Ground wave radar data quality control method based on CNN-LSTM model

Abstract

This invention discloses a ground wave radar data quality control method based on a CNN-LSTM model. The method includes: determining the radar detection data range affected by terrain based on the accuracy differences of radar echo data; training a CNN-LSTM neural network based on the radar detection data range to obtain a target model; and correcting the radar data in the terrain-affected area based on the target model to obtain corrected radar data. This invention can combine actual marine environmental information to correct radar detection data affected by coastlines, islands, and other terrain features, improving the accuracy of ocean current direction in synthesized radar detection data. It can be widely applied in the field of computer technology.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to a method for controlling the data quality of ground wave radar based on a CNN-LSTM model. Background Technology

[0002] Currently, ocean current data is primarily acquired through numerical simulations and actual observations. While numerical simulations can model ocean current conditions across an entire sea area in real time, their accuracy is relatively low. Actual point-based observations, though accurate, often only provide data for a single point or vertical profile, making it difficult to obtain observational data for large areas of the sea, and are easily affected by environmental factors. High-frequency ground-wave radar (GWR) offers advantages such as beyond-line-of-sight capability, wide coverage, all-weather operation, and low cost, and is less affected by weather. The use of GWR to invert and calculate ocean currents is constantly being improved and is gradually maturing. Although existing technologies can obtain highly accurate ocean current information in high-precision areas observed by GWR, traditional calculation methods are based on radar echo information and do not consider the influence of the actual marine environment. Therefore, the area of ​​this high-precision area is much smaller than the actual sea area acquired by the radar.

[0003] In the field of acquiring ocean current data using high-frequency ground wave radar data, the following main problems exist:

[0004] 1. Most existing inversion algorithms are based on electronic information theory, which has certain limitations. In coastal areas, radar electromagnetic waves are easily interfered with by the coastline and islands, the seabed topography is also relatively complex, and factors such as tides, runoff, and wind also have an impact. Radar electromagnetic echoes cannot effectively analyze the interaction between these physical factors, so the accuracy of ocean current information obtained using electronic information theory is relatively low.

[0005] 2. Current technologies for correcting radar data using neural networks mostly use echo parameters from high-precision radar zones in their input layers, neglecting marine environmental information such as the physical processes influencing ocean currents. A few incorporate tides and wind fields, but when correcting radar data in topographically affected areas, the lack of consideration for topographic influence and extraction of spatial features results in slightly lower correction accuracy. Furthermore, each data point requires a separate neural network, making it impossible to apply a single trained neural network to the correction of radar data across the entire affected area, thus reducing its practicality. Summary of the Invention

[0006] The main objective of this invention is to propose a ground wave radar data quality control method based on a CNN-LSTM model, which can improve the accuracy of ocean current direction in radar detection data synthesis.

[0007] To achieve the above objectives, one aspect of this invention proposes a ground wave radar data quality control method based on a CNN-LSTM model, comprising the following steps:

[0008] Based on the differences in the accuracy of radar echo data, the range of radar detection data affected by terrain is determined;

[0009] Based on the radar detection data range, a CNN-LSTM neural network is trained to obtain a target model;

[0010] The radar data of the terrain-affected area is corrected based on the target model to obtain the corrected radar data.

[0011] In some embodiments, determining the radar detection data range affected by terrain based on the accuracy difference of radar echo data includes the following steps:

[0012] Based on the differences in the accuracy of radar echo data, data is detected in radar detection areas near coastlines and islands;

[0013] Regions where the elliptic direction angle error of the empirical orthogonal function of the synthetic ocean current is greater than 40° are identified as radar detection data ranges affected by terrain.

[0014] In some embodiments, training a CNN-LSTM neural network to obtain a target model based on the radar detection data range includes the following steps:

[0015] An initial model is built based on the CNN-LSTM neural network;

[0016] A 9×9 matrix centered at the data points to be corrected is used as the input data for the CNN-LSTM neural network;

[0017] The training cycle of the CNN-LSTM neural network was set to 200 rounds, and the weight parameters of the network model were updated using the Adam optimization algorithm. The initial learning rate was set to 0.001.

[0018] During the training of the network model, the number of convolutional layers, the size of the convolutional kernel, and the number of hidden layers in the LSTM module are adjusted.

[0019] The final network structure is determined, and all input training data are used as the training set to train the neural network. Training ends when the target minimum error is reached and remains relatively stable, generating a radar data correction model as the target model.

