Intelligent source tracing monitoring system and method for water quality of rivers and lakes

Abstract

The invention belongs to the technical field of river and lake water quality monitoring, and discloses a river and lake water quality intelligent traceability monitoring system and method.Dissolved organic matter fluorescent fingerprints are analyzed in real time through a three-dimensional fluorescent monitoring unit, pollution markers are detected, and parameters such as COD, ammonia nitrogen and total phosphorus are synchronously detected through a full-spectrum monitoring unit; the video monitoring unit is used for collecting section images / videos of rivers and lakes, other sensors are used for collecting real-time values of chlorophyll, pH, conductivity, turbidity and dissolved oxygen in a water body, then the data processing module is used for carrying out normalization processing on multi-source data, and the analysis and early warning module is used for analyzing water quality early warning scores, spectrum early warning scores and video early warning scores. All scores are synthesized to obtain a comprehensive early warning score, traceability is started when comprehensive early warning is larger than or equal to the grade II, pollution sources are recognized based on three-dimensional fluorescence spectrum comparison and upstream patrol, and a river and lake water body monitoring-early warning-traceability-verification closed loop is achieved.

Description

Technical Field

[0001] This invention belongs to the field of river and lake water quality monitoring technology, specifically relating to an intelligent source tracing monitoring system and method for river and lake water quality. Background Technology

[0002] With increasing attention being paid to environmental issues, river and lake water quality monitoring has become a crucial link in protecting water resources and maintaining ecological balance. Among these, unmanned surface vessel (USV) monitoring equipment for river and lake water quality is the cornerstone of ecological monitoring and the core of efficient environmental governance, playing a vital role.

[0003] Existing unmanned surface vessels (USVs) for water quality monitoring have the following problems: 1. Most mainstream USVs are equipped with conventional five-parameter sensors (such as pH, dissolved oxygen, and turbidity) or single-spectrum devices, lacking the ability to detect dissolved organic matter (DOM) in real time; 2. Mainstream USVs can only monitor and cannot simultaneously support water quality early warning and pollution source tracing functions; 3. Insufficient depth of data analysis: The data systems of most water quality monitoring USVs only realize data collection, transmission, and visualization, lacking the ability to fuse multi-source data and perform intelligent analysis, making it difficult to provide scientific decision-making based on data analysis for pollution source location and river and lake supervision. Summary of the Invention

[0004] The purpose of this invention is to provide an intelligent source tracing and monitoring system and method for river and lake water quality, thereby solving the above-mentioned technical problems.

[0005] The objective of this invention can be achieved through the following technical solutions: A smart source tracing and monitoring system for river and lake water quality, the system comprising: a smart monitoring module, a data processing module, an analysis and early warning module, and a pollution source tracing module; The intelligent monitoring module includes a three-dimensional fluorescence monitoring unit, a full-spectrum monitoring unit, a video monitoring unit, and other sensor monitoring units, used to collect characteristic parameters of river and lake water; The data processing module is used to normalize the collected parameters so that different parameters are at the same standard. The analysis and early warning module is used to analyze the processed data and issue different types of early warning scores based on the analysis results. If the early warning scores of different types reach or exceed the corresponding thresholds, an early warning is issued immediately. The pollution source tracing module is used to identify the source of pollution in the water body at the warning location when the unmanned vessel issues an early warning.

[0006] As a further description of the present invention, the three-dimensional fluorescence monitoring unit uses a miniaturized fluorescence probe to analyze the fluorescence characteristic fingerprint of dissolved organic matter in water in real time and collect real-time values ​​of various pollutant markers. The full-spectrum monitoring unit covers the ultraviolet-visible-near-infrared bands and simultaneously detects real-time values ​​of COD, ammonia nitrogen, total phosphorus, total nitrogen, and nitrate nitrogen parameters; The video monitoring unit is used to collect images and videos of river and lake cross sections; The other sensor monitoring unit integrates chlorophyll, pH, conductivity, dissolved oxygen, and turbidity sensor probes to collect real-time values ​​of chlorophyll, pH, conductivity, turbidity, and dissolved oxygen in the water.

