An intelligent unmanned ship-based water quality pollution traceability analysis method and system

Abstract

The application relates to the field of water quality pollution tracing, and in particular to a water quality pollution tracing analysis method and system based on an intelligent unmanned ship. The method comprises the following steps: acquiring real-time position multi-modal network monitoring parameters based on a multi-mode communication module, and performing dynamic multi-mode network intelligent switching decision and constructing an intelligent multi-mode network switching strategy; acquiring real-time water area multi-dimensional water quality monitoring parameters of the intelligent unmanned ship according to multi-dimensional water quality sensors, and performing abnormal pollution parameter change detection and marking abnormal change pollution parameters; performing reverse water quality space change analysis and gradient descent iteration tracing calculation on the abnormal change pollution parameters to generate a potential pollution diffusion path; performing intelligent tracing cruise according to the potential pollution diffusion path, and performing pollution water source tracking positioning to obtain a pollution water source point. The application realizes efficient and accurate water quality pollution tracing.

Description

Technical Field

[0001] This invention relates to the field of water pollution source tracing technology, and in particular to a water pollution source tracing analysis method and system based on intelligent unmanned vessels. Background Technology

[0002] With the advancement of industrialization and the continuous improvement of urbanization, water pollution has become an increasingly serious problem, posing a significant threat to the ecological environment and human health. This is especially urgent in water-scarce regions. Traditional methods for tracing and analyzing water pollution sources rely primarily on manual inspections and fixed monitoring stations, which suffer from problems such as delayed response, limited monitoring point distribution, and incomplete data. These traditional methods are not only limited in spatial scope and timeliness but also have significant blind spots in locating and tracing pollution sources, leading to delays in early warning and effective handling of pollution incidents.

[0003] In recent years, with the rapid development of intelligent technology and unmanned systems, intelligent unmanned surface vessels (USVs) have gradually attracted widespread attention as a new type of water quality monitoring tool. Intelligent USVs offer advantages such as high efficiency, flexibility, and real-time monitoring, enabling them to conduct long-term, continuous, automated monitoring across a wide range of water areas and collect water quality-related data. Compared to traditional manual patrols and fixed monitoring stations, USVs can achieve dynamic monitoring and automated operation, effectively improving the efficiency of water pollution detection and the accuracy of source tracing. By carrying various sensors (such as water quality sensors and GPS positioning devices), USVs can collect water quality data in real time and obtain precise location coordinates, thus providing basic data for the location and tracking of pollution sources.

[0004] However, despite the enormous potential of intelligent unmanned surface vessels (USVs) in water quality monitoring, challenges remain in tracing the source of water pollution, including accurately determining the location of pollution sources, identifying pollution diffusion paths, and analyzing pollution components. Traditional pollution tracing methods often rely on experience-based judgment, failing to provide timely and accurate analysis of the spatial distribution and diffusion trends of pollution sources. Therefore, utilizing real-time monitoring data from intelligent USVs in conjunction with advanced data analysis methods for pollution tracing has become a significant technical challenge in water pollution prevention and control. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention proposes a water pollution source tracing and analysis method and system based on intelligent unmanned vessels, thereby resolving at least one of the aforementioned technical issues.

[0006] To achieve the above objectives, this invention provides a method for tracing and analyzing water pollution sources based on an intelligent unmanned surface vessel (USV). The USV is equipped with multi-dimensional water quality sensors, a multi-mode communication module, and an infrared camera. The method includes the following steps:

[0007] Step S1: Obtain real-time location multimodal network monitoring parameters based on the multimodal communication module, make dynamic multimodal network intelligent switching decisions, and construct an intelligent multimodal network switching strategy;

[0008] Step S2: Obtain real-time multi-dimensional water quality monitoring parameters of the intelligent unmanned vessel based on multi-dimensional water quality sensors, and detect and mark abnormal pollution parameter changes.

[0009] Step S3: Perform reverse spatial variation analysis of water quality and gradient descent iterative source tracing calculation on abnormally changing pollution parameters to generate potential pollution diffusion paths;

[0010] Step S4: Conduct intelligent source tracing patrols based on potential pollution diffusion paths and locate polluted water sources to obtain the pollution source points;

[0011] Step S5: Identify hidden sewage outlets based on the source of pollution and perform quantitative analysis of the pollution source components to obtain information on the water pollution components;

[0012] Step S6: Based on the information on water pollution components, conduct pollution source diffusion and evolution analysis, and based on the intelligent multi-mode network switching strategy, conduct intelligent early warning decision-making to construct a water pollution source tracing and early warning strategy.

[0013] This invention combines multi-mode communication technologies (such as 5G, WiFi, and BeiDou) to ensure stable connectivity for unmanned vessels even in complex environments or areas far from network signals. By monitoring network status in real time (such as bandwidth, latency, and packet loss rate), the optimal communication link can be dynamically selected based on different environmental conditions, thereby maximizing data transmission speed and reliability and ensuring accurate and real-time transmission of water quality monitoring data. Traditional communication technologies (such as single networks) may experience signal interruptions or instability in certain environments. This intelligent switching strategy optimizes the network switching mechanism to ensure the continuity of the communication link, avoiding data loss or monitoring interruptions due to network failures, which is particularly important in complex water areas and areas with weak signals. High-precision real-time water quality data is acquired using multiple water quality sensors (such as pH, dissolved oxygen, turbidity, and conductivity), providing rich monitoring data support for subsequent pollution source tracing. Comprehensive collection of multi-dimensional water quality data allows for a more comprehensive assessment of the water quality situation, reducing the number of potential pollution sources missed. Detection of abnormal pollution parameter changes can quickly identify sudden changes in water quality and promptly mark areas with pollution. This function effectively improves the response speed and accuracy of water quality monitoring, enabling timely identification of pollution sources and reducing pollution spread. Through inverse spatial change analysis of water quality and gradient descent iterative source tracing calculations, the potential location of pollution sources and pollution spread paths can be inversely calculated based on changes in marked abnormal pollution parameters. This process can quickly identify pollution sources, avoiding misjudgment or omission. The gradient descent algorithm optimizes the source tracing path, making the process more efficient. This step not only improves the accuracy of pollution source tracing but also reduces errors caused by environmental complexity and sensor errors, enhancing the intelligence and reliability of the entire system. Based on potential pollution spread paths, the intelligent unmanned surface vessel (USV) can dynamically adjust its cruise path according to the environment, accurately tracking pollution along the path. Combining intelligent obstacle avoidance and path planning technologies, the system can avoid collisions with obstacles, adapt to changes in water conditions, and optimize its navigation route, thereby efficiently and accurately locating pollution sources. Through autonomous cruising and intelligent path planning, the USV eliminates the inefficiency and errors of manual operation, ensuring continuous and efficient monitoring of water areas. This automated process significantly improves the efficiency and accuracy of water pollution monitoring. Identifying hidden sewage outlets using equipment such as infrared imaging, and combining this with quantitative analysis of pollutant types and concentrations based on sampling and monitoring data from pollution sources, can provide more in-depth information for pollution source identification. This analysis can provide a scientific basis for pollution source control and legal enforcement. Detailed analysis of the sewage outlet and its surrounding water quality can accurately determine the composition (such as heavy metals and chemicals) and concentration distribution of the pollution source, providing crucial data support for water quality management and environmental protection. Based on pollution source composition information, it is possible to simulate the diffusion and evolution process of the pollution source and predict the potential impact of pollution events on surrounding water areas. This analysis can provide forward-looking decision support for pollution prevention and control, helping to quickly implement emergency measures and reduce pollution spread.By incorporating a dynamic switching strategy for the multi-mode communication module, the system can rapidly send early warning signals to relevant departments or personnel in the event of a pollution incident. Through intelligent early warning decision-making, the system can automatically assess pollution risks and generate appropriate emergency response strategies. This not only improves the system's response speed but also provides timely guidance for pollution control.

[0014] This specification provides a water pollution source tracing and analysis system based on an intelligent unmanned vessel, used to execute the water pollution source tracing and analysis method based on an intelligent unmanned vessel as described above, including:

[0015] The multimodal network switching module is used to obtain real-time location multimodal network monitoring parameters based on the multimodal communication module, make dynamic multimodal network intelligent switching decisions, and construct intelligent multimodal network switching strategies.

[0016] The pollution change detection module is used to acquire real-time multi-dimensional water quality monitoring parameters of the intelligent unmanned vessel based on multi-dimensional water quality sensors, and to detect and mark abnormal pollution parameter changes.

[0017] The reverse water quality source tracing module is used to perform reverse water quality spatial change analysis and gradient descent iterative source tracing calculation on abnormally changing pollution parameters in order to generate potential pollution diffusion paths.

[0018] The source tracing and navigation module is used to conduct intelligent source tracing and navigation based on potential pollution diffusion paths, and to track and locate polluted water sources to obtain the source of pollution.

[0019] The quantitative analysis module is used to identify hidden sewage outlets based on the source of pollution and to perform quantitative analysis of the pollution source components in order to obtain information on the components of water pollution.

[0020] The intelligent early warning module is used to analyze the evolution of pollution source diffusion based on water pollution component information and to make intelligent early warning decisions based on intelligent multi-mode network switching strategies, thereby constructing a water pollution source tracing and early warning strategy.

