A marine vortex prediction method and device, electronic equipment and storage medium

Abstract

The application discloses a marine vortex prediction method and device, electronic equipment and storage medium, and relates to the technical field of data processing. The method comprises the following steps: collecting monitoring parameters of a marine vortex by an unmanned device; constructing a moving path prediction model of a vortex center position according to a time sequence of the moving characteristics of the marine vortex in the monitoring parameters, and then determining the moving path of the marine vortex; constructing a moving speed prediction model of the marine vortex according to the change amount of the center position of the marine vortex with time and the reanalysis wind field information of the region where the marine vortex is located in the monitoring parameters, and then determining the moving speed of the marine vortex; and constructing a life cycle prediction model of the marine vortex according to the vortex state parameters and the moving track parameters of the marine vortex in the monitoring parameters, and then determining the life cycle of the marine vortex. The application can more accurately predict the moving path, moving speed and life cycle of the marine vortex by using more comprehensive monitoring data.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, and particularly relates to an ocean vortex prediction method and device, electronic equipment and storage medium. BACKGROUND

[0002] The prior art predicts the future characteristics of ocean vortexes by monitoring the ocean vortexes through satellites, but the satellites can only obtain the information of sea currents on the surface of the ocean. However, the ocean vortexes are three-dimensional structures, and are constantly evolving under the influence of sea-air interaction, topography and the like. Therefore, the vortexes predicted by the prior art are still not accurate enough. SUMMARY

[0003] The main purpose of the embodiments of the present application is to provide an ocean vortex prediction method, device, electronic equipment and storage medium to improve the accuracy of predicting the vortexes.

[0004] To achieve the above purpose, one aspect of the embodiments of the present application provides an ocean vortex prediction method, which comprises the following steps:

[0005] Collecting monitoring parameters of the ocean vortex through an unmanned device;

[0006] Constructing a moving path prediction model of the vortex center position according to the time sequence of the moving characteristics of the ocean vortex in the monitoring parameters;

[0007] Determining the moving path of the ocean vortex according to the moving path prediction model;

[0008] Constructing a moving speed prediction model of the ocean vortex according to the change amount of the center position of the ocean vortex with time in the monitoring parameters and the reanalysis wind field information of the region where the ocean vortex is located;

[0009] Determining the moving speed of the ocean vortex according to the moving speed prediction model;

[0010] Constructing a life cycle prediction model of the ocean vortex according to the vortex state parameters and moving trajectory parameters of the ocean vortex in the monitoring parameters;

[0011] Determining the life cycle of the ocean vortex according to the life cycle prediction model.

[0012] In some embodiments, the step of constructing a moving path prediction model of the vortex center position according to the time sequence of the moving characteristics of the ocean vortex in the monitoring parameters comprises the following steps:

[0013] Drawing a spatial flow field map according to the sea currents observed by the SWOT satellite and the ground wave radar;

[0014] determining the longitude and latitude of the center point of the clockwise or counterclockwise rotating flow field as the initial position of the marine eddy center according to the spatial flow field map by using the flow field geometric feature method;

[0015] determining the longitude difference or latitude difference between adjacent two time instants as the position difference information of each marine eddy;

[0016] calculating the included angle between the path connecting vectors of adjacent two time instants;

[0017] determining the position difference information, the included angle and the eddy attribute feature value as the movement feature of the marine eddy, wherein the eddy attribute feature value includes the absolute value of vorticity, deformation rate and eddy energy;

[0018] constructing the movement path prediction model by taking the movement features of the previous three time instants as input and taking the longitude or latitude of the center position of the marine eddy at the current time instant as output, wherein the movement path prediction model includes a convolutional neural network, a bidirectional long short-term memory network model and an attention mechanism;

[0019] constructing a time series dataset according to the movement features, wherein the time series dataset includes the movement features of seven consecutive time instants;

[0020] standardizing the time series dataset, and then training the movement path prediction model by using the standardized time series dataset.

[0021] In some embodiments, the method of determining the movement path of the marine eddy according to the movement path prediction model includes the following steps:

[0022] inputting the position difference information, the included angle and the eddy attribute feature value of the marine eddy at any time instant into the trained movement path prediction model to obtain the output value of the movement path prediction model;

[0023] de-normalizing the output value to obtain the longitude or latitude of the center position of the marine eddy.

[0024] In some embodiments, the method of constructing the movement speed prediction model of the marine eddy according to the change amount of the center position of the marine eddy over time in the monitoring parameters and the reanalysis wind field information of the region where the marine eddy is located includes the following steps:

[0025] drawing a spatial flow field map according to the ocean current observed by the SWOT satellite and the ground wave radar;

[0026] The method comprises the following steps: determining the initial longitude and latitude of the center of the marine eddy according to the spatial flow field diagram by using a flow field geometric feature method, and determining the position of the center of the clockwise or counterclockwise rotating flow field as the initial longitude and latitude of the marine eddy center;

[0027] Starting from the initial longitude and latitude, the change amount of the center position of the marine eddy between adjacent time points is calculated in the longitude direction and the latitude direction, respectively.

[0028] The reanalysis wind field information provided by ECMWF is collected, and the spatially averaged zonal wind speed and meridional wind speed of the sea area where the marine eddy is located are determined according to the reanalysis wind field information; wherein the area in the reanalysis wind field information covers the sea area where the maximum radius of the marine eddy is located, and the time sequence covers seven consecutive time points.

[0029] Based on the reanalysis wind field information, the information entropy method is used to calculate the information entropy values of the zonal wind speed and the meridional wind speed in the sea area where the marine eddy is located.

[0030] The calculation formula of the information entropy value is as follows:

[0031]

[0032] Wherein, H(X) represents the information entropy value, x i The component values of the wind speed in the longitude or latitude direction corresponding to the same time point and different spatial positions, n represents the number of spatial wind speed values provided by ECMWF in the marine eddy coverage area;

[0033] The moving speed prediction model is constructed by using a causal convolutional neural network method.

[0034] The expression of the convolution operation in the causal convolutional neural network is as follows:

[0035]

[0036] Wherein, x[t] is the value of the input time sequence at time t, ω[k] is the weight of the convolution kernel, K is the size of the convolution kernel, and y[t] is the output after the convolution operation.

