Pollution Source Tracing System and Method Based on Water Quality and Quantity Monitoring and Analysis of Drainage System

Abstract

The present invention discloses a pollution source tracing system and method based on water quality and quantity monitoring and analysis of a drainage system, specifically relating to the technical field of water quality and quantity monitoring; real-time collection of water quality and quantity parameters is carried out through a distributed sensor network, and combined with denoising processing and anomaly detection algorithms to ensure data quality; a multivariate dynamic analysis method and a spatio-temporal correlation model are used to identify the abnormal change characteristics of pollutant concentrations; a non-linear superposition effect correction model is constructed to eliminate the interference of false peaks at intersections and locate the real pollution sources; combined with a geographic information system and a drainage basin model, the pollution propagation process is simulated based on the diffusion path and flow characteristics to accurately determine the geographical location of the pollution sources; finally, the tracing results are visually output to a user terminal, a pollution source tracing report is generated, and scientific treatment suggestions and multi-level alarm signals are provided to provide effective technical support for pollution treatment and emergency response.

Description

Technical Field

[0001] The present invention relates to the technical field of water quality and quantity monitoring, and particularly to a pollution source tracing system and method based on water quality and quantity monitoring and analysis of a drainage system. Background Art

[0002] With the acceleration of the urbanization process, the problem of water resource pollution has become increasingly serious. As an important water resource collection and treatment facility, the water quality and quantity of the urban drainage system directly affect the water environment quality of the downstream water body. Traditional drainage system monitoring usually adopts the methods of decentralized sampling and regular detection, but this method has problems such as poor real-time performance, limited coverage, and low monitoring efficiency, and it is difficult to meet the requirements of modern environmental monitoring. In recent years, water quality and quantity monitoring systems based on the Internet of Things, big data analysis, and sensor technology have gradually emerged. The combination of these technologies makes real-time and automated water quality and quantity monitoring possible, providing important technical support for the rapid identification and tracing of pollutant emission sources.

[0003] The pollution source tracing system is an important extension of drainage system monitoring. Through comprehensive analysis of water quality and quantity data, combined with geographic information system (GIS) and watershed models, it can accurately locate pollution sources and formulate scientific pollution control measures. However, most current tracing systems rely on a single water quality index or simple emission path analysis, lacking a comprehensive analysis of the multi-variable and multi-scale pollution diffusion process in complex drainage systems, which limits the accuracy and reliability of pollution source tracing.

[0004] The existing technologies have the following deficiencies:

[0005] In the current pollution source tracing system based on the drainage system, due to the non-linear superposition effect generated by the emissions of multiple pollution sources in the complex drainage network at the intersection, it is easy to form a "false peak" of local pollutant concentration, interfering with the accuracy of pollution source tracing. This superposition effect will cover up the real main pollution source, making it difficult for pollution source tracing to focus on the actual emission point. In addition, the dynamic characteristics of industrial and domestic sewage emissions (such as changes during peak and trough periods) and emergencies such as rainfall will further exacerbate this complexity, resulting in the monitoring system being difficult to distinguish short-term anomalies from long-term pollution trends. At the same time, outliers in the monitoring data (such as sensor errors or temporary sediment disturbances) may be misjudged as pollution source signals, and the existing analysis methods cannot effectively filter these interferences, resulting in the tracing conclusion deviating from the actual situation and affecting the scientificity and timeliness of governance decisions. Summary of the Invention

[0006] The purpose of the present invention is to provide a pollution source tracing system and method based on water quality and quantity monitoring and analysis of a drainage system to solve the deficiencies in the background art.

[0007] To achieve the above object, the present invention provides the following technical solutions: A pollution source tracing method based on water quality and quantity monitoring and analysis of a drainage system, comprising the following steps:

[0008] S1: Real-time collect water quality parameters and water quantity parameters in the drainage system through a distributed sensor network, where the water quality parameters include pH value, dissolved oxygen, chemical oxygen demand, ammonia nitrogen concentration, and total phosphorus concentration, and the water quantity parameters include flow velocity and flow rate;

[0009] S2: Denoise the collected water quality parameters and water quantity parameters, and filter non-pollution source signals through an anomaly detection algorithm, including abnormal data caused by sensor errors and sudden sediment disturbances;

[0010] S3: Use multivariate analysis methods to dynamically analyze the processed data, identify abnormal changes in pollutant concentrations in the drainage system, and at the same time determine the correlation between pollutant concentration changes and emission time periods and regions through a spatio-temporal correlation analysis model;

[0011] S4: Construct a non-linear superposition effect correction model, and identify the potential locations of real pollution sources by simulating the superposition effects of emissions from different pollution sources in a complex drainage network;

[0012] S5: Combine a geographic information system and a drainage basin model, and based on the pollutant diffusion path and the flow characteristics of the drainage network, simulate the propagation process of pollutants and locate the geographical location of the pollution source;

[0013] S6: Output the pollution source tracing results to the user terminal, generate a pollution source tracing report, and provide pollution control suggestions and alarm signals according to the tracing results.

[0014] Preferably, in S4, obtain the topological structure of the drainage network based on a geographic information system, including pipeline nodes, pipeline attributes, and pollution source locations; use the one-dimensional Saint-Venant equation to simulate the fluid motion in the pipeline, describe the changes in the velocity and flow rate of the water flow in the pipeline, and the expression is: where: Q is the instantaneous flow rate in the pipeline; A is the cross-sectional area of the flow; H is the water head height; g is the acceleration due to gravity;

[0015] Simulate the diffusion and mixing of pollutants in the pipe network based on the convection-diffusion equation: where: C is the pollutant concentration; u is the water flow velocity, D is the diffusion coefficient; R(C) is the reaction term for pollutant degradation or transformation;

[0016] For each known pollution source, separately simulate its emission characteristics, including emission intensity, emission duration, and emission cycle; output the pollution concentration contribution value of the single source at each node in the pipe network, denoted as C i(x, t); within the same space-time range, the emission concentrations of all pollution sources are superimposed to calculate the total pollution concentration: If the monitored value C obs (x, t) at a certain node is higher than C total (x, t), it indicates that there are unrecognized potential pollution sources.

