Marine ecological early warning and forecasting method based on multi-source data fusion
By using multi-source data fusion and mechanism-data driven techniques, the data and model problems in marine ecological early warning and forecasting have been solved, achieving fully automated and efficient early warning and forecasting, and improving the accuracy and timeliness of early warnings.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- POWERCHINA HUADONG ENG CORP LTD
- Filing Date
- 2026-03-20
- Publication Date
- 2026-06-23
AI Technical Summary
Existing marine ecological early warning and forecasting technologies suffer from problems such as single data sources, insufficient coverage, limited model accuracy, poor timeliness, difficulty in data fusion, and low level of intelligence, making it difficult to meet the requirements for accuracy and timeliness of early warnings.
By acquiring multi-source heterogeneous marine environmental data, performing standardized processing, intelligent feature extraction, and feature fusion, and combining data assimilation optimization with a mechanism- and data-driven fusion forecasting model, we can achieve fully automated early warning throughout the entire process.
It has achieved comprehensive acquisition and deep integration of marine ecological data, improved the accuracy, timeliness and intelligence of early warning and forecasting, and enabled rapid response to rapidly evolving ecological disasters.
Smart Images
Figure CN122264203A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to a method for marine ecological early warning and forecasting through multi-source data fusion. It is applicable to the interdisciplinary field of marine environmental monitoring and information technology. Background Technology
[0002] Current marine ecological early warning and forecasting technologies mainly suffer from the following four problems and shortcomings: First, at the data level, there are shortcomings such as limited data sources and insufficient coverage. Most existing technologies rely on a single or a few data sources, such as satellite remote sensing data or scattered on-site monitoring data. Satellite remote sensing is easily affected by cloud cover and weather conditions, has monitoring blind spots, and struggles to acquire underwater parameters; while on-site monitoring, although highly accurate, is costly and has limited coverage, making it impossible to comprehensively and promptly capture the full picture and development process of ecological disasters, resulting in a one-sided early warning perspective.
[0003] Secondly, at the model level, there are problems with limited accuracy and poor timeliness. Existing forecast models are mostly based on historical statistical patterns or simplified physical-ecological models, failing to fully utilize real-time / near-real-time multi-dimensional heterogeneous data, and have limited ability to characterize complex nonlinear processes. At the same time, traditional mechanistic models or single models often lack effective data assimilation mechanisms, and cannot dynamically correct the model trajectory through observational data, resulting in errors accumulating over time. Their generalization ability and adaptability are significantly insufficient when facing sudden and new disasters.
[0004] Third, there is a prominent problem of "information silos" at the data fusion level. Marine environmental data comes from different management departments, observation platforms, and sensors, with varying data formats, standards, and spatiotemporal resolutions, lacking a unified and efficient multi-source heterogeneous data fusion framework and technical standards. This situation leads to a significant expenditure of human and material resources on data preprocessing and integration, and data barriers between departments and systems prevent the data value from being maximized in collaborative analysis, severely restricting the overall improvement of early warning capabilities.
[0005] Finally, at the system level, there are shortcomings such as low intelligence and delayed early warning response. Existing technologies lack sufficient automation and intelligence in data processing, feature analysis, and early warning dissemination. Feature extraction largely relies on manual experience or traditional algorithms, making it difficult to automatically mine deep, key features from massive amounts of high-dimensional data. Excessive manual intervention in the early warning process results in an overly long chain from data acquisition to early warning information dissemination, leading to slow response times and failing to meet the practical needs for rapid early warning of rapidly evolving ecological disasters.
[0006] These problems make it difficult for existing technologies to meet the growing application demands in terms of accuracy, timeliness, and reliability of marine ecological early warning and forecasting. Summary of the Invention
[0007] The technical problem to be solved by the present invention is to provide a method for marine ecological early warning and forecasting based on multi-source data fusion, in order to address the above-mentioned problems.
