Water environment pollution monitoring method and system based on data analysis

Abstract

The invention discloses a water environment pollution monitoring method and system based on data analysis, and belongs to the technical field of pollution monitoring. The system comprises an environment monitoring module, an influence analysis module, an abnormity identification module and a visualization and storage module. The environment monitoring module collects a GIS map and historical records and collects monitoring indexes and hydrological indexes of a river water body through monitoring equipment. The influence analysis module marks the position of each monitoring device on the GIS map, analyzes the upstream and downstream relationship and sets a relative time difference; establishing an influence relation chain for each monitoring index, and calculating an influence coefficient between the monitoring devices; the anomaly recognition module takes the monitoring indexes currently collected by the monitoring devices as measured values, calculates predicted values according to the influence relation chain, calculates anomaly indexes of the monitoring devices in combination with influence coefficients and divides anomaly regions; and the visualization and storage module displays the position of each abnormal area, prompts a worker to check each abnormal area one by one according to an abnormal index in an inverted order, and stores the collected index data into a historical record.

Description

Technical Field

[0001] This invention relates to the field of pollution monitoring technology, specifically to a water environment pollution monitoring method and system based on data analysis. Background Technology

[0002] With the increasing global demand for water resource protection, continuous and accurate monitoring of the aquatic environment has become a core task of environmental management and ecological security. Currently, the industry generally relies on widely deployed automated sensing terminals and spatial information platforms to build the ability to collect and visualize key water quality parameters in real time, forming the basic technological paradigm for water environment monitoring.

[0003] Currently, existing technological paradigms still have certain limitations. On the one hand, existing technologies typically treat discretely distributed monitoring nodes as independent units for isolated threshold judgments, completely ignoring the inherent spatiotemporal correlations and physical transmission processes of the water environment as a dynamically interconnected organic whole. This leads to the system's inability to distinguish between parameter changes caused by normal upstream transport and local sudden anomalies, generating a large number of invalid alarms. On the other hand, existing methods also fail to effectively identify and isolate external non-local inputs such as water replenishment from inlets during rainfall and rainfall interference, resulting in continuous disturbances to the monitoring node data. Such external inputs can significantly alter the water quality characteristics of local water bodies, making the monitoring data unable to accurately reflect the actual pollution status of the target river section, potentially causing pollution events to be masked or clean water bodies to be misjudged, thus leading to systematic distortion of monitoring conclusions. Due to the lack of quantitative modeling of historical correlation patterns between nodes and the ability to intelligently analyze the sources of data anomalies, pollution source tracing work relies heavily on human experience, resulting in low investigation efficiency and a lack of the ability to self-evolve and optimize decision-making models through continuous operation, making it difficult to adapt to long-term dynamic changes in the environment. Summary of the Invention

[0004] The purpose of this invention is to provide a water environment pollution monitoring method and system based on data analysis to solve the problems mentioned in the background art.

[0005] To address the aforementioned technical problems, this invention provides a water environment pollution monitoring method based on data analysis, comprising:

[0006] S100: Collect GIS maps and historical records, and collect monitoring indicators and hydrological indicators of river water bodies through monitoring equipment.

[0007] GIS maps are used to describe the topographical structure and distribution of rivers and waterways within a specified area.

[0008] The designated area refers to the spatial range that is pre-delineated based on the boundaries of the hydrological units of the river basin and the needs of administrative management.

[0009] Historical records include various monitoring indicators collected by each monitoring device at different historical moments.

[0010] Monitoring equipment refers to specialized equipment used to collect river water samples and perform automated analysis, obtaining various monitoring indicators and hydrological indicators based on the analysis results.

[0011] Monitoring indicators are parameters used to quantify the pollution characteristics of river water bodies. Hydrological indicators are physical quantities that describe the dynamic characteristics and transport capacity of water bodies.

[0012] In the field of water environment, monitoring indicators are the core basis for judging water quality, pollution level and ecological health.

[0013] By constructing a multi-dimensional data foundation for monitoring and integrating GIS maps, abstract monitoring data is anchored to specific spatial geographic environments, thus realizing the spatialization of the monitoring network.

[0014] At the same time, collecting and storing historical records introduces a time dimension to the system, enabling subsequent analysis to be based on historical patterns.

[0015] This has changed the situation of isolated data and weak spatiotemporal correlation in traditional monitoring, and made it possible to establish a dynamic and traceable watershed pollution analysis model.

[0016] S200. Mark the location of each monitoring device on the GIS map, analyze the upstream and downstream relationships, and set relative time differences. Establish influence chains for each monitoring indicator based on historical records, thereby calculating the influence coefficient between every two monitoring devices. Specifically, this includes:

[0017] S201. Mark the location of each monitoring device on the GIS map and analyze the flow velocity and direction in the hydrological indicators collected by the monitoring devices. Based on the flow direction and relative position of each monitoring device, analyze the upstream and downstream relationships between the monitoring devices.

[0018] S202, monitoring equipment located upstream As downstream monitoring equipment The parent node, As The child nodes. Each monitoring device is treated as a node, and a flow relationship tree diagram is established based on the parent-child node relationship.

[0019] S203. Based on the parent-child node relationship between each monitoring device in the flow direction relationship tree diagram, construct a smooth flow velocity evolution curve according to the flow velocity data collected by each pair of parent-child nodes, and analyze the relative position of each pair of parent-child nodes in the GIS map.

[0020] S204. Based on the relative position, plan the shortest flow path for the corresponding parent and child nodes, analyze the propagation time difference of the flow velocity evolution curve within the shortest flow path under the parent and child nodes, and use it as the relative time difference of the monitoring equipment corresponding to the parent and child nodes.

[0021] The relative time difference is defined as the time difference between parent and child nodes monitoring the same hydrological event, reflecting the actual propagation time of water flow between nodes. Its setting is constrained by the topological relationship of the tree diagram to avoid erroneous associations across levels or between non-parent and child nodes.

[0022] The calculation of relative time difference is not simply the geographical distance divided by the average flow velocity, but is based on the topological constraints of parent and child nodes and the shortest flow path. This effectively avoids misalignment caused by complex terrain such as meandering rivers and tributary confluence, making time alignment more consistent with hydrodynamic realities.

[0023] S205. Analyze the flow velocity of each monitoring device in the historical records and calculate the relative time difference, then align the time series according to the relative time difference. Based on the flow direction relationship tree diagram, establish the influence relationship formula and influence relationship chain for each monitoring indicator. Specifically, this includes:

[0024] S2051. Starting from the node without a parent node in the flow relationship tree diagram and ending at the node without a child node, count the number of nodes. Starting from the origin and ending at the destination, the flow relationship tree diagram is broken down into... A separate flow relationship chain.

[0025] Disassembly does not modify the original flow relationship tree diagram, but rather regenerates multiple flow relationship chains without changing the original flow relationship tree diagram.

