A Big Data-Based Method for Marine Environmental Risk Identification

Abstract

This invention relates to the field of artificial intelligence technology and discloses a method for identifying marine environmental risks based on big data. The method includes: standardizing marine environmental monitoring data to obtain a standardized dataset of the ocean; performing correlation analysis on the standardized dataset based on a pre-set risk knowledge base to obtain risk correlation features of the ocean; constructing a topology of the risk correlation features based on historical risk events of the ocean to obtain an environmental risk feature map of the ocean; tracking the trajectory of the environmental risk feature map to obtain the spatiotemporal evolution path of the ocean; simulating the spatiotemporal evolution path to obtain a risk evolution trajectory report of the ocean; and comprehensively evaluating the risk evolution trajectory report to obtain the risk level and early warning information of the ocean. This invention can improve the efficiency of marine environmental risk identification based on big data.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence technology, and in particular to a method for identifying marine environmental risks based on big data. Background Technology

[0002] In the field of marine environmental risk identification, existing technologies lack efficient standardized processing mechanisms for multi-source heterogeneous marine environmental monitoring data, making it difficult to achieve in-depth data integration and unified standardization. Due to the lack of a systematic risk correlation analysis system, it is impossible to accurately mine the risk correlation characteristics hidden in the data, resulting in insufficient comprehensiveness of risk identification and difficulty in covering various potential risk factors in complex marine environments.

[0003] Meanwhile, traditional methods have significant limitations in capturing the evolution patterns of risks. They lack the ability to dynamically quantify and construct topologies of risk transmission paths, and cannot effectively track the spatiotemporal evolution of risks. Their risk assessment and early warning rely on static data and empirical models, which are difficult to adapt to the dynamic characteristics of the marine environment. This results in poor accuracy in risk level determination and insufficient timeliness and relevance of early warning information, failing to provide reliable decision support for marine environmental risk management. Therefore, how to improve the accuracy, timeliness, and relevance of marine environmental risk identification and early warning has become an urgent problem to be solved. Summary of the Invention

[0004] This invention provides a method for identifying marine environmental risks based on big data, in order to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, this invention provides a marine environmental risk identification method based on big data, comprising:

[0006] S1. Standardize the marine environmental monitoring data to obtain a standardized dataset of the ocean.

[0007] S2. Based on a preset risk knowledge base, perform correlation analysis on the standardized dataset to obtain the risk correlation characteristics of the ocean;

[0008] S3. Based on the historical risk events of the ocean, perform topological construction on the risk correlation features to obtain the environmental risk feature map of the ocean;

[0009] S4. Track the trajectory of the environmental risk feature map to obtain the spatiotemporal evolution path of the ocean;

[0010] S5. Perform path simulation on the spatiotemporal evolution path to obtain a risk evolution trajectory report of the ocean;

[0011] S6. Conduct a comprehensive analysis of the risk evolution trajectory report to obtain the risk level and early warning information of the ocean.

[0012] In a preferred embodiment, the standardization process of the marine environmental monitoring data to obtain a standardized dataset of the ocean includes:

[0013] Receive marine environmental monitoring data streams;

[0014] The environmental monitoring data of the ocean is cleaned in real time to obtain preliminary purification data of the ocean;

[0015] Spatiotemporal alignment is performed on the preliminary cleanup data of the ocean to obtain intermediate standardized data of the ocean;

[0016] Multimodal fusion is performed on the intermediate standard data to obtain the standard dataset of the ocean.

[0017] In a preferred embodiment, the step of performing correlation analysis on the standardized dataset based on a preset risk knowledge base to obtain the risk correlation characteristics of the ocean includes:

[0018] The standardized dataset is mapped to a preset risk knowledge base to obtain the mapping relationship of the ocean;

[0019] Based on the mapping relationship, the risk knowledge base is searched to obtain the initial risk association features of the ocean;

[0020] The confidence level of the initial risk association features is verified to obtain the refining risk association features of the ocean.

[0021] The initial risk association features and the refined risk association features are fused to obtain the risk association features of the ocean.

[0022] In a preferred embodiment, the step of performing topological construction on the risk correlation features based on the historical risk events of the ocean to obtain an environmental risk feature map of the ocean includes:

[0023] By performing spatiotemporal matching between the historical risk events of the ocean and the risk-related features, the correspondence between the event features of the ocean can be obtained.

[0024] The network is reconstructed based on the correspondence of the event features to obtain the initial risk topology network of the ocean;

[0025] Redundancy elimination is performed on the initial risk topology network to obtain the refined risk topology structure of the ocean;

[0026] The refined risk topology is mapped to the geographic spatiotemporal coordinate system of the ocean to obtain the spatiotemporal topology layer of the ocean;

[0027] The transmission path in the refined risk topology is quantitatively tracked to obtain the real-time dynamic risk transmission intensity of the transmission path;

[0028] The real-time dynamic risk transmission intensity value is visualized and enhanced to obtain an environmental risk characteristic map of the ocean.

[0029] In a preferred embodiment, the step of quantifying and tracking the transmission paths in the refined risk topology to obtain the real-time dynamic risk transmission intensity of the transmission paths includes:

[0030] The node sequence and connecting edges of the transmission path are serialized and extracted to obtain the path composition data of the ocean;

[0031] The effectiveness of the node sequence and the connecting edges is evaluated to obtain the node influence degree and edge influence degree of the ocean.

[0032] Based on the node influence, the edge influence, and the time difference corresponding to the path composition data, the real-time dynamic risk transmission intensity of the transmission path is calculated, wherein the calculation formula for the real-time dynamic risk transmission intensity is as follows:

[0033] ;

[0034] in, This represents the real-time dynamic risk transmission intensity along the transmission path. For the first The influence degree of each node. For the first The influence degree of the edge mentioned in the article, The time difference is... The preset time decay constant, For the summation function, It is a logarithmic function. It is a natural exponential function;

[0035] Based on the real-time dynamic risk transmission intensity, the spatiotemporal topology layer is integrated and rendered to obtain the environmental risk feature map of the ocean.