[0021] This invention utilizes multi-mode communication technology to acquire the real-time location of unmanned surface vessels (USVs) and dynamically optimizes network switching decisions based on real-time monitoring parameters of the multi-mode network. This enables USVs to continuously transmit data through optimal network connections even in complex environments, ensuring the stability and smoothness of data communication. By dynamically selecting the optimal communication method, USVs can avoid data loss due to communication instability, improving the reliability of mission execution, especially in complex aquatic environments, ensuring the real-time nature and accuracy of data. Through multi-dimensional water quality sensors, real-time water quality data (such as temperature, pH, dissolved oxygen, turbidity, etc.) is collected, allowing the module to effectively track dynamic changes in water quality. This module can promptly detect abnormal water quality changes and mark pollution change parameters, providing data support for subsequent pollution source tracing and treatment. This allows for the rapid identification of potential pollution risks and the prevention of pollution spread. Through reverse source tracing analysis of abnormal pollution parameters, the spatial location and possible diffusion paths of pollution sources can be inferred, providing a basis for accurately locating pollution sources and helping to accurately identify the root cause of water quality anomalies. Using a gradient descent algorithm for calculation optimization enables efficient and accurate identification of potential pollution sources. Through iterative model calculations, the accuracy and precision of pollution source tracing will be improved. Based on potential pollution diffusion paths obtained from reverse analysis, unmanned surface vessels (USVs) can be guided to conduct intelligent patrols and track pollution sources in real time. Real-time patrols enable USVs to effectively track the location of pollution sources and provide data support for subsequent pollution treatment. Intelligent source tracing patrols not only ensure efficient tracking but also avoid resource waste, making the patrol process more precise and energy-efficient. Based on information about polluted water sources, potentially hidden discharge outlets can be identified. Quantitative analysis of pollution source components can accurately identify the types and components of pollution sources (such as organic matter, heavy metals, pesticides, etc.), providing precise data support for treatment measures. Analysis of pollution sources can identify the specific types of pollutants, providing a basis for decision-making in water quality treatment and pollution source control. Accurate pollution component information will help formulate more effective treatment measures. Based on the prediction of pollution source diffusion evolution, combined with intelligent multi-mode network switching strategies, early warning signals of water pollution can be issued in advance. By predicting the diffusion trend of pollutants in advance, a time window is provided for emergency response and treatment measures. The intelligent early warning module, combined with an intelligent multi-mode network switching strategy, can make rapid early warning decisions based on real-time data, network conditions, and pollution source information, helping relevant departments to take timely action to deal with the spread of pollution. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the steps of a water pollution source tracing and analysis method based on an intelligent unmanned vessel according to the present invention;

[0023] Figure 2 This is a detailed flowchart illustrating the implementation steps of step S1.

[0024] Figure 3This is a detailed flowchart illustrating the implementation steps of step S2;

[0025] Figure 4 This is a flowchart illustrating the detailed implementation steps of step S3. Detailed Implementation

[0026] It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the invention.

[0027] This application provides a method and system for tracing and analyzing water pollution sources based on an intelligent unmanned vessel. The executing entities of the method and system include, but are not limited to, the following: mechanical equipment, data processing platform, cloud server node, network upload device, etc., which can be considered as general computing nodes in this application. The data processing platform includes, but is not limited to, at least one of the following: an audio-visual management system, an information management system, and a cloud data management system.

[0028] Please see Figures 1 to 4 This invention provides a method for tracing and analyzing water pollution sources based on an intelligent unmanned surface vessel (USV). The USV is equipped with a multi-dimensional water quality sensor, a multi-mode communication module, and an infrared camera. The method includes the following steps:

[0029] Step S1: Obtain real-time location multimodal network monitoring parameters based on the multimodal communication module, make dynamic multimodal network intelligent switching decisions, and construct an intelligent multimodal network switching strategy;

[0030] Step S2: Obtain real-time multi-dimensional water quality monitoring parameters of the intelligent unmanned vessel based on multi-dimensional water quality sensors, and detect and mark abnormal pollution parameter changes.

[0031] Step S3: Perform reverse spatial variation analysis of water quality and gradient descent iterative source tracing calculation on abnormally changing pollution parameters to generate potential pollution diffusion paths;

[0032] Step S4: Conduct intelligent source tracing patrols based on potential pollution diffusion paths and locate polluted water sources to obtain the pollution source points;

[0033] Step S5: Identify hidden sewage outlets based on the source of pollution and perform quantitative analysis of the pollution source components to obtain information on the water pollution components;

[0034] Step S6: Based on the information on water pollution components, conduct pollution source diffusion and evolution analysis, and based on the intelligent multi-mode network switching strategy, conduct intelligent early warning decision-making to construct a water pollution source tracing and early warning strategy.

[0035] This invention combines multi-mode communication technologies (such as 5G, WiFi, and BeiDou) to ensure stable connectivity for unmanned vessels even in complex environments or areas far from network signals. By monitoring network status in real time (such as bandwidth, latency, and packet loss rate), the optimal communication link can be dynamically selected based on different environmental conditions, thereby maximizing data transmission speed and reliability and ensuring accurate and real-time transmission of water quality monitoring data. Traditional communication technologies (such as single networks) may experience signal interruptions or instability in certain environments. This intelligent switching strategy optimizes the network switching mechanism to ensure the continuity of the communication link, avoiding data loss or monitoring interruptions due to network failures, which is particularly important in complex water areas and areas with weak signals. High-precision real-time water quality data is acquired using multiple water quality sensors (such as pH, dissolved oxygen, turbidity, and conductivity), providing rich monitoring data support for subsequent pollution source tracing. Comprehensive collection of multi-dimensional water quality data allows for a more comprehensive assessment of the water quality situation, reducing the number of potential pollution sources missed. Detection of abnormal pollution parameter changes can quickly identify sudden changes in water quality and promptly mark areas with pollution. This function effectively improves the response speed and accuracy of water quality monitoring, enabling timely identification of pollution sources and reducing pollution spread. Through inverse spatial change analysis of water quality and gradient descent iterative source tracing calculations, the potential location of pollution sources and pollution spread paths can be inversely calculated based on changes in marked abnormal pollution parameters. This process can quickly identify pollution sources, avoiding misjudgment or omission. The gradient descent algorithm optimizes the source tracing path, making the process more efficient. This step not only improves the accuracy of pollution source tracing but also reduces errors caused by environmental complexity and sensor errors, enhancing the intelligence and reliability of the entire system. Based on potential pollution spread paths, the intelligent unmanned surface vessel (USV) can dynamically adjust its cruise path according to the environment, accurately tracking pollution along the path. Combining intelligent obstacle avoidance and path planning technologies, the system can avoid collisions with obstacles, adapt to changes in water conditions, and optimize its navigation route, thereby efficiently and accurately locating pollution sources. Through autonomous cruising and intelligent path planning, the USV eliminates the inefficiency and errors of manual operation, ensuring continuous and efficient monitoring of water areas. This automated process significantly improves the efficiency and accuracy of water pollution monitoring. Identifying hidden sewage outlets using equipment such as infrared imaging, and combining this with quantitative analysis of pollutant types and concentrations based on sampling and monitoring data from pollution sources, can provide more in-depth information for pollution source identification. This analysis can provide a scientific basis for pollution source control and legal enforcement. Detailed analysis of the sewage outlet and its surrounding water quality can accurately determine the composition (such as heavy metals and chemicals) and concentration distribution of the pollution source, providing crucial data support for water quality management and environmental protection. Based on pollution source composition information, it is possible to simulate the diffusion and evolution process of the pollution source and predict the potential impact of pollution events on surrounding water areas. This analysis can provide forward-looking decision support for pollution prevention and control, helping to quickly implement emergency measures and reduce pollution spread.By incorporating a dynamic switching strategy for the multi-mode communication module, the system can rapidly send early warning signals to relevant departments or personnel in the event of a pollution incident. Through intelligent early warning decision-making, the system can automatically assess pollution risks and generate appropriate emergency response strategies. This not only improves the system's response speed but also provides timely guidance for pollution control.

[0036] In the embodiments of the present invention, see Figure 1 This is a flowchart illustrating the steps of a water pollution source tracing and analysis method based on an intelligent unmanned vessel according to the present invention. In this example, the steps of the water pollution source tracing and analysis method based on an intelligent unmanned vessel include:

[0037] Step S1: Obtain real-time location multimodal network monitoring parameters based on the multimodal communication module, make dynamic multimodal network intelligent switching decisions, and construct an intelligent multimodal network switching strategy;

[0038] In this embodiment, a suitable multi-mode communication module is selected, typically including multiple communication methods such as 4G, 5G, Wi-Fi, and Bluetooth. The module is designed to support rapid switching and flexible data transmission. The multi-mode communication module is integrated into the intelligent unmanned surface vessel's (USV) control system, and the communication interface is configured to ensure seamless interoperability with other system modules (such as GPS positioning modules and sensor modules). The multi-mode communication module is activated and performs a self-test to ensure all communication functions are normal. The self-test includes checking signal strength, network availability, and connection stability. The initial state of the device is recorded, including available network types, signal strength, and current GPS coordinates for subsequent data analysis and decision-making. The USV's GPS positioning system is activated to acquire the USV's geographical location information in real time. The GPS module's accuracy is ensured to be within the meter range for accurate location information. A data acquisition frequency is set, for example, acquiring location data once per second, and recording the timestamp, latitude and longitude, speed, and heading of each data point. The acquired real-time location information is transmitted through the multi-mode communication module to ensure data real-time performance. The communication method with the strongest current signal is selected for data transmission to reduce latency. Simultaneously, the location data is stored in the USV's local database for subsequent analysis and backtracking. Each record should include information such as time, location, signal strength, and network type. Real-time monitoring of various network environment parameters, including signal strength, latency, bandwidth, and network availability, is crucial for determining the optimal communication method. A monitoring frequency should be set (e.g., once every 5 seconds) to record performance indicators under different network conditions for subsequent analysis. Performance data for each network type should be collected in real-time using a multi-mode communication module. The accuracy and completeness of the data must be ensured, including signal strength (in dBm), latency (in milliseconds), and bandwidth (in Mbps) for each network. Preliminary analysis of the collected data should be performed to identify trends in the current network environment, such as signal fluctuations and network switching frequency. A dynamic switching decision algorithm should be designed based on the real-time network monitoring parameters. This algorithm should be able to automatically select the optimal communication method based on network performance indicators. A weighted decision model should be used, assigning weights to parameters such as signal strength, latency, and bandwidth. For example, signal strength accounts for 60%, latency for 30%, and bandwidth for 10%, to ensure that the selected network has the best overall performance. During algorithm operation, the current network performance indicators should be evaluated in real-time and compared with preset thresholds. When the performance of a network falls below a set threshold, a handover decision is triggered. Based on the decision, the system automatically switches to a network with a stronger signal and better performance. For example, if the Wi-Fi signal strength is below -70dBm, it automatically switches to a 4G network. The decision-making process and results of each handover are recorded, including network performance metrics before and after the handover, handover time, and handover frequency. This data will be used for subsequent algorithm optimization and adjustment.Regularly evaluate the switching strategy and, based on feedback from practical applications, optimize the decision-making algorithm and weight settings to improve the intelligence level of the switching process.

[0039] Step S2: Obtain real-time multi-dimensional water quality monitoring parameters of the intelligent unmanned vessel based on multi-dimensional water quality sensors, and detect and mark abnormal pollution parameter changes.