[0017] Preferably, in S4, for the unrecognized potential pollution sources, a non-linear correction factor f(x, t) is introduced to simulate the non-linear superposition behavior of pollutants in the complex pipe network: C corrected (x, t) = C total (x, t) + f(x, t); where f(x, t) is obtained by fitting historical monitoring data and the flow-concentration relationship; C corrected (x, t) represents the pollutant concentration distribution corrected by the non-linear superposition effect at the spatial position x and time t; an inversion algorithm is used to calculate the emission characteristics of the pollution source: min||C 0bs (x, t) - C simulated (x, t)|| 2 ; where C simulated (x, t) represents the simulated concentration distribution calculated by the pollutant diffusion model based on the assumed pollution source location and emission characteristics at the spatial position x and time t. The calculation formula is: M i is the emission flow rate of the i-th potential pollution source; C source,i is the emission concentration of the i-th pollution source; L i is the pipeline length from the pollution source to the monitoring point; D is the diffusion coefficient; is the Gaussian diffusion function, which describes the diffusion process of pollutants;

[0018] Based on the pollutant concentration gradient, calculate the potential location of the pollution source: If the concentration at the downstream monitoring point is higher than that at the upstream point, combined with the flow velocity and concentration change, locate the potential pollution source; identify the key nodes where the pollutant concentration suddenly increases along the path, and evaluate the contribution of each pollution source to the total concentration by assigning weights w i : where n is the total number of pollution sources. If the weight of a certain pollution source is higher than that of other pollution sources, it is marked as a potential pollution source.

[0019] Preferably, in S5, based on the pollutant diffusion path and the flow characteristics of the drainage network, simulate the propagation process of pollutants and locate the geographical location of the pollution source, specifically:

[0020] Divide the drainage network into several segments s i according to the flow direction, and each segment corresponds to a monitoring point; quantitatively calculate the contribution of the pollutant concentration of each segment: where: C i is the path segment si Contribution rate to the total pollutant concentration at the monitoring point, Q i is the flow rate of path segment s i ; C in，i is the input pollutant concentration of path segment s i ; For continuous path segments s i and s i+1 , calculate the gradient change of the contribution rate: where PCG is the pollution contribution rate gradient between segment s i and S i+1 ; L i+1 , L i is the spatial distance between segment S i+1 and s i .

[0021] Preferably, for each monitoring point M i , calculate the time delay Δt i for the pollutant to propagate from the upstream node U i to the monitoring point, and the expression is: where: L i is the distance from the upstream node U i to the monitoring point M i ; u is the water flow velocity; C in is the pollutant concentration; k is the diffusion rate coefficient; D is the pipe diameter; Normalize the lag time to obtain the diffusion lag coefficient, and the expression is: where: Δt max is the maximum lag time in the drainage network.

[0022] Preferably, classify each region in the drainage network by applying the K-Means clustering algorithm to the joint distribution of the pollution contribution rate gradient PCG and the diffusion lag coefficient DLC. Specifically: for each monitoring point i, form a two-dimensional feature vector: X i = [PCG i , DLC i ; Normalize the features: where: μX is the feature mean; σX is the feature standard deviation;

[0023] Use the K-Means clustering algorithm to classify the standardized PCG and DLC data: K-Means divides the data points into k clusters by minimizing the Euclidean distance, and the expression is: where: C j is the j-th cluster; μ j is the centroid of the j-th cluster;

[0024] Use the elbow rule to determine the optimal number of clusters k: gradually increase the value of k, calculate the sum of squared errors within the cluster under each k; find the point where the error decreases significantly slower as the optimal k value;

[0025] The monitoring points are divided into k clusters, each cluster corresponds to a pollution characteristic area: high PCG, high DLC: areas with severe distal diffusion; high PCG, low DLC: areas with potential pollution sources; low PCG, high DLC: downstream areas affected by dilution; low PCG, low DLC: background pollution areas;

[0026] A heat map of the joint distribution of PCG and DLC is generated based on GIS: high PCG and low DLC areas are marked as red high priority areas on the map; medium PCG and DLC areas are marked as yellow medium priority areas; low PCG and high DLC areas are marked as green low priority areas; in the red high priority areas, the location of the pollution source is confirmed by combining historical emission data, pipeline network structure and actual investigation.

[0027] Preferably, in S6, an alarm signal is generated according to the traceability result, specifically:

[0028] The mean μ and standard deviation σ of PCG and DLC data of all monitoring points are calculated respectively; PCG and DLC of each monitoring point are standardized to the degree of deviation, and the expression is: In the formula, is the PCG deviation degree of monitoring point i, is the DLC deviation degree of monitoring point i, μ PCG is the mean of PCG, σ PCG is the standard deviation of PCG; μ DLC is the mean value of DLC, σ DLC is the standard deviation of DLC;

[0029] Based on the degree of deviation between PCG and DLC, define the warning level rules:

[0030] Level 1 warning: and The concentration changes steadily and the diffusion hysteresis is normal, indicating slight pollution and no emergency response is required;

[0031] Level 2 warning: or Significant changes in concentration or large diffusion lag indicate moderate pollution, requiring enhanced monitoring and preliminary investigation of pollution sources;

[0032] Level 3 warning or Or both exceed the standard at the same time; the concentration changes dramatically and the diffusion lag is serious, indicating that the peak concentration of pollutants is abnormal or the propagation speed is fast, and emergency response measures need to be initiated immediately.