[0008] The technical solution adopted in this invention is: a marine ecological early warning and forecasting method based on multi-source data fusion, comprising: Acquire multi-source heterogeneous marine environmental data for the target sea area, including marine environmental data from space-based, air-based, sea-based, and shore-based monitoring and social perception. Multi-source heterogeneous marine environmental data are sequentially standardized, intelligently extracted, and fused to obtain a fused data cube. Based on the multi-source heterogeneous marine environmental data, the marine ecological dynamics model is optimized through data assimilation. Using a fused data cube and an assimilated and optimized marine ecological dynamics model as input, the system performs calculations through a fusion forecasting model that couples mechanism and data, outputting forecast results for marine ecological disasters. Based on a warning indicator library, the forecast results are matched into graded warning information and released.
[0009] Through the aforementioned technical means, we can achieve comprehensive acquisition, in-depth integration, and model optimization of marine ecological data in the target sea area, construct a forecasting system driven by both mechanism and data, and complete the closed-loop operation of the entire process from data collection to early warning release. We can break through the limitations of a single data source from the source, solve the problems of poor integration of multi-source data and accumulation of model errors, realize the accurate output and automated release of early warning and forecast results, and improve the systematicness and practicality of marine ecological early warning and forecasting as a whole.
[0010] As a preferred embodiment, the multi-source heterogeneous marine environmental data includes: Marine environmental data collected by satellite remote sensing modules, including chlorophyll a concentration and sea surface temperature data in the target sea area; Marine environmental data collected through the on-site monitoring module includes profile data of water temperature, nutrients, and chlorophyll a in the target sea area; Marine environmental data predicted by ocean dynamic models, including current field and temperature field data of the target sea area; Marine environmental data collected through the social perception module includes text and image information related to the marine ecology of the target sea area on the Internet platform.
[0011] As a preferred option, the standardization process includes: unifying the spatiotemporal grid and coordinate system of multi-source heterogeneous data, performing quality control processing such as data calibration, denoising, and completion, and realizing the normalization of multi-source heterogeneous data formats and standards.
[0012] By employing the aforementioned technical means, barriers between data from different sources, in different formats, and with different spatiotemporal resolutions are eliminated, the compatibility issues of multi-source heterogeneous data are resolved, the quality of the original data is optimized, and the consistency and effectiveness of the data are improved. This provides a standardized, high-quality data source for subsequent intelligent feature extraction, feature fusion, and data assimilation, avoids fusion errors caused by chaotic data formats, and improves the efficiency of the entire data processing workflow.
[0013] As a preferred embodiment, the intelligent feature extraction includes: using a convolutional neural network (CNN) to extract deep spatial features from image-type spatial data, and using a long short-term memory network (LSTM) to extract dynamic change features from monitoring time-series data.
[0014] Through the above-mentioned technical means, we can automatically mine deep key features in massive high-dimensional marine environmental data, accurately capture the spatial distribution patterns of image-type spatial data and the temporal evolution patterns of monitoring data, and extract features that are more in line with the actual changes in marine ecosystems. This provides a high-value feature foundation for feature fusion to generate fused data cubes and improves the depth and effectiveness of multi-source data feature fusion.
[0015] As a preferred embodiment, the data assimilation optimization employs an ensemble Kalman filter algorithm to inject in-situ true observation data into the marine ecological dynamics model, quantify the error distribution between the observation data and the model simulation data, update the model's state variables, dynamically correct the model's initial field and trajectory, eliminate model error accumulation, and obtain an assimilated and optimized marine ecological dynamics model.
[0016] By using the above-mentioned technical means, the true value data measured in the field is deeply integrated with the marine ecological dynamics model, and the operational deviation of the model is dynamically corrected. This fundamentally solves the problem of initial field error accumulation in traditional mechanistic models, making the optimized model more consistent with the actual marine ecological conditions of the target sea area, improving the accuracy of the model in describing the evolution of the marine ecosystem, and providing a high-precision mechanistic basis for subsequent fusion forecast models.