[0007] As a further description of the present invention, the working process of the analysis and early warning module includes: Obtain the real-time values ​​of various water quality indicators per unit volume of river and lake water at the current location of the unmanned vessel. Compare the real-time concentration of each water quality indicator with the concentration thresholds set by the system. If the real-time concentration of any water quality indicator is lower than the corresponding threshold set by the system, no warning is triggered; otherwise, construct a water quality warning scoring calculation model, with the expression: ; In the formula, For water quality early warning scoring, The number of water quality indicators monitored. and These are the warning scores for the i-th and j-th water quality indicators, respectively. belong , belong , is the weighting coefficient corresponding to the i-th water quality measurement indicator.

[0008] As a further description of the present invention, the process for obtaining the early warning score of the i-th water quality measurement indicator includes: Construct a risk index calculation model for the i-th water quality indicator, with the following expression: ,in, Let i be the measured concentration of the i-th water quality indicator. Let i be the standard limit value for the i-th water quality indicator. The risk index is the i-th water quality indicator. The risk index of the i-th water quality indicator Compared with the various threshold ranges set by the system, when When the value is less than A, then the water quality warning score for the i-th item is... for ,when If A is greater than or equal to A and less than B, then the water quality early warning score for the i-th item is... for ,when If the water quality warning score is greater than or equal to B and less than C, then the i-th item's score is... for ,when When C is greater than or equal to C, then the water quality early warning score for the i-th item is... for .

[0009] As a further description of the present invention, the working process of the analysis and early warning module also includes: Obtain the full spectrum and fluorescence spectrum of river and lake water quality at the current location of the unmanned vessel, extract characteristic peaks related to water pollution from the full spectrum and fluorescence spectrum, and read the peak intensity of each characteristic peak and the offset of each characteristic peak relative to the standard peak. Construct a spectral early warning scoring calculation model, the expression of which is: ; In the formula, For spectral early warning scoring, The number of characteristic peak terms, and The first The characteristic peak early warning score and the first Each characteristic peak early warning score, among which... belong , belong , For the first The weighting coefficients corresponding to the early warning scores of each characteristic peak.

[0010] As a further description of the present invention, the first... The process of obtaining the early warning score for each characteristic peak includes: Construct the first The calculation model for the early warning score of a feature peak is expressed as follows: ; In the formula, and These are the weighting coefficients corresponding to peak intensity and offset. For the first The intensity of each characteristic peak, For the first The offset of the position of each characteristic peak compared to the standard peak. and These are the peak intensity and offset reference values ​​set by the system, respectively. For the first Early warning score for each characteristic peak; The first Early warning score for individual characteristic peaks Compared with the various threshold ranges set by the system, when When it is less than D, then the first Early warning score for individual characteristic peaks for ,when If D is greater than or equal to D and less than E, then the first... Early warning score for individual characteristic peaks for ,when If it is greater than or equal to E and less than F, then the first... Early warning score for individual characteristic peaks for ,when When F is greater than or equal to F, then the first... Early warning score for individual characteristic peaks for .

[0011] As a further description of the present invention, the working process of the analysis and early warning module also includes: Based on video monitoring units, a score is issued when an abnormal water surface color is identified that differs from the normal color of river and lake water. A score is issued when drainage behavior is detected at the outlet. A score is issued if trash or blue-green algae are detected on the water surface. ; Construct a video early warning scoring calculation model, with the following expression: ; In the formula, Rate the video warning. , and They are respectively , and The corresponding weighting coefficients.

[0012] As a further description of the present invention, water quality early warning scoring is performed respectively. Spectral early warning score and video warning score The weighted summation yields the comprehensive early warning score for the unmanned vessel's patrol process, and the early warning levels are divided based on the range of the early warning scores, including Level I, Level II, and Level III early warnings from high to low.

