River pollutant tracing method and system

Abstract

The invention provides a river pollutant traceability method and a river pollutant traceability system. The method comprises the following steps: constructing a potential pollution source feature fingerprint database fused with an LSTM time sequence feature extraction model, carrying out abnormal water quality fingerprint identification on a monitored river reach, if identification is abnormal, collecting an upstream water sample, detecting to obtain a water quality fingerprint, comparing the water quality fingerprint with the abnormal water quality fingerprint, determining a target river reach according to a comparison result, and determining the target river reach according to the comparison result. The method comprises the following steps: sampling all enterprises of a target river reach in real time, comparing detection result data with detection result data of an abnormal water sample, determining potential pollution sources according to a comparison result, determining high-matching candidate sources from the potential pollution sources by utilizing an LSTM (Long Short Term Memory) time sequence feature extraction model, carrying out space transition verification on the high-matching candidate sources, and carrying out space transition verification on the high-matching candidate sources. And determining the pollution source according to the verification result. According to the method, the accuracy, efficiency and result reliability of tracing the river pollutants can be effectively improved, and the scenes of multi-source pollution, intermittent emission and the like of complex rivers can be effectively handled.

Description

Technical Field

[0001] The present application relates to the technical field of river pollutant source tracing, and in particular, to a river pollutant source tracing method and system thereof. Background Art

[0002] Tracing the source of river pollutants is a core technology in water environment management, and various methods have been developed in existing technologies. For example, some studies identify the source of pollutants by collecting stable isotope ratios in water bodies, using the "fingerprint" characteristics of isotopes to distinguish different emission sources. However, such methods often focus on single-dimensional characteristics and are difficult to deal with multi-source mixed pollution scenarios. Other technologies build a pollution source feature database to store the types and concentration ratios of pollutants discharged by enterprises, and use similarity algorithms to compare abnormal water samples with database features to identify the pollution source. However, these technologies ignore the temporal changes in enterprise pollution discharge (such as periodic emissions caused by production cycles and trend fluctuations caused by changes in equipment operating status), resulting in the inability to capture dynamic characteristics such as intermittent factory pollution discharge. This can easily lead to misjudgment or omission due to static feature matching errors. In addition, to locate the scope of pollution, existing technologies often use multi-point sampling along the river and trace the source through concentration gradient analysis. However, this method requires sampling and comparison section by section, which is particularly time-consuming and labor-intensive in river sections with complex tributaries. The investigation is inefficient and it is difficult to accurately define the boundaries of pollution spread. At the same time, the verification methods of existing technologies are relatively simple, mostly relying on feature matching of static data, lacking spatial verification of the dynamic migration process of pollutants, and making it difficult to confirm the direct causal relationship between pollution sources and pollution incidents.

[0003] In summary, existing technologies have problems such as insufficient multi-dimensional feature fusion, lack of temporal regularity mining, inefficient pollution range positioning, and single verification methods. They are unable to meet the needs of rapid and accurate tracing of pollutants in complex rivers. Therefore, there is an urgent need for an efficient tracing method that integrates temporal feature extraction, intelligent hierarchical investigation and spatial trajectory verification. Summary of the Invention

[0004] The purpose of this application is to provide a river pollutant tracing method and system, which can solve the problems of insufficient multi-dimensional feature fusion, lack of temporal regularity mining, inefficient pollution range positioning and single verification means in the existing technology, and form a complete technical chain of "feature storage-anomaly identification-range locking-source screening-verification and confirmation". By integrating temporal feature extraction, intelligent hierarchical investigation and spatial trajectory verification, the accuracy, efficiency and reliability of river pollutant tracing are significantly improved, and it can effectively deal with scenarios such as multi-source pollution in complex rivers and intermittent emissions.

[0005] This application also provides a method for tracing the source of river pollutants, comprising the following steps: Build a potential pollution source feature fingerprint library integrated with the LSTM time series feature extraction model; Conduct abnormal water quality fingerprint identification in monitored river sections; If an abnormality is identified, water samples from upstream are collected and tested to obtain a water quality fingerprint, which is then compared with the abnormal water quality fingerprint. The target river section is determined based on the comparison results. Conduct real-time sampling of all enterprises in the target river section, compare the test results with the test results of abnormal water samples, and identify potential pollution sources based on the comparison results; Determine high-matching candidate sources from potential pollution sources using the LSTM time series feature extraction model; Perform spatial shift verification on high-matching candidate sources and confirm the pollution source based on the verification results.

[0006] Optionally, in the river pollutant source tracing method described in the present application, the construction of a potential pollution source feature fingerprint library integrated with the LSTM time series feature extraction model includes: Collect pollutant discharge samples from enterprises in the monitored river sections at different discharge times to obtain pollutant characteristic data, including pollutant types, pollutant concentration ratios, pollutant concentration time series data, and stable isotope ratios; The preset LSTM algorithm is used to train the pollutant concentration time series data in the sample, extract the time series feature vector, including emission periodicity feature data and trend feature data, and obtain the LSTM time series feature extraction model; The pollutant characteristic data, time series feature vectors and LSTM time series feature extraction model are stored in the database to generate a potential pollution source feature fingerprint library.

[0007] Optionally, in the river pollutant source tracing method described in this application, the abnormal water quality fingerprint identification of the monitored river section includes: The water quality fingerprint of the monitored river section is collected in real time, and the fluorescence peak position offset, new peak intensity ratio and abnormal duration are obtained by combining the water quality fingerprint of normal water quality; If the fluorescence peak position offset is greater than the preset offset threshold, the newly added peak intensity ratio is greater than the preset intensity ratio threshold, and the abnormality duration is greater than the preset time threshold, the water quality is determined to be abnormal.

[0008] Optionally, in the river pollutant source tracing method described in the present application, if an abnormality is identified, collecting upstream water samples and testing to obtain a water quality fingerprint, comparing it with the abnormal water quality fingerprint, and determining the target river section based on the comparison result, including: If an abnormality is identified, the section where the water quality is abnormal is taken as the starting point, and the water quality fingerprint of the upstream section is collected and compared with the water quality fingerprint of the abnormal water quality section; If the comparison is successful, continue to push sampling upstream; If the comparison is unsuccessful, water samples are collected again at the dichotomous position between the upstream section and the section with abnormal water quality and compared with the water quality fingerprint of the section with abnormal water quality; The same cycle of pollution source tracing and investigation is carried out based on the water quality fingerprint comparison results until the similarity change rate of the water quality fingerprints of two adjacent sampling points exceeds the preset change rate threshold; The river section between two adjacent sampling points was taken as the target river section.

