A method and system for real-time monitoring of maritime information and early warning

Abstract

This invention relates to the field of marine disaster early warning technology. It discloses a method and system for real-time monitoring of marine information, comprising: acquiring surface detection data collected by each surface monitoring device within an outer tier in real time; triggering each surface monitoring device within an inner tier to collect surface detection data when the pollutant type is determined based on the surface detection data; obtaining pollution data of the pollutant based on the surface detection data collected by the outer and inner tiers; using the pollution data as input features of a pre-configured pollutant prediction model to obtain the estimated time of pollutant arrival at the target sea area; and providing important decision support for maintaining the health and sustainability of nearshore fishery resources by predicting the estimated time of pollutant arrival at the target sea area.

Description

Technical Field

[0001] This invention relates to the field of marine disaster early warning technology, and more specifically, to an early warning method and system for real-time monitoring of marine information. Background Technology

[0002] In recent years, with the increasing demand for marine resources, nearshore aquaculture has developed rapidly. Unlike inland freshwater aquaculture, nearshore waters are in direct contact with offshore waters, making them susceptible to water quality fluctuations. Meanwhile, the spread of marine disasters (such as oil tanker spills and offshore drilling leaks) from offshore and adjacent areas also threatens the safety and production stability of nearshore aquaculture farms. Therefore, an efficient and accurate early warning method is needed to take timely measures to maintain the health and sustainability of nearshore fishery resources.

[0003] Currently, existing marine disaster early warning methods or systems for nearshore areas are generally designed for early warning of marine biological disasters in nearshore areas, such as red tide early warning to protect fishery resources. Although there is some literature on early warning of nearshore pollution disasters, such as Chinese Patent Publication No. CN113177183B which discloses a method and system for monitoring and early warning of seawater pollution based on marine remote sensing images, this type of method achieves early warning of nearshore pollution disasters by monitoring the pollution trend of nearshore rivers; however, research and practical application of the above methods and existing technologies have revealed at least the following shortcomings:

[0004] (1) It is impossible to make timely predictions and warnings for pollution events from outside the target sea area, which can easily cause the escape or death of organisms in nearshore aquaculture areas, thus making it difficult to provide important decision support for maintaining the health and sustainability of nearshore fishery resources;

[0005] (2) Monitoring is delayed and the prediction characteristics and results are easily interfered with, making it difficult to apply to early warning of small-scale seawater pollution events. Summary of the Invention

[0006] The purpose of this invention is to provide a method and system for real-time monitoring of maritime information to solve one or more technical problems existing in the prior art, and to at least provide a beneficial option or create conditions.

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] A real-time monitoring and early warning method for maritime information, the method relying on multiple surface monitoring devices deployed in a tiered manner to form a monitoring zone consisting of an outer tier and an inner tier, the method comprising:

[0009] The system acquires water surface monitoring data from each water surface monitoring device within the outer tier in real time. When the pollutant type is determined based on the water surface monitoring data, it triggers each water surface monitoring device within the inner tier to collect water surface monitoring data. The water surface monitoring data includes a set of material elements and a water surface image.

[0010] Pollution data of pollutants are obtained based on water surface monitoring data collected from the outer and inner tiers; the pollution data includes the location through which the pollutants flow, the time of arrival at the location, the flow velocity, the flow direction angle, and the diffusion coefficient.

[0011] Pollution data is used as input features for a pre-configured pollutant prediction model, which is then used to obtain the estimated time when pollutants arrive at the target sea area.

[0012] Furthermore, before determining the type of pollutant based on surface monitoring data, this includes:

[0013] Extract the set of material elements from the water surface detection data, wherein the set of material elements contains a variety of detected material elements;

[0014] Retrieve the pollutant element database, and query and traverse the pollutant element database for each substance element in the substance element set to obtain the query results;

[0015] If the query results show an empty set, it is determined that no marine pollution disaster event has occurred; if the query results show a non-empty set, it is determined that a marine pollution disaster event has occurred.

[0016] Furthermore, the determination of pollutant type based on water surface detection data includes:

[0017] The corresponding surface monitoring equipment that detects marine pollution disasters within the outer tiers is marked as the first target equipment;

[0018] Acquire the set of material elements and water surface images collected by the first target device, and extract the texture and color features from the water surface images;

[0019] The material element set, texture features, and color features are combined into a combined feature, and the combined feature is input into a pre-configured pollutant identification model to obtain the pollutant type.

[0020] Furthermore, the generation logic of the pre-configured pollutant identification model is as follows:

[0021] Historical pollutant identification data is acquired and divided into a pollutant identification training set and a pollutant identification test set; the historical pollutant identification data includes combined features and their corresponding labels.

[0022] A classifier is constructed by using the combined features in the pollutant identification training set as the input data of the classifier and the corresponding labeled data in the pollutant identification training set as the output data of the classifier. The classifier is then trained to obtain an initial classification network.

[0023] The initial classification network is validated using the input data of the classifier. The initial classification network with an output accuracy greater than or equal to the preset test accuracy is used as the pre-configured pollutant identification model.

