A method for detecting water pollution in a buried pipe drainage well and a water pollution monitor

Abstract

The application provides a method for detecting water pollution in a buried pipe drainage well and a water pollution monitor. The water pollution concentration of each detection node in the buried pipe drainage well is detected by an intelligent water quality sensor. The mutual support degree of the water pollution detection results between the intelligent water quality sensors is determined according to the cross-correlation coefficient of the water pollution concentration between different detection nodes and the spatial topology structure of the intelligent water quality sensor. The interference evaluation index of the water pollution detection at different depths is obtained by fuzzy evaluation of the interference effect intensity of the stratified water pollution detection at different depths according to the vertical concentration gradient of the water pollution in the buried pipe drainage well and the local flow velocity of the water body at different depths. The water pollution concentration of each detection node is cost fused based on all the mutual support degrees and the interference evaluation index of the water pollution detection at different depths, and the fusion pollution degree of the water pollution is obtained. The cost fusion of multi-source information in the water pollution detection process of the buried pipe drainage well can be realized by adopting the scheme.

Description

Technical Field

[0001] This application relates to the field of water pollution detection technology, and more specifically, to a method for detecting water pollution in a concealed drainage well and a water pollution monitoring instrument. Background Technology

[0002] Water pollution detection is a crucial link in ensuring the quality of water bodies and public health and safety. Especially in urban drainage systems, pollutants are discharged into drainage wells through underground pipes, often causing local water quality deterioration and even triggering ecological disasters. To promptly grasp the spread of pollution, conducting efficient and accurate water pollution monitoring has become a key task in current urban water environment management. Traditional water quality testing mostly relies on fixed-point sampling or manual inspection, which is inefficient and has limited coverage, making it difficult to cope with the spatial heterogeneity and dynamic changes of pollutants in underground drainage wells. Therefore, establishing a multi-point pollution detection technology system based on intelligent sensing and information fusion has become an important direction for improving the intelligence and precision of pollution monitoring.

[0003] Existing water pollution detection methods generally suffer from a strong reliance on single-point detection. Most technologies collect data from only one or a few detection nodes, lacking comprehensive consideration of the detection relationships between sensors and spatial interference factors. When faced with complex groundwater flow disturbances and vertical distribution variations of pollutants, these methods are prone to significant deviations in overall detection results due to node anomalies or positional shifts, failing to accurately reflect the overall pollution status of drainage wells. Furthermore, the lack of crucial information exchange and fusion mechanisms between detection nodes prevents the full exploitation of the correlation characteristics and complementary advantages between multi-source detection data, leading to one-sided assessments of pollution concentrations and hindering reliable support for accurate identification, source tracing, and remediation of water pollution. Therefore, achieving cost-effective fusion of multi-source information during the detection of water pollution in underground drainage wells to enhance the accuracy of overall detection results has become a challenging problem for the industry. Summary of the Invention

[0004] This application provides a method for detecting water pollution in underground drainage wells and a water pollution monitoring instrument, which can achieve cost fusion of multi-source information during the detection of water pollution in underground drainage wells.

[0005] In a first aspect, this application provides a method for detecting water pollution in a concealed drainage well, comprising the following steps:

[0006] Multiple detection nodes and intelligent water quality sensors are deployed in the target underground drainage well. Water pollution data of each detection node is obtained by detecting the water pollution data of each detection node through the intelligent water quality sensors, and then the water pollution concentration of each detection node is extracted.

[0007] Based on the water pollution data of each detection node within a historical time period, the cross-correlation coefficients of water pollution concentration between different detection nodes are determined. Then, the mutual support of water pollution detection results between each smart water quality sensor is determined by all the cross-correlation coefficients and the spatial topology of the smart water quality sensor.

[0008] The vertical concentration gradient of water pollution in the target underground drainage well is extracted. Based on the vertical concentration gradient and the local flow velocity of water bodies at different depths, the intensity of the stratified interference effect of water pollution detection at different depths is evaluated in a fuzzy manner to obtain the interference evaluation index of water pollution detection at different depths.

[0009] Based on all mutual support and interference evaluation indices of water pollution detection at different depths, the water pollution concentration at each detection node is fused at a cost to obtain the fused pollution degree of water pollution, and then the fused pollution degree is used as the pollution degree of the water body in the target underground drainage well.

[0010] Preferably, the water pollution data at each detection node is obtained through intelligent water quality sensors, and the water pollution concentration at each detection node is then extracted, specifically including:

[0011] For each detection node, water pollution data is collected in real time using intelligent water quality sensors.

[0012] The collected water pollution data is smoothed to remove outliers and noise interference, resulting in smoothed water pollution data.

[0013] The water pollution concentration of each detection node is extracted from the smoothed water pollution data by using a preset pollutant concentration calculation model.

[0014] Preferably, the correlation coefficients between water pollution concentrations at different monitoring nodes, determined based on water pollution data from various monitoring nodes within a historical time period, specifically include:

[0015] For every two detection nodes, the water pollution data of the two detection nodes within the historical time period are time-aligned to obtain a time-aligned water pollution data set.

[0016] Based on the time-aligned water pollution data set, cross-correlation analysis was performed on the water pollution concentration between two detection nodes to obtain the cross-correlation coefficient of water pollution concentration between the two detection nodes, and then the cross-correlation coefficient of water pollution concentration between each pair of detection nodes was obtained.

[0017] Preferably, determining the mutual support of water pollution detection results among various smart water quality sensors by using all cross-correlation coefficients and the spatial topology of the smart water quality sensors specifically includes:

[0018] For every two smart water quality sensors, the confidence distance between the two smart water quality sensors is determined based on the spatial topology of the smart water quality sensors;

[0019] The consistency of water pollution detection results between two smart water quality sensors is evaluated by using all cross-correlation coefficients and the confidence distance to obtain the cross-support degree of water pollution detection results between the two smart water quality sensors, and then the cross-support degree of water pollution detection results between each pair of smart water quality sensors is obtained.

