Residual circulation monitoring method, system and device and storage medium

Abstract

The invention discloses a residual circulation monitoring method, system and device and a storage medium, and the method comprises the steps: obtaining basic sea area data and multi-source environment data of a target sea area, and building an FVCOM model according to the basic sea area data; performing ocean numerical simulation on the target sea area through an FVCOM model by taking the multi-source environment data as a boundary condition to obtain ocean dynamics simulation data of the target sea area; inputting the ocean dynamics simulation data into a pre-trained residual circulation prediction model of a target sea area to obtain residual circulation distribution prediction data of the target sea area; and according to the residual circulation distribution prediction data, updating the boundary condition, and returning to perform ocean numerical simulation on the target sea area through the FVCOM model until a preset simulation frequency is reached, thereby obtaining a residual circulation distribution prediction condition of the target sea area. The method improves the accuracy and efficiency of residual circulation distribution prediction, also improves the accuracy and efficiency of wind field distribution prediction, and can be widely applied to the technical field of ocean engineering.

Description

Technical Field

[0001] The present invention relates to the field of marine engineering technology, and in particular to a residual circulation monitoring method, system, device and storage medium. Background Art

[0002] As one of the main driving forces of ocean circulation, wind field has a particularly significant impact on residual circulation. Wind field changes the velocity and direction of the ocean surface, thereby affecting the temperature and salinity distribution of the ocean, thereby driving the generation and change of residual circulation. Therefore, in-depth research on the response mechanism of residual circulation to wind field is of great theoretical and practical significance for understanding the evolution of ocean circulation, predicting the changing trend of the marine environment, and formulating scientific and reasonable marine management strategies.

[0003] Domestic scholars have made some achievements in studying the impact of wind fields on ocean circulation. For example, some scholars have studied the impact of different types of wind fields such as monsoons and typhoons on the South China Sea circulation through numerical simulation methods, and found that wind fields can significantly change the circulation structure of the South China Sea; the response mechanism of the residual circulation to the wind field has also been studied. For example, some scholars have studied the response mechanism of the Yellow Sea residual circulation to the wind field through observation data and numerical simulation methods, and found that the residual circulation will change under the influence of the wind field, and this change will also affect the wind field in the surrounding sea areas. Foreign scholars have also made some progress in studying the impact of wind fields on ocean circulation. For example, some scholars have analyzed the response of the North Atlantic circulation to the wind field through long-term observation data, and found that the wind field can change the intensity and direction of the North Atlantic circulation; the response mechanism of the residual circulation to the wind field has also been studied. For example, some scholars have studied the response mechanism of the Pacific residual circulation to the wind field through numerical simulation methods.

[0004] At present, some studies have attempted to explore the response mechanism of the residual circulation to the wind field through field observation and data analysis. Among them, the most typical implementation scheme is to use high-frequency ground wave radar to observe the velocity and direction of the ocean surface, and combine meteorological data to analyze the impact of the wind field on the ocean circulation. These schemes usually include the following steps: first, deploy high-frequency ground wave radar stations in the target sea area to obtain real-time data on the velocity and direction of the ocean surface; second, collect meteorological data of the same period, including wind speed, wind direction, etc.; then, extract the key characteristic parameters of the residual circulation and wind field through data processing and analysis; finally, use statistical analysis and physical models to study the response law of the residual circulation to changes in the wind field.

[0005] These implementation schemes have revealed the mechanism of the response of the residual circulation to the wind field to a certain extent, but there are still some limitations. For example, the observation range of high-frequency ground wave radar is limited and may not cover the entire target sea area; at the same time, the acquisition of meteorological data may also be affected by factors such as weather conditions and the distribution of observation sites. In addition, these schemes mainly focus on the statistical relationship between the residual circulation and the wind field, and lack in-depth discussion on the formation and evolution mechanism of the residual circulation. In addition, the existing schemes also have the problems of low efficiency and insufficient stability, and may not be able to accurately capture the circulation changes caused by complex physical processes. Summary of the invention

[0006] The purpose of the present invention is to solve one of the technical problems existing in the prior art to at least a certain extent.

[0007] To this end, an object of an embodiment of the present invention is to provide a residual circulation monitoring method, which improves the accuracy and efficiency of residual circulation distribution prediction, and also improves the accuracy and efficiency of wind field distribution prediction.

[0008] Another object of an embodiment of the present invention is to provide a residual circulating flow monitoring system.

[0009] In order to achieve the above technical objectives, the technical solutions adopted by the embodiments of the present invention include:

[0010] In a first aspect, an embodiment of the present invention provides a residual circulating current monitoring method, comprising the following steps:

[0011] Acquire basic sea area data and multi-source environmental data of the target sea area, and establish an FVCOM model based on the basic sea area data;

[0012] Using the multi-source environmental data as boundary conditions, performing ocean numerical simulation on the target sea area through the FVCOM model, and obtaining ocean dynamics simulation data of the target sea area;

[0013] Inputting the ocean dynamics simulation data into a pre-trained residual circulation prediction model of the target sea area to obtain residual circulation distribution prediction data of the target sea area;

[0014] The boundary conditions are updated according to the residual circulation distribution prediction data, and the FVCOM model is used to perform ocean numerical simulation on the target sea area until a preset number of simulations is reached to obtain the residual circulation distribution prediction of the target sea area.

[0015] Furthermore, in one embodiment of the present invention, the basic sea area data includes seabed topography data, coastline shape data and sea water level data of the target sea area, and the multi-source environmental data includes meteorological data, ocean current data, sea water temperature data and sea water salinity data.

[0016] Furthermore, in one embodiment of the present invention, the residual circulation prediction model is trained by the following steps:

[0017] Acquire first ocean dynamics sample data of multiple historical periods of the target sea area and corresponding residual circulation distribution labels, wherein the residual circulation distribution labels include the distribution of residual circulation velocity, residual circulation direction and residual circulation intensity of the target sea area;

[0018] Constructing a labeled data set according to the first ocean dynamics sample data and the residual circulation distribution label;

[0019] Acquire second ocean dynamics sample data of multiple historical periods of the target sea area, and construct an unlabeled data set according to the second ocean dynamics sample data;

[0020] Semi-supervised learning is performed on a pre-constructed convolutional neural network according to the labeled data set and the unlabeled data set to obtain the trained residual circulation prediction model.

[0021] Further, in one embodiment of the present invention, the semi-supervised learning of the convolutional neural network is performed according to the labeled data set and the unlabeled data set to obtain the trained residual circulation prediction model, which specifically includes:

[0022] Inputting the labeled data set into the convolutional neural network to obtain a trained initial model;

[0023] Predicting the unlabeled data set using the initial model, and selecting a high-confidence prediction result as a pseudo label according to a confidence threshold;

[0024] Annotate the unlabeled data according to the pseudo-label to obtain a pseudo-label data set;

[0025] The initial model is fine-tuned according to the pseudo-label data set and the labeled data set, and the unlabeled data set is predicted by the initial model again until a preset number of model fine-tuning times is reached to obtain the trained residual circulation prediction model.