[0026] S2052, Settings For each historical moment, the flow rate of each monitoring device at that historical moment is retrieved from the historical records. The relative time difference is calculated based on the flow rates of each monitoring device at the same historical moment and mapped to each flow direction relationship chain.

[0027] S2053. Based on the relative time difference between nodes in the flow relationship chain, align the time sequence of collected monitoring indicators. Establish the influence relationship formula for each monitoring indicator between each node in each flow relationship chain, and then calculate the difference coefficient. Specifically, this includes:

[0028] S2053-1, Statistical Flow Relationship Chain Total number of nodes By pairwise association of these nodes, a total of There are node pairs. Among them, .

[0029] S2053-2, Based on the nodes at each historical moment Align time series with relative time difference, and obtain node pairs after alignment time series at the same historical time. Collected monitoring indicators They were used as independent and dependent variables, respectively, and packaged into a sample.

[0030] S2053-3, Setting the intercept and regression coefficients And establish a linear regression model. This The independent variables in each sample are used as input values. The output value of each sample The difference between the dependent variable and the dependent variable is taken as the gap value. The expression is as follows:

[0031] ;

[0032] S2053-4. The sum of the differences among all samples is used as the difference coefficient. The minimum difference coefficient is obtained by adjusting the intercept and regression coefficient. After determining the values ​​of the intercept and regression coefficients, the monitoring indicators are obtained. Next node pair The influence relationship.

[0033] S2053-5, and so on, establish monitoring indicators for each node in each flow relationship chain. And the influence relationships of other monitoring indicators, and classify these influence relationships according to the different monitoring indicators.

[0034] The relative time difference between nodes needs to be obtained by summing the relative time differences of all parent and child nodes contained in the flow relationship chain.

[0035] When there are several other nodes between node A and node B in the flow relationship chain, the relative time differences of these nodes are summed to obtain the relative time difference between node A and node B.

[0036] S2054, Based on monitoring indicators between nodes The difference coefficient is used to set a midpoint in each flow relationship chain, and influence relationship chains are then defined. A midpoint is set for each monitoring indicator in each flow relationship chain, and influence relationship chains are defined accordingly. Specifically, this includes:

[0037] S2054-1, Setting the fluctuation coefficient Analyze the flow relationship chain Middle starting point with its child nodes Monitoring indicators between Difference coefficient According to the formula: Calculate reference coefficient .

[0038] S2054-2, Analyzing the flow relationship chain The difference coefficient between every two nodes is used to calculate the flow relationship chain sequentially from the starting point to the ending point. internal nodes Then each node Gap Index :

[0039] ;

[0040] In the formula, To flow to the relationship chain Middle node The total number of all previous nodes, For nodes Compared to the previous one The difference coefficient between nodes.

[0041] The gap index is used to assess the overall attenuation of the correlation between a node along the water flow direction and all its upstream historical nodes under a specific monitoring indicator.

[0042] A comprehensive gap index is obtained by calculating the arithmetic mean of the gap coefficients between this node and each upstream node. This index reflects the cumulative change intensity of the historical correlation pattern of this monitoring indicator from the start of the process to the current node.

[0043] When the gap index exceeds the dynamic threshold set based on the basic gap coefficient between the starting point and its immediate downstream node, it is determined that the correlation has significantly weakened, thus providing a quantitative criterion for automatically dividing independent influence relationship chains.

[0044] S2054-3, When node Gap Index Greater than the reference coefficient When selecting a node The previous node is used as the midpoint. The difference index among all nodes. None are greater than the reference coefficient When choosing a node, select the node corresponding to the endpoint as the midpoint.

[0045] S2054-4, From Flow to Relationship Chain The starting point and the midpoint are used to divide the monitoring indicators. The influence relationship chain. And so on, monitoring indicators are set for each flow relationship chain. Set the midpoint and divide the influence relationship chain.

[0046] The midpoint is dynamically determined by the gap coefficient and the fluctuation coefficient. In essence, it is to automatically identify and cut out the monitoring segments with strong correlation based on the degree of decay of the correlation between indicators in historical data.

[0047] It can adapt to the characteristics of different river sections, improving the precision and accuracy of calculation and analysis.

[0048] S206. Analyze the gap coefficients between nodes in each influence chain, and the monitoring equipment corresponding to each node. Based on the influence chain of each monitoring equipment, analyze the influence coefficients of various monitoring indicators between the monitoring equipment. Specifically, this includes:

[0049] S2061. Obtaining Monitoring Indicators The entire influence relationship chain, marked nodes and Simultaneous influence chains. Analyze and label monitoring indicators between nodes in each influence chain. The difference coefficient.

[0050] S2062, Statistical markers affect the number of relationship chains Get node pairs The difference coefficient in each label-affected relation chain is substituted into the formula to calculate the node pair. Monitoring indicators under two monitoring devices Influence coefficient :

[0051] ;

[0052] In the formula, For node pairs In the The bar markers affect the gap coefficient in the relation chain. For the first The bar marker affects the maximum difference coefficient between nodes in the relation chain.

[0053] The influence coefficient is used to quantify the strength of the influence correlation between two specific monitoring devices on a certain monitoring indicator. It is used to comprehensively consider the performance of the corresponding node pair of the monitoring device in all influence relationship chains that include them.

[0054] The correlation contribution of a single chain is obtained by calculating the relative ratio of the difference coefficient in each labeled relation chain to the maximum difference coefficient in that chain, and then calculating the difference between the ratio and a fixed value.

[0055] The arithmetic mean of the correlations of all relevant relationship chains is used to derive a normalized comprehensive influence coefficient. The closer the node pair is to a fixed value, the more stable the historical influence pattern of the two monitoring devices on a certain monitoring indicator and the stronger the predictability.

[0056] S2063. Obtain the complete influence relationship chain for each monitoring indicator and analyze the difference coefficients of different nodes on each monitoring indicator. Based on the correspondence between nodes and monitoring equipment, calculate the influence coefficients of each monitoring indicator between monitoring equipment.

[0057] This method deeply integrates physical hydrological processes with statistical data models. By establishing a flow direction relationship tree diagram and calculating relative time differences, it simulates the actual convection and diffusion processes of pollutants in water bodies, transforming static spatial location relationships into a dynamic, time-series dependent causal relationship network.

[0058] The final calculated impact coefficient quantifies the statistical correlation strength and direction between any two monitoring points on various water quality indicators, providing a scientific and quantifiable reference benchmark for anomaly judgment.

[0059] S300. Obtain the currently collected monitoring indicators from each monitoring device as measured values, and calculate the predicted values ​​based on the influence relationship chain. Analyze the difference between the measured values ​​and the predicted values, and calculate the anomaly index for each monitoring device based on the influence coefficient, and delineate the anomaly zone.