[0024] Furthermore, after triggering each surface monitoring device within the internal tier to collect surface monitoring data, the process includes:

[0025] Within a set time frame, the corresponding surface monitoring equipment that detects marine pollution disasters within the inner perimeter tiers will be marked as the second target equipment;

[0026] Count the number of the second target devices and compare the number of the second target devices with a preset number threshold;

[0027] If the number of second target devices is less than or equal to a preset threshold, it is determined that the marine pollution disaster will not spread to the target sea area; if the number of second target devices is greater than the preset threshold, it is determined that the marine pollution disaster will spread to the target sea area.

[0028] Furthermore, the location through which the pollutant flows is determined based on the coordinates of the first target device and the second target device; the time of arrival at the location through which the pollutant flows is determined based on the time of the marine pollution disaster event determined by the first target device and the second target device.

[0029] Furthermore, the generation logic of the pre-configured pollutant prediction model is as follows:

[0030] Acquire historical estimated time training data, and divide the historical estimated time training data into an estimated time training set and an estimated time test set; the historical estimated time training data includes input features and the corresponding pollutant arrival time in the target sea area.

[0031] A regression network is constructed by using the input features from the estimated time training set as the input data of the regression network and the arrival time of pollutants in the target sea area from the estimated time training set as the output data of the regression network. The regression network is then trained to obtain the initial pollutant prediction network.

[0032] The initial pollutant prediction network is validated using the estimated time test set. The initial pollutant prediction network whose output is less than or equal to the test error is used as the pre-configured pollutant prediction model.

[0033] A real-time monitoring and early warning system for maritime information includes:

[0034] The data collection module is used to acquire water surface detection data collected by each water surface monitoring device in the outer tier in real time. When the pollutant type is determined based on the water surface detection data, it triggers each water surface monitoring device in the inner tier to collect water surface detection data. The water surface detection data includes a set of material elements and a water surface image.

[0035] The feature extraction module is used to obtain pollution data of pollutants based on the water surface detection data collected from the outer and inner tiers; the pollution data includes the location through which the pollutants flow, the time of arrival at the location, the flow velocity, the flow direction angle, and the diffusion coefficient.

[0036] The disaster early warning module uses pollution data as input features for a pre-configured pollutant prediction model, and then uses the model to obtain the estimated time when pollutants arrive at the target sea area.

[0037] An electronic device includes a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the early warning method for real-time monitoring of maritime information as described above.

[0038] A computer-readable storage medium storing a computer program, which, when executed, implements the early warning method for real-time monitoring of maritime information as described above.

[0039] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0040] This application discloses a real-time monitoring and early warning method and system for marine information, comprising: acquiring surface monitoring data collected by each surface monitoring device in the outer tier in real time; when the pollutant type is determined based on the surface monitoring data, triggering each surface monitoring device in the inner tier to collect surface monitoring data; acquiring pollution data of the pollutants based on the surface monitoring data collected by the outer and inner tiers; using the pollution data as input features of a pre-configured pollutant prediction model, and using the pre-configured pollutant prediction model to obtain the estimated time of arrival of the pollutants in the target sea area; based on the above process, this invention is beneficial for timely prediction and early warning of pollution events from outside the target sea area, thereby helping to avoid the escape or death of organisms in nearshore aquaculture areas, and providing important decision support for maintaining the health and sustainability of nearshore fishery resources; in addition, compared with the prior art, by deploying gradient monitoring zones, this invention can be applied to early warning of small-scale marine pollution disaster events. Attached Figure Description

[0041] Figure 1 A flowchart illustrating an early warning method for real-time monitoring of maritime information provided by the present invention;

[0042] Figure 2 This invention provides a schematic diagram of a module for a real-time monitoring and early warning system for maritime information.

[0043] Figure 3 This invention provides a schematic diagram of the deployment of surface monitoring equipment and the division of sea areas.

[0044] Figure 4 A schematic diagram of the structure of an electronic device provided by the present invention;

[0045] Figure 5 This is a schematic diagram of the structure of a computer-readable storage medium provided by the present invention. Detailed Implementation

[0046] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0047] Example 1

[0048] Please see Figure 1 As shown, this embodiment discloses an early warning method for real-time monitoring of maritime information. The method relies on multiple surface monitoring devices, which are deployed in a tiered manner to form a monitoring zone consisting of an outer tier and an inner tier. The method includes:

[0049] S101: Real-time acquisition of water surface detection data collected by each water surface monitoring device in the outer tier; when the pollutant type is determined based on the water surface detection data, triggering each water surface monitoring device in the inner tier to collect water surface detection data; the water surface detection data includes a set of material elements and a water surface image;

[0050] It should be noted that the target sea area is specifically the nearshore aquaculture waters or nearshore fishery resource management waters, while the non-target sea area is relative to the target sea area and includes nearshore sea areas adjacent to the open sea or the target sea area. The scope and area of ​​the target sea area are determined according to the specific circumstances of the actual target sea area.