[0013] As a further description of the present invention, the working process of the pollution source tracing module includes: When the unmanned surface vessel (USV) issues a comprehensive warning at the survey point Y and the warning level is greater than or equal to Level II, the USV’s source tracing function is automatically triggered. The USV performs convolutional neural network analysis on the three-dimensional fluorescence spectrum data of the water body at point Y and tens of thousands of three-dimensional fluorescence spectra of different types of water bodies in the cloud to identify the suspected pollution source of the polluted water body at point Y. The cross-section and outlet within a radius of L meters upstream of point Y are sequentially scanned and monitored. Multidimensional spectral data of water from outlet A to outlet B and from cross-section 1 to cross-section N are collected. The similarity is compared and analyzed with the three-dimensional fluorescence spectral data of water at point Y. If the most similar data comes from the outlet, the source tracing is terminated and water samples are collected from the outlet under video surveillance for retention, which is convenient for source tracing law enforcement. If the point with the highest similarity is a cross section, then the multidimensional spectral data of the water body at outlets B and C of the upstream cross section 1 will be further moved and collected. The similarity analysis will be performed with the three-dimensional fluorescence spectral data of the water body at cross section 1 to determine the final source of pollution outlet. At the same time, the source tracing will end and a water sample will be collected from the outlet under video surveillance for retention, which will facilitate source tracing law enforcement.

[0014] A method for intelligent source tracing and monitoring of river and lake water quality, the method comprising the following steps: Step S1: River and lake water quality monitoring begins; Step S2: The real-time monitoring data is analyzed by the multimodal water quality early warning analysis module to determine whether an early warning is triggered; Step S3: If an early warning is triggered and the level is greater than or equal to Level II, activate the pollution source tracing analysis model, trace upstream to identify the pollution path, and collect samples from suspected pollution sources for evidence. Then, continue along the original monitoring route. Step S4: If no warning is triggered, continue along the original inspection route; Step S5: End the survey and generate a survey report.

[0015] The beneficial effects of this invention are: This invention, based on water quality early warning information transmitted from the front end, uses three-dimensional fluorescence spectral data to conduct source tracing analysis of polluted areas and identify the sources of water pollution. Simultaneously, an unmanned surface vessel (USV) for intelligent source tracing monitoring of river and lake water quality performs source tracing analysis of upstream water quality, identifying the direction of pollution sources and discharge paths based on a multi-dimensional spectral pollution path identification model. Based on real-time water quality monitoring, early warning information, and source tracing information, the USV autonomously navigates to the vicinity of suspected pollution sources to conduct witnessed sampling under video surveillance for verification, achieving a closed loop of "monitoring-early warning-source tracing-verification" for river and lake water bodies.

[0016] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic diagram of part of the working logic of the intelligent source tracing and monitoring system for river and lake water quality of the present invention; Figure 2 This is a schematic diagram of pollution source tracing analysis in the intelligent river and lake water quality tracing and monitoring system of the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] Please see Figures 1-2 As shown, a smart source tracing and monitoring system for river and lake water quality is disclosed. The system includes: a smart monitoring module, a data processing module, an analysis and early warning module, and a pollution source tracing module. The intelligent monitoring module includes a three-dimensional fluorescence monitoring unit, a full-spectrum monitoring unit, a video monitoring unit, and other sensor monitoring units, used to collect characteristic parameters of river and lake water; The data processing module is used to normalize the collected parameters so that different parameters are at the same standard. The analysis and early warning module is used to analyze the processed data and issue different types of early warning scores based on the analysis results. If the early warning scores of different types reach or exceed the corresponding thresholds, an early warning is issued immediately. The pollution source tracing module is used to identify the source of pollution in the water body at the warning location when the unmanned vessel issues an early warning.