[0060] Specifically, it includes:

[0061] S301. Obtain the current monitoring indicators collected by each monitoring device and use them as measured values, then filter out those indicators that contain data from the monitoring devices. The complete influence chain of the corresponding node. Analyze the relationship between each node in each influence chain and the monitoring equipment. The relationship between the influence of various monitoring indicators between corresponding nodes.

[0062] S302. Substitute the measured values ​​of each node into the influence relationship formula to obtain the monitoring equipment. The different predicted values ​​for each monitoring indicator are analyzed. Monitoring equipment is also analyzed. The differences between the predicted and measured values ​​of various monitoring indicators are used to calculate the monitoring equipment's performance in conjunction with the corresponding influence coefficients. Abnormal index :

[0063] ;

[0064] In the formula, For the number of all monitored indicators, For the first In the entire impact chain of the monitoring indicators, excluding monitoring equipment The number of duplicates after removing all other nodes.

[0065] For the first The first monitoring indicator Each node corresponds to a monitoring device and a monitoring equipment. The influence coefficient between them.

[0066] For the first The first monitoring indicator Substituting each node into the influence relationship formula, the monitoring equipment obtained The predicted value. For monitoring equipment Next The measured values ​​of the monitoring indicators.

[0067] The anomaly index is used to calculate the overall anomaly level of a single monitoring device. Its calculation process involves several levels:

[0068] First, for each monitoring indicator, the relative deviation between the measured value of the device and the multiple predicted values ​​obtained based on each associated node and the corresponding influence relationship is calculated.

[0069] Secondly, the relative deviations are weighted using the influence coefficients between the device and each associated node to reflect the differences in the reliability of prediction results of different associated nodes. The average value of all weighted relative deviations under the indicator is then calculated to obtain the abnormal contribution value of the indicator.

[0070] Finally, the arithmetic mean of the abnormal contribution values ​​of all monitoring indicators is calculated again to obtain the comprehensive abnormality index of the equipment. This index comprehensively reflects the degree of deviation of the equipment from the overall monitoring network across all monitoring indicators.

[0071] The calculation of the anomaly index of each monitoring device requires using the time corresponding to the measured value as the base time, analyzing the relative time difference of all other monitoring devices in the same influence chain, and adding the maximum relative time difference to the base time to obtain the cutoff time. Only when the time exceeds the cutoff time can the anomaly index of the monitoring device be calculated completely.

[0072] S303. Calculate the anomaly index for each monitoring device, and classify monitoring devices with an anomaly index greater than the threshold as abnormal devices. Preset base distance. Based on the abnormality index of abnormal equipment Calculate reference distance On the GIS map, using the location of each abnormal device as the center, and referring to the distance... Circular anomaly regions are divided according to their radii.

[0073] Reference distance The calculation formula is as follows:

[0074] ;

[0075] In the formula, The preset attenuation index, This represents the maximum anomaly index among all abnormal devices.

[0076] The reference distance is used to non-linearly map the anomaly index of the monitoring equipment to the radius of the suspicious area in geospatial space to guide the scope of on-site investigation.

[0077] By using a saturation growth function, the ratio of the device's anomaly index to the maximum anomaly index among all devices is converted into a reference distance value between zero and a preset base distance.

[0078] When the anomaly index is low, the radius increases gradually, making it easier to focus on key locations for detailed investigation. As the anomaly index approaches its maximum value, the radius gradually approaches its theoretical maximum value to cover a wider area of ​​potential pollution impact.

[0079] It enables adaptive adjustment of the investigation scope based on the severity of the anomaly.

[0080] By comparing real-time measured values ​​with predicted values ​​calculated based on historical correlation models, it is possible to effectively distinguish between "changes caused by normal upstream propagation" and "real anomalies that suddenly occur locally".

[0081] By combining the impact coefficient with a weighted average to calculate the anomaly index, the judgment result takes into account not only the magnitude of the deviation but also the credibility of the source of the deviation.

[0082] Finally, anomaly zones are dynamically generated based on the anomaly index, transforming the abstract numerical index into intuitive spatial action instructions with priority and scope guidance.

[0083] The S400 monitoring center displays the specific location of each anomaly area on the GIS map through a visual screen, prompting staff to check each anomaly area one by one in reverse order of the anomaly index, and storing the collected monitoring indicators and hydrological indicators into the historical record.

[0084] The monitoring loop has been completed and the system has been given learning capabilities. The visualization display presents complex analysis results intuitively on the GIS map, greatly improving the situational awareness and decision-making efficiency of managers.

[0085] The prompt to sort by abnormal index in reverse order enables intelligent sorting of handling strategies.

[0086] Storing this data in the historical records will enrich the historical database with the experience of each monitoring, analysis and handling, which will be used to update and optimize the future impact relationship chain and impact coefficient, so that the prediction and anomaly detection capabilities can be continuously improved over time.

[0087] The present invention also provides a water environment pollution monitoring system based on data analysis, including an environmental monitoring module, an impact analysis module, an anomaly identification module, and a visualization and storage module.

[0088] The environmental monitoring module collects GIS maps and historical records, and collects monitoring indicators and hydrological indicators of river water bodies through monitoring equipment.

[0089] Collect multi-source data, including GIS maps describing the river topography and distribution, historical records of monitoring indicators and hydrological indicators collected by various monitoring devices, and collect water body indicator data in real time through dedicated monitoring devices deployed in the river.

[0090] The spatial, temporal, and real-time data foundation required to build the system combines geographic information, historical patterns, and field measurements, providing complete and dynamic data input for subsequent analysis and ensuring the spatiotemporal continuity and data traceability of the monitoring work.

[0091] The impact analysis module marks the location of each monitoring device on the GIS map, analyzes upstream and downstream relationships, and sets relative time differences. It establishes impact relationship chains for various monitoring indicators and calculates the impact coefficients between monitoring devices.

[0092] First, mark the location of the equipment on the GIS map, and establish the upstream and downstream relationships between the monitoring equipment based on the flow direction and relative position to form a flow relationship tree diagram.

[0093] Then, the relative time difference between parent and child nodes is calculated based on flow rate and path, and the time series of historical data is aligned according to this time difference.

[0094] Finally, by using linear regression, quantitative influence relationships and influence chains are established for each monitoring indicator across different equipment nodes, and the influence coefficients between any two monitoring devices for each monitoring indicator are calculated.

[0095] By deeply integrating and mathematically modeling the spatial topology, hydrodynamic characteristics, and historical water quality change patterns of river systems, the degree of mutual influence and propagation delay between monitoring points were quantified, and a predictive benchmark model based on spatial correlation was established for anomaly detection.

[0096] The anomaly identification module uses the current monitoring indicators collected by each monitoring device as measured values, calculates predicted values ​​based on the influence relationship chain, and calculates the anomaly index of each monitoring device and divides the anomaly area based on the influence coefficient.