[0020] Preferably, extracting the vertical concentration gradient of water pollution in the target underground drainage well specifically includes:

[0021] The detection node is divided into at least three detection layers in the vertical direction to obtain the pollutant concentration of each layer.

[0022] Based on the concentration difference between adjacent detection layers, cubic spline interpolation is performed on the water pollution concentration in the vertical direction to generate a continuous vertical concentration distribution curve.

[0023] The first derivative of the vertical concentration distribution curve within a preset depth interval is determined, and the maximum gradient value is extracted as the vertical concentration gradient of water pollution in the target underground drainage well.

[0024] Preferably, the fuzzy evaluation of the stratification interference effect intensity of water pollution detection at different depths is performed based on the vertical concentration gradient and the local flow velocity of water bodies at different depths, resulting in an interference evaluation index for water pollution detection at different depths. Specifically, this includes:

[0025] Initialize the initial membership degree of the interference effect of different detection layers;

[0026] An interference effect intensity matrix between each detection layer is constructed based on the vertical concentration gradient and the local flow velocity of water at different depths.

[0027] The numerical values ​​in the interference effect intensity matrix are mapped to fuzzy sets using fuzzy membership functions;

[0028] Construct a fuzzy interference evaluation matrix based on all initial membership degrees and the fuzzy set;

[0029] Based on the fuzzy interference evaluation matrix, the interference effect intensity of each detection layer is evaluated using fuzzy methods to obtain the interference evaluation index for water pollution detection corresponding to each detection layer.

[0030] Preferably, the intelligent water quality sensor refers to an integrated sensing device that has multi-parameter water quality sensing capabilities, data preprocessing functions, and a communication module.

[0031] Secondly, this application provides a water pollution monitoring instrument, which includes a water pollution detection unit, the water pollution detection unit comprising:

[0032] The detection module is used to detect water pollution data at each detection node through intelligent water quality sensors, and then extract the water pollution concentration at each detection node.

[0033] The processing module is used to determine the cross-correlation coefficients of water pollution concentrations between different detection nodes based on water pollution data from each detection node within a historical time period, and then determine the mutual support of water pollution detection results between each smart water quality sensor through all the cross-correlation coefficients and the spatial topology of the smart water quality sensor.

[0034] The processing module is also used to extract the vertical concentration gradient of water pollution in the target underground pipe drainage well, and to perform a fuzzy evaluation of the stratification interference effect intensity of water pollution detection at different depths based on the vertical concentration gradient and the local flow velocity of water bodies at different depths, so as to obtain the interference evaluation index of water pollution detection at different depths.

[0035] The execution module is used to perform cost fusion of the water pollution concentration at each detection node based on all mutual support and interference evaluation indices of water pollution detection at different depths, to obtain the fused pollution degree of water pollution, and then use the fused pollution degree as the pollution degree of the water body in the target underground drainage well.

[0036] Thirdly, this application provides a computer device, the computer device including a memory and a processor, the memory storing code, and the processor being configured to acquire the code and execute the above-described method for detecting water pollution in underground drainage wells.

[0037] Fourthly, this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for detecting water pollution in underground drainage wells.

[0038] The technical solutions provided by the embodiments disclosed in this application have the following beneficial effects:

[0039] In this embodiment, multiple detection nodes and intelligent water quality sensors are deployed in the target underground drainage well. Water pollution data from each detection node is obtained through the intelligent water quality sensors, and the water pollution concentration at each node is extracted. Based on the water pollution data from each detection node over a historical period, cross-correlation coefficients of water pollution concentrations between different detection nodes are determined. Then, the mutual support degree of water pollution detection results among the intelligent water quality sensors is determined using all cross-correlation coefficients and the spatial topology of the intelligent water quality sensors. The vertical concentration gradient of water pollution in the target underground drainage well is extracted. Based on the vertical concentration gradient and the local flow velocity of water at different depths, a fuzzy evaluation of the stratified interference effect of water pollution detection at different depths is performed to obtain an interference evaluation index for water pollution detection at different depths. Based on all mutual support degrees and the interference evaluation index for water pollution detection at different depths, the water pollution concentrations at each detection node are fused at a cost to obtain a fused pollution degree, which is then used as the pollution degree of the water in the target underground drainage well.

[0040] Therefore, this application performs cost fusion of water pollution concentrations at each detection node based on all mutual support and interference evaluation indices of water pollution detection at different depths, to obtain the fused pollution degree of water pollution. First, by deploying multiple intelligent water quality sensor nodes in drainage wells, the problem of strong dependence on single-point detection in traditional methods is overcome, ensuring broad coverage of water quality information at different spatial locations and providing a basic guarantee for subsequent multi-source data fusion. Second, by introducing historical data to calculate the cross-correlation coefficients between each detection node and combining the spatial topology of the sensors, a mutual support evaluation system for water quality detection results between nodes is further established, effectively revealing the correlation and consistency between detection results between nodes, and enabling quantitative evaluation of the reliability of detection data. Then, by extracting vertical concentration gradient and local water flow velocity information, interference effects are constructed. The strength matrix is ​​calculated, and the interference evaluation index is calculated using a fuzzy evaluation method, thereby achieving quantitative discrimination of the reliability of pollution data of water bodies under vertical disturbance conditions and enhancing the model's response capability to the stratified pollution characteristics of water bodies. Finally, all mutual support and interference evaluation indices of water pollution detection at different depths are introduced into the fusion cost function, with the fusion weight of water pollution concentration at the detection node as the optimization variable. The fusion cost function is iteratively solved using a genetic algorithm to achieve cost fusion between detection data. This not only solves the redundancy and conflict problems existing in multi-source data, but also fully retains the effective information provided by sensors at different depths and locations, thereby obtaining pollution concentration expression results that are more consistent with reality. In summary, the proposed solution can realize cost fusion of multi-source information in the process of water pollution detection in underground pipe drainage wells, thereby improving the detection accuracy of water pollution detection in underground pipe drainage wells. Attached Figure Description

[0041] Figure 1 This is an exemplary flowchart of a water pollution detection method in a concealed drainage well according to some embodiments of this application;

[0042] Figure 2 This is a schematic diagram illustrating an application scenario for detecting water pollution in concealed drainage wells, based on some embodiments of this application.