[0037] The expression of the prediction value of the moving speed prediction model is as follows:

[0038]

[0039] Wherein, is the prediction value, and b is the bias term.

[0040] In some embodiments, the method comprises the following steps:

[0041] The change amount of the center position of the marine eddy between adjacent time instants, the change amount of the information entropy of the wind speed of the adjacent time instants in the sea area where the marine eddy is located, and the spatially averaged wind speed and wind volume of the marine eddy at seven time instants are input into the moving speed prediction model to obtain the moving speed of the marine eddy.

[0042] In some embodiments, the step of constructing the life cycle prediction model of the marine eddy according to the eddy state parameters and the moving trajectory parameters of the marine eddy in the monitoring parameters comprises the following steps:

[0043] The eddy state parameters and the moving trajectory parameters are used as input parameters, and the life cycle of the marine eddy is used as an output parameter, and a residual network model is constructed as a life cycle prediction model.

[0044] The residual network model is implemented by using a Pytortch third-party deep learning library; in the residual network model, a StandardScaler class in a scikit-learn library is used for data preprocessing of the input parameters, so as to convert each input parameter into normally distributed data with a mean of 0 and a standard deviation of 1; and ReLUs are used as nonlinear activation functions in residual blocks of the residual network model.

[0045] In some embodiments, the step of determining the life cycle of the marine eddy according to the life cycle prediction model comprises the following steps:

[0046] The age, vorticity, eddy kinetic energy, deformation rate and instantaneous propagation speed of the marine eddy are used as the eddy state parameters, and the latitude and longitude of the center position of the marine eddy and the displacement distance of the current center position relative to the initial time instant are used as the moving trajectory parameters, which are input into the life cycle prediction model to predict the life cycle of the marine eddy.

[0047] To achieve the above object, another aspect of the embodiment of the present application proposes a marine eddy prediction device, which comprises:

[0048] A data acquisition unit is configured to acquire monitoring parameters of a marine eddy by using an unmanned device.

[0049] A first model construction unit is configured to construct a moving path prediction model of an eddy center position according to a time sequence of moving features of the marine eddy in the monitoring parameters.

[0050] A moving path prediction unit is configured to determine a moving path of the marine eddy according to the moving path prediction model.

[0051] a second model construction unit, configured to construct a moving speed prediction model of the ocean eddy according to a change amount of a center position of the ocean eddy over time in the monitoring parameters and reanalysis wind field information of a region where the ocean eddy is located;

[0052] a moving speed prediction unit, configured to determine a moving speed of the ocean eddy according to the moving speed prediction model;

[0053] a third model construction unit, configured to construct a life cycle prediction model of the ocean eddy according to an eddy state parameter and a moving track parameter of the ocean eddy in the monitoring parameters;

[0054] a life cycle prediction unit, configured to determine a life cycle of the ocean eddy according to the life cycle prediction model.

[0055] To achieve the above object, another aspect of the embodiments of the present application provides an electronic device, which comprises a memory and a processor, the memory stores a computer program, and the processor implements the above method when executing the computer program.

[0056] To achieve the above object, another aspect of the embodiments of the present application provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the above method.

[0057] The embodiments of the present application at least have the following beneficial effects:

[0058] The present application can collect monitoring parameters of ocean eddies by unmanned equipment, construct a moving path prediction model of an eddy center position according to a time sequence of moving characteristics of the ocean eddy in the monitoring parameters, determine a moving path of the ocean eddy according to the moving path prediction model, construct a moving speed prediction model of the ocean eddy according to a change amount of a center position of the ocean eddy over time in the monitoring parameters and reanalysis wind field information of a region where the ocean eddy is located, determine a moving speed of the ocean eddy according to the moving speed prediction model, construct a life cycle prediction model of the ocean eddy according to an eddy state parameter and a moving track parameter of the ocean eddy in the monitoring parameters, and determine a life cycle of the ocean eddy according to the life cycle prediction model. The present application can collect comprehensive monitoring data of ocean eddies by unmanned equipment, and then predict the moving path, moving speed and life cycle of the ocean eddy more accurately by using more comprehensive monitoring data, thereby providing a reliable scheme for predicting ocean eddies. BRIEF DESCRIPTION OF DRAWINGS

[0059] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the embodiments will be briefly introduced. Obviously, the drawings in the following description only represent some of the embodiments of the present application, and for those skilled in the art, other drawings can be obtained without creative effort.

[0060] Figure 1 A flowchart of a marine vortex prediction method provided by an embodiment of the present application is shown in FIG. 1.

[0061] Figure 2 A structural diagram of a marine vortex prediction device provided by an embodiment of the present application is shown in FIG. 2.

[0062] Figure 3 A hardware structural diagram of an electronic device provided by an embodiment of the present application is shown in FIG. 3. DETAILED DESCRIPTION

[0063] In order to make the objects, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and not to limit the present application. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with embodiments of the present application, but are only examples of devices and methods consistent with some aspects of the embodiments of the present application as detailed in the appended claims.

[0064] It can be understood that the terms "first", "second", and the like used in the present application can be used herein to describe various concepts, but unless specifically stated, these concepts are not limited by these terms. These terms are only used to distinguish one concept from another. For example, without departing from the scope of the embodiments of the present application, the first information can also be referred to as the second information, and similarly, the second information can also be referred to as the first information. Depending on the context, the word "if" as used herein can be interpreted as "when" or "when" or "in response to determining".

[0065] The terms "at least one", "multiple", "each", "any" and the like used in the present application include one, two or more than two, multiple includes two or more than two, each refers to each of the corresponding multiple, and any refers to any one of the multiple.

[0066] 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 to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to be limiting of this application.