[0017] As a preferred embodiment, the mechanism-driven and data-driven coupled fusion forecast model includes: using an assimilated and optimized marine ecological dynamics model as the mechanism-driven part to characterize the physical, chemical, and biological intrinsic evolutionary laws of the marine ecosystem; and using a machine learning algorithm as the data-driven part to learn the nonlinear mapping relationship between the high-dimensional fusion features in the fusion data cube and the probability and evolution law of marine ecological disasters, and to complete the coupling operation under the physical law constraints of the mechanism-driven part.
[0018] Through the above technical means, the deep coupling of the mechanism model and the machine learning algorithm is achieved, so that the forecast model has the dual advantages of physical interpretability and high AI accuracy. The mechanism-driven part defines the physical law boundary for the forecast, avoiding the distortion problem of pure data fitting. The data-driven part makes up for the shortcomings of the mechanism model in characterizing complex nonlinear processes, accurately captures the evolution law of marine ecological disasters, and greatly improves the accuracy and generalization ability of the model for marine ecological disaster forecasting, and achieves effective adaptation to sudden and new disasters.
[0019] As a preferred embodiment, the step of matching the forecast results into graded early warning information based on the early warning indicator library and issuing it includes: automatically matching the forecast results with the preset marine ecological disaster classification judgment standards in the early warning indicator library, identifying the disaster early warning level, and generating standardized early warning information corresponding to the early warning level.
[0020] A marine ecological early warning and forecasting device based on multi-source data fusion, comprising: The data acquisition module is used to acquire multi-source heterogeneous marine environmental data of the target sea area, including marine environmental data from space-based, air-based, sea-based, and shore-based monitoring and social perception. The data fusion module is used to perform standardization, intelligent feature extraction, and feature fusion on multi-source heterogeneous marine environmental data in sequence to obtain a fused data cube. Based on the multi-source heterogeneous marine environmental data, the marine ecological dynamics model is optimized through data assimilation. The intelligent early warning module is used to take the fused data cube and the assimilated and optimized marine ecological dynamics model as input, perform calculations through the fusion forecast model that couples mechanism and data, output the forecast results of marine ecological disasters, and match the forecast results into graded early warning information and release them based on the early warning indicator library.
[0021] A storage medium storing a computer program executable by a processor, wherein the computer program, when executed, implements the steps of the marine ecological early warning and forecasting method based on multi-source data fusion.
[0022] A marine ecological early warning and forecasting device includes a memory and a processor. The memory stores a computer program that can be executed by the processor. When the computer program is executed, it implements the steps of the marine ecological early warning and forecasting method based on multi-source data fusion.
[0023] The beneficial effects of this invention are as follows: By acquiring multi-source heterogeneous marine environmental data from space-based, air-based, sea-based, shore-based monitoring and social perception, this invention achieves three-dimensional and all-round collection of marine ecological data, breaks through the spatiotemporal coverage limitations and monitoring blind spots of a single data source, and makes up for the deficiencies of traditional monitoring in terms of data dimensions and coverage. It can comprehensively and timely capture the full picture and development process of marine ecological disasters, making the early warning perspective more comprehensive.
[0024] This invention achieves data assimilation optimization through ensemble Kalman filtering algorithm, eliminating the error accumulation problem of marine ecological dynamics model. At the same time, it constructs a fusion forecast model that couples mechanism and data-driven approaches, making full use of real-time / near real-time multi-dimensional heterogeneous data, improving the model's ability to characterize complex nonlinear processes, significantly improving the accuracy and timeliness of marine ecological disaster forecasts, and enhancing the model's generalization ability and adaptability to sudden and new disasters.
[0025] This invention constructs a unified and efficient multi-source heterogeneous data fusion framework by standardizing, intelligently extracting and fusing multi-source heterogeneous data. It achieves the normalization and deep fusion of data with different formats, standards and spatiotemporal resolutions, eliminates data barriers between departments and systems, maximizes the collaborative analysis value of multi-source data, reduces the manpower and material resources consumed in data preprocessing and integration, and improves the early warning capability of marine ecology.
[0026] This invention achieves automatic feature extraction through deep learning algorithms and automatic generation of tiered early warning information based on an early warning indicator library. It constructs a fully automated system from data acquisition and processing to early warning release, significantly reducing manual intervention in the early warning process, shortening the chain from data acquisition to early warning information release, improving the automation and intelligence level of the early warning and forecasting process, and realizing rapid early warning of rapidly evolving marine ecological disasters, meeting the needs of practical applications.
