CN115877377A - A radar network vector flow field synthesis method, system, device and storage medium - Google Patents

Abstract

The invention discloses a method, a system, equipment and a storage medium for synthesizing a radar networking vector flow field, wherein the method comprises the following steps: acquiring radar radial data of a radar network; performing fixed-point observation and navigation observation processing on the coverage area of the radar network to obtain marine observation data; inputting the ocean observation data and the radar radial data into an untrained radar networking model for training to obtain a trained radar networking model; and acquiring radar observation data to be synthesized, and inputting the radar observation data to be synthesized into the trained radar networking model to obtain a synthesized vector flow field. The embodiment of the invention can improve the accuracy of vector flow field synthesis and can be widely applied to the technical field of ground wave radars.

Description

A radar network vector flow field synthesis method, system, device and storage medium

Technical Field

The present invention relates to the field of ground wave radar technology, and in particular to a radar network vector flow field synthesis method, system, equipment and storage medium.

Background Art

Ocean current is one of the most important dynamic parameters in seawater, and accurate observation of sea surface flow field is of great significance to human production and life. Compared with existing fixed-point current meter observations, running observations and satellite remote sensing observations, high-frequency ground wave radar has the advantages of all-weather, large-scale and high temporal and spatial resolution. With the development of ground wave radar technology, radars observing the same sea area have gradually developed from single-station observation of radial flow to dual-base station synthetic vector flow, and then to multi-base station network observation. In related technologies, ground wave radar networking technology usually uses echo data quality to screen observations of different radars at the same point, and then synthesizes vector flow fields. However, this method relies on the quality of echo data to select synthesis, without taking into account the physical process of the ocean flow field, and the accuracy of the synthesized vector flow field is low. In summary, the technical problems existing in related technologies need to be solved urgently.

Summary of the invention

In view of this, an embodiment of the present invention provides a radar network vector flow field synthesis method, system, device and storage medium to improve the accuracy of the synthesized flow field.

On the one hand, the present invention provides a radar network vector flow field synthesis method, comprising:

Obtain radar radial data of radar network;

Performing fixed-point observation and cruise observation processing on the coverage area of the radar network to obtain ocean observation data;

Inputting the ocean observation data and the radar radial data into an untrained radar networking model for training to obtain a trained radar networking model;

The radar observation data to be synthesized is obtained, and the radar observation data to be synthesized is input into the trained radar networking model to obtain a synthesized vector flow field.

Optionally, performing fixed-point observation and cruise observation processing on the coverage area of the radar network to obtain ocean observation data includes:

The ocean observation data includes fixed-point observation data and cruise observation data;

Performing fixed-point observation processing on the areas where the detection ranges overlap in the radar network to obtain fixed-point observation data;

The high-precision area and the edge area in the radar network are subjected to cruise observation processing to obtain cruise observation data.

Optionally, inputting the ocean observation data and the radar radial data into an untrained radar networking model for training to obtain a trained radar networking model includes:

Performing data preprocessing on the ocean observation data and the radar radial data to obtain a training data set;

Synthesize the radar radial data between every two radars in the training data set to obtain a radar synthetic flow field data set;

The radar synthetic flow field data set is input into an untrained radar networking model to obtain a trained radar networking model.

Optionally, the performing data preprocessing on the ocean observation data and the radar radial data to obtain a training data set includes:

Performing spatial gridding processing on the radar radial data according to an inverse distance weighted algorithm to obtain gridded data;

The ocean observation data and the gridded data are averaged hourly to obtain a training data set.

Optionally, synthesizing the radar radial data between every two radars in the training data set to obtain a radar synthetic flow field data set includes:

Acquire first radial data and second radial data, where the first radial data and the second radial data are radar radial data of any two radars in the training data set;

A radar synthetic flow field data set is obtained by calculation according to the radial velocity and azimuth of the first radial data and the radial velocity and azimuth of the second radial data.

Optionally, inputting the radar synthetic flow field data set into an untrained radar networking model to obtain a trained radar networking model includes:

Performing standardization processing on the radar synthetic flow field data set to obtain standardized data;

Performing parameter initialization processing on the untrained radar networking model according to a grid search method to obtain an initialized radar networking model;

The standardized data is input into the initialized radar networking model to obtain a trained radar networking model.