[0067] Embodiments of the present application provide a marine vortex prediction method and device, electronic equipment and storage medium, relating to the technical field of data processing. The marine vortex prediction method and device, electronic equipment and storage medium provided by the embodiments of the present application can be applied in a terminal, can also be applied in a server, and can also be software running in a terminal or a server. In some embodiments, the terminal can be a smart phone, a tablet computer, a notebook computer, a desktop computer, a smart speaker, a smart watch, a vehicle-mounted terminal, and the like, but is not limited thereto; the server end can be configured as a stand-alone physical server, can also be configured as a server cluster or a distributed system composed of multiple physical servers, can also be configured as a cloud server providing basic cloud computing services such as cloud service, cloud database, cloud computing, cloud function, cloud storage, network service, cloud communication, middleware service, domain name service, security service, CDN, and big data and artificial intelligence platform, and the server can also be a node server in a blockchain network; the software can be an application for implementing a marine vortex prediction method, and the like, but is not limited to the above forms.

[0068] The present application can be used in many general or special computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, and the like. The present application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, and the like that perform specific tasks or implement specific abstract data types. The present application can also be practiced in a distributed computing environment, in which tasks are performed by remote processing devices connected by a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media, including storage devices.

[0069] Reference Figure 1 The marine vortex prediction method provided by the embodiments of the present application can include, but is not limited to, S100 to S160, and specifically as follows:

[0070] S100: Collecting monitoring parameters of marine vortex by unmanned equipment.

[0071] Specifically, the embodiment can monitor ocean eddies at close range by unmanned equipment to collect monitoring data.

[0072] S110: constructing a moving path prediction model of the eddy center position according to a time sequence of the moving feature of the ocean eddy in the monitoring parameter.

[0073] Further, S110 can include the following steps S111-S117:

[0074] S111: drawing a spatial flow field map according to the ocean current observed by the SWOT satellite and the ground wave radar;

[0075] S112: determining the central point position longitude and latitude of the clockwise rotating or counterclockwise rotating flow field as the initial position of the ocean eddy center according to the spatial flow field map by applying the flow field geometric feature method;

[0076] S113: determining the longitude difference or latitude difference between adjacent two time points in the first three time points of each ocean eddy as position difference information;

[0077] S114: calculating the included angle between the path connecting vectors of adjacent two time points;

[0078] determining the position difference information, the included angle, and the eddy attribute feature value as the moving feature of the ocean eddy; wherein the eddy attribute feature value includes the absolute value of vorticity, deformation rate, and eddy energy;

[0079] S115: constructing the moving path prediction model by taking the moving feature of the first three time points as input and taking the longitude or latitude of the center position of the ocean eddy at the current time as output; wherein the moving path prediction model includes convolutional neural network, bidirectional long short-term memory network model, and attention mechanism;

[0080] S116: constructing a time sequence data set according to the moving feature; wherein the time sequence data set includes the moving feature of seven consecutive time points;

[0081] S117: standardizing the time sequence data set, and then training the moving path prediction model by using the standardized time sequence data set.

[0082] S120: determining the moving path of the ocean eddy according to the moving path prediction model.

[0083] Further, S120 can include S121-S122:

[0084] S121: input the position difference information, the included angle and the vortex attribute characteristic value of the marine vortex at any moment into the trained mobile path prediction model to obtain an output value of the mobile path prediction model;

[0085] S122: reverse-normalize the output value to obtain the longitude or latitude of the center position of the marine vortex.

[0086] S130: construct a mobile speed prediction model of the marine vortex according to the change amount of the center position of the marine vortex in the monitoring parameter over time and the reanalysis wind field information of the region where the marine vortex is located.

[0087] Further, S130 can include the following steps S131-S136:

[0088] S131: draw a spatial flow field map according to the ocean current observed by the SWOT satellite and the ground wave radar;

[0089] S132: apply a flow field geometric feature method to determine the center point position longitude and latitude of the clockwise rotating or counterclockwise rotating flow field as the initial longitude and latitude of the marine vortex center according to the spatial flow field map;

[0090] S133: starting from the initial longitude and latitude, calculate the change amount of the center position of the marine vortex between adjacent moments in the longitude direction and the latitude direction, respectively;

[0091] S134: collect the reanalysis wind field information provided by ECMWF, and then determine the spatially averaged meridional wind speed and wind amount and the latitudinal wind speed and wind amount in the sea area where the marine vortex is located according to the reanalysis wind field information; wherein the region in the reanalysis wind field information covers the sea area where the maximum radius of the marine vortex is located, and the time series covers seven consecutive moments;

[0092] S135: based on the reanalysis wind field information, use an information entropy method to calculate the information entropy values of the wind speed and wind amount in the meridional direction and the wind speed and wind amount in the latitudinal direction in the sea area where the marine vortex is located;

[0093] The calculation formula of the information entropy value is as follows:

[0094]

[0095] Wherein, H(X) represents the information entropy value, x i The component values of the wind speed in the longitude or latitude direction corresponding to the same moment and different spatial positions, n represents the number of spatial wind speed values provided by ECMWF in the coverage area of the marine vortex;

[0096] S136: use a causal convolutional neural network method to construct the mobile speed prediction model;

[0097] The expression of the convolution operation in the causal convolutional neural network is:

[0098]

[0099] wherein x[t] is the value of the input time sequence at time t, ω[k] is the weight of the convolution kernel, K is the size of the convolution kernel, and y[t] is the output after the convolution operation;

[0100] The expression of the prediction value of the moving speed prediction model is:

[0101]

[0102] wherein, is the prediction value, and b is a bias term.

[0103] S140: determining the moving speed of the marine eddy according to the moving speed prediction model.

[0104] Further, S140 can include the following step S141:

[0105] S141 inputs the change amount of the center position of the marine eddy between adjacent time instants, the change amount of the wind speed information entropy of the marine eddy in the sea area where the marine eddy is located, and the spatially averaged wind speed and wind volume of the marine eddy in the sea area at 7 time instants into the moving speed prediction model, to obtain the moving speed of the marine eddy.

[0106] S150: constructing a life cycle prediction model of the marine eddy according to the eddy state parameters and the moving trajectory parameters of the marine eddy in the monitoring parameters.

[0107] Further, S150 can include the following step S151:

[0108] S151: constructing a residual network model as the life cycle prediction model by taking the eddy state parameters and the moving trajectory parameters as input parameters and taking the life cycle of the marine eddy as an output parameter;

[0109] wherein the residual network model is implemented through a Pytortch third-party deep learning library; the StandardScaler class in the scikit-learn library is used in the residual network model to perform data preprocessing on the input parameters, so as to convert each input parameter into normally distributed data with a mean of 0 and a standard deviation of 1; and ReLUs are used as the nonlinear activation function in the residual blocks of the residual network model.