[0021] Through the above technical solution, this application uses a three-dimensional fluorescence monitoring unit to analyze the fluorescent fingerprint of dissolved organic matter in real time and detect pollution markers. A full-spectrum monitoring unit simultaneously detects parameters such as COD, ammonia nitrogen, and total phosphorus. A video monitoring unit acquires images / videos of river and lake sections to identify abnormal colors, discharge outlets, garbage, or algae. Other sensors, including integrated probes for chlorophyll, pH, dissolved oxygen, and turbidity, collect real-time values ​​of chlorophyll, pH, conductivity, turbidity, and dissolved oxygen in the water. The data processing module then normalizes the multi-source data. An analysis and early warning module analyzes water quality early warning scores, spectral early warning scores, and video early warning scores, and combines all scores to obtain a comprehensive early warning score. When the comprehensive early warning level is ≥ Level II, source tracing is initiated. Based on three-dimensional fluorescence spectral comparison and upstream monitoring, the pollution source is identified, achieving a closed loop of "monitoring-early warning-source tracing-verification" for river and lake water bodies.

[0022] The three-dimensional fluorescence monitoring unit uses a miniaturized fluorescence probe (excitation wavelength 200-600nm, emission wavelength 250-700nm) to analyze the fluorescence characteristic fingerprint of dissolved organic matter in water in real time and collect real-time values ​​of various pollutant markers. The full-spectrum monitoring unit covers the ultraviolet-visible-near-infrared bands and simultaneously detects real-time values ​​of COD, ammonia nitrogen, total phosphorus, total nitrogen, and nitrate nitrogen parameters; The video monitoring unit is used to collect images and videos of river and lake cross sections; The other sensor monitoring unit integrates chlorophyll, pH, conductivity, dissolved oxygen, and turbidity sensor probes to collect real-time values ​​of chlorophyll, pH, conductivity, turbidity, and dissolved oxygen in the water.

[0023] The working process of the analysis and early warning module includes: Obtain the real-time values ​​of various water quality indicators per unit volume of river and lake water at the current location of the unmanned vessel. Compare the real-time concentration of each water quality indicator with the concentration thresholds set by the system. If the real-time concentration of any water quality indicator is lower than the corresponding threshold set by the system, no warning is triggered; otherwise, construct a water quality warning scoring calculation model, with the expression: ; In the formula, For water quality early warning scoring, The number of water quality indicators monitored. and These are the warning scores for the i-th and j-th water quality indicators, respectively. belong , belong , is the weighting coefficient corresponding to the i-th water quality measurement indicator.

[0024] The process for obtaining the early warning score for the i-th water quality measurement indicator includes: Construct a risk index calculation model for the i-th water quality indicator, with the following expression: ,in, Let i be the measured concentration of the i-th water quality indicator. Let i be the standard limit value for the i-th water quality indicator. The risk index is the i-th water quality indicator. The risk index of the i-th water quality indicator Compared with the various threshold ranges set by the system, when When the value is less than A, then the water quality warning score for the i-th item is... for ,when If A is greater than or equal to A and less than B, then the water quality early warning score for the i-th item is... for ,when If the water quality warning score is greater than or equal to B and less than C, then the i-th item's score is... for ,when When C is greater than or equal to C, then the water quality early warning score for the i-th item is... for .

[0025] Based on the above technical solution, this embodiment constructs a water quality early warning scoring calculation model. In the formula, Water quality early warning score (comprehensive pollution level). and These are the warning scores for the i-th and j-th water quality indicators, respectively (based on the ratio of measured concentration to the standard limit, and assigned in segments). - ), The term represents the synergistic effect of pollutants (quantifying the interaction between different pollutants). In the formula, the influence of extreme values ​​is reduced by the sum of squares and the square root, while the interaction term is introduced to reflect the synergistic effect of pollutants.