[0097] Obtain the current measured values ​​of each monitoring device, and use the established influence relationship chain and corresponding influence formula of each monitoring indicator to substitute the measured values ​​of other nodes in the chain to calculate the different predicted values ​​of multiple monitoring indicators of the target device.

[0098] By comparing the deviations between the measured values ​​and the predicted values, and by weighting the results with the influence coefficients between the devices, the anomaly index of each monitoring device is calculated.

[0099] Based on the magnitude of the anomaly index, and using the device location as the center and a dynamically calculated radius, different anomaly zones are divided on the GIS map.

[0100] It has enabled a shift from single-point data judgment to collaborative diagnosis based on spatial correlation networks, which can more accurately locate real pollution anomalies. Through anomaly indices and dynamic regional division, it can rank suspected pollution areas by severity and predict their spatial extent, greatly improving the targeting and efficiency of the investigation.

[0101] The visualization and storage module displays the location of each abnormal area, prompts staff to check each abnormal area one by one in reverse order of the abnormal index, and stores the collected index data into the historical record.

[0102] The results output by the anomaly identification module, namely the location and extent of each anomaly area, are overlaid and displayed on the visualization screen of the GIS map. The system will prompt staff to conduct on-site investigations one by one according to the anomaly index from high to low. At the same time, all monitoring indicators and hydrological indicators collected in this round are automatically saved to the historical record.

[0103] The visual interface provides managers with an intuitive and comprehensive understanding of the pollution situation and decision support, while prioritizing investigation paths optimizes human resource allocation. Data storage enables the continuous accumulation of historical records, providing data support for subsequent iterative updates to impact relationship chains and impact coefficients, thus giving the entire monitoring system the ability to continuously learn and self-optimize.

[0104] Compared with the prior art, the beneficial effects achieved by the present invention are:

[0105] From isolated threshold judgment to spatiotemporal correlation diagnosis: This solution abandons the traditional model of treating monitoring points as independent units for isolated threshold alarms. By constructing an influence relationship network based on water flow topology and historical data, it achieves collaborative analysis and anomaly diagnosis of monitoring data. It can effectively distinguish between parameter changes caused by normal upstream pollution plume transport and genuine local anomalies, thereby significantly reducing false alarms and missed alarms and improving the accuracy and reliability of alarms.

[0106] From manual experience-based source tracing to quantitative relationship-based source tracing: This solution uses mathematical modeling to integrate the historical correlations between monitoring points with hydrological processes, generating quantitative impact coefficients and impact relationship chains. When anomalies occur, the system can automatically assess the correlation contribution of each upstream and downstream node, providing clear scientific evidence and investigation directions for pollution source tracing, greatly reducing the blind spots of manual investigation and significantly improving emergency response efficiency.

[0107] From static alarm status to dynamic decision support: This solution deeply integrates the calculated anomaly index with the geographic information system, not only displaying the locations of anomalies but also dynamically generating and visualizing the scope and investigation priority of suspicious areas based on the severity of the anomalies. This provides on-site command with intuitive and actionable spatial decision support, achieving a leap from "informing where the standards are exceeded" to "guiding how to investigate," and optimizing the scheduling and allocation of human resources.

[0108] From a fixed and rigid system to adaptive learning: This solution designs a complete data closed loop, feeding the results of each monitoring and response as new samples back into the historical record. It can continuously iterate and update its core impact relationship model using new data, enabling diagnostic and source tracing capabilities to continuously optimize and evolve as the system operates over time. This adapts to the long-term dynamic changes in watershed hydrological conditions and pollution characteristics, overcoming the shortcomings of traditional system models that are rigid and degrade in performance. Attached Figure Description

[0109] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0110] Figure 1 This is a flowchart illustrating the water environment pollution monitoring method based on data analysis according to the present invention.

[0111] Figure 2 This is a schematic diagram of the water environment pollution monitoring system based on data analysis according to the present invention. Detailed Implementation

[0112] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0113] Example 1: Please refer to Figure 1 This invention provides a water environment pollution monitoring method based on data analysis, comprising:

[0114] S100: Collect GIS maps and historical records, and collect monitoring indicators and hydrological indicators of river water bodies through monitoring equipment.

[0115] GIS maps are used to describe the topographical structure and distribution of rivers and waterways within a specified area.

[0116] The designated area refers to the spatial range that is pre-delineated based on the boundaries of the hydrological units of the river basin and the needs of administrative management.

[0117] Historical records include various monitoring indicators collected by each monitoring device at different historical moments.

[0118] In the specific implementation process, monitoring equipment refers to specialized equipment used to collect river water samples and perform automated analysis, and obtain various monitoring indicators and hydrological indicators based on the analysis results.

[0119] Monitoring indicators are parameters used to quantify the pollution characteristics of river water bodies. Hydrological indicators are physical quantities that describe the dynamic characteristics and transport capacity of water bodies.

[0120] In the field of water environment, monitoring indicators are the core basis for judging water quality, pollution level and ecological health.

[0121] In the specific implementation process, a multi-dimensional data foundation for monitoring is constructed. By integrating GIS maps, abstract monitoring data is anchored to specific spatial geographic environments, thus realizing the spatialization of the monitoring network.

[0122] At the same time, collecting and storing historical records introduces a time dimension to the system, enabling subsequent analysis to be based on historical patterns.

[0123] This has changed the situation of isolated data and weak spatiotemporal correlation in traditional monitoring, and made it possible to establish a dynamic and traceable watershed pollution analysis model.

[0124] S200. Mark the location of each monitoring device on the GIS map, analyze the upstream and downstream relationships, and set relative time differences. Establish influence chains for each monitoring indicator based on historical records, thereby calculating the influence coefficient between every two monitoring devices. Specifically, this includes:

[0125] S201. Mark the location of each monitoring device on the GIS map and analyze the flow velocity and direction in the hydrological indicators collected by the monitoring devices. Based on the flow direction and relative position of each monitoring device, analyze the upstream and downstream relationships between the monitoring devices.

[0126] S202, monitoring equipment located upstream As downstream monitoring equipment The parent node, As The child nodes. Each monitoring device is treated as a node, and a flow relationship tree diagram is established based on the parent-child node relationship.

[0127] S203. Based on the parent-child node relationship between each monitoring device in the flow direction relationship tree diagram, construct a smooth flow velocity evolution curve according to the flow velocity data collected by each pair of parent-child nodes, and analyze the relative position of each pair of parent-child nodes in the GIS map.

[0128] S204. Based on the relative position, plan the shortest flow path for the corresponding parent and child nodes, analyze the propagation time difference of the flow velocity evolution curve within the shortest flow path under the parent and child nodes, and use it as the relative time difference of the monitoring equipment corresponding to the parent and child nodes.