[0110] S160: determining the life cycle of the marine eddy according to the life cycle prediction model.

[0111] Further, S160 can include the following step S161:

[0112] S161: input the age, vorticity, eddy kinetic energy, deformation rate and instantaneous propagation speed of the marine eddy as the eddy state parameters, and the latitude and longitude of the center position of the marine eddy and the displacement distance of the current center position relative to the initial moment as the movement trajectory parameters into the life cycle prediction model, and predict the life cycle of the marine eddy.

[0113] Next, the scheme of the embodiments of the present application will be described in detail with reference to specific application examples.

[0114] In view of the complexity of the evolution of marine eddies, the embodiments focus on the construction of prediction models for representative characteristic elements (movement path, movement speed and life cycle) of marine eddies.

[0115] First eddy characteristic element: prediction model of eddy movement path.

[0116] The generation, evolution and extinction of eddies are closely related to the structure of the wind field. The eddy path prediction steps proposed in the embodiments are as follows:

[0117] (1) After the eddy is generated, the spatial flow field map is drawn according to the ocean current observed by the SWOT satellite and the ground wave radar, and the center point position latitude and longitude of the clockwise or counterclockwise rotating flow field is determined by applying the flow field geometric feature (VG) method, which is regarded as the most initial position WZ0 of the eddy center.

[0118] (2) The position difference information (the difference value of longitude or latitude between adjacent two time points, i.e. the difference value of longitude and latitude between t-2 and t-3 time points, and the difference value of longitude and latitude between t-2 and t-1 time points) of each eddy at the previous three time points (t-3, t-2, t-1), the included angle a between the path connecting vector of (t-1, t-2) two time points and the path connecting vector of (t-2, t-3) two time points, and the attribute characteristic value of the eddy (including the absolute value of vorticity, deformation rate and eddy kinetic energy) are taken as input variables, and the longitude or latitude data of the eddy center at the current time (t0) is taken as output variable, and the longitude prediction model and the latitude prediction model of the eddy center position are established respectively.

[0119] (3) Based on the convolutional neural network (CNN) with high efficient data feature mining ability, the bidirectional long short-term memory network model (BiLSTM) and the attention mechanism (Attention) for allocating feature weight information, a vortex trajectory CNN-BiLSTM-Attention prediction model is constructed.

[0120] The training of the model is as follows: ① input data. The input variables required for the training of the CNN-BiLSTM-Attention model are input, including the longitude and latitude difference, deformation rate, vorticity and eddy energy of the previous three time points. ② Construct time series. The sample data is constructed into a time series corresponding to the feature parameters of the previous three time points and the t0 time point of each vortex. ③ Data set division. Randomly select 70% of the data set as the training set and the remaining 30% as the test set. ④ Data standardization. Considering that different attributes have different dimensions, resulting in large differences in numerical values, the data is normalized to [0, 1]. ⑤ Initialize parameters. Initialize the parameter configuration of the CNN-BiLSTM-Attention model. ⑥ Feature attention layer calculation. The standardized data is input into the attention layer for calculation, which is used to learn the importance of each input factor in the input sequence so that the model can focus on relatively important feature parameters. ⑦ Sequence folding layer calculation. The input sample data is converted into a form suitable for convolution operation by the convolution layer. ⑧ CNN layer calculation. The folded data is input into the CNN model for calculation. After the stacking of the convolution layer and the activation function, the model can learn more complex feature expressions. ⑨ Sequence unfolding layer calculation. The sample data is restored to the time series state. ⑩ Flatten layer calculation. The sample data is converted into a one-dimensional vector data that can be read by BiLSTM. BiLSTM model calculation. The output data of the Flatten layer is input into the LSTM unit in the forward and backward directions of the BiLSTM layer for training, and the output data is obtained. The output data and the true value are calculated by the loss function to determine whether it is lower than the threshold value. If yes, the model is saved, otherwise the parameters are adjusted repeatedly for model training.

[0121] (4) Vortex movement path prediction process:

[0122] Load data. The saved trained CNN-BiLSTM-Attention model is loaded to keep the trained weights and biases of each layer of neural network; 2) input data. The divided test set is input into the model for prediction, and finally the output value of the CNN-BiLSTM-Attention model is obtained; 3) inverse normalization calculation. The output value is restored to the latitude / longitude data in the motion trajectory; 4) output results. The restored data is output and compared with the true results to evaluate the regression effect of the model. The related parameters of the CNN-BiLSTM-Attention model are: the total number of times of network training Epoch is set to 500, the initial adjustment step of network parameters in the training stage is set to 0.03, the setting for updating network parameters in the training process is 0.45, the learning rate drop factor is set to 0.01, and the learning rate drop period is 120.

[0123] (5) Prediction result analysis:

[0124] In the embodiment, the latitude prediction model and the longitude prediction model of the vortex motion trajectory are constructed respectively, the parameters of the two models are set the same, and the same data set is used for the test set and the training set to ensure that the prediction effects of the two models are the same.

[0125] (6) Prediction result evaluation:

[0126] To test the prediction effect of the CNN-BiLSTM-Attention model on the vortex moving trajectory, the distance difference between the prediction result and the true value and the direction angle are evaluated. From the direction angle of the vortex motion, the prediction effects of the latitude and longitude parameter prediction models are evaluated. The direction of the vortex position at t0 relative to the position at t-1 is divided into eight directions, i.e. east, south, west, north, southeast, northeast, southwest and northwest. The direction of the real and predicted vortex center position (t0 time) relative to the previous time (t-1 time) position is compared to obtain the density distribution diagram of the vortex moving path prediction.