[0026] Further, in one embodiment of the present invention, the semi-supervised learning of the convolutional neural network is performed according to the labeled data set and the unlabeled data set to obtain the trained residual circulation prediction model, which specifically includes:

[0027] Dividing the labeled data set into a first data subset and a second data subset;

[0028] Inputting the first data subset and the second data subset into the convolutional neural network respectively to obtain a trained first initial model and a second initial model;

[0029] Predicting the unlabeled data set using the first initial model and the second initial model respectively, selecting a prediction result with high confidence as a pseudo label according to a confidence threshold, and obtaining a first pseudo label and a second pseudo label;

[0030] The unlabeled data are labeled according to the first pseudo-label and the second pseudo-label to obtain a first pseudo-label data set and a second pseudo-label data set;

[0031] Fine-tune the first initial model according to the second pseudo-label data set and the first data subset, fine-tune the second initial model according to the first pseudo-label data set and the second data subset, and return to predict the unlabeled data set through the first initial model and the second initial model respectively until a preset number of model fine-tuning times is reached, so as to obtain a trained first cocirculating flow prediction model and a second cocirculating flow prediction model;

[0032] The first residual circulating flow prediction model and the second residual circulating flow prediction model are integrated and learned to obtain the trained residual circulating flow prediction model.

[0033] Furthermore, in one embodiment of the present invention, the step of inputting the labeled data set into the convolutional neural network to obtain a trained initial model specifically includes:

[0034] Dividing the labeled data set into a training set, a validation set, and a test set;

[0035] Inputting the training set into the convolutional neural network to obtain a residual circulation distribution prediction result;

[0036] Determine the loss value according to the residual circulation distribution prediction result and the residual circulation distribution label;

[0037] Updating internal parameters of the convolutional neural network according to the loss value, and inputting the validation set into the updated convolutional neural network to obtain a model performance of the convolutional neural network on the validation set;

[0038] Updating the hyperparameters of the convolutional neural network according to the model performance, and returning the training set to the convolutional neural network until the model performance no longer improves in a plurality of consecutive iterations, thereby obtaining a first initial model;

[0039] Inputting the test set into the first initial model to obtain a model evaluation index of the first initial model;

[0040] When the model evaluation index meets the preset requirements, the first initial model is used as the trained initial model.

[0041] Furthermore, in one embodiment of the present invention, the residual circulating flow monitoring method further comprises the following steps:

[0042] Obtaining historical residual circulation distribution data and historical wind field observation data of the target area;

[0043] Performing regression analysis on the historical residual circulation distribution data and the historical wind field observation data to obtain a data relationship between the residual circulation distribution and the wind field parameters;

[0044] The wind field distribution prediction of the target area is determined according to the residual circulation distribution prediction and the data relationship.

[0045] In a second aspect, an embodiment of the present invention provides a residual circulating flow monitoring system, including:

[0046] A data acquisition module is used to acquire basic sea area data and multi-source environmental data of the target sea area, and establish an FVCOM model based on the basic sea area data;

[0047] An ocean numerical simulation module, used to use the multi-source environmental data as boundary conditions, perform ocean numerical simulation on the target sea area through the FVCOM model, and obtain ocean dynamics simulation data of the target sea area;

[0048] A residual circulation distribution prediction module is used to input the ocean dynamics simulation data into a pre-trained residual circulation prediction model of the target sea area to obtain residual circulation distribution prediction data of the target sea area;

[0049] The boundary condition updating module is used to update the boundary conditions according to the residual circulation distribution prediction data, and return to perform ocean numerical simulation of the target sea area through the FVCOM model until a preset number of simulations is reached to obtain the residual circulation distribution prediction of the target sea area.

[0050] In a third aspect, an embodiment of the present invention provides a residual circulating flow monitoring device, comprising:

[0051] at least one processor;

[0052] at least one memory for storing at least one program;

[0053] When the at least one program is executed by the at least one processor, the at least one processor implements the above-mentioned residual circulating current monitoring method.

[0054] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium, in which a program executable by a processor is stored. When the program executable by the processor is executed by the processor, it is used to execute the above-mentioned residual circulating current monitoring method.

[0055] The advantages and beneficial effects of the present invention will be partly given in the following description, partly become apparent from the following description, or be understood through the practice of the present invention:

[0056] The embodiment of the present invention obtains basic sea area data and multi-source environmental data of the target sea area, establishes an FVCOM model according to the basic sea area data, takes the multi-source environmental data as boundary conditions, performs ocean numerical simulation on the target sea area through the FVCOM model, obtains ocean dynamics simulation data of the target sea area, inputs the ocean dynamics simulation data into a pre-trained residual circulation prediction model of the target sea area, obtains residual circulation distribution prediction data of the target sea area, updates the boundary conditions according to the residual circulation distribution prediction data, and returns to perform ocean numerical simulation on the target sea area through the FVCOM model until a preset number of simulations is reached to obtain the residual circulation distribution prediction of the target sea area. An embodiment of the present invention combines the FVCOM model with a deep neural network, uses the FVCOM model to generate ocean dynamics simulation data as the input of the residual circulation prediction model, updates the boundary conditions of the FVCOM model according to the residual circulation distribution prediction data output by the residual circulation prediction model, and improves the model's prediction ability for the residual circulation distribution field in a data-driven manner, thereby improving the accuracy and efficiency of the residual circulation distribution prediction; in addition, the wind field distribution prediction can be determined based on the obtained residual circulation distribution prediction, thereby improving the accuracy and efficiency of the wind field distribution prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the technical solution in the embodiments of the present invention, the following introduction is made to the drawings required for use in the embodiments of the present invention. It should be understood that the drawings introduced below are only for the convenience of clearly describing some embodiments of the technical solution of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0058] Figure 1 A flowchart of a residual circulating flow monitoring method provided by an embodiment of the present invention;

[0059] Figure 2 A structural block diagram of a residual circulating flow monitoring system provided by an embodiment of the present invention;

[0060] Figure 3 A structural block diagram of a residual circulating flow monitoring device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0061] The embodiments of the present invention are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and are not to be construed as limitations of the present invention. For the step numbers in the following embodiments, they are only provided for the convenience of explanation, and the order between the steps is not limited in any way, and the execution order of each step in the embodiment can be adaptively adjusted according to the understanding of those skilled in the art.

[0062] In the description of the present invention, the meaning of "a plurality" is two or more than two. If there is a description of "a first" or "a second", it is only used for the purpose of distinguishing technical features, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features or implicitly indicating the order of the indicated technical features. In addition, unless otherwise defined, all technical and scientific terms used in this document have the same meaning as those commonly understood by those skilled in the art.