[0036] In a preferred embodiment, the step of tracking the trajectory of the environmental risk feature map to obtain the spatiotemporal evolution path of the ocean includes:

[0037] Source tracing analysis of the environmental risk characteristic map yields the set of risk transmission origins in the ocean;

[0038] The transmission patterns in the historical risk events and the intensity of the real-time dynamic risk transmission are used as dynamic transmission probability parameters of the risk transmission starting point set.

[0039] Based on the dynamic transmission probability parameters, the transmission state of the risk transmission starting point set is deduced to obtain the preliminary transmission trajectory of the ocean;

[0040] The initial transmission trajectory is subjected to spatiotemporal consistency verification to obtain the optimized trajectory set of the ocean;

[0041] Cluster analysis is performed on the optimized trajectory set to obtain the spatiotemporal evolution path of the ocean.

[0042] In a preferred embodiment, the step of performing spatiotemporal consistency verification on the preliminary propagation trajectory to obtain the optimized trajectory set of the ocean includes:

[0043] The continuity of the preliminary transmission trajectory is verified to obtain the spatially continuous trajectory of the ocean;

[0044] The credibility of the continuous spatial trajectory is evaluated to obtain the overall transmission confidence of the ocean trajectory;

[0045] Based on the comprehensive transmission confidence of the trajectory, trajectory optimization is performed on the spatial continuous trajectory to obtain the optimized trajectory set of the ocean.

[0046] In a preferred embodiment, the step of performing path simulation on the spatiotemporal evolution path to obtain the risk evolution trajectory report of the ocean includes:

[0047] The starting point of the main transmission path, key turning points and transmission direction sequence are extracted from the spatiotemporal evolution path to obtain the path skeleton data of the ocean;

[0048] The path skeleton data is parameterized and encapsulated to obtain a parameterized path framework for the ocean.

[0049] Based on the parameterized path framework, the main transmission path is coupled and simulated to obtain a set of simulated trajectories of the ocean;

[0050] Based on the set of simulated trajectories, a risk evolution trajectory report for the ocean is generated.

[0051] In a preferred embodiment, based on the parameterized path model, a set of simulated trajectories of the ocean is obtained by performing random perturbation superposition simulation on the main transmission path, including:

[0052] Based on the parameterized path framework, the node sequence of the main transmission path is constructed in a benchmark manner to obtain the benchmark transmission trajectory of the ocean.

[0053] Based on the parameterized path framework, the baseline transmission trajectory is subjected to perturbation superposition processing to obtain the perturbed transmission trajectory of the ocean.

[0054] The reference conduction trajectory and the disturbed conduction trajectory are spatiotemporally aligned to obtain a set of simulated trajectories of the ocean.

[0055] In a preferred embodiment, the step of comprehensively analyzing the risk evolution trajectory report to obtain the risk level and early warning information of the ocean includes:

[0056] The trend values, impact range, and path characteristics of the trajectory in the risk evolution trajectory report are fused in multiple dimensions to obtain the structured analysis input of the ocean;

[0057] Based on the structured analysis input, the historical risk case database of the ocean is matched and retrieved to obtain the case matching results of the ocean;

[0058] Based on the case matching results, the risk evolution trajectory report is quantitatively evaluated to obtain the comprehensive risk index of the risk evolution trajectory report;

[0059] The risk index is mapped to a level to obtain the risk level of the ocean;

[0060] By integrating the risk level and the structured analysis input, early warning information for the ocean is obtained.

[0061] Compared with the prior art, the present invention has the following beneficial effects:

[0062] 1. This invention achieves deep integration and unified standardization of data from different types and sources by systematically standardizing marine environmental monitoring data, forming a well-structured and comprehensive standardized dataset, providing high-quality data support for subsequent risk correlation analysis; based on a pre-set risk knowledge base, it performs precise correlation analysis, efficiently mining the risk correlation characteristics hidden in the data, and constructs an environmental risk characteristic map that accurately reflects the risk correlation relationship and spatiotemporal distribution by combining historical risk events, accurately tracking the spatiotemporal evolution path of risks, and comprehensively covering multiple possible scenarios of risk evolution.

[0063] 2. This invention objectively derives a comprehensive risk index and a clear risk level through a series of coherent comprehensive analysis operations, such as multi-dimensional fusion and case matching retrieval. It generates early warning information containing key information such as risk trends, scope of impact, and transmission paths, which significantly improves the accuracy, timeliness, and pertinence of marine environmental risk identification and provides comprehensive and reliable decision support for marine environmental risk management, ensuring the scientific nature and effectiveness of risk response measures. Attached Figure Description

[0064] Figure 1A flowchart illustrating a marine environmental risk identification method based on big data, provided as an embodiment of the present invention;

[0065] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0066] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0067] This application provides a method for identifying marine environmental risks based on big data. The executing entity of this method includes, but is not limited to, at least one of the following: a server, a terminal, or other electronic devices configured to execute the method provided in this application. In other words, the method can be executed by software or hardware installed on a terminal device or a server-side device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cluster of cloud servers. The server can be an independent server or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0068] Reference Figure 1 The diagram shown is a flowchart illustrating a marine environmental risk identification method based on big data, according to an embodiment of the present invention. In this embodiment, the marine environmental risk identification method based on big data includes:

[0069] S1. Standardize the marine environmental monitoring data to obtain a standardized dataset of the ocean.

[0070] In this embodiment of the invention, the standardization processing of marine environmental monitoring data to obtain a standardized dataset of the ocean includes:

[0071] Receive marine environmental monitoring data streams;

[0072] The environmental monitoring data of the ocean is cleaned in real time to obtain preliminary purification data of the ocean;

[0073] Spatiotemporal alignment is performed on the preliminary cleanup data of the ocean to obtain intermediate standardized data of the ocean;

[0074] Multimodal fusion is performed on the intermediate standard data to obtain the standard dataset of the ocean.

[0075] The reception of marine environmental monitoring data streams is carried out through various specialized monitoring devices deployed in nearshore, offshore, and offshore areas. These devices include water quality sensors, hydrological monitoring buoys, meteorological observation stations, and marine ecological monitoring devices. All devices continuously capture marine environmental data at a pre-set fixed acquisition frequency. After data acquisition, the data is transmitted in real time to the data processing terminal via wireless communication networks or undersea wired transmission links. The data processing terminal establishes a dedicated data receiving interface to verify and confirm each transmitted data in real time, ensuring that all raw data collected by the monitoring devices enters the subsequent processing stage without loss or damage, ultimately forming a continuous and complete marine environmental monitoring data stream covering different sea areas and monitoring dimensions.

