Dangerous chemical leakage and diffusion prediction method and system based on unmanned ship, and storage medium

Abstract

The invention relates to the technical field of data processing, and discloses a hazardous chemical substance leakage and diffusion prediction method and system based on an unmanned ship, and a storage medium. The method comprises the following steps: acquiring water quality, weather and geographic data through an unmanned ship sensor, and processing to obtain an initial data set; calculating a diffusion range according to the data, optimizing a monitoring area, and generating a cooperative monitoring track; performing exception elimination and smoothing processing on the data, and constructing spatial distribution; solving a convection diffusion equation, and establishing a three-dimensional flow field model; analyzing a hazardous chemical substance diffusion path, and predicting pollutant distribution; and establishing a risk assessment system, and generating emergency disposal suggestions and decision-making schemes. According to the application, rapid and accurate monitoring of the polluted area is realized through the multi-parameter sensor array carried by the unmanned ship, and accurate prediction of the hazardous chemical substance pollutant diffusion process is realized in combination with multi-source data fusion processing and a three-dimensional flow field simulation technology.

Description

Technical Field

[0001] The present application relates to the field of data processing, and in particular to a method, system and storage medium for predicting the leakage and spread of hazardous chemicals based on an unmanned ship. Background Art

[0002] In the process of emergency response to hazardous chemical leakage accidents, timely understanding of the diffusion trend and distribution characteristics of pollutants in water areas is the key to formulating scientific disposal plans. At present, water pollution monitoring mainly relies on fixed monitoring stations and manual sampling. Fixed monitoring stations have high deployment costs and limited coverage, while manual sampling has problems such as low efficiency and poor timeliness. At the same time, traditional pollutant diffusion prediction methods are mainly based on empirical formulas or simplified models, which are difficult to accurately reflect the diffusion process of pollutants in complex water environments. Existing emergency decision support methods are often based on static data and empirical judgments, and lack the ability to analyze and predict the dynamic diffusion characteristics of pollutants in real time.

[0003] The existing technology has the following shortcomings: First, fixed monitoring and manual sampling are difficult to achieve rapid and comprehensive coverage of polluted areas, and it is impossible to obtain key data in the process of pollutant diffusion in a timely manner; second, the existing diffusion prediction method fails to fully consider the comprehensive impact of environmental factors such as hydrology and meteorology, and there is a large deviation between the prediction results and the actual situation; third, the traditional emergency decision-making method lacks effective integration of real-time monitoring data and prediction results, and it is difficult to provide scientific and reliable decision-making basis for emergency disposal. In addition, the existing technology also has problems such as single data processing method, low accuracy of prediction model, and insufficient intelligence of decision support. Summary of the invention

[0004] The present application provides a method, system and storage medium for predicting the leakage and diffusion of hazardous chemicals based on an unmanned ship, which is used to achieve rapid and accurate monitoring of contaminated areas through a multi-parameter sensor array carried by the unmanned ship, and to accurately predict the diffusion process of hazardous chemical pollutants by combining multi-source data fusion processing and three-dimensional flow field simulation technology.

[0005] In the first aspect, the present application provides a method for predicting the leakage and diffusion of hazardous chemicals based on an unmanned ship, and the method for predicting the leakage and diffusion of hazardous chemicals based on an unmanned ship includes: collecting water quality parameters, meteorological data and geographic information through a sensor array carried on the unmanned ship, and processing them through a data acquisition controller to obtain an initial monitoring data set; calculating the potential diffusion range based on the initial monitoring data set, optimizing the divided grid monitoring area, and obtaining a multi-point collaborative monitoring trajectory plan; according to the multi-point collaborative monitoring trajectory plan, the initial monitoring data set is subjected to outlier removal and data smoothing processing, and spatial distribution construction is performed to obtain a fused processing data set; based on the fused processing data set, a three-dimensional flow field model of the water area is obtained by constructing and solving a group of convection-diffusion equations; through the three-dimensional flow field model of the water area, the diffusion path of hazardous chemicals in the water area is analyzed to obtain the spatiotemporal distribution prediction results of hazardous chemical pollutants; based on the spatiotemporal distribution prediction results of hazardous chemical pollutants, a pollution risk level assessment system is established, and emergency disposal recommendations are generated to obtain a visual decision support plan.

[0006] In a second aspect, the present application provides a hazardous chemical leakage and diffusion prediction system based on an unmanned ship, and the hazardous chemical leakage and diffusion prediction system based on an unmanned ship includes:

[0007] The acquisition module is used to collect water quality parameters, meteorological data and geographic information through the sensor array carried on the unmanned ship, and the data acquisition controller processes them to obtain the initial monitoring data set;

[0008] The calculation module is used to calculate the potential diffusion range based on the initial monitoring data set, optimize the divided grid monitoring area, and obtain a multi-point collaborative monitoring trajectory plan;

[0009] A smoothing module is used to remove outliers and smooth the data on the initial monitoring data set according to the multi-point collaborative monitoring trajectory scheme, and to construct a spatial distribution to obtain a fused processing data set;

[0010] A construction module is used to construct and solve a group of convection-diffusion equations based on the fused processing data set to obtain a three-dimensional flow field model of the water area;

[0011] An analysis module is used to analyze the diffusion path of hazardous chemicals in the water area through the three-dimensional flow field model of the water area, and obtain the prediction results of the spatiotemporal distribution of hazardous chemical pollutants;

[0012] The generation module is used to establish a pollution risk level assessment system based on the spatiotemporal distribution prediction results of hazardous chemical pollutants, generate emergency disposal suggestions, and obtain a visual decision support plan.

[0013] A third aspect of the present application provides a computer-readable storage medium, which stores instructions. When the computer-readable storage medium is run on a computer, the computer executes the above-mentioned method for predicting the leakage and diffusion of hazardous chemicals based on an unmanned ship.

[0014] In the technical solution provided in the present application, multi-parameter synchronous monitoring is carried out by a sensor array carried on an unmanned ship, so as to realize the rapid and accurate collection of water quality parameters, meteorological data and geographic information, and overcome the problem of limited coverage of traditional fixed monitoring points; by processing and analyzing the initial monitoring data set, the accurate calculation of the potential diffusion range and the dynamic optimization of the grid monitoring area are realized, and the monitoring efficiency and data quality are improved; the multi-point collaborative monitoring trajectory scheme is adopted, combined with the outlier removal and data smoothing processing technology, to ensure the reliability and continuity of the monitoring data; by constructing and solving the convection-diffusion equation group, a three-dimensional flow field model that accurately reflects the movement characteristics of the water area is obtained, which provides a reliable basis for the prediction of pollutant diffusion; based on the three-dimensional flow field model of the water area, the diffusion path analysis is carried out to realize the accurate prediction of the spatiotemporal distribution of hazardous chemical pollutants, and effectively improve the accuracy of the prediction results; by establishing a pollution risk level assessment system and generating emergency disposal recommendations, an intuitive and efficient visual decision support scheme is formed, which provides a scientific basis for emergency command decision-making and improves the emergency disposal capability and efficiency of hazardous chemical leakage accidents in the water area. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying any creative work.

[0016] Figure 1 A schematic diagram of an embodiment of a method for predicting leakage and diffusion of hazardous chemicals based on an unmanned ship in an embodiment of the present application;

[0017] Figure 2 This is a schematic diagram of an embodiment of a hazardous chemical leakage and diffusion prediction system based on an unmanned ship in an embodiment of the present application. DETAILED DESCRIPTION

[0018] The embodiments of the present application provide a method, system and storage medium for predicting the leakage and diffusion of hazardous chemicals based on an unmanned ship. The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of this application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments described here can be implemented in an order other than that illustrated or described here. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0019] For ease of understanding, the specific process of the embodiment of the present application is described below. Figure 1 In the embodiment of the present application, an embodiment of a method for predicting the leakage and diffusion of hazardous chemicals based on an unmanned ship includes:

[0020] Step S101, collecting water quality parameters, meteorological data and geographic information through a sensor array carried on an unmanned ship, and processing them through a data acquisition controller to obtain an initial monitoring data set;

[0021] Step S102: Calculate the potential diffusion range according to the initial monitoring data set, optimize the divided grid monitoring area, and obtain a multi-point collaborative monitoring trajectory plan;

[0022] Step S103: According to the multi-point collaborative monitoring trajectory scheme, outliers are removed and data is smoothed on the initial monitoring data set, and spatial distribution is constructed to obtain a fused processing data set;

[0023] Step S104: Based on the fused data set, a three-dimensional flow field model of the water area is obtained by constructing and solving a group of convection-diffusion equations;

[0024] Step S105: using the three-dimensional flow field model of the water area, the diffusion path of hazardous chemicals in the water area is analyzed to obtain the spatiotemporal distribution prediction results of hazardous chemical pollutants;

[0025] Step S106: Based on the spatiotemporal distribution prediction results of hazardous chemical pollutants, a pollution risk level assessment system is established, and emergency disposal recommendations are generated to obtain a visual decision support plan.

[0026] It is understandable that the execution subject of the present application can be a hazardous chemical leakage and diffusion prediction system based on an unmanned ship, or a terminal or a server, which is not limited here. The present application embodiment is described by taking a server as the execution subject as an example.