[0063] With the in-depth development of marine engineering, the importance of residual circulation research in the fields of marine engineering, wind power generation, etc. has become increasingly prominent. The FVCOM model has a wide range of applications in the fields of marine science and marine engineering, and can accurately simulate the circulation characteristics of the ocean, including ocean currents, vortices, and boundary flows. The present invention adopts a method combining the FVCOM model with a deep neural network model, and innovatively proposes a new residual circulation monitoring method to achieve accurate prediction of the spatial residual flow field.

[0064] Reference Figure 1 The embodiment of the present invention provides a residual circulating current monitoring method, which specifically includes the following steps:

[0065] S101, obtaining basic sea area data and multi-source environmental data of the target sea area, and establishing an FVCOM model based on the basic sea area data;

[0066] S102, using the multi-source environmental data as boundary conditions, performing ocean numerical simulation on the target sea area through the FVCOM model, and obtaining ocean dynamics simulation data of the target sea area;

[0067] S103, inputting the ocean dynamics simulation data into a pre-trained residual circulation prediction model of the target sea area to obtain residual circulation distribution prediction data of the target sea area;

[0068] S104, updating the boundary conditions according to the residual circulation distribution prediction data, and returning to perform ocean numerical simulation of the target sea area through the FVCOM model until a preset number of simulations is reached to obtain the residual circulation distribution prediction of the target sea area.

[0069] Specifically, FVCOM is an unstructured triangular grid framework, finite volume, free surface, three-dimensional primitive equation ocean numerical model. The primitive equations of FVCOM mainly include momentum equation, mass continuity equation, and temperature, salinity and density equations. In physics and mathematics, the Mellor-Yamada 2.5-order vertical turbulence closure model and Smagorinsky horizontal turbulence closure model are used to close the equation group. In the vertical direction, the general vertical coordinate system is used to fit the irregular bottom terrain, and in the horizontal direction, the unstructured triangular grid is used to spatially discretize the horizontal calculation area. In numerical calculation, the control equations are discretized and solved by performing flux finite volume integration on the horizontal triangular control volume. The finite volume integration method combines the free geometric fitting characteristics of the finite element method and the simple discrete structure and efficient calculation characteristics of the finite difference method, thereby combining the advantages of the two methods. The finite volume integration format can better ensure the conservation of mass, momentum, salinity, temperature and heat in the calculation of estuaries, bays and oceans with complex geometric structures. The advantages of the FVCOM model in numerical processing methods and shoreline topography fitting have made it widely used in estuary and coastal areas.

[0070] FVCOM includes a variety of physical, water quality, and ecological computing modules. The FVCOM model input and output are based on the standardized NETCDF format, which has universal platform compatibility, and can use VISIT visualization software to quickly visualize the input and output structures in 2D / 3D. The model is based on the Fortran 90 / 95 standard and implements computational parallelism under the framework of MPI (Message Passing Interface). It can achieve parallel fast simulation on high-performance computers with shared memory and distributed memory multi-computing nodes.

[0071] The overall idea of ​​the embodiment of the present invention is as follows:

[0072] 1) Use the FVCOM model to generate simulated data of ocean dynamics, such as temperature, salinity, current velocity, etc.; standardize or normalize these data to meet the input requirements of the deep neural network model.

[0073] 2) Select a suitable deep neural network architecture, such as AlexNet, VGGNet or a custom architecture; design network layers, including convolutional layers, pooling layers, fully connected layers, etc. Use historical data sets to train deep neural networks, learn patterns in ocean data, and use back-propagation algorithms to optimize network weights to obtain a residual circulation prediction model.

[0074] 3) Using the output data of the FVCOM model, the spatial features are extracted through the residual circulation prediction model to capture the local features in the ocean data, such as vortices and fronts.

[0075] 4) Use the output of the residual circulation prediction model as the input of the FVCOM model or the basis for adjusting boundary conditions. For example, the residual circulation prediction model can predict the flow velocity changes under specific conditions, and these prediction results can be fed back to the FVCOM model to adjust its boundary conditions.

[0076] 5) Repeatedly use the FVCOM model to generate ocean dynamics simulation data as the input of the residual circulation prediction model, and update the boundary conditions of the FVCOM model according to the residual circulation distribution prediction data output by the residual circulation prediction model. This process is repeated until the preset number of simulations is reached, and the residual circulation distribution prediction can be obtained.

[0077] An embodiment of the present invention combines the FVCOM model with a deep neural network, uses the FVCOM model to generate ocean dynamics simulation data as the input of the residual circulation prediction model, updates the boundary conditions of the FVCOM model according to the residual circulation distribution prediction data output by the residual circulation prediction model, and improves the model's prediction ability for the residual circulation distribution field in a data-driven manner, thereby improving the accuracy and efficiency of the residual circulation distribution prediction; in addition, the wind field distribution prediction can be determined based on the obtained residual circulation distribution prediction, thereby improving the accuracy and efficiency of the wind field distribution prediction.

[0078] It should be noted that the FVCOM model is a well-known model in the prior art. The focus of the embodiment of the present invention is to use the model to obtain the ocean dynamics simulation data of the target sea area. The specific FVCOM model construction and the ocean numerical simulation process are not described in detail, and only the update of its boundary conditions (input data) and the output ocean dynamics simulation data are focused on.

[0079] As a further optional implementation, the basic sea area data includes seabed topography data, coastline shape data and sea water level data of the target sea area, and the multi-source environmental data includes meteorological data, ocean current data, sea water temperature data and sea water salinity data.

[0080] The following is a detailed description of the training process of the residual circulation prediction model according to the embodiment of the present invention.

[0081] First, the relevant design of the residual circulation prediction model of the embodiment of the present invention is introduced.

[0082] 1) Data preprocessing and enhancement

[0083] Standardization / Normalization: Scale the input data to a uniform range (such as 0 to 1 or -1 to 1) to speed up convergence.

[0084] Data enhancement: Increase data diversity and reduce overfitting by rotating, scaling, cropping, flipping, etc.

[0085] The embodiments of the present invention use automated data enhancement technology, such as using a neural network to predict optimal enhancement parameters; and customize data enhancement strategies based on data characteristics of a specific sea area to better simulate changes in the actual marine environment.

[0086] 2) Network architecture design

[0087] Choose a suitable infrastructure: Choose a suitable CNN architecture (such as AlexNet, VGGNet, ResNet, etc.).

[0088] Custom layer design: Custom layers are designed based on the characteristics of ocean dynamics data. Attention mechanisms are introduced to enhance the model's ability to recognize key features. Residual connections are designed to solve the gradient vanishing problem in deep network training to improve the model's ability to recognize key ocean features.

[0089] 3) Hyperparameter selection

[0090] Learning rate: Choose an appropriate learning rate, which can be determined through cross-validation to ensure the speed and stability of the loss function decrease during model training.