[0076] When performing real-time cleaning of marine environmental monitoring data streams, the reasonable value range and standard data format for each monitoring data item are first determined based on the natural attributes and historical statistical data of each marine environmental monitoring indicator. Then, each data item in the data stream is checked one by one. If the value of a data item exceeds the reasonable value range of the corresponding indicator, it is directly judged as abnormal data and removed. If some fields are found to be missing, they are supplemented according to the trend of effective data from adjacent collection periods of the monitoring equipment. If identical duplicate data is detected, only the earliest collected data is retained. At the same time, the format of entries that do not meet the standard is converted and adjusted to ensure that the format of all data is consistent. After the above series of operations, preliminary cleaned data is obtained by removing abnormal, missing, duplicate, and non-standard data.

[0077] When performing spatiotemporal alignment on preliminary purification data, a unified time scale and spatial coordinate system are first determined. The time scale is set to a fixed time interval. Preliminary purification data with different timestamps collected by different monitoring devices are uniformly adjusted to the time node corresponding to this fixed time interval. This ensures that all valid monitoring data from all monitoring devices within that time period can be collected at the same time node. The spatial coordinate system adopts the nationally unified geographic coordinate system. All geographical location information attached to the preliminary purification data collected from different monitoring points is converted into coordinate data under this unified geographic coordinate system. This ensures that each piece of preliminary purification data corresponds to a unique time identifier and spatial coordinates, achieving unified calibration of data in both time and space dimensions, thereby forming intermediate standardized data.

[0078] When performing multimodal fusion on intermediate standard data, the various modal types included in the intermediate standard data are first sorted out, including monitoring data of different dimensions such as water quality index data, hydrological and dynamic data, meteorological element data, and marine ecological data. Then, using a unified time identifier and spatial coordinates as the core association basis, intermediate standard data of different modalities corresponding to the same time and spatial location are accurately matched, and the single-dimensional data scattered in various modalities are integrated into an organic whole. The information of water quality, hydrology, meteorology, ecology and other aspects under the same spatiotemporal background is presented in a centralized manner, forming a unified data set in which each data contains complete spatiotemporal information and multi-dimensional environmental parameters. This data set is the standard dataset of the ocean.

[0079] The beneficial effects are that this standardized processing flow can comprehensively and accurately process marine environmental monitoring data. Through systematic receiving, cleaning, spatiotemporal alignment, and multimodal fusion operations, it effectively removes invalid interference information from the data, ensuring the accuracy, integrity, and consistency of the data. At the same time, it achieves unified and standardized integration of data of different types and sources, forming a standardized dataset with a clear structure and comprehensive information. This provides high-quality and reliable data support for subsequent correlation analysis based on a preset risk knowledge base, ensures the accuracy of risk correlation feature extraction, and lays a solid foundation for the efficient advancement of the entire marine environmental risk identification process.

[0080] S2. Based on a preset risk knowledge base, perform correlation analysis on the standardized dataset to obtain the risk correlation characteristics of the ocean;

[0081] In this embodiment of the invention, the step of performing correlation analysis on the standardized dataset based on a preset risk knowledge base to obtain the risk correlation characteristics of the ocean includes:

[0082] The standardized dataset is mapped to a preset risk knowledge base to obtain the mapping relationship of the ocean;

[0083] Based on the mapping relationship, the risk knowledge base is searched to obtain the initial risk association features of the ocean;

[0084] The confidence level of the initial risk association features is verified to obtain the refining risk association features of the ocean.

[0085] The initial risk association features and the refined risk association features are fused to obtain the risk association features of the ocean.

[0086] When mapping a standardized dataset to a pre-defined risk knowledge base, the core structure of the pre-defined risk knowledge base is first clarified. This knowledge base contains fixed content such as various types of marine environmental risks, corresponding characteristic indicators, data dimension divisions, and association rules. Then, for each data item in the standardized dataset, its monitoring dimension, numerical attributes, and spatiotemporal identifier are compared and matched one by one with the corresponding classification entries in the knowledge base. For example, the seawater temperature data in the standardized dataset corresponds to the temperature characteristic indicators under the temperature anomaly risk in the knowledge base, and the dissolved oxygen data corresponds to the dissolved oxygen characteristic indicators under the water quality deterioration risk. Through this precise entry correspondence, a one-to-one correspondence between the standardized dataset and the risk knowledge base is established. This correspondence is the mapping relationship of the ocean.

[0087] Based on the established mapping relationship, the core retrieval scope and direction of the standardized dataset in the risk knowledge base are determined. Then, according to the knowledge base entries pointed to by the mapping relationship, all risk feature information associated with each entry is retrieved layer by layer, including the manifestation of risk features, influencing factors, and related data thresholds. All retrieved risk feature information associated with the standardized dataset is collected and organized to form a set covering various potential associated risk features. This set is the initial risk association feature of the ocean.

[0088] Before verifying the confidence level of the initial risk association features, a clear confidence level judgment standard is formulated based on historical valid risk association cases and marine environmental risk patterns. This standard clarifies the minimum matching requirements for the effective association between risk features and the standardized dataset, including core judgment dimensions such as data fit and spatiotemporal correlation. Then, each feature in the initial risk association features is checked one by one to determine whether its association with the standardized dataset meets the confidence level judgment standard. All invalid association features that do not meet the standard are eliminated, and only valid association features that fully meet the standard are retained to form the refined risk association features of the ocean.

[0089] When fusing the initial risk association features and the refined risk association features, the effective features that were not eliminated from the initial risk association features are first screened out and merged with all the features in the refined risk association features. Only one copy of the feature entries that appear repeatedly is retained. At the same time, the internal relationship logic between different features is sorted out, the primary and secondary relationships and interaction mechanisms between features are clarified, so that the merged feature set forms a unified whole with complete structure, clear association and no redundancy. This whole is the risk association feature of the ocean.

[0090] The beneficial effects are that this association analysis process, through precise mapping, retrieval, verification and fusion operations, can efficiently mine real and effective risk association information from standardized datasets, ensure the accuracy and completeness of risk association features, avoid interference from invalid features, provide high-quality basic data for subsequent topology construction based on historical risk events, and effectively improve the accuracy of marine environmental risk identification and the overall process efficiency.