Optionally, inputting the standardized data into the initialized radar networking model to obtain a trained radar networking model includes:

Obtain ocean observation data;

Inputting the standardized data into the initialized radar networking model to obtain an output result;

Calculate the model error according to the output result and the ocean observation data;

The radar networking model is updated according to the model error by using the chain rule to obtain a trained radar networking model.

On the other hand, an embodiment of the present invention further provides a radar network vector flow field synthesis system, including:

The first module is used to obtain radar radial data of the radar network;

The second module is used to perform fixed-point observation and cruise observation processing on the coverage area of the radar network to obtain ocean observation data;

The third module is used to input the ocean observation data and the radar radial data into an untrained radar networking model for training to obtain a trained radar networking model;

The fourth module is used to obtain radar observation data to be synthesized, input the radar observation data to be synthesized into the trained radar networking model, and obtain a synthetic vector flow field.

On the other hand, an embodiment of the present invention further discloses an electronic device, including a processor and a memory;

The memory is used to store programs;

The processor executes the program to implement the method described above.

On the other hand, an embodiment of the present invention further discloses a computer-readable storage medium, wherein the storage medium stores a program, and the program is executed by a processor to implement the method described above.

On the other hand, an embodiment of the present invention further discloses a computer program product or a computer program, which includes a computer instruction stored in a computer-readable storage medium. A processor of a computer device can read the computer instruction from the computer-readable storage medium, and the processor executes the computer instruction, so that the computer device executes the above method.

Compared with the prior art, the present invention adopts the above technical solution and has the following technical effects: the embodiment of the present invention obtains the radar radial data of the radar network; performs fixed-point observation and cruise observation processing on the coverage area of the radar network to obtain ocean observation data; inputs the ocean observation data and the radar radial data into an untrained radar networking model for training to obtain a trained radar networking model; obtains the radar observation data to be synthesized, and inputs the radar observation data to be synthesized into the trained radar networking model to obtain a synthetic vector flow field. This method can improve the accuracy of the synthetic vector flow field through the radar networking model obtained by training with ocean observation data, which is conducive to obtaining a more accurate radar networking vector flow field.

BRIEF DESCRIPTION OF THE DRAWINGS

In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

FIG1 is a flow chart of a radar network vector flow field synthesis method provided in an embodiment of the present application;

FIG2 is a radar network coverage diagram provided by an embodiment of the present application;

FIG3 is a schematic diagram of a synthesis principle of radar radial data provided by an embodiment of the present application;

FIG4 is a framework diagram of a radar networking model provided in an embodiment of the present application.

DETAILED DESCRIPTION

In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

First, some nouns involved in this application are analyzed:

BP neural network: BP neural network is a multi-layer feedforward network trained according to error back propagation, including input layer, hidden layer and output layer. If the actual output of forward propagation does not match the expected output, error back propagation is performed to adjust the weights and biases in the network to change the loss function value.

High-frequency ground wave radar: Ground wave radar is a major means of sea detection. Its detection principle is to use the characteristic of small diffraction propagation attenuation of the conductive ocean surface to emit high-frequency radio waves, which can break through the horizon and detect targets 300 kilometers away with high detection accuracy.

In the related art, the method of synthesizing vector flow field is generally based on bistatic radar, and the signal-to-noise ratio of the echo signal is used as the standard for selecting the networking scheme, and the radars used at different spatial points are selected and synthesized to obtain the vector flow field. However, the applicability of the related vector flow field synthesis method is limited when applied to multi-base radar, and it does not take into account the physical process of the ocean flow field and the constraints of the field observation data on the networking scheme. The accuracy of the synthesized vector flow field is not high.

In view of this, a radar networking vector flow field synthesis method is provided in an embodiment of the present application. The synthesis method in the embodiment of the present application can be applied to a terminal, can be applied to a server, or can be software running in a terminal or a server. The terminal can be a tablet computer, a laptop computer, a desktop computer, etc., but is not limited thereto. The server can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms.