[0026] The working process of the analysis and early warning module also includes: Obtain the full spectrum and fluorescence spectrum of river and lake water quality at the current location of the unmanned vessel, extract characteristic peaks related to water pollution from the full spectrum and fluorescence spectrum, and read the peak intensity of each characteristic peak and the offset of each characteristic peak relative to the standard peak. Construct a spectral early warning scoring calculation model, the expression of which is: ; In the formula, For spectral early warning scoring, The number of characteristic peak terms, and The first The characteristic peak early warning score and the first Each characteristic peak early warning score, among which... belong , belong , For the first The weighting coefficients corresponding to the early warning scores of each characteristic peak.

[0027] The first The process of obtaining the early warning score for each characteristic peak includes: Construct the first The calculation model for the early warning score of a feature peak is expressed as follows: ; In the formula, and These are the weighting coefficients corresponding to peak intensity and offset. For the first The intensity of each characteristic peak, For the first The offset of the position of each characteristic peak compared to the standard peak. and These are the peak intensity and offset reference values ​​set by the system, respectively. For the first Early warning score for each characteristic peak; The first Early warning score for individual characteristic peaks Compared with the various threshold ranges set by the system, when When it is less than D, then the first Early warning score for individual characteristic peaks for ,when If D is greater than or equal to D and less than E, then the first... Early warning score for individual characteristic peaks for ,when If it is greater than or equal to E and less than F, then the first... Early warning score for individual characteristic peaks for ,when When F is greater than or equal to F, then the first... Early warning score for individual characteristic peaks for .

[0028] Based on the above technical solution, this embodiment constructs a spectral early warning scoring calculation model. , used to quantify spectral anomalies (such as shifts in fluorescence peaks of organic matter, enhancement of chlorophyll peaks in algae, etc.), where, For spectral early warning scoring, and The first The characteristic peak early warning score and the first Early warning score for each characteristic peak (based on) The values ​​are obtained by weighted summation of the ratio of actual peak intensity to standard value and the ratio of actual offset to standard offset, and then assigned in segments. - The influence of extreme values ​​is reduced by using the sum of squares and square roots, while an interaction term is introduced to reflect the synergistic effect of different characteristic peaks. This is a synergistic effect between characteristic peaks. The working process of the analysis and early warning module also includes: Based on video monitoring units, a score is issued when an abnormal water surface color is identified that differs from the normal color of river and lake water. A score is issued when drainage behavior is detected at the outlet. A score is issued if trash or blue-green algae are detected on the water surface. ; Construct a video early warning scoring calculation model, with the following expression: ; In the formula, Rate the video warning. , and They are respectively , and The corresponding weighting coefficients.

[0029] Using the above technical solution, this embodiment constructs a video early warning scoring model, which supplements the deficiencies of chemical / spectral data by linearly weighted integration of visual information. In the formula, This indicates an abnormal color on the water surface (such as industrial wastewater discharge). This indicates the discharge of wastewater at the outlet (suspected source of pollution). This indicates surface debris or blue-green algae (a marker of eutrophication).

[0030] Water quality early warning scores were assigned separately. Spectral early warning score and video warning score The weighted summation yields the comprehensive early warning score for the unmanned vessel's patrol process, and the early warning levels are divided based on the range of the early warning scores, including Level I, Level II, and Level III early warnings from high to low.

[0031] Using the above technical solution, this embodiment performs a comprehensive scoring calculation based on the water quality early warning score. Spectral early warning score and video warning score The results are obtained by weighted summation, and then the comprehensive early warning score is compared with the set threshold to classify the warnings into three levels: Level I, Level II, and Level III, from highest to lowest.