[0009] Optionally, in the river pollutant source tracing method described in the present application, real-time sampling is performed on all enterprises in the target river section, the test result data is compared with the test result data of abnormal water samples, and the potential pollution source is determined based on the comparison results, including: The test result data include pollutant types, pollutant concentration ratios and stable isotope ratios; Compare the pollutant types, pollutant concentration ratios and stable isotope ratios of enterprise water samples and abnormal water samples one by one; Enterprises whose pollutant types, pollutant concentration ratios and stable isotope ratios were successfully matched were selected as potential pollution sources.

[0010] Optionally, in the river pollutant source tracing method described in the present application, determining highly matching candidate sources from potential pollution sources using the LSTM time series feature extraction model includes: Obtain pollutant concentration time series data of potential pollution source water samples and abnormal water samples, and input them into the LSTM time series feature extraction model to obtain emission periodicity feature data and trend feature data; The discharge periodic characteristic data and trend characteristic data of the water samples of potential pollution sources are compared with those of abnormal water samples, and the potential pollution sources that have been successfully matched are regarded as high-match candidate sources.

[0011] Optionally, in the river pollutant source tracing method described in the present application, performing spatial shift verification on highly matching candidate sources and confirming the pollution source based on the verification results include: Starting from the location of the high-matching candidate source, concentration time series data of each section along the downstream of the river are collected; The pollutant concentration time series data, pollutant types, pollutant concentration ratios, and stable isotope ratios of potential pollution sources are input into the pre-built hydrodynamic-water quality coupling model to generate pollutant concentration time series data at different spatial points. These data are then compared with the concentration time series data of each section actually monitored. If the comparison is successful, the high-match candidate source is determined to be a pollution source.

[0012] In a second aspect, the present application provides a river pollutant source tracing system, the system comprising: a memory and a processor, wherein the memory stores a program for a river pollutant source tracing method, and when the program for the river pollutant source tracing method is executed by the processor, the following steps are implemented: Build a potential pollution source feature fingerprint library integrated with the LSTM time series feature extraction model; Conduct abnormal water quality fingerprint identification in monitored river sections; If an abnormality is identified, water samples from upstream are collected and tested to obtain a water quality fingerprint, which is then compared with the abnormal water quality fingerprint. The target river section is determined based on the comparison results. Conduct real-time sampling of all enterprises in the target river section, compare the test results with the test results of abnormal water samples, and identify potential pollution sources based on the comparison results; Determine high-matching candidate sources from potential pollution sources using the LSTM time series feature extraction model; Perform spatial shift verification on high-matching candidate sources and confirm the pollution source based on the verification results.

[0013] Optionally, in the river pollutant source tracing system described in the present application, the construction of the potential pollution source feature fingerprint library integrated with the LSTM time series feature extraction model includes: Collect pollutant discharge samples from enterprises in the monitored river sections at different discharge times to obtain pollutant characteristic data, including pollutant types, pollutant concentration ratios, pollutant concentration time series data, and stable isotope ratios; The preset LSTM algorithm is used to train the pollutant concentration time series data in the sample, extract the time series feature vector, including emission periodicity feature data and trend feature data, and obtain the LSTM time series feature extraction model; The pollutant characteristic data, time series feature vectors and LSTM time series feature extraction model are stored in the database to generate a potential pollution source feature fingerprint library.

[0014] Optionally, in the river pollutant tracing system described in this application, the abnormal water quality fingerprint identification of the monitored river section includes: The water quality fingerprint of the monitored river section is collected in real time, and the fluorescence peak position offset, new peak intensity ratio and abnormal duration are obtained by combining the water quality fingerprint of normal water quality; If the fluorescence peak position offset is greater than the preset offset threshold, the newly added peak intensity ratio is greater than the preset intensity ratio threshold, and the abnormality duration is greater than the preset time threshold, the water quality is determined to be abnormal.

[0015] From the above, it can be seen that the river pollutant tracing method and system provided by this application systematically constructs a potential pollution source fingerprint library that integrates static characteristics (pollutant types, concentration ratios, stable isotope ratios) and dynamic time series laws (emission periodicity characteristics, trend characteristics). During the pollution incident period, it relies on multi-dimensional fingerprint information (covering pollutant composition, quantitative ratios, isotope identification and time series feature vectors) to conduct comprehensive comparison, and combines the hydrodynamic-water quality coupling model to analyze and deduce the migration trajectory of pollutants. Even in the face of complex scenarios such as the coexistence of multiple suspected enterprises and intermittent pollution discharge, this application can still effectively eliminate interference sources and accurately lock the real pollution source, solving the problems of low tracing efficiency and insufficient matching accuracy of traditional methods in complex pollution scenarios.

[0016] Other features and advantages of the present application will be described in the following description, and in part will become apparent from the description, or understood by practicing the embodiments of the present application. The objectives and other advantages of the present application can be achieved and obtained through the structures particularly pointed out in the written description and the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without creative work.

[0018] Figure 1 A flow chart of a method for tracing the source of river pollutants provided in an embodiment of the present application; Figure 2 A flow chart of the method for tracing the source of river pollutants provided in an embodiment of the present application for constructing a potential pollution source feature fingerprint library; Figure 3 A flow chart for confirming pollution sources in the river pollutant tracing method provided in an embodiment of the present application. DETAILED DESCRIPTION

[0019] The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work fall within the scope of protection of the present application.

[0020] It should be noted that similar reference numerals and letters represent similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined or explained in subsequent drawings. At the same time, in the description of this application, the terms "first", "second", etc. are only used to distinguish the description and should not be understood as indicating or implying relative importance.