[0020] In some embodiments, the training process of the CNN-LSTM neural network further includes the following steps:

[0021] Based on the EOF ellipse orientation angle error, the average flow orientation angle error, and the EOF ellipse major and minor axis errors, evaluation indicators are determined, and the target model is validated; this step specifically involves:

[0022] A data matrix is ​​constructed, with the first and second columns representing the u and v components, respectively. Then, the velocity matrix is ​​decomposed into a first mode and a second mode through EOF analysis. The major axis of the ellipse represents the first mode, which is the eigenvalue of the maximum standard deviation of the velocity, and the minor axis represents the second mode, which is the eigenvalue of the maximum standard deviation of the velocity.

[0023] The direction angle of an ellipse is calculated using the formula. The calculation yields the feature vectors of the first and second modes, respectively:

[0024] The formulas for calculating the EOF elliptical direction angle error and the average flow direction angle error are as follows:

[0025] Δθ=|(θ1-θ2+180°)mod360°-180°|, where θ1 and θ2 represent the angles of the EOF ellipse direction or the average flow direction of the observed and corrected values, respectively, mod is the modulo operation, and Δθ is the error angle;

[0026] The formula for calculating the major and minor axis error E of the EOF ellipse is:

[0027] Among them, Lmaj radar The major axis of the EOF ellipse representing radar data; Lmaj observation The major axis of the EOF ellipse representing the observed data; Lmin radar The minor axis of the ellipse representing the EOF of radar data; Lmin observation The minor axis of the ellipse EF represents the observed data.

[0028] In some embodiments, the initial model established based on the CNN-LSTM neural network includes the following steps:

[0029] Features are extracted from Convolutional Neural Networks (CNNs) through convolution operations. The convolution operation extracts local features step-by-step by sliding a convolution kernel across the input data. The formula for feature extraction is as follows: Where Y(x, y) represents the value at position (x, y) in the feature matrix extracted by the integrative neural network; r is the ReLU activation function; the size of the convolution kernel is (m×n); I(x+i, y+j) is the value at position (i, j) of the convolution kernel in the input data matrix; K(i, j) is the value at position (i, j) in the convolution kernel; and b is the bias term.

[0030] The formula for calculating average pooling is as follows:

[0031]

[0032] Where O(x, y) represents the value at position (x, y) in the output feature matrix; the size of the pooling window is (m×n); s is the step size, representing the step size of each sliding of the pooling window; Y(x·s+i, y·s+j) represents the elements in the input feature map covered by the pooling window;

[0033] The formula for calculating the network parameters of an LSTM neural network is as follows:

[0034] f t =σ(W f ·[h t-1 x t ]+b f )

[0035] i t =σ(W i ·[h t-1 x t ]+b i )

[0036]

[0037] o t =σ(W o ·[h t-1 x t ]+b o )

[0038] h t =o t *tanh(C t )

[0039] Where σ represents the sigmoid activation function; tanh represents the hyperbolic tangent activation function; · represents the dot product; * represents the Hadamard product; t is the time step; x t It is the input data at time t; h t It is the hidden state at time t; Indicates the candidate cell state; C t f is the cell state at time t; t It's a forget gate, which determines the state from cell C. t What information is discarded? t It is the input gate, which controls how much information about the candidate state at the current moment needs to be saved into the cell state; t It is an output gate, and the output gate controls the current state C. t How much information is output to the hidden state h? t W f W iW c and W o These are the weights of the corresponding gates and units; b f b i b c and b o These are the bias terms corresponding to gates and units;

[0040] The formula for calculating the loss function used during neural network training is:

[0041]

[0042] Where N is the number of samples; y i Let be the value of the observation on the i-th sample; Let be the predicted value for the i-th sample.

[0043] In some embodiments, using a 9×9 matrix centered at the data points to be corrected as input data for the CNN-LSTM neural network includes the following steps:

[0044] Based on the observed ocean current u and v component composite velocity, tidal current velocity u component, tidal current velocity v component, wind speed u component, wind speed v component, and altitude data, an input data matrix for the CNN module is constructed. Each batch of input data contains matrix data for 10 time steps: t, t+1, ..., t+9. After two convolution and pooling operations, the input matrix data is flattened and input into the LSTM module. The output of the LSTM module is combined with the flattened altitude data, and finally, the corrected u and v components for 10 time steps are output through a fully connected layer (FC).