[0043] Figure 3 This is a schematic flowchart illustrating the process of determining a vertical concentration gradient according to some embodiments of this application;

[0044] Figure 4 This is a schematic diagram of the structure of a water pollution detection unit according to some embodiments of this application;

[0045] Figure 5 This is a schematic diagram of the structure of a computer device for implementing a method for detecting water pollution in a concealed drainage well, according to some embodiments of this application. Detailed Implementation

[0046] To better understand the technical solution of this application, the technical solution of this application will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0047] refer to Figure 1 The figure is an exemplary flowchart of a water pollution detection method in a concealed pipe drainage well according to some embodiments of this application. The water pollution detection method 100 in the concealed pipe drainage well mainly includes the following steps:

[0048] In step 101, multiple detection nodes and intelligent water quality sensors are deployed in the target underground drainage well. Water pollution data of each detection node is obtained by detecting the intelligent water quality sensors, and then the water pollution concentration of each detection node is extracted.

[0049] It should be noted that, in this application, the underground drainage well refers to an inspection well or collection well that collects and discharges water through an underground closed pipe, mainly used for the collection, guidance, and detection of water flow in urban drainage systems. Specifically, the deployment of multiple detection nodes and intelligent water quality sensors in the target underground drainage well can be achieved in the following way: based on the structural form of the drainage well, the water flow path, and the pollution distribution characteristics, representative detection locations are selected, and nodes are deployed in layers in the vertical direction to cover water bodies of different depths. In the horizontal direction, nodes are deployed in conjunction with key locations such as the main water flow channel, bends, and inlets to capture pollution sources and trends. Each detection node is equipped with an intelligent water quality sensor, which specifically includes: and is connected to a unified data transmission system through a data acquisition module.

[0050] In some embodiments, reference Figure 2As shown in the figure, this figure is a schematic diagram of the application scenario of water pollution detection in a concealed pipe drainage well in some embodiments of this application. In this case, multiple smart water quality sensors 120 are installed in the concealed pipe drainage well 110. The smart water quality sensors are responsible for collecting water quality data in the drainage well. The collected water quality data is then transmitted to the data processor 130 via a data cable. The data processor executes the data processing program of the water pollution detection unit to obtain the detection results of water pollution in the concealed pipe drainage well. The detection results are then visualized through the visualization module 140, making it convenient for users to view and analyze them through computer devices.

[0051] It should be noted that the intelligent water quality sensor in this application refers to an integrated sensing device with multi-parameter water quality sensing capabilities, data preprocessing functions, and a communication module, which can monitor pollution-related physical or chemical parameters (such as chemical oxygen demand, ammonia nitrogen, and turbidity) in water bodies in real time. The intelligent water quality sensor includes a fluorescence chemical oxygen demand sensor module, an electrochemical ammonia nitrogen sensor module, and an optical turbidity sensor module.

[0052] In some embodiments, the water pollution data of each detection node is obtained by using a smart water quality sensor, and the water pollution concentration of each detection node is then extracted. This can be achieved through the following steps:

[0053] For each detection node, water pollution data is collected in real time using intelligent water quality sensors.

[0054] The collected water pollution data is smoothed to remove outliers and noise interference, resulting in smoothed water pollution data.

[0055] The water pollution concentration of each detection node is extracted from the smoothed water pollution data by using a preset pollutant concentration calculation model.

[0056] In practice, the collected water pollution data is smoothed to remove outliers and noise interference. The smoothed water pollution data can be obtained in the following way: The collected water pollution data is transmitted to an edge computing device or a remote processing center. The original data is then smoothed, and a common method is the moving average method, which replaces the current data point with the mean within a set time window to suppress short-term fluctuations. At the same time, the 3σ principle (mean ± 3 times the standard deviation) is used to remove abrupt outliers, resulting in smoothed water pollution data. The water pollution concentration of each detection node is extracted from the smoothed water pollution data through a preset pollutant concentration calculation model. The water pollution concentration of each detection node can be obtained by inputting the smoothed water pollution data into a preset pollutant concentration calculation model. This pollutant concentration calculation model is built based on a large amount of experimental data and can use a linear regression model or a multivariate mapping model. For example, by establishing a regression equation between "chemical oxygen demand-ammonia nitrogen-turbidity", the physical parameters are converted into the water pollution concentration of the detection node.

[0057] It should be noted that the pollutant concentration calculation model in this application refers to a mapping model used to convert multi-parameter raw water quality data collected by intelligent water quality sensors into pollutant concentration values. Its technical principle is based on the stable functional relationship between pollutant concentration and multiple sensing parameters (such as chemical oxygen demand, ammonia nitrogen, and turbidity). Statistical or machine learning models are established through a large number of experimental samples to complete parameter fitting and feature extraction. Commonly used models include multiple linear regression, support vector regression, or neural network models. The input is multiple water quality parameter vectors, and the output is the concentration estimate of the target pollutant. This model is trained offline and deployed to devices or servers to achieve rapid conversion of sensor data into pollutant concentration. In addition, this application can integrate multiple pollution parameters into a single water pollution evaluation index through the pollutant concentration calculation model.

[0058] In step 102, the cross-correlation coefficients of water pollution concentrations between different detection nodes are determined based on the water pollution data of each detection node within the historical time period. Then, the mutual support of water pollution detection results between each smart water quality sensor is determined by all the cross-correlation coefficients and the spatial topology of the smart water quality sensor.