[0032] The working process of the pollution source tracing module includes: When the unmanned surface vessel (USV) issues a comprehensive early warning at monitoring point Y, and the warning level is greater than or equal to Level II, the USV's source tracing function is automatically triggered. The USV performs convolutional neural network analysis on the three-dimensional fluorescence spectrum data of the water body at point Y and tens of thousands of three-dimensional fluorescence spectra of different types of water bodies in the cloud (including domestic sewage, agricultural non-point source water bodies, surface water, fluoride-containing wastewater, electroplating wastewater, medical wastewater, livestock and poultry breeding wastewater, etc.) to identify the suspected pollution sources of the polluted water body at point Y. The cross-section and outlet within a radius of L meters upstream of point Y are sequentially scanned and monitored. Multidimensional spectral data of water from outlet A to outlet B and from cross-section 1 to cross-section N are collected. The similarity is compared and analyzed with the three-dimensional fluorescence spectral data of water at point Y. If the most similar data comes from the outlet, the source tracing is terminated and water samples are collected from the outlet under video surveillance for retention, which is convenient for source tracing law enforcement. If the point with the highest similarity is a cross section, then the multidimensional spectral data of the water body at outlets B and C of the upstream cross section 1 will be further moved and collected. The similarity analysis will be performed with the three-dimensional fluorescence spectral data of the water body at cross section 1 to determine the final source of pollution outlet. At the same time, the source tracing will end and a water sample will be collected from the outlet under video surveillance for retention, which will facilitate source tracing law enforcement.

[0033] Through the above technical solution, this embodiment automatically activates the source tracing function when the unmanned vessel triggers a Level II or higher comprehensive warning at the patrol point. It performs convolutional neural network matching between the three-dimensional fluorescence data of the warning point and the cloud-based pollution source spectral library (including domestic sewage, industrial wastewater, etc.) to initially identify the pollution type. It moves and detects within a radius L upstream of the point, collecting multi-dimensional spectral data of the discharge outlet and cross-section. The similarity is compared with the data of the warning point. If the maximum similarity matches the discharge outlet, the pollution source is locked and video evidence is recorded. If it matches the cross-section, the upstream discharge outlet is tracked until the final source is determined. By combining spectral analysis and mobile monitoring, the pollution source can be quickly and accurately located.

[0034] A method for intelligent source tracing and monitoring of river and lake water quality, the method comprising the following steps: Step S1: River and lake water quality monitoring begins; Step S2: The real-time monitoring data is analyzed by the multimodal water quality early warning analysis module to determine whether an early warning is triggered; Step S3: If an early warning is triggered and the level is greater than or equal to Level II, activate the pollution source tracing analysis model, trace upstream to identify the pollution path, and collect samples from suspected pollution sources for evidence. Then, continue along the original monitoring route. Step S4: If no warning is triggered, continue along the original inspection route; Step S5: End the survey and generate a survey report.

[0035] This system also includes an intelligent monitoring platform with functions such as spatial heat maps of river and lake water quality, spatiotemporal health trends of water quality, water quality early warning messages, source tracing results and location, water quality sampling, and water environment AI experts. Among them, the water environment AI expert is based on a knowledge graph-driven vertical large-scale model expert system for water environment AI. It has completed domain fine-tuning based on the Qwen 2.5-VL and DeepseekR1 large models, and embedded parameters of hydrodynamic models such as SWMM and EFDC to build a dual-driven architecture of "mechanism-AI large model". This breaks through the limitations of the vertical application of general large models, realizes water quality parameter prediction, semantic parsing of pollution events, intelligent analysis of multimodal data and automatic generation of charts, strengthens data governance in the field of water environment, awakens "dormant" data, releases data value, and enhances the digital capabilities of environmental management.

[0036] It should be noted that the formulas in this application are all dimensionless and numerical calculations. The formulas are obtained by software simulation based on a large amount of data and are the closest to the real situation. The thresholds, weight coefficients and scoring assignments involved in this application are all empirical values ​​and are selected by those skilled in the art according to the actual situation.

[0037] The above description is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described or use similar methods to replace them, as long as they do not deviate from the concept of the invention or exceed the scope defined in the claims, they should all fall within the protection scope of the present invention.