[0051] It is worth noting that the multiple water surface monitoring devices are arranged in a tiered manner, such as... Figure 3(Diagram of Surface Monitoring Equipment Deployment and Sea Area Division) As shown in the diagram, the shaded area represents the target sea area, and the unshaded area represents the non-target sea area. In this diagram, two tiers of surface monitoring equipment are deployed within the non-target sea area and close to the target sea area (i.e., the target sea area is surrounded by two tiers of surface monitoring equipment), namely the outer tier and the inner tier. Each outer tier of surface monitoring equipment includes several surface monitoring devices. It should be further noted that the tier furthest from the target sea area (i.e., the outermost) is designated as the outer tier, while the tier between the outer tier and the boundary of the target sea area is designated as the inner tier. It should also be noted that there is only one outer tier, while multiple inner tiers can be set up. The tiers are deployed at equal intervals, such as 1 nautical mile, 2 nautical miles, ..., 3 nautical miles or K nautical miles, where K is a positive integer. The surface monitoring devices within each tier are also deployed at equal intervals, such as 100 meters, 200 meters, ..., 300 meters or R meters, where R is a positive integer.

[0052] Specifically, the water surface monitoring equipment is either a monitoring buoy or a monitoring unmanned vessel, and the water surface monitoring equipment is equipped with at least an electrochemical sensor, an optical sensor, a biological sensor, a chemical sensor, a heavy metal sensor, an organic pollutant sensor, and a camera device, etc.

[0053] It is understandable that before collecting surface monitoring data, monitoring buoys are deployed in advance to designated monitoring locations in non-target sea areas, or several monitoring unmanned vessels are pre-controlled to arrive at designated monitoring locations in non-target sea areas to form a multi-tiered monitoring zone. When pollutants are detected by any surface monitoring device in the outer tier, it indicates that a marine pollution disaster event has occurred, and a marine pollution disaster event warning needs to be issued to determine the estimated time when pollutants will arrive in the target sea area, so as to take timely measures to maintain the health and sustainability of nearshore fishery resources.

[0054] During implementation, before determining the type of pollutant based on surface monitoring data, the following steps are included:

[0055] Extract the set of material elements from the water surface detection data, wherein the set of material elements contains a variety of detected material elements;

[0056] Retrieve the pollutant element database, and query and traverse the pollutant element database for each substance element in the substance element set to obtain the query results;

[0057] It should be noted that a pollution element database has been established in advance, and identified pollution elements such as organophosphorus, lead, mercury, cadmium, chromium, sulfur, and nitrogen have been included in the database beforehand.

[0058] If the query results show an empty set, it is determined that no marine pollution disaster event has occurred; if the query results show a non-empty set, it is determined that a marine pollution disaster event has occurred.

[0059] Understandably, when the query results show an empty set, the collected set of material elements does not contain any polluting elements that have been pre-identified in the polluting element database, which means that no marine water pollution has occurred. Conversely, when the query results show a non-empty set, the collected set of material elements contains any polluting elements that have been pre-identified in the polluting element database, which means that marine water pollution has occurred.

[0060] In practice, determining the type of pollutant based on water surface monitoring data includes:

[0061] The corresponding surface monitoring equipment that detects marine pollution disasters within the outer tiers is marked as the first target equipment;

[0062] Acquire the set of material elements and water surface images collected by the first target device, and extract the texture and color features from the water surface images;

[0063] It should be noted that the texture and color features in the water surface images are extracted using existing methods such as histograms, convolutional neural networks (CNNs), or pre-trained models (such as ResNet, VGG, and Inception). It is understood that these models have been trained on large-scale image data and have powerful feature extraction capabilities.

[0064] The material element set, texture features, and color features are combined into a combined feature, and the combined feature is input into a pre-configured pollutant identification model to obtain the pollutant type;

[0065] Specifically, the generation logic of the pre-configured pollutant identification model is as follows:

[0066] Historical pollutant identification data is acquired and divided into a pollutant identification training set and a pollutant identification test set; the historical pollutant identification data includes combined features and their corresponding labels.

[0067] Specifically, the label indicates the type of pollutant, which includes, but is not limited to, petroleum pollutants, pesticide pollutants, fertilizer pollutants, heavy metal pollutants, and plastic pollutants.

[0068] It should be noted that the labeling is achieved through manual or automatic labeling. For example, the material elements of petroleum pollutants include organic matter, sulfur, nitrogen, oxygen and other elements. The color characteristics of petroleum pollutants are dark brown or black, and their texture characteristics are that they are flaky in water. Therefore, the combination of characteristics that meet the above conditions is labeled as petroleum pollutants. Similarly, the labeling principle for other types of pollutants is the same, and will not be elaborated on further.

[0069] A classifier is constructed by using the combined features in the pollutant identification training set as the input data of the classifier and the corresponding labeled data in the pollutant identification training set as the output data of the classifier. The classifier is then trained to obtain an initial classification network.

[0070] The initial classification network is validated using the input data of the classifier. The initial classification network with an output accuracy greater than or equal to the preset test accuracy is used as the pre-configured pollutant identification model.