[0045] The input data is normalized, and the expression for the normalization process is:

[0046]

[0047] Among them, X i Represents the normalization result; X ioriginal Represents the raw data in the input field; X max X represents the maximum value of the original data; min This represents the minimum value of the original data;

[0048] The normalized input training data is randomly divided into two parts, 80% and 20%, which are used as the training set and the test set, respectively, to train network parameters and test network performance.

[0049] In some embodiments, the step of correcting the radar data of the terrain-affected area according to the target model to obtain corrected radar data specifically involves:

[0050] The radar data of the terrain-affected area is corrected according to the target model. The input data matrix during correction includes the composite velocity of ocean current u and v components, tidal current velocity u component, tidal current velocity v component, wind speed u component, wind speed v component, and altitude data obtained by radar detection. Finally, the corrected radar data is obtained.

[0051] Another aspect of this invention provides a ground wave radar data quality control system based on a CNN-LSTM model, comprising:

[0052] The first module is used to determine the range of radar detection data affected by terrain based on the accuracy differences of radar echo data.

[0053] The second module is used to train the CNN-LSTM neural network based on the radar detection data range to obtain the target model;

[0054] The third module is used to correct the radar data of the terrain-affected area according to the target model, so as to obtain the corrected radar data.

[0055] To achieve the above objectives, another aspect of the present invention provides an electronic device, which includes a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the method described above.

[0056] To achieve the above objectives, another aspect of the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the methods described above.

[0057] This invention also discloses a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device can read the computer instructions from the computer-readable storage medium and execute the computer instructions, causing the computer device to perform the aforementioned method.

[0058] The embodiments of this invention include at least the following beneficial effects: This invention provides a ground wave radar data quality control method based on a CNN-LSTM model. This method determines the radar detection data range affected by terrain based on the accuracy differences of radar echo data; trains a CNN-LSTM neural network based on the radar detection data range to obtain a target model; and corrects the radar data in the terrain-affected area based on the target model to obtain corrected radar data. The embodiments of this invention can combine actual marine environmental information to correct radar detection data affected by terrain such as coastlines and islands, improving the accuracy of ocean current direction in the synthesized radar detection data. Attached Figure Description

[0059] Figure 1 This is a schematic diagram of an implementation environment provided by an embodiment of the present invention;

[0060] Figure 2 This is a flowchart of the overall steps provided in the embodiments of the present invention;

[0061] Figure 3 This is a flowchart illustrating the specific implementation steps provided in the embodiments of the present invention;

[0062] Figure 4 This is a flowchart of the steps for training a CNN-LSTM neural network provided in an embodiment of the present invention;

[0063] Figure 5 This is a schematic diagram of the framework structure of the network model provided in an embodiment of the present invention;

[0064] Figure 6 This is a schematic diagram of the LSTM neural network structure provided in an embodiment of the present invention;

[0065] Figure 7 This is a schematic diagram of the hardware structure of the electronic device provided in an embodiment of the present invention. Detailed Implementation

[0066] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. In the following description, when referring to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the embodiments of this invention; they are merely examples of apparatuses and methods consistent with some aspects of the embodiments of this invention as detailed in the appended claims.

[0067] It is understood that the terms “first,” “second,” etc., used in this invention may be used herein to describe various concepts, but unless specifically stated otherwise, these concepts are not limited by these terms. These terms are used only to distinguish one concept from another. For example, first information may also be referred to as second information without departing from the scope of embodiments of the invention, and similarly, second information may also be referred to as first information. Depending on the context, the words “if,” “when,” or “in response to determination” as used herein may be interpreted as “when…” or “when…” or “in response to determination.”

[0068] The terms “at least one,” “multiple,” “each,” “any,” etc., used in this invention, “at least one” includes 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.

[0069] 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 invention pertains. The terminology used herein is for the purpose of describing embodiments of the invention only and is not intended to limit the invention.

[0070] Before providing a detailed description of the embodiments of the present invention, some related technologies involved in the embodiments of the present invention will be described first, as follows:

[0071] The CNN-LSTM neural network model combines convolutional neural networks (CNNs) and long short-term memory (LSTM) networks. CNNs are feedforward neural networks that incorporate convolutional computations, effectively extracting local features from input data using convolution and pooling operations. They are primarily used to process spatially structured data (such as images and videos). LSTMs are a special type of recurrent neural network (RNN) mainly used to process time-series data, addressing the vanishing and exploding gradient problems that traditional RNNs encounter when processing long sequences. The CNN-LSTM model combines the spatial feature extraction advantages of CNNs with the temporal feature capture advantages of LSTMs, making it suitable for processing complex spatiotemporal data.