[0021] Please refer to Figure 1 , Figure 1 This is a flow chart of a river pollutant source tracing method in some embodiments of the present application. The river pollutant source tracing method is used in a terminal device, such as a computer, a mobile phone terminal, etc. The river pollutant source tracing method includes the following steps: S11. Build a potential pollution source feature fingerprint library integrating LSTM time series feature extraction model; S12. Conduct fingerprint identification of abnormal water quality in the monitored river section; S13. If an abnormality is identified, collect water samples from upstream and test to obtain a water quality fingerprint, compare it with the abnormal water quality fingerprint, and determine the target river section based on the comparison result; S14. Conduct real-time sampling of all enterprises in the target river section, compare the test results with the test results of abnormal water samples, and identify potential pollution sources based on the comparison results; S15, using the LSTM time series feature extraction model to determine a high-matching candidate source from potential pollution sources; S16. Perform spatial shift verification on the highly matched candidate sources and confirm the pollution source based on the verification results.

[0022] It should be noted that this application realizes the all-round storage of enterprise pollution discharge characteristics by constructing a potential pollution source feature fingerprint library that integrates the LSTM (long short-term memory network) time series feature extraction model - not only static features such as pollutant types, concentration ratios, stable isotope ratios, etc., but also dynamic time series vectors such as emission periodicity and trend characteristics are extracted through the LSTM algorithm to provide a multi-dimensional benchmark for traceability; through the joint judgment of fluorescence peak position offset, new peak intensity ratio and abnormal duration, accurate identification of water quality anomalies is achieved to avoid misjudgment of a single indicator; the intelligent strategy of "upstream promotion sampling + dichotomy supplementary sampling" is adopted The target river section is roughly located, and the threshold of the change rate of the water quality fingerprint similarity of adjacent sampling points is combined to quickly narrow the scope of pollution and greatly improve the efficiency of investigation. After screening potential pollution sources based on the triple comparison of pollutant types, concentration ratios, and stable isotope ratios, high-match candidate sources are further determined through LSTM time series feature vector comparison, improving the screening accuracy from both static characteristics and dynamic laws. Finally, the river channel refined hydrodynamic-water quality coupling model is used to simulate the migration trajectory of pollutants from candidate sources to monitoring sections, and the simulated concentration time series data is compared and verified with the actual monitoring data to establish a direct causal relationship between the pollution source and the pollution incident. A complete traceability chain of "feature storage-anomaly identification-range locking-source screening-verification and confirmation" is formed to solve the problems of missing time series features, low matching accuracy, and long traceability cycle of traditional methods, and to achieve accurate and efficient tracing of river pollutants. It is especially suitable for rivers in industrial clusters and can efficiently trace complex scenarios such as intermittent illegal discharge and multi-source complex pollution.

[0023] Please refer to Figure 2 , Figure 2 This is a flow chart of constructing a potential pollution source feature fingerprint library for a river pollutant source tracing method in some embodiments of the present application. According to an embodiment of the present invention, constructing a potential pollution source feature fingerprint library integrated with an LSTM time series feature extraction model includes: S21. Collect pollutant discharge samples from enterprises in the monitored river sections at different discharge times to obtain pollutant characteristic data, including pollutant types, pollutant concentration ratios, pollutant concentration time series data, and stable isotope ratios; S22. Use a preset LSTM algorithm to train the pollutant concentration time series data in the sample, extract time series feature vectors, including emission periodicity feature data and trend feature data, and obtain an LSTM time series feature extraction model; S23. Store the pollutant feature data, time series feature vectors and LSTM time series feature extraction model into the database to generate a potential pollution source feature fingerprint library.

[0024] It's important to note that by collecting pollution samples from enterprises over different time periods, static features such as pollutant type, concentration ratio, and stable isotope ratio are integrated with time series features (periodicity and trend) extracted by LSTM to generate a fingerprint library of potential pollution source characteristics, providing a multi-dimensional benchmark for subsequent comparisons. Compared to traditional databases that only store static features, this fingerprint library captures the temporal patterns of enterprise pollution emissions, laying the data foundation for accurate pollution source matching, reducing misjudgments caused by ignoring temporal features, and solving the problem of distinguishing similar pollutant emissions from similar enterprises. Pollutant concentration time series data refers to a time series of pollutant concentration values ​​collected at fixed intervals.

[0025] Among them, the emission periodic characteristic data include peak frequency, peak interval, phase and peak-to-base ratio. Peak frequency refers to the number of times the emission concentration peak occurs per unit time, which directly reflects the activity level of periodic emissions and can distinguish between continuous and intermittent pollution discharge. Peak interval refers to the time difference between two adjacent emission peaks (such as a peak occurs every 8 hours). It is the core quantitative indicator of periodicity and can accurately identify fixed interval patterns such as daily cycles and weekly cycles. Phase refers to the relative time position of the peak within the cycle (such as a peak occurs at 12 o'clock every day), which can distinguish the "time imprint" of emissions from different pollution sources (for example, the peak phase of similar factories may differ by 2 hours due to different production schedules). The peak-to-base ratio refers to the ratio of the peak concentration to the baseline concentration within the cycle (the average concentration during the non-peak period) (such as the peak is 5 times the baseline value), which can reflect the intensity fluctuation amplitude of periodic emissions and enhance feature recognition.

[0026] Trend characteristic data includes peak slope, rate of change, trend duration, and mutation points. Peak slope refers to the rate at which emission concentration rises from the baseline value to the peak value (e.g., a slope of 4 mg / (L·h) for a rise from 1 mg / L to 5 mg / L within one hour). It can reflect the "startup characteristics" of the pollution discharge process (e.g., the difference in slope between instantaneous and slowly accumulated emissions). Rate of change refers to the rate of change in concentration over time within the overall trend (e.g., a daily increase of 0.2 mg / L). It quantifies the "steepness" of the trend and distinguishes between slow deterioration and sudden exceedance. Trend duration refers to the duration of the same trend (increasing / decreasing) (e.g., a decreasing trend for 10 consecutive days). It can eliminate short-term interference (e.g., instantaneous leaks) and focus on long-term stable emission characteristics. Mutation points are the time points when the concentration trend significantly changes (e.g., a sudden jump from 2 mg / L to 8 mg / L on the 5th of a certain month). They can be linked to abnormal events at the pollution source (e.g., equipment failure, illegal discharge), enhancing the timeliness of source tracing.