[0071] It should be noted that the classifier is specifically an RNN recurrent neural network model;

[0072] In one alternative implementation, after triggering each surface monitoring device within the internal tier to collect surface monitoring data, the process includes:

[0073] Within a set time frame, the corresponding surface monitoring equipment that detects marine pollution disasters within the inner perimeter tiers will be marked as the second target equipment;

[0074] It should be understood that: how to determine a marine pollution disaster event, and how to mark the corresponding surface monitoring equipment that detected a marine pollution disaster event within the inner perimeter as the second target equipment; the logic for marine pollution disaster events and marking the first target equipment is the same as described above, and details can be found in the description above, so we will not elaborate further here;

[0075] It should be noted that the set time range is determined based on the specific experimental results of the technicians, such as 1 hour, 2 hours, ..., 3 hours or W hours, where W is an integer greater than zero;

[0076] Count the number of the second target devices and compare the number of the second target devices with a preset number threshold;

[0077] If the number of the second target devices is less than or equal to a preset threshold, it is determined that the marine pollution disaster will not spread to the target sea area; if the number of the second target devices is greater than the preset threshold, it is determined that the marine pollution disaster will spread to the target sea area.

[0078] It is understandable that by using a monitoring zone consisting of outer and inner tiers, i.e., by deploying equipment in multiple tiers, the steps in this embodiment can make a targeted judgment on whether a marine pollution disaster event will affect the target sea area. To further explain, by determining whether a certain number of second target devices will appear in the inner tier within a certain period of time, it can be known whether pollutants will spread to the target sea area, which is conducive to deciding whether to trigger or not to trigger the early warning mechanism, and also helps to save computing resources.

[0079] It should be understood that when a determination is made that a marine pollution disaster will not spread to the target sea area, the subsequent early warning mechanism will not be triggered, that is, it is assumed that the pollutants will not spread to the target sea area; conversely, when a determination is made that a marine pollution disaster will spread to the target sea area, the subsequent early warning mechanism will be triggered, that is, it is assumed that the pollutants will spread to the target sea area, and the estimated time for the pollutants to arrive at the target sea area is predicted.

[0080] S102: Based on the water surface monitoring data collected from the outer and inner tiers, obtain the pollution data of the pollutants; the pollution data includes the location through which the pollutants flow, the time of arrival at the location, the flow velocity, the flow direction angle, and the diffusion coefficient.

[0081] Specifically, the location through which the pollutant flows is determined based on the coordinates of the first target device and the second target device; the time of arrival at the location through which the pollutant flows is determined based on the time of the marine pollution disaster event determined by the first target device and the second target device.

[0082] It is understandable that when the first target device or the second target device is obtained, it means that pollutants have appeared in the area of ​​the corresponding water surface monitoring device, and the coordinates of the corresponding water surface monitoring device are taken as the location through which the pollutants flow; similarly, when the first target device or the second target device is obtained, it is considered that a marine pollution disaster event has occurred, and the time when the first target device or the second target device causes the marine pollution disaster event is taken as the time when it arrives at the location through which the pollutants flow.

[0083] It should be noted that the average flow velocity of the pollutants is calculated based on each set of data. For example, assuming there are three target devices in the outer tier (A1, A2, and A3) and two target devices in the inner tier (B1 and B2), there are six possible combinations: (A1, A2), (A1, A3), (A2, A3), (A1, B1), (A1, B2), and (B1, B2). The flow positions and arrival positions of the two target devices in each combination are then used as a set of data. The formula for calculating the average flow velocity of the pollutants is: In the formula: V represents the average flow velocity of the pollutants. iLet be the flow velocity of the i-th pollutant group, Q be the total number of data groups, (X1,Y1) and (X2,Y2) be the flow positions of the two pollutants, and T2-T1 be the absolute difference in the time taken to reach the two flow positions. Correspondingly, the logic for obtaining the flow direction angle of the pollutants is also similar, such as using the arctangent function to calculate the flow direction angle of the angular pollutants.

[0084] It should also be noted that the water surface detection data also includes the concentration values ​​of pollutants; the diffusion coefficient is calculated based on the concentration values ​​of pollutants, and its calculation formula is as follows: In the formula: J is the diffusion flux, representing the amount of matter passing through a unit area per unit time, with units of kg / m². 2 / s; Let C be the gradient of pollutant concentration C with respect to spatial X; D is the diffusion coefficient.

[0085] It is understandable that, unlike traditional marine disaster prediction models, this implementation step uses a first target device in the outer tier and a second target device in the inner tier to obtain the location, time of arrival, velocity, and direction angle of pollutants. This helps to eliminate or weaken the variability of parameters such as actual ocean current velocity and wind speed, thereby improving the accuracy of the final model prediction.

[0086] S103: Use pollution data as input features of a pre-configured pollutant prediction model to obtain the estimated time when pollutants arrive at the target sea area.

[0087] Specifically, the generation logic of the pre-configured pollutant prediction model is as follows:

[0088] Acquire historical estimated time training data, and divide the historical estimated time training data into an estimated time training set and an estimated time test set; the historical estimated time training data includes input features and the corresponding pollutant arrival time in the target sea area.