[0091] S3. Based on the historical risk events of the ocean, perform topological construction on the risk correlation features to obtain the environmental risk feature map of the ocean;

[0092] In this embodiment of the invention, the step of performing topological construction on the risk correlation features based on the historical risk events of the ocean to obtain the environmental risk feature map of the ocean includes:

[0093] By performing spatiotemporal matching between the historical risk events of the ocean and the risk-related features, the correspondence between the event features of the ocean can be obtained.

[0094] The network is reconstructed based on the correspondence of the event features to obtain the initial risk topology network of the ocean;

[0095] Redundancy elimination is performed on the initial risk topology network to obtain the refined risk topology structure of the ocean;

[0096] The refined risk topology is mapped to the geographic spatiotemporal coordinate system of the ocean to obtain the spatiotemporal topology layer of the ocean;

[0097] The transmission path in the refined risk topology is quantitatively tracked to obtain the real-time dynamic risk transmission intensity of the transmission path;

[0098] The real-time dynamic risk transmission intensity value is visualized and enhanced to obtain an environmental risk characteristic map of the ocean.

[0099] The step of quantifying and tracking the transmission paths in the refined risk topology to obtain the real-time dynamic risk transmission intensity of the transmission paths includes:

[0100] The node sequence and connecting edges of the transmission path are serialized and extracted to obtain the path composition data of the ocean;

[0101] The effectiveness of the node sequence and the connecting edges is evaluated to obtain the node influence degree and edge influence degree of the ocean.

[0102] Based on the node influence, the edge influence, and the time difference corresponding to the path composition data, the real-time dynamic risk transmission intensity of the transmission path is calculated, wherein the calculation formula for the real-time dynamic risk transmission intensity is as follows:

[0103] ;

[0104] in, This represents the real-time dynamic risk transmission intensity along the transmission path. For the first The influence degree of each node. For the first The influence degree of the edge mentioned in the article, The time difference is... The preset time decay constant, For the summation function, It is a logarithmic function. It is a natural exponential function;

[0105] Based on the real-time dynamic risk transmission intensity, the spatiotemporal topology layer is integrated and rendered to obtain the environmental risk feature map of the ocean.

[0106] This process involves meticulously analyzing historical marine risk events, including their exact timing, location, type, and severity. It also clarifies the specific attributes and manifestations of risk-related characteristics. By comparing the timing of each historical risk event with the occurrence of its associated characteristics, a temporal correspondence is ensured. Furthermore, the location of each historical risk event is precisely matched with the corresponding marine region, allowing each event to identify risk-related characteristics that perfectly align with its temporal and spatial dimensions. Ultimately, this yields a comprehensive understanding of the correspondence between marine event characteristics.

[0107] Based on the correspondence of event characteristics, each historical risk event is taken as a core node, and the risk association characteristics that match the event in time and space are taken as the association attributes of the node. According to the order of events and the closeness of the association of characteristics, the initial connection relationship between nodes is established, the association logic and transmission direction between different nodes are clarified, and an initial risk topology network of the ocean that can reflect the preliminary association structure between marine risk events and associated characteristics is formed.

[0108] A comprehensive investigation of nodes and connecting edges in the initial risk topology network was conducted to identify duplicate nodes and connecting edges with identical functions. For duplicate nodes, only one core node was retained and all its associated information was integrated. Connecting edges with redundant functions were removed directly. At the same time, the existence of invalid structures such as closed loops in the network was checked. Invalid loops were eliminated by cutting off non-critical connecting edges in closed loops. Finally, a refined risk topology structure for the ocean was obtained, which is simple in structure, accurate in information, and free of redundancy.

[0109] The specific criteria for dividing the geographic spatiotemporal coordinate system of the ocean are clearly defined, including the precision of latitude and longitude division and the setting of time scale. Each node in the refined risk topology is accurately located to its corresponding position in the geographic spatiotemporal coordinate system according to its corresponding marine region latitude and longitude coordinates. At the same time, the time scale in the coordinate system is matched according to the time attributes of the node association characteristics, so that the spatial position and time attributes of the refined risk topology are completely consistent with the geographic spatiotemporal coordinate system, thereby obtaining the spatiotemporal topology layer of the ocean.

[0110] For the clearly defined transmission paths in the refined risk topology, the node arrangement order in each path is extracted sequentially according to the extension order of the path. At the same time, the specific distribution of the connecting edges between the nodes is recorded to ensure the integrity of the node sequence and the accuracy of the connecting edges. The extracted node arrangement order and connecting edge distribution are systematically integrated to form complete ocean path composition data.

[0111] Based on the path composition data, the role and influence of each node in the transmission path are analyzed. By statistically analyzing the number of transmissions a node participates in and the triggering frequency of subsequent nodes, the importance of each node, i.e., the node influence degree, is determined. At the same time, the smoothness and stability of each connecting edge in the information transmission between nodes are examined. Based on the usage frequency and transmission efficiency of the connecting edges, the effect strength of each connecting edge, i.e., the edge influence degree, is judged. Finally, the node influence degree and edge influence degree of the ocean are obtained.

[0112] By combining the order of node sequence occurrence in the path composition data, the time interval between adjacent nodes, i.e., the time difference corresponding to the path composition data, is calculated. The magnitude of node influence, the strength of side influence, and the length of time difference are comprehensively considered. Through intuitive comparative analysis, the comprehensive effect of each factor on risk transmission is determined, thereby obtaining the real-time dynamic risk transmission intensity of the transmission path. Then, the information of real-time dynamic risk transmission intensity is integrated into the spatiotemporal topology layer. The layer is differentiated by color and brightness according to the intensity level. After the integrated rendering is completed, the marine environmental risk feature map is obtained.

[0113] The real-time dynamic risk transmission intensity value is classified and divided into different level ranges according to the intensity value. Each level range is given a unique visualization style, including unique color labels and clear graphic symbols. The real-time dynamic risk transmission intensity value of different intensity levels is accurately matched with the corresponding visualization style, so that the difference in intensity value can be presented in an intuitive visual form, and finally the marine environmental risk characteristic map is obtained.