[0040] In this embodiment, suitable multi-dimensional water quality sensors are selected, typically including pH sensors, turbidity sensors, dissolved oxygen sensors, ammonia nitrogen sensors, and total phosphorus sensors. These sensors should possess high sensitivity and rapid response capabilities to ensure the accuracy of real-time monitoring. The water quality sensors are integrated into the intelligent unmanned surface vessel's control system, ensuring seamless interface with the data acquisition module and performing necessary calibrations to improve measurement accuracy. The water quality sensors are activated and self-tested to confirm that all sensors are functioning correctly. The self-test process should include checking the sensor's response speed, sensitivity, and calibration status. Initial calibration data and sensor status are recorded, including the measurement range and sensitivity of each sensor, for subsequent analysis and adjustment. The real-time data acquisition frequency of the water quality sensors is set, for example, acquiring water quality parameters once per second. The system is ensured to stably collect key water quality indicators such as pH, turbidity, dissolved oxygen, ammonia nitrogen, and total phosphorus. Using a data logging module, each collected water quality parameter is stored in real-time in the unmanned surface vessel's database, ensuring data integrity and traceability. During monitoring, changes in each water quality parameter are continuously recorded, including timestamps, sensor readings, and environmental conditions (such as water temperature and flow rate). Regularly check the sensor's operational status to ensure proper functioning and promptly identify and address any potential malfunctions or anomalies. Develop detection standards for abnormal changes based on historical water quality monitoring data and standards. These standards typically include the normal range and warning values ​​for each water quality parameter. For example, the normal pH range is 6.5-8.5, and the turbidity exceedance limit is 5 NTU. Set dynamic thresholds; when a water quality parameter's change exceeds the set standard (e.g., exceeding the normal range by 10%), it is considered an abnormal change. Use data analysis algorithms (such as the standard deviation method or Z-score method) to analyze water quality parameter changes in real time. For each monitoring cycle, calculate the deviation of the current parameter from the historical average and determine if it exceeds the set threshold. When an abnormal change is detected, the system automatically marks the parameter and records relevant information, including the abnormal value, timestamp, and sensor location. Store the marked abnormal pollution parameters in a database and generate detailed abnormal change records. These records should include the specific value of the abnormal parameter, the magnitude of the change, and its potential impact. Regularly generate abnormal change reports to summarize anomalies during the monitoring period, providing a basis for subsequent environmental governance and decision-making. Once an abnormal change in pollution parameters is detected, the system will automatically trigger an alarm mechanism, notifying relevant personnel for further analysis and processing. The alarm information should include key information such as the abnormal parameter, location, and time. Ensure that alarm information can be quickly transmitted to the monitoring center via a multi-mode communication module for timely response. After detecting an anomaly, increase the monitoring frequency (e.g., once every 5 seconds) to more closely track water quality changes and ensure rapid response to potential pollution sources. Combine the abnormal data with environmental conditions for comprehensive analysis to assess the potential impact range and consequences of the pollution source, providing a scientific basis for developing subsequent remediation plans.

[0041] Step S3: Perform reverse spatial variation analysis of water quality and gradient descent iterative source tracing calculation on abnormally changing pollution parameters to generate potential pollution diffusion paths;

[0042] In this embodiment, data on previously monitored abnormal pollution parameters are collected and organized, including outliers, timestamps, geographical locations, and water quality parameter trends. This data will serve as the basis for reverse analysis. Historical data of abnormal parameters are recorded for subsequent analysis, ensuring the data includes sufficient time span and spatial distribution, such as hourly water quality parameter readings over the past 24 hours. The collected data is cleaned to remove potential noise and errors, ensuring the accuracy of subsequent analysis. Missing and outlier values ​​are checked and processed using interpolation or other data imputation methods. The data is standardized to eliminate the influence of different parameter units and magnitudes, making it suitable for subsequent spatial variation analysis. Spatial interpolation techniques (such as Kriging interpolation, inverse distance weighting, etc.) are used to perform spatial variation analysis on the abnormal pollution parameters. A suitable interpolation method is selected, and the spatial distribution of pollutants in the water area is calculated based on known outlier data. Interpolation parameters, such as search radius and weights, are set to ensure that the generated spatial variation model accurately reflects the distribution characteristics of pollutants. Based on the spatial model, reverse analysis is performed to identify possible source points of pollutants. By tracing the trajectory of pollutant changes in time and space, the potential location of pollution sources is determined. This study analyzes the diffusion paths and velocities of pollutants in water bodies by considering the flow characteristics (such as flow velocity and direction). A fluid dynamics model is used to simulate the impact of water flow on pollutant diffusion. A spatial variation map of pollutants is generated, showing the distribution of abnormally changing pollution parameters and their possible source locations. Data visualization is achieved through heatmaps or contour maps for easier analysis. Parameter settings and results for each analysis step are recorded for subsequent evaluation and validation. A gradient descent algorithm is designed to optimize the calculation of pollutant diffusion paths. The algorithm should consider the influence of abnormal changes in water quality parameters, flow velocity, flow direction, and other environmental factors. A loss function, such as the goodness of fit based on pollutant concentration and spatial variation, is determined as the optimization objective. The goal is to minimize the loss function to find the optimal diffusion path. Iterative calculations are performed using the gradient descent method based on the initial path. In each iteration, the path is adjusted according to the current path's loss value, and the direction and step size are updated to find a better solution. In each iteration, the loss value of the current path is calculated, and the direction and range of the path are dynamically adjusted based on the fluid dynamics model and changes in water quality parameters to ensure the rationality of the calculated path. When the predetermined number of iterations or the loss value converges, stop the iteration and record the final pollution diffusion path. Ensure the path reflects the potential pollution sources and their diffusion characteristics. Validate the results by comparing them with historical data and field monitoring results to assess the accuracy of the estimated path. If necessary, further adjust the model parameters and algorithm settings based on the assessment results. Compile the estimated potential pollution diffusion paths and their analysis results into a report. The report should include a visualization of the path, analysis of key parameters, the possible locations of pollution sources, and their impact range.The report should describe in detail the implementation process, parameter settings, and results of each step to make it operable and valuable for reference.

[0043] Step S4: Conduct intelligent source tracing patrols based on potential pollution diffusion paths and locate polluted water sources to obtain the pollution source points;

[0044] In this embodiment, a cruise route for the intelligent unmanned surface vessel (USV) is designed based on the previously calculated pollution diffusion path. The cruise route should cover potential pollution sources and affected areas to ensure real-time monitoring and location of pollution sources. The cruise route is divided into multiple monitoring points, each including specific latitude and longitude information and a predetermined dwell time for detailed water quality testing and data collection. The cruise speed and data collection frequency are determined. For example, the USV is set to cruise at a speed of 5 kilometers per hour and stop at each monitoring point for 5 minutes to conduct water quality testing. The cruise route is optimized by considering environmental factors such as water flow speed and wind direction to ensure the USV can reach key monitoring areas in the shortest possible time. The water quality monitoring module on the USV, including multi-dimensional water quality sensors, is activated. The sensors are ensured to be calibrated and undergo a self-check before the cruise begins to confirm their normal operating status. During the cruise, water quality parameters such as pH, dissolved oxygen, turbidity, ammonia nitrogen, and total phosphorus are collected in real time to ensure data accuracy and reliability. The data collection frequency is set, for example, recording water quality parameters once per minute, and the collected data is transmitted in real time to the USV's control system for storage. Each data record should include a timestamp, location, monitoring parameters, and corresponding environmental conditions for subsequent analysis and tracking. Based on the collected water quality parameter data, a pollution source tracing and location algorithm is designed. This algorithm should be able to identify abnormal changes in water quality parameters and analyze them in conjunction with potential pollution diffusion paths. Multivariate analysis methods (such as principal component analysis or cluster analysis) are used to identify patterns of water quality parameter changes to determine the possible location of the pollution source. Water quality data is analyzed in real time during the patrol. When a water quality parameter exceeds a set threshold, the system automatically marks the location and performs further data collection and analysis. The patrol route is dynamically adjusted by combining historical monitoring data and real-time parameter changes, prioritizing detailed detection at possible pollution source locations. The precise location of each monitoring point is recorded using a GPS positioning system and compared with changes in water quality parameters to verify the accuracy of pollution source location. By comparing water quality data from different monitoring points, the actual location of the pollution source is confirmed, and the characteristic parameters of the pollution source are recorded. Once the existence of a pollution source is confirmed at a monitoring point, more detailed sampling and analysis are immediately conducted to verify the type and concentration of pollutants. The sampling location, time, and water sample parameters are recorded. Identified contaminated water sources and their characteristics should be marked to ensure data integrity and traceability. The results of the contaminated water source identification should be compiled into a report, which should include the specific location of the pollution source, water quality parameters, types of potential pollutants, and their concentrations. The report should detail the implementation process, data collection, and analysis methods for each step, providing a scientific basis for subsequent remediation plans.

[0045] Step S5: Identify hidden sewage outlets based on the source of pollution and perform quantitative analysis of the pollution source components to obtain information on the water pollution components;

[0046] In this embodiment, all data related to the polluted water source are collected, including water quality monitoring results, historical pollution records, watershed characteristics, and environmental conditions. This data will serve as the basis for identifying hidden discharge outlets. Using Geographic Information System (GIS) tools, the location information of the polluted water source is integrated and analyzed with the watershed's hydrological data, land use, and historical distribution of discharge outlets to identify potential discharge outlet locations. Statistical analysis methods (such as control charts or outlier detection algorithms) are used to process the water quality data to identify the spatiotemporal distribution characteristics of abnormal pollutant concentrations. Emphasis is placed on the patterns of pollutant diffusion in the water body to find clues to unidentified pollution sources. Spatial analysis techniques are combined, and hotspot analysis (such as Getis-Ord Gi* statistics) is used to identify areas with abnormal pollutant concentrations, providing direction for identifying hidden discharge outlets. Based on the data analysis results, potential discharge outlet locations are selected for on-site investigation. Through field investigation, possible discharge outlet outlets, sewage flow, and surrounding environmental characteristics are observed. Multi-dimensional water quality sensors on intelligent unmanned surface vessels are used to take real-time samples at suspected locations to verify whether the water quality parameters match the data from the polluted water source. Based on on-site data, further confirm the location of hidden sewage outlets. Collect water samples near the confirmed hidden sewage outlets, ensuring the sampling process follows standard operating procedures (SOPs) to avoid sample contamination. Typically, collect water samples from multiple points for comparative analysis. Record the collection time, location, and environmental conditions (e.g., water temperature, flow rate) for each sample to ensure data integrity and traceability. Based on the characteristics of the water samples, select appropriate analytical methods for quantitative analysis of pollutant components. Commonly used methods include gas chromatography-mass spectrometry (GC-MS), high-performance liquid chromatography (HPLC), and atomic absorption spectrometry (AAS). Determine the analytical targets and standard methods for each pollutant to ensure data accuracy and reliability. For example, for heavy metal pollution, AAS can be used to analyze the concentration of metal ions such as lead, cadmium, and mercury. Compile the analytical results of water pollution components into a comprehensive report, which should include detailed information on the pollutant components, possible pollutant types, concentrations, and potential sources. Combining the characteristics of the polluted water source and the identification results of hidden sewage outlets, analyze the causes and diffusion pathways of pollution to provide a scientific basis for remediation plans. Based on information on pollutant composition, corresponding remediation recommendations are proposed, including the selection of pollutant removal technologies, monitoring frequency, and remediation measures. The recommendations should be practical and specific. Regular monitoring of pollution sources is recommended to assess the effectiveness of remediation efforts and adjust remediation measures based on monitoring data to ensure continuous improvement in water quality.