[0089] It should be understood that the input features in the historical time prediction training data include the location through which the pollutant flows, the time of arrival at the location, the flow velocity, the angle of flow direction, and the diffusion coefficient; while the time of arrival of the corresponding pollutant in the target sea area in the historical time prediction training data is obtained based on the actual situation or experimental records.

[0090] A regression network is constructed by using the input features from the estimated time training set as the input data of the regression network and the arrival time of pollutants in the target sea area from the estimated time training set as the output data of the regression network. The regression network is then trained to obtain the initial pollutant prediction network.

[0091] The initial pollutant prediction network was validated using the estimated time test set. The initial pollutant prediction network whose output was less than or equal to the test error was used as the pre-configured pollutant prediction model.

[0092] It should be noted that the regression network mentioned is specifically one of the following models: random forest regression, multinomial regression, support vector machine regression, or neural network.

[0093] By predicting the estimated time when pollutants will arrive at the target sea area, this embodiment can provide important decision support for maintaining the health and sustainability of nearshore fishery resources. In addition, compared with the prior art, by deploying gradient monitoring strips, this invention can be applied to early warning of small-scale marine pollution disasters.

[0094] Example 2

[0095] Please see Figure 2 As shown, based on the same inventive concept, this embodiment discloses an early warning system for real-time monitoring of maritime information, including:

[0096] The data collection module 210 is used to acquire water surface detection data collected by each water surface monitoring device in the outer tier in real time. When the pollutant type is determined based on the water surface detection data, it triggers each water surface monitoring device in the inner tier to collect water surface detection data. The water surface detection data includes a set of material elements and a water surface image.

[0097] It should be noted that the target sea area is specifically the nearshore aquaculture waters or nearshore fishery resource management waters, while the non-target sea area is relative to the target sea area and includes nearshore sea areas adjacent to the open sea or the target sea area. The scope and area of ​​the target sea area are determined according to the specific circumstances of the actual target sea area.

[0098] It is worth noting that the multiple water surface monitoring devices are arranged in a tiered manner, such as... Figure 3(Diagram of Surface Monitoring Equipment Deployment and Sea Area Division) As shown in the diagram, the shaded area represents the target sea area, and the unshaded area represents the non-target sea area. In this diagram, two tiers of surface monitoring equipment are deployed within the non-target sea area and close to the target sea area (i.e., the target sea area is surrounded by two tiers of surface monitoring equipment), namely the outer tier and the inner tier. Each outer tier of surface monitoring equipment includes several surface monitoring devices. It should be further noted that the tier furthest from the target sea area (i.e., the outermost) is designated as the outer tier, while the tier between the outer tier and the boundary of the target sea area is designated as the inner tier. It should also be noted that there is only one outer tier, while multiple inner tiers can be set up. The tiers are deployed at equal intervals, such as 1 nautical mile, 2 nautical miles, ..., 3 nautical miles or K nautical miles, where K is a positive integer. The surface monitoring devices within each tier are also deployed at equal intervals, such as 100 meters, 200 meters, ..., 300 meters or R meters, where R is a positive integer.

[0099] Specifically, the water surface monitoring equipment is either a monitoring buoy or a monitoring unmanned vessel, and the water surface monitoring equipment is equipped with at least an electrochemical sensor, an optical sensor, a biological sensor, a chemical sensor, a heavy metal sensor, an organic pollutant sensor, and a camera device, etc.

[0100] It is understandable that before collecting surface monitoring data, monitoring buoys are deployed in advance to designated monitoring locations in non-target sea areas, or several monitoring unmanned vessels are pre-controlled to arrive at designated monitoring locations in non-target sea areas to form a multi-tiered monitoring zone. When pollutants are detected by any surface monitoring device in the outer tier, it indicates that a marine pollution disaster event has occurred, and a marine pollution disaster event warning needs to be issued to determine the estimated time when pollutants will arrive in the target sea area, so as to take timely measures to maintain the health and sustainability of nearshore fishery resources.

[0101] During implementation, before determining the type of pollutant based on surface monitoring data, the following steps are included:

[0102] Extract the set of material elements from the water surface detection data, wherein the set of material elements contains a variety of detected material elements;

[0103] Retrieve the pollutant element database, and query and traverse the pollutant element database for each substance element in the substance element set to obtain the query results;

[0104] It should be noted that a pollution element database has been established in advance, and identified pollution elements such as organophosphorus, lead, mercury, cadmium, chromium, sulfur, and nitrogen have been included in the database beforehand.

[0105] If the query results show an empty set, it is determined that no marine pollution disaster event has occurred; if the query results show a non-empty set, it is determined that a marine pollution disaster event has occurred.

[0106] Understandably, when the query results show an empty set, the collected set of material elements does not contain any polluting elements that have been pre-identified in the polluting element database, which means that no marine water pollution has occurred. Conversely, when the query results show a non-empty set, the collected set of material elements contains any polluting elements that have been pre-identified in the polluting element database, which means that marine water pollution has occurred.