[0059] In some embodiments, determining the cross-correlation coefficients of water pollution concentrations between different monitoring nodes based on water pollution data from various monitoring nodes within a historical time period can be achieved using the following steps:

[0060] For every two detection nodes, the water pollution data of the two detection nodes within the historical time period are time-aligned to obtain a time-aligned water pollution data set.

[0061] Based on the time-aligned water pollution data set, cross-correlation analysis was performed on the water pollution concentration between two detection nodes to obtain the cross-correlation coefficient of water pollution concentration between the two detection nodes, and then the cross-correlation coefficient of water pollution concentration between each pair of detection nodes was obtained.

[0062] It should be noted that the cross-correlation coefficient in this application is an indicator that measures the similarity of the water pollution concentration change trends between two detection nodes.

[0063] In practical implementation, for any two detection nodes, firstly, the water pollution data of the two detection nodes within a historical time period are time-aligned. The resulting time-aligned water pollution data set can be achieved as follows: retrieve the water pollution data sequences of each detection node within a unified historical time period, and achieve time synchronization through timestamp alignment. Specifically, select a common time reference point, and perform interpolation or downsampling on the data of the two nodes at a fixed sampling period (e.g., once every 2 minutes) to ensure that each pair of nodes forms a one-to-one corresponding time sequence. The processed data is then used as the time-aligned water pollution data set. Pollution data set; Based on the time-aligned water pollution data set, cross-correlation analysis is performed on the water pollution concentration between two detection nodes to obtain the cross-correlation coefficient between the two detection nodes. This can be achieved in the following way: For each time-aligned data sequence, the Spearman rank correlation coefficient is used to calculate the mean, variance, and covariance of the two sequences respectively. The covariance is then divided by the product of the two standard deviations to obtain the Pearson correlation coefficient with a value range between -1 and 1. The calculated Pearson correlation coefficient is then used as the cross-correlation coefficient between the water pollution concentrations of the two detection nodes.

[0064] In some embodiments, determining the mutual support of water pollution detection results among various smart water quality sensors by using all cross-correlation coefficients and the spatial topology of the smart water quality sensors can be achieved through the following steps:

[0065] For every two smart water quality sensors, the confidence distance between the two smart water quality sensors is determined based on the spatial topology of the smart water quality sensors;

[0066] The consistency of water pollution detection results between two smart water quality sensors is evaluated by using all cross-correlation coefficients and the confidence distance to obtain the cross-support degree of water pollution detection results between the two smart water quality sensors, and then the cross-support degree of water pollution detection results between each pair of smart water quality sensors is obtained.

[0067] It should be noted that the mutual support in this application is a reliability evaluation index for measuring the spatial proximity and concentration correlation of water pollution detection results between two smart water quality sensors.

[0068] In specific implementation, for each pair of smart water quality sensors, firstly, the confidence distance between the two smart water quality sensors can be determined based on their spatial topology. This can be achieved by obtaining the spatial distance between the two smart water quality sensors from their spatial topology, and then using the natural exponential function value of the inverse of the spatial distance as the confidence distance between the two smart water quality sensors. This confidence distance reflects the spatial proximity of the sensors; water quality measured by neighboring sensors should be more relevant. Secondly, the consistency of water pollution detection results between the two smart water quality sensors is evaluated by using all cross-correlation coefficients and the confidence distance to obtain the mutual support degree of water pollution detection results between the two smart water quality sensors. This can be achieved by obtaining the cross-correlation coefficient of water pollution concentration between the corresponding detection nodes of the two smart water quality sensors, and using the product of the cross-correlation coefficient and the confidence distance as the mutual support degree of water pollution detection results between the two smart water quality sensors.

[0069] In step 103, the vertical concentration gradient of water pollution in the target underground drainage well is extracted. Based on the vertical concentration gradient and the local flow velocity of water bodies at different depths, the intensity of the stratified interference effect of water pollution detection at different depths is fuzzy evaluated to obtain the interference evaluation index of water pollution detection at different depths.

[0070] In some embodiments, reference Figure 3 As shown in the figure, this is a schematic flowchart of determining the vertical concentration gradient in some embodiments of this application. In this embodiment, the extraction of the vertical concentration gradient of water pollution in the target underground drainage well can be achieved by the following steps:

[0071] In step 1031, the detection node is divided into at least three detection layers in the vertical direction, and the pollutant concentration of each layer is obtained;

[0072] In step 1032, cubic spline interpolation is performed on the water pollution concentration in the vertical direction based on the concentration difference between adjacent detection layers to generate a continuous vertical concentration distribution curve.

[0073] In step 1033, the first derivative of the vertical concentration distribution curve within a preset depth interval is determined, and the maximum gradient value is extracted as the vertical concentration gradient of water pollution in the target underground drainage well.

[0074] It should be noted that the cubic spline interpolation in this application ensures the continuity of the interpolation curve in terms of numerical values ​​and first and second derivatives by constructing a cubic polynomial between adjacent data points, thus achieving a smooth transition of the curve and effectively eliminating abrupt changes caused by uneven spacing between detection nodes or data fluctuations. The vertical concentration gradient in this application is an indicator of the rate of change of pollutant concentration in the vertical direction of the water body. The vertical concentration distribution curve in this application refers to a smooth curve that represents the continuous change of pollutant concentration in the water body with depth.