[0114] The node influence is derived from the effectiveness assessment of the node sequence of the transmission path in the refined risk topology. By analyzing the role and influence of each node in the risk transmission process, counting the number of times a node participates in the transmission and the triggering frequency of subsequent nodes, the importance of each node is determined, and finally the node influence of each node is obtained.

[0115] The edge influence degree originates from the effectiveness evaluation of the connecting edges of the transmission path in the refined risk topology. It examines the smoothness and stability of each connecting edge in the information transmission between nodes. Based on the usage frequency and transmission efficiency of the connecting edges, the strength of the effect of each connecting edge is judged, thereby deriving the edge influence degree corresponding to each connecting edge.

[0116] The time difference is determined based on the path composition data, which is obtained by serializing and extracting the node sequence and connecting edges of the transmission path. Specifically, the time difference is the time interval between the occurrence of adjacent nodes in the path composition data. By recording the occurrence time of each node in the transmission process, the difference between the occurrence times of two adjacent nodes is calculated to obtain the time difference corresponding to the path composition data.

[0117] The preset time decay constant is a fixed value pre-set based on historical data and actual conditions of marine environmental risk transmission. This value comprehensively considers the influence of various factors in the marine environment on the decay rate of risk transmission intensity. After statistical analysis and verification of a large number of historical risk events, a fixed value that can accurately reflect the natural decay law of risk transmission intensity over time has been determined.

[0118] The formula's significance lies in its ability to accurately quantify the real-time dynamic risk transmission intensity of the transmission path by comprehensively considering the synergistic effect of the influence of all nodes and edges in the transmission path, combined with the attenuation effect caused by the time difference.

[0119] The calculation first multiplies the influence degree of each connecting edge with the influence degree of the node connected to that edge pairwise. Then, all the multiplication results are summed to obtain a total. A fixed value of one is added to this total. The logarithm of the result is then calculated, and the ratio of the time difference to the preset time decay constant is calculated and taken as negative. The natural exponent is then calculated for this negative result. Finally, the logarithmic result and the natural exponent result are multiplied together to obtain the real-time dynamic risk transmission intensity of the transmission path, which fully presents the real-time dynamic changes of risk during the transmission process.

[0120] The greater the influence of a node, the greater the result of multiplying the influence of each edge by the influence of the corresponding node. The sum of all multiplication results will also be greater, the logarithmic calculation result will increase accordingly, and the real-time dynamic risk transmission intensity of the final transmission path will be greater.

[0121] The greater the edge influence, the greater the result of multiplying the influence of each edge by the influence of the corresponding node, and the greater the sum of all multiplication results. The logarithmic calculation result increases accordingly, thereby increasing the real-time dynamic risk transmission intensity of the transmission path.

[0122] The larger the time difference, the larger the ratio of the time difference to the preset time decay constant, the smaller the negative value of this ratio, the smaller the natural exponent calculation result, and the smaller the real-time dynamic risk transmission intensity of the final transmission path.

[0123] The larger the preset time decay constant, the smaller the ratio of the time difference to the preset time decay constant, the larger the negative value of the ratio, the larger the natural exponential calculation result, and the greater the real-time dynamic risk transmission intensity of the transmission path.

[0124] The beneficial effect is that, through the above specific implementation process, a marine environmental risk characteristic map can be systematically and accurately constructed. The entire process starts from the spatiotemporal matching of historical risk events and risk correlation characteristics, and through multiple standardized operations such as network reconstruction and redundancy elimination, it realizes the quantitative tracking and visualization of risk transmission paths, ensuring that the map can comprehensively and accurately reflect the distribution characteristics, transmission patterns and real-time dynamic intensity of marine environmental risks, and provide a reliable basis for monitoring, early warning and prevention and control decisions of marine environmental risks.

[0125] S4. Track the trajectory of the environmental risk feature map to obtain the spatiotemporal evolution path of the ocean;

[0126] In this embodiment of the invention, the step of tracking the trajectory of the environmental risk feature map to obtain the spatiotemporal evolution path of the ocean includes:

[0127] Source tracing analysis of the environmental risk characteristic map yields the set of risk transmission origins in the ocean;

[0128] The transmission patterns in the historical risk events and the intensity of the real-time dynamic risk transmission are used as dynamic transmission probability parameters of the risk transmission starting point set.

[0129] Based on the dynamic transmission probability parameters, the transmission state of the risk transmission starting point set is deduced to obtain the preliminary transmission trajectory of the ocean;

[0130] The initial transmission trajectory is subjected to spatiotemporal consistency verification to obtain the optimized trajectory set of the ocean;

[0131] Cluster analysis is performed on the optimized trajectory set to obtain the spatiotemporal evolution path of the ocean.

[0132] The process of performing a spatiotemporal consistency check on the initial propagation trajectory to obtain the optimized trajectory set for the ocean includes:

[0133] The continuity of the preliminary transmission trajectory is verified to obtain the spatially continuous trajectory of the ocean;

[0134] The credibility of the continuous spatial trajectory is evaluated to obtain the overall transmission confidence of the ocean trajectory;

[0135] Based on the comprehensive transmission confidence of the trajectory, trajectory optimization is performed on the spatial continuous trajectory to obtain the optimized trajectory set of the ocean.

[0136] By tracing back all risk transmission paths in the environmental risk characteristic map, searching in reverse along the transmission direction of each path, the initial risk origin location and start time of each path are determined. The origin locations and start times corresponding to these different paths are summarized and sorted, and duplicate origin information is removed to form a set containing all unique risk origin points. This set is the risk transmission origin set of the ocean.

[0137] By analyzing the transmission processes and methods of various risk events recorded in the marine historical risk event database, common and unique transmission patterns of different types of risk events are extracted. At the same time, the real-time dynamic risk transmission intensity of each transmission path, which has been calculated before, is extracted. These transmission patterns and real-time dynamic risk transmission intensity are integrated together as dynamic transmission probability parameters to determine the subsequent transmission possibility of each starting point in the risk transmission starting point set.

[0138] Based on dynamic transmission probability parameters, for each starting point in the risk transmission starting point set, we analyze the probability of risk transmission under the current marine environment conditions according to the historical transmission pattern, as well as the transmission trend corresponding to different real-time dynamic risk transmission intensities. We deduce the changes in the transmission state of the risk at different time nodes and different spatial locations after it starts from the starting point, and record the complete transmission route of each deduced route. These transmission routes together constitute the preliminary transmission trajectory of the ocean.