1 , an embodiment of the present invention provides a radar network vector flow field synthesis method, including:

S101, obtaining radar radial data of the radar network;

S102, performing fixed-point observation and cruise observation processing on the coverage area of the radar network to obtain ocean observation data;

S103, inputting the ocean observation data and the radar radial data into an untrained radar networking model for training, to obtain a trained radar networking model;

S104, obtaining radar observation data to be synthesized, and inputting the radar observation data to be synthesized into the trained radar networking model to obtain a synthesized vector flow field.

In an embodiment of the present invention, by acquiring radar radial data of the radar network and ocean observation data obtained by ocean observation technology, the radar radial data and the ocean observation data are input into the radar network model for training, and the optimal parameters combining the radar data and the ocean information are trained by machine learning, so as to minimize the error between the corrected radar echo data and the real ocean information, and obtain the trained radar network model, and synthesize the trained radar network model to obtain a more accurate vector flow field. The radar network in the embodiment of the present invention includes more than two high-frequency ground wave radars, and can be compatible with multiple sets of radars of different frequencies and models.

As a further preferred implementation, the performing of fixed-point observation and cruise observation processing on the coverage area of the radar network to obtain ocean observation data includes:

The ocean observation data includes fixed-point observation data and cruise observation data;

Performing fixed-point observation processing on the areas where the detection ranges overlap in the radar network to obtain fixed-point observation data;

The high-precision area and the edge area in the radar network are subjected to cruise observation processing to obtain cruise observation data.

In an embodiment of the present invention, the ocean observation data obtained can make the synthesized vector flow field data more accurate. By combining the machine learning method, the radar networking model is trained with the ocean observation data as the target, and the weight parameters of the radar networking model are adjusted with error minimization as the objective function, so that the synthesis result is more accurate. In an embodiment of the present invention, the coverage area of the radar network is first fixed-point observation and cruise observation. The fixed-point observation needs to be fixed in the area where the detection range overlaps in the radar network, so as to obtain the fixed-point observation data. Referring to Figure 2, Figure 2 is a radar network coverage diagram provided by an embodiment of the present application, in which DGDA represents the radar coverage of Danggan Island Station, GUIS represents the radar coverage of Guishan Island Station, HEQI represents the radar coverage of Hengqin Station, MWDA represents the radar coverage of Miaowan Island Station, HESD represents the radar coverage of Hengshan Island Station, WSDL represents the radar coverage of Wu Shunde Station, Radar_station represents the radar station position, ADCP represents the acoustic Doppler current profiler, and path represents the route. Cruise observation needs to pass through the high-precision area and edge area in the radar network. The high-precision area refers to the area where the angle between any two radars is greater than 30° and less than 120° or the data acquisition rate is greater than 80%; the edge area refers to the edge position of the coverage area of the radar network. Cruise observation data is obtained through cruise observation technology, and fixed-point observation data and cruise observation data are unified into ocean observation data.

Further as a preferred implementation manner, the inputting the ocean observation data and the radar radial data into an untrained radar networking model for training to obtain a trained radar networking model comprises:

Performing data preprocessing on the ocean observation data and the radar radial data to obtain a training data set;

Synthesize the radar radial data between every two radars in the training data set to obtain a radar synthetic flow field data set;

The radar synthetic flow field data set is input into an untrained radar networking model to obtain a trained radar networking model.

In an embodiment of the present invention, the ocean observation data and the radar radial data are preprocessed, and the temporal and spatial resolutions of the radar radial data and the ocean observation data detected by different radars in the radar network are unified to obtain a training data set. All radar radial data with unified resolution in the training data set are synthesized in pairs to obtain a radar synthetic flow field data set. The radar synthetic flow field data set is input into an untrained radar networking model for training processing, and finally a trained radar networking model is obtained. In an embodiment of the present invention, the radar networking model is constructed using a BP neural network model.

As a further preferred implementation, the data preprocessing of the ocean observation data and the radar radial data to obtain a training data set includes:

Performing spatial gridding processing on the radar radial data according to an inverse distance weighted algorithm to obtain gridded data;

The ocean observation data and the gridded data are averaged hourly to obtain a training data set.