[0091] Batch size: determines the number of samples in each batch, which affects the convergence speed and stability of the model.

[0092] Optimizer: Select a suitable optimization algorithm, such as SGD, Adam, RMSprop, etc.

[0093] The embodiment of the present invention uses Bayesian optimization or genetic algorithm to automatically find the optimal hyperparameters, and dynamically adjusts the learning rate and other hyperparameters during the training process to adapt to the learning progress of the model.

[0094] 4) Model training

[0095] Training strategy: Use early stopping (stop training when the performance on the validation set no longer improves) to avoid overfitting.

[0096] Model evaluation: Evaluate model performance on an independent test set and use metrics such as accuracy and recall to measure the model’s predictive ability.

[0097] The embodiment of the present invention trains the model on multiple related tasks simultaneously to improve the generalization ability of the model; and adds uncertainty estimation to the model training to better evaluate the credibility of the model prediction.

[0098] 5) Model optimization

[0099] Transfer learning: Use a pre-trained model for transfer learning, using the features learned on similar tasks to improve the initial performance of the model.

[0100] Hyperparameter tuning: Based on the preliminary training results, further adjust the hyperparameters using methods such as grid search, random search, or Bayesian optimization.

[0101] The embodiment of the present invention uses a meta-learning method to quickly adapt to new tasks and reduce the dependence on a large amount of labeled data; through model fusion technology, such as Bagging or Boosting, the stability and accuracy of the model are improved.

[0102] 6) Model verification and optimization

[0103] Cross-validation: K-fold cross-validation is used to evaluate the stability and generalization ability of the model.

[0104] Performance evaluation: Perform a detailed analysis of the model’s performance on different datasets to identify the model’s strengths and weaknesses.

[0105] 7) Model deployment and application

[0106] Model deployment: Deploy the trained model to the actual ocean monitoring system to ensure that the model can receive real-time data and make predictions.

[0107] Real-time monitoring: Use the model to monitor and predict residual circulation in real time, and update the model in time to adapt to environmental changes.

[0108] Association response mechanism: Establish an association response mechanism for monitoring points, identify abnormal monitoring points and divide them into different abnormal monitoring point groups.

[0109] 8) Result analysis and evaluation

[0110] Characteristic analysis: Based on the model prediction results, analyze the characteristics and movement laws of the residual circulation field, such as speed, direction, temperature and salinity distribution.

[0111] Movement pattern assessment: Assess the movement pattern of the residual circulation and its impact on the wind field.

[0112] Analysis of influencing factors: Explore the factors that affect the characteristics of the residual circulation, such as the marine environment, topography and climate change.

[0113] Risk assessment: Based on the model prediction results, assess the risks that may be brought about by the residual circulation and propose corresponding countermeasures.

[0114] 9) Multi-source data fusion

[0115] Ensemble learning: Ensemble learning methods are used to combine information from different sensors and data sources to improve the accuracy of data.

[0116] Data Fusion Algorithms: Develop new data fusion algorithms that can process and integrate multimodal data from high-frequency ground wave radar, satellite remote sensing, buoys, and meteorological stations.

[0117] 10) Adaptive data preprocessing

[0118] Smart Denoising: Develop smart denoising algorithms that can adaptively identify and remove noise from data without losing important signal features.

[0119] Dynamic Data Cleaning: Implement dynamic data cleaning technology that can identify outliers and missing data based on the real-time characteristics and historical patterns of the data.

[0120] 11) Efficient data compression

[0121] Lossless and Lossless Compression Techniques: Research and apply efficient data compression techniques to reduce storage requirements and increase data transfer speeds while maximizing data integrity and availability.

[0122] Feature selection: Advanced feature selection techniques, such as model-based feature selection, automatically determine which data features are most critical for model training, thereby reducing unnecessary data storage and processing.

[0123] 12) Application of augmented reality (AR) and virtual reality (VR) technologies

[0124] Interactive Data Visualization: Using AR and VR technologies to provide more intuitive and interactive data visualization tools to help researchers gain a deeper understanding of complex ocean data.

[0125] Virtual Environment Simulation: Simulating the ocean environment in a VR environment allows researchers to explore and analyze data in a virtual three-dimensional space to discover potential patterns and associations.

[0126] 13) Using blockchain technology to ensure data integrity

[0127] Data cannot be tampered with: Blockchain technology is used to ensure the integrity and non-tamperability of data during collection, storage and processing, thereby increasing the credibility of the data.

[0128] Decentralized data management: Through the decentralized characteristics of blockchain, distributed storage and management of data can be achieved, improving data security and access efficiency.

[0129] As an optional implementation, the residual circulation prediction model is trained by the following steps:

[0130] S201, obtaining first ocean dynamics sample data of multiple historical periods of the target sea area and corresponding residual circulation distribution labels, the residual circulation distribution labels including the distribution of residual circulation speed, residual circulation direction and residual circulation intensity of the target sea area;

[0131] S202, constructing a labeled data set according to the first ocean dynamics sample data and the residual circulation distribution label;

[0132] S203, obtaining second ocean dynamics sample data of multiple historical periods in the target sea area, and constructing an unlabeled data set according to the second ocean dynamics sample data;

[0133] S204, performing semi-supervised learning on the pre-built convolutional neural network according to the labeled data set and the unlabeled data set to obtain a trained residual circulation prediction model.

[0134] Specifically, the embodiment of the present invention uses a semi-supervised learning method to reduce the workload of data labeling, automatically labels a large amount of unlabeled data through an algorithm, and then has experts review and correct it.

[0135] Using semi-supervised learning can utilize a large amount of unlabeled data to improve model performance and reduce the workload of manual labeling. In the field of marine engineering, it is often expensive and time-consuming to obtain a large amount of accurately labeled data, while unlabeled data such as sensor data is relatively easy to obtain. Among them, the self-training method is suitable for situations where the relationship between features and labels in the data set is relatively clear, while collaborative training is suitable for situations where the data set can be divided by two sets of different features and each set of features has a certain degree of discriminative ability.

[0136] As an optional implementation method, semi-supervised learning is performed on the convolutional neural network according to the labeled data set and the unlabeled data set to obtain a trained residual circulation prediction model, which specifically includes:

[0137] S2040, inputting the labeled data set into the convolutional neural network to obtain a trained initial model;

[0138] S2041, predicting the unlabeled data set through the initial model, and selecting the prediction results with high confidence as pseudo labels according to the confidence threshold;

[0139] S2042, labeling the unlabeled data according to the pseudo-labels to obtain a pseudo-label data set;

[0140] S2043, fine-tuning the initial model according to the pseudo-label data set and the labeled data set, and returning to predict the unlabeled data set through the initial model until a preset number of model fine-tuning times is reached to obtain a trained residual circulation prediction model.