[0139] Each preliminary transmission trajectory is examined to determine whether the spatial distribution of adjacent time points on the trajectory conforms to the spatial logic of natural risk transmission in the marine environment. That is, whether there is a reasonable geographical connection and environmental transmission condition between adjacent spatial locations. Those broken trajectories with spatial jumps and that do not conform to the natural transmission logic are eliminated, and trajectories that are continuous and smooth in space and conform to the law of risk transmission are retained. These retained trajectories are the continuous spatial trajectories of the ocean.

[0140] Information related to risk transmission corresponding to continuous spatial trajectories is collected, including the types of risks involved in the trajectory, the corresponding real-time dynamic risk transmission intensity, and the number of successful records of similar risk transmission in history. This information is then comprehensively analyzed to assess the rationality and reliability of each continuous spatial trajectory in the actual marine environment. Based on the assessment results, a specific numerical confidence level is given for each trajectory. This numerical confidence level is the overall trajectory transmission confidence level in the ocean.

[0141] A fixed confidence threshold is set, and the overall transmission confidence of each spatial continuous trajectory is compared with the threshold. Spatial continuous trajectories with confidence levels that reach or exceed the threshold are retained, while trajectories with confidence levels below the threshold are removed. The retained trajectories are then organized and summarized to form an optimized set of ocean trajectories with a reasonable structure and high reliability.

[0142] Feature extraction is performed on all trajectories in the optimized trajectory set. The core features of each trajectory are analyzed, such as the direction of transmission, key nodes passed through, and spatial range covered. Based on the similarity of these core features, trajectories with highly similar features are classified into the same category. Each category represents a major risk transmission trend. Trajectories in each category are comprehensively refined to form a unified path that reflects the core transmission law of that category. The unified paths of all categories together constitute the spatiotemporal evolution path of the ocean.

[0143] The beneficial effects are that this implementation process, through systematic source tracing analysis, parameter setting, state simulation, consistency verification, and cluster analysis, fully reconstructs the transmission process of risks in the marine environment. It not only ensures the accuracy of the risk transmission starting point, but also improves the scientific nature of the transmission trajectory simulation through dynamic transmission probability parameters. Spatiotemporal consistency verification further filters out reliable trajectories, and cluster analysis ultimately extracts the core spatiotemporal evolution path. The entire process is logically rigorous and operationally specific, effectively solving the problem that traditional methods are difficult to track the spatiotemporal evolution of risks. It provides accurate and comprehensive basic data for subsequent path simulation and risk assessment, and significantly improves the timeliness and accuracy of marine environmental risk identification.

[0144] S5. Perform path simulation on the spatiotemporal evolution path to obtain a risk evolution trajectory report of the ocean;

[0145] In this embodiment of the invention, the step of performing path simulation on the spatiotemporal evolution path to obtain the risk evolution trajectory report of the ocean includes:

[0146] The starting point of the main transmission path, key turning points and transmission direction sequence are extracted from the spatiotemporal evolution path to obtain the path skeleton data of the ocean;

[0147] The path skeleton data is parameterized and encapsulated to obtain a parameterized path framework for the ocean.

[0148] Based on the parameterized path framework, the main transmission path is coupled and simulated to obtain a set of simulated trajectories of the ocean;

[0149] Based on the set of simulated trajectories, a risk evolution trajectory report for the ocean is generated.

[0150] Based on the parameterized path model, a set of simulated trajectories of the ocean is obtained by performing random perturbation superposition simulation on the main transmission path, including:

[0151] Based on the parameterized path framework, the node sequence of the main transmission path is constructed in a benchmark manner to obtain the benchmark transmission trajectory of the ocean.

[0152] Based on the parameterized path framework, the baseline transmission trajectory is subjected to perturbation superposition processing to obtain the perturbed transmission trajectory of the ocean.

[0153] The reference conduction trajectory and the disturbed conduction trajectory are spatiotemporally aligned to obtain a set of simulated trajectories of the ocean.

[0154] By deeply analyzing the established spatiotemporal evolution paths of the ocean, the dominant main transmission path that runs through the entire risk transmission process is identified. The starting point of this main transmission path is precisely located. Nodes that change the direction of risk transmission and play a key role in the trajectory are selected as key turning points. The changes in the direction of risk transmission along the main transmission path are recorded in chronological order to form a transmission direction sequence. The starting point of the main transmission path, key turning points, and transmission direction sequence are summarized and integrated to finally obtain the ocean's path skeleton data.

[0155] The core elements in the path skeleton data are decomposed and sorted out to clarify the specific attributes of the starting point of the main transmission path, the characteristic parameters of key turning nodes, and the expression specifications of the transmission direction sequence. These elements are organized and encapsulated according to a unified format standard to construct a standardized framework that includes the basic structure of the path, key node information, and transmission direction rules. This framework is the parameterized path framework of the ocean.

[0156] Based on the parameterized path framework, the characteristics of the main transmission path are matched with the standard parameters in the framework to simulate various relevant factors in the marine environment that may affect risk transmission, such as changes in ocean currents and weather conditions. These factors are coupled with the transmission laws of the main transmission path, and multiple risk transmission trajectories under different scenarios are generated through step-by-step deduction. These trajectories together constitute a set of simulated ocean trajectories.

[0157] Based on the explicit node sequence rules in the parameterized path framework, and taking the actual node distribution of the main transmission path as a blueprint, a standard trajectory is constructed according to the order and connection relationship of the nodes. This trajectory can fully reflect the core characteristics of the main transmission path and is not affected by other interference factors. This trajectory is the benchmark transmission trajectory of the ocean.

[0158] Within the constraints of the parameterized path framework, random variations that conform to the actual fluctuation range of the marine environment are introduced for key attributes such as the node position and transmission speed of the baseline transmission trajectory. The baseline transmission trajectory is then fine-tuned so that reasonable deviations are generated while maintaining the core transmission trend. After such perturbation superposition processing, multiple new trajectories with slight differences from the baseline transmission trajectory are formed. These new trajectories are the disturbed transmission trajectories of the ocean.