[0027] Specifically, comprehensive data collection is carried out through the sensor array carried on the unmanned ship. The sensor array consists of a variety of professional sensing equipment, including a water quality monitoring sensor group, meteorological data collection equipment and a geographic information collection device. The water quality monitoring sensor group integrates pH sensors, dissolved oxygen sensors, conductivity sensors and turbidity sensors, which are responsible for collecting key parameters such as water pH, dissolved oxygen content, conductivity and turbidity. The meteorological data collection equipment includes anemometers, wind vanes and temperature and humidity sensors for real-time monitoring of wind speed, wind direction and environmental temperature and humidity information. The geographic information collection device is equipped with a high-precision GPS positioning system and a sonar depth sounder to obtain accurate geographic coordinates and water depth data.

[0028] The data acquisition controller performs preliminary processing on the collected raw data. First, it performs time synchronization marking to ensure that all data has a unified time reference. Then, the data is formatted in a unified manner, and the data collected by different sensors are converted into a standard format. Next, the dimension normalization is performed to unify the data of different physical quantities into the same numerical range. Finally, a data integrity check is performed to preliminarily screen for possible data missing or anomalies to form an initial monitoring data set.

[0029] Based on the initial monitoring data set obtained, the potential diffusion range is calculated and the grid monitoring area is optimized. First, the water quality parameters are gradient analyzed, and the main diffusion direction of pollutants is determined by calculating the spatial change rate of pollutant concentration. At the same time, the meteorological data is correlated with the geographic information to evaluate the impact of wind and water flow on pollutant diffusion. On this basis, the boundaries of the potential diffusion range are delineated, and the grid is divided according to the diffusion characteristics of pollutants.

[0030] In the grid optimization stage, the grid is encrypted in areas where the pollutant concentration changes dramatically, while the grid is sparse in areas where the change is gentle. This adaptive grid division strategy can improve the calculation efficiency while ensuring the calculation accuracy. For the optimized grid monitoring area, the sampling points are prioritized based on the pollutant concentration gradient and diffusion rate, and the optimal multi-point collaborative monitoring trajectory is generated in combination with the navigation constraints of the unmanned ship.

[0031] According to the multi-point collaborative monitoring trajectory plan, the initial monitoring data is further processed and analyzed. First, outliers are identified and eliminated. The mean and standard deviation of the data are calculated by statistical methods, and a reasonable confidence interval is set. Data outside the confidence interval is marked as an outlier. For the marked outliers, by comparing them with the data of adjacent sampling points, it is determined whether they are real anomalies, and whether they should be corrected or eliminated.

[0032] The data smoothing process uses the sliding window method to perform weighted average calculation on the data in each time window to eliminate random fluctuations in the data. The spatial distribution construction uses the Kriging interpolation method to generate the pollutant distribution field by spatially interpolating the data of discrete sampling points. The resulting fusion processing data set retains the key features of the original data and has good continuity and reliability.

[0033] After obtaining the fused data set, we began to construct and solve the convection-diffusion equations. First, we extracted the three-dimensional coordinates of the water boundary points, stratified the water space according to longitude, latitude and depth, and generated computational grid node data with spatial positioning information. Based on these node data, combined with the temperature, salinity and other physical quantity data in the water quality parameters, we determined the vertical stratification structure and constructed the hydrodynamic boundary conditions including the thermocline and density stratification.

[0034] The momentum equation of water body movement is spatially differentiated, and the continuity equation and momentum equation are converted into a set of algebraic equations. The velocity components in the horizontal and vertical directions are decomposed and calculated respectively, and the velocity field and pressure field are iteratively solved by the pressure correction method. At the same time, the initial flow field data are coupled with the wind speed and wind direction information in the meteorological data to calculate the water surface wind stress and internal shear stress. The obtained three-dimensional flow field model of the water area contains the velocity vector and flow direction information of the water body at different depths and different positions.

[0035] Based on the three-dimensional flow field model of the water area, the diffusion path analysis of hazardous chemicals was started. First, the velocity vector and flow direction data were extracted, and the diffusion coefficient of pollutants in the water body was calculated in combination with the physical and chemical characteristic parameters of hazardous chemicals. The diffusion rate of pollutants in the horizontal and vertical directions was decomposed and calculated, and the influence of the thermocline on vertical diffusion was considered. Combined with the water body density distribution data, the migration law of pollutants in different water layers was analyzed.

[0036] The pollutant migration process is divided into time steps, and the concentration gradient of pollutants is calculated in each time step, and the influence of degradation reaction kinetics is considered. By superimposing time series and considering the cumulative effects of water disturbance and mixing, the spatiotemporal distribution prediction results of hazardous chemical pollutants are generated.

[0037] Based on the prediction results of the spatiotemporal distribution of hazardous chemical pollutants, a risk assessment and emergency response system is established. Pollutant concentration data is extracted from the prediction results and graded and calibrated according to the toxicity threshold. The spatial distribution range and exposure time of each concentration interval are counted, and sensitive areas of the water environment are marked. A detailed analysis is conducted on high-risk areas, and their location coordinates and pollution characteristic parameters are extracted. Combined with the navigation conditions of the waters and the distribution of rescue resources, an emergency response zoning plan is formulated.

[0038] According to the characteristics of different regions, emergency disposal measures are determined according to the difficulty of pollutant control and disposal priority. The control effects of different disposal measures are evaluated through multi-scenario simulation calculations to form a scientific and reasonable list of emergency disposal suggestions. Finally, the distribution of risk areas and disposal suggestions are visualized to form an intuitive decision support plan, providing strong support for emergency command and disposal work.

[0039] For example, a hazardous chemical leak occurred at a chemical terminal, and the leaked substance was an organic solvent. The unmanned boat first collected data in the area around the leak point, and monitored that the water quality pH value was 5.8, the dissolved oxygen content was 3.2 mg / L, and the conductivity increased significantly. Meteorological data showed that the wind speed was 2.5 m / s, the wind direction was southeast, and the ambient temperature was 25 ° C. Through the analysis and processing of the collected data, it was determined that the pollutants mainly diffused to the southeast. According to the diffusion trend, a dense monitoring grid of 50m×50m was laid in the high-concentration area, and a sparse grid of 100m×100m was used in the peripheral area. Through data fusion processing, an accurate three-dimensional flow field distribution was constructed, and the prediction showed that the pollution group would reach the water intake 3 kilometers downstream in 2 hours. Based on this, the system divided the risk level areas into three levels and generated a zoning disposal plan including emergency enclosures, adsorption treatment and other measures.

[0040] In the embodiment of the present application, multi-parameter synchronous monitoring is carried out by a sensor array mounted on an unmanned vessel, so that rapid and accurate collection of water quality parameters, meteorological data and geographic information is achieved, overcoming the problem of limited coverage of traditional fixed monitoring points; by processing and analyzing the initial monitoring data set, accurate calculation of the potential diffusion range and dynamic optimization of the grid monitoring area are achieved, thereby improving monitoring efficiency and data quality; a multi-point collaborative monitoring trajectory scheme is adopted, combined with outlier removal and data smoothing processing technology, to ensure the reliability and continuity of the monitoring data; by constructing and solving a group of convection-diffusion equations, a three-dimensional flow field model that accurately reflects the movement characteristics of the water area is obtained, providing a reliable basis for pollutant diffusion prediction; diffusion path analysis is performed based on the three-dimensional flow field model of the water area, so as to achieve accurate prediction of the spatiotemporal distribution of hazardous chemical pollutants, effectively improving the accuracy of the prediction results; by establishing a pollution risk level assessment system and generating emergency disposal recommendations, an intuitive and efficient visual decision support scheme is formed, which provides a scientific basis for emergency command decision-making and improves the emergency disposal capability and efficiency of hazardous chemical leakage accidents in the water area.

[0041] In a specific embodiment, the process of executing step S101 may specifically include the following steps:

[0042] (1) Scan and sample the water area through the pH sensor, dissolved oxygen sensor, conductivity sensor and turbidity sensor in the sensor array to obtain water quality parameter data;

[0043] (2) The meteorological environment above the water area is sampled through the anemometer, wind vane, and temperature and humidity sensors in the sensor array to obtain meteorological data;

[0044] (3) Use the GPS positioning device and sonar depth sounder in the sensor array to locate and measure the water area to obtain geographic information;

[0045] (4) Synchronize the water quality parameter data, meteorological data and geographic information according to the sampling time to obtain the original data with time stamp;

[0046] (5) The original data with time stamps are classified and sorted according to water quality parameter indicators, meteorological parameter indicators and geographical parameter indicators, and each indicator is linearly normalized and dimensionally unified. The processed indicators are then merged and cross-validated to obtain standardized data;

[0047] (6) The initial monitoring data set is obtained by performing a data integrity scan on the standardized data, identifying missing data points and outliers, interpolating the missing points with adjacent values, replacing the outliers with medians, and then extracting time series features and checking data consistency on the processed data.