[0072] Deep learning: Deep learning is a new research direction in the field of machine learning. It discovers distributed feature representations of data by combining low-level features to form more abstract high-level representations of attribute categories or features.

[0073] Ground wave radar: Ground wave radar is a major means of sea surface detection. Its detection principle is based on the characteristic that the propagation of conductive ocean surfaces has low diffraction attenuation. It emits high-frequency radio waves that can penetrate the horizon and detect targets up to 300 kilometers away, with high detection accuracy.

[0074] The ground wave radar data quality control method based on a CNN-LSTM model provided in this invention relates to the field of computer technology. This method can be applied to terminals, servers, or software running on either a terminal or a server. In some embodiments, the terminal can be a smartphone, tablet, laptop, desktop computer, smart speaker, smartwatch, or vehicle terminal, but is not limited to these. The server can be configured as an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms. The server can also be a node server in a blockchain network. The software can be an application implementing the ground wave radar data quality control method based on a CNN-LSTM model, but is not limited to the above forms.

[0075] This invention can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. Examples include: 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, and distributed computing environments including any of the above systems or devices. This invention can be described in the general context of computer-executable instructions, such as program modules, that are executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This invention can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0076] like Figure 1 The diagram shown is a schematic representation of an implementation environment provided by an embodiment of the present invention. (Refer to...) Figure 1 The implementation environment includes at least one terminal 102 and a server 101. The terminal 102 and the server 101 can be connected via a network, either wirelessly or via a wired connection, to complete data transmission and exchange.

[0077] Server 101 can be a standalone physical server, 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 communication, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms.

[0078] Additionally, server 101 can also be a node server in a blockchain network. Blockchain is a novel application model of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanisms, and encryption algorithms.

[0079] Terminal 102 can be a smartphone, tablet, laptop, desktop computer, smart speaker, smartwatch, etc. It can also be a vehicle-mounted terminal of the various device types described above, but is not limited to these. Terminal 102 and server 101 can be directly or indirectly connected via wired or wireless communication, and this embodiment of the invention does not impose any limitations.

[0080] Exemplary based on Figure 1 The implementation environment shown in this embodiment of the invention provides a ground wave radar data quality control method based on a CNN-LSTM model. The following description uses the application of this ground wave radar data quality control method based on a CNN-LSTM model in server 101 as an example. It can be understood that this method can also be applied to terminal 102.

[0081] Reference Figure 2 , Figure 2 The flowchart illustrates a method for controlling the quality of ground wave radar data based on a CNN-LSTM model applied to a server, as provided in this embodiment of the invention. The execution entity of this method can be any of the aforementioned computer devices (including servers or terminals). (Refer to...) Figure 2 The method may include the following steps:

[0082] Based on the differences in the accuracy of radar echo data, the range of radar detection data affected by terrain is determined;

[0083] Based on the radar detection data range, a CNN-LSTM neural network is trained to obtain a target model;

[0084] The radar data of the terrain-affected area is corrected based on the target model to obtain the corrected radar data.

[0085] In some embodiments, determining the radar detection data range affected by terrain based on the accuracy difference of radar echo data includes the following steps:

[0086] Based on the differences in the accuracy of radar echo data, data is detected in radar detection areas near coastlines and islands;

[0087] Regions where the elliptic direction angle error of the empirical orthogonal function of the synthetic ocean current is greater than 40° are identified as radar detection data ranges affected by terrain.

[0088] In some embodiments, training a CNN-LSTM neural network to obtain a target model based on the radar detection data range includes the following steps:

[0089] An initial model is built based on the CNN-LSTM neural network;

[0090] A 9×9 matrix centered at the data points to be corrected is used as the input data for the CNN-LSTM neural network;

[0091] The training cycle of the CNN-LSTM neural network was set to 200 rounds, and the weight parameters of the network model were updated using the Adam optimization algorithm. The initial learning rate was set to 0.001.

[0092] During the training of the network model, the number of convolutional layers, the size of the convolutional