In an embodiment of the present invention, radar radial data and ocean observation data are preprocessed, specifically, by spatially gridding the radar radial data through an inverse distance weighted algorithm, selecting data within the effective area covered by the radar network, that is, selecting an area with a data acquisition rate greater than 60% for gridding, and approximating the coordinates of the radar radial data to the coordinates of the nearest grid point. The spatial resolution of the gridding can be determined according to specific needs to obtain gridded data. In terms of time, the gridded data and ocean observation data are averaged hourly, and the gridded data and ocean observation data are unified into data with a resolution of 1 hour to obtain a training data set.

As a further preferred implementation, synthesizing the radar radial data between every two radars in the training data set to obtain a radar synthetic flow field data set includes:

Acquire first radial data and second radial data, where the first radial data and the second radial data are radar radial data of any two radars in the training data set;

A radar synthetic flow field data set is obtained by calculation according to the radial velocity and azimuth of the first radial data and the radial velocity and azimuth of the second radial data.

In an embodiment of the present invention, a radar synthetic flow field data set is obtained by synthesizing the radar radial data between every two radars in the training data set. Based on the principle of vector projection, the embodiment of the present invention converts the radial flow field between every two radars into a vector flow field, and the synthesis method is shown in FIG3 . In FIG3 , r _v and r _c are radial velocities observed by two radar stations, respectively, θ _v and θ _c are the azimuth angles of radial velocities, Vel is the synthesized vector flow, and u and v are the components of the vector flow in the latitudinal and longitudinal directions. In an embodiment of the present invention, the radar radial data of any two radars in the training data set are referred to as first radial data and second radial data, and the radar synthetic flow field is calculated based on the radial velocity and azimuth angle of the first radial data and the radial velocity and azimuth angle of the second radial data, thereby obtaining a radar synthetic flow field data set by synthesizing all the data in the training data set. The synthesis calculation formula is as follows:

Wherein, u represents the latitudinal component of the resultant vector flow, v represents the meridional component of the resultant vector flow, r _v and _rc represent the radial velocities of the first radial data and the second radial data, respectively, and θ _v and θ _c represent the direction angles of the radial velocities of the first radial data and the second radial data, respectively.

Further as a preferred implementation manner, the step of inputting the radar synthetic flow field data set into an untrained radar networking model to obtain a trained radar networking model includes:

Performing standardization processing on the radar synthetic flow field data set to obtain standardized data;

Performing parameter initialization processing on the untrained radar networking model according to a grid search method to obtain an initialized radar networking model;

The standardized data is input into the initialized radar networking model to obtain a trained radar networking model.

In an embodiment of the present invention, the radar synthetic flow field data set is input into the machine learning model for training. The radar networking model is constructed by the BP neural network in the embodiment of the present invention. As shown in FIG4 , the radar networking model is obtained by training. The embodiment of the present invention first performs standardization processing on the radar synthetic flow field data set to obtain standardized data. The formula of the standardization method is as follows:

Where _Xm represents the standardized data, _Xoriginal represents the data in the radar synthetic flow field dataset, _Xmin represents the minimum value in the radar synthetic flow field dataset, and _Xmax represents the maximum value in the radar synthetic flow field dataset.

The embodiment of the present invention performs parameter initialization processing on the untrained radar networking model by means of a grid search method, and the empirical formula used is as follows:

Among them, p, m and n are the number of neurons in the hidden layer, input layer and output layer respectively, and q is a constant between 1 and 10.

In the embodiment of the present invention, an initialized radar networking model is obtained by initializing through a grid search method, and standardized data after standardization processing is input into the initialized radar networking model for training to obtain a trained radar networking model.

Further as a preferred implementation manner, the inputting the standardized data into the initialized radar networking model to obtain a trained radar networking model includes:

Obtain ocean observation data;

Inputting the standardized data into the initialized radar networking model to obtain an output result;

Calculate the model error according to the output result and the ocean observation data;

The radar networking model is updated according to the model error by using the chain rule to obtain a trained radar networking model.