[0107] In practice, determining the type of pollutant based on water surface monitoring data includes:

[0108] The corresponding surface monitoring equipment that detects marine pollution disasters within the outer tiers is marked as the first target equipment;

[0109] Acquire the set of material elements and water surface images collected by the first target device, and extract the texture and color features from the water surface images;

[0110] It should be noted that the texture and color features in the water surface images are extracted using existing methods such as histograms, convolutional neural networks (CNNs), or pre-trained models (such as ResNet, VGG, and Inception). It is understood that these models have been trained on large-scale image data and have powerful feature extraction capabilities.

[0111] The material element set, texture features, and color features are combined into a combined feature, and the combined feature is input into a pre-configured pollutant identification model to obtain the pollutant type;

[0112] Specifically, the generation logic of the pre-configured pollutant identification model is as follows:

[0113] Historical pollutant identification data is acquired and divided into a pollutant identification training set and a pollutant identification test set; the historical pollutant identification data includes combined features and their corresponding labels.

[0114] Specifically, the label indicates the type of pollutant, which includes, but is not limited to, petroleum pollutants, pesticide pollutants, fertilizer pollutants, heavy metal pollutants, and plastic pollutants.

[0115] It should be noted that the labeling is achieved through manual or automatic labeling. For example, the material elements of petroleum pollutants include organic matter, sulfur, nitrogen, oxygen and other elements. The color characteristics of petroleum pollutants are dark brown or black, and their texture characteristics are that they are flaky in water. Therefore, the combination of characteristics that meet the above conditions is labeled as petroleum pollutants. Similarly, the labeling principle for other types of pollutants is the same, and will not be elaborated on further.

[0116] A classifier is constructed by using the combined features in the pollutant identification training set as the input data of the classifier and the corresponding labeled data in the pollutant identification training set as the output data of the classifier. The classifier is then trained to obtain an initial classification network.

[0117] The initial classification network is validated using the input data of the classifier. The initial classification network with an output accuracy greater than or equal to the preset test accuracy is used as the pre-configured pollutant identification model.

[0118] It should be noted that the classifier is specifically an RNN recurrent neural network model;

[0119] In one alternative implementation, after triggering each surface monitoring device within the internal tier to collect surface monitoring data, the process includes:

[0120] Within a set time frame, the corresponding surface monitoring equipment that detects marine pollution disasters within the inner perimeter tiers will be marked as the second target equipment;

[0121] It should be understood that: how to determine a marine pollution disaster event, and how to mark the corresponding surface monitoring equipment that detected a marine pollution disaster event within the inner perimeter as the second target equipment; the logic for marine pollution disaster events and marking the first target equipment is the same as described above, and details can be found in the description above, so we will not elaborate further here;

[0122] It should be noted that the set time range is determined based on the specific experimental results of the technicians, such as 1 hour, 2 hours, ..., 3 hours or W hours, where W is an integer greater than zero;

[0123] Count the number of the second target devices and compare the number of the second target devices with a preset number threshold;

[0124] If the number of the second target devices is less than or equal to a preset threshold, it is determined that the marine pollution disaster will not spread to the target sea area; if the number of the second target devices is greater than the preset threshold, it is determined that the marine pollution disaster will spread to the target sea area.

[0125] It is understandable that by using a monitoring zone consisting of outer and inner tiers, i.e., by deploying equipment in multiple tiers, the steps in this embodiment can make a targeted judgment on whether a marine pollution disaster event will affect the target sea area. To further explain, by determining whether a certain number of second target devices will appear in the inner tier within a certain period of time, it can be known whether pollutants will spread to the target sea area, which is conducive to deciding whether to trigger or not to trigger the early warning mechanism, and also helps to save computing resources.

[0126] It should be understood that when a determination is made that a marine pollution disaster will not spread to the target sea area, the subsequent early warning mechanism will not be triggered, that is, it is assumed that the pollutants will not spread to the target sea area; conversely, when a determination is made that a marine pollution disaster will spread to the target sea area, the subsequent early warning mechanism will be triggered, that is, it is assumed that the pollutants will spread to the target sea area, and the estimated time for the pollutants to arrive at the target sea area is predicted.

[0127] The feature extraction module 220 is used to obtain pollution data of pollutants based on the water surface detection data collected from the outer and inner tiers; the pollution data includes the location through which the pollutants flow, the time of arrival at the location, the flow velocity, the flow direction angle, and the diffusion coefficient.

[0128] Specifically, the location through which the pollutant flows is determined based on the coordinates of the first target device and the second target device; the time of arrival at the location through which the pollutant flows is determined based on the time of the marine pollution disaster event determined by the first target device and the second target device.

[0129] It is understandable that when the first target device or the second target device is obtained, it means that pollutants have appeared in the area of ​​the corresponding water surface monitoring device, and the coordinates of the corresponding water surface monitoring device are taken as the location through which the pollutants flow; similarly, when the first target device or the second target device is obtained, it is considered that a marine pollution disaster event has occurred, and the time when the first target device or the second target device causes the marine pollution disaster event is taken as the time when it arrives at the location through which the pollutants flow.