[0048] Specifically, multiple sensors are used to collect data in all directions. The water quality monitoring equipment in the sensor array includes pH sensors, dissolved oxygen sensors, conductivity sensors and turbidity sensors. These sensors use a scanning sampling strategy to continuously monitor the water area along the preset sampling path. The pH sensor reflects the acidity and alkalinity of the water body by measuring the hydrogen ion concentration, the dissolved oxygen sensor determines the dissolved oxygen content in the water, the conductivity sensor detects changes in the conductivity of the water body, and the turbidity sensor monitors the content of suspended matter in the water body. These sensors collect data at a fixed sampling frequency (usually once per second) to form a water quality parameter data stream. In terms of meteorological environment monitoring above the water area, the anemometer uses the ultrasonic principle to measure the air flow speed and has the characteristics of fast response. The wind vane can accurately indicate the wind direction through aerodynamic design. The temperature and humidity sensor uses a resistive temperature measuring element and a capacitive humidity sensor to monitor the air temperature and relative humidity respectively. The sampling frequency of these meteorological monitoring equipment is usually set to once every 10 seconds to capture the dynamic changes in meteorological conditions.

[0049] Geographic information collection adopts dual means. The GPS positioning device provides centimeter-level plane positioning accuracy and achieves precise positioning by receiving multiple satellite signals. The sonar depth sounder uses the propagation characteristics of sound waves in water to measure water depth, emits ultrasonic waves and receives echo signals, and calculates the water depth value based on the round-trip time of the sound waves. The sampling frequency of GPS data and water depth data is set to 5 times per second to ensure the continuous recording of position and water depth information during navigation. The data synchronization marking link uses a unified time base to align data streams from different sources according to the sampling time. The specific operation is to add a millisecond timestamp to each data record, including the sampling year, month, day, hour, minute, second and millisecond information. This synchronization marking mechanism ensures the time correspondence between different parameters and lays the foundation for subsequent data fusion.

[0050] Data standardization is first classified and sorted according to parameter types, and water quality parameters (pH, dissolved oxygen, conductivity, turbidity), meteorological parameters (wind speed, wind direction, temperature, humidity) and geographical parameters (longitude, latitude, water depth) are classified separately. Then linear normalization is performed to uniformly map parameters of different dimensions to the [0,1] interval. Taking water quality parameters as an example, the original range of pH value is 0-14, which is mapped to the standard interval through linear transformation; the unit of dissolved oxygen concentration is mg / L, which is also normalized; conductivity and turbidity data also undergo a similar standardization process. In meteorological data, the unit of wind speed is m / s, the unit of temperature is ℃, and the relative humidity is originally in percentage form. These data have the same numerical range after normalization. Geographic data standardizes the longitude and latitude coordinates and water depth values ​​separately.

[0051] In the data merging and cross-validation phase, the standardized parameters are reorganized according to the timestamp to form a multidimensional data matrix. Each row represents a complete observation record at a time point, including all monitoring parameters. The cross-validation process checks the rationality of the data through correlation analysis between parameters. For example, the relationship between wind speed and water surface fluctuations, the relationship between temperature and dissolved oxygen, etc. Through this physical correlation verification, potential data anomalies can be discovered. In the data integrity processing phase, the data matrix is ​​first scanned to identify missing data points and outliers. Missing data is usually caused by temporary sensor failure or communication interruption. For such missing points, the adjacent value interpolation method is used to fill them. Specifically, if the time interval before and after the missing point is small (such as less than 10 seconds), the linear interpolation method is used; if the interval is large, the distance weighted average method is used to consider the data of multiple time points before and after. Outlier identification is based on statistical characteristics, the mean and standard deviation of each parameter are calculated, and data points that deviate from the mean by more than three times the standard deviation are marked as abnormal. For the marked outliers, the median replacement method is used for processing, and the median selection range is the data within the time window (usually 1 minute) before and after the outlier point.

[0052] In the time series feature extraction phase, the processed data is subjected to sliding window analysis, with the window size set to 5 minutes and the step length set to 1 minute. Statistical features are calculated in each window, including mean, variance, maximum value, minimum value, etc. These time series features reflect the dynamic change characteristics of various parameters. Data consistency verification ensures the continuity and rationality of the data by comparing the feature values ​​of adjacent time windows.

[0053] For example, after a hazardous chemical leak occurred at a chemical terminal, an unmanned boat carried a sensor array to monitor the accident waters. Water quality monitoring showed a sharp drop in pH near the leak point, while conductivity increased significantly. This trend was clearly reflected in the data record through time synchronization markers. Meteorological data recorded southeast winds, and the wind speed fluctuated during the monitoring period. GPS positioning and sonar sounding data showed that the water depth in the monitoring area varied between 5 and 15 meters. During the data standardization process, the pH value range of 6-8 was mapped to the standard range of 0-1, and the conductivity was converted from μS / cm units to dimensionless values. After the data was merged, it was found that the dissolved oxygen data at individual time points were missing, and they were supplemented by the weighted average of nearby time points. Outlier processing found several turbidity readings that were significantly deviated, and replaced them with the median in the corresponding time window.

[0054] In a specific embodiment, the process of executing step S102 may specifically include the following steps:

[0055] (1) Perform gradient analysis on the water quality parameters in the initial monitoring data set, extract the trend of pollutant concentration changes, and obtain preliminary diffusion direction determination results;

[0056] (2) Conduct correlation analysis between meteorological data and geographic information in the initial monitoring data set to obtain water environment impact factors;

[0057] (3) Based on the preliminary diffusion direction determination results and water environment impact factors, the potential diffusion range is delineated, and the delineated range is gridded according to the pollutant diffusion characteristics to obtain the initial grid monitoring area;

[0058] (4) Optimize the density distribution of the initial grid monitoring area, increase the density of the grid in areas where pollutant concentrations change dramatically, and reduce the density of the grid in areas where the concentrations change slowly, to obtain the optimized grid monitoring area;

[0059] (5) Prioritize the sampling points in the optimized grid monitoring area, assign a weight value to each sampling point based on the pollutant concentration gradient and diffusion rate, and combine the navigation constraints of the unmanned ship to obtain a weighted sampling sequence;

[0060] (6) Perform spatiotemporal continuity analysis and path smoothing on the weighted sampling sequence, perform curve fitting and steering angle optimization on the navigation path, and fine-tune the path considering obstacle avoidance requirements to obtain a multi-point collaborative monitoring trajectory plan.

[0061] Specifically, the water quality parameters in the initial monitoring data set are subjected to gradient analysis, and the concentration gradient between each monitoring point is calculated using the central difference method. For each monitoring point, the concentration difference of the surrounding adjacent points is calculated to form a concentration gradient vector. By synthesizing these vectors, the main diffusion direction of pollutants on the plane is obtained. At the same time, the concentration distribution in the vertical direction is analyzed to determine the vertical migration trend of pollutants in the water body.

[0062] The determination of the factors affecting the water environment requires the correlation analysis of meteorological data and geographic information. The wind field data is used to generate a continuous wind field distribution through spatial interpolation, and the formation of wind-driven currents is analyzed in combination with the terrain characteristics of the water area. The water depth data in the geographic information is used to calculate the vertical mixing intensity, and the shoreline information is used to determine the boundary constraints. Through multi-factor coupling analysis, the correlation between the hydrodynamic field and pollutant transport is constructed.

[0063] The mathematical model for defining the boundaries of the potential diffusion range is expressed as follows:

[0064]

[0065] in: represents the diffusion boundary decision function; is the concentration gradient weight coefficient; is the pollutant concentration gradient; is the wind field impact weight coefficient; is the wind field velocity vector; is the water depth influence weight coefficient; is the water depth distribution function.

[0066] When the judgment function When it is greater than the threshold, the point is included in the diffusion range. After the boundary is determined, the grid is divided according to the diffusion characteristics of the pollutants, and a denser grid is used in areas with large concentration gradients.

[0067] Adaptive grid technology is used to optimize the grid density distribution. In areas where pollutant concentrations change dramatically, the grid size is dynamically adjusted according to the size of the concentration gradient. For areas with gentle changes, a relatively sparse grid layout is used to balance calculation accuracy and efficiency. The optimized grid structure fully considers the spatial characteristics of pollutant diffusion. The priority ranking of sampling points is determined based on a combination of multiple factors. First, the pollutant concentration gradient of each grid point is calculated. The points with larger gradients have higher sampling priorities. At the same time, the diffusion rate is considered, and the sampling points are encrypted in areas where pollutants diffuse rapidly. It is also necessary to combine the navigation constraints of the unmanned ship, including factors such as maximum speed and turning radius, to ensure the feasibility of the sampling path.

[0068] Finally, the temporal and spatial continuity of the sampling sequence is optimized. The cubic spline interpolation method is used to smooth the navigation path, reduce sharp turns, and make the path more consistent with the motion characteristics of the unmanned ship. At the same time, the distribution of obstacles in the water area is considered, and necessary fine-tuning of the path is performed to ensure navigation safety.

[0069] For example, after a hazardous chemical leak occurred at a chemical terminal, the initial water quality data was analyzed and it was found that the pH value decreased radially at the leak point, and the preliminary diffusion direction was obtained by calculating the spatial gradient of the pH value. The wind speed at the site was 3m / s, southeast wind, and the analysis of the water depth and terrain data showed that the pollutants mainly diffused to the southeast. Within the determined diffusion range, the monitoring area is divided into three levels of grids: a dense grid of 5m×5m is used within 50 meters around the leak point, a medium-density grid of 10m×10m is used within 50-200 meters, and a sparse grid of 20m×20m is used in the peripheral area. According to the diffusion rate and concentration gradient of pollutants, the sampling points are prioritized and the sampling path is generated. Such a sampling strategy not only ensures the monitoring accuracy of key areas, but also takes into account the monitoring efficiency.