[0141] As an optional implementation method, semi-supervised learning is performed on the convolutional neural network according to the labeled data set and the unlabeled data set to obtain a trained residual circulation prediction model, which specifically includes:

[0142] S2044, dividing the labeled data set into a first data subset and a second data subset;

[0143] S2045, inputting the first data subset and the second data subset into the convolutional neural network respectively to obtain a trained first initial model and a second initial model;

[0144] S2046, predicting the unlabeled data set using the first initial model and the second initial model respectively, selecting a prediction result with high confidence as a pseudo label according to a confidence threshold, and obtaining a first pseudo label and a second pseudo label;

[0145] S2047, labeling the unlabeled data according to the first pseudo label and the second pseudo label, respectively, to obtain a first pseudo label data set and a second pseudo label data set;

[0146] S2048, fine-tuning the first initial model according to the second pseudo-label data set and the first data subset, fine-tuning the second initial model according to the first pseudo-label data set and the second data subset, and returning to predict the unlabeled data set through the first initial model and the second initial model respectively until a preset number of model fine-tuning times is reached, thereby obtaining a trained first cocirculation prediction model and a second cocirculation prediction model;

[0147] S2049, perform integrated learning on the first residual circulation prediction model and the second residual circulation prediction model to obtain a trained residual circulation prediction model.

[0148] Specifically, the implementation steps of the semi-supervised learning in the embodiment of the present invention are as follows:

[0149] 1) Data preprocessing: Clean, standardize and normalize the collected ocean data.

[0150] 2) Initial model training: Use labeled data to train the initial model until it converges.

[0151] 3) Pseudo label generation:

[0152] Self-training: Use the initial model to predict unlabeled data and select high-confidence prediction results as pseudo labels based on the confidence threshold.

[0153] Co-training: If co-training is adopted, the dataset is divided into two sets of features, and two models are trained separately, providing pseudo labels to each other.

[0154] 4) Model fine-tuning: Add pseudo-labeled data to labeled data and continue training the model.

[0155] 5) Iterative optimization: Repeat the pseudo-label generation step and model fine-tuning step until the model performance no longer improves or reaches the predetermined number of iterations.

[0156] 6) Model evaluation: Evaluate model performance on the validation set to ensure that the newly added pseudo-labeled data actually improves model performance.

[0157] 7) Manual review: For the results of automatic labeling, a manual review process can be set up to further improve the accuracy of data labeling.

[0158] In addition, semi-supervised learning can be performed by combining self-training and collaborative training. The specific ideas are as follows:

[0159] 1) Feature selection: First, the features in the dataset are analyzed to determine whether the features can be divided into two groups, each with good discriminative power.

[0160] 2) Collaborative training: If the dataset meets the conditions for collaborative training, the collaborative training method can be used first, allowing the two models to provide pseudo labels independently.

[0161] 3) Self-training as a supplement: Based on collaborative training, for those samples with high confidence predicted by each model, self-training methods can be used to further generate pseudo labels.

[0162] 4) Dynamic Adjustment: Dynamically adjust the confidence threshold and other hyperparameters during self-training and co-training based on the model’s performance on the validation set.

[0163] 5) Ensemble learning: The models obtained by self-training and collaborative training are integrated to improve the robustness and accuracy of the final model.

[0164] As an optional implementation, the labeled data set is input into a convolutional neural network to obtain a trained initial model, which specifically includes:

[0165] S20401, dividing the labeled data set into a training set, a validation set, and a test set;

[0166] S20402, inputting the training set into the convolutional neural network to obtain the residual circulation distribution prediction result;

[0167] S20403. Determine the loss value according to the residual circulation distribution prediction result and the residual circulation distribution label;

[0168] S20404, updating the internal parameters of the convolutional neural network according to the loss value, and inputting the verification set into the updated convolutional neural network to obtain the model performance of the convolutional neural network on the verification set;

[0169] S20405. Update the hyperparameters of the convolutional neural network according to the model performance, and return to input the training set into the convolutional neural network until the model performance no longer improves in multiple consecutive iterations, thereby obtaining a first initial model;

[0170] S20406. Input the test set into the first initial model to obtain a model evaluation index of the first initial model;

[0171] S20407. When the model evaluation index meets the preset requirements, the first initial model is used as the trained initial model.

[0172] Specifically, the specific steps and details of using a labeled dataset to train the initial model are as follows:

[0173] 1. Dataset selection and preparation

[0174] 1) Data collection: Select multi-source data sets including high-frequency ground wave radar data, meteorological station data, satellite remote sensing data, etc.

[0175] 2) Data labeling: Accurately label the data, especially in the identification of residual circulation features, to ensure the accuracy of the training data.

[0176] 3) Data segmentation: The dataset is divided into training set, validation set and test set, with a ratio of 70% training set, 15% validation set and 15% test set.

[0177] 4) Data balance: If the number of samples in some categories in the dataset is unbalanced, the method of oversampling the minority class or undersampling the majority class is used to balance the data.

[0178] 2. Consideration of influencing variables

[0179] 1) Environmental factors: including wind speed, wind direction, temperature, salinity, tide, etc.

[0180] 2) Time Series: Consider the time series characteristics of the data to capture seasonal and cyclical changes.

[0181] 3) Spatial characteristics: Analyze the spatial distribution characteristics of the sea area, such as seabed topography, coastline shape, etc.

[0182] 3. Specific steps of model training

[0183] 1) Determine input parameters: Input parameters usually include ocean dynamic characteristics, such as temperature, salinity, current velocity, flow direction, etc., as well as possible temporal and spatial characteristics.

[0184] 2) Determine output parameters: Output parameters are the targets of model prediction, such as changes in the speed, direction or intensity of the residual circulation.

[0185] 3) Select CNN architecture: Select a CNN architecture that is suitable for the characteristics of ocean data, using existing architectures such as AlexNet, VGGNet, or ResNet.

[0186] 4) Define the loss function: Select an appropriate loss function such as mean squared error (MSE) or cross entropy loss according to the task type (such as classification or regression).

[0187] 5) Choose an optimizer: Choose an optimization algorithm to update the network weights, such as SGD, Adam, or RMSprop.

[0188] 6) Set learning rate and hyperparameters: Determine the learning rate, batch size, number of training rounds (epochs), and other hyperparameters.

[0189] 7) Training process: Use the training data set to train the CNN model. In each epoch, the data is input into the model in batches, and forward propagation and back propagation are performed to update the network weights.

[0190] 8) Validation and early stopping: After each epoch, the validation set is used to evaluate the model performance. If the performance on the validation set does not improve for multiple consecutive epochs, training is stopped to avoid overfitting.

[0191] 9) Model evaluation: Evaluate model performance on an independent test set using accuracy, recall, F1 score, or other relevant metrics.