[0027] In summary, this invention, through multi-source data fusion and mechanism-data dual-driven technical design, comprehensively solves the core pain points of existing marine ecological early warning and forecasting technologies, significantly improves the accuracy, timeliness, intelligence level and reliability of early warning and forecasting, and provides strong technical support for marine ecological environment protection and disaster prevention and mitigation. Attached Figure Description
[0028] Figure 1 This is a flowchart of the marine ecological early warning and forecasting method in the embodiment. Detailed Implementation
[0029] The embodiments of the present invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention. The step numbers in the following embodiments are set only for ease of explanation, and there is no limitation on the order between the steps. The execution order of each step in the embodiments can be adaptively adjusted according to the understanding of those skilled in the art.
[0030] In the description of this invention, "multiple" means two or more. The use of "first" and "second" is for distinguishing technical features only and should not be construed as indicating or implying relative importance, or implicitly indicating the number of indicated technical features, or the order of the indicated technical features. Furthermore, unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art.
[0031] Example 1: This example is a method for marine ecological early warning and forecasting based on multi-source data fusion, specifically including the following: S100. Acquire multi-source heterogeneous marine environmental data for the target sea area, including marine environmental data from space-based, air-based, sea-based, and shore-based monitoring and social perception.
[0032] In this embodiment, the multi-source heterogeneous marine environmental data specifically includes marine environmental data collected by a satellite remote sensing module, such as chlorophyll a concentration and sea surface temperature data of the target sea area; marine environmental data collected by a field monitoring module, such as water temperature, nutrient salts, and chlorophyll a profile data of the target sea area; marine environmental data predicted by a marine dynamic model, such as current field and temperature field data of the target sea area; and marine environmental data collected by a social perception module, such as text and image information related to the marine ecology of the target sea area on an internet platform.
[0033] The characteristics of space-based, air-based, sea-based, and shore-based monitoring technologies in this example are detailed in the table below: S200. The multi-source heterogeneous marine environmental data are sequentially standardized, intelligently extracted, and fused to obtain a fused data cube (a high-dimensional deep fused feature set). Based on the multi-source heterogeneous marine environmental data, the marine ecological dynamics model is optimized through data assimilation.
[0034] The standardization process in this embodiment includes: unifying the spatiotemporal grid and coordinate system of multi-source heterogeneous data, performing quality control processing such as data calibration, denoising, and completion, and realizing the normalization of multi-source heterogeneous data formats and standards.
[0035] In this example, intelligent feature extraction includes: using a convolutional neural network (CNN) to extract deep spatial features from image-type spatial data, and using a long short-term memory network (LSTM) to extract dynamic change features from monitoring time-series data.
[0036] In this embodiment, feature fusion includes fusing the deep spatial features and dynamic change features to generate a fused data cube containing multi-dimensional related features of the marine ecosystem.
[0037] In this embodiment, the data assimilation optimization adopts the ensemble Kalman filter algorithm, which injects the in-situ true observation data into the physical-chemical-biological coupled three-dimensional marine ecological dynamics model, quantifies the error distribution between the observation data and the model simulation data, updates the model's state variables, dynamically corrects the model's initial field and trajectory, eliminates the accumulation of model errors, and obtains the assimilated and optimized marine ecological dynamics model.
[0038] S300: Taking the fused data cube and the assimilated and optimized marine ecological dynamics model as input, the system performs calculations through a fusion forecast model that couples mechanism and data, outputs forecast results of marine ecological disasters, and matches the forecast results into graded early warning information and releases it based on the early warning indicator library.
[0039] The S310 fusion forecast model uses an assimilated and optimized marine ecological dynamics model as the mechanism-driven part to characterize the physical, chemical, and biological evolutionary laws of the marine ecosystem; and a machine learning algorithm as the data-driven part to learn the nonlinear mapping relationship between the high-dimensional fusion features in the fusion data cube and the probability and evolution law of marine ecological disasters. Under the physical law constraints of the mechanism-driven part, it completes the coupling operation and outputs the forecast results of the probability of disaster occurrence, impact range, and disaster intensity of the target sea area in the next 72 hours.