As a further preferred embodiment, referring to FIG. 4 , the standardized data _Xm is input into the input layer of the initialized radar networking model, wherein the standardized data is all synthetic flow fields in the radar synthetic flow field data set, and Xmn represents the radar synthetic flow field synthesized by the mth radar and the nth radar in the radar synthetic flow field data set. The calculation is performed through the hidden layer in the radar synthesis model, and the output formula of the hidden layer is as follows:

为隐藏层U_n第j个神经元的阈值。Where _Uj is the output of the jth neuron in the hidden layer, f() is the mapping of the neuron activation function, and _vij is the weight of the i-th input variable _Xi and the j-th hidden layer neuron _Uj .

is the threshold of the jth neuron in the hidden layer _Un .

The output result of the hidden layer is input into the output layer Y of the radar networking model. The output formula of the output layer is as follows:

Where wj is the weight of the connection between the jth neuron and the output layer, Θy is the threshold of the output layer neuron, and y is the output result of the output layer.

The embodiment of the present invention calculates the error between the output result of the radar networking model and the ocean observation data through the selected objective function, and after updating the weight value of each layer in the radar networking model according to the chain rule, the output value is calculated again, and the training is repeated until the specified number of times or the objective function converges, and the training is stopped to obtain the radar networking model. Among them, the objective function is measured by the loss function, which is defined on a single training data and is used to measure the prediction error of a training data. Specifically, the loss value of the training data is determined by the label of a single training data and the prediction result of the training data by the model. In actual training, a training data set has a lot of training data, so the cost function is generally used to measure the overall error of the training data set. The cost function is defined on the entire training data set and is used to calculate the average value of the prediction error of all training data, which can better measure the prediction effect of the model. For a general machine learning model, based on the aforementioned cost function, plus the regularization term that measures the complexity of the model, it can be used as the objective function of the training, and the loss value of the entire training data set can be calculated based on the objective function. There are many types of commonly used loss functions, such as 0-1 loss function, square loss function, absolute loss function, logarithmic loss function, cross entropy loss function, etc., which can all be used as loss functions of machine learning models, which will not be elaborated one by one here. In an embodiment of the present invention, any loss function can be selected to determine the loss value of training. Based on the loss value of the training, the parameters of the model are updated by the back propagation algorithm, and a trained radar networking model can be obtained after several rounds of iteration. Specifically, the number of iterations can be set in advance, or the training is considered to be completed when the test set meets the accuracy requirements. In an embodiment of the present invention, a radar networking model can be built based on a BP neural network. The radar networking model obtained by training converts the radial flow field observed by multiple radars into a complete vector flow field, which can improve the accuracy of vector flow field synthesis.

On the other hand, an embodiment of the present invention further provides a radar network vector flow field synthesis system, including:

The first module is used to obtain radar radial data of the radar network;

The second module is used to perform fixed-point observation and cruise observation processing on the coverage area of the radar network to obtain ocean observation data;

The third module is used to input the ocean observation data and the radar radial data into an untrained radar networking model for training to obtain a trained radar networking model;

The fourth module is used to obtain radar observation data to be synthesized, input the radar observation data to be synthesized into the trained radar networking model, and obtain a synthetic vector flow field.

Corresponding to the method of FIG. 1 , an embodiment of the present invention further provides an electronic device, including a processor and a memory; the memory is used to store a program; the processor executes the program to implement the method described above.

Corresponding to the method of FIG. 1 , an embodiment of the present invention further provides a computer-readable storage medium, wherein the storage medium stores a program, and the program is executed by a processor to implement the method described above.

The embodiment of the present invention also discloses a computer program product or a computer program, which includes a computer instruction stored in a computer-readable storage medium. A processor of a computer device can read the computer instruction from the computer-readable storage medium, and the processor executes the computer instruction, so that the computer device executes the method shown in FIG1.

In summary, the embodiments of the present invention have the following advantages: the embodiments of the present invention use a machine learning method combined with an algorithm for synthesizing vector currents constrained by ocean observation data, and take minimizing the error between ocean observation data and radar data as the objective function, rather than only selecting a radar for synthesizing vector flow fields based on the quality of radar echo data. This fully considers the actual ocean environment and physical processes, thereby improving the accuracy of vector flow field synthesis.

In some selectable 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 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. Selectable embodiments are expected, wherein the order of various operations is changed and the sub-operation of a part for which is described as a larger operation is performed independently.