[0159] By selecting a unified time node and spatial coordinate reference system, spatiotemporal positioning calibration is performed on the baseline transmission trajectory and each node on each disturbed transmission trajectory to ensure that all trajectories maintain a consistent reference standard in the same time and spatial dimensions, eliminating deviations between different trajectories caused by different spatiotemporal references. The calibrated baseline transmission trajectory and all disturbed transmission trajectories are then integrated to form a set of simulated ocean trajectories.

[0160] Each trajectory in the simulated trajectory set is analyzed in detail, and core information such as the transmission starting point, key nodes passed through, transmission direction, scope of influence, and evolution trend of each trajectory is recorded. This information is systematically organized and summarized according to a unified report format, clearly presenting the complete process and key characteristics of risk evolution under different simulation scenarios, and finally generating a risk evolution trajectory report for the ocean.

[0161] The beneficial effects are that this implementation process, through a series of coherent operations from path skeleton extraction to final report generation, constructs a standardized parametric path framework based on spatiotemporal evolution paths. By constructing and integrating baseline and disturbed transmission trajectories, it comprehensively covers multiple possible scenarios of risk evolution. The generated set of simulated trajectories is comprehensive and reasonable. The risk evolution trajectory report formed based on this set can clearly and in detail present various key information on risk evolution, providing rich and reliable basis for subsequent comprehensive assessment. It effectively improves the scientificity and accuracy of path simulation in marine environmental risk identification and makes up for the limitations of traditional methods in risk evolution prediction.

[0162] S6. Conduct a comprehensive analysis of the risk evolution trajectory report to obtain the risk level and early warning information of the ocean.

[0163] In this embodiment of the invention, the step of comprehensively analyzing the risk evolution trajectory report to obtain the risk level and early warning information of the ocean includes:

[0164] The trend values, impact range, and path characteristics of the trajectory in the risk evolution trajectory report are fused in multiple dimensions to obtain the structured analysis input of the ocean;

[0165] Based on the structured analysis input, the historical risk case database of the ocean is matched and retrieved to obtain the case matching results of the ocean;

[0166] Based on the case matching results, the risk evolution trajectory report is quantitatively evaluated to obtain the comprehensive risk index of the risk evolution trajectory report;

[0167] The risk index is mapped to a level to obtain the risk level of the ocean;

[0168] By integrating the risk level and the structured analysis input, early warning information for the ocean is obtained.

[0169] By deeply analyzing the risk evolution trajectory report, the trend value of each trajectory is extracted. This trend value reflects the strength and trend of risk changes over time. The marine area covered by each trajectory, i.e. the scope of influence, is clarified. The unique attributes such as key nodes and transmission direction of the trajectory, i.e. path characteristics, are sorted out. These trend values, scope of influence, and path characteristics extracted from different dimensions are organized according to a unified classification standard. Duplicate information is removed and related information is added to form a structured and comprehensive marine analysis input that can be directly used for subsequent assessments.

[0170] A historical marine risk case database was established, storing complete information on all past marine environmental risk events, including key information such as the trend characteristics, affected areas, and transmission paths of the events. Using structured analysis input as the retrieval basis, each piece of information in the structured analysis input was compared with the corresponding information of each case in the historical risk case database. The database identified several historical cases with the highest similarity to the structured analysis input in terms of trend values, scope of impact, and path characteristics. These selected historical cases and their related matching details were then compiled and summarized to obtain the marine case matching results.

[0171] For each historical case in the case matching results, we analyze its correlation with the current risk evolution trajectory report in terms of risk severity, scope of impact, and transmission speed. Based on the correlation, we assign corresponding weights to each matched case. Combining the historical risk level and actual impact of each matched case, we conduct a comprehensive quantitative assessment of the current risk evolution trajectory report. Through comprehensive calculation, we obtain a specific value that can objectively reflect the overall level of current risk. This value is the comprehensive risk index of the risk evolution trajectory report.

[0172] A clear risk level classification standard is pre-defined, dividing the numerical range of the comprehensive risk index into several intervals. Each interval corresponds to a fixed risk level, such as low risk, medium risk, and high risk. The calculated comprehensive risk index is compared with these pre-defined intervals to determine the specific interval to which the comprehensive risk index belongs, thereby obtaining the corresponding risk level of the ocean.

[0173] Collect all key information from the structured analysis input, including risk trends, affected marine areas, and specific transmission paths. Integrate this information with the established marine risk levels, and clearly and explicitly describe the current marine environmental risk level, potential impacts, risk development trends, and necessary countermeasures according to the standardized format of early warning information, ultimately forming marine early warning information.

[0174] The beneficial effects are that this implementation process, through the coherent operation of multi-dimensional integration, case matching retrieval, quantitative assessment, level mapping, and information integration, achieves a comprehensive and in-depth analysis of risk evolution trajectory reports. It not only makes full use of the experience data of historical risk cases, but also combines the specific characteristics of current risks, ensuring the objectivity of the comprehensive risk index and the accuracy of the risk level. The generated early warning information contains rich key information and has strong pertinence and practicality. It effectively solves the problems of insufficient timeliness and pertinence of early warning information in traditional methods, provides reliable decision support for marine environmental risk management, and significantly improves the scientificity and effectiveness of risk response.

[0175] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0176] This application embodiment can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0177] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for identifying marine environmental risks based on big data, characterized in that, The method includes: S1. Standardize the marine environmental monitoring data to obtain a standardized dataset of the ocean. S2. Based on a preset risk knowledge base, perform correlation analysis on the standardized dataset to obtain the risk correlation characteristics of the ocean; S3. Based on the historical risk events of the ocean, perform topological construction on the risk correlation features to obtain the environmental risk feature map of the ocean; S4. Track the trajectory of the environmental risk feature map to obtain the spatiotemporal evolution path of the ocean; S5. Perform path simulation on the spatiotemporal evolution path to obtain a risk evolution trajectory report of the ocean; S6. Conduct a comprehensive analysis of the risk evolution trajectory report to obtain the risk level and early warning information of the ocean.