[0070] In terms of data processing details, the calculation of pollutant concentration gradients uses the fourth-order central difference format, which has higher accuracy than simple first-order differences. For any point (i, j) in the monitoring area, the concentration gradient calculation involves the data of the surrounding 16 grid points. The gradients in the horizontal and vertical directions are calculated separately and then synthesized into a total gradient vector. This high-order difference method can more accurately capture the spatial variation characteristics of pollutant concentrations. In the process of constructing the water environment impact factor, the wind field data is first interpolated by Kriging to generate a continuous distribution. The influence of terrain is taken into account during interpolation, and virtual observation points are added in complex terrain areas to improve the interpolation accuracy. The water depth data is also processed by spatial interpolation to generate a continuous water depth distribution field. These environmental factors comprehensively affect the diffusion process of pollutants through weight superposition.

[0071] The grid optimization process uses a recursive quadtree decomposition method. Starting from the initial uniform grid, the areas that need to be encrypted are subdivided step by step according to the concentration gradient and hydrodynamic characteristics. The criteria for grid encryption include concentration gradient threshold, velocity change rate, and water depth change rate. The size ratio of adjacent grids does not exceed 2:1 to ensure the stability of the calculation. A multi-criteria scoring mechanism is used to determine the priority of sampling points. The score of each grid point consists of the following components: concentration gradient contribution, diffusion rate contribution, location importance (distance from sensitive areas), and sampling operation difficulty. These factors are comprehensively scored through a weight system, and the final score determines the sampling priority. The improved The algorithm performs global path planning and then uses cubic B-spline curves to smooth the path. The smoothing process must ensure the continuity of the path while satisfying the motion constraints of the unmanned ship. For the obstacle avoidance requirements during navigation, a dynamic safety distance is set, which is dynamically adjusted with the navigation speed and water conditions.

[0072] For example, when a leakage accident occurred in a chemical storage tank area, the pH gradient analysis found that the pollution group expanded in an elliptical shape, and the long axis direction was basically consistent with the wind direction. The environmental impact factor analysis showed that the surface flow velocity generated by the wind field was about 3% of the wind speed, and considering the 5-15 meter water depth change on site, the pollutants showed obvious stratification in the vertical direction. According to these characteristics, the monitoring area is divided into three nested elliptical areas, and the inner area adopts the densest monitoring grid, and the grid side length increases with the distance from the leakage source. Each sampling point is assigned a comprehensive weight score, which reflects the importance of monitoring at the point. The weight calculation takes into account factors such as the pH gradient of the point, the distance from sensitive waters, and the water flow rate. The generated sampling path shows the characteristics of spiral advancement, starting from the leakage source and expanding outward, ensuring the priority monitoring of high-concentration areas. The path turning points are all arc transition processed, and the minimum arc radius is not less than the minimum turning radius of the unmanned ship.

[0073] In a specific embodiment, the process of executing step S103 may specifically include the following steps:

[0074] (1) The initial monitoring data set is divided into time series based on the sampling point location information in the multi-point collaborative monitoring trajectory scheme to form a data sequence corresponding to the spatial location;

[0075] (2) Based on the mean, standard deviation and coefficient of variation, the statistical characteristics of the data sequence are quantified, outliers exceeding the threshold are identified based on the confidence interval, and outlier labeling data is generated;

[0076] (3) Through time and space correlation analysis, the outliers in the outlier marked data are compared with the adjacent sampling values ​​to correct and eliminate the outliers and obtain the cleaned data sequence;

[0077] (4) Establish a sliding window on the cleaned data sequence, perform weighted average calculation on the data in the window, verify the data continuity, and output the smoothed data;

[0078] (5) According to the spatial coordinates of the sampling points, the smoothed data is distributed mapped and filled with interval interpolation to construct preliminary spatial distribution data;

[0079] (6) Through spatial autocorrelation analysis and boundary constraint processing, the consistency of the preliminary spatial distribution data is corrected and optimized to obtain a fused processed data set.

[0080] Specifically, the initial monitoring data set is segmented in time series according to the sampling point location information in the multi-point collaborative monitoring trajectory scheme. The process of time series segmentation is to reorganize the continuously collected data according to the spatial position, and each sampling point position corresponds to a time series. This sequence contains the changes of all monitoring parameters at the point position over time, such as water quality parameters such as pH, dissolved oxygen, conductivity, turbidity, and time series of meteorological parameters such as wind speed and wind direction. For the data sequence at each spatial position, statistical methods are used to quantify the features. The mean calculation reflects the central tendency of the data, the standard deviation represents the degree of dispersion of the data, and the coefficient of variation measures the relative fluctuation of the data by the ratio of the standard deviation to the mean. Based on these statistical features, a confidence interval is established to identify outliers. The confidence interval is usually set to a range of three standard deviations above and below the mean, and data points outside this range are marked as potential outliers. This method can effectively identify abnormal data caused by sensor failure or external interference.

[0081] For each marked outlier point, the correlation coefficient with its temporal neighbor and spatial neighbor is calculated. Temporal correlation analysis focuses on the continuity of data and determines whether the outlier conforms to the law of parameter change. Spatial correlation analysis considers the physical characteristics of pollutant diffusion and checks whether the outlier violates the law of diffusion. Through this dual correlation analysis, the validity of the outlier can be accurately judged and whether it should be corrected or eliminated.

[0082] The data smoothing process uses sliding window technology. The size of the sliding window is determined by the data sampling frequency, and the typical window size is 5-10 sampling points. In each window, the data is weighted averaged, and the weight coefficient decreases as the distance from the center point increases. This processing method not only retains the main characteristics of the data, but also effectively eliminates the influence of random fluctuations. At the same time, the continuity of the data is verified by comparing the rate of change of the average values ​​of adjacent windows. The construction of spatial distribution data is achieved by spatial mapping and interpolation of the smoothed data. According to the spatial coordinates of the sampling points, the monitoring data is mapped into three-dimensional space. Since there are blank areas between the sampling points, the data in these areas need to be filled by interpolation methods. The interpolation process uses the Kriging method, which takes into account spatial autocorrelation and can give the optimal unbiased estimate.

[0083] The final data consistency correction is completed through spatial autocorrelation analysis and boundary constraint processing. Spatial autocorrelation analysis is based on the calculation of the Moran index and is used to evaluate the spatial distribution pattern of the data. For the data in the boundary area, it is necessary to consider the actual physical boundary conditions, such as the shoreline, water depth changes and other constraints for correction. Through these processes, the final fusion processing data set meets the continuity requirements of the data and conforms to the physical laws of pollutant diffusion.

[0084] For example, a leak of hazardous chemicals in a chemical enterprise is used as an example for detailed description. After the accident, the unmanned boat continuously monitored the leak area according to the predetermined monitoring trajectory. For each monitoring point, a time series data is recorded. Taking pH monitoring as an example, suppose a fixed position point records 120 data points during continuous monitoring, with a sampling interval of 30 seconds. Through statistical analysis, it is found that the mean pH value of this point is 6.5, the standard deviation is 0.3, and the coefficient of variation is 0.046. According to these statistical characteristics, the confidence interval is set to [5.6, 7.4]. Three data points are found to be beyond this range in the data sequence, of which the pH values ​​of two points suddenly rise to 8.2, and one point drops to 5.0. The spatiotemporal correlation analysis of these outliers shows that the two points with a pH value of 8.2 appear at two adjacent moments, but compared with the data of other surrounding monitoring points, they obviously deviate from the diffusion trend, and the rate of change far exceeds the physical diffusion rate of pollutants. Therefore, they are judged as outliers and are eliminated. Although the point with a pH value of 5.0 deviates from the mean, it is consistent with the diffusion direction of the pollutants and is consistent with the change trend of the adjacent points, so this data point is retained. In the sliding window processing stage, 7 data points are selected as the window size. Different weights are assigned to the data points in each window, with the maximum weight of the center point being 1.0, and the weights on both sides decreasing to 0.8, 0.6, and 0.4 respectively. Through this weighted average processing, the high-frequency fluctuations in the data are effectively eliminated, and a smoother change trend is obtained. In the process of constructing the spatial distribution, the processed data of all monitoring points are mapped to the three-dimensional space according to their GPS coordinates. Since the spacing between the monitoring points is 50 meters, the middle area needs to be interpolated and filled. Using the Kriging interpolation method, the directionality of pollutant diffusion is taken into account, and a larger correlation distance is used in the main diffusion direction, and a smaller correlation distance is used perpendicular to the diffusion direction, so as to obtain a more realistic spatial distribution feature. Through spatial autocorrelation analysis, it is found that some data distributions that do not conform to physical laws appear in areas close to the coastline. Through boundary constraint processing, the data in these areas are corrected considering the blocking effect of the coastline.

[0085] In a specific embodiment, the process of executing step S104 may specifically include the following steps:

[0086] (1) Extract the three-dimensional coordinates of the water boundary points from the fused data set, stratify the extracted coordinate points according to longitude, latitude and depth, and then generate computational grid node data with spatial location information through rasterization;

[0087] (2) Based on the computational grid node data, the temperature, salinity, and density of the water quality parameters are calibrated in layers according to the depth. The vertical layered structure is determined in combination with the water depth gradient, and the hydrodynamic boundary condition parameters including the thermocline and density layering are constructed.