Claims

1. A smart source tracing and monitoring system for river and lake water quality, characterized in that, The system includes: an intelligent monitoring module, a data processing module, an analysis and early warning module, and a pollution source tracing module; The intelligent monitoring module includes a three-dimensional fluorescence monitoring unit, a full-spectrum monitoring unit, a video monitoring unit, and other sensor monitoring units, used to collect characteristic parameters of river and lake water; The data processing module is used to normalize the collected parameters so that different parameters are at the same standard. The analysis and early warning module is used to analyze the processed data and issue different types of early warning scores based on the analysis results. If the early warning scores of different types reach or exceed the corresponding thresholds, an early warning is issued immediately. The pollution source tracing module is used to identify the source of pollution in the water body at the warning location when the unmanned vessel issues an early warning.

2. The intelligent source tracing and monitoring system for river and lake water quality according to claim 1, characterized in that, The three-dimensional fluorescence monitoring unit uses a miniaturized fluorescence probe to analyze the fluorescence characteristic fingerprint of dissolved organic matter in water in real time and collect real-time values ​​of various pollutant markers. The full-spectrum monitoring unit covers the ultraviolet-visible-near-infrared bands and simultaneously detects real-time values ​​of COD, ammonia nitrogen, total phosphorus, total nitrogen, and nitrate nitrogen parameters; The video monitoring unit is used to collect images and videos of river and lake cross sections; The other sensor monitoring unit integrates chlorophyll, pH, conductivity, dissolved oxygen, and turbidity sensor probes to collect real-time values ​​of chlorophyll, pH, conductivity, turbidity, and dissolved oxygen in the water.

3. The intelligent source tracing and monitoring system for river and lake water quality according to claim 1, characterized in that, The working process of the analysis and early warning module includes: Obtain the real-time values ​​of various water quality indicators per unit volume of river and lake water at the current location of the unmanned vessel. Compare the real-time concentration of each water quality indicator with the concentration thresholds set by the system. If the real-time concentration of any water quality indicator is lower than the corresponding threshold set by the system, no warning is triggered; otherwise, construct a water quality warning scoring calculation model, with the expression: ； In the formula, For water quality early warning scoring, The number of water quality indicators monitored. and These are the warning scores for the i-th and j-th water quality indicators, respectively. belong , belong , is the weighting coefficient corresponding to the i-th water quality measurement indicator.

4. The intelligent source tracing and monitoring system for river and lake water quality according to claim 3, characterized in that, The process for obtaining the early warning score for the i-th water quality measurement indicator includes: Construct a risk index calculation model for the i-th water quality indicator, with the following expression: ,in, Let i be the measured concentration of the i-th water quality indicator. Let i be the standard limit value for the i-th water quality indicator. The risk index is the i-th water quality indicator. The risk index of the i-th water quality indicator Compared with the various threshold ranges set by the system, when When the value is less than A, then the water quality warning score for the i-th item is... for ,when If A is greater than or equal to A and less than B, then the water quality warning score for the i-th item is... for ,when If the water quality warning score is greater than or equal to B and less than C, then the i-th item's score is... for ,when When C is greater than or equal to C, then the water quality early warning score for the i-th item is... for .

5. The intelligent source tracing and monitoring system for river and lake water quality according to claim 1, characterized in that, The working process of the analysis and early warning module also includes: Obtain the full spectrum and fluorescence spectrum of river and lake water quality at the current location of the unmanned vessel, extract characteristic peaks related to water pollution from the full spectrum and fluorescence spectrum, and read the peak intensity of each characteristic peak and the offset of each characteristic peak relative to the standard peak. Construct a spectral early warning scoring calculation model, the expression of which is: ； In the formula, For spectral early warning scoring, The number of characteristic peak terms, and The first The characteristic peak early warning score and the first Each characteristic peak early warning score, among which... belong , belong , For the first The weighting coefficients corresponding to the early warning scores of each characteristic peak.