[0075] In specific implementation, firstly, the detection nodes are divided into at least three detection layers vertically. The pollutant concentration of each layer can be obtained as follows: the detection nodes are divided into at least three detection layers according to depth, based on the vertical direction of the drainage well. The pollutant concentration of all detection nodes in each layer is statistically analyzed, and the average value is taken as the pollutant concentration of the corresponding detection layer. Then, cubic spline interpolation is performed on the water pollution concentration in the vertical direction based on the concentration difference between adjacent detection layers to generate a continuous vertical concentration distribution curve. This can be achieved as follows: the concentration difference between adjacent detection layers is used as the mean feature of the data items of the cubic spline interpolation algorithm nodes. The water pollution concentration of the vertical detection nodes is calculated using the cubic spline interpolation algorithm, and the interpolated curve is taken as the vertical concentration distribution curve. This cubic spline interpolation algorithm constructs a cubic multinomial between every two adjacent data points. This method ensures that the interpolation curve is continuous not only in terms of concentration values ​​but also in terms of the continuity of the first and second derivatives, thereby generating a smooth and continuous vertical concentration distribution curve. This eliminates abrupt changes caused by uneven node spacing or data fluctuations. Finally, the first derivative of the vertical concentration distribution curve within a preset depth interval is determined, and the maximum gradient value is extracted as the vertical concentration gradient of water pollution in the target underground drainage well. This can be achieved in the following way: Based on the vertical concentration distribution curve, concentration values ​​are uniformly sampled at preset depth intervals (e.g., every 0.1 meters). The first derivative is calculated by dividing the concentration difference between adjacent sampling points by the depth interval. This first derivative reflects the rate of change of concentration with depth. The first derivatives of all sampling points are iterated, and the maximum value is extracted as the vertical concentration gradient of water pollution in the underground drainage well. The vertical concentration gradient represents the maximum change in water pollution concentration in the vertical direction.

[0076] In some embodiments, the fuzzy evaluation of the stratification interference effect intensity of water pollution detection at different depths based on the vertical concentration gradient and the local flow velocity of water bodies at different depths, to obtain the interference evaluation index for water pollution detection at different depths, can be achieved through the following steps:

[0077] Initialize the initial membership degree of the interference effect of different detection layers;

[0078] An interference effect intensity matrix between each detection layer is constructed based on the vertical concentration gradient and the local flow velocity of water at different depths.

[0079] The numerical values ​​in the interference effect intensity matrix are mapped to fuzzy sets using fuzzy membership functions;

[0080] Construct a fuzzy interference evaluation matrix based on all initial membership degrees and the fuzzy set;

[0081] Based on the fuzzy interference evaluation matrix, the interference effect intensity of each detection layer is evaluated using fuzzy methods to obtain the interference evaluation index for water pollution detection corresponding to each detection layer.

[0082] It should be noted that the initial membership degree in this application refers to the preliminary quantitative degree of attribution representing the intensity of interference effects of each detection layer; the interference effect intensity matrix in this application refers to the relationship matrix reflecting the degree of mutual interference between different detection layers in water pollution detection; the fuzzy interference evaluation matrix in this application refers to the evaluation matrix used to comprehensively evaluate the interference impact of each detection layer after the interference effect intensity is converted into a fuzzy set through the fuzzy membership function; and the interference evaluation index in this application refers to the index that quantitatively represents the intensity of interference to water pollution detection of each detection layer based on the fuzzy evaluation results.

[0083] In specific implementation, firstly, the initial membership degree of the interference effect at different depth detection layers can be initialized as follows: retrieve historical water quality monitoring records of the target underground drainage well at different water levels, extract index data related to the interference intensity such as concentration fluctuation frequency, disturbance amplitude, and abrupt change time points, and then set a set of empirical rules based on the number of times significant concentration changes occur in the historical data for each detection layer (e.g., the higher the change frequency, the greater the initial membership degree). Then, combine this with the level division in the rule base (e.g., low, medium, and high levels correspond to different membership degree value ranges), and use a membership degree assignment function (e.g., a linear piecewise function) to initialize the interference effect at each layer. Initial membership mapping is performed to obtain the initial membership of the interference effect of each detection layer. Secondly, the interference effect intensity matrix between each detection layer can be constructed based on the vertical concentration gradient and the local flow velocity of water at different depths. This can be achieved by dividing the deployed detection nodes into at least three detection layers according to depth along the vertical direction of the drainage well, and obtaining the corresponding vertical concentration gradient value and local flow velocity value for each layer. Based on a preset interference effect evaluation rule, the potential interference intensity caused by the difference in concentration variation and flow velocity between any two detection layers is calculated. This rule can use a linear superposition model, where the interference intensity is a weighted synthesis of the concentration gradient difference and the flow velocity difference. The interference intensity values ​​between each detection layer and other layers are calculated sequentially, and all calculation results are combined by layer and filled into a two-dimensional matrix. This two-dimensional matrix is ​​used as the interference effect intensity matrix. It should be further noted that the interference effect intensity matrix is ​​normalized in this application to ensure that its values ​​adapt to the input requirements of the fuzzy evaluation system. Furthermore, mapping the values ​​in the interference effect intensity matrix to fuzzy sets using fuzzy membership functions can be achieved in the following way: based on the qualitative evaluation level (such as "weak interference", "medium interference", "strong interference") corresponding to each value in the interference effect intensity matrix, a fuzzy membership function (e.g., a triangular membership function) is used. Each value in the interference effect intensity matrix is ​​mapped to the membership value of the corresponding evaluation level in the fuzzy set, and a fuzzy membership set is constructed to represent the degree of membership of the interference intensity level. Then, the fuzzy interference evaluation matrix can be constructed based on all the initial memberships and the fuzzy set in the following way: according to the general construction process of the fuzzy comprehensive evaluation model, the initial membership of each detection layer is used as the weight, and the fuzzy membership set of the corresponding layer is used as the evaluation index. The final fuzzy interference evaluation matrix is ​​obtained by combining the rows. Each row in the fuzzy interference evaluation matrix represents the membership distribution of a certain detection layer under different interference levels and has the input format of fuzzy comprehensive evaluation.Finally, based on the fuzzy interference evaluation matrix, the interference effect intensity of each detection layer is evaluated using fuzzy methods to obtain the interference evaluation index for water pollution detection corresponding to each detection layer. This can be achieved as follows: Based on the constructed fuzzy interference evaluation matrix, each row is extracted as the fuzzy evaluation vector for the corresponding detection layer. Each element in this vector represents the membership degree of the corresponding detection layer to different interference levels (e.g., weak, medium, strong). The fuzzy comprehensive evaluation method is used to normalize the evaluation vector and introduce a quantitative weight vector (e.g., corresponding to "weak = 0.2", "medium = 0.5", "strong = 0.8"). The fuzzy membership degree is multiplied by the corresponding weight item by item and summed to obtain the fuzzy expected value for each detection layer. This fuzzy expected value is then used as the interference evaluation index for that detection layer.