[0033] The present invention also provides a pollution source tracing system based on the monitoring and analysis of water quality and quantity in a drainage system, including a data acquisition module, a data processing module, a correlation analysis module, a non-linear superposition effect correction module, a pollution propagation simulation and positioning module, and an early warning module:

[0034] Data acquisition module: It collects water quality parameters and water quantity parameters in the drainage system in real time through a distributed sensor network. The water quality parameters include pH value, dissolved oxygen, chemical oxygen demand, ammonia nitrogen concentration, and total phosphorus concentration, and the water quantity parameters include flow velocity and flow rate;

[0035] Data processing module: It performs denoising processing on the collected water quality parameters and water quantity parameters, and filters out non-pollution source signals through an anomaly detection algorithm, including abnormal data caused by sensor errors and sudden sediment disturbances;

[0036] Correlation analysis module: It uses a multivariate analysis method to dynamically analyze the processed data, identify abnormal changes in pollutant concentrations in the drainage system, and at the same time determine the correlation between changes in pollutant concentrations and emission time periods and regions through a spatio-temporal correlation analysis model;

[0037] Non-linear superposition effect correction module: It constructs a non-linear superposition effect correction model, and identifies the potential locations of real pollution sources by simulating the superposition effects of emissions from different pollution sources in a complex drainage network;

[0038] Pollution propagation simulation and positioning module: It combines a geographic information system and a drainage basin model, and based on the pollutant diffusion path and the flow characteristics of the drainage network, simulates the pollutant propagation process and locates the geographical location of the pollution source;

[0039] Early warning module: It outputs the pollution source tracing results to the user terminal, generates a pollution source tracing report, and provides pollution control suggestions and alarm signals according to the tracing results.

[0040] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0041] 1. The present invention collects water quality and water quantity data in real time through a distributed sensor network, combines denoising processing and an anomaly detection algorithm, and improves the accuracy of monitoring data; uses multivariate analysis, spatio-temporal correlation analysis, and a non-linear superposition effect correction model to effectively identify real pollution sources and solve the problem of interference from "false peaks"; combines GIS and a drainage basin model to simulate the pollutant diffusion path and accurately locate the geographical location of the pollution source; through the joint analysis of the pollution contribution rate gradient (PCG) and the diffusion lag coefficient (DLC), it further classifies the pollution areas of the drainage network and provides a scientific basis for the priority investigation of pollution sources.

[0042] 2. The present invention significantly improves the accuracy and reliability of pollution source tracing in complex drainage networks, realizing full-process intelligence from data collection to precise positioning of pollution sources. By generating visual source tracing reports and multi-level warning signals, the present invention can provide real-time decision-making support for management departments, guiding pollution control and emergency response; its precise pollution source positioning and pollution area priority division help narrow the investigation scope and reduce treatment costs; in addition, through dynamic monitoring of potential pollution sources and analysis of emission behaviors, major pollution events can be pre-warned in advance, reducing environmental risks and protecting the ecological environment and public health. Brief Description of the Drawings

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required to be used in the embodiments. Obviously, the drawings described below are only some embodiments recorded in the present invention, and those of ordinary skill in the art can also obtain other drawings based on these drawings.

[0044] Figure 1 It is a flowchart of the method of the present invention.

[0045] Figure 2 It is a system module diagram of the present invention. Detailed Embodiments

[0046] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.

[0047] Example 1, please refer to Figure 1 As shown, the pollution source tracing method based on water quality and water volume monitoring and analysis of the drainage system in this embodiment includes the following steps:

[0048] S1: Real-time collect water quality parameters and water volume parameters in the drainage system through a distributed sensor network, where the water quality parameters include pH value, dissolved oxygen, chemical oxygen demand, ammonia nitrogen concentration, and total phosphorus concentration, and the water volume parameters include flow velocity and flow rate;

[0049] S2: Denoise the collected water quality parameters and water volume parameters, and filter non-pollution source signals through an anomaly detection algorithm, including abnormal data caused by sensor errors and sudden sediment disturbances;

[0050] S3: Use the multivariate analysis method to dynamically analyze the processed data, identify abnormal changes in pollutant concentrations in the drainage system, and at the same time determine the correlation between pollutant concentration changes and emission time periods and regions through a spatio-temporal correlation analysis model;

[0051] S4: Construct a non-linear superposition effect correction model, and identify the potential locations of real pollution sources by simulating the superposition effects of emissions from different pollution sources in a complex drainage network;

[0052] S5: Combine the geographic information system and the drainage basin model, and based on the pollutant diffusion path and the flow characteristics of the drainage network, simulate the pollutant propagation process to locate the geographical location of the pollution source;

[0053] S6: Output the pollution source tracing results to the user terminal, generate a pollution source tracing report, and provide pollution control suggestions and alarm signals according to the tracing results.

[0054] In S1, install sensor nodes at the main pipelines, junctions and key areas (such as industrial sewage outlets and domestic sewage outlets) of the drainage system;

[0055] Each sensor node includes a water quality monitoring module and a water volume monitoring module. Water quality monitoring module: Configure pH value, electrochemistry dissolved oxygen (DO) sensor, optical chemical oxygen demand (COD) sensor, ion selective electrode ammonia nitrogen sensor, and total phosphorus (TP) photometric sensor; Water volume monitoring module: Use an ultrasonic flow velocity sensor and a weir flowmeter to jointly monitor the flow velocity and flow rate.

[0056] Real-time collect water quality parameters in the drainage system through sensors, including pH value, dissolved oxygen (DO), chemical oxygen demand (COD), ammonia nitrogen concentration, and total phosphorus concentration (TP); Synchronously collect water volume parameters, including flow velocity and flow rate; The data collection frequency is set to once every 5 minutes to meet the requirements of real-time monitoring.

[0057] Each sensor node transmits the collected data to the central data processing platform in real time through a wireless communication module (such as LoRa or NB-IoT); The data is preliminarily encrypted during the transmission process to ensure data integrity and security; On the central platform, the data is stored in time series and metadata such as sensor location and monitoring time is marked.