[0027] According to an embodiment of the present invention, the abnormal water quality fingerprint identification of the monitored river section includes: The water quality fingerprint of the monitored river section is collected in real time, and the fluorescence peak position offset, new peak intensity ratio and abnormal duration are obtained by combining the water quality fingerprint of normal water quality; If the fluorescence peak position offset is greater than the preset offset threshold, the newly added peak intensity ratio is greater than the preset intensity ratio threshold, and the abnormality duration is greater than the preset time threshold, the water quality is determined to be abnormal.

[0028] It should be noted that the multi-parameter judgment of abnormal water quality based on the fluorescence peak position offset, the new peak intensity ratio and the duration of the abnormality can capture water quality abnormalities more sensitively and accurately than the traditional single indicator identification (such as the concentration of a certain pollutant exceeds the standard), reduce the risk of misjudgment due to background fluctuations or accidental interference, and achieve rapid locking of pollution events.

[0029] According to an embodiment of the present invention, if an abnormality is identified, upstream water samples are collected and tested to obtain a water quality fingerprint, which is compared with the abnormal water quality fingerprint, and the target river section is determined based on the comparison result, including: If an abnormality is identified, the section where the water quality is abnormal is taken as the starting point, and the water quality fingerprint of the upstream section is collected and compared with the water quality fingerprint of the abnormal water quality section; If the comparison is successful, continue to push sampling upstream; If the comparison is unsuccessful, water samples are collected again at the dichotomous position between the upstream section and the section with abnormal water quality and compared with the water quality fingerprint of the section with abnormal water quality; The same cycle of pollution source tracing and investigation is carried out based on the water quality fingerprint comparison results until the similarity change rate of the water quality fingerprints of two adjacent sampling points exceeds the preset change rate threshold; The river section between two adjacent sampling points was taken as the target river section.

[0030] It should be noted that by locating the target river section through the strategy of "upstream advance sampling + binary supplementary sampling" and combining the threshold of the change rate of water quality fingerprint similarity of adjacent sampling points, the scope of pollution sources can be quickly narrowed down. Compared with traditional non-targeted full-river section sampling, this method greatly reduces the sampling volume and analysis costs, and improves the accuracy and efficiency of positioning the target river section.

[0031] According to an embodiment of the present invention, real-time sampling is performed on all enterprises in the target river section, the test result data is compared with the test result data of abnormal water samples, and the potential pollution source is determined based on the comparison results, including: The test result data include pollutant types, pollutant concentration ratios and stable isotope ratios; Compare the pollutant types, pollutant concentration ratios and stable isotope ratios of enterprise water samples and abnormal water samples one by one; Enterprises whose pollutant types, pollutant concentration ratios and stable isotope ratios were successfully matched were selected as potential pollution sources.

[0032] It should be noted that by screening potential pollution sources through triple comparison of pollutant types, concentration ratios, and stable isotope ratios, compared with single indicator comparison, multi-dimensional feature matching can significantly improve the specificity of pollution source screening (such as stable isotope ratios have "fingerprint" characteristics and are difficult to tamper with artificially), reducing misjudgments caused by similar pollutant emissions from different enterprises.

[0033] According to an embodiment of the present invention, determining a high-matching candidate source from potential pollution sources using the LSTM time series feature extraction model includes: Obtain pollutant concentration time series data of potential pollution source water samples and abnormal water samples, and input them into the LSTM time series feature extraction model to obtain emission periodicity feature data and trend feature data; The discharge periodic characteristic data and trend characteristic data of the water samples of potential pollution sources are compared with those of abnormal water samples, and the potential pollution sources that have been successfully matched are regarded as high-match candidate sources.

[0034] It's important to note that the LSTM model extracts and compares time series features, such as emission periodicity and trends, between potential pollution sources and abnormal water samples, further screening for highly compatible candidate sources based on their temporal patterns. Compared to screening based solely on static features, supplementing these temporal features can capture behavioral patterns in enterprise pollution discharges (such as periodic emissions due to production shifts), eliminate interference sources that don't align with the temporal patterns of abnormal events, and improve the reliability of candidate sources.

[0035] Please refer to Figure 3 , Figure 3 This is a flow chart of the method for confirming the pollution source of river pollutant source tracing in some embodiments of the present application. According to the embodiment of the present invention, the spatial shift verification of the high-matching candidate source and the confirmation of the pollution source based on the verification results include: S31. Starting from the location of the high-matching candidate source, collect concentration time series data at each section along the downstream of the river; S32. Input the pollutant concentration time series data, pollutant types, pollutant concentration ratios, and stable isotope ratios of potential pollution sources into a pre-built hydrodynamic-water quality coupling model to generate pollutant concentration time series data at different spatial points, and compare them with the concentration time series data of each section actually monitored; S33: If the comparison is successful, the high-match candidate source is determined to be a pollution source.

[0036] It should be noted that the pollutant migration trajectory is simulated through the river hydrodynamic-water quality coupling model, and the concentration time series data (concentration change curve over time) and static characteristics of the highly matched candidate source are input into the model. The model is used to generate the concentration change curve (concentration time series data) of the pollutant at different spatial points (such as 1km, 3km, and 5km downstream of the candidate source). These curves are compared with the actual monitored concentration time series data of each section, focusing on the spatial order of the concentration peaks (whether they decrease along the direction of water flow) and whether the decay rate is consistent with the model prediction. If the trends of the two are consistent, it means that the spatial migration path of the pollutant is consistent with the emission logic of the candidate source. Through "theoretical simulation + actual data verification", the misjudgment of "similar features but different sources" caused by relying solely on feature comparison is avoided, ensuring that the pollution source finally locked in has a clear spatial migration correlation with the pollution incident, and improving the accuracy of the traceability results. For example: Enterprise A matches the water quality fingerprint, static characteristics and dynamic time series vector of the polluted water body, but the river hydrodynamic-water quality coupling model shows that its emissions take 5 hours to reach the pollution point, and the actual pollution breaks out within 1 hour. The actual pollution source is the unregistered enterprise B upstream.