[0088] (3) Based on the parameters of the hydrodynamic boundary conditions, the momentum equation of water motion is spatially differentiated, the continuity equation and momentum equation are transformed into a set of algebraic equations, and a flow field calculation matrix describing the water flow velocity and pressure distribution is established;

[0089] (4) For the flow field calculation matrix, the velocity components in the horizontal and vertical directions are decomposed and calculated respectively. The velocity field and pressure field are iteratively solved by the pressure correction method to obtain the initial flow field data that characterizes the three-dimensional motion characteristics of the water body.

[0090] (5) The initial flow field data is coupled with the wind speed and direction information in the meteorological data to calculate the water surface wind stress and internal shear stress, taking into account the effects of Coriolis force and gravity, and outputting multi-level flow field data including the influence of external forces;

[0091] (6) The multi-layer flow field data are processed near the wall boundary layer and the turbulence intensity is calculated. The interaction between the layers is corrected by the vertical mixing coefficient to obtain a three-dimensional flow field model of the water area. The three-dimensional flow field model of the water area contains the velocity vector and flow direction information of the water body at different depths and different positions, which is used to characterize the three-dimensional motion state of any point in the water area.

[0092] Specifically, the three-dimensional coordinate information of the water boundary points, including longitude, latitude and depth data, is first extracted from the fusion processing data set. The longitude and latitude data are obtained through GPS, and the depth data is provided by the sonar depth sounder. These boundary points are divided into multiple water layers in the vertical direction. The thickness of each layer is determined according to the change of water depth. Usually, a thinner layer thickness is used in shallow water areas and a thicker layer thickness is used in deep water areas. Through this layering strategy, three-dimensional space grid nodes are generated, and each node has accurate spatial position information. After determining the spatial structure of the calculation grid, the construction of hydrodynamic boundary conditions begins. The temperature, salinity and density data in the water quality parameters are calibrated according to the depth. The changes of these parameters with depth reflect the stratification characteristics of the water body. The thermocline refers to the area where the temperature changes sharply with depth, usually between the surface and the deep layer. The density stratification is formed by the combined action of temperature and salinity, which directly affects the vertical diffusion of pollutants. By analyzing the vertical distribution characteristics of these parameters, the key stratification interface is determined and the hydrodynamic boundary conditions are established.

[0093] The momentum equation for water motion is described by the three-dimensional incompressible Navier-Stokes equation:

[0094]

[0095] in: represents the fluid velocity vector; For time; is the fluid density; For pressure; is the kinematic viscosity coefficient; For the external force term.

[0096] After discretization, the continuity equation and momentum equation are transformed into a set of algebraic equations in the following form:

[0097]

[0098] in: is the adjacent node coefficient; is the variable value of the adjacent node; is the central node coefficient; is the variable value of the central node; is the source term; It means to sum all the nodes adjacent to the center point P.

[0099] The iterative solution of the pressure field and velocity field adopts the SIMPLE algorithm, and its core equation is:

[0100]

[0101]

[0102] in: is the new pressure value after correction; is the pressure value of the current iteration step; is the pressure correction; is the relaxation factor, which is used to control the amplitude of pressure correction and ensure calculation stability; is the new speed value after correction; is the velocity value of the current iteration step; is the speed correction; is the relaxation factor for velocity.

[0103] After obtaining the initial flow field data, the influence of external environmental factors is considered. The coupling of wind field data is an important link. The wind generates shear stress on the water surface. This stress is transmitted downward through molecular viscosity, affecting the movement of surface water. The wind stress on the water surface changes with the square of the wind speed, and its direction is consistent with the wind direction. At the same time, due to the influence of the earth's rotation, the effect of the Coriolis force needs to be considered. The Coriolis force is related to the speed and latitude of the water body and has a significant impact on the movement of large-scale water bodies. The gravity effect is mainly reflected in the density stratification effect, which causes the retardation of vertical movement. The near-wall boundary layer treatment adopts the wall function method, and a special calculation grid is set in the area close to the solid boundary to capture the rapid change of speed. The calculation of turbulence intensity adopts the k-ε two-equation model to solve the turbulent kinetic energy and dissipation rate respectively. The correction of the vertical mixing coefficient takes into account the inhibitory effect of the thermocline, reduces the mixing intensity in the thermocline area, and more accurately describes the vertical diffusion process of pollutants. The three-dimensional flow field model of the water area is a numerical expression system that contains the velocity vector and flow direction information of each point in space. This model can reflect the movement of water at different depths and locations, and provide basic data for pollutant diffusion prediction. The accuracy of the flow field model directly affects the accuracy of subsequent diffusion predictions, so in the process of model construction, the calculation accuracy needs to be strictly controlled at every stage.

[0104] For example, after a hazardous chemical leak occurred in a chemical park at the estuary, data collected by unmanned boats showed that the water depth in the monitoring area ranged from 3 to 15 meters. When constructing the computational grid, the water body is divided into 10 layers vertically, with the surface layer thickness set to 0.5 meters to accurately describe the surface flow, and the deep layer thickness is 1 to 2 meters. The water temperature data shows that there is an obvious thermocline at a depth of 4 to 6 meters, and the temperature drops by 3 degrees within 0.5 meters, resulting in a sudden change in density. This jump layer significantly affects the vertical transport of pollutants. When solving the flow field, the time step is set to 60 seconds, and the implicit iteration format is used for calculation. During the pressure correction process, the relaxation factor is set to 0.3 to ensure the stability of the calculation. The initial flow field calculation results show that the maximum surface velocity appears above the thermocline, and the velocity gradually decreases with increasing depth. The northeast wind field data observed on site is brought into the calculation, with a wind speed of 5 meters per second. The surface shear stress generated causes the surface water to transport in the wind direction. After considering the Coriolis force, the surface flow direction is deflected by 15 degrees from the wind direction. The near-wall area is processed by standard wall function, and the dimensionless distance of the first grid point from the solid boundary is kept between 30 and 100. Turbulence calculations show that the turbulence intensity changes significantly near the thermocline, and the turbulent mixing below the thermocline is suppressed. Through numerical simulation, the three-dimensional flow field distribution is obtained, and this flow field model accurately reflects the movement characteristics of the water body.

[0105] In a specific embodiment, the process of executing step S105 may specifically include the following steps:

[0106] (1) Extract the velocity vector and flow direction data from the three-dimensional flow field model of the water area, combine the physical and chemical characteristic parameters of hazardous chemicals, calculate the diffusion coefficient of pollutants in the water body, and obtain the initial diffusion parameters;

[0107] (2) Based on the initial diffusion parameters, the diffusion rates of pollutants in the horizontal and vertical directions are decomposed and calculated, and the influence of the thermocline on vertical diffusion is considered to obtain the layered diffusion rate data;

[0108] (3) Coupling the layered diffusion rate data with the water density distribution data, analyzing the migration patterns of pollutants in different water layers, and obtaining multi-level pollutant migration data;

[0109] (4) Divide the multi-level pollutant transport data into time steps, calculate the pollutant concentration gradient change in each time step, and consider the impact of degradation reaction kinetics to obtain instantaneous concentration distribution data;

[0110] (5) Superimpose the instantaneous concentration distribution data in time series, consider the cumulative effects of water disturbance and mixing, and obtain the pollutant distribution data in a continuous time series;

[0111] (6) The pollutant distribution data under the continuous time series is projected in time and space, and the concentration field is numerically smoothed to obtain the prediction results of the spatiotemporal distribution of hazardous chemical pollutants.

[0112] Specifically, velocity vectors and flow direction data are extracted from the three-dimensional flow field model of the water area. These data contain information about the movement of water bodies at different locations and depths. The diffusion coefficient of pollutants in water bodies is calculated by combining the physical and chemical characteristic parameters of hazardous chemicals, including molecular diffusion coefficient, solubility, volatility, etc. It is particularly important to note that the diffusion coefficient is not a fixed value, but a dynamic parameter that changes with environmental factors such as temperature and salinity.

[0113] For the decomposition of diffusion rate in horizontal and vertical directions, the following mathematical model is used:

[0114]

[0115]

[0116] in: is the horizontal diffusion flux; is the vertical diffusion flux; is the horizontal response coefficient;

[0117] is the vertical response coefficient; is the horizontal transport flux; is the vertical transport flux; is the horizontal diffusion coefficient; is the vertical diffusion coefficient.

[0118] The coupling calculation of stratified diffusion rate and water density distribution uses the following equation:

[0119]

[0120] in: is the pollutant concentration; is the diffusion weight tensor; is the density-dependent vertical mixing function; is the water density.

[0121] The space-time projection and numerical smoothing process adopt the following formula:

[0122]

[0123] in: is the smoothed concentration field; is the weight coefficient; is the Gaussian smoothing kernel function; is the spatial position vector; is the sampling point position vector; is the original concentration time series.