[0192] 10) Model tuning: Based on the performance of the model on the test set, adjust the network architecture, hyperparameters or training process to optimize the model performance.

[0193] 11) Feature Importance Analysis: Analyze the importance of each feature in the model to understand which input parameters contribute most to the model predictions.

[0194] 4. Input parameters and output parameters

[0195] Input parameters: including pre-processed ocean current velocity, direction, temperature, salinity and other multi-dimensional data.

[0196] Output parameters: residual circulation distribution characteristics predicted by the model, such as flow velocity, flow direction, vortex intensity, etc.

[0197] 5. Improvement and optimization

[0198] 1) Transfer learning and fine-tuning

[0199] Pre-trained models: Use models pre-trained on similar tasks as a starting point. These models have been trained on a large amount of data and can capture common characteristics.

[0200] Fine-tuning: Based on the pre-trained model, fine-tune it according to the characteristics of the current task. This may include replacing the last few layers of the network or adding new layers to the existing network.

[0201] Optimization measures: Make detailed adjustments to the structure of the CNN model, such as increasing or decreasing the number of layers, adjusting the size of the convolution kernel, and introducing or optimizing residual connections.

[0202] 2) Hyperparameter Optimization

[0203] Automated search: Use methods such as Bayesian optimization and genetic algorithms to automatically search for the optimal hyperparameter combination.

[0204] Hyperparameter space definition: Define the search space for hyperparameters such as learning rate, batch size, number of network layers, number of neurons, etc.

[0205] Optimization: Use automated hyperparameter optimization techniques such as grid search, random search, or Bayesian optimization to find the best model parameters.

[0206] 3) Network architecture innovation

[0207] Custom layers: Design custom layers or modules, such as attention mechanisms and residual connections, to enhance the model’s ability to capture key information.

[0208] Multi-scale feature fusion: Combining features of different scales to capture information from local to global.

[0209] 4) Loss function optimization

[0210] Custom loss function: Design the loss function according to the task requirements. For example, for an unbalanced dataset, a weighted loss function can be used to improve the model's ability to recognize minority classes.

[0211] Optimization measures: Design or select a loss function suitable for the prediction of residual circulation characteristics, which may include weighting the loss function to deal with imbalance problems in the data.

[0212] 5) Regularization Techniques

[0213] L1 / L2 Regularization: Add L1 or L2 regularization during model training to reduce model complexity and overfitting.

[0214] Dropout: Randomly drop some network connections during training to prevent the network from overfitting the training data.

[0215] Optimization measures: Apply techniques such as Dropout and L1 / L2 regularization to reduce the overfitting of the model and improve the generalization ability of the model.

[0216] 6) Ensemble learning strategy

[0217] Optimization measures: Use ensemble learning methods, such as bagging or boosting, to combine the prediction results of multiple models to improve the stability and accuracy of the overall prediction.

[0218] 7) Real-time performance optimization

[0219] Optimization measures: Optimize the model to meet real-time or near real-time prediction requirements, including model pruning, quantization, and hardware acceleration.

[0220] In some optional embodiments, the trained model can be deployed on edge computing devices, such as offshore buoys or offshore platforms, to enable real-time processing and analysis of data and reduce data transmission delays.

[0221] Specifically, the model is optimized to adapt to the computing power of edge devices; data preprocessing and model reasoning are implemented on edge devices; and key prediction results are transmitted in real time through wireless communication technology.

[0222] In some optional embodiments, an interactive ocean environment simulation and visualization platform may be developed to enable users to intuitively understand the residual circulation characteristics and their impact on wind field response.

[0223] Specifically, the model prediction results are integrated into the marine environment simulation software; a three-dimensional visualization interface is provided to display the dynamic changes of the residual circulation; and users are allowed to explore the impact of different environmental factors on the residual circulation through interactive tools.

[0224] In some optional embodiments, an adaptive data collection system can be developed to dynamically adjust the frequency and scope of data collection based on the model's predicted needs.

[0225] Specifically, the integrated model feeds back into the data collection strategy, enabling the system to automatically adjust the data collection plan based on the predicted uncertainty; enable remote control and scheduling of data collection equipment; and optimize data storage and processing processes to support adaptive data collection.

[0226] In some optional embodiments, an interdisciplinary decision support system can be constructed to integrate data and models from multiple fields such as oceanography, meteorology, and environmental science to provide support for ocean management and policy making.

[0227] Specifically, integrate models and data sources from different fields to establish a comprehensive decision-making platform; develop decision support tools such as scenario analysis, risk assessment, etc.; and provide decision makers with a friendly user interface and report generation tools.

[0228] As an optional implementation, the residual circulating flow monitoring method further includes the following steps:

[0229] S105, obtaining historical residual circulation distribution data and historical wind field observation data of the target area;

[0230] S106, performing regression analysis on the historical residual circulation distribution data and the historical wind field observation data to obtain the data relationship between the residual circulation distribution and the wind field parameters;

[0231] S107. Determine the wind field distribution forecast of the target area according to the residual circulation distribution forecast and data relationship.

[0232] Specifically, based on the prediction results, the characteristics and movement laws of the residual circulation field and its impact on the wind field are analyzed and evaluated, as follows:

[0233] 1) Determine evaluation objectives and indicators

[0234] Property evaluation: Evaluate the physical properties of the residual circulation, such as speed, direction, temperature, salinity, etc.

[0235] Assessment of movement patterns: Analyze the periodicity, stability and spatial distribution pattern of the residual circulation.

[0236] Wind farm impact assessment: Quantify the impact of residual circulation on wind farm parameters such as wind speed, wind direction, and wind energy density.

[0237] 2) Data collection and processing

[0238] Collect historical observation data of residual circulation and wind field, as well as model prediction results.

[0239] Perform data cleaning, standardization and integration to ensure data quality.

[0240] 3) Statistical analysis

[0241] Descriptive statistics: Calculate the mean, median, standard deviation, etc. to describe the basic characteristics of the residual circulation and wind field.

[0242] Correlation analysis: Use correlation coefficients (such as the Pearson correlation coefficient) to assess the linear relationship between variables.

[0243] Regression analysis: Use linear regression or nonlinear regression models to evaluate the relationship between residual circulation characteristics and wind field parameters.

[0244] 4) Model Validation

[0245] Cross-validation: Use methods such as K-fold cross-validation to evaluate the prediction accuracy and generalization ability of the model.

[0246] Error analysis: Calculate indicators such as mean square error (MSE), root mean square error (RMSE), and mean absolute error (MAE) to evaluate the prediction error.

[0247] 5) Kinetic analysis

[0248] Flow Visualization: Visualize the flow characteristics of the residual annular flow using tools such as streamline plots and velocity vector fields.

[0249] Vorticity analysis: Calculate vorticity and eddies, and analyze the dynamic structure of residual circulation.