[0084] It should be noted that this application constructs an interference effect intensity matrix by combining the vertical concentration gradient with the local flow velocity of water bodies at different depths, and uses fuzzy membership functions to achieve fuzzy evaluation of interference in water pollution detection. This effectively solves the problem in traditional technologies that cannot accurately quantify and dynamically reflect the complex interference relationships between water layers at different depths. First, initializing the membership degree provides a reasonable starting point for subsequent fuzzy evaluation, improving the accuracy and stability of the evaluation. Second, using the vertical concentration gradient and flow velocity to construct the interference intensity matrix scientifically reflects the diffusion and disturbance mechanisms of pollutants at different depths, overcoming the limitations of single-index analysis. Third, by converting the interference intensity into a fuzzy set through fuzzy membership functions, the model's adaptability to uncertainty and data fluctuations is enhanced. Finally, a comprehensive evaluation is performed based on the fuzzy evaluation matrix to obtain the interference evaluation index of each detection layer, realizing a quantitative expression of the stratified interference effect of water pollution detection. In summary, the scheme of this application can improve the accuracy and reliability of water quality monitoring, optimize the fusion processing of water pollution data at different depths, and help to more scientifically guide pollution control and risk assessment.

[0085] In step 104, the water pollution concentration at each detection node is fused at a cost based on all mutual support and interference evaluation indices of water pollution detection at different depths to obtain the fused pollution degree of water pollution, and then the fused pollution degree is used as the pollution degree of the water body in the target underground drainage well.

[0086] In some embodiments, the water pollution concentration at each detection node is fused at a cost based on all mutual support and interference evaluation indices of water pollution detection at different depths to obtain the fused pollution degree of water pollution. This can be achieved through the following steps:

[0087] The fusion weights of water pollution concentrations at each detection node are used as optimization variables. A fusion cost function is constructed with the goal of minimizing the differences in mutual support between detection nodes and the deviation of the interference evaluation index.

[0088] All mutual support and interference evaluation indices of water pollution detection at different depths are introduced into the fusion cost function as weighted penalty terms;

[0089] The fusion cost function is iteratively solved using a genetic algorithm until the pollution fusion weight of each detection node converges to the optimal state, thereby obtaining the fusion pollution degree of water pollution in the target underground drainage well.

[0090] It should be noted that the fusion cost function in this application refers to the objective function used to quantify the differences in mutual support and interference evaluation bias between the water pollution concentrations of the detection nodes; the fusion pollution degree in this application is a comprehensive evaluation index for measuring the water pollution concentration of the target underground pipe drainage well.

[0091] In practical implementation, firstly, the fusion weights of water pollution concentrations at each detection node are used as optimization variables. The goal is to minimize the differences in mutual support between detection nodes and the deviation of the interference evaluation index. The fusion cost function can be constructed as follows: A cost function model is built using the concentration weights of each node as variables. The optimization objective consists of two parts: firstly, nodes with strong mutual support should have higher concentration consistency, i.e., a weighted sum of the square of the concentration difference between any two nodes and the product of their mutual support is constructed, reflecting the degree of matching between concentration distribution and node synergy; secondly, the response of each node's concentration value to the interference evaluation index of its detection layer, i.e., the lower the degree of interference, the higher the concentration consistency. For layer-by-layer measurements, the concentration values ​​should have higher reliability. It should also be noted that the fusion cost function in this application can be described as follows: This cost function consists of two main weighted sub-terms. The first term represents the deviation between the water pollution concentration and its corresponding mutual support among all detection nodes. That is, the higher the support of a node pair, the more its pollution concentration difference should be suppressed. This term is reflected by summing the squares of the concentration differences between all node pairs and their support values. The second term represents the deviation between the water pollution concentration of each detection node and the interference evaluation index of its depth layer. That is, the lower the interference index of a layer, the more its concentration value should be retained with confidence. This term is reflected by summing the squares of the concentration differences between each node and its interference index. The overall fusion cost function is composed of the sum of the squares of the weighted biases, and is represented by the sum of the two terms mentioned above. Two weighting factors are set to control the optimization intensity of the concentration consistency target and the interference suppression target, respectively, thereby achieving the globally optimal fusion estimate of the fusion pollution degree. Then, the fusion cost function can be introduced into the form of weighted penalty terms for all mutual support and interference evaluation indices of water pollution detection at different depths. This can be achieved by integrating the two parts into a unified cost function as weighted penalty terms, setting weighting factors for the support difference term and the interference index bias term to control their contribution to the optimization objective. Finally, the fusion cost function is optimized using a genetic algorithm. The fusion pollution degree of water pollution in the target underground drainage well can be obtained by iteratively solving the problem until the pollution fusion weight of each detection node converges to the optimal state. This can be achieved by using a genetic algorithm, specifically by initializing the concentration population, calculating the cost function value of each concentration combination as the fitness, iteratively generating a new population through selection, crossover, and mutation, and continuously optimizing the cost function value. Finally, under the condition of convergence within the preset number of iterations or error threshold, the concentration result with the minimum cost (i.e., the fusion result obtained by weighting and fusing the water pollution concentrations of all detection nodes through the optimized fusion weight) is output as the fusion pollution degree of water pollution in the target underground drainage well.