[0124] Specifically, determine the division of time steps. The selection of time steps needs to meet the requirements of computational stability and accuracy at the same time. An adaptive time step strategy is usually adopted, using a smaller time step during periods of drastic changes in pollutant concentrations and a larger time step during periods of gentle changes. In each time step, the spatial gradient of pollutant concentrations is calculated, while considering the influence of chemical degradation reactions. Chemical degradation includes multiple processes such as hydrolysis, photolysis and biodegradation. The reaction rate constants of these processes need to be corrected according to environmental conditions such as water temperature and pH value on site. The disturbance and mixing of water bodies mainly include turbulent mixing, wave action and temperature convection. The intensity of turbulent mixing varies with the change of water depth. It is greatly affected by wind and waves in the surface layer and gradually weakens with increasing depth. Wave action mainly affects the mixing process of surface water bodies, and its intensity is related to wind speed and wind zone length. Temperature convection has a significant impact near the thermocline, affecting the vertical transport of pollutants.

[0125] The prediction of spatiotemporal distribution needs to take into account the combined effects of all these physical and chemical processes. Numerical smoothing is used to eliminate false oscillations generated during the calculation process to ensure the physical rationality of the prediction results. The following example illustrates the prediction process:

[0126] After a hazardous chemical leak occurred at a chemical terminal, monitoring data showed that the leaked substance was an organic solvent, and the molecular diffusion coefficient of the substance increased with increasing temperature. Data extracted from the three-dimensional flow field model showed that the maximum flow velocity on the surface was 0.5 m / s, which gradually decreased with increasing depth. The initial diffusion parameters calculated based on these data show that the diffusion coefficient in the horizontal direction is one order of magnitude larger than the vertical diffusion coefficient. When calculating the layered diffusion rate, it was found that there was a significant thermocline at a depth of 4-6 meters, which led to a significant inhibition of vertical diffusion. The water body density distribution data showed that the positions of the density jump layer and the thermocline basically coincided, further strengthening this inhibitory effect. Through coupled calculations, it was found that the diffusion rate of pollutants above the thermocline was much greater than that below the jump layer. The time step division adopted an adaptive strategy, using a small time step of 60 seconds at the beginning of the leak, and gradually increasing to 300 seconds as the diffusion process stabilized. In each time step, the degradation loss of pollutants was calculated, and it was found that the degradation rate in the surface layer was significantly higher than that in the deep layer due to the effect of light. The time series superposition shows that the pollution group expands in an elliptical shape in the horizontal direction, and the main axis direction is consistent with the mainstream direction. The prediction results after numerical smoothing show that within 2 hours after the leak, the high concentration area is mainly concentrated within 5 meters of the surface layer, and gradually spreads to the deep layer over time. This layered structure feature is in good agreement with the on-site monitoring data, verifying the reliability of the prediction method.

[0127] In a specific embodiment, the process of executing step S106 may specifically include the following steps:

[0128] (1) Extract pollutant concentration data from the spatiotemporal distribution prediction results of hazardous chemicals pollutants, calibrate them according to toxicity thresholds, and obtain initial risk classification data;

[0129] (2) Based on the initial risk classification data, the spatial distribution range and exposure time of each concentration range are counted, the sensitive areas of the water environment are marked, and a risk area distribution map is obtained;

[0130] (3) For high-risk areas in the risk area distribution map, extract their location coordinates and pollution characteristic parameters, and combine the navigation conditions of the waters and the distribution of rescue resources to obtain emergency response zoning data;

[0131] (4) Sort the emergency response zoning data according to the difficulty of pollutant control and disposal priority, determine the emergency disposal measures for different areas, and generate zoning disposal plans;

[0132] (5) Conduct multi-scenario simulation calculations on the zoning disposal plan, evaluate the control effects of different disposal measures, and output a list of emergency disposal recommendations;

[0133] (6) Integrate the risk area distribution map and the emergency response recommendation list, perform data visualization through the geographic information display platform, and obtain a visual decision support plan.

[0134] Specifically, firstly, the pollutant concentration data are extracted from the prediction results of the spatiotemporal distribution of hazardous chemicals pollutants. These data contain the concentration values ​​of pollutants at different times and different spatial locations. According to the toxicological characteristics of hazardous chemicals, key concentration values ​​such as acute toxicity threshold, chronic toxicity threshold and lethal concentration are determined, and the prediction results are graded and calibrated according to these thresholds. The risk level is usually divided into four levels: red represents extremely high risk areas (concentration exceeds the lethal concentration), orange represents high risk areas (concentration exceeds the acute toxicity threshold), yellow represents medium risk areas (concentration exceeds the chronic toxicity threshold), and green represents low risk areas (concentration is lower than the chronic toxicity threshold). For in-depth analysis of the initial risk classification data, it is necessary to count the spatial distribution range and exposure time of each concentration interval. The spatial distribution range is quantified by calculating the area and volume of areas with different risk levels, and the exposure time is determined based on the duration of the pollutant concentration exceeding the corresponding threshold. Water environmental sensitive areas include important water functional areas such as water intakes, aquaculture areas, and nature reserves. These areas need to be specially marked in the risk area distribution map. Through overlay analysis, the spatial overlap relationship between environmental sensitive areas and areas of different risk levels is determined.

[0135] For high-risk areas in the risk area distribution map, the spatial range, center coordinates, maximum concentration value, duration and other characteristic parameters are extracted in detail. At the same time, the navigation conditions of the waters are considered, including factors such as channel distribution, navigation density, water depth conditions, and the distribution of rescue resources, including the location and quantity of emergency ships, oil booms, adsorption materials and other equipment. The comprehensive analysis of this information forms emergency response zoning data, which provides a basis for the formulation of emergency disposal plans.

[0136] The processing of emergency response zoning data adopts a multi-criteria evaluation method, which comprehensively considers the two dimensions of pollutant control difficulty and disposal priority. The difficulty of pollutant control is determined by multiple factors, including pollutant concentration, hydrodynamic conditions, water depth conditions, etc. The disposal priority is determined based on environmental sensitivity, threat level and rescue feasibility. By constructing a scoring matrix, each area is comprehensively scored and ranked according to the score. According to the ranking results, corresponding emergency disposal measures are formulated for different areas to form a zoning disposal plan.

[0137] The multi-scenario simulation calculation of the zoning disposal plan takes into account different combinations of emergency disposal measures. Disposal measures include physical methods (enclosure, salvage), chemical methods (neutralization, flocculation) and biological methods (microbial degradation). For each combination of measures, a corresponding mathematical model is constructed to predict its effect on the diffusion and degradation of pollutants. The simulation calculation needs to consider the constraints of on-site conditions, including weather conditions, hydrological conditions, equipment performance and other factors. By comparing the control effects under different scenarios, the optimal combination of disposal plans is screened out.

[0138] The decision support solution is visualized through the geographic information display platform. The platform needs to support the overlay display of multi-level data, including the base layer (displaying the water terrain), the risk distribution layer (displaying areas of different risk levels), the sensitive area layer (displaying environmentally sensitive targets) and the emergency resource layer (displaying the distribution of rescue equipment). Through the interactive operation interface, emergency commanders can quickly obtain detailed information at any location.

[0139] For example, a chemical enterprise had a hazardous chemical leakage accident, and the leaked substance was a benzene compound. According to toxicological data, the acute toxicity threshold was set at 50 mg / L, the chronic toxicity threshold was set at 10 mg / L, and the lethal concentration was set at 200 mg / L. The spatiotemporal distribution prediction results showed that there were areas with a lethal concentration within 200 meters around the leakage source. After graded calibration, these data were displayed as red high-risk areas. In the spatial distribution analysis, it was found that the high-risk area overlapped with the protection range of an important water intake, which required priority treatment. Through the analysis of the characteristic parameters of the area, it was found that the pollution value continued to exceed the standard and was near the main navigation channel. Considering the distribution of emergency resources on site, including two emergency disposal ships and 1,000 meters of oil booms, a "containment + adsorption" disposal plan was formulated. Multi-scenario simulations compared the effects of different containment plans, including single-point containment, multi-point containment, and segmented containment. The simulation results show that the scheme of upstream double-point containment combined with downstream single-point interception can effectively control the spread of pollutants. At the same time, placing adsorption materials at key locations can significantly reduce the concentration of pollutants in water bodies.

[0140] In a specific embodiment, the process of performing multi-scenario simulation calculation steps on the partition treatment scheme may specifically include the following steps:

[0141] (1) Divide the disposal scenarios for each disposal area in the zoning disposal plan, extract the regional area, pollutant type and concentration data, and establish a scenario disposal parameter set;

[0142] (2) Analyze the pollution control process of the scenario treatment parameter set, calculate the time consumption and resource requirements of each treatment measure, and form a treatment resource allocation table;

[0143] (3) Based on the disposal resource allocation table and the on-site deployable resource data, a matching analysis is performed to determine the optimal disposal sequence for each disposal area and output a zoned disposal sequence plan;

[0144] (4) For each treatment step in the zoning treatment sequence plan, calculate the pollutant reduction rate and treatment cost, and generate effect evaluation indicators;

[0145] (5) Comprehensively rank the effect evaluation indicators according to the reduction rate, cost and time consumption, screen out the optimal treatment combination, and form the optimal treatment plan result;

[0146] (6) Convert the optimization results of the disposal plan into specific operational instructions, including the disposal sequence, resource allocation and emergency response points, and generate an emergency disposal recommendation list.