In addition, although the present invention is described in the context of functional modules, it should be understood that, unless otherwise specified, one or more of the functions and/or features described may be integrated into a single physical device and/or software module, or one or more functions and/or features may be implemented in separate physical devices or software modules. 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.

If the 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 enabling 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 methods described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

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.

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 programmable 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 program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

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.

In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" 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 representations of the above terms do 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.

Although the 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.

The above is a specific description of the preferred implementation of the present invention, but the present invention is not limited to the described 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 (10)

1. A radar network vector flow field synthesis method, characterized in that the method comprises: Obtain radar radial data of radar network; Performing fixed-point observation and cruise observation processing on the coverage area of the radar network to obtain ocean observation data; Inputting the ocean observation data and the radar radial data into an untrained radar networking model for training to obtain a trained radar networking model; The radar observation data to be synthesized is obtained, and the radar observation data to be synthesized is input into the trained radar networking model to obtain a synthesized vector flow field. 2. The method according to claim 1, characterized in that the performing of fixed-point observation and cruise observation processing on the coverage area of the radar network to obtain ocean observation data comprises: The ocean observation data includes fixed-point observation data and cruise observation data; Performing fixed-point observation processing on the areas where the detection ranges overlap in the radar network to obtain fixed-point observation data; The high-precision area and the edge area in the radar network are subjected to cruise observation processing to obtain cruise observation data. 3. The method according to claim 1, characterized in that the step of inputting the ocean observation data and the radar radial data into an untrained radar networking model for training to obtain a trained radar networking model comprises: Performing data preprocessing on the ocean observation data and the radar radial data to obtain a training data set; Synthesize the radar radial data between every two radars in the training data set to obtain a radar synthetic flow field data set; The radar synthetic flow field data set is input into an untrained radar networking model to obtain a trained radar networking model. 4. The method according to claim 3, characterized in that the preprocessing of the ocean observation data and the radar radial data to obtain a training data set comprises: Performing spatial gridding processing on the radar radial data according to an inverse distance weighted algorithm to obtain gridded data; The ocean observation data and the gridded data are averaged hourly to obtain a training data set. 5. The method according to claim 3, characterized in that the step of synthesizing the radar radial data between every two radars in the training data set to obtain a radar synthetic flow field data set comprises: Acquire first radial data and second radial data, where the first radial data and the second radial data are radar radial data of any two radars in the training data set; A radar synthetic flow field data set is obtained by calculation according to the radial velocity and azimuth of the first radial data and the radial velocity and azimuth of the second radial data. 6. The method according to claim 3, characterized in that the step of inputting the radar synthetic flow field data set into an untrained radar networking model to obtain a trained radar networking model comprises: Performing standardization processing on the radar synthetic flow field data set to obtain standardized data; Performing parameter initialization processing on the untrained radar networking model according to a grid search method to obtain an initialized radar networking model; The standardized data is input into the initialized radar networking model to obtain a trained radar networking model. 7. The method according to claim 6, characterized in that the step of inputting the standardized data into the initialized radar networking model to obtain a trained radar networking model comprises: Obtain ocean observation data; Inputting the standardized data into the initialized radar networking model to obtain an output result; Calculate the model error according to the output result and the ocean observation data; The radar networking model is updated according to the model error by using the chain rule to obtain a trained radar networking model. 8. A radar network vector flow field synthesis system, characterized in that the system comprises: The first module is used to obtain radar radial data of the radar network; The second module is used to perform fixed-point observation and cruise observation processing on the coverage area of the radar network to obtain ocean observation data; The third module is used to input the ocean observation data and the radar radial data into an untrained radar networking model for training to obtain a trained radar networking model; The fourth module is used to obtain radar observation data to be synthesized, input the radar observation data to be synthesized into the trained radar networking model, and obtain a synthetic vector flow field. 9. An electronic device, characterized in that the electronic device comprises a memory and a processor; The memory is used to store programs; The processor executes the program to implement the method according to any one of claims 1 to 7. 10. A computer-readable storage medium storing a computer program, wherein the computer program implements the method according to any one of claims 1 to 7 when executed by a processor.

Images