[0130] It should be noted that the average flow velocity of the pollutants is calculated based on each set of data. For example, assuming there are three target devices in the outer tier (A1, A2, and A3) and two target devices in the inner tier (B1 and B2), there are six possible combinations: (A1, A2), (A1, A3), (A2, A3), (A1, B1), (A1, B2), and (B1, B2). The flow positions and arrival positions of the two target devices in each combination are then used as a set of data. The formula for calculating the average flow velocity of the pollutants is: In the formula: V represents the average flow velocity of the pollutants. iLet be the flow velocity of the i-th pollutant group, Q be the total number of data groups, (X1,Y1) and (X2,Y2) be the flow positions of the two pollutants, and T2-T1 be the absolute difference in the time taken to reach the two flow positions. Correspondingly, the logic for obtaining the flow direction angle of the pollutants is also similar, such as using the arctangent function to calculate the flow direction angle of the angular pollutants.

[0131] It should also be noted that the water surface detection data also includes the concentration values ​​of pollutants; the diffusion coefficient is calculated based on the concentration values ​​of pollutants, and its calculation formula is as follows: In the formula: J is the diffusion flux, representing the amount of matter passing through a unit area per unit time, with units of kg / m². 2 / s; Let C be the gradient of pollutant concentration C with respect to spatial X; D is the diffusion coefficient.

[0132] It is understandable that, unlike traditional marine disaster prediction models, this implementation step uses a first target device in the outer tier and a second target device in the inner tier to obtain the location, time of arrival, velocity, and direction angle of pollutants. This helps to eliminate or weaken the variability of parameters such as actual ocean current velocity and wind speed, thereby improving the accuracy of the final model prediction.

[0133] The disaster early warning module 230 is used to take pollution data as input features of a pre-configured pollutant prediction model and use the pre-configured pollutant prediction model to obtain the estimated time when pollutants arrive at the target sea area.

[0134] Specifically, the generation logic of the pre-configured pollutant prediction model is as follows:

[0135] Acquire historical estimated time training data, and divide the historical estimated time training data into an estimated time training set and an estimated time test set; the historical estimated time training data includes input features and the corresponding pollutant arrival time in the target sea area.

[0136] It should be understood that the input features in the historical time prediction training data include the location through which the pollutant flows, the time of arrival at the location, the flow velocity, the angle of flow direction, and the diffusion coefficient; while the time of arrival of the corresponding pollutant in the target sea area in the historical time prediction training data is obtained based on the actual situation or experimental records.

[0137] A regression network is constructed by using the input features from the estimated time training set as the input data of the regression network and the arrival time of pollutants in the target sea area from the estimated time training set as the output data of the regression network. The regression network is then trained to obtain the initial pollutant prediction network.

[0138] The initial pollutant prediction network was validated using the estimated time test set. The initial pollutant prediction network whose output was less than or equal to the test error was used as the pre-configured pollutant prediction model.

[0139] It should be noted that the regression network mentioned is specifically one of the following models: random forest regression, multinomial regression, support vector machine regression, or neural network.

[0140] By predicting the estimated time when pollutants will arrive at the target sea area, this embodiment can provide important decision support for maintaining the health and sustainability of nearshore fishery resources. In addition, compared with the prior art, by deploying gradient monitoring strips, this invention can be applied to early warning of small-scale marine pollution disasters.

[0141] Example 3

[0142] Please see Figure 4 As shown, this embodiment discloses an electronic device, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the early warning method for real-time monitoring of maritime information as described in any of the above methods.

[0143] Since the electronic device described in this embodiment is an electronic device used to implement the early warning method for real-time monitoring of maritime information in the embodiments of this application, those skilled in the art can understand the specific implementation methods and various variations of the electronic device in this embodiment based on the early warning method for real-time monitoring of maritime information described in the embodiments of this application. Therefore, how the electronic device implements the method in the embodiments of this application will not be described in detail here. Any electronic device used by those skilled in the art to implement the early warning method for real-time monitoring of maritime information in the embodiments of this application falls within the scope of protection of this application.

[0144] Example 4

[0145] Please see Figure 5 As shown, this embodiment discloses a computer-readable storage medium storing a computer program, which, when executed, implements the early warning method for real-time monitoring of maritime information as described above.

[0146] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters, weights, and thresholds in the formulas are set by those skilled in the art according to the actual situation.

[0147] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via a wired or wireless network. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0148] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed in this invention can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0149] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0150] In the several embodiments provided by this invention, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only one method, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0151] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0152] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0153] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

[0154] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A warning method for real-time monitoring of maritime information, the method relying on a plurality of water surface monitoring devices, characterized in that, The multiple water surface monitoring devices are arranged in echelon to form a monitoring belt composed of an outer echelon and an inner echelon, and the method comprises the following steps: Real-time acquisition of water surface detection data collected by each water surface monitoring device in the outer echelon, triggering collection of water surface detection data by each water surface monitoring device in the inner echelon when the type of the pollutant is determined according to the water surface detection data; the water surface detection data comprises a set of material elements and a water surface image; Acquisition of pollution data of the pollutant according to the water surface detection data collected by the outer echelon and the inner echelon; the pollution data comprises a flowing position of the pollutant, a time of arrival at the flowing position, a flow rate, a flow direction angle and a diffusion coefficient; Taking the pollution data as input features of a preconfigured pollutant prediction model, and obtaining an estimated time of arrival of the pollutant at a target sea area by using the preconfigured pollutant prediction model.