[0147] Specifically, each area in the zoning disposal plan is first divided into disposal scenarios, which include a combination of different disposal methods such as physical interception, chemical treatment and biodegradation. The basic parameters of each area include area size, physical and chemical properties of pollution (such as solubility, volatility), current concentration level and diffusion trend. These parameters are sorted to form a scenario disposal parameter set as the basic data for subsequent analysis. When analyzing the pollution control process based on the scenario disposal parameter set, it is necessary to consider the implementation conditions and efficiency of various disposal measures. For physical interception, calculate the deployment time, coverage area and required length of the oil boom; for chemical treatment, estimate the amount of agent released, reaction time and mixing efficiency; for biodegradation, evaluate the required number of strains, culture time and degradation cycle. Through these analyses, the specific time and resource quantity required for each disposal measure are obtained to form a disposal resource allocation table.

[0148] Match and analyze the disposal resource allocation table with the actual resources that can be deployed on site, including the number of emergency ships, professional personnel, equipment and material reserves, etc. Through the optimization algorithm, the time sequence and spatial distribution of resource allocation are considered to determine the most reasonable disposal sequence. This process needs to balance the disposal priorities and resource utilization efficiency of different regions, and finally output a detailed zoning disposal sequence plan. In the disposal effect evaluation stage, quantitative analysis is performed on each disposal step in the sequence plan. The calculation of pollutant reduction rate needs to consider the actual efficiency of disposal measures, the impact of environmental conditions and the risk of secondary pollution. Disposal costs include equipment usage fees, labor costs, material consumption and transportation costs. By establishing an evaluation index system, the effect of each disposal step is quantitatively evaluated.

[0149] The comprehensive ranking of effect evaluation indicators adopts a multi-objective decision-making method. After standardizing the three dimensions of reduction rate, cost and time consumption, different weight coefficients are assigned. For areas with high urgency, the reduction rate weight is larger; for environmentally sensitive areas, the cost weight is relatively small. Through comprehensive scoring, the optimal disposal combination plan is screened out.

[0150] The emergency response recommendation list needs to transform the preferred options into specific and actionable instructions. The instructions include detailed disposal sequence, personnel division of labor, equipment deployment and emergency response process. Each disposal step has a clear time node, operation requirements and precautions. At the same time, an emergency plan needs to be formulated to deal with possible emergencies.

[0151] For example, a solvent leak occurred at a chemical terminal, and the contaminated area was divided into three disposal areas, namely the high-concentration core area, the medium-concentration diffusion area, and the low-concentration impact area. For the core area, the disposal parameters show an area of ​​2,000 square meters, and the main disposal solution is oil boom combined with adsorbent. Resource demand analysis shows that 200 meters of oil boom and 500 kilograms of adsorbent material are needed, and the disposal time is expected to be 4 hours.

[0152] Based on the on-site emergency resource reserves, it was found that the oil boom reserves were sufficient but the adsorption materials needed to be transported and replenished. Through time optimization, it was decided to contain the core area first, and at the same time transport the adsorption materials, and immediately carry out the adsorption treatment after the materials were in place. The effect evaluation showed that this treatment method can reduce the pollutant concentration in the core area by 90% within 6 hours, and the disposal cost is within an acceptable range. For the diffusion area, a chemical flocculation combined with salvage solution is adopted, which requires the addition of flocculants and the provision of salvage equipment. The resource allocation plan stipulates the amount of flocculant added and the operating range of the salvage ship, and the optimal addition ratio and salvage frequency are determined through field tests. The disposal sequence is arranged to be carried out at the same time as the core area disposal to prevent the continued spread of pollutants.

[0153] The above describes the method for predicting the spread of hazardous chemicals based on an unmanned ship in the embodiment of the present application. The following describes the system for predicting the spread of hazardous chemicals based on an unmanned ship in the embodiment of the present application. Figure 2 In the embodiment of the present application, an embodiment of a hazardous chemical leakage and diffusion prediction system based on an unmanned ship includes:

[0154] The acquisition module 201 is used to collect water quality parameters, meteorological data and geographic information through a sensor array carried on the unmanned ship, and process them through a data acquisition controller to obtain an initial monitoring data set;

[0155] The calculation module 202 is used to calculate the potential diffusion range according to the initial monitoring data set, optimize the divided grid monitoring area, and obtain a multi-point collaborative monitoring trajectory plan;

[0156] The smoothing module 203 is used to remove outliers and smooth the data on the initial monitoring data set according to the multi-point collaborative monitoring trajectory scheme, and to construct a spatial distribution to obtain a fused processing data set;

[0157] A construction module 204 is used to construct and solve a group of convection-diffusion equations based on the fused processing data set to obtain a three-dimensional flow field model of the water area;

[0158] The analysis module 205 is used to analyze the diffusion path of hazardous chemicals in the water area through the three-dimensional flow field model of the water area, and obtain the prediction results of the spatiotemporal distribution of hazardous chemical pollutants;

[0159] The generation module 206 is used to establish a pollution risk level assessment system based on the spatiotemporal distribution prediction results of the hazardous chemical pollutants, generate emergency disposal suggestions, and obtain a visual decision support plan.

[0160] Through the coordinated cooperation of the above-mentioned components, the multi-parameter synchronous monitoring is carried out by the sensor array carried on the unmanned ship, which realizes the rapid and accurate collection of water quality parameters, meteorological data and geographic information, and overcomes the problem of limited coverage of traditional fixed monitoring points; through the processing and analysis of the initial monitoring data set, the accurate calculation of the potential diffusion range and the dynamic optimization of the grid monitoring area are realized, which improves the monitoring efficiency and data quality; the multi-point collaborative monitoring trajectory scheme is adopted, combined with the outlier removal and data smoothing processing technology, to ensure the reliability and continuity of the monitoring data; by constructing and solving the convection-diffusion equation group, a three-dimensional flow field model that accurately reflects the movement characteristics of the water area is obtained, which provides a reliable basis for the prediction of pollutant diffusion; based on the three-dimensional flow field model of the water area, the diffusion path analysis is carried out to realize the accurate prediction of the spatiotemporal distribution of hazardous chemical pollutants, and effectively improve the accuracy of the prediction results; by establishing a pollution risk level assessment system and generating emergency disposal recommendations, an intuitive and efficient visual decision support plan is formed, which provides a scientific basis for emergency command decision-making and improves the emergency disposal capacity and efficiency of hazardous chemical leakage accidents in the water area.

[0161] The present application also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. Instructions are stored in the computer-readable storage medium. When the instructions are executed on a computer, the computer executes the steps of the method for predicting the leakage and diffusion of hazardous chemicals based on an unmanned ship.

[0162] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0163] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program codes.

[0164] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for predicting the leakage and diffusion of hazardous chemicals based on an unmanned ship, characterized in that: The hazardous chemicals leakage and diffusion prediction method based on unmanned ships includes: The sensor array carried on the unmanned boat collects water quality parameters, meteorological data and geographic information, which are processed by the data acquisition controller to obtain the initial monitoring data set; The potential diffusion range is calculated based on the initial monitoring data set, and the divided grid monitoring area is optimized to obtain a multi-point collaborative monitoring trajectory plan; According to the multi-point collaborative monitoring trajectory scheme, outliers are removed and data smoothing is performed on the initial monitoring data set, and spatial distribution is constructed to obtain a fused processing data set; Based on the fused processed data set, a three-dimensional flow field model of the water area is obtained by constructing and solving a group of convection-diffusion equations; Through the three-dimensional flow field model of the water area, the diffusion path of hazardous chemicals in the water area is analyzed to obtain the spatiotemporal distribution prediction results of hazardous chemical pollutants; Based on the spatiotemporal distribution prediction results of hazardous chemical pollutants, a pollution risk level assessment system is established, and emergency disposal recommendations are generated to obtain a visual decision support plan.

2. The method for predicting the leakage and diffusion of hazardous chemicals based on an unmanned ship according to claim 1 is characterized in that: The water quality parameters, meteorological data and geographic information are collected by the sensor array carried on the unmanned ship, and processed by the data acquisition controller to obtain the initial monitoring data set, including: Scanning and sampling the water area through the pH sensor, dissolved oxygen sensor, conductivity sensor and turbidity sensor in the sensor array to obtain water quality parameter data; The meteorological environment above the water area is sampled by the anemometer, wind vane and temperature and humidity sensors in the sensor array to obtain meteorological data; The GPS positioning device and the sonar depth sounder in the sensor array are used to locate and measure the water area to obtain geographic information; Synchronously marking the water quality parameter data, the meteorological data and the geographic information according to the sampling time to obtain raw data with a timestamp; The raw data with time stamps are classified and sorted according to water quality parameter indicators, meteorological parameter indicators and geographical parameter indicators, and each indicator is linearly normalized and dimensionally unified, and then the processed indicators are merged and cross-validated to obtain standardized data; The initial monitoring data set is obtained by performing a data integrity scan on the standardized data, identifying data missing points and outliers, interpolating adjacent values ​​for missing points, replacing outliers with medians, and then extracting time series features and checking data consistency on the processed data.