[0058] The central data processing platform performs integrity verification on the transmitted data and eliminates abnormal or missing data; Apply a denoising algorithm to smooth the possible short-term fluctuations (such as sensor interference and external environmental noise) in the data and extract the long-term stable trend.

[0059] Through the above implementation steps, the present application realizes the efficient real-time monitoring of water quality and water volume at key points of the drainage system. The monitoring data shows that the water quality parameters and water volume parameters of different drainage outlets can be uploaded to the platform in real time with high resolution, providing a reliable data basis for pollution source tracing and subsequent treatment. For example, in the sewage monitoring of an industrial park, the pH value shows a continuous low level (exceeding the acid standard), and the concentrations of COD and ammonia nitrogen increase significantly. The possible pollution sources of abnormal emissions are quickly locked through real-time data analysis.

[0060] S2: Use the moving average method to smooth the original data to reduce the influence of random fluctuations on water quality parameters and water volume parameters. Suppress high-frequency noise, and use Butterworth low-pass filters to process flow rate and velocity data to filter out short-term abnormal fluctuations caused by environmental disturbances (such as wind and vibration). Use the locally weighted regression (LOESS) method to extract the long-term trends of water quality data (such as pH value, dissolved oxygen, COD), while retaining the signal characteristics of sudden pollution.

[0061] Establish a baseline model of the normal range based on historical monitoring data. For example, the normal range of pH value is 6.5–8.5; the normal range of dissolved oxygen is >5mg / L; the normal value of total phosphorus concentration is <0.5mg / L. If the current monitoring data deviates from the baseline by more than the preset threshold (such as ±10%), it is marked as a suspected outlier.

[0062] Use principal component analysis (PCA) combined with Mahalanobis distance to detect multivariate anomalies: Take all monitoring parameters (such as pH, COD, flow velocity, etc.) as input variables, reduce the dimension through PCA, and extract the main features; calculate the Mahalanobis distance between the current data point and the normal feature distribution. If it exceeds the set threshold (such as the 99% confidence interval), it is determined as an outlier.

[0063] Make a judgment based on the correlation between water volume and water quality: If the flow velocity increases significantly within a short period of time (such as during a rainstorm), and the concentrations of COD and total phosphorus show a short-term abnormally high value at the same time, it is determined as sediment disturbance; adjust the flow weight to correct the relevant data to eliminate the influence of sediment disturbance.

[0064] For single-point outliers, use time interpolation method for correction; for continuous outliers, reconstruct through historical trends and adjacent node data.

[0065] Classify and label the processed data: normal data, abnormal data (after correction), and possible pollution signals; all data are stored in the central database according to the timestamp and sensor location for subsequent analysis.

[0066] In this application, the above-mentioned denoising and anomaly detection steps effectively improve the reliability and accuracy of the data. For example, in a drainage system monitoring, due to a sensor failure, the dissolved oxygen data showed a short-term abnormally high value (>20mg / L), which was marked as a sensor error and corrected by the anomaly detection algorithm. At the same time, during the rainstorm, the short-term increase in COD caused by sediment disturbance was captured, and the pollution tracing data was corrected through correlation analysis, providing accurate monitoring results for subsequent analysis.

[0067] S3: Input data include: water quality parameters: pH value, dissolved oxygen (DO), chemical oxygen demand (COD), ammonia nitrogen concentration and total phosphorus concentration; water volume parameters: flow rate and flow; time information: timestamp of data collection; spatial information: geographical location of sensor monitoring point and its node number in the pipe network. Align the multivariate data collected by each sensor according to the timestamp to ensure the consistency of multivariate data points in the same time period; for missing data, use adjacent time period data or spatial interpolation algorithm to complete. Perform basic statistical analysis on the denoised data, remove outliers that exceed the upper and lower interquartile range (IQR) to avoid misjudging non-polluting signals as abnormal changes.

[0068] STL decomposition (Seasonal-Trend Decomposition using Loess) is used to decompose the time series of monitoring parameters into three parts: X(t)=T(t)+S(t)+R(t); where: T(t): long-term trend (trend term); S(t): cyclical fluctuation (seasonal term); R(t): random fluctuation or anomaly (residual term).

[0069] Analyze the slowly changing pollution characteristics, for example: the total phosphorus concentration gradually accumulates, slowly rising from 0.2 mg / L to 0.5 mg / L, indicating that there may be a persistent low-intensity pollution source; the COD concentration gradually increases in the industrial emission area, reflecting long-term emission behavior.

[0070] Determine the daily cycle changes in pollutant concentrations. For example, COD concentrations increase significantly during daily peak hours (such as 10:00-12:00), which may correspond to industrial sewage discharge; ammonia nitrogen concentrations fluctuate greatly during the morning and evening peak hours in living areas (such as 7:00-9:00, 17:00-19:00), which may correspond to domestic sewage discharge. By analyzing the residual term R(t), identify abnormal changes in a short period of time: Example: The ammonia nitrogen concentration at a certain node soared from 0.3mg / L to 2.5mg / L in 20 minutes, indicating that there may be a sudden emission event.

[0071] Correlation analysis determines the possible types and characteristics of pollution sources by calculating the relationship between pollutant parameters. nCorrelatio n (Coefficient) Calculate the linear correlation between each parameter: Where: X and Y are two pollutant parameters (such as chemical oxygen demand and ammonia nitrogen concentration); r X，Y ranges from [-1, 1]. The closer it is to 1, the stronger the positive correlation. Example 1: If the COD and ammonia nitrogen concentration both increase significantly (the correlation coefficient is close to 1), it may be industrial sewage discharge; Example 2: If the pH value decreases and the total phosphorus concentration increases, it may correspond to the characteristics of domestic sewage discharge.