[0127] (7) To explore the influence of the feature parameters of the prediction model on the prediction results of the moving path, the importance of the input variables is evaluated. To evaluate the reliability and accuracy of the CNN-BiLSTM-Attention model in prediction, the model is compared with the LSTM model in terms of prediction effect. The evaluation index is the goodness-of-fit "R"2. The same vortex data set is used, with 70% of the sample data as the training set and the same parameter settings. The input sample is the vortex attribute and trajectory data at the previous three time steps (t-3, t-2, t-1), and the output data is the latitude at t0. The initial learning rate is changed to compare the prediction accuracy of the two models under different initial learning rates. The initial learning rate ranges from "0.5x"10"5" to 0.5, and the goodness-of-fit under different initial learning rates is plotted to evaluate the advantages of the CNN-BiLSTM-Attention model in prediction effect.

[0128] (8) Time step sensitivity experiment: To optimize and evaluate the constructed prediction model, further analysis of the influence of different time steps on the prediction of the next time output vortex movement trajectory position is conducted. Latitude and longitude prediction models are established for different input time steps. The effect of increasing time step on the vortex movement path prediction model is evaluated by listing the correlation coefficient R, root mean square error RMSE, and prediction residual PRD. Based on the principle that the larger the R value, the smaller the RMSE value, and the higher the PRD value, the better the model prediction effect, the optimal input step is determined.

[0129] The second vortex feature element is the prediction of the vortex moving speed.

[0130] The moving speed of the vortex refers to the distance of the vortex center changing in unit time.

[0131] (1) After the vortex is generated, the spatial flow field map of the sea area where the vortex is located is drawn based on the sea current observed by SWOT satellite and ground wave radar, and the initial longitude and latitude of the vortex center position are determined by applying the flow field geometric feature (VG) method.

[0132] (2) From the generation of the vortex, the change in the vortex center position between adjacent observation times (t, t-1) is calculated in the longitude direction and the latitude direction, respectively, which are respectively denoted as:

[0133] △Lon(t, t-1) = Lon(t) - Lon(t-1);

[0134] △Lat(t, t-1) = Lat(t) - Lat(t-1);

[0135] (3) Collect the reanalysis wind field information provided by ECMWF, covering the maximum radius of the vortex sea area, and covering 3 time steps (t-1, t-2, t-3) before the current time t0 and 3 time steps (t+1, t+2, t+3) after the current time, a total of 7 time wind field information. The wind field (wind speed) information obtained is averaged in the longitude and latitude directions, respectively, to obtain the meridional wind speed and the zonal wind speed of the vortex sea area.

[0136] (4) Information entropy is used to measure the average value of self-information brought by the whole random distribution. Based on the collected ECMWF wind field information, the information entropy method is used to calculate the information entropy value of the meridional wind speed and the zonal wind speed in the vortex sea area, and the calculation formula is as follows:

[0137]

[0138] In the above formula, xi corresponds to the component value of the wind speed at different spatial positions in the longitude or latitude direction at the same time, and n represents the number of spatial wind speed values provided by ECMWF in the vortex coverage area.

[0139] For the components of wind speed in longitude and latitude, the change in information entropy in the vortex area between adjacent time steps (t, t-1) is calculated respectively as:

[0140] In the longitude direction: △H Lon (t, t-1) = H Lon (t) - H Lon (t-1);

[0141] In the latitude direction: △H Lat (t, t-1) = H Lat (t) - H Lat (t-1);

[0142] (5) A causal convolutional neural network method is used to construct a prediction model of the moving speed of the vortex. Temporal Convolutional Network (TCN) is a network structure specially designed for time series prediction. It can effectively process time series data and capture long-term dependencies in time series through causal convolutional layers. The core is convolution operation, which extracts features and makes predictions by sliding convolution kernel on time series.

[0143] The convolution operation formula is as follows:

[0144]

[0145] where x[t] is the value of the input time series at time t, ω[k] is the weight of the convolution kernel, K is the size of the convolution kernel, and y[t] is the output after the convolution operation.

[0146] An important property of causal convolution is that the output only depends on the current input and past inputs, not on future inputs. This makes causal convolutional neural networks particularly suitable for time series prediction tasks.

[0147] Extended formula: In a causal convolutional neural network, multiple convolutional layers are usually used to extract features. The output of each convolutional layer serves as the input to the next convolutional layer. By stacking multiple convolutional layers, the network can capture long-term dependencies in the time series.

[0148] Activation function and residual connection: After each convolutional layer, an activation function such as ReLU is usually used to increase the nonlinearity of the network. Residual connections are also commonly used in causal convolutional neural networks to help the network learn deeper levels.

[0149] Prediction formula: When performing time series prediction, the output layer of the network usually uses a linear activation function to obtain the prediction value. The prediction value can be expressed as follows:

[0150]

[0151] where b is the bias term.

[0152] (6) For the moving speed components of the vortex (meridional and latitudinal), a prediction model for the moving speed components of the vortex is constructed. The input variables include the change in the vortex center position between adjacent observation times (t, t-1), the change in the information entropy of the wind speed in the vortex region between adjacent times (t, t-1), and the spatially averaged wind speed and volume at 7 times (t-3, t-2, t-1, t0, t+1, t+2, t+3) in the vortex region. The prediction variables are the meridional and latitudinal moving speed components of the vortex.

[0153] (7) By calculating the correlation coefficient R, the root mean square error RMSE, and the mean absolute error MAE between the prediction results and the observations, based on the principle that the larger the R value, the smaller the RMSE value, and the smaller the MAE value, the better the prediction effect of the model, the model parameters are optimized, and the model is quantitatively evaluated, and the optimal model is selected as the prediction model of the moving speed of the vortex.

[0154] (8) The predicted moving speed of the vortex is visualized on the vortex moving path map through vector annotation, and the spatial points to which the vortex will move at different time intervals (30 minutes, 1 hour, 2 hours, 4 hours, 8 hours, 12 hours, 24 hours, etc.) are calculated.

[0155] Third vortex characteristic element: prediction of vortex life cycle.

[0156] (1) Select typical characteristic parameters of vortex to represent the state of vortex and as input variables, construct network model expressing the relationship between vortex state quantity and life cycle, and then predict and analyze the life cycle of vortex. The selected input variables are divided into two categories: one category is the characteristic parameters related to the state of vortex, reflecting the morphology and dynamic characteristics of vortex, including the age of vortex, vorticity, eddy kinetic energy, deformation rate, instantaneous propagation speed, and the second category is the propagation trajectory parameters of vortex, including the latitude and longitude of vortex center and the displacement distance of current vortex center relative to the initial time. The output variable is the life cycle of vortex. In order to eliminate the influence of different characteristic parameters on the accuracy of the model due to different magnitudes and unit differences, normalization processing is performed on all sample data.