[0129] In the specific implementation process, the relative time difference is defined as the time difference between parent and child nodes monitoring the same hydrological event, reflecting the actual propagation time of water flow between nodes. Its setting is constrained by the topological relationship of the tree diagram to avoid erroneous association across levels or between non-parent and child nodes.

[0130] The calculation of relative time difference is not simply the geographical distance divided by the average flow velocity, but is based on the topological constraints of parent and child nodes and the shortest flow path. This effectively avoids misalignment caused by complex terrain such as meandering rivers and tributary confluence, making time alignment more consistent with hydrodynamic realities.

[0131] S205. Analyze the flow velocity of each monitoring device in the historical records and calculate the relative time difference, then align the time series according to the relative time difference. Based on the flow direction relationship tree diagram, establish the influence relationship formula and influence relationship chain for each monitoring indicator. Specifically, this includes:

[0132] S2051. Starting from the node without a parent node in the flow relationship tree diagram and ending at the node without a child node, count the number of nodes. Starting from the origin and ending at the destination, the flow relationship tree diagram is broken down into... A separate flow relationship chain.

[0133] In practice, the dismantling does not involve modifying the original flow relationship tree diagram, but rather regenerating multiple flow relationship chains without changing the original flow relationship tree diagram.

[0134] S2052, Settings For each historical moment, the flow rate of each monitoring device at that historical moment is retrieved from the historical records. The relative time difference is calculated based on the flow rates of each monitoring device at the same historical moment and mapped to each flow direction relationship chain.

[0135] S2053. Based on the relative time difference between nodes in the flow relationship chain, align the time sequence of collected monitoring indicators. Establish the influence relationship formula for each monitoring indicator between each node in each flow relationship chain, and then calculate the difference coefficient. Specifically, this includes:

[0136] S2053-1, Statistical Flow Relationship Chain Total number of nodes By pairwise association of these nodes, a total of There are node pairs. Among them, .

[0137] S2053-2, Based on the nodes at each historical moment Align time series with relative time difference, and obtain node pairs after alignment time series at the same historical time. Collected monitoring indicators They were used as independent and dependent variables, respectively, and packaged into a sample.

[0138] S2053-3, Setting the intercept and regression coefficients And establish a linear regression model. This The independent variables in each sample are used as input values. The output value of each sample The difference between the dependent variable and the dependent variable is taken as the gap value. The expression is as follows:

[0139] ;

[0140] S2053-4. The sum of the differences among all samples is used as the difference coefficient. The minimum difference coefficient is obtained by adjusting the intercept and regression coefficient. After determining the values ​​of the intercept and regression coefficients, the monitoring indicators are obtained. Next node pair The influence relationship.

[0141] S2053-5, and so on, establish monitoring indicators for each node in each flow relationship chain. And the influence relationships of other monitoring indicators, and classify these influence relationships according to the different monitoring indicators.

[0142] In the actual implementation process, the relative time difference between nodes needs to be obtained by summing the relative time differences of all parent and child nodes contained in the relationship chain.

[0143] When there are several other nodes between node A and node B in the flow relationship chain, the relative time differences of these nodes are summed to obtain the relative time difference between node A and node B.

[0144] S2054, Based on monitoring indicators between nodes The difference coefficient is used to set a midpoint in each flow relationship chain, and influence relationship chains are then defined. A midpoint is set for each monitoring indicator in each flow relationship chain, and influence relationship chains are defined accordingly. Specifically, this includes:

[0145] S2054-1, Setting the fluctuation coefficient Analyze the flow relationship chain Middle starting point with its child nodes Monitoring indicators between Difference coefficient According to the formula: Calculate reference coefficient .

[0146] S2054-2, Analyzing the flow relationship chain The difference coefficient between every two nodes is used to calculate the flow relationship chain sequentially from the starting point to the ending point. internal nodes Then each node Gap Index :

[0147] ;

[0148] In the formula, To flow to the relationship chain Middle node The total number of all previous nodes, For nodes Compared to the previous one The difference coefficient between nodes.

[0149] In practice, the gap index is used to assess the overall attenuation of the correlation between a node along the water flow direction and all its upstream historical nodes under a specific monitoring indicator.

[0150] A comprehensive gap index is obtained by calculating the arithmetic mean of the gap coefficients between this node and each upstream node. This index reflects the cumulative change intensity of the historical correlation pattern of this monitoring indicator from the start of the process to the current node.

[0151] When the gap index exceeds the dynamic threshold set based on the basic gap coefficient between the starting point and its immediate downstream node, it is determined that the correlation has significantly weakened, thus providing a quantitative criterion for automatically dividing independent influence relationship chains.

[0152] S2054-3, When node Gap Index Greater than the reference coefficient When selecting a node The previous node is used as the midpoint. The difference index among all nodes. None are greater than the reference coefficient When choosing a node, select the node corresponding to the endpoint as the midpoint.

[0153] S2054-4, From Flow to Relationship Chain The starting point and the midpoint are used to divide the monitoring indicators. The influence relationship chain. And so on, monitoring indicators are set for each flow relationship chain. Set the midpoint and divide the influence relationship chain.

[0154] In the specific implementation process, the midpoint is dynamically determined by the gap coefficient and the fluctuation coefficient. In essence, it is based on the degree of decay of the correlation between indicators in historical data to automatically identify and cut out the monitoring segments with strong correlation.

[0155] It can adapt to the characteristics of different river sections (such as the large proportion of water inflow and outflow between upstream and downstream river channels, strong connection and high correlation), thus improving the precision and accuracy of calculation and analysis.

[0156] S206. Analyze the gap coefficients between nodes in each influence chain, and the monitoring equipment corresponding to each node. Based on the influence chain of each monitoring device, analyze the influence coefficients of various monitoring indicators between the monitoring devices. Specifically, this includes:

[0157] S2061. Obtaining Monitoring Indicators The entire influence relationship chain, marked nodes and Simultaneous influence chains. Analyze and label monitoring indicators between nodes in each influence chain. The difference coefficient.

[0158] S2062, Statistical markers affect the number of relationship chains Get node pairs The difference coefficient in each label-affected relation chain is substituted into the formula to calculate the node pair. Monitoring indicators under two monitoring devices Influence coefficient :

[0159] ;

[0160] In the formula, For node pairs In the The bar markers affect the gap coefficient in the relation chain. For the first The bar marker affects the maximum difference coefficient between nodes in the relation chain.

[0161] In practical implementation, the influence coefficient is used to quantify the strength of the influence correlation between two specific monitoring devices on a certain monitoring indicator. It is used to comprehensively consider the performance of the corresponding node pair of the monitoring device in all influence relationship chains that include them.

[0162] The correlation contribution of a single chain is obtained by calculating the relative ratio of the difference coefficient in each labeled relation chain to the maximum difference coefficient in that chain, and then calculating the difference between the ratio and a fixed value.