[0037] The refined hydrodynamic-water quality coupling model of the river channel can simultaneously simulate hydrodynamic parameters and pollutant migration trajectories. When constructing the model, it first uses high-precision topographic data of the monitored river section (including cross-sections, underwater DEM and roughness zoning), hydrological data (upstream flow, downstream water level), pollution source data (outlet location, discharge volume and time series characteristics) and meteorological data, and adopts a two-dimensional or three-dimensional hydrodynamic model (such as the shallow water equation) to simulate the river flow field (flow velocity, water depth, turbulence intensity). The water quality module is synchronously coupled to simulate the convection, diffusion and degradation process of pollutants. The key parameters such as roughness and degradation coefficient are calibrated by measured water level, flow velocity and pollutant concentration data, so that the deviation between the model simulation results and the measured data is controlled within the preset threshold (such as water level error ≤10%), and finally a coupling model that can accurately reproduce the pollutant migration trajectory is formed.

[0038] According to an embodiment of the present invention, the further embodiment includes: Extract historical pollutant characteristic data of pollution sources from the potential pollution source characteristic fingerprint database, including pollutant types, pollutant concentration ratios, stable isotope ratios and time series characteristic vectors; Compare historical pollutant characteristic data with the real-time pollutant characteristic data of the current pollution source to obtain the pollutant type matching degree, pollutant concentration ratio deviation rate, stable isotope ratio deviation rate and time series characteristic vector cosine similarity; Calculate the weighted average of the pollutant concentration ratio deviation rate and the stable isotope ratio deviation rate to determine whether the calculation result is less than or equal to the preset deviation rate threshold; Perform weighted average calculation on the pollutant type matching degree and the cosine similarity of the time series feature vector to determine whether the calculation result is greater than or equal to the preset matching degree threshold; If any of the above judgments is no, a multi-dimensional investigation and prompt will be conducted on the potential pollution source.

[0039] It should be noted that when calling the historical pollutant characteristic data in the potential pollution source characteristic fingerprint library and comparing and analyzing the "historical emission characteristics" and "current emission characteristics", if the difference is too large, it is necessary to further investigate whether the enterprise has made process changes, illegal emissions, etc.

[0040] Among them, the stable isotope ratio deviation rate can be calculated by (real-time value - historical value) / historical value × 100%; the pollutant type matching degree is expressed as the ratio of the number of characteristic pollutant types shared by real-time data and historical data to the number of all characteristic pollutant types in historical data; pollutant concentration ratio deviation rate: for multiple characteristic pollutants, calculate (real-time ratio - historical ratio) / historical ratio × 100% separately, and add up the calculation results of all characteristic pollutants to calculate the average.

[0041] The present invention also discloses a river pollutant source tracing system, comprising a memory and a processor. The memory stores a river pollutant source tracing method program, and when the river pollutant source tracing method program is executed by the processor, the following steps are implemented: Build a potential pollution source feature fingerprint library integrated with the LSTM time series feature extraction model; Conduct abnormal water quality fingerprint identification in monitored river sections; If an abnormality is identified, water samples from upstream are collected and tested to obtain a water quality fingerprint, which is then compared with the abnormal water quality fingerprint. The target river section is determined based on the comparison results. Conduct real-time sampling of all enterprises in the target river section, compare the test results with the test results of abnormal water samples, and identify potential pollution sources based on the comparison results; Determine high-matching candidate sources from potential pollution sources using the LSTM time series feature extraction model; Perform spatial shift verification on high-matching candidate sources and confirm the pollution source based on the verification results.

[0042] It should be noted that this application realizes the all-round storage of enterprise pollution discharge characteristics by constructing a potential pollution source feature fingerprint library that integrates the LSTM (long short-term memory network) time series feature extraction model - not only static features such as pollutant types, concentration ratios, stable isotope ratios, etc., but also dynamic time series vectors such as emission periodicity and trend characteristics are extracted through the LSTM algorithm to provide a multi-dimensional benchmark for traceability; through the joint judgment of fluorescence peak position offset, new peak intensity ratio and abnormal duration, accurate identification of water quality anomalies is achieved to avoid misjudgment of a single indicator; the intelligent strategy of "upstream promotion sampling + dichotomy supplementary sampling" is adopted The target river section is roughly located, and the threshold of the change rate of the water quality fingerprint similarity of adjacent sampling points is combined to quickly narrow the scope of pollution and greatly improve the efficiency of investigation. After screening potential pollution sources based on the triple comparison of pollutant types, concentration ratios, and stable isotope ratios, high-match candidate sources are further determined through LSTM time series feature vector comparison, improving the screening accuracy from both static characteristics and dynamic laws. Finally, the river channel refined hydrodynamic-water quality coupling model is used to simulate the migration trajectory of pollutants from candidate sources to monitoring sections, and the simulated concentration time series data is compared and verified with the actual monitoring data to establish a direct causal relationship between the pollution source and the pollution incident. A complete traceability chain of "feature storage-anomaly identification-range locking-source screening-verification and confirmation" is formed to solve the problems of missing time series features, low matching accuracy, and long traceability cycle of traditional methods, and to achieve accurate and efficient tracing of river pollutants. It is especially suitable for rivers in industrial clusters and can efficiently trace complex scenarios such as intermittent illegal discharge and multi-source complex pollution.

[0043] According to an embodiment of the present invention, the construction of a potential pollution source feature fingerprint library integrating an LSTM time series feature extraction model includes: Collect pollutant discharge samples from enterprises in the monitored river sections at different discharge times to obtain pollutant characteristic data, including pollutant types, pollutant concentration ratios, pollutant concentration time series data, and stable isotope ratios; The preset LSTM algorithm is used to train the pollutant concentration time series data in the sample, extract the time series feature vector, including emission periodicity feature data and trend feature data, and obtain the LSTM time series feature extraction model; The pollutant characteristic data, time series feature vectors and LSTM time series feature extraction model are stored in the database to generate a potential pollution source feature fingerprint library.

[0044] It's important to note that by collecting pollution samples from enterprises over different time periods, static features such as pollutant type, concentration ratio, and stable isotope ratio are integrated with time series features (periodicity and trend) extracted by LSTM to generate a fingerprint library of potential pollution source characteristics, providing a multi-dimensional benchmark for subsequent comparisons. Compared to traditional databases that only store static features, this fingerprint library captures the temporal patterns of enterprise pollution emissions, laying the data foundation for accurate pollution source matching, reducing misjudgments caused by ignoring temporal features, and solving the problem of distinguishing similar pollutant emissions from similar enterprises. Pollutant concentration time series data refers to a time series of pollutant concentration values ​​collected at fixed intervals.