[0072] Use a multiple linear regression model to evaluate the comprehensive impact of multiple parameters on pollutant concentration and determine the main influencing factors. Abnormal event detection identifies events of significant deviation from the normal range of pollutants in real time by defining dynamic thresholds.

[0073] The dynamic threshold is calculated based on historical data and environmental baselines: Normal range of pH value: 6.5 - 8.5; Normal range of total phosphorus concentration: < 0.5 mg / L; Normal range of COD concentration: < 50 mg / L. Dynamically adjust the threshold range considering factors such as real-time flow and seasonal variations.

[0074] Set a sliding window (such as data in the past 30 minutes) and calculate the mean and standard deviation: Upper Limit = μ + k·σ; Lower Limit = μ - k·σ; where μ is the mean within the window, σ is the standard deviation, and k is an adjustment coefficient (usually 2).

[0075] When the current monitored value exceeds the dynamic threshold range (upper and lower limits), it is marked as an abnormal event: Example: If the total phosphorus concentration jumps from 0.3 mg / L to 1.2 mg / L (exceeding 2 times the standard deviation of the normal range) during a certain period, it is determined as an abnormal change event. Record the occurrence time, monitoring point location, and pollutant type of the abnormal event for subsequent analysis.

[0076] S4: Obtain the topological structure of the drainage network based on Geographic Information System (GIS), including: pipe nodes (such as junctions, bifurcations, outlets, etc.); pipe attributes (such as length, diameter, slope); source location (known or potential).

[0077] Use the one-dimensional Saint-Venant equations to simulate the fluid motion in the pipeline and describe the changes in water velocity and flow rate in the pipeline: Where: Q is the instantaneous flow rate in the pipeline; A is the cross-sectional area of the flow; H is the water head height; g is the acceleration due to gravity.

[0078] Simulate the diffusion and mixing of pollutants in the pipe network based on the convection-diffusion equation: Where: C is the pollutant concentration; u is the water flow velocity, D is the diffusion coefficient; R(C) is the reaction term for pollutant degradation or transformation (such as COD oxidation).

[0079] For each known or potential pollution source, simulate its emission characteristics separately, including: emission intensity (such as COD concentration); emission duration (such as scheduled emission or intermittent emission); emission cycle (such as peak emission during a certain period of time every day). Output the pollution concentration contribution value of a single source at each node in the pipeline network, recorded as C i (x, t). In the same time and space, the emission concentrations of all pollution sources are superimposed to calculate the total pollution concentration: If the monitoring value C of a node obs (x, t) is higher than C total (x, t), indicating the presence of unidentified potential sources of contamination.

[0080] The nonlinear correction factor f(x, t) is introduced to simulate the nonlinear superposition behavior of pollutants in complex pipe networks: C corrected (x, t) = C total (x, t) + f(x, t); where f(x, t) is obtained by fitting the historical monitoring data and the flow-concentration relationship; C corrected (x, t) represents the pollutant concentration distribution after the nonlinear superposition effect correction at the spatial position x and time t. Example: At the intersection of a tributary and a main road, the correction value may cause an abnormal increase in local concentration due to the intensified turbulent mixing.

[0081] Use an inversion algorithm (such as an optimization method based on gradient descent or genetic algorithm) to calculate the emission characteristics of the pollution source so that the simulated concentration distribution matches the monitored concentration as closely as possible: min||C obs (x, t)-C simulated (x, t)|| 2 Where, C simulated (x, t) represents the simulated concentration distribution calculated by the pollutant diffusion model at spatial position x and time t based on the assumed pollution source location and emission characteristics. It is a key variable in the inversion of pollutant propagation path and is used to compare with actual monitoring data. The calculation formula is: M i is the emission flow of the ith potential pollution source; C source,i is the emission concentration of the i-th pollution source; L i is the length of the pipeline from the pollution source to the monitoring point; D is the diffusion coefficient; is a Gaussian diffusion function, which describes the diffusion process of pollutants.

[0082] Calculating the potential location of the pollution source based on the pollutant concentration gradient: If the concentration at the downstream monitoring point is higher than that at the upstream point, the potential pollution source is located by combining the flow velocity and the concentration change; identifying the key nodes (such as illegal discharge outlets) where the pollutant concentration suddenly increases along the path.

[0083] By assigning weights w i Evaluating the contribution of each pollution source to the total concentration: In the formula, n is the total number of pollution sources. If the weight of a certain pollution source is higher than that of other pollution sources, it is marked as a potential pollution source.

[0084] Comparing the pollutant concentration distribution output by the model with the actual monitoring data to verify the simulation accuracy; if the deviation is large, optimizing the model by adjusting the diffusion coefficient, correction factor or pollution source parameters.

[0085] S5: Combining the geographic information system and the drainage basin model, based on the pollutant diffusion path and the flow characteristics of the drainage network, simulating the pollutant propagation process and locating the geographical location of the pollution source.

[0086] Defining the pollution contribution rate gradient as PCG, which is used to measure the change rate of the contribution of different nodes along the drainage path to the pollutant concentration at the monitoring point, reflecting the spatial distribution characteristics of the pollution source.

[0087] Dividing the drainage network into several segments s i (the minimum path unit between nodes), and each segment corresponds to a monitoring point; quantitatively calculating the pollutant concentration contribution of each segment: Where: C i Is the contribution rate of the path segment s i To the total pollutant concentration at the monitoring point, Q i Is the flow rate of the path segment s i ; C in，i Is the input pollutant concentration of the path segment s i .

[0088] For consecutive path segments s i And s i+1 , calculating the gradient change of the contribution rate: Where PCG is the pollution contribution rate gradient between segment s i And S i+1 ; L i+1 , L i Is the spatial distance of segment S i+1 And s i (calculated based on GIS). A larger positive PCG value indicates a rapid increase in pollutant concentration, which may be the location of the pollution source; a negative PCG value indicates that the pollutant diffuses and dilutes or the influence of the pollution source gradually weakens.