6. The intelligent source tracing and monitoring system for river and lake water quality according to claim 5, characterized in that, The first The process of obtaining the early warning score for each characteristic peak includes: Construct the first The calculation model for the early warning score of a feature peak is expressed as follows: ； In the formula, and These are the weighting coefficients corresponding to peak intensity and offset. For the first The intensity of each characteristic peak, For the first The offset of the position of each characteristic peak compared to the standard peak. and These are the peak intensity and offset reference values ​​set by the system, respectively. For the first Early warning score for each characteristic peak; The first Early warning score for individual characteristic peaks Compared with the various threshold ranges set by the system, when When it is less than D, then the first Early warning score for individual characteristic peaks for ,when If D is greater than or equal to D and less than E, then the first... Early warning score for individual characteristic peaks for ,when If it is greater than or equal to E and less than F, then the first... Early warning score for individual characteristic peaks for ,when When F is greater than or equal to F, then the first... Early warning score for individual characteristic peaks for .

7. The intelligent source tracing and monitoring system for river and lake water quality according to claim 1, characterized in that, The working process of the analysis and early warning module also includes: Based on video monitoring units, a score is issued when an abnormal water surface color is identified that differs from the normal color of river and lake water. A score is issued when drainage behavior is detected at the outlet. A score is issued if trash or blue-green algae are detected on the water surface. ; Construct a video early warning scoring calculation model, with the following expression: ； In the formula, Rate the video warning. , and They are respectively , and The corresponding weighting coefficients.

8. The intelligent source tracing and monitoring system for river and lake water quality according to any one of claims 3-7, characterized in that, Water quality early warning scores were assigned separately. Spectral early warning score and video warning score The weighted summation yields the comprehensive early warning score for the unmanned vessel's patrol process, and the early warning levels are divided based on the range of the early warning scores, including Level I, Level II, and Level III early warnings from high to low.

9. The intelligent source tracing and monitoring system for river and lake water quality according to claim 1, characterized in that, The working process of the pollution source tracing module includes: When the unmanned surface vessel (USV) issues a comprehensive warning at the survey point Y and the warning level is greater than or equal to Level II, the USV’s source tracing function is automatically triggered. The USV performs convolutional neural network analysis on the three-dimensional fluorescence spectrum data of the water body at point Y and tens of thousands of three-dimensional fluorescence spectra of different types of water bodies in the cloud to identify the suspected pollution source of the polluted water body at point Y. The cross-section and outlet within a radius of L meters upstream of point Y are sequentially scanned and monitored. Multidimensional spectral data of water from outlet A to outlet B and from cross-section 1 to cross-section N are collected. The similarity is compared and analyzed with the three-dimensional fluorescence spectral data of water at point Y. If the most similar data comes from the outlet, the source tracing is terminated and water samples are collected from the outlet under video surveillance for retention, which is convenient for source tracing law enforcement. If the point with the highest similarity is a cross section, then the multidimensional spectral data of the water body at outlets B and C of the upstream cross section 1 will be further moved and collected. The similarity analysis will be performed with the three-dimensional fluorescence spectral data of the water body at cross section 1 to determine the final source of pollution outlet. At the same time, the source tracing will end and a water sample will be collected from the outlet under video surveillance for retention, which will facilitate source tracing law enforcement.

10. A method for intelligent source tracing and monitoring of river and lake water quality, wherein the method is implemented using the intelligent source tracing and monitoring system for river and lake water quality as described in any one of claims 1-9, characterized in that, The method includes the following steps: Step S1: River and lake water quality monitoring begins; Step S2: The real-time monitoring data is analyzed by the multimodal water quality early warning analysis module to determine whether an early warning is triggered; Step S3: If an early warning is triggered and the level is greater than or equal to Level II, activate the pollution source tracing analysis model, trace upstream to identify the pollution path, and collect samples from suspected pollution sources for evidence. Then, continue along the original monitoring route. Step S4: If no warning is triggered, continue to execute the original inspection route; Step S5: End the survey and generate a survey report.

Images