[0045] Among them, the emission periodic characteristic data include peak frequency, peak interval, phase and peak-to-base ratio. Peak frequency refers to the number of times the emission concentration peak occurs per unit time, which directly reflects the activity level of periodic emissions and can distinguish between continuous and intermittent pollution discharge. Peak interval refers to the time difference between two adjacent emission peaks (such as a peak occurs every 8 hours). It is the core quantitative indicator of periodicity and can accurately identify fixed interval patterns such as daily cycles and weekly cycles. Phase refers to the relative time position of the peak within the cycle (such as a peak occurs at 12 o'clock every day), which can distinguish the "time imprint" of emissions from different pollution sources (for example, the peak phase of similar factories may differ by 2 hours due to different production schedules). The peak-to-base ratio refers to the ratio of the peak concentration to the baseline concentration within the cycle (the average concentration during the non-peak period) (such as the peak is 5 times the baseline value), which can reflect the intensity fluctuation amplitude of periodic emissions and enhance feature recognition.

[0046] Trend characteristic data includes peak slope, rate of change, trend duration, and mutation points. Peak slope refers to the rate at which emission concentration rises from the baseline value to the peak value (e.g., a slope of 4 mg / (L·h) for a rise from 1 mg / L to 5 mg / L within one hour). It can reflect the "startup characteristics" of the pollution discharge process (e.g., the difference in slope between instantaneous and slowly accumulated emissions). Rate of change refers to the rate of change in concentration over time within the overall trend (e.g., a daily increase of 0.2 mg / L). It quantifies the "steepness" of the trend and distinguishes between slow deterioration and sudden exceedance. Trend duration refers to the duration of the same trend (increasing / decreasing) (e.g., a decreasing trend for 10 consecutive days). It can eliminate short-term interference (e.g., instantaneous leaks) and focus on long-term stable emission characteristics. Mutation points are the time points when the concentration trend significantly changes (e.g., a sudden jump from 2 mg / L to 8 mg / L on the 5th of a certain month). They can be linked to abnormal events at the pollution source (e.g., equipment failure, illegal discharge), enhancing the timeliness of source tracing.

[0047] According to an embodiment of the present invention, the abnormal water quality fingerprint identification of the monitored river section includes: The water quality fingerprint of the monitored river section is collected in real time, and the fluorescence peak position offset, new peak intensity ratio and abnormal duration are obtained by combining the water quality fingerprint of normal water quality; If the fluorescence peak position offset is greater than the preset offset threshold, the newly added peak intensity ratio is greater than the preset intensity ratio threshold, and the abnormality duration is greater than the preset time threshold, the water quality is determined to be abnormal.

[0048] It should be noted that the multi-parameter judgment of abnormal water quality based on the fluorescence peak position offset, the new peak intensity ratio and the duration of the abnormality can capture water quality abnormalities more sensitively and accurately than the traditional single indicator identification (such as the concentration of a certain pollutant exceeds the standard), reduce the risk of misjudgment due to background fluctuations or accidental interference, and achieve rapid locking of pollution events.

[0049] According to an embodiment of the present invention, if an abnormality is identified, upstream water samples are collected and tested to obtain a water quality fingerprint, which is compared with the abnormal water quality fingerprint, and the target river section is determined based on the comparison result, including: If an abnormality is identified, the section where the water quality is abnormal is taken as the starting point, and the water quality fingerprint of the upstream section is collected and compared with the water quality fingerprint of the abnormal water quality section; If the comparison is successful, continue to push sampling upstream; If the comparison is unsuccessful, water samples are collected again at the dichotomous position between the upstream section and the section with abnormal water quality and compared with the water quality fingerprint of the section with abnormal water quality; The same cycle of pollution source tracing and investigation is carried out based on the water quality fingerprint comparison results until the similarity change rate of the water quality fingerprints of two adjacent sampling points exceeds the preset change rate threshold; The river section between two adjacent sampling points was taken as the target river section.

[0050] It should be noted that by locating the target river section through the strategy of "upstream advance sampling + binary supplementary sampling" and combining the threshold of the change rate of water quality fingerprint similarity of adjacent sampling points, the scope of pollution sources can be quickly narrowed down. Compared with traditional non-targeted full-river section sampling, this method greatly reduces the sampling volume and analysis costs, and improves the accuracy and efficiency of positioning the target river section.

[0051] According to an embodiment of the present invention, real-time sampling is performed on all enterprises in the target river section, the test result data is compared with the test result data of abnormal water samples, and the potential pollution source is determined based on the comparison results, including: The test result data include pollutant types, pollutant concentration ratios and stable isotope ratios; Compare the pollutant types, pollutant concentration ratios and stable isotope ratios of enterprise water samples and abnormal water samples one by one; Enterprises whose pollutant types, pollutant concentration ratios and stable isotope ratios were successfully matched were selected as potential pollution sources.

[0052] It should be noted that by screening potential pollution sources through triple comparison of pollutant types, concentration ratios, and stable isotope ratios, compared with single indicator comparison, multi-dimensional feature matching can significantly improve the specificity of pollution source screening (such as stable isotope ratios have "fingerprint" characteristics and are difficult to tamper with artificially), reducing misjudgments caused by similar pollutant emissions from different enterprises.

[0053] According to an embodiment of the present invention, determining a high-matching candidate source from potential pollution sources using the LSTM time series feature extraction model includes: Obtain pollutant concentration time series data of potential pollution source water samples and abnormal water samples, and input them into the LSTM time series feature extraction model to obtain emission periodicity feature data and trend feature data; The discharge periodic characteristic data and trend characteristic data of the water samples of potential pollution sources are compared with those of abnormal water samples, and the potential pollution sources that have been successfully matched are regarded as high-match candidate sources.