[0089] Define the diffusion lag coefficient as DLC, which is used to describe the time delay for pollutants to propagate to the downstream monitoring point in the drainage network and is related to the pipe flow characteristics and pollutant diffusion characteristics.

[0090] For each monitoring point M i , calculate the time delay Δt for pollutants to propagate from the upstream node U i to the monitoring point. The expression is: i where: L is the distance from the upstream node U i to the monitoring point M i ; u is the water flow velocity; C i is the pollutant concentration; k is the diffusion rate coefficient; D is the pipe diameter. in

[0091] Normalize the lag time to obtain the diffusion lag coefficient. The expression is: where: Δt max is the maximum lag time in the drainage network. A larger DLC value indicates a slow pollutant propagation speed, which may be due to the dominant diffusion effect; a smaller DLC value indicates a fast pollutant propagation speed, which may be affected by a high-intensity pollution source.

[0092] By applying a clustering algorithm (such as K-Means) to the joint distribution of the pollution contribution rate gradient PCG and the diffusion lag coefficient DLC, classify each region in the drainage network to mark the priority areas of potential pollution sources and improve the accuracy of pollution source location.

[0093] For each monitoring point i, form a two-dimensional feature vector: X i = [PCG i , DLC i ; normalize the features to avoid the influence of dimensional differences on clustering: where: μX is the feature mean; σX is the feature standard deviation.

[0094] Use the K-Means clustering algorithm to classify the standardized PCG and DLC data: K-Means divides the data points into k clusters by minimizing the Euclidean distance. The expression is: where: C j is the jth cluster; μ j is the centroid of the jth cluster.

[0095] Use the elbow method to determine the optimal number of clusters k: gradually increase the value of k, calculate the sum of squared errors within the cluster (SSE) for each k; find the point where the error decrease significantly slows down as the optimal k value (elbow point).

[0096] The monitoring points are divided into k clusters, and each cluster corresponds to a pollution characteristic area: high PCG, high DLC: areas with severe distal diffusion; high PCG, low DLC: areas where potential pollution sources are located; low PCG, high DLC: downstream areas that may be affected by dilution; low PCG, low DLC: background pollution areas.

[0097] Focus on analyzing the high PCG, low DLC cluster: High PCG indicates a rapid increase in pollutant concentration, suggesting the possible existence of a pollution source; low DLC indicates a small delay in pollutant propagation, suggesting that the monitoring point is close to the source.

[0098] Based on GIS, a heat map of the joint distribution of PCG and DLC is generated: Areas with high PCG and low DLC are marked as red high-priority areas on the map; areas with medium PCG and DLC are marked as yellow medium-priority areas; areas with low PCG and high DLC are marked as green low-priority areas. In the red high-priority areas, combined with historical emission data, pipe network structure, and actual investigations, the location of the pollution source is confirmed.

[0099] S6: Output the pollution source tracing results to the user terminal, generate a pollution source tracing report, and provide pollution control suggestions and alarm signals according to the tracing results.

[0100] Calculate the mean μ and standard deviation σ of the PCG and DLC data for all monitoring points respectively; Standardize the PCG and DLC of each monitoring point to the degree of deviation (z-score), and the expression is: In the formula, is the degree of deviation of PCG for monitoring point i, is the degree of deviation of DLC for monitoring point i, μ PCG is the mean of PCG, σ PCG is the standard deviation of PCG; μ DLC is the mean of DLC, σ DLC is the standard deviation of DLC.

[0101] Based on the degree of deviation of PCG and DLC, define the early warning level rules:

[0102] Level 1 early warning: and The concentration change is stable and the diffusion lag is normal, indicating mild pollution and no emergency response is required.

[0103] Level 2 early warning: or The concentration change is significant or the diffusion lag is large, indicating moderate pollution, and it is necessary to strengthen monitoring and initially investigate the pollution source.

[0104] Level 3 early warning or Or both exceed the standard simultaneously; the concentration changes violently and the diffusion lag is serious, indicating that the pollutant concentration peak is abnormal or the propagation speed is fast, and emergency response measures need to be initiated immediately.

[0105] Output a pollution source tracing report, and the report content includes: Pollution source location: Combine GIS to mark the areas with high PCG and low DLC; Pollutant diffusion path: Show the influence range of pollutants based on flow characteristics; Warning level: Generate a warning level heat map for the monitoring points and their surrounding areas (green - level 1, yellow - level 2, red - level 3).

[0106] The warning level response measures include: Level 1 warning: Regularly monitor, record the change trend of pollutant concentration, and observe whether it further develops. Level 2 warning: Strengthen monitoring, dispatch inspection teams to the possible pollution source areas, conduct on - site investigations, and evaluate the treatment needs. Level 3 warning: Immediately initiate emergency response measures, including temporarily closing the sewage outlets, protecting the downstream water bodies (such as adding purification agents), and law enforcement actions to prevent pollution from spreading.

[0107] In this embodiment, the water quality parameters (such as pH value, dissolved oxygen, chemical oxygen demand, ammonia nitrogen concentration, and total phosphorus concentration) and water volume parameters (such as flow velocity and flow rate) in the drainage system are collected in real - time through a distributed sensor network. The collected data is denoised and anomaly - detected to eliminate interference signals such as sensor errors and sudden sediment disturbances. Multivariate analysis methods are used to dynamically analyze the processed data, and combined with the spatio - temporal correlation model to identify the abnormal changes in pollutant concentration and their relationship with the emission time period and area. By constructing a non - linear superposition effect correction model, the superposition effects of different pollution sources in a complex drainage network are simulated to locate the potential positions of real pollution sources. Combining with the Geographic Information System (GIS) and the drainage basin model, based on the pollutant diffusion path and flow characteristics, the propagation process of pollutants is further simulated to accurately determine the geographical location of the pollution source. Finally, the tracing results are output to the user terminal, a tracing report is generated, and pollution control suggestions and multi - level warning signals are provided, providing a scientific basis for pollution prevention and emergency response.