[0047] Step S6: Based on the information on water pollution components, conduct pollution source diffusion and evolution analysis, and based on the intelligent multi-mode network switching strategy, conduct intelligent early warning decision-making to construct a water pollution source tracing and early warning strategy.

[0048] In this embodiment, the water pollution component information obtained from previous analysis is collected, including pollutant types, concentrations and spatial distribution data. This data will provide basic input for the pollution source diffusion model.

[0049] Choose a suitable diffusion model, such as a Gaussian diffusion model, a Lagrangian particle tracing model, or a hydrodynamic model, to describe the diffusion process of pollutants in the water body. When selecting a model, consider environmental factors such as water flow velocity, wind speed, and temperature. Based on historical data and field monitoring results, set model parameters, such as water flow velocity (e.g., 0.5 m / s) and diffusion coefficient (e.g., 0.1 m / s). 2 The model is calibrated to improve its accuracy. Preliminary simulations are performed to verify the model's rationality and accuracy. Model parameters are continuously adjusted by comparing with actual monitoring data to ensure the model accurately reflects the diffusion characteristics of pollutants. Based on the determined parameters and model, the evolution of pollution source diffusion is simulated. Through numerical simulation, the spatial distribution changes of pollutants over different time periods are predicted, generating a time series of pollutant concentrations. Pollutant concentration distribution maps are recorded at each time step to ensure the completeness and visualization of the simulation results, enabling relevant personnel to intuitively understand the dynamics of pollutant diffusion. Simultaneously with the pollution source diffusion simulation, the performance of the multi-mode network environment is continuously monitored, including signal strength, bandwidth, and latency, to ensure the real-time performance and stability of data transmission. A regular monitoring frequency (e.g., every 5 minutes) is set to record network performance data for subsequent intelligent switching decisions. An intelligent switching algorithm based on real-time network performance data is designed. When a network performance degradation is detected (e.g., signal strength below -70dBm), the system automatically switches to a stronger network (e.g., switching from Wi-Fi to 4G or 5G). Priority settings are introduced into the algorithm, such as prioritizing networks with higher bandwidth during data transmission to ensure real-time transmission of pollution diffusion simulation data and early warning information. Based on the intelligent switching algorithm's decisions, pollution source diffusion simulation results and water quality monitoring data are transmitted to the control center in real time. This ensures that early warning information can be sent to relevant management personnel quickly and accurately. During data transmission, the system automatically generates early warning information, including pollutant concentration, diffusion range, and potential affected areas. Clear early warning standards are established based on water quality monitoring data and pollution source diffusion simulation results. These typically include early warning thresholds for different pollutant concentrations; for example, an early warning is triggered when ammonia nitrogen concentration exceeds 0.5 mg / L. A three-level early warning mechanism is set: Level 1 for minor pollution, Level 2 for moderate pollution, and Level 3 for severe pollution, ensuring timely and effective tiered responses. During monitoring, the system compares current water quality parameters with the early warning standards in real time. When a parameter exceeds a set threshold, the corresponding early warning level is automatically triggered. Early warning information is rapidly transmitted to the decision-making center via a multi-mode network, ensuring that relevant personnel can promptly understand the pollution situation and take appropriate measures.

[0050] In this embodiment, see Figure 2 The diagram below illustrates the detailed implementation steps of step S1. In this embodiment, the detailed implementation steps of step S1 include:

[0051] The system locates the real-time position of the intelligent unmanned vessel and acquires multimodal network monitoring parameters based on the multimodal communication module.

[0052] Calculate the latency, packet loss rate, and bandwidth parameters of the multimodal network monitoring parameters;

[0053] Based on latency, packet loss rate and bandwidth parameters, real-time multi-index parameter change analysis is performed to generate multi-monitoring index change characteristics for each type of network.

[0054] Real-time network status assessment is performed based on the changing characteristics of multiple monitoring indicators for each type of network to generate a real-time network status assessment value for each type of network.

[0055] Dynamic multi-mode network intelligent switching decisions are made based on the real-time network status assessment values ​​of each network, and an intelligent multi-mode network switching strategy is constructed.

[0056] In this embodiment, a multi-mode communication module is installed on the unmanned surface vessel (USV), supporting multiple communication protocols (such as 4G, 5G, Wi-Fi, and satellite communication). The module is designed to switch between different network environments to obtain real-time location information. A Global Navigation Satellite System (GNSS) receiver is configured to acquire the USV's geographic location data in real time and integrate it with the multi-mode communication module, ensuring the accuracy and real-time nature of the positioning information. Real-time location data of the USV, including longitude, latitude, altitude, and speed, is collected periodically (e.g., every second) via the multi-mode communication module, while network monitoring parameters are also collected. The collected location data and network monitoring parameters are transmitted to the ground control center via a pre-defined communication protocol for subsequent analysis. During data transmission, the timestamps of each sent and received data packet are recorded to calculate network latency. The latency is calculated by subtracting the sending time from the receiving time, and the latency of each data packet is recorded. Statistical methods (such as mean and variance) are used to process the latency data, identifying latency trends and outliers to ensure data accuracy and reliability. The packet loss rate is calculated by comparing the number of successfully received data packets with the total number of sent data packets. Packet loss rate = (Number of lost packets / Total number of sent packets) × 100%. Record the packet loss rate under different network modes and analyze its impact on the real-time positioning of the unmanned surface vessel (USV). For example, the packet loss rate may increase significantly in areas with weak signals. Regularly monitor network bandwidth using network bandwidth testing tools (such as iperf) and record bandwidth changes over different time periods. Analyze the stability and trends of bandwidth to identify periods of network congestion or instability for subsequent network status assessment. Organize the latency, packet loss rate, and bandwidth parameter data into tables, arranged chronologically for comprehensive analysis. Ensure that the timestamps of the data correspond to the USV's location information so that network performance and location data can be correlated in subsequent analysis. Use data mining and statistical analysis methods (such as time series analysis) to extract the changing characteristics of multiple monitoring indicators for each network. Identify the correlation and changing trends between latency, packet loss rate, and bandwidth. Generate charts (such as line charts and bar charts) to display the changes in monitoring indicators under different network conditions for intuitive analysis. Analyze the relationship between the changing characteristics of network monitoring indicators and the operational status of the USV. For example, is an increase in packet loss rate related to an increase in latency within a specific area? Record and analyze the results to provide a basis for subsequent network status assessments. Establish assessment formulas by weighting network status assessment indicators based on latency, packet loss rate, and bandwidth. For example, a comprehensive score of latency, packet loss rate, and bandwidth can be used as the primary basis for network status assessment. Set assessment thresholds to determine the assessment levels (e.g., excellent, good, poor) for different network states. Calculate the real-time status assessment value for each network using the established formulas based on the real-time collected monitoring parameters. Record the assessment results for each network over different time periods. Generate an assessment report detailing the status assessment value and its changes for each network to provide a basis for subsequent decision-making.Based on real-time network status assessment values, a dynamic switching decision algorithm is designed. This algorithm should be able to intelligently determine when to switch networks based on changes in the assessment values. Switching conditions are set; for example, when the assessment value of a certain network falls below a set threshold, the algorithm automatically switches to another network to maintain the positioning accuracy and communication stability of the unmanned surface vessel (USV). An intelligent multi-mode network switching strategy is implemented in the USV's control system, monitoring network status in real time and performing dynamic switching. The switching process is ensured to be seamless, avoiding any impact on real-time positioning. Each step of the switching operation is recorded, including the switching time, the network status before and after the switch, and its impact, for subsequent analysis. The implemented switching strategy is systematically tested to monitor the positioning accuracy and stability of the USV under different network conditions. Based on the test results, the strategy is optimized to ensure that the switching decisions can adapt to changing network environments, improving the USV's real-time positioning capabilities and communication quality.

[0057] In this embodiment, see Figure 3 The diagram below illustrates the detailed implementation steps of step S2. In this embodiment, the detailed implementation steps of step S2 include:

[0058] Real-time multi-dimensional water quality monitoring parameters of the intelligent unmanned vessel are obtained from multi-dimensional water quality sensors.

[0059] The average values ​​of pH, dissolved oxygen, turbidity, and conductivity are calculated based on the real-time multi-dimensional water quality monitoring parameters of the water area.

[0060] Water quality changes at multiple sampling points are analyzed based on the average values ​​of pH, dissolved oxygen, turbidity, and conductivity to generate water quality change characteristics at multiple sampling points.

[0061] The parameters of water quality change characteristics at multiple sampling points are analyzed to determine the evolution of change trends, and multi-parameter change trend curves are generated.

[0062] Anomaly changes in pollution parameters are detected on the multi-parameter trend curves, and abnormally changing pollution parameters are marked.