[0054] It's important to note that the LSTM model extracts and compares time series features, such as emission periodicity and trends, between potential pollution sources and abnormal water samples, further screening for highly compatible candidate sources based on their temporal patterns. Compared to screening based solely on static features, supplementing these temporal features can capture behavioral patterns in enterprise pollution discharges (such as periodic emissions due to production shifts), eliminate interference sources that don't align with the temporal patterns of abnormal events, and improve the reliability of candidate sources.

[0055] According to an embodiment of the present invention, performing spatial shift verification on a highly matching candidate source and confirming the pollution source based on the verification result includes: Starting from the location of the high-matching candidate source, concentration time series data of each section along the downstream of the river are collected; The pollutant concentration time series data, pollutant types, pollutant concentration ratios, and stable isotope ratios of potential pollution sources are input into the pre-built hydrodynamic-water quality coupling model to generate pollutant concentration time series data at different spatial points. These data are then compared with the concentration time series data of each section actually monitored. If the comparison is successful, the high-match candidate source is determined to be a pollution source.

[0056] It should be noted that the pollutant migration trajectory is simulated through the river hydrodynamic-water quality coupling model, and the concentration time series data (concentration change curve over time) and static characteristics of the highly matched candidate source are input into the model. The model is used to generate the concentration change curve (concentration time series data) of the pollutant at different spatial points (such as 1km, 3km, and 5km downstream of the candidate source). These curves are compared with the actual monitored concentration time series data of each section, focusing on the spatial order of the concentration peaks (whether they decrease along the direction of water flow) and whether the decay rate is consistent with the model prediction. If the trends of the two are consistent, it means that the spatial migration path of the pollutant is consistent with the emission logic of the candidate source. Through "theoretical simulation + actual data verification", the misjudgment of "similar features but different sources" caused by relying solely on feature comparison is avoided, ensuring that the pollution source finally locked in has a clear spatial migration correlation with the pollution incident, and improving the accuracy of the traceability results. For example: Enterprise A matches the water quality fingerprint, static characteristics and dynamic time series vector of the polluted water body, but the river hydrodynamic-water quality coupling model shows that its emissions take 5 hours to reach the pollution point, and the actual pollution breaks out within 1 hour. The actual pollution source is the unregistered enterprise B upstream.

[0057] The refined hydrodynamic-water quality coupling model of the river channel can simultaneously simulate hydrodynamic parameters and pollutant migration trajectories. When constructing the model, it first uses high-precision topographic data of the monitored river section (including cross-sections, underwater DEM and roughness zoning), hydrological data (upstream flow, downstream water level), pollution source data (outlet location, discharge volume and time series characteristics) and meteorological data, and adopts a two-dimensional or three-dimensional hydrodynamic model (such as the shallow water equation) to simulate the river flow field (flow velocity, water depth, turbulence intensity). The water quality module is synchronously coupled to simulate the convection, diffusion and degradation process of pollutants. The key parameters such as roughness and degradation coefficient are calibrated by measured water level, flow velocity and pollutant concentration data, so that the deviation between the model simulation results and the measured data is controlled within the preset threshold (such as water level error ≤10%), and finally a coupling model that can accurately reproduce the pollutant migration trajectory is formed.

[0058] According to an embodiment of the present invention, the further embodiment includes: Extract historical pollutant characteristic data of pollution sources from the potential pollution source characteristic fingerprint database, including pollutant types, pollutant concentration ratios, stable isotope ratios and time series characteristic vectors; Compare historical pollutant characteristic data with the real-time pollutant characteristic data of the current pollution source to obtain the pollutant type matching degree, pollutant concentration ratio deviation rate, stable isotope ratio deviation rate and time series characteristic vector cosine similarity; Calculate the weighted average of the pollutant concentration ratio deviation rate and the stable isotope ratio deviation rate to determine whether the calculation result is less than or equal to the preset deviation rate threshold; Perform weighted average calculation on the pollutant type matching degree and the cosine similarity of the time series feature vector to determine whether the calculation result is greater than or equal to the preset matching degree threshold; If any of the above judgments is no, a multi-dimensional investigation and prompt will be conducted on the potential pollution source.

[0059] It should be noted that when calling the historical pollutant characteristic data in the potential pollution source characteristic fingerprint library and comparing and analyzing the "historical emission characteristics" and "current emission characteristics", if the difference is too large, it is necessary to further investigate whether the enterprise has made process changes, illegal emissions, etc.

[0060] Among them, the stable isotope ratio deviation rate can be calculated by (real-time value - historical value) / historical value × 100%; the pollutant type matching degree is expressed as the ratio of the number of characteristic pollutant types shared by real-time data and historical data to the number of all characteristic pollutant types in historical data; pollutant concentration ratio deviation rate: for multiple characteristic pollutants, calculate (real-time ratio - historical ratio) / historical ratio × 100% separately, and add up the calculation results of all characteristic pollutants to calculate the average.

[0061] The present invention discloses a river pollutant source tracing method and system thereof, which systematically constructs a potential pollution source fingerprint library that integrates static characteristics (pollutant types, concentration ratios, stable isotope ratios) and dynamic time series laws (emission periodicity characteristics, trend characteristics), relies on multi-dimensional fingerprint information (covering pollutant composition, quantitative ratios, isotope identification and time series feature vectors) to conduct comprehensive comparison during the pollution event period, and combines the hydrodynamic-water quality coupling model to analyze and deduce the migration trajectory of pollutants. Even in the face of complex scenarios such as the coexistence of multiple suspected enterprises and intermittent pollution discharge, the present application can still effectively eliminate interference sources and accurately lock the real pollution source, solving the problems of low tracing efficiency and insufficient matching accuracy of traditional methods in complex pollution scenarios.

[0062] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0063] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed across multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the scheme of this embodiment.

[0064] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

Claims

1. A method for tracing the source of river pollutants, characterized in that: The following steps are involved: Build a potential pollution source feature fingerprint library integrated with the LSTM time series feature extraction model; Conduct abnormal water quality fingerprint identification in monitored river sections; If an abnormality is identified, water samples from upstream are collected and tested to obtain a water quality fingerprint, which is then compared with the abnormal water quality fingerprint. The target river section is determined based on the comparison results. Conduct real-time sampling of all enterprises in the target river section, compare the test results with the test results of abnormal water samples, and identify potential pollution sources based on the comparison results; Determine high-matching candidate sources from potential pollution sources using the LSTM time series feature extraction model; Perform spatial shift verification on high-matching candidate sources and confirm the pollution source based on the verification results.