[0040] S320. The forecast result is automatically matched with the pre-set red tide disaster classification criteria in the early warning indicator database to identify the early warning level of the red tide disaster and generate corresponding standardized early warning information. Finally, the early warning information is released through multiple channels such as websites, mobile text messages, and mobile apps via the early warning release terminal, completing the entire process of early warning and forecasting of red tide disasters.
[0041] Example 2: This example is a marine ecological early warning and forecasting device based on multi-source data fusion, specifically including: The data acquisition module is used to acquire multi-source heterogeneous marine environmental data of the target sea area, including marine environmental data from space-based, air-based, sea-based, and shore-based monitoring and social perception. The data fusion module is used to perform standardization, intelligent feature extraction, and feature fusion on multi-source heterogeneous marine environmental data in sequence to obtain a fused data cube. Based on the multi-source heterogeneous marine environmental data, the marine ecological dynamics model is optimized through data assimilation. The intelligent early warning module is used to take the fused data cube and the assimilated and optimized marine ecological dynamics model as input, perform calculations through the fusion forecast model that couples mechanism and data, output the forecast results of marine ecological disasters, and match the forecast results into graded early warning information and release them based on the early warning indicator library.
[0042] Example 3: This example is a storage medium that stores a computer program that can be executed by a processor. When the computer program is executed, it implements the steps of the marine ecological early warning and forecasting method based on multi-source data fusion described in Example 1.
[0043] Example 4: This example is a marine ecological early warning and forecasting device, which has a memory and a processor. The memory stores a computer program that can be executed by the processor. When the computer program is executed, it implements the steps of the marine ecological early warning and forecasting method of multi-source data fusion described in Example 1.
[0044] Furthermore, although the invention has been described in the context of functional modules, it should be understood that, unless otherwise stated, one or more of the aforementioned functions and / or features may be integrated into a single physical device and / or software module, or one or more functions and / or features may be implemented in a separate physical device or software module. It is also understood that a detailed discussion of the actual implementation of each module is unnecessary for understanding the invention. Rather, given the properties, functions, and internal relationships of the various functional modules in the apparatus disclosed herein, the actual implementation of the module will be understood within the scope of conventional skill of an engineer. Therefore, those skilled in the art can implement the invention as set forth in the claims using ordinary techniques without excessive experimentation. It is also understood that the specific concepts disclosed are merely illustrative and not intended to limit the scope of the invention, which is determined by the full scope of the appended claims and their equivalents.
[0045] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.
[0046] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-including system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.
[0047] More specific examples (a non-exhaustive list) of computer-readable media include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which the aforementioned program can be printed, because the aforementioned program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.
[0048] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.
[0049] In the foregoing description of this specification, references to terms such as "one embodiment," "another embodiment," or "some embodiments" indicate that a specific feature, structure, material, or characteristic described in connection with an embodiment or example is included in at least one embodiment or example of the present invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.
[0050] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.
[0051] The above is a detailed description of the preferred embodiments of the present invention. However, the present invention is not limited to the above embodiments. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the present invention. All such equivalent modifications or substitutions are included within the scope defined by the claims of this application.
Claims
1. A method for marine ecological early warning and forecasting through multi-source data fusion, characterized in that, include: Acquire multi-source heterogeneous marine environmental data for the target sea area, including marine environmental data from space-based, air-based, sea-based, and shore-based monitoring and social perception. Multi-source heterogeneous marine environmental data are sequentially standardized, intelligently extracted, and fused to obtain a fused data cube. Based on the multi-source heterogeneous marine environmental data, the marine ecological dynamics model is optimized through data assimilation. Using a fused data cube and an assimilated and optimized marine ecological dynamics model as input, the system performs calculations through a fusion forecasting model that couples mechanism and data, outputting forecast results for marine ecological disasters. Based on a warning indicator library, the forecast results are matched into graded warning information and released.