[0157] (2) The prediction of vortex life cycle is constructed in Python residual network model, and realized through Pytortch third-party deep learning library. At the same time, in order to make the network converge faster and better, StandardScaler class in scikit-learn library is used for data preprocessing in the model. Each characteristic parameter is converted to normal distribution data with mean of 0 and standard deviation of 1. ReLUs is used as the nonlinear activation function in the residual block. After debugging, the learning rate is set to 0.0008 and the number of iterations is set to 300.

[0158] (3) In order to compare and analyze the prediction effect of the selected residual network model, residual network model and random forest model are used to model, train and test based on the same sample data. Through repeated debugging and optimization of random forest parameters, the parameters of random forest model are set as follows: the number of decision trees is 950 and the maximum depth is 12.

[0159] (4) Compare the results of random forest and residual network model, calculate the correlation coefficient R between the predicted results and the observations, root mean square error RMSE, and mean absolute error MAE, based on the principle that the larger the R value, the smaller the RMSE value and the MAE value, the better the prediction effect of the model, and carry out quantitative evaluation of the model.

[0160] In summary, the technical features included in this embodiment are as follows:

[0161] (1) Mobile path prediction model: the difference between the moving speeds between two time points is taken as one of the input variables, and the angle a between the path connecting vectors of (t-1, t-2) and (t-2, t-3) is calculated, which is taken as one of the input variables.

[0162] (2) Moving path prediction model: two types of parameters are considered as input variables, one is the vortex attribute characteristic parameter (including absolute value of vorticity, deformation rate and eddy energy), and the other is the vortex position information parameter (including the longitude and latitude of the vortex at different times, the longitude and latitude difference between adjacent times, and the angle of the line connecting the vortex center positions of adjacent times).

[0163] (3) Moving speed prediction model: vortex moving speed component models are constructed in the longitude and latitude directions respectively, and the adjacent time longitude and latitude change and the vortex region information entropy change are considered as input variables of the prediction model.

[0164] (4) Moving speed prediction model: seven time wind speed components averaged in the vortex region space are selected as input variables, and the wind field evolution rate is included in the wind speed component information.

[0165] (5) Life cycle prediction model: the vortex shape and dynamic characteristic information and the vortex propagation trajectory parameters are selected as input variables to construct the prediction model.

[0166] Reference Figure 2 The embodiment of the application further provides a marine vortex prediction device, which can realize the marine vortex prediction method.

[0167] A data acquisition unit is configured to acquire monitoring parameters of a marine vortex by using an unmanned device.

[0168] A first model construction unit is configured to construct a moving path prediction model of a vortex center position according to a time sequence of moving characteristics of the marine vortex in the monitoring parameters.

[0169] A moving path prediction unit is configured to determine a moving path of the marine vortex according to the moving path prediction model.

[0170] A second model construction unit is configured to construct a moving speed prediction model of the marine vortex according to a change of the center position of the marine vortex with time and reanalysis wind field information of a region where the marine vortex is located in the monitoring parameters.

[0171] A moving speed prediction unit is configured to determine a moving speed of the marine vortex according to the moving speed prediction model.

[0172] A third model construction unit is configured to construct a life cycle prediction model of the marine vortex according to vortex state parameters and moving trajectory parameters of the marine vortex in the monitoring parameters.

[0173] A life cycle prediction unit is configured to determine a life cycle of the marine vortex according to the life cycle prediction model.

[0174] It can be understood that the contents in the above method embodiments are all applicable to the present device embodiments, the present device embodiments specifically implement the functions same as those of the above method embodiments, and achieve the same beneficial effects as those of the above method embodiments.

[0175] The present application also provides an electronic device, which comprises a memory and a processor. The memory stores a computer program, and the processor implements the above method for predicting oceanic eddy when executing the computer program. The electronic device can be any intelligent terminal, such as a tablet computer or a vehicle-mounted computer.

[0176] It can be understood that the contents in the above method embodiments are all applicable to the present device embodiments, the present device embodiments specifically implement the functions same as those of the above method embodiments, and achieve the same beneficial effects as those of the above method embodiments.

[0177] Please refer to Figure 3 , Figure 3 The hardware structure of the electronic device of another embodiment is illustrated, which comprises:

[0178] The processor 301 can be implemented in the form of a general-purpose CPU (Central Processing Unit), a microprocessor, an ASIC (Application Specific Integrated Circuit), or one or more integrated circuits, and is used to execute related programs to implement the technical solutions provided by the present application;

[0179] The memory 302 can be implemented in the form of a ROM (Read Only Memory), a static storage device, a dynamic storage device, or a RAM (Random Access Memory). The memory 302 can store an operating system and other application programs. When the technical solutions provided by the present application are implemented by software or firmware, the related program codes are stored in the memory 302 and are called and executed by the processor 301 to implement the method for predicting oceanic eddy according to the present application;

[0180] The input / output interface 303 is used to realize information input and output.

[0181] The communication interface 304 is used to realize the communication interaction between the present device and other devices. The communication can be realized in a wired manner (for example, USB, network cable, etc.) or in a wireless manner (for example, mobile network, WIFI, Bluetooth, etc.).

[0182] A bus 305 transmits information between various components (for example, the processor 301, the memory 302, the input / output interface 303, and the communication interface 304) in the device.

[0183] The processor 301, the memory 302, the input / output interface 303, and the communication interface 304 are communicatively connected to each other within the device through the bus 305.

[0184] The embodiment of the present application further provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to realize the marine eddy prediction method.

[0185] It can be understood that the contents in the above method embodiments are applicable to the storage medium embodiment, the storage medium embodiment specifically realizes the functions of the above method embodiments, and achieves the same beneficial effects as the above method embodiments.