2. The early warning method of real-time monitoring of maritime information according to claim 1, characterized in that, Before determining the type of the pollutant according to the water surface detection data, the following steps are included: Extracting a set of material elements in the water surface detection data, wherein the set of material elements contains multiple detected material elements; Calling a pollution element database, and placing each material element in the set of material elements in the pollution element database for query iteration to obtain a query result; When the query result shows an empty set, it is determined that no marine pollution disaster event occurs; when the query result shows a non-empty set, it is determined that a marine pollution disaster event occurs.

3. The early warning method of real-time monitoring of marine information according to claim 2, characterized in that, The determination of the type of the pollutant according to the water surface detection data comprises the following steps: Marking a corresponding water surface monitoring device in the outer echelon that monitors the occurrence of the marine pollution disaster event as a first target device; Obtaining a set of material elements and a water surface image collected by the first target device, and extracting texture features and color features in the water surface image; Combining the set of material elements, the texture features and the color features into combined features, and inputting the combined features into a preconfigured pollutant identification model to obtain the type of the pollutant.

4. The early warning method of real-time monitoring of marine information according to claim 3, characterized in that, The generation logic of the preconfigured pollutant identification model is as follows: Obtaining historical pollutant identification data, and dividing the historical pollutant identification data into a pollutant identification training set and a pollutant identification test set; the historical pollutant identification data comprises combined features and corresponding labeled labels; Building a classifier, taking the combined features in the pollutant identification training set as input data of the classifier, and taking the corresponding labeled labels in the pollutant identification training set as output data of the classifier, training the classifier to obtain an initial classification network; Model verification of the initial classification network by using the input data of the classifier, and outputting an initial classification network with a test accuracy greater than or equal to a preset test accuracy as the preconfigured pollutant identification model.

5. The early warning method of real-time monitoring of maritime information according to claim 4, characterized in that, After triggering collection of water surface detection data by each water surface monitoring device in the inner echelon, the following steps are included: Within a set time range, marking a corresponding water surface monitoring device in the inner echelon that monitors the occurrence of the marine pollution disaster event as a second target device; Counting the number of the second target devices, and comparing the number of the second target devices with a preset number threshold; If the number of the second target devices is less than or equal to the preset number threshold, it is determined that the marine pollution disaster event will not spread to the target sea area; if the number of the second target devices is greater than the preset number threshold, it is determined that the marine pollution disaster event will spread to the target sea area.

6. The early warning method of real-time monitoring of marine information according to claim 5, characterized in that, The flowing position of the pollutant is determined according to coordinate values of the first target device and the second target device; The time of reaching the flowing position is determined according to a time determined by the first target device and the second target device according to the marine pollution disaster event.

7. The early warning method of real-time monitoring of maritime information according to claim 6, characterized in that, The generation logic of the preconfigured pollutant prediction model is as follows: Historical estimated time training data is obtained, and the historical estimated time training data is divided into an estimated time training set and an estimated time test set; the historical estimated time training data includes input features and corresponding times of the pollutant reaching a target sea area; A regression network is constructed, the input features in the estimated time training set are used as input data of the regression network, the times of the pollutant reaching the target sea area in the estimated time training set are used as output data of the regression network, the regression network is trained, and an initial pollutant prediction network is obtained; The initial pollutant prediction network is verified by using the estimated time test set, and the initial pollutant prediction network with a test error less than or equal to a test error is output as the preconfigured pollutant prediction model.

8. An early warning system for real-time monitoring of maritime information, characterized in that, It comprises: A data collection module is configured to acquire water surface detection data collected by each water surface monitoring device in the outer echelon in real time, and trigger each water surface monitoring device in the inner echelon to collect water surface detection data when the type of the pollutant is determined according to the water surface detection data; the water surface detection data includes a set of material elements and a water surface image; A feature extraction module is configured to acquire pollution data of the pollutant according to the water surface detection data collected by the outer echelon and the inner echelon; the pollution data includes a flowing position of the pollutant, a time of reaching the flowing position, a flow rate, a flow direction angle, and a diffusion coefficient; A disaster warning module is configured to use the pollution data as input features of a preconfigured pollutant prediction model, and acquire an estimated time of the pollutant reaching a target sea area by using the preconfigured pollutant prediction model.

9. An electronic device comprising a memory, a processor, and a computer program stored on the memory and running on the processor, characterized in that, The processor executes the computer program to realize the early warning method for real-time monitoring of marine information according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a computer program, and the computer program is executed to realize the early warning method for real-time monitoring of marine information according to any one of claims 1 to 7.

Images