[0063] In this embodiment, multi-dimensional water quality sensors, including pH, dissolved oxygen, turbidity, and conductivity sensors, are installed on the intelligent unmanned surface vessel (USV). These sensors should have real-time data acquisition capabilities to ensure accurate monitoring of water quality parameters under different aquatic environments. During installation, ensure a good connection between the sensors and the USV's control system, enabling real-time data transmission and necessary calibration to improve measurement accuracy. Configure a data acquisition system to periodically (e.g., every second or every minute) acquire water quality monitoring parameters from the sensors. Record data from each sensor, including pH value, dissolved oxygen concentration (mg / L), turbidity (NTU), and conductivity (μS / cm). Transmit the collected data in real-time to a ground control center or cloud platform via a wireless communication module for subsequent analysis and processing. Store the real-time monitoring data in a secure database to ensure data integrity and traceability. Record the timestamp, sensor location, and environmental conditions (e.g., temperature, weather) for each data point to provide a basis for subsequent analysis. Process the collected water quality monitoring data, removing outliers and noise to ensure accuracy. Statistical methods such as moving average or median filtering can be used to smooth the data. Data is categorized by sampling point and time to facilitate subsequent analysis of water quality changes at different sampling points. For each sampling point, the average values ​​of pH, dissolved oxygen, turbidity, and conductivity are calculated. For example, if a sampling point records 60 data points within 10 minutes, the arithmetic mean of these data points is calculated. The average value for each sampling point is recorded and the results are stored in a database for later analysis. Change characteristics are extracted based on the average water quality parameters from different sampling points. By comparing the average values ​​of each sampling point, the spatial distribution characteristics and change patterns of water quality are identified. Statistical analysis methods (such as analysis of variance) are used to compare water quality parameters at different sampling points to determine which sampling points show significant differences. Water quality change characteristic maps are plotted for each sampling point. For example, heatmaps or line graphs can be used to display the changing trends of pH, dissolved oxygen, turbidity, and conductivity at each sampling point. The trends and characteristics of the changes are recorded, and possible causes are analyzed, such as the influence of the surrounding environment and the presence of pollution sources, providing a basis for subsequent water quality monitoring. Data visualization tools are used to plot the water quality parameters of each sampling point as multi-parameter change trend curves. Each curve should clearly indicate the time axis and parameter values ​​for easy observation of trends. During plotting, ensure that curves for different parameters use different colors or styles for easy differentiation. By observing the trend curves of multiple parameters, analyze the correlation between different water quality parameters. For example, is an increase in turbidity related to a decrease in dissolved oxygen, exploring the intrinsic connections between water quality changes? Record the results of trend analysis to provide a reference for anomaly detection. Select appropriate anomaly detection methods, such as statistical control charts, machine learning algorithms, or threshold-based detection methods. Set reasonable thresholds to identify abnormal changes.For each water quality parameter (pH, dissolved oxygen, turbidity, and conductivity), a normal range is set based on historical data; values ​​exceeding this range are considered abnormal. During real-time monitoring, anomalies are continuously detected at the sampling points. Whenever an abnormal change is detected, the parameter is immediately recorded and marked, along with a timestamp and sampling point information. An alarm system is generated to promptly alert the control center when an abnormality occurs, ensuring a rapid response. Anomaly detection results are summarized and a report is generated, detailing the abnormal changes at each sampling point, including the abnormal parameter, the time of occurrence, and its potential impact. Based on the anomaly detection results, potential pollution sources and their impact on water quality are analyzed, providing a scientific basis for subsequent water quality management.

[0064] In this embodiment, see Figure 4 The diagram below illustrates the detailed implementation steps of step S3. In this embodiment, the detailed implementation steps of step S3 include:

[0065] A gradient analysis of water pollution changes is performed on abnormally changing pollution parameters to generate gradients of water pollution parameter changes.

[0066] Inverse spatial variation analysis of water quality is performed based on the gradient of water pollution parameter changes to generate spatial distribution data of water pollution.

[0067] The fastest pollution concentration direction is identified from the spatial distribution data of water pollution, and the fastest pollution concentration direction is extracted.

[0068] Based on the direction of the fastest pollution concentration, gradient descent iterative source tracing calculation is performed to generate a water pollution source tracing trajectory.

[0069] Water flow dynamics evolution prediction is performed based on water pollution source tracing trajectories to generate potential pollution diffusion paths.

[0070] In this embodiment, data on labeled abnormally changing pollution parameters are collected, including water quality parameters (such as pH, dissolved oxygen, turbidity, and conductivity) and their timestamps from different sampling points. Data integrity is ensured by removing outliers and noise to guarantee the accuracy of subsequent analysis. Data normalization is performed to facilitate comparison and analysis between different parameters and avoid analytical biases caused by different dimensions. Spatial interpolation methods (such as Kriging interpolation) are used to spatially interpolate the water quality parameters at each sampling point, generating a continuous distribution map of the water quality parameters. By comparing the pollution parameters at adjacent sampling points, the gradient of pollution change is calculated. Based on the pollution change gradient data, a spatial variation model of water pollution is established. The model needs to consider the influence of environmental factors such as water flow direction, wind speed, and temperature on water quality changes. A fluid dynamics model is used to simulate the diffusion characteristics of pollutants in the water body, and the trajectory of pollutants in the water body is analyzed by combining water flow velocity and direction. The source location of pollutants is determined through reverse analysis. Based on the gradient change, the diffusion path of pollutants is calculated backward from high-concentration areas to low-concentration areas, and the pollution concentration at each sampling point is calculated. Record the concentration changes at each point to generate spatial distribution data of pollutants in the water body, indicating the highest concentration and diffusion range of pollutants. Summarize the spatial distribution data of water pollution obtained from the reverse analysis to generate a complete water pollution distribution map. Ensure the map shows the trend of pollution concentration changes and the location of pollution sources, providing a foundation for subsequent pollution diffusion analysis. Analyze the concentration direction using the generated spatial distribution data of water pollution. Identify the fastest direction of pollution diffusion by calculating the rate of change of concentration in different directions. Use the directional derivative calculation method to identify the direction of most significant concentration change. This process can be achieved through methods such as local weighted regression to improve the accuracy of direction identification. Mark the fastest pollution concentration direction on the pollution distribution map, using arrows or other symbols to represent the diffusion trend and speed. Ensure the graphics are intuitive for convenient subsequent analysis and decision-making. Record the specific values ​​of the fastest pollution concentration direction and the corresponding sampling point information, generating a detailed report describing the dynamic characteristics of pollution diffusion and its potential impact. Based on the fastest pollution concentration direction, construct a gradient descent source tracing model. This model should be able to simulate the trajectory of pollutants in the water flow and perform reverse calculations. To ensure the accuracy and effectiveness of the model, source-tracing parameters such as pollutant concentration, diffusion rate, and environmental conditions are set. Iterative calculations are performed to progressively advance towards the source. At each step, the pollutant concentration distribution and diffusion direction are updated based on hydrodynamics and concentration changes. The results of each iteration are recorded, generating a source-tracing trajectory for the pollutants, including concentration changes and time information at each point. The generated water pollution source-tracing trajectory is plotted to show the estimated path of the pollution source. Visualization provides an intuitive representation of the pollutant diffusion process and its environmental impact. Based on the source-tracing trajectory data, a hydrodynamic evolution prediction model is constructed. The model must consider the influence of water flow characteristics, temperature, wind speed, and other environmental factors on pollution diffusion.Numerical simulation methods (such as the finite element method or finite difference method) are used to model water flow dynamics and predict future diffusion paths of pollutants. Evolutionary prediction models are employed to calculate the diffusion paths and concentration changes of pollutants over future time periods. Potential pollution diffusion path maps are generated, indicating areas that may be affected by pollution. The prediction results are recorded, and the diffusion characteristics of pollutants under different environmental conditions are analyzed to provide a scientific basis for countermeasures.

[0071] In this embodiment, step S4 includes the following steps:

[0072] Intelligent source tracing patrols are conducted based on potential pollution diffusion paths, and real-time patrol infrared imaging images are collected.

[0073] Image brightness enhancement processing is performed on real-time cruise infrared imaging images to obtain brightness-optimized cruise infrared images;

[0074] Visual recognition of water obstacles is performed on brightness-optimized cruise infrared images, and water obstacle nodes are marked;

[0075] Calculate the target water area based on potential pollution diffusion paths;

[0076] Intelligent path planning is performed based on water obstacle nodes and target water points to generate intelligent source tracing and patrol routes;

[0077] The polluted water source is located by tracking and locating the polluted water source based on the intelligent source tracing and navigation route.

[0078] In this embodiment, a cruise plan for the intelligent unmanned surface vessel (USV) is formulated based on previously generated potential pollution diffusion paths. This plan should consider factors such as water flow characteristics, wind speed, weather conditions, and water depth to optimize cruise efficiency and safety. During planning, the cruise path is ensured to cover all potential pollution sources and their diffusion directions for comprehensive monitoring of water quality changes. The USV's infrared imaging equipment is configured to enable real-time imaging under various environmental conditions. The equipment should possess high sensitivity and high resolution to capture minute temperature changes and water obstacles. A data acquisition frequency is set, typically one infrared image per second, to obtain continuous monitoring image data and ensure real-time assessment of water conditions. During the cruise, infrared images are acquired in real-time and stored in the USV's computing system. Each image should include a timestamp and location coordinates for subsequent analysis and processing. The image format and resolution are ensured to be suitable for subsequent processing, typically using TIFF or PNG formats to preserve image quality. Image processing algorithms are used to enhance the brightness of the acquired infrared images. Common methods include histogram equalization and gamma correction, which effectively improve image contrast and brightness, making water features more prominent. During processing, the brightness changes of each image are monitored to ensure that the enhanced image clearly displays obstacles and other features of interest in the water area. The processed images are visually inspected to ensure the brightness enhancement effect meets expectations. Several key images can be selected for quantitative analysis to evaluate the quality and stability of the enhancement effect. Based on the enhanced infrared images, image recognition algorithms (such as convolutional neural networks, CNNs) are applied to visually identify obstacles in the water area. These obstacles may include floating objects, shoreline vegetation, and other objects that may potentially affect the navigation of the unmanned vessel. The dataset used to train the model should include infrared images of various water area obstacles to improve the accuracy and robustness of recognition. During recognition, the algorithm should automatically mark the coordinates of obstacles and generate corresponding node information, including the type, size, location, and relative distance of the obstacle to the unmanned vessel. Relevant information for each obstacle node is recorded and combined with real-time cruise data for subsequent path planning and decision support. Target water areas are calculated based on potential pollution diffusion paths. These points should be located in the core area of ​​pollution diffusion, typically where pollution concentration is highest. Using previous water quality monitoring data and a flow model, the pollution characteristics of the target water areas are evaluated to determine their impact on water quality. This study integrates data on potential pollution paths, obstacle nodes, and target water areas, analyzing the relationships between these points. It ensures that the selection of target water areas best reflects the dynamic changes in pollution diffusion. Based on obstacle nodes and target water areas, intelligent path planning is performed using path planning algorithms (such as A* or Dijkstra's algorithm). The goal is to generate an optimal cruising path that avoids obstacles and quickly reaches the target water area.During the planning process, factors such as water flow characteristics, the relative positions of obstacles, the speed of the unmanned surface vessel (USV), and safe distances are considered. The generated intelligent source-tracing cruise path is visualized and displayed on the USV's control interface. This graphical display facilitates real-time monitoring and route adjustment by the operator. During actual cruises, the path is updated in real time to address unforeseen obstacles or environmental changes, ensuring the safety and efficiency of the USV. A high-precision positioning system is configured on the USV to acquire its position in real time and integrate it with the path planning system for accurate tracking of polluted water sources. Combined with real-time water quality monitoring data, the pollution level of the area traversed by the USV is continuously assessed to determine if it is approaching a pollution source. During the cruise, potential polluted water sources are gradually located based on changes in water quality monitoring data and their relationship with the intelligent source-tracing cruise path. The pollution characteristics of each point are recorded, including concentration, type, and impact range. By comparing water quality data from different time periods, the existence and changes of pollution sources are confirmed, and detailed monitoring reports are generated. The tracking and location results are fed back to the control center for environmental remediation and subsequent monitoring. The accuracy of all data records and analysis results is ensured to support decision-making. Based on the location results, formulate corresponding governance plans and take timely measures to reduce or eliminate the impact of water pollution on the ecological environment.