3. The method for predicting the leakage and diffusion of hazardous chemicals based on an unmanned ship according to claim 1 is characterized in that: The potential diffusion range is calculated according to the initial monitoring data set, and the divided grid monitoring area is optimized to obtain a multi-point collaborative monitoring trajectory plan, including: Performing gradient analysis on the water quality parameters in the initial monitoring data set, extracting the trend of pollutant concentration changes, and obtaining preliminary diffusion direction determination results; Performing correlation analysis on the meteorological data and geographic information in the initial monitoring data set to obtain water area environment impact factors; Delimiting the potential diffusion range based on the preliminary diffusion direction determination result and the water area environmental impact factor, and dividing the delimited range into grids according to the pollutant diffusion characteristics to obtain an initial grid monitoring area; Density distribution optimization is performed on the initial grid monitoring area, grid density is performed in areas where pollutant concentration changes dramatically, and grid sparseness is performed in areas where the change is gentle, to obtain an optimized grid monitoring area; Prioritizing the sampling points in the optimized grid monitoring area, assigning a weight value to each sampling point based on the pollutant concentration gradient and diffusion rate, and obtaining a weighted sampling sequence in combination with the navigation constraints of the unmanned ship; The weighted sampling sequence is subjected to spatiotemporal continuity analysis and path smoothing processing, the navigation path is subjected to curve fitting and steering angle optimization, and the path is fine-tuned considering obstacle avoidance requirements, thereby obtaining the multi-point collaborative monitoring trajectory solution.

4. The method for predicting the leakage and diffusion of hazardous chemicals based on an unmanned ship according to claim 1 is characterized in that: According to the multi-point collaborative monitoring trajectory scheme, outliers are removed and data smoothing is performed on the initial monitoring data set, and spatial distribution is constructed to obtain a fused processing data set, including: The initial monitoring data set is divided into time series based on the sampling point position information in the multi-point collaborative monitoring trajectory scheme to form a data sequence corresponding to the spatial position; Based on the mean, standard deviation and coefficient of variation, the statistical characteristics of the data sequence are quantified, outliers exceeding the threshold are identified based on the confidence interval, and outlier marking data is generated; Through time and space correlation analysis, the abnormal points in the abnormal value marked data are compared with the adjacent sampling values, the abnormal data are corrected and eliminated, and the cleaned data sequence is obtained; Establishing a sliding window on the cleaned data sequence, performing a weighted average operation on the data in the window and verifying the data continuity, and outputting smoothed data; According to the spatial coordinates of the sampling points, distribution mapping and interval interpolation filling are performed on the smoothed data to construct preliminary spatial distribution data; The consistency of the preliminary spatial distribution data is corrected and optimized through spatial autocorrelation analysis and boundary constraint processing to obtain the fused processing data set.

5. The method for predicting the leakage and diffusion of hazardous chemicals based on an unmanned ship according to claim 1, characterized in that: Based on the fusion processing data set, a three-dimensional flow field model of the water area is obtained by constructing and solving a group of convection-diffusion equations, including: Extracting three-dimensional coordinates of water boundary points from the fused processing data set, stratifying the extracted coordinate points according to longitude, latitude and depth, and then generating computational grid node data with spatial position information through rasterization processing; Based on the computational grid node data, the temperature, salinity and density of the water quality parameters are calibrated in layers according to the depth, the vertical layered structure is determined in combination with the water depth change gradient, and the hydrodynamic boundary condition parameters including the thermocline and density layering are constructed; According to the hydrodynamic boundary condition parameters, the momentum equation of water body motion is spatially differentiated, the continuity equation and the momentum equation are converted into a group of algebraic equations, and a flow field calculation matrix describing the water flow velocity and pressure distribution is established; For the flow field calculation matrix, the velocity components in the horizontal direction and the vertical direction are respectively decomposed and calculated, and the velocity field and the pressure field are iteratively solved by a pressure correction method to obtain initial flow field data representing the three-dimensional motion characteristics of the water body; The initial flow field data is coupled with the wind speed and wind direction information in the meteorological data, the water surface wind stress and internal shear stress are calculated, and the Coriolis force and gravity are considered to output multi-level flow field data including the influence of external forces; The multi-level flow field data are subjected to near-wall boundary layer processing and turbulence intensity calculation, and the interaction between the layers is corrected by the vertical mixing coefficient to obtain the three-dimensional flow field model of the water area, wherein the three-dimensional flow field model of the water area includes the velocity vector and flow direction information of the water body at different depths and different positions, and is used to characterize the three-dimensional motion state of any point in the water area.

6. The method for predicting the leakage and diffusion of hazardous chemicals based on an unmanned ship according to claim 1 is characterized in that: The three-dimensional flow field model of the water area is used to analyze the diffusion path of hazardous chemicals in the water area to obtain the spatiotemporal distribution prediction results of hazardous chemical pollutants, including: Extracting velocity vector and flow direction data from the three-dimensional flow field model of the water area, combining the physical and chemical characteristic parameters of hazardous chemicals, calculating the diffusion coefficient of pollutants in the water body, and obtaining initial diffusion parameters; Based on the initial diffusion parameters, the diffusion rates of pollutants in the horizontal and vertical directions are decomposed and calculated, and the influence of the thermocline on the vertical diffusion is considered to obtain the layered diffusion rate data; The layered diffusion rate data is coupled with the water body density distribution data to analyze the migration law of pollutants in different water layers and obtain multi-level pollutant migration data; Dividing the multi-level pollutant transport data into time steps, calculating the pollutant concentration gradient change in each time step, and considering the influence of degradation reaction kinetics to obtain instantaneous concentration distribution data; Performing time series superposition on the instantaneous concentration distribution data, taking into account the cumulative effects of water disturbance and mixing, to obtain pollutant distribution data under continuous time series; The pollutant distribution data under the continuous time series is projected in time and space, and the concentration field is numerically smoothed to obtain the prediction results of the time and space distribution of the hazardous chemical pollutants.

7. The method for predicting the leakage and diffusion of hazardous chemicals based on an unmanned ship according to claim 1 is characterized in that: According to the spatiotemporal distribution prediction results of hazardous chemicals pollutants, a pollution risk level assessment system is established, and emergency disposal suggestions are generated to obtain a visual decision support plan, including: Extract pollutant concentration data from the spatiotemporal distribution prediction results of the hazardous chemicals pollutants, perform classification and calibration according to toxicity thresholds, and obtain initial risk classification data; Based on the initial risk classification data, the spatial distribution range and exposure time of each concentration interval are counted, the water environment sensitive areas are marked, and a risk area distribution map is obtained; For the high-risk areas in the risk area distribution map, extract their location coordinates and pollution characteristic parameters, and obtain emergency response zoning data in combination with the navigation conditions of the waters and the distribution of rescue resources; Sorting the emergency response zoning data according to the difficulty of pollutant control and disposal priority, determining emergency disposal measures for different areas, and generating zoning disposal plans; Conduct multi-scenario simulation calculations on the zoning disposal plan, evaluate the control effects of different disposal measures, and output a list of emergency disposal recommendations; The risk area distribution map and the emergency response recommendation list are integrated, and data visualization is performed through a geographic information display platform to obtain the visualized decision support solution.

8. The method for predicting the leakage and diffusion of hazardous chemicals based on an unmanned ship according to claim 7 is characterized in that: The multi-scenario simulation calculation is performed on the partition disposal plan to evaluate the control effects of different disposal measures and output an emergency disposal suggestion list, including: Divide each disposal area in the zoning disposal plan into disposal scenarios, extract regional area, pollutant type and concentration data, and establish a scenario disposal parameter set; Perform pollution control process analysis on the scenario disposal parameter set, calculate the time consumption and resource requirements of each disposal measure, and form a disposal resource allocation table; Based on the disposal resource configuration table and the on-site deployable resource data, a matching analysis is performed to determine the optimal disposal sequence for each disposal area, and a partition disposal sequence plan is output; For each treatment step in the zoning treatment sequence plan, the pollutant reduction rate and treatment cost are calculated to generate effect evaluation indicators; The effect evaluation indicators are comprehensively ranked according to the reduction rate, cost and time consumption, and the optimal treatment combination is screened out to form the treatment plan optimization result; The optimization result of the disposal plan is converted into specific operation instructions, including disposal sequence, resource allocation and emergency response points, to generate the emergency disposal recommendation list.

9. A hazardous chemical leakage and diffusion prediction system based on an unmanned ship, used to implement the hazardous chemical leakage and diffusion prediction method based on an unmanned ship as described in any one of claims 1 to 8, characterized in that: The hazardous chemical leakage and diffusion prediction system based on unmanned ship includes: The acquisition module is used to collect water quality parameters, meteorological data and geographic information through the sensor array carried on the unmanned ship, and the data acquisition controller processes them to obtain the initial monitoring data set; The calculation module is used to calculate the potential diffusion range based on the initial monitoring data set, optimize the divided grid monitoring area, and obtain a multi-point collaborative monitoring trajectory plan; A smoothing module is used to remove outliers and smooth the data on the initial monitoring data set according to the multi-point collaborative monitoring trajectory scheme, and to construct a spatial distribution to obtain a fused processing data set; A construction module is used to construct and solve a group of convection-diffusion equations based on the fused processing data set to obtain a three-dimensional flow field model of the water area; An analysis module is used to analyze the diffusion path of hazardous chemicals in the water area through the three-dimensional flow field model of the water area, and obtain the prediction results of the spatiotemporal distribution of hazardous chemical pollutants; The generation module is used to establish a pollution risk level assessment system based on the spatiotemporal distribution prediction results of hazardous chemical pollutants, generate emergency disposal suggestions, and obtain a visual decision support plan.

10. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instruction is executed by the processor, the method for predicting the leakage and diffusion of hazardous chemicals based on an unmanned ship as described in any one of claims 1-8 is implemented.

Images