[0163] The arithmetic mean of the correlations of all relevant relationship chains is used to derive a normalized comprehensive influence coefficient. The closer the node pair is to a fixed value, the more stable the historical influence pattern of the two monitoring devices on a certain monitoring indicator and the stronger the predictability.

[0164] S2063. Obtain the complete influence relationship chain for each monitoring indicator and analyze the difference coefficients of different nodes on each monitoring indicator. Based on the correspondence between nodes and monitoring equipment, calculate the influence coefficients of each monitoring indicator between monitoring equipment.

[0165] In the specific implementation process, physical hydrological processes are deeply integrated with data statistical models. By establishing a flow direction relationship tree diagram and calculating relative time differences, the actual convection and diffusion processes of pollutants in water bodies are simulated, transforming static spatial location relationships into dynamic, time-series dependent causal relationship networks.

[0166] The final calculated impact coefficient quantifies the statistical correlation strength and direction between any two monitoring points on various water quality indicators, providing a scientific and quantifiable reference benchmark for anomaly judgment.

[0167] S300. Obtain the currently collected monitoring indicators from each monitoring device as measured values, and calculate the predicted values ​​based on the influence relationship chain. Analyze the difference between the measured values ​​and the predicted values, and calculate the anomaly index for each monitoring device based on the influence coefficient, and delineate the anomaly zone.

[0168] Specifically, it includes:

[0169] S301. Obtain the current monitoring indicators collected by each monitoring device and use them as measured values, then filter out those indicators that contain data from the monitoring devices. The complete influence chain of the corresponding node. Analyze the relationship between each node in each influence chain and the monitoring equipment. The relationship between the influence of various monitoring indicators between corresponding nodes.

[0170] S302. Substitute the measured values ​​of each node into the influence relationship formula to obtain the monitoring equipment. The different predicted values ​​for each monitoring indicator are analyzed. Monitoring equipment is also analyzed. The differences between the predicted and measured values ​​of various monitoring indicators are used to calculate the monitoring equipment's performance in conjunction with the corresponding influence coefficients. Abnormal index :

[0171] ;

[0172] In the formula, For the number of all monitored indicators, For the first In the entire impact chain of the monitoring indicators, excluding monitoring equipment The number of duplicates after removing all other nodes.

[0173] For the first The first monitoring indicator Each node corresponds to a monitoring device and a monitoring equipment. The influence coefficient between them.

[0174] For the first The first monitoring indicator Substituting each node into the influence relationship formula, the monitoring equipment obtained The predicted value. For monitoring equipment Next The measured values ​​of the monitoring indicators.

[0175] In practice, the anomaly index is used to calculate the overall anomaly level of a single monitoring device. Its calculation process involves several levels:

[0176] First, for each monitoring indicator, the relative deviation between the measured value of the device and the multiple predicted values ​​obtained based on each associated node and the corresponding influence relationship is calculated.

[0177] Secondly, the relative deviations are weighted using the influence coefficients between the device and each associated node to reflect the differences in the reliability of prediction results of different associated nodes. The average value of all weighted relative deviations under the indicator is then calculated to obtain the abnormal contribution value of the indicator.

[0178] Finally, the arithmetic mean of the abnormal contribution values ​​of all monitoring indicators is calculated again to obtain the comprehensive abnormality index of the equipment. This index comprehensively reflects the degree of deviation of the equipment from the overall monitoring network across all monitoring indicators.

[0179] The calculation of the anomaly index of each monitoring device requires using the time corresponding to the measured value as the base time, analyzing the relative time difference of all other monitoring devices in the same influence chain, and adding the maximum relative time difference to the base time to obtain the cutoff time. Only when the time exceeds the cutoff time can the anomaly index of the monitoring device be calculated completely.

[0180] S303. Calculate the anomaly index for each monitoring device, and classify monitoring devices with an anomaly index greater than the threshold as abnormal devices. Preset base distance. Based on the abnormality index of abnormal equipment Calculate reference distance On the GIS map, using the location of each abnormal device as the center, and referring to the distance... Circular anomaly regions are divided according to their radii.

[0181] Reference distance The calculation formula is as follows:

[0182] ;

[0183] In the formula, The preset attenuation index, This represents the maximum anomaly index among all abnormal devices.

[0184] In practice, the reference distance is used to non-linearly map the anomaly index of the monitoring equipment to the radius of the suspicious area in geospatial space, so as to guide the scope of on-site investigation.

[0185] By using a saturation growth function, the ratio of the device's anomaly index to the maximum anomaly index among all devices is converted into a reference distance value between zero and a preset base distance.

[0186] When the anomaly index is low, the radius increases gradually, making it easier to focus on key locations for detailed investigation. As the anomaly index approaches its maximum value, the radius gradually approaches its theoretical maximum value to cover a wider area of ​​potential pollution impact.

[0187] It enables adaptive adjustment of the investigation scope based on the severity of the anomaly.

[0188] In practice, by comparing real-time measured values ​​with predicted values ​​calculated based on historical correlation models, it is possible to effectively distinguish between "changes caused by normal upstream propagation" and "real anomalies that suddenly occur locally".

[0189] By combining the impact coefficient with weighted calculation of the anomaly index, the judgment result not only considers the magnitude of the deviation, but also the credibility of the source of the deviation (the prediction of nodes with strong correlation is more credible, and their deviation weight is greater).

[0190] Finally, anomaly zones are dynamically generated based on the anomaly index, transforming the abstract numerical index into intuitive spatial action instructions with priority (reverse sorting) and range guidance (reference distance).

[0191] The S400 monitoring center displays the specific location of each anomaly area on the GIS map through a visual screen, prompting staff to check each anomaly area one by one in reverse order of the anomaly index, and storing the collected monitoring indicators and hydrological indicators into the historical record.

[0192] The monitoring loop has been completed and the system has been given learning capabilities. The visualization display presents complex analysis results (abnormal indices, abnormal areas) intuitively on the GIS map, which greatly improves the situational awareness and decision-making efficiency of managers.

[0193] In the specific implementation process, the prompts for checking in reverse order of abnormal indices enabled the intelligent sorting of handling strategies.

[0194] Storing this data in the historical records will enrich the historical database with the experience of each monitoring, analysis and handling, which will be used to update and optimize the future impact relationship chain and impact coefficient, so that the prediction and anomaly detection capabilities can be continuously improved over time.

[0195] Example 2: Please refer to Figure 2 The present invention also provides a water environment pollution monitoring system based on data analysis, including an environmental monitoring module, an impact analysis module, an anomaly identification module, and a visualization and storage module.

[0196] The environmental monitoring module collects GIS maps and historical records, and collects monitoring indicators and hydrological indicators of river water bodies through monitoring equipment.