2. The marine ecological early warning and forecasting method based on multi-source data fusion according to claim 1, characterized in that: The multi-source heterogeneous marine environmental data includes: Marine environmental data collected by satellite remote sensing modules, including chlorophyll a concentration and sea surface temperature data of the target sea area; Marine environmental data collected through the on-site monitoring module includes profile data of water temperature, nutrients, and chlorophyll a in the target sea area; Marine environmental data predicted by ocean dynamic models, including current field and temperature field data of the target sea area; Marine environmental data collected through the social perception module includes text and image information related to the marine ecology of the target sea area on the Internet platform.
3. The marine ecological early warning and forecasting method based on multi-source data fusion according to claim 1, characterized in that, The standardization process includes: unifying the spatiotemporal grid and coordinate system of multi-source heterogeneous data, performing quality control processing such as data calibration, denoising, and completion, and realizing the normalization of multi-source heterogeneous data formats and standards.
4. The marine ecological early warning and forecasting method based on multi-source data fusion according to claim 1, characterized in that, The intelligent feature extraction includes: using a convolutional neural network (CNN) to extract deep spatial features from image-type spatial data, and using a long short-term memory network (LSTM) to extract dynamic change features from monitoring time-series data.
5. The marine ecological early warning and forecasting method based on multi-source data fusion according to claim 1, characterized in that, The data assimilation optimization employs an ensemble Kalman filter algorithm, which injects in-situ true observation data into the marine ecological dynamics model, quantifies the error distribution between the observation data and the model simulation data, updates the model's state variables, dynamically corrects the model's initial field and trajectory, eliminates the accumulation of model errors, and obtains an assimilated and optimized marine ecological dynamics model.
6. The marine ecological early warning and forecasting method based on multi-source data fusion according to claim 1, characterized in that, The mechanism-driven and data-driven coupled fusion forecast model includes: using an assimilated and optimized marine ecological dynamics model as the mechanism-driven part to characterize the physical, chemical, and biological intrinsic evolutionary laws of the marine ecosystem; and using a machine learning algorithm as the data-driven part to learn the nonlinear mapping relationship between the high-dimensional fusion features in the fusion data cube and the probability and evolution law of marine ecological disasters, and to complete the coupling operation under the physical law constraints of the mechanism-driven part.
7. The marine ecological early warning and forecasting method based on multi-source data fusion according to claim 1, characterized in that, The step of matching the forecast results into graded early warning information based on the early warning indicator library and issuing it includes: automatically matching the forecast results with the preset marine ecological disaster classification judgment standards in the early warning indicator library, identifying the disaster early warning level, and generating standardized early warning information corresponding to the early warning level.
8. A marine ecological early warning and forecasting device based on multi-source data fusion, characterized in that, include: The data acquisition module is used to acquire multi-source heterogeneous marine environmental data of the target sea area, including marine environmental data from space-based, air-based, sea-based, and shore-based monitoring and social perception. The data fusion module is used to perform standardization, intelligent feature extraction, and feature fusion on multi-source heterogeneous marine environmental data in sequence to obtain a fused data cube. Based on the multi-source heterogeneous marine environmental data, the marine ecological dynamics model is optimized through data assimilation. The intelligent early warning module is used to take the fused data cube and the assimilated and optimized marine ecological dynamics model as input, perform calculations through the fusion forecast model that couples mechanism and data, output the forecast results of marine ecological disasters, and match the forecast results into graded early warning information and release them based on the early warning indicator library.
9. A storage medium having a computer program stored thereon that can be executed by a processor, characterized in that, When the computer program is executed, it implements the steps of the marine ecological early warning and forecasting method based on multi-source data fusion as described in any one of claims 1 to 7.
10. A marine ecological early warning and forecasting device, comprising a memory and a processor, wherein the memory stores a computer program executable by the processor, characterized in that, When the computer program is executed, it implements the steps of the marine ecological early warning and forecasting method based on multi-source data fusion as described in any one of claims 1 to 7.