[0186] The memory is a non-transitory computer readable storage medium, and can be used to store non-transitory software programs and non-transitory computer executable programs. In addition, the memory can include a high-speed random access memory, and can also include a non-transitory memory, for example, at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state memory device. In some embodiments, the memory can optionally include a memory remotely arranged relative to the processor, and the remote memory can be connected to the processor through a network. Examples of the above network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0187] The embodiments described in the embodiments of the present application are used to more clearly illustrate the technical solutions of the embodiments of the present application, and do not constitute a limitation on the technical solutions provided by the embodiments of the present application. Those skilled in the art can know that, with the evolution of technology and the appearance of new application scenarios, the technical solutions provided by the embodiments of the present application are also applicable to similar technical problems.

[0188] Those skilled in the art can understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of the present application, and can include more or fewer steps than the figures, or combine certain steps, or different steps.

[0189] The device embodiments described above are only schematic, and the units described as separate components can or can not be physically separate, that is, can be located in one place, or can be distributed on multiple network units. According to actual needs, part or all of the modules can be selected to achieve the purpose of the embodiments of the present application.

[0190] Those skilled in the art can understand that all or some of the steps in the method disclosed above, the function modules / units in the system and the device can be implemented as software, firmware, hardware or appropriate combination thereof.

[0191] The terms "first", "second", "third", "fourth" etc. (if any) in the description of the application and in the claims that follow are used for distinguishing between similar elements and not necessarily for describing a sequential or chronological order. It is to be understood that the use of these terms herein is to be construed to cover the embodiments of the application whether or not the embodiments are described using the same term. Furthermore, the terms "comprise", "comprising", "include", "including", and "has", "having" and variants thereof are to be construed in a non-exclusive manner when used in this description and in the claims that follow. For example, when used in the context of a process, method, system, product or apparatus, the term "comprising" means that the process, method, system, product or apparatus includes the recited steps or units, but can also include additional steps or units not specifically recited.

[0192] It should be understood that, in the present application, "at least one" means one or more, and "multiple" means two or more. "And / or" is used to describe the relationship between associated objects, which means that there can be three relationships, for example, "A and / or B" can mean that there are three cases: only A, only B, and A and B at the same time, where A and B can be singular or plural. The character " / " generally represents an "or" relationship between the associated objects. "At least one of the following" or similar expressions means any combination of these items, including any combination of single or multiple items. For example, at least one of a, b or c can mean a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be singular or plural.

[0193] In several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the above-mentioned units is only a logical function division, and actual implementation can have another division manner, for example, a plurality of units or components can be combined or integrated into another system, or some features can be omitted or not executed. In addition, the coupling or direct coupling or communication connection between the displayed or discussed objects can be indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0194] The units described as separate components above can or can not be physically separate, and the components shown as units can or can not be physical units, that is, can be located in one place, or can be distributed to multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment scheme.

[0195] In addition, each functional unit in each embodiment of the present application can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0196] If the integrated unit is realized in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer readable storage medium. Based on this understanding, the technical scheme of the present application or the part that contributes to the prior art or the whole or part of the technical scheme can be embodied in the form of a software product. The computer software product is stored in a storage medium, including multiple instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method of each embodiment of the present application. The foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various program storage media.

[0197] The preferred embodiments of the embodiments of the present application are described above with reference to the accompanying drawings, but this does not limit the scope of the embodiments of the present application. Any modifications, equivalent replacements and improvements made by those skilled in the art without departing from the scope and essence of the embodiments of the present application shall be within the scope of the embodiments of the present application.

Claims

1. A method for predicting ocean vortexes, characterized in that: The method comprises the following steps: Collect monitoring parameters of ocean eddies through unmanned equipment; A movement path prediction model for the vortex center position is constructed based on the time series of the movement characteristics of the ocean vortex in the monitoring parameters; wherein the movement characteristics include position difference information, angle and vortex attribute characteristic values; a spatial flow field map is drawn based on the ocean currents observed by the SWOT satellite and the ground wave radar; a flow field geometric characteristic method is applied to determine the longitude and latitude of the center point of the clockwise or counterclockwise rotating flow field according to the spatial flow field map as the initial position of the ocean vortex center; the longitude difference or latitude difference between two adjacent moments in the first three moments of each ocean vortex is determined as position difference information; the angle between the path connection vectors of two adjacent moments is calculated; the position difference information, the angle and the vortex attribute characteristic values ​​are determined as the movement characteristics of the ocean vortex; wherein the vortex attribute characteristic values ​​include the absolute value of vorticity, deformation rate and eddy kinetic energy; determining a movement path of the ocean vortex according to the movement path prediction model; A prediction model for the movement speed of the ocean vortex is constructed based on the change in the central position of the ocean vortex over time in the monitoring parameters and the reanalysis wind field information of the area where the ocean vortex is located; wherein the reanalysis wind field information includes the change in the wind speed information entropy in the sea area where the ocean vortex is located at adjacent moments and the wind speed and wind volume of the sea area where the ocean vortex is located at seven moments in space average; determining a moving speed of the ocean vortex according to the moving speed prediction model; A life cycle prediction model for the ocean vortex is constructed based on the vortex state parameters and movement trajectory parameters of the ocean vortex in the monitoring parameters; wherein the age, vorticity, vortex kinetic energy, deformation rate and instantaneous propagation speed of the ocean vortex are used as the vortex state parameters, and the latitude and longitude of the center position of the ocean vortex and the displacement distance of the current center position relative to the initial time are used as the movement trajectory parameters; The life cycle of the ocean vortex is determined according to the life cycle prediction model.

2. A method for predicting ocean vortexes according to claim 1, characterized in that: The method of constructing a movement path prediction model of the vortex center position according to the time series of the movement characteristics of the ocean vortex in the monitoring parameters comprises the following steps: The movement features of the previous three moments are used as input, and the longitude or latitude of the center position of the ocean vortex at the current moment is used as output to construct the movement path prediction model; wherein the movement path prediction model includes a convolutional neural network, a bidirectional long short-term memory network model, and an attention mechanism; Constructing a time series data set according to the movement features; wherein the time series data set includes the movement features at seven consecutive moments; The time series data set is standardized, and then the movement path prediction model is trained using the standardized time series data set.