[0108] Example 2, please refer to Figure 2 As shown, the pollution source tracing system based on the water quality and water volume monitoring and analysis of the drainage system in this embodiment includes a data acquisition module, a data processing module, a correlation analysis module, a non - linear superposition effect correction module, a pollution propagation simulation and positioning module, and a warning module:

[0109] Data acquisition module: Real - time collect water quality parameters and water volume parameters in the drainage system through a distributed sensor network, where the water quality parameters include pH value, dissolved oxygen, chemical oxygen demand, ammonia nitrogen concentration, and total phosphorus concentration, and the water volume parameters include flow velocity and flow rate;

[0110] Data processing module: Denoise the collected water quality parameters and water volume parameters, and filter out non-pollution source signals through anomaly detection algorithms, including abnormal data caused by sensor errors and sudden sediment disturbances;

[0111] Correlation analysis module: Dynamically analyze the processed data using multivariate analysis methods to identify abnormal changes in pollutant concentrations in the drainage system. At the same time, determine the correlation between pollutant concentration changes and emission time periods and regions through a spatio-temporal correlation analysis model;

[0112] Nonlinear superposition effect correction module: Construct a nonlinear superposition effect correction model to identify the potential locations of real pollution sources by simulating the superposition effects of emissions from different pollution sources in complex drainage networks;

[0113] Pollution propagation simulation and location module: Combine geographic information systems and drainage basin models to simulate the pollutant propagation process based on the pollutant diffusion path and the flow characteristics of the drainage network, and locate the geographical location of the pollution source;

[0114] Early warning module: Output the pollution source tracing results to the user terminal, generate a pollution source tracing report, and provide pollution control suggestions and alarm signals according to the tracing results.

[0115] The above formulas are all dimensionless and take their numerical calculations. The formulas are obtained by software simulation of a large amount of collected data to get a formula closest to the real situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0116] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. These three situations, where A and B can be singular or plural. In addition, the character " / " in this article generally represents an "or" relationship between the front and back associated objects, but it may also represent an "and / or" relationship. Specifically, it can be understood by referring to the context before and after.

[0117] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed in this article can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in hardware or software depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0118] As described above, it is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of changes or substitutions, which should all be covered within the protection scope of the present application.

Claims

1. A pollution source tracing method based on monitoring and analysis of water quality and quantity in drainage systems, characterized by: The following steps are involved: S1: Real-time collection of water quality parameters and water quantity parameters in the drainage system through a distributed sensor network. Water quality parameters include pH value, dissolved oxygen, chemical oxygen demand, ammonia nitrogen concentration and total phosphorus concentration, and water quantity parameters include flow rate and flow rate. S2: De-noise the collected water quality parameters and water quantity parameters, and filter out non-pollution source signals through anomaly detection algorithms, including abnormal data caused by sensor errors and sudden sediment disturbances; S3: Use multivariate analysis methods to dynamically analyze the processed data to identify abnormal changes in pollutant concentrations in the drainage system, and determine the correlation between pollutant concentration changes and discharge periods and regions through spatiotemporal correlation analysis models; S4: Construct a nonlinear superposition effect correction model to identify the potential location of the real pollution source by simulating the superposition effect of emissions from different pollution sources in a complex drainage network, including: The topological structure of the drainage network is obtained based on the geographic information system, including pipeline nodes, pipeline attributes and pollution source locations; the one-dimensional Saint-Venant equation is used to simulate the fluid movement in the pipeline to describe the velocity and flow changes of the water flow in the pipeline. The expression is: ; ; Where: Q is the instantaneous flow rate in the pipe; A is the flow cross-sectional area; H is the water head height; g is the gravitational acceleration; Simulate the diffusion and mixing of pollutants in the pipe network based on the convection-diffusion equation: ; Where: C is the pollutant concentration; u is the water flow velocity, D is the diffusion coefficient; R(C) is the reaction term for pollutant degradation or transformation; For each known pollution source, simulate its emission characteristics separately, including emission intensity, emission duration and emission cycle; output the pollution concentration contribution value of the single source at each node in the pipeline network, recorded as ; In the same time and space range, the emission concentrations of all pollution sources are superimposed to calculate the total pollution concentration: ; If the monitoring value of a node Higher than , indicating the presence of unidentified potential sources of contamination; For unidentified potential pollution sources, a nonlinear correction factor f(x, t) is introduced to simulate the nonlinear superposition behavior of pollutants in complex pipe networks: ; In the formula, f(x,t) is obtained by fitting the historical monitoring data and the flow-concentration relationship; It represents the pollutant concentration distribution after the nonlinear superposition effect correction at the spatial position x and time t; the emission characteristics of the pollution source are calculated using the inversion algorithm: ; In the formula, It represents the simulated concentration distribution calculated by the pollutant diffusion model at the spatial position x and time t based on the assumed pollution source location and emission characteristics. The calculation formula is: ; is the emission flow of the ith potential pollution source; is the emission concentration of the i-th pollution source; is the length of the pipeline from the pollution source to the monitoring point; D is the diffusion coefficient; is the Gaussian diffusion function, which describes the diffusion process of pollutants; Calculate the potential location of pollution sources based on the pollutant concentration gradient: If the concentration at the downstream monitoring point is higher than that at the upstream point, locate the potential pollution source by combining the flow rate and concentration changes; identify the key nodes along the path where the pollutant concentration increases suddenly, and assign weights to the nodes. Assess the contribution of each pollution source to the total concentration: ; In the formula, n is the total number of pollution sources. If the weight of a pollution source is higher than that of other pollution sources, it is marked as a potential pollution source; S5: Combined with geographic information system and drainage basin model, based on the pollutant diffusion path and flow characteristics of drainage network, the pollutant propagation process is simulated and the geographical location of pollution source is located; S6: Output the pollution source tracing results to the user terminal, generate a pollution source tracing report, and provide pollution control suggestions and alarm signals based on the tracing results.