[0197] In the specific implementation process, multi-source data is collected, including GIS maps describing the river topography and distribution, historical records of monitoring indicators (such as pollutant concentration) and hydrological indicators (such as flow velocity and flow direction) collected by various monitoring devices, and real-time data of various water indicators are collected through dedicated monitoring devices deployed in the river.

[0198] The spatial, temporal, and real-time data foundation required to build the system combines geographic information, historical patterns, and field measurements, providing complete and dynamic data input for subsequent analysis and ensuring the spatiotemporal continuity and data traceability of the monitoring work.

[0199] The impact analysis module marks the location of each monitoring device on the GIS map, analyzes upstream and downstream relationships, and sets relative time differences. It establishes impact relationship chains for various monitoring indicators and calculates the impact coefficients between monitoring devices.

[0200] In the specific implementation process, the equipment locations are first marked on the GIS map, and the upstream and downstream relationships between monitoring equipment are established based on the flow direction and relative position, forming a flow relationship tree diagram.

[0201] Then, the relative time difference between parent and child nodes is calculated based on flow rate and path, and the time series of historical data is aligned according to this time difference.

[0202] Finally, by using linear regression, quantitative influence relationships and influence chains are established for each monitoring indicator across different equipment nodes, and the influence coefficients between any two monitoring devices for each monitoring indicator are calculated.

[0203] By deeply integrating and mathematically modeling the spatial topology, hydrodynamic characteristics, and historical water quality change patterns of river systems, the degree of mutual influence and propagation delay between monitoring points were quantified, and a predictive benchmark model based on spatial correlation was established for anomaly detection.

[0204] The anomaly identification module uses the current monitoring indicators collected by each monitoring device as measured values, calculates predicted values ​​based on the influence relationship chain, and calculates the anomaly index of each monitoring device and divides the anomaly area based on the influence coefficient.

[0205] In the specific implementation process, the current measured values ​​of each monitoring device are obtained. Using the established influence relationship chain of various monitoring indicators and the corresponding influence relationship formula, the measured values ​​of other nodes in the chain are substituted to calculate the different predicted values ​​of multiple monitoring indicators of the target device.

[0206] By comparing the deviations between the measured values ​​and the predicted values, and by weighting the results with the influence coefficients between the devices, the anomaly index of each monitoring device is calculated.

[0207] Based on the magnitude of the anomaly index, and using the device location as the center and a dynamically calculated radius, different anomaly zones are divided on the GIS map.

[0208] It has enabled a shift from single-point data judgment to collaborative diagnosis based on spatial correlation networks, which can more accurately locate real pollution anomalies. Through anomaly indices and dynamic regional division, it can rank suspected pollution areas by severity and predict their spatial extent, greatly improving the targeting and efficiency of the investigation.

[0209] The visualization and storage module displays the location of each abnormal area, prompts staff to check each abnormal area one by one in reverse order of the abnormal index, and stores the collected index data into the historical record.

[0210] During implementation, the results output by the anomaly identification module—namely, the location and extent of each anomaly area—are overlaid on the visualization screen of the GIS map. The system prompts staff to conduct on-site investigations one by one, following the anomaly index from highest to lowest (in reverse order). Simultaneously, all monitoring and hydrological data collected in this round are automatically stored in the historical record.

[0211] The visual interface provides managers with an intuitive and comprehensive understanding of the pollution situation and decision support, while prioritizing investigation paths optimizes human resource allocation. Data storage enables the continuous accumulation of historical records, providing data support for subsequent iterative updates to impact relationship chains and impact coefficients, thus giving the entire monitoring system the ability to continuously learn and self-optimize.

[0212] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0213] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A water environment pollution monitoring method based on data analysis, characterized in that: The method includes: S100: Collect GIS maps and historical records, and collect monitoring indicators and hydrological indicators of river water bodies through monitoring equipment; S200. Mark the location of each monitoring device on the GIS map, analyze the upstream and downstream relationships and set the relative time difference; combine historical records to establish an influence relationship chain for each monitoring indicator, and then calculate the influence coefficient between each pair of monitoring devices. S300: Obtain the current monitoring indicators collected by each monitoring device as measured values, calculate the predicted values ​​based on the influence relationship chain; analyze the difference between the measured values ​​and the predicted values, calculate the anomaly index of each monitoring device in combination with the influence coefficient, and divide the anomaly area; The S400 monitoring center displays the specific location of each anomaly area on the GIS map through a visual screen, prompting staff to check each anomaly area one by one in reverse order of the anomaly index, and storing the collected monitoring indicators and hydrological indicators into the historical record.

2. The water environment pollution monitoring method based on data analysis according to claim 1, characterized in that: In S100, GIS maps are used to describe the topographic structure and distribution of rivers and waterways within a specified area; The designated area refers to the spatial range pre-delineated based on the boundaries of hydrological units in a river basin and administrative management needs; Historical records include various monitoring indicators collected by each monitoring device at different historical moments; Monitoring equipment refers to specialized equipment used to collect river water samples and perform automated analysis, obtaining various monitoring indicators and hydrological indicators based on the analysis results; Monitoring indicators are parameters used to quantify the pollution characteristics of river water bodies; hydrological indicators are physical quantities that describe the dynamic characteristics and transport capacity of water bodies.

3. The water environment pollution monitoring method based on data analysis according to claim 2, characterized in that: S200 includes: S201. Mark the location of each monitoring device on the GIS map, and analyze the flow velocity and direction in the hydrological indicators collected by the monitoring devices; based on the flow direction and relative position of each monitoring device, analyze the upstream and downstream relationships between the monitoring devices. S202, monitoring equipment located upstream As downstream monitoring equipment The parent node, As The child nodes; each monitoring device is treated as a node, and a flow relationship tree diagram is established based on the parent-child node relationship; S203. Based on the parent-child node relationship between each monitoring device in the flow direction relationship tree diagram, construct a smooth flow velocity evolution curve according to the flow velocity data collected by each pair of parent-child nodes, and analyze the relative position of each pair of parent-child nodes in the GIS map. S204. Based on the relative position, plan the shortest flow path for the corresponding parent and child nodes, analyze the propagation time difference of the flow velocity evolution curve within the shortest flow path under the parent and child nodes, and use it as the relative time difference of the monitoring equipment corresponding to the parent and child nodes. S205. Analyze the flow velocity of each monitoring device in the historical record and calculate the relative time difference, then align the time series according to the relative time difference; establish the influence relationship formula and influence relationship chain for each monitoring indicator based on the flow direction relationship tree diagram. S206. Analyze the gap coefficients between nodes in each influence relationship chain, and the monitoring equipment corresponding to each node; based on the influence relationship chain of each monitoring equipment, analyze the influence coefficients of various monitoring indicators between the monitoring equipment.

