matrix as constraints, a multi-objective optimization function is constructed, including: the first objective function is the water quality compliance rate; the second objective function is the total carbon emissions; and the third objective function is the operation and maintenance cost.

[0114] Pre-set hydraulic structure operating parameters, such as gate opening degree θ∈[0°,90°] and pump station start-stop frequency n∈[0,5 times / day], are used as decision variables to coordinate multi-objective conflicts. For example, improving water quality requires increasing pump station frequency but also increases carbon emissions. Candidate scheduling strategies are generated, such as candidate scheduling strategy S={S1: gate opening degree 70° + pump station start-stop 3 times, S2: gate opening degree 50° + pump station start-stop 4 times}. Each candidate scheduling strategy is accompanied by a quantitative benefit assessment, such as S1: compliance rate 85%, carbon emissions 120kg, cost 8000 yuan. This breaks through the limitations of single-objective regional optimization. Under the constraint of spatial heterogeneity of pollution load, a set of engineering scheduling schemes that take into account the triple objectives of water quality, carbon emissions, and cost is generated. This upgrades management decision-making from experience-based judgment to scientific quantitative game theory, effectively improving the comprehensive benefits of water quality management.

[0115] Step S242: Perform water quality management simulation based on candidate scheduling strategies to obtain an environmental benefit simulation report.

[0116] Candidate scheduling strategies are input into a digital twin platform to drive a coupled water quality migration model. Based on the surface water drainage path topology (field → tributary ditch → main canal → ecological pond), the model simulates the pollutant migration process after strategy execution. Specifically, the water quality migration model calculates the concentration decay of nitrogen and phosphorus in the ditch network, such as ammonia nitrogen decreasing from 8 mg / L at the field outlet to 2.5 mg / L at the ecological pond inlet. Simultaneously, equipment energy consumption (e.g., pump station power consumption P = 0.8 kW) and a carbon emission factor database (e.g., diesel pump carbon emissions of 2.6 kg) are integrated to quantify the strategy's energy consumption and carbon footprint. Next, the key nodes of the water discharge path are analyzed, such as the changes in plankton characteristics at the outlet of the ecological pond, to assess the ecological restoration effect. The final environmental benefit simulation report includes core indicators such as the total reduction rate of pollutants in surface water discharge, the compliance rate of sensitive indicators of receiving water bodies, carbon emission reduction per unit area, and purification efficiency of ecological ditches. It also includes water quality change trend charts at key nodes, such as the peak change curve of ammonia nitrogen at field outlets 6 hours after a rainstorm. This breaks through the experience-based reliance on surface water non-point source pollution control, accurately simulates its environmental benefits before the implementation of the strategy, and provides a scientific basis for the coordinated decision-making of surface water source measures optimization and engineering regulation.

[0117] Step S243: Based on the environmental benefit simulation report and plankton characteristic data, perform deviation analysis to obtain deviation results. Adjust the weights of the multi-objective optimization function based on the deviation results and output a water quality management plan.

[0118] By comparing the environmental benefit simulation report with the plankton characteristic data, such as comparing the predicted ammonia nitrogen concentration of 1.2 mg / L at the outlet of the ecological ditch reflected in the environmental benefit simulation report with the measured ammonia nitrogen value of 1.8 mg / L corresponding to the abnormal chlorophyll value of sensitive algae reflected in the plankton characteristic data, the relative deviation was calculated (relative deviation = (measured value - predicted value) / predicted value × 100%).

[0119] When the relative deviation exceeds the preset relative deviation, the weight adjustment of the multi-objective optimization function is triggered, that is, the weight coefficient is updated based on the gradient descent method. For example, the weight of water quality compliance rate is set to 0.5 and then adjusted to 0.55; the weight of total carbon emissions is simultaneously reduced to 0.3 and the weight of operation and maintenance cost is adjusted to 0.16.

[0120] The adjusted weights are input into the multi-objective optimization model for re-solution, generating a water quality management plan.

[0121] In this embodiment, by constructing a closed-loop learning mechanism of "prediction-execution-verification-optimization", the water quality management plan can be continuously iterated and upgraded in the dynamic surface water environment, reducing prediction bias, lowering management costs, and ensuring the sustainable achievement of ecological restoration indicators.

[0122] The above are all preferred embodiments of this application, and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. A water quality monitoring and management platform, characterized in that, Includes network monitoring and management modules; The monitoring network includes imaging equipment deployed at each drainage path node, wherein plankton sensitive to pollutants are present at each drainage path node; The monitoring network is used to acquire plankton characteristic data at the corresponding nodes of the water receding path through each of the aforementioned imaging devices; The management module is used to invert the spatial distribution of pollutant concentrations based on the plankton characteristic data and real-time environmental factor data, and generate a pollutant intensity heat map. Based on the pollutant intensity thermogram and preset hydrodynamic parameters, an initial diffusion trajectory set is obtained; Based on the initial diffusion trajectory set, the probability of directional transfer of pollutants between adjacent drainage path nodes is calculated, and a probability matrix of pollutant propagation between nodes is generated. Based on the probability matrix and real-time hydrological disturbance factors, path diffusion simulation is performed to obtain a set of pollution diffusion paths; Based on the set of pollution diffusion paths, a probability distribution surface is generated, and combined with a preset risk threshold, a pollution diffusion path probability map is output. In the pollution diffusion path probability map, target areas with risk diffusion probability values ​​exceeding preset risk diffusion probability values ​​are extracted, and a pollution load spatial allocation matrix is ​​generated by combining the preset hydrodynamic parameters corresponding to the target areas and the real-time environmental factor data. After constructing a multi-objective optimization function based on the pollution load spatial allocation matrix, candidate scheduling strategies are obtained by coordinating the pre-set hydraulic structure operation parameters. Water quality management simulation is performed based on the candidate scheduling strategies to obtain an environmental benefit simulation report; Based on the environmental benefit simulation report and the plankton characteristic data, a deviation analysis is performed to obtain the deviation results. The weights of the multi-objective optimization function are then adjusted using the deviation results to output a water quality management plan.

2. The water quality monitoring and management platform according to claim 1, characterized in that, The step of acquiring plankton characteristic data at the corresponding receding water path nodes using the aforementioned imaging devices includes: The imaging device captures images of plankton at the corresponding receding water path nodes to obtain initial images; The initial image is preprocessed to output the planktonic feature data.

3. The water quality monitoring and management platform according to claim 2, characterized in that, The step of preprocessing the initial image and outputting the planktonic feature data includes: After converting the initial image into a grayscale image, the grayscale image is denoised to obtain a denoised grayscale image. Based on the denoised grayscale image, a binary image is generated, and then the binary image is segmented according to the target contour of the plankton to obtain a segmented image. After extracting an initial image of interest from the segmented image using a preset shape, the initial image of interest is calibrated to obtain a calibrated image of interest; Preset morphological feature parameters are extracted from the calibrated image of interest to obtain the planktonic feature data.

Images