2. The pollution source tracing method based on drainage system water quality and quantity monitoring and analysis according to claim 1 is characterized by: In S5, based on the pollutant diffusion path and the flow characteristics of the drainage network, the pollutant propagation process is simulated and the geographical location of the pollution source is located, specifically: Divide the drainage network into several sections according to flow direction , each section corresponds to a monitoring point; the pollutant concentration contribution of each section is quantitatively calculated: ;in: For path segment Contribution rate to the total concentration of pollutants at the monitoring point, For path segment Traffic volume; For path segment Input pollutant concentration; for continuous path segments and , calculate the gradient change of contribution rate: ; Among them, PCG is segment and The pollution contribution rate gradient between them; , For segment and spatial distance.

3. The pollution source tracing method based on drainage system water quality and quantity monitoring and analysis according to claim 2 is characterized by: For each monitoring point , calculate the pollutants from the upstream node Time delay of propagation to monitoring point , the expression is: ;in: Upstream node To the monitoring point distance; u is the water flow speed; is the pollutant concentration; k is the diffusion rate coefficient; D is the pipe diameter; the diffusion hysteresis coefficient is obtained by normalizing the lag time, and the expression is: ;in: is the maximum delay time in the drainage network.

4. The pollution source tracing method based on drainage system water quality and quantity monitoring and analysis according to claim 3 is characterized by: The K-Means clustering algorithm is applied to the joint distribution of the pollution contribution gradient PCG and the diffusion hysteresis coefficient DLC to classify the areas in the drainage network. Specifically, for each monitoring point i, a two-dimensional feature vector is formed: ; Normalize the features: ;in: is the characteristic mean; σX is the characteristic standard deviation; The K-Means clustering algorithm is used to classify the standardized PCG and DLC data: K-Means divides the data points into k clusters by minimizing the Euclidean distance, which is expressed as: in: is the jth cluster; is the centroid of the jth cluster; Use the elbow rule to determine the optimal number of clusters k: gradually increase the value of k, calculate the sum of squared errors within the cluster under each k; find the point where the error decreases significantly slower as the optimal k value; The monitoring points are divided into k clusters, each cluster corresponds to a pollution characteristic area: high PCG, high DLC: areas with severe distal diffusion; high PCG, low DLC: areas with potential pollution sources; low PCG, high DLC: downstream areas affected by dilution; low PCG, low DLC: background pollution areas; A heat map of the joint distribution of PCG and DLC is generated based on GIS: high PCG and low DLC areas are marked as red high priority areas on the map; medium PCG and DLC areas are marked as yellow medium priority areas; low PCG and high DLC areas are marked as green low priority areas; in the red high priority areas, the location of the pollution source is confirmed by combining historical emission data, pipeline network structure and actual investigation.

5. The pollution source tracing method based on drainage system water quality and quantity monitoring and analysis according to claim 4 is characterized by: In S6, an alarm signal is generated according to the traceability result, specifically: The mean μ and standard deviation σ of PCG and DLC data of all monitoring points are calculated respectively; PCG and DLC of each monitoring point are standardized to the degree of deviation, and the expression is: ; In the formula, is the PCG deviation degree of monitoring point i, is the DLC deviation degree of monitoring point i, is the mean value of PCG, is the standard deviation of PCG; is the mean value of DLC, is the standard deviation of DLC; Based on the degree of deviation between PCG and DLC, define the warning level rules: Level 1 warning: ; The concentration changes steadily and the diffusion hysteresis is normal, indicating slight pollution and no emergency response is required; Level 2 warning: or ; Significant changes in concentration or large diffusion lag indicate moderate pollution, requiring enhanced monitoring and preliminary investigation of pollution sources; Level 3 warning or Or both exceed the standard at the same time; the concentration changes dramatically and the diffusion lag is serious, indicating that the peak concentration of pollutants is abnormal or the propagation speed is fast, and emergency response measures need to be initiated immediately.

6. A pollution source tracing system based on monitoring and analysis of water quality and quantity of a drainage system, used to implement the pollution source tracing method based on monitoring and analysis of water quality and quantity of a drainage system according to any one of claims 1 to 5, characterized in that: It includes data acquisition module, data processing module, correlation analysis module, nonlinear superposition effect correction module, pollution propagation simulation and positioning module and early warning module: Data acquisition module: collects water quality parameters and water quantity parameters in the drainage system in real time through a distributed sensor network. Water quality parameters include pH value, dissolved oxygen, chemical oxygen demand, ammonia nitrogen concentration and total phosphorus concentration, and water quantity parameters include flow rate and flow rate; Data processing module: denoises the collected water quality parameters and water quantity parameters, and filters non-pollution source signals through anomaly detection algorithms, including abnormal data caused by sensor errors and sudden sediment disturbances; Correlation analysis module: Use multivariate analysis methods to dynamically analyze the processed data to identify abnormal changes in pollutant concentrations in the drainage system, and determine the correlation between pollutant concentration changes and discharge periods and regions through spatiotemporal correlation analysis models; Nonlinear superposition effect correction module: Construct a nonlinear superposition effect correction model to identify the potential location of the real pollution source by simulating the superposition effect of emissions from different pollution sources in a complex drainage network; Pollution propagation simulation and positioning module: Combining geographic information system and drainage basin model, based on the pollutant diffusion path and flow characteristics of the drainage network, it simulates the pollutant propagation process and locates the geographical location of the pollution source; Early warning module: outputs pollution source tracing results to the user terminal, generates pollution source tracing reports, and provides pollution control suggestions and alarm signals based on the tracing results.

Images