[0250] 6) Analysis of influencing factors

[0251] Identify and quantify the key factors that affect the characteristics and movement of the residual circulation, such as topography, climatic conditions, seasonal changes, etc.

[0252] 7) Climate impact assessment

[0253] Climate model integration: Integrate residual circulation models with climate models to assess their impact on regional climate patterns.

[0254] Extreme event analysis: Evaluate the response of the residual circulation to extreme climate events (e.g. typhoons, storm surges).

[0255] 8) Model prediction and risk assessment

[0256] Scenario analysis: Simulate different wind field scenarios to predict changes in residual circulation and their potential impacts.

[0257] Risk Matrix: Construct a risk matrix to assess the potential risks and impacts under different scenarios.

[0258] 9) Comprehensive evaluation report

[0259] Prepare an assessment report summarizing the analysis results of the residual circulation characteristics, movement patterns and wind field impacts.

[0260] Provide charts, data analysis and professional explanations to make the report easy to understand.

[0261] 10) Application and decision support

[0262] The evaluation results will be applied to fields such as marine engineering and wind power generation to provide a scientific basis for decision-making.

[0263] The relevant optimization measures are as follows:

[0264] 1) Multi-dimensional feature analysis:

[0265] A multidimensional characteristic analysis method is developed, which not only considers the flow velocity and direction, but also includes ocean physical parameters such as temperature, salinity, and pressure to comprehensively evaluate the characteristics of the residual circulation.

[0266] 2) Dynamic time series analysis:

[0267] Dynamic time series analysis techniques, such as long short-term memory networks (LSTM) or temporal convolutional networks (TCN), are introduced to capture the temporal dynamic characteristics of cocirculation flows.

[0268] 3) Weight assessment of influencing factors:

[0269] Use machine learning algorithms, such as random forests or gradient boosted decision trees, to assess the weight of different environmental factors on the residual circulation characteristics.

[0270] 4) Interactive visualization tools:

[0271] Develop interactive visualization tools that allow users to dynamically adjust parameters according to different needs and scenarios and view changes in residual circulation characteristics in real time.

[0272] 5) Quantification of forecast uncertainty:

[0273] Introduce prediction uncertainty quantification techniques, such as Bayesian methods or ensemble learning, to provide a confidence assessment of the prediction results.

[0274] 6) Interdisciplinary impact assessment:

[0275] Combining research findings from multiple disciplines such as oceanography, meteorology, and ecology, the impact of residual circulation on wind fields and the broader environment is assessed.

[0276] 7) Real-time data feedback mechanism:

[0277] A real-time data feedback mechanism is established to feed the latest observation data into the model to dynamically adjust and optimize the prediction results.

[0278] The method steps of the embodiment of the present invention are described above. It can be recognized that the embodiment of the present invention combines the FVCOM model with a deep neural network, uses the FVCOM model to generate ocean dynamics simulation data as the input of the residual circulation prediction model, and updates the boundary conditions of the FVCOM model according to the residual circulation distribution prediction data output by the residual circulation prediction model, thereby improving the model's prediction ability for the residual circulation distribution field in a data-driven manner, thereby improving the accuracy and efficiency of the residual circulation distribution prediction; in addition, the wind field distribution prediction can be determined based on the obtained residual circulation distribution prediction, thereby improving the accuracy and efficiency of the wind field distribution prediction. The embodiment of the present invention has broad application prospects in the fields of marine engineering, wind power generation, etc., and is expected to provide strong technical support for the development of related fields.

[0279] Reference Figure 2 , an embodiment of the present invention provides a residual circulating current monitoring system, comprising:

[0280] The data acquisition module is used to obtain the basic sea area data and multi-source environmental data of the target sea area, and establish the FVCOM model based on the basic sea area data;

[0281] The ocean numerical simulation module is used to use multi-source environmental data as boundary conditions, perform ocean numerical simulation on the target sea area through the FVCOM model, and obtain the ocean dynamics simulation data of the target sea area;

[0282] A residual circulation distribution prediction module is used to input the ocean dynamics simulation data into the pre-trained residual circulation prediction model of the target sea area to obtain the residual circulation distribution prediction data of the target sea area;

[0283] The boundary condition update module is used to update the boundary conditions according to the residual circulation distribution prediction data, and return to perform ocean numerical simulation of the target sea area through the FVCOM model until the preset number of simulations is reached to obtain the residual circulation distribution prediction of the target sea area.

[0284] The contents of the above method embodiments are all applicable to the present system embodiments. The functions specifically implemented by the present system embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0285] Reference Figure 3 , an embodiment of the present invention provides a residual circulating current monitoring device, comprising:

[0286] at least one processor;

[0287] at least one memory for storing at least one program;

[0288] When the at least one program is executed by the at least one processor, the at least one processor implements the residual circulating current monitoring method.

[0289] The contents of the above method embodiments are all applicable to the present device embodiments. The functions specifically implemented by the present device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0290] An embodiment of the present invention further provides a computer-readable storage medium, in which a program executable by a processor is stored. When the program executable by the processor is executed by the processor, it is used to execute the above-mentioned residual circulating current monitoring method.

[0291] A computer-readable storage medium of an embodiment of the present invention can execute a residual circulating current monitoring method provided by an embodiment of the method of the present invention, can execute any combination of implementation steps of the method embodiment, and has the corresponding functions and beneficial effects of the method.

[0292] The embodiment of the present invention also discloses a computer program product or a computer program, wherein the computer program product or the computer program includes computer instructions, and the computer instructions are stored in a computer-readable storage medium. A processor of a computer device can read the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes Figure 1 The method shown.

[0293] In some optional embodiments, the function / operation mentioned in the block diagram may not occur in the order mentioned in the operation diagram. For example, depending on the function / operation involved, the two boxes shown in succession can actually be executed substantially simultaneously or the above-mentioned boxes can sometimes be executed in reverse order. In addition, the embodiment presented and described in the flow chart of the present invention is provided by way of example, for the purpose of providing a more comprehensive understanding of technology. The disclosed method is not limited to the operation and logic flow presented herein. The optional embodiment is expected, wherein the order of various operations is changed and the sub-operation described as a part of a larger operation is performed independently.

[0294] In addition, although the present invention is described in the context of functional modules, it should be understood that, unless otherwise specified to the contrary, one or more of the above-mentioned functions and / or features can be integrated into a single physical device and / or software module, or one or more functions and / or features can 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 present invention. More specifically, in view of the properties, functions and internal relationships of the various functional modules in the device disclosed herein, the actual implementation of the module will be understood within the conventional skills of the engineer. Therefore, those skilled in the art can implement the present invention set forth in the claims without excessive experimentation using ordinary techniques. It is also understood that the specific concepts disclosed are merely illustrative and are not intended to limit the scope of the present invention, which is determined by the full scope of the appended claims and their equivalents.