2. The method for tracing the source of river pollutants according to claim 1, characterized in that: The potential pollution source feature fingerprint library constructed by integrating the LSTM time series feature extraction model includes: Collect pollutant discharge samples from enterprises in the monitored river sections at different discharge times to obtain pollutant characteristic data, including pollutant types, pollutant concentration ratios, pollutant concentration time series data, and stable isotope ratios; The preset LSTM algorithm is used to train the pollutant concentration time series data in the sample, extract the time series feature vector, including emission periodicity feature data and trend feature data, and obtain the LSTM time series feature extraction model; The pollutant characteristic data, time series feature vectors and LSTM time series feature extraction model are stored in the database to generate a potential pollution source feature fingerprint library.

3. The method for tracing the source of river pollutants according to claim 2, characterized in that: The abnormal water quality fingerprint identification of the monitored river section includes: The water quality fingerprint of the monitored river section is collected in real time, and the fluorescence peak position offset, new peak intensity ratio and abnormal duration are obtained by combining the water quality fingerprint of normal water quality; If the fluorescence peak position offset is greater than the preset offset threshold, the newly added peak intensity ratio is greater than the preset intensity ratio threshold, and the abnormality duration is greater than the preset time threshold, the water quality is determined to be abnormal.

4. The method for tracing the source of river pollutants according to claim 3, characterized in that: If an abnormality is identified, upstream water samples are collected and tested to obtain a water quality fingerprint, which is compared with the abnormal water quality fingerprint. The target river section is determined based on the comparison results, including: If an abnormality is identified, the section where the water quality is abnormal is taken as the starting point, and the water quality fingerprint of the upstream section is collected and compared with the water quality fingerprint of the abnormal water quality section; If the comparison is successful, continue to push sampling upstream; If the comparison is unsuccessful, water samples are collected again at the dichotomous position between the upstream section and the section with abnormal water quality and compared with the water quality fingerprint of the section with abnormal water quality; The same cycle of pollution source tracing and investigation is carried out based on the water quality fingerprint comparison results until the similarity change rate of the water quality fingerprints of two adjacent sampling points exceeds the preset change rate threshold; The river section between two adjacent sampling points was taken as the target river section.

5. The method for tracing the source of river pollutants according to claim 4, characterized in that: The aforementioned real-time sampling is conducted on all enterprises in the target river section, and the test results are compared with the test results of abnormal water samples. The potential pollution sources are determined based on the comparison results, including: The test result data include pollutant types, pollutant concentration ratios and stable isotope ratios; Compare the pollutant types, pollutant concentration ratios and stable isotope ratios of enterprise water samples and abnormal water samples one by one; Enterprises whose pollutant types, pollutant concentration ratios and stable isotope ratios were successfully matched were selected as potential pollution sources.

6. The method for tracing the source of river pollutants according to claim 5, characterized in that: Determining a high-match candidate source from potential pollution sources using the LSTM time series feature extraction model includes: Obtain pollutant concentration time series data of potential pollution source water samples and abnormal water samples, and input them into the LSTM time series feature extraction model to obtain emission periodicity feature data and trend feature data; The discharge periodic characteristic data and trend characteristic data of the water samples of potential pollution sources are compared with those of abnormal water samples, and the potential pollution sources that have been successfully matched are regarded as high-match candidate sources.

7. The method for tracing the source of river pollutants according to claim 6, characterized in that: The spatial shift verification of the highly matched candidate sources and confirmation of the pollution source based on the verification results include: Starting from the location of the high-matching candidate source, concentration time series data of each section along the downstream of the river are collected; The pollutant concentration time series data, pollutant types, pollutant concentration ratios, and stable isotope ratios of potential pollution sources are input into the pre-built hydrodynamic-water quality coupling model to generate pollutant concentration time series data at different spatial points. These data are then compared with the concentration time series data of each section actually monitored. If the comparison is successful, the high-match candidate source is determined to be a pollution source.

8. A river pollutant tracing system, characterized in that: The system comprises a memory and a processor, wherein the memory stores a program for a method for tracing the source of river pollutants, and when the program for tracing the source of river pollutants is executed by the processor, the following steps are implemented: Build a potential pollution source feature fingerprint library integrated with the LSTM time series feature extraction model; Conduct abnormal water quality fingerprint identification in monitored river sections; If an abnormality is identified, water samples from upstream are collected and tested to obtain a water quality fingerprint, which is then compared with the abnormal water quality fingerprint. The target river section is determined based on the comparison results. Conduct real-time sampling of all enterprises in the target river section, compare the test results with the test results of abnormal water samples, and identify potential pollution sources based on the comparison results; Determine high-matching candidate sources from potential pollution sources using the LSTM time series feature extraction model; Perform spatial shift verification on high-matching candidate sources and confirm the pollution source based on the verification results.

9. The river pollutant tracing system according to claim 8, characterized in that: The potential pollution source feature fingerprint library constructed by integrating the LSTM time series feature extraction model includes: Collect pollutant discharge samples from enterprises in the monitored river sections at different discharge times to obtain pollutant characteristic data, including pollutant types, pollutant concentration ratios, pollutant concentration time series data, and stable isotope ratios; The preset LSTM algorithm is used to train the pollutant concentration time series data in the sample, extract the time series feature vector, including emission periodicity feature data and trend feature data, and obtain the LSTM time series feature extraction model; The pollutant characteristic data, time series feature vectors and LSTM time series feature extraction model are stored in the database to generate a potential pollution source feature fingerprint library.

10. The river pollutant tracing system according to claim 9, characterized in that: The abnormal water quality fingerprint identification of the monitored river section includes: The water quality fingerprint of the monitored river section is collected in real time, and the fluorescence peak position offset, new peak intensity ratio and abnormal duration are obtained by combining the water quality fingerprint of normal water quality; If the fluorescence peak position offset is greater than the preset offset threshold, the newly added peak intensity ratio is greater than the preset intensity ratio threshold, and the abnormality duration is greater than the preset time threshold, the water quality is determined to be abnormal.

Images