4. The water environment pollution monitoring method based on data analysis according to claim 3, characterized in that: S205 includes: S2051. Starting from the node without a parent node in the flow relationship tree diagram and ending at the node without a child node, count the number of nodes. Starting from the origin and ending at the destination, the flow relationship tree diagram is decomposed into... A separate flow relationship chain; S2052, Settings For each historical moment, retrieve the flow rate of each monitoring device at each historical moment from the historical records; calculate the relative time difference based on the flow rate of each monitoring device at the same historical moment, and map it to each flow direction relationship chain respectively; S2053. Based on the relative time difference between nodes in the flow relationship chain, align the time sequence of the collected monitoring indicators; establish the influence relationship between each monitoring indicator and each node in each flow relationship chain, and thus calculate the difference coefficient. ; S2054, Based on monitoring indicators between nodes The difference coefficient is set at the midpoint in each flow relationship chain, and the influence relationship chain is divided; the midpoint is set for each monitoring indicator in each flow relationship chain, and the influence relationship chain is divided accordingly.

5. The water environment pollution monitoring method based on data analysis according to claim 4, characterized in that: S2053 includes: S2053-1, Statistical Flow Relationship Chain Total number of nodes By pairwise association of these nodes, a total of 1 pair of nodes; among them ; S2053-2, Based on the nodes at each historical moment Align time series with relative time difference, and obtain node pairs after alignment time series at the same historical time. Collected monitoring indicators , respectively, were used as independent and dependent variables and packaged into a sample; S2053-3, Setting the intercept and regression coefficients And establish a linear regression model; The independent variables in each sample are used as input values. The output value of each sample The difference between the dependent variable and the dependent variable is taken as the gap value; the expression is as follows: ； S2053-4. The sum of the differences among all samples is used as the difference coefficient. The minimum difference coefficient is obtained by adjusting the intercept and regression coefficient. After determining the values ​​of the intercept and regression coefficients, the monitoring indicators are obtained. Next node pair The influence relationship; S2053-5, and so on, establish monitoring indicators for each node in each flow relationship chain. And the influence relationships of other monitoring indicators, and classify these influence relationships according to the different monitoring indicators.

6. The water environment pollution monitoring method based on data analysis according to claim 4, characterized in that: S2054 includes: S2054-1, Setting the fluctuation coefficient Analyze the flow relationship chain Middle starting point with its child nodes Monitoring indicators between Difference coefficient According to the formula: Calculate reference coefficient ; S2054-2, Analyzing the flow relationship chain The difference coefficient between every two nodes is used to calculate the flow relationship chain sequentially from the starting point to the ending point. internal nodes Then each node Gap Index : ； In the formula, To flow to the relationship chain Middle node The total number of all previous nodes, For nodes Compared to the previous one The difference coefficient between nodes; S2054-3, When node Gap Index Greater than the reference coefficient When selecting a node The previous node is used as the midpoint; the difference index of all nodes. None are greater than the reference coefficient When choosing a point, select the node corresponding to the endpoint as the midpoint. S2054-4, From Flow to Relationship Chain The starting point and the midpoint are used to divide the monitoring indicators. The influence relationship chain; and so on, with monitoring indicators for each flow relationship chain. Set the midpoint and divide the influence relationship chain.

7. The water environment pollution monitoring method based on data analysis according to claim 3, characterized in that: S206 includes: S2061. Obtaining Monitoring Indicators The entire influence relationship chain, marked nodes and Simultaneous influence relationship chains; analysis and labeling of monitoring indicators between nodes in each influence relationship chain. The difference coefficient; S2062, Statistical markers affect the number of relationship chains Get node pairs The difference coefficient in each label-affected relation chain is substituted into the formula to calculate the node pair. Monitoring indicators under two monitoring devices Influence coefficient : ； In the formula, For node pairs In the The bar markers affect the gap coefficient in the relation chain. For the first The bar marker affects the maximum difference coefficient between nodes in the relation chain; S2063. Obtain the complete influence relationship chain of each monitoring indicator, and analyze the difference coefficient of different nodes on each monitoring indicator; combine the correspondence between nodes and monitoring equipment, and calculate the influence coefficient of each monitoring indicator between monitoring equipment respectively.

8. The water environment pollution monitoring method based on data analysis according to claim 3, characterized in that: The S300 includes: S301. Obtain the current monitoring indicators collected by each monitoring device and use them as measured values, then filter out those indicators that contain data from the monitoring devices. The complete influence chain of the corresponding node; analyze the relationship between each node in each influence chain and the monitoring equipment. The influence relationship between various monitoring indicators between corresponding nodes; S302. Substitute the measured values ​​of each node into the influence relationship formula to obtain the monitoring equipment. The different predicted values ​​of various monitoring indicators; analysis of monitoring equipment. The differences between the predicted and measured values ​​of various monitoring indicators are used to calculate the monitoring equipment's performance in conjunction with the corresponding influence coefficients. Abnormal index ; S303. Calculate the anomaly index for each monitoring device, and classify monitoring devices with an anomaly index greater than the threshold as abnormal devices; preset the base distance. Based on the abnormality index of abnormal equipment Calculate reference distance On the GIS map, using the location of each abnormal device as the center, and referring to the distance... Divide the area into circular anomaly regions based on the radius; Reference distance The calculation formula is as follows: ； In the formula, The preset attenuation index, This represents the maximum anomaly index among all abnormal devices.

9. The water environment pollution monitoring method based on data analysis according to claim 8, characterized in that: In S302, the anomaly index The calculation formula is: ； In the formula, For the number of all monitored indicators, For the first In the entire impact chain of the monitoring indicators, excluding monitoring equipment The number of duplicates after removing all other nodes; For the first The first monitoring indicator Each node corresponds to a monitoring device and a monitoring equipment. The influence coefficient between them; For the first The first monitoring indicator Substituting each node into the influence relationship formula, the monitoring equipment obtained The predicted value; For monitoring equipment Next The measured values ​​of the monitoring indicators.

10. A water environment pollution monitoring system based on data analysis, characterized in that: The system includes an environmental monitoring module, an impact analysis module, an anomaly identification module, and a visualization and storage module; The environmental monitoring module collects GIS maps and historical records, and collects monitoring indicators and hydrological indicators of river water bodies through monitoring equipment; The impact analysis module marks the location of each monitoring device on the GIS map, analyzes the upstream and downstream relationships and sets the relative time difference; it establishes an impact relationship chain for each monitoring indicator and calculates the impact coefficient between monitoring devices. The anomaly identification module uses the current monitoring indicators collected by each monitoring device as measured values, calculates predicted values ​​based on the influence relationship chain, and calculates the anomaly index of each monitoring device and divides the anomaly area based on the influence coefficient. The visualization and storage module displays the location of each abnormal area, prompts staff to check each abnormal area one by one in reverse order of the abnormal index, and stores the collected indicator data in the history.

Images