3. A method for predicting ocean vortexes according to claim 2, characterized in that: Determining the movement path of the ocean vortex according to the movement path prediction model comprises the following steps: Inputting the position difference information, the angle, and the vortex attribute characteristic value of the ocean vortex at any time into the trained movement path prediction model to obtain an output value of the movement path prediction model; The output value is denormalized to obtain the longitude or latitude of the center position of the ocean vortex.

4. A method for predicting ocean vortexes according to claim 1, characterized in that: The method of constructing a prediction model for the movement speed of the ocean vortex according to the change in the center position of the ocean vortex over time in the monitoring parameters and the reanalyzed wind field information in the area where the ocean vortex is located comprises the following steps: Draw spatial flow field maps based on ocean currents observed by SWOT satellites and ground wave radar; Applying a flow field geometric feature method to determine the longitude and latitude of the center point of the clockwise or counterclockwise rotating flow field according to the spatial flow field diagram as the initial longitude and latitude of the center of the ocean vortex; Starting from the initial longitude and latitude, calculating the change in the center position of the ocean vortex between adjacent moments in the longitude and latitude directions respectively; Collect the reanalysis wind field information provided by ECMWF, and then determine the spatially averaged meridional wind speed and volume and zonal wind speed and volume of the sea area where the ocean vortex is located based on the reanalysis wind field information; wherein the area in the reanalysis wind field information covers the sea area where the maximum radius of the ocean vortex is located, and the time series covers seven consecutive moments; Based on the reanalyzed wind field information, the information entropy method is used to calculate the information entropy values ​​of the wind speed and volume in the longitudinal direction and the wind speed and volume in the latitude in the sea area where the ocean vortex is located; The information entropy value is calculated as follows: ； in, represents the information entropy value, Corresponding to the component values ​​of wind speed in the longitude or latitude direction at different spatial locations at the same time, n represents the number of spatial wind speed values ​​provided by ECMWF in the ocean vortex coverage area; The moving speed prediction model is constructed using a causal convolutional neural network method; The expression of the convolution operation in the causal convolutional neural network is: ； in, is the input time series at time The value of is the weight of the convolution kernel, is the size of the convolution kernel, is the output after the convolution operation; The expression of the predicted value of the moving speed prediction model is: ； in, is the predicted value, is the bias term.

5. A method for predicting ocean vortexes according to claim 4, characterized in that: Determining the moving speed of the ocean vortex according to the moving speed prediction model comprises the following steps: The change in the center position of the ocean vortex between adjacent moments, the change in the wind speed information entropy in the sea area where the ocean vortex is located at adjacent moments, and the wind speed and wind volume of the sea area where the ocean vortex is located at 7 spatially averaged moments are input into the moving speed prediction model to obtain the moving speed of the ocean vortex.

6. A method for predicting ocean vortexes according to claim 1, characterized in that: The method of constructing a life cycle prediction model of the ocean vortex according to the vortex state parameters and movement trajectory parameters of the ocean vortex in the monitoring parameters comprises the following steps: Taking the vortex state parameter and the movement trajectory parameter as input parameters and the life cycle of the ocean vortex as output parameters, a residual network model is constructed as a life cycle prediction model; The residual network model is implemented using the Pytorch third-party deep learning library. The StandardScaler class in the scikit-learn library is used in the residual network model to perform data preprocessing on the input parameters to convert each input parameter into normally distributed data with a mean of 0 and a standard deviation of 1. The residual network model uses ReLUs as the nonlinear activation function in the residual block.

7. A method for predicting ocean vortexes according to claim 6, characterized in that: Determining the life cycle of the ocean vortex according to the life cycle prediction model comprises the following steps: The age, vorticity, eddy kinetic energy, deformation rate and instantaneous propagation speed of the ocean vortex are used as the vortex state parameters, and the latitude and longitude of the center position of the ocean vortex and the displacement distance of the current center position relative to the initial moment are used as the movement trajectory parameters to be input into the life cycle prediction model to predict the life cycle of the ocean vortex.

8. An ocean vortex prediction device, characterized in that: The device comprises: A data acquisition unit, used to collect monitoring parameters of ocean eddies through unmanned equipment; A first model building unit is used to build a movement path prediction model of the vortex center position according to the time series of the movement characteristics of the ocean vortex in the monitoring parameters; wherein the movement characteristics include position difference information, angle and vortex attribute characteristic values; draw a spatial flow field map based on the ocean currents observed by the SWOT satellite and the ground wave radar; apply the flow field geometric characteristic method to determine the longitude and latitude of the center point of the clockwise or counterclockwise rotating flow field according to the spatial flow field map as the initial position of the ocean vortex center; determine the longitude difference or latitude difference between two adjacent moments in the first three moments of each ocean vortex as position difference information; calculate the angle between the path connection vectors of two adjacent moments; determine the position difference information, the angle and the vortex attribute characteristic values ​​as the movement characteristics of the ocean vortex; wherein the vortex attribute characteristic values ​​include the absolute value of vorticity, deformation rate and eddy kinetic energy; a movement path prediction unit, configured to determine the movement path of the ocean vortex according to the movement path prediction model; A second model building unit is used to build a prediction model of the movement speed of the ocean vortex based on the change in the center position of the ocean vortex in the monitoring parameters over time and the reanalysis wind field information of the area where the ocean vortex is located; wherein the reanalysis wind field information includes the change in the wind speed information entropy in the sea area where the ocean vortex is located at adjacent moments and the wind speed and wind volume of the sea area where the ocean vortex is located at seven moments in space average; a moving speed prediction unit, configured to determine the moving speed of the ocean vortex according to the moving speed prediction model; a third model building unit, configured to build a life cycle prediction model for the ocean vortex based on the vortex state parameters and movement trajectory parameters of the ocean vortex in the monitoring parameters; wherein the age, vorticity, eddy kinetic energy, deformation rate, and instantaneous propagation velocity of the ocean vortex are used as the vortex state parameters, and the latitude and longitude of the center position of the ocean vortex and the displacement distance of the current center position relative to the initial moment are used as the movement trajectory parameters; A life cycle prediction unit is used to determine the life cycle of the ocean vortex according to the life cycle prediction model.

9. An electronic device, characterized in that: The electronic device includes a memory and a processor, the memory stores a computer program, and the processor implements the method according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images