2. The marine environmental risk identification method based on big data as described in claim 1, characterized in that, The standardization process of the marine environmental monitoring data to obtain the standardized dataset of the ocean includes: Receive marine environmental monitoring data streams; The environmental monitoring data of the ocean is cleaned in real time to obtain preliminary purification data of the ocean; Spatiotemporal alignment is performed on the preliminary cleanup data of the ocean to obtain intermediate standardized data of the ocean; Multimodal fusion is performed on the intermediate standard data to obtain the standard dataset of the ocean.

3. The marine environmental risk identification method based on big data as described in claim 1, characterized in that, The method, based on a pre-defined risk knowledge base, performs correlation analysis on the standardized dataset to obtain the risk correlation characteristics of the ocean, including: The standardized dataset is mapped to a preset risk knowledge base to obtain the mapping relationship of the ocean; Based on the mapping relationship, the risk knowledge base is searched to obtain the initial risk association features of the ocean; The confidence level of the initial risk association features is verified to obtain the refining risk association features of the ocean. The initial risk association features and the refined risk association features are fused to obtain the risk association features of the ocean.

4. The marine environmental risk identification method based on big data as described in claim 1, characterized in that, The method of constructing a topology of the risk association features based on historical risk events of the ocean to obtain an environmental risk feature map of the ocean includes: By performing spatiotemporal matching between the historical risk events of the ocean and the risk-related features, the correspondence between the event features of the ocean can be obtained. The network is reconstructed based on the correspondence of the event features to obtain the initial risk topology network of the ocean; Redundancy elimination is performed on the initial risk topology network to obtain the refined risk topology structure of the ocean; The refined risk topology is mapped to the geographic spatiotemporal coordinate system of the ocean to obtain the spatiotemporal topology layer of the ocean; The transmission path in the refined risk topology is quantitatively tracked to obtain the real-time dynamic risk transmission intensity of the transmission path; The real-time dynamic risk transmission intensity value is visualized and enhanced to obtain an environmental risk characteristic map of the ocean.

5. The marine environmental risk identification method based on big data as described in claim 4, characterized in that, The step of quantifying and tracking the transmission paths in the refined risk topology to obtain the real-time dynamic risk transmission intensity of the transmission paths includes: The node sequence and connecting edges of the transmission path are serialized and extracted to obtain the path composition data of the ocean; The effectiveness of the node sequence and the connecting edges is evaluated to obtain the node influence degree and edge influence degree of the ocean. Based on the node influence, the edge influence, and the time difference corresponding to the path composition data, the real-time dynamic risk transmission intensity of the transmission path is calculated, wherein the calculation formula for the real-time dynamic risk transmission intensity is as follows: ； in, This represents the real-time dynamic risk transmission intensity along the transmission path. For the first The influence degree of each node. For the first The influence degree of the edge mentioned in the article, The time difference is... The preset time decay constant, For the summation function, It is a logarithmic function. It is a natural exponential function; Based on the real-time dynamic risk transmission intensity, the spatiotemporal topology layer is integrated and rendered to obtain the environmental risk feature map of the ocean.

6. The marine environmental risk identification method based on big data as described in claim 1, characterized in that, The process of tracking the environmental risk feature map to obtain the spatiotemporal evolution path of the ocean includes: Source tracing analysis of the environmental risk characteristic map yields the set of risk transmission origins in the ocean; The transmission patterns in the historical risk events and the intensity of the real-time dynamic risk transmission are used as dynamic transmission probability parameters of the risk transmission starting point set. Based on the dynamic transmission probability parameters, the transmission state of the risk transmission starting point set is deduced to obtain the preliminary transmission trajectory of the ocean; The initial transmission trajectory is subjected to spatiotemporal consistency verification to obtain the optimized trajectory set of the ocean; Cluster analysis is performed on the optimized trajectory set to obtain the spatiotemporal evolution path of the ocean.

7. The marine environmental risk identification method based on big data as described in claim 6, characterized in that, The process of performing a spatiotemporal consistency check on the initial propagation trajectory to obtain the optimized trajectory set for the ocean includes: The continuity of the preliminary transmission trajectory is verified to obtain the spatially continuous trajectory of the ocean; The credibility of the continuous spatial trajectory is evaluated to obtain the overall transmission confidence of the ocean trajectory; Based on the comprehensive transmission confidence of the trajectory, trajectory optimization is performed on the spatial continuous trajectory to obtain the optimized trajectory set of the ocean.

8. The marine environmental risk identification method based on big data as described in claim 1, characterized in that, The process of simulating the spatiotemporal evolution path to obtain the risk evolution trajectory report of the ocean includes: The starting point of the main transmission path, key turning points and transmission direction sequence are extracted from the spatiotemporal evolution path to obtain the path skeleton data of the ocean; The path skeleton data is parameterized and encapsulated to obtain a parameterized path framework for the ocean. Based on the parameterized path framework, the main transmission path is coupled and simulated to obtain a set of simulated trajectories of the ocean; Based on the set of simulated trajectories, a risk evolution trajectory report for the ocean is generated.

9. The marine environmental risk identification method based on big data as described in claim 8, characterized in that, Based on the parameterized path model, a set of simulated trajectories of the ocean is obtained by performing random perturbation superposition simulation on the main transmission path, including: Based on the parameterized path framework, the node sequence of the main transmission path is constructed in a benchmark manner to obtain the benchmark transmission trajectory of the ocean. Based on the parameterized path framework, the baseline transmission trajectory is subjected to perturbation superposition processing to obtain the perturbed transmission trajectory of the ocean. The reference conduction trajectory and the disturbed conduction trajectory are spatiotemporally aligned to obtain a set of simulated trajectories of the ocean.

10. The marine environmental risk identification method based on big data as described in claim 1, characterized in that, The comprehensive analysis of the risk evolution trajectory report yields the risk level and early warning information for the ocean, including: The trend values, impact range, and path characteristics of the trajectory in the risk evolution trajectory report are fused in multiple dimensions to obtain the structured analysis input of the ocean; Based on the structured analysis input, the historical risk case database of the ocean is matched and retrieved to obtain the case matching results of the ocean; Based on the case matching results, the risk evolution trajectory report is quantitatively evaluated to obtain the comprehensive risk index of the risk evolution trajectory report; The risk index is mapped to a level to obtain the risk level of the ocean; By integrating the risk level and the structured analysis input, early warning information for the ocean is obtained.

Images