[0092] It should be noted that this application constructs a fusion cost function, incorporating the differences in mutual support between detection nodes and the deviation of the hierarchical interference evaluation index into a unified optimization objective. This avoids fusion bias caused by a single indicator, improving the comprehensiveness and accuracy of the fusion model. The scheme in this application uses a genetic algorithm for iterative solution, effectively overcoming the defect of traditional optimization methods that are prone to getting trapped in local optima, and ensuring the global optimality and stability of the fusion results. Overall, the scheme in this application solves the problem of information conflict and trade-offs in multi-node, multi-depth fusion, realizes the reasonable fusion of water pollution concentration data, improves the accuracy and reliability of fused data, and has strong technical innovation and application value.

[0093] On the other hand, in some embodiments, this application provides a water pollution monitoring instrument, which includes a water pollution detection unit, with reference to... Figure 4 The figure is a schematic diagram of the structure of a water pollution detection unit according to some embodiments of this application. The water pollution detection unit 400 includes: a detection module 401, a processing module 402, and an execution module 403, which are described below:

[0094] Detection module 401, in this application, is mainly used to detect water pollution data of each detection node through intelligent water quality sensor, and then extract the water pollution concentration of each detection node.

[0095] Processing module 402 in this application is used to determine the cross-correlation coefficients of water pollution concentrations between different detection nodes based on water pollution data of each detection node within a historical time period, and then determine the mutual support degree of water pollution detection results between each smart water quality sensor through all cross-correlation coefficients and the spatial topology of the smart water quality sensor.

[0096] In this application, the processing module 402 is also used to extract the vertical concentration gradient of water pollution in the target underground pipe drainage well, and to perform a fuzzy evaluation of the stratification interference effect intensity of water pollution detection at different depths based on the vertical concentration gradient and the local flow velocity of water bodies at different depths, so as to obtain the interference evaluation index of water pollution detection at different depths.

[0097] The execution module 403 in this application is mainly used to perform cost fusion of the water pollution concentration of each detection node based on all mutual support and interference evaluation index of water pollution detection at different depths, to obtain the fused pollution degree of water pollution, and then use the fused pollution degree as the pollution degree of the water body in the target underground pipe drainage well.

[0098] In addition, this application also provides a computer device, which includes a memory and a processor. The memory stores code, and the processor is configured to acquire the code and execute the above-described method for detecting water pollution in underground drainage wells.

[0099] In some embodiments, reference Figure 5 The figure is a schematic diagram of the structure of a computer device for implementing a method for detecting water pollution in a concealed pipe drainage well, according to some embodiments of this application. The water pollution detection method for concealed pipe drainage wells described in the above embodiments can... Figure 5 The computer device shown is used to implement this, and the computer device 500 includes at least one processor 501, a communication bus 502, a memory 503, and at least one communication interface 504.

[0100] Processor 501 can be a general-purpose central processing unit (CPU) or an application-specific integrated circuit (ASIC).

[0101] The communication bus 502 can be used to transmit information between the aforementioned components.

[0102] Memory 503 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disks or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto. Memory 503 may exist independently and be connected to processor 501 via communication bus 502. Memory 503 may also be integrated with processor 501.

[0103] The memory 503 stores program code for executing the solution of this application, and its execution is controlled by the processor 501. The processor 501 executes the program code stored in the memory 503. The program code may include one or more software modules. In the above embodiment, the water pollution detection method in the concealed pipe drainage well can be implemented by the processor 501 and one or more software modules in the program code in the memory 503.

[0104] Communication interface 504 uses any transceiver-like device to communicate with other devices or communication networks, such as Ethernet, radio access network (RAN), wireless local area networks (WLAN), etc.

[0105] In a specific implementation, as one example, a computer device may include multiple processors, each of which may be a single-core (single CPU) processor or a multi-core (multi CPU) processor. Here, a processor may refer to one or more devices, circuits, and / or processing cores used to process data (e.g., computer program instructions).

[0106] The aforementioned computer device can be a general-purpose computer device or a special-purpose computer device. In specific implementations, the computer device can be a desktop computer, a portable computer, a web server, a handheld digital assistant (PDA), a mobile phone, a tablet computer, a wireless terminal device, a communication device, or an embedded device. This application does not limit the type of computer device.

[0107] In addition, this application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for detecting water pollution in underground drainage wells.

[0108] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0109] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A method for detecting water pollution in concealed drainage wells, characterized in that, Includes the following steps: Multiple detection nodes and intelligent water quality sensors are deployed in the target underground drainage well. Water pollution data of each detection node is obtained by detecting the water pollution data of each detection node through the intelligent water quality sensors, and then the water pollution concentration of each detection node is extracted. Based on water pollution data from various monitoring nodes over a historical period, the cross-correlation coefficients of water pollution concentrations between different monitoring nodes are determined. Then, using all cross-correlation coefficients and the spatial topology of the smart water quality sensors, the mutual support degree of water pollution detection results between each smart water quality sensor is determined. Specifically, for every two smart water quality sensors, the confidence distance between them is determined based on their spatial topology. The consistency of water pollution detection results between the two smart water quality sensors is evaluated using all cross-correlation coefficients and the confidence distance to obtain the mutual support degree of water pollution detection results between the two smart water quality sensors, thus obtaining the mutual support degree of water pollution detection results between each pair of smart water quality sensors. The vertical concentration gradient of water pollution in the target underground drainage well is extracted. Based on the vertical concentration gradient and the local flow velocity of water at different depths, a fuzzy evaluation of the stratified interference effect intensity of water pollution detection at different depths is performed to obtain the interference evaluation index for water pollution detection at different depths. Specifically, this includes: initializing the initial membership degrees of the interference effect of different detection layers; constructing an interference effect intensity matrix between each detection layer based on the vertical concentration gradient and the local flow velocity of water at different depths; mapping the values ​​in the interference effect intensity matrix to fuzzy sets using fuzzy membership functions; constructing a fuzzy interference evaluation matrix based on all the initial membership degrees and the fuzzy sets; and performing a fuzzy evaluation of the interference effect intensity of each detection layer based on the fuzzy interference evaluation matrix to obtain the interference evaluation index for water pollution detection corresponding to each detection layer. Based on all mutual support and interference evaluation indices of water pollution detection at different depths, the water pollution concentration at each detection node is fused at a cost to obtain the fused pollution degree of water pollution. Specifically, this includes: using the fusion weight of water pollution concentration at each detection node as an optimization variable, constructing a fusion cost function with the goal of minimizing the difference in mutual support and the deviation of interference evaluation indices between detection nodes; introducing all mutual support and interference evaluation indices of water pollution detection at different depths into the fusion cost function in the form of weighted penalty terms; iteratively solving the fusion cost function using a genetic algorithm until the pollution fusion weight of each detection node converges to the optimal state, thereby obtaining the fused pollution degree of water pollution in the target underground drainage well, and then using the fused pollution degree as the pollution degree of the water body in the target underground drainage well.