[0295] If the above functions are implemented in the form of 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 the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the above methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0296] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0297] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and editable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the above-mentioned program is printed, since the above-mentioned program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or processing in other suitable ways as necessary, and then stored in a computer memory.

[0298] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0299] In the above description of this specification, the description with reference to the terms "one embodiment / example", "another embodiment / example" or "certain embodiments / examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0300] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the claims and their equivalents.

[0301] The above is a specific description of the preferred implementation of the present invention, but the present invention is not limited to the above embodiments. Those skilled in the art may make various equivalent modifications or substitutions without violating the spirit of the present invention. These equivalent modifications or substitutions are all included in the scope defined by the claims of this application.

Claims

1. A residual circulation monitoring method, characterized in that: The following steps are involved: Acquire basic sea area data and multi-source environmental data of the target sea area, and establish an FVCOM model based on the basic sea area data; Using the multi-source environmental data as boundary conditions, performing ocean numerical simulation on the target sea area through the FVCOM model, and obtaining ocean dynamics simulation data of the target sea area; Inputting the ocean dynamics simulation data into a pre-trained residual circulation prediction model of the target sea area to obtain residual circulation distribution prediction data of the target sea area; The boundary conditions are updated according to the residual circulation distribution prediction data, and the FVCOM model is used to perform ocean numerical simulation on the target sea area until a preset number of simulations is reached to obtain the residual circulation distribution prediction of the target sea area.

2. A residual circulating flow monitoring method according to claim 1, characterized in that: The basic sea area data includes seabed topography data, coastline shape data and seawater level data of the target sea area, and the multi-source environmental data includes meteorological data, ocean current data, seawater temperature data and seawater salinity data.

3. A residual circulating flow monitoring method according to claim 1, characterized in that: The residual circulation prediction model is trained by the following steps: Acquire first ocean dynamics sample data of multiple historical periods of the target sea area and corresponding residual circulation distribution labels, wherein the residual circulation distribution labels include the distribution of residual circulation velocity, residual circulation direction, and residual circulation intensity of the target sea area; Constructing a labeled data set according to the first ocean dynamics sample data and the residual circulation distribution label; Acquire second ocean dynamics sample data of multiple historical periods of the target sea area, and construct an unlabeled data set according to the second ocean dynamics sample data; Semi-supervised learning is performed on a pre-constructed convolutional neural network according to the labeled data set and the unlabeled data set to obtain the trained residual circulation prediction model.

4. A residual circulating flow monitoring method according to claim 3, characterized in that: The semi-supervised learning of the convolutional neural network is performed according to the labeled data set and the unlabeled data set to obtain the trained residual circulation prediction model, which specifically includes: Inputting the labeled data set into the convolutional neural network to obtain a trained initial model; Predicting the unlabeled data set using the initial model, and selecting a high-confidence prediction result as a pseudo label according to a confidence threshold; Annotate the unlabeled data according to the pseudo-label to obtain a pseudo-label data set; The initial model is fine-tuned according to the pseudo-label data set and the labeled data set, and the unlabeled data set is predicted by the initial model again until a preset number of model fine-tuning times is reached to obtain the trained residual circulation prediction model.

5. A residual circulating flow monitoring method according to claim 3, characterized in that: The semi-supervised learning of the convolutional neural network is performed according to the labeled data set and the unlabeled data set to obtain the trained residual circulation prediction model, which specifically includes: Dividing the labeled data set into a first data subset and a second data subset; Inputting the first data subset and the second data subset into the convolutional neural network respectively to obtain a trained first initial model and a second initial model; Predicting the unlabeled data set using the first initial model and the second initial model respectively, selecting a prediction result with high confidence as a pseudo label according to a confidence threshold, and obtaining a first pseudo label and a second pseudo label; The unlabeled data are labeled according to the first pseudo-label and the second pseudo-label to obtain a first pseudo-label data set and a second pseudo-label data set; Fine-tune the first initial model according to the second pseudo-label data set and the first data subset, fine-tune the second initial model according to the first pseudo-label data set and the second data subset, and return to predict the unlabeled data set through the first initial model and the second initial model respectively until a preset number of model fine-tuning times is reached, so as to obtain a trained first cocirculating flow prediction model and a second cocirculating flow prediction model; The first residual circulating flow prediction model and the second residual circulating flow prediction model are integrated and learned to obtain the trained residual circulating flow prediction model.

6. A residual circulating flow monitoring method according to claim 4, characterized in that: The step of inputting the labeled data set into the convolutional neural network to obtain a trained initial model specifically includes: Dividing the labeled data set into a training set, a validation set, and a test set; Inputting the training set into the convolutional neural network to obtain a residual circulation distribution prediction result; Determine the loss value according to the residual circulation distribution prediction result and the residual circulation distribution label; Updating internal parameters of the convolutional neural network according to the loss value, and inputting the validation set into the updated convolutional neural network to obtain a model performance of the convolutional neural network on the validation set; Updating the hyperparameters of the convolutional neural network according to the model performance, and returning the training set to the convolutional neural network until the model performance no longer improves in a plurality of consecutive iterations, thereby obtaining a first initial model; Inputting the test set into the first initial model to obtain a model evaluation index of the first initial model; When the model evaluation index meets the preset requirements, the first initial model is used as the trained initial model.

7. A residual circulating flow monitoring method according to any one of claims 1 to 6, characterized in that: The residual circulating flow monitoring method further comprises the following steps: Obtaining historical residual circulation distribution data and historical wind field observation data of the target area; Performing regression analysis on the historical residual circulation distribution data and the historical wind field observation data to obtain a data relationship between the residual circulation distribution and the wind field parameters; The wind field distribution prediction of the target area is determined according to the residual circulation distribution prediction and the data relationship.

8. A residual circulation monitoring system, characterized in that: include: A data acquisition module is used to acquire basic sea area data and multi-source environmental data of the target sea area, and establish an FVCOM model based on the basic sea area data; An ocean numerical simulation module, used to use the multi-source environmental data as boundary conditions, perform ocean numerical simulation on the target sea area through the FVCOM model, and obtain ocean dynamics simulation data of the target sea area; A residual circulation distribution prediction module is used to input the ocean dynamics simulation data into a pre-trained residual circulation prediction model of the target sea area to obtain residual circulation distribution prediction data of the target sea area; The boundary condition updating module is used to update the boundary conditions according to the residual circulation distribution prediction data, and return to perform ocean numerical simulation of the target sea area through the FVCOM model until a preset number of simulations is reached to obtain the residual circulation distribution prediction of the target sea area.

9. A residual circulating current monitoring device, characterized in that: include: at least one processor; at least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements a residual circulating current monitoring method as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a program executable by a processor, characterized in that: The processor-executable program is used to execute a residual circulating current monitoring method as claimed in any one of claims 1 to 7 when executed by the processor.

Images