[0079] In this embodiment, the specific steps for conducting intelligent source tracing patrols based on potential pollution diffusion paths and acquiring real-time patrol infrared imaging images are as follows:

[0080] The intelligent unmanned vessel activates its covert tracking mode, turns off visible light, and enables infrared cameras and lidar scanning. It then conducts sequential source tracing cruises along potential pollution diffusion paths and collects real-time cruise infrared imaging images.

[0081] In this embodiment, the "stealth tracking mode" is selected in the unmanned surface vessel's control system. This mode aims to reduce interference with the aquatic environment and decrease detectability. In this mode, all visible light sources are turned off, ensuring the unmanned vessel is difficult to detect at night or in complex environments. The unmanned vessel's system status is checked to ensure its battery is fully charged and all sensors (including the infrared camera and lidar) are functioning normally. A system self-test is performed to ensure the sensors can acquire data and process it in real time. The infrared camera is configured to acquire high-resolution images in low-light conditions. A high sensitivity and an appropriate frame rate (e.g., 30 frames per second) are set to capture subtle temperature changes and underwater features. The lidar system should be calibrated to ensure accurate distance and environmental feature measurements. The laser emission frequency and scanning angle are set to obtain comprehensive aquatic environmental data. Based on previous pollution diffusion path data, a cruise route for the unmanned vessel is planned. This route should cover all potential pollution sources and diffusion areas to ensure comprehensive monitoring and information collection. The route planning considers water flow direction, wind speed, obstacle locations, and the unmanned vessel's speed to optimize cruise efficiency and ensure safety. The unmanned vessel's navigation system is activated, and it cruises along the planned route. Utilizing autonomous driving technology, the unmanned surface vessel (USV) is able to autonomously avoid obstacles and adjust its course in real time. During cruise, the USV's position and speed are continuously monitored to ensure it stays within the designated path. Precise positioning is achieved through GPS and an inertial navigation system (INS), providing real-time feedback. An infrared camera is activated to begin acquiring real-time infrared images of the water area. Appropriate exposure time and gain are set to improve image clarity and contrast. The camera is ensured to be highly sensitive to temperature changes at the water surface and underwater. The parameters of the infrared camera are dynamically adjusted according to environmental conditions (such as water temperature, air temperature, and weather changes) to adapt to different monitoring needs. The acquired infrared images are stored in real-time in the USV's computing system, ensuring the image format (such as TIFF or PNG) is suitable for subsequent processing. The timestamp and location coordinates of each image are recorded for later analysis and retrospection. A reasonable storage strategy is implemented to ensure the system can effectively manage storage space and avoid data loss during long-term cruises. A lidar system is activated to scan the environment. The lidar's scanning frequency and range are set to obtain comprehensive information on the water terrain and obstacles. Typically, the scanning range of the lidar is set within 30 meters to ensure high-precision measurements. During patrol, the lidar should collect distance data in real time and generate environmental point cloud data to construct a 3D model of the water area. Infrared images are fused with lidar data to generate a comprehensive environmental information map. Data fusion algorithms (such as Kalman filtering) are used to integrate data from both sensors, improving the accuracy of environmental identification. By analyzing thermal features in the infrared images and the point cloud data generated by the lidar, obstacles and potential pollution sources in the water area can be further identified, ensuring the safe operation of the unmanned vessel during patrol.At the control center of the unmanned surface vessel (USV), a real-time monitoring interface is set up to display the current cruise status, infrared images, lidar data, and the USV's location. This ensures that operators can be aware of environmental changes and potential risks during the cruise. The monitoring system should have an alarm function, automatically issuing an alert when abnormalities are detected (such as suddenly appearing obstacles or abnormal temperatures) so that operators can take timely action.

[0082] In this embodiment, the specific steps of step S5 are as follows:

[0083] When the intelligent unmanned vessel arrives at the polluted water source, it activates the infrared camera to perform a panoramic scan of the target water area and obtain a panoramic infrared image of the target water area.

[0084] Use panoramic infrared imaging of the target water area to identify and mark hidden sewage outlets;

[0085] Water quality samples were taken from the surrounding area of ​​the hidden sewage outlet, and water quality sampling parameters of the hidden sewage outlet were extracted.

[0086] Water pollution is classified based on water quality sampling parameters from hidden sewage outlets to generate water pollution types;

[0087] Pollutant concentration values ​​are calculated based on water quality sampling parameters from hidden sewage outlets.

[0088] Quantitative analysis of pollution source components is conducted based on water pollution type and pollutant concentration values ​​to obtain information on water pollution components.

[0089] In this embodiment, after the unmanned surface vessel (USV) reaches the polluted water source, it activates the infrared camera to perform a panoramic scan. The camera is ensured to be adjusted to its optimal operating state, including appropriate exposure time and gain settings, to capture clear images under different lighting conditions. The infrared camera is set to wide-angle mode to cover a larger area of ​​water, ensuring comprehensive information about hidden sewage outlets and their surrounding environment is obtained. The infrared camera is activated for an all-around scan. A scanning cycle is set, for example, once per second, and a complete infrared imaging dataset is generated, which will be used for subsequent analysis. The collected data should include the timestamp of each frame, the USV's location, and the environmental conditions of the water area (such as water temperature, air temperature, etc.) for subsequent processing and analysis. The acquired panoramic infrared images are stored in real-time in the USV's computing system, ensuring the image format is suitable for subsequent processing (such as TIFF or PNG) to preserve image quality. Relevant information for each image is recorded, and the data is securely stored for subsequent analysis and retrospection. The panoramic infrared images are processed to identify hidden sewage outlets. Image processing techniques (such as edge detection, region growing algorithms, etc.) are used to analyze areas of abnormal temperature in the images, which are often associated with sewage outlets. Machine learning models are applied to train the model to identify the features of hidden sewage outlets, ensuring the model can accurately identify the location and shape of the outlets. The dataset should include images of different types of sewage outlets to improve the model's robustness. Once a hidden sewage outlet is identified, its coordinates and relevant information (such as temperature and relative location) are immediately marked on the infrared imaging image. Graphical marking and annotation can be used to add markers and annotations to the image. Detailed information for each marker is recorded, including location, identification criteria, and environmental conditions, for subsequent water quality sampling and analysis. Based on the marked locations of hidden sewage outlets, a water quality sampling plan for the surrounding area is developed. The selection of sampling points should reflect the impact of the sewage outlets on water quality; multiple sampling points near the outlets should generally be selected. The sampling depth and location are determined, considering factors such as water flow direction, wind speed, and water temperature to avoid interference during the sampling process. A high-precision water sampler is used to collect water samples at the predetermined locations. Multiple water samples should be collected at each sampling point to ensure data reliability and accuracy. The specific location, sampling time, and sampling depth of each sampling point are recorded for subsequent data analysis. Collected water sample information is recorded in a database, including sample number, sampling location, time, and relevant environmental data. The integrity and traceability of all data are ensured. Water quality sampling parameters around hidden sewage outlets are analyzed, typically including pH, dissolved oxygen, turbidity, and conductivity. These parameters are measured in real time using high-precision water quality sensors, and the data is recorded. Data analysis employs statistical methods to calculate the mean, standard deviation, etc., of each parameter to assess the overall water quality. Based on the collected water quality parameters, classification algorithms (such as Support Vector Machines (SVM) or decision trees) are used to classify water pollution. These algorithms should be trained to identify different types of water pollution sources.The pollution type of the sample is determined by comparing sample data with known pollution types, and pollution classification results are generated. The pollution classification results for each sampling point are recorded, and a detailed water pollution classification report is generated. The report should include information such as pollution type, possible pollution sources, and their impacts, providing a basis for subsequent analysis. Pollutant concentration values ​​are calculated based on water quality sampling parameters. Calculations are typically performed using known formulas or models, such as extrapolation based on the chemical components and concentrations in the water sample. Multiple methods are used for verification, such as comparison with standard solutions, to ensure the accuracy of the calculation results. The calculated pollutant concentration values ​​are recorded in a database and linked to the sampling point information for subsequent analysis. Data integrity and traceability are ensured. A detailed pollutant concentration value report is generated, including the concentration value and corresponding pollution type for each sampling point, providing basic data for subsequent pollution source analysis. Based on the water pollution type and pollutant concentration values, an appropriate component analysis method, such as gas chromatography (GC) or high-performance liquid chromatography (HPLC), is selected to quantitatively analyze the pollutant components in the water sample. The experimental conditions required for the analysis are determined, including sample processing, instrument calibration, and analytical parameter settings. Pollutant composition analysis is conducted in the laboratory using selected instruments to analyze water samples. Based on the analysis results, the concentration and relative proportion of each pollutant are quantified. During the analysis, the conditions and results of each step are recorded to ensure accuracy and repeatability. The results of the pollution source composition analysis are summarized to generate a detailed water pollution composition information report. The report should include the concentration, composition ratio, and potential impact of each pollutant. Based on the pollution composition information, a scientific basis is provided for subsequent remediation measures to ensure the restoration and protection of the aquatic environment.

[0090] In this embodiment, the specific steps of step S6 are as follows:

[0091] Analyze the sewage flow distribution of hidden sewage outlets to obtain sewage flow distribution data;

[0092] Based on the data of sewage outlet water flow distribution and water pollution composition, the diffusion evolution of pollution sources is analyzed to generate water pollution diffusion evolution characteristics.