2. The method as described in claim 1, characterized in that, Water pollution data from various detection nodes is obtained through intelligent water quality sensors, and the water pollution concentration at each detection node is then extracted, specifically including: For each detection node, water pollution data is collected in real time using intelligent water quality sensors. The collected water pollution data is smoothed to remove outliers and noise interference, resulting in smoothed water pollution data. The water pollution concentration of each detection node is extracted from the smoothed water pollution data by using a preset pollutant concentration calculation model.

3. The method as described in claim 1, characterized in that, The cross-correlation coefficients between different monitoring nodes are determined based on water pollution data from various monitoring nodes over a historical period. Specifically, this includes: For every two detection nodes, the water pollution data of the two detection nodes within the historical time period are time-aligned to obtain a time-aligned water pollution data set. Based on the time-aligned water pollution data set, cross-correlation analysis was performed on the water pollution concentration between two detection nodes to obtain the cross-correlation coefficient of water pollution concentration between the two detection nodes, and then the cross-correlation coefficient of water pollution concentration between each pair of detection nodes was obtained.

4. The method as described in claim 1, characterized in that, Extracting the vertical concentration gradient of water pollution in the target underground drainage well specifically includes: The detection node is divided into at least three detection layers in the vertical direction to obtain the pollutant concentration of each layer. Based on the concentration difference between adjacent detection layers, cubic spline interpolation is performed on the water pollution concentration in the vertical direction to generate a continuous vertical concentration distribution curve. The first derivative of the vertical concentration distribution curve within a preset depth interval is determined, and the maximum gradient value is extracted as the vertical concentration gradient of water pollution in the target underground drainage well.

5. The method as described in claim 1, characterized in that, The intelligent water quality sensor refers to an integrated sensing device that has multi-parameter water quality sensing capabilities, data preprocessing functions, and a communication module.

6. A water pollution monitoring instrument, comprising a water pollution detection unit, characterized in that, The water pollution detection unit includes: The detection module is used to detect water pollution data at each detection node through intelligent water quality sensors, and then extract the water pollution concentration at each detection node. The processing module is used to determine the cross-correlation coefficients of water pollution concentrations between different detection nodes based on water pollution data from each detection node within a historical time period. Then, it determines the mutual support degree of water pollution detection results between each smart water quality sensor using all cross-correlation coefficients and the spatial topology of the smart water quality sensors. Specifically, this includes: for every two smart water quality sensors, determining the confidence distance between them based on their spatial topology; evaluating the consistency of water pollution detection results between the two smart water quality sensors using all cross-correlation coefficients and the confidence distance to obtain the mutual support degree of water pollution detection results between the two smart water quality sensors, and thus obtaining the mutual support degree of water pollution detection results between each pair of smart water quality sensors. The processing module is further configured to extract the vertical concentration gradient of water pollution in the target underground drainage well, and to perform a fuzzy evaluation of the stratified interference effect intensity of water pollution detection at different depths based on the vertical concentration gradient and the local flow velocity of water bodies at different depths, thereby obtaining an interference evaluation index for water pollution detection at different depths. Specifically, this includes: initializing the initial membership degrees of the interference effect of different detection layers; constructing an interference effect intensity matrix between each detection layer based on the vertical concentration gradient and the local flow velocity of water bodies at different depths; mapping the values ​​in the interference effect intensity matrix to fuzzy sets using fuzzy membership functions; constructing a fuzzy interference evaluation matrix based on all the initial membership degrees and the fuzzy sets; and performing a fuzzy evaluation of the interference effect intensity of each detection layer based on the fuzzy interference evaluation matrix, thereby obtaining an interference evaluation index for water pollution detection corresponding to each detection layer. The execution module is used to perform cost fusion of water pollution concentration at each detection node based on all mutual support and interference evaluation indices of water pollution detection at different depths, to obtain the fused pollution degree of water pollution. Specifically, it includes: using the fusion weight of water pollution concentration at each detection node as an optimization variable, constructing a fusion cost function with the goal of minimizing the difference in mutual support and the deviation of interference evaluation indices between detection nodes; introducing all mutual support and interference evaluation indices of water pollution detection at different depths into the fusion cost function in the form of weighted penalty terms; iteratively solving the fusion cost function through a genetic algorithm until the pollution fusion weight of each detection node converges to the optimal state, thereby obtaining the fused pollution degree of water pollution in the target underground pipe drainage well, and then using the fused pollution degree as the pollution degree of the water body in the target underground pipe drainage well.

7. A computer device comprising a memory and a processor, the memory storing code, characterized in that, The processor is configured to acquire the code and execute the water pollution detection method in a concealed drainage well as described in any one of claims 1 to 5.

8. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the water pollution detection method in the underground drainage well as described in any one of claims 1 to 5.

Images