[0093] The pollution risk level is assessed based on the characteristics of water pollution diffusion and evolution, and the water pollution diffusion risk assessment results are obtained.

[0094] Intelligent early warning decisions are made based on the risk assessment results of water pollution spread and the intelligent multi-mode network switching strategy, and a water pollution source tracing and early warning strategy is constructed.

[0095] In this embodiment, a flow sensor is installed near the hidden sewage outlet to monitor the water flow rate and volume in real time. The sensor should have high accuracy and reliability to ensure data accuracy. Based on the water flow characteristics, a suitable flow measurement device (such as an electromagnetic flowmeter or ultrasonic flowmeter) is selected. Multi-point sampling is performed to record the flow data of the sewage outlet at different time periods to capture the changing trend of the water flow. Typically, the sampling period is once per minute, and the data record includes a timestamp, flow value, and environmental conditions (such as water level and air temperature). Based on the collected flow data, a hydrodynamic model (such as a two-dimensional flow model) is used to model the water flow distribution at the sewage outlet. The model should consider the direction and velocity of the water flow and the geometry of the sewage outlet to generate a distribution map of the sewage flow. Numerical simulation methods (such as the finite element method or finite difference method) are used to simulate the water flow, identifying areas of velocity variation, especially the water flow dynamics around the sewage outlet. Based on the water flow distribution data of the sewage outlet, a mathematical model of pollutant diffusion is established. Diffusion equations (such as convection-diffusion equations) are used to describe the movement and diffusion process of pollutants in the water body. The model should integrate water flow rate, pollutant concentration, and their physicochemical properties (such as solubility and settling rate) to achieve accurate diffusion simulation. Numerical simulation techniques (such as Monte Carlo simulation or the Lagrangian method) should be used to simulate the pollutant diffusion process. Initial conditions (such as pollutant concentration at the discharge outlet) and boundary conditions should be set to simulate the dynamic changes of pollutants in the water body. The distribution of pollutant concentrations at different time points should be recorded, and their spatial variation characteristics should be analyzed to determine the rate and extent of pollutant diffusion. A pollution source diffusion evolution characteristic map should be generated to show the concentration changes of pollutants in the water body and their diffusion paths. The main directions of pollutant diffusion and potential impact areas should be identified through analysis of the results. Key parameters (such as maximum concentration and diffusion range) should be recorded to provide a basis for subsequent risk assessment and early warning decisions. Pollution risk level assessment standards should be developed based on the characteristics and diffusion properties of water pollution. The assessment standards should include indicators such as pollutant concentration, diffusion rate, and environmental sensitivity of the affected area. Thresholds for different risk levels should be set, such as low, medium, and high risk, to ensure the operability and scientific validity of the assessment results. The pollution risk level for each area should be calculated based on the diffusion evolution characteristic data. Using established standards, pollutant concentrations and diffusion characteristics are compared with risk levels to assess the risk of specific areas. Statistical analysis methods (such as multiple regression analysis) are employed to quantify the impact of different factors on the risk level, ensuring the accuracy of the assessment results. The risk assessment results for each area are recorded, and a detailed risk assessment report is generated. The report should include the risk level, potential impacts, and corresponding management recommendations to support subsequent decision-making. Based on the pollution diffusion risk assessment results, an intelligent early warning decision-making model is designed. The model should combine real-time monitoring data and historical data to assess the dynamic changes in pollution risk.A warning threshold is set; when the pollution risk exceeds the set value, the system automatically issues an alarm to facilitate timely response. The warning decision model integrates an intelligent multi-mode network switching strategy to ensure stable and real-time data transmission across different network environments (such as 4G, 5G, and Wi-Fi). This ensures that warning information is promptly delivered to relevant management departments and the public, supporting emergency response. Once the system issues a warning, the corresponding emergency response mechanism is immediately activated, including monitoring of the affected area, control of pollution sources, and implementation of remediation measures. Feedback data after the warning implementation is collected to evaluate the effectiveness of the warning strategy, and adjustments and optimizations are made based on the feedback results to ensure that subsequent warning work is more scientific and effective.

[0096] In this embodiment, the specific steps for constructing a water pollution source tracing and early warning strategy based on the water pollution diffusion risk assessment results and the intelligent multi-mode network switching strategy are as follows:

[0097] The risk assessment results for the spread of water pollution are specifically divided into three levels of early warning;

[0098] The intelligent early warning decision-making specifically refers to:

[0099] Level 1 alarm: Water quality parameters slightly exceed the standard, data is recorded but no remote alarm is triggered;

[0100] Level 2 Alarm: Clear signs of pollution are detected. Based on the intelligent multi-mode network switching strategy, the optimal network is selected, and an early warning is sent to the monitoring center.

[0101] Level 3 alarm: A major pollution incident is confirmed, and the emergency response mechanism is activated.

[0102] In this embodiment, a clear three-level early warning standard is established based on water quality monitoring data and pollution diffusion characteristics. The standard for each early warning level should be specific and quantifiable. For example: Level 1 alarm: Water quality parameters slightly exceed the standard, exceeding the normal range but not reaching the warning line (e.g., pH slightly exceeding the range of 6.5-8.5). Level 2 alarm: Obvious signs of pollution are detected, such as the concentration of certain key pollutants (e.g., ammonia nitrogen, total phosphorus) exceeding the preset warning value (e.g., ammonia nitrogen exceeding 0.5 mg / L). Level 3 alarm: A major pollution event is confirmed, with pollutant concentrations reaching severely excessive levels (e.g., heavy metals exceeding national standards by 10 times). The water quality monitoring system collects water quality parameter data in real time, including pH, dissolved oxygen, turbidity, and specific pollutant concentrations. An automated data analysis system is set up to continuously evaluate the real-time data. Based on set thresholds and standards, the current water quality status is automatically determined, triggering the corresponding early warning level. Automatic evaluation is performed after each data collection (e.g., once per minute) to ensure timely response to changes in water quality. When water quality parameters slightly exceed standards (e.g., pH values ​​between 6.3-6.5 or 7.5-7.8), the system automatically records relevant data without triggering a remote alarm. This process should ensure data integrity for subsequent analysis. A data recording module is set up to automatically store the exceeded data in the database, including timestamps, exceedance values, and corresponding environmental conditions (e.g., temperature, flow rate). The system periodically generates data analysis reports summarizing the exceedance situations. These reports should include the exceeded parameters, frequency of occurrence, and potential environmental impacts; this data will provide a basis for subsequent risk assessments. Reports are archived to ensure data traceability and support possible subsequent investigations and remediation. Water quality parameters continue to be monitored. If a worsening trend is detected (e.g., continued exceedance), the system will enter a level-two alarm monitoring state to ensure dynamic tracking of water quality changes. When clear signs of pollution are detected (e.g., the concentration of a pollutant exceeds the set warning value), the system will immediately trigger a level-two alarm. At this time, the system will select the optimal network for data transmission based on an intelligent multi-mode network switching strategy. The selection of the optimal network is based on factors such as network stability, bandwidth, and latency, ensuring that alarm information can be sent to the monitoring center in a timely and accurate manner. The system automatically generates early warning information, including the exceeding water quality parameters, concentration values, location, and time, and sends the warning to the monitoring center via the selected network (e.g., 4G / 5G / Wi-Fi). This ensures that the monitoring center can react quickly upon receiving the warning information and take appropriate monitoring and response measures. Once a Level 2 alarm is triggered, the system will increase the monitoring frequency (e.g., once every 10 seconds) and continuously record changes in water quality parameters for rapid assessment of the pollution source. The monitoring center will analyze pollution trends based on real-time data and assess whether an escalation to a Level 3 alarm is necessary. When a major pollution event is confirmed (e.g., heavy metal concentrations exceeding the standard by 10 times, or the detection of toxic substances), the system will trigger a Level 3 alarm and immediately activate the emergency response mechanism.This mechanism includes various emergency measures, such as notifying relevant environmental protection departments and initiating on-site sampling and monitoring procedures. Once a Level 3 alarm is triggered, the system will automatically send an emergency notification to relevant management departments to ensure they are promptly informed of the severity and urgency of the incident. An on-site emergency monitoring team will be activated to quickly proceed to the site for sampling and analysis, confirming the pollution source and assessing the pollution extent. Following the emergency response, a detailed environmental investigation and water quality assessment will be conducted. Sampling data, analysis results, and emergency response measures at each step will be recorded to ensure data integrity and accuracy. A detailed emergency response report will be generated, including pollution source analysis, environmental impact assessment, and subsequent remediation recommendations, providing a reference for long-term environmental protection work.

[0103] In this embodiment, a water pollution source tracing and analysis system based on an intelligent unmanned vessel is provided, used to execute the water pollution source tracing and analysis method based on an intelligent unmanned vessel as described above, including:

[0104] The multimodal network switching module is used to obtain real-time location multimodal network monitoring parameters based on the multimodal communication module, make dynamic multimodal network intelligent switching decisions, and construct intelligent multimodal network switching strategies.

[0105] The pollution change detection module is used to acquire real-time multi-dimensional water quality monitoring parameters of the intelligent unmanned vessel based on multi-dimensional water quality sensors, and to detect and mark abnormal pollution parameter changes.

[0106] The reverse water quality source tracing module is used to perform reverse water quality spatial change analysis and gradient descent iterative source tracing calculation on abnormally changing pollution parameters in order to generate potential pollution diffusion paths.

[0107] The source tracing and navigation module is used to conduct intelligent source tracing and navigation based on potential pollution diffusion paths, and to track and locate polluted water sources to obtain the source of pollution.

[0108] The quantitative analysis module is used to identify hidden sewage outlets based on the source of pollution and to perform quantitative analysis of the pollution source components in order to obtain information on the components of water pollution.

[0109] The intelligent early warning module is used to analyze the evolution of pollution source diffusion based on water pollution component information and to make intelligent early warning decisions based on intelligent multi-mode network switching strategies, thereby constructing a water pollution source tracing and early warning strategy.

[0110] This invention utilizes multi-mode communication technology to acquire the real-time location of unmanned surface vessels (USVs) and dynamically optimizes network switching decisions based on real-time monitoring parameters of the multi-mode network. This enables USVs to continuously transmit data through optimal network connections even in complex environments, ensuring the stability and smoothness of data communication. By dynamically selecting the optimal communication method, USVs can avoid data loss due to communication instability, improving the reliability of mission execution, especially in complex aquatic environments, ensuring the real-time nature and accuracy of data. Through multi-dimensional water quality sensors, real-time water quality data (such as temperature, pH, dissolved oxygen, turbidity, etc.) is collected, allowing the module to effectively track dynamic changes in water quality. This module can promptly detect abnormal water quality changes and mark pollution change parameters, providing data support for subsequent pollution source tracing and treatment. This allows for the rapid identification of potential pollution risks and the prevention of pollution spread. Through reverse source tracing analysis of abnormal pollution parameters, the spatial location and possible diffusion paths of pollution sources can be inferred, providing a basis for accurately locating pollution sources and helping to accurately identify the root cause of water quality anomalies. Using a gradient descent algorithm for calculation optimization enables efficient and accurate identification of potential pollution sources. Through iterative model calculations, the accuracy and precision of pollution source tracing will be improved. Based on potential pollution diffusion paths obtained from reverse analysis, unmanned surface vessels (USVs) can be guided to conduct intelligent patrols and track pollution sources in real time. Real-time patrols enable USVs to effectively track the location of pollution sources and provide data support for subsequent pollution treatment. Intelligent source tracing patrols not only ensure efficient tracking but also avoid resource waste, making the patrol process more precise and energy-efficient. Based on information about polluted water sources, potentially hidden discharge outlets can be identified. Quantitative analysis of pollution source components can accurately identify the types and components of pollution sources (such as organic matter, heavy metals, pesticides, etc.), providing precise data support for treatment measures. Analysis of pollution sources can identify the specific types of pollutants, providing a basis for decision-making in water quality treatment and pollution source control. Accurate pollution component information will help formulate more effective treatment measures. Based on the prediction of pollution source diffusion evolution, combined with intelligent multi-mode network switching strategies, early warning signals of water pollution can be issued in advance. By predicting the diffusion trend of pollutants in advance, a time window is provided for emergency response and treatment measures. The intelligent early warning module, combined with an intelligent multi-mode network switching strategy, can make rapid early warning decisions based on real-time data, network conditions, and pollution source information, helping relevant departments to take timely action to deal with the spread of pollution.

[0111] Therefore, the embodiments should be considered as exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of the equivalents of the application are intended to be included within the invention.

[0112] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement it. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features of the invention herein.

Claims

1. A method for tracing and analyzing water pollution sources based on intelligent unmanned vessels, characterized in that, The intelligent unmanned vessel is equipped with multi-dimensional water quality sensors, a multi-mode communication module, and an infrared camera, and includes the following steps: Step S1: Obtain real-time location multimodal network monitoring parameters based on the multimodal communication module, make dynamic multimodal network intelligent switching decisions, and construct an intelligent multimodal network switching strategy; Step S2: Obtain real-time multi-dimensional water quality monitoring parameters of the intelligent unmanned vessel based on multi-dimensional water quality sensors, and detect and mark abnormal pollution parameter changes. Step S3: Perform reverse spatial variation analysis of water quality and gradient descent iterative source tracing calculation on abnormally changing pollution parameters to generate potential pollution diffusion paths; Step S4: Conduct intelligent source tracing patrols based on potential pollution diffusion paths and locate polluted water sources to obtain the pollution source points; Step S5: Identify hidden sewage outlets based on the source of pollution and perform quantitative analysis of the pollution source components to obtain information on the water pollution components; Step S6: Based on the information on water pollution components, conduct pollution source diffusion and evolution analysis, and based on the intelligent multi-mode network switching strategy, conduct intelligent early warning decision-making to construct a water pollution source tracing and early warning strategy; The specific steps of step S2 are as follows: Real-time multi-dimensional water quality monitoring parameters of the intelligent unmanned vessel are obtained from multi-dimensional water quality sensors. The average values ​​of pH, dissolved oxygen, turbidity, and conductivity are calculated based on the real-time multi-dimensional water quality monitoring parameters of the water area. Water quality changes at multiple sampling points are analyzed based on the average values ​​of pH, dissolved oxygen, turbidity, and conductivity to generate water quality change characteristics at multiple sampling points. The parameters of water quality change characteristics at multiple sampling points are analyzed to determine the evolution of change trends, and multi-parameter change trend curves are generated. Detect abnormal pollution parameter changes in multi-parameter trend curves and mark abnormally changing pollution parameters. Step S3 is as follows: A gradient analysis of water pollution changes is performed on abnormally changing pollution parameters to generate gradients of water pollution parameter changes. Inverse spatial variation analysis of water quality is performed based on the gradient of water pollution parameter changes to generate spatial distribution data of water pollution. The fastest pollution concentration direction is identified from the spatial distribution data of water pollution, and the fastest pollution concentration direction is extracted. Based on the direction of the fastest pollution concentration, gradient descent iterative source tracing calculation is performed to generate a water pollution source tracing trajectory. Based on the water pollution source tracing trajectory, predict the hydrodynamic evolution to generate potential pollution diffusion paths; The specific steps of step S4 are as follows: Intelligent source tracing patrols are conducted based on potential pollution diffusion paths, and real-time patrol infrared imaging images are collected. Image brightness enhancement processing is performed on real-time cruise infrared imaging images to obtain brightness-optimized cruise infrared images; Visual recognition of water obstacles is performed on brightness-optimized cruise infrared images, and water obstacle nodes are marked; Calculate the target water area based on potential pollution diffusion paths; Intelligent path planning is performed based on water obstacle nodes and target water points to generate intelligent source tracing and patrol routes; The polluted water source is located by tracking and locating the polluted water source based on the intelligent source tracing and navigation route.

2. The water pollution source tracing and analysis method based on intelligent unmanned vessels according to claim 1, characterized in that, The specific steps of step S1 are as follows: The system locates the real-time position of the intelligent unmanned vessel and acquires multimodal network monitoring parameters based on the multimodal communication module. Calculate the latency, packet loss rate, and bandwidth parameters of the multimodal network monitoring parameters; Based on latency, packet loss rate and bandwidth parameters, real-time multi-index parameter change analysis is performed to generate multi-monitoring index change characteristics for each type of network. Real-time network status assessment is performed based on the changing characteristics of multiple monitoring indicators for each type of network to generate a real-time network status assessment value for each type of network. Dynamic multi-mode network intelligent switching decisions are made based on the real-time network status assessment values ​​of each network, and an intelligent multi-mode network switching strategy is constructed.

3. The water pollution source tracing and analysis method based on intelligent unmanned vessels according to claim 1, characterized in that, The specific steps for conducting intelligent source tracing patrols based on potential pollution diffusion paths and acquiring real-time patrol infrared imaging images are as follows: The intelligent unmanned vessel activates its covert tracking mode, turns off visible light, and enables infrared cameras and lidar scanning. It then conducts sequential source tracing cruises along potential pollution diffusion paths and collects real-time cruise infrared imaging images.

4. The water pollution source tracing and analysis method based on intelligent unmanned vessels according to claim 1, characterized in that, The specific steps of step S5 are as follows: When the intelligent unmanned vessel arrives at the polluted water source, it activates the infrared camera to perform a panoramic scan of the target water area and obtain a panoramic infrared image of the target water area. Use panoramic infrared imaging of the target water area to identify and mark hidden sewage outlets; Water quality samples were taken from the surrounding area of ​​the hidden sewage outlet, and water quality sampling parameters of the hidden sewage outlet were extracted. Water pollution is classified based on water quality sampling parameters from hidden sewage outlets to generate water pollution types; Pollutant concentration values ​​are calculated based on water quality sampling parameters from hidden sewage outlets. Quantitative analysis of pollution source components is conducted based on water pollution type and pollutant concentration values ​​to obtain information on water pollution components.

5. The water pollution source tracing and analysis method based on intelligent unmanned vessels according to claim 1, characterized in that, The specific steps of step S6 are as follows: Analyze the sewage flow distribution of hidden sewage outlets to obtain sewage flow distribution data; Based on the data of sewage outlet water flow distribution and water pollution composition, the diffusion evolution of pollution sources is analyzed to generate water pollution diffusion evolution characteristics. The pollution risk level is assessed based on the characteristics of water pollution diffusion and evolution, and the water pollution diffusion risk assessment results are obtained. Intelligent early warning decisions are made based on the risk assessment results of water pollution spread and the intelligent multi-mode network switching strategy, and a water pollution source tracing and early warning strategy is constructed.

6. The water pollution source tracing and analysis method based on intelligent unmanned vessels according to claim 5, characterized in that, The specific steps for constructing a water pollution source tracing and early warning strategy based on the water pollution diffusion risk assessment results and intelligent multi-mode network switching strategy are as follows: The risk assessment results for the spread of water pollution are specifically divided into three levels of early warning; The intelligent early warning decision-making specifically refers to: Level 1 alarm: Water quality parameters slightly exceed the standard, data is recorded but no remote alarm is triggered; Level 2 Alarm: Clear signs of pollution are detected. Based on the intelligent multi-mode network switching strategy, the optimal network is selected, and an early warning is sent to the monitoring center. Level 3 alarm: A major pollution incident is confirmed, and the emergency response mechanism is activated.

7. A water pollution source tracing and analysis system based on intelligent unmanned vessels, characterized in that, The method for performing water pollution source tracing analysis based on intelligent unmanned vessels as described in claim 1 includes: The multimodal network switching module is used to obtain real-time location multimodal network monitoring parameters based on the multimodal communication module, make dynamic multimodal network intelligent switching decisions, and construct intelligent multimodal network switching strategies. The pollution change detection module is used to acquire real-time multi-dimensional water quality monitoring parameters of the intelligent unmanned vessel based on multi-dimensional water quality sensors, and to detect and mark abnormal pollution parameter changes. The reverse water quality source tracing module is used to perform reverse water quality spatial change analysis and gradient descent iterative source tracing calculation on abnormally changing pollution parameters in order to generate potential pollution diffusion paths. The source tracing and navigation module is used to conduct intelligent source tracing and navigation based on potential pollution diffusion paths, and to track and locate polluted water sources to obtain the source of pollution. The quantitative analysis module is used to identify hidden sewage outlets based on the source of pollution and to perform quantitative analysis of the pollution source components in order to obtain information on the components of water pollution. The intelligent early warning module is used to analyze the evolution of pollution source diffusion based on water pollution component information and to make intelligent early warning decisions based on intelligent multi-mode network switching strategies, thereby constructing a water pollution source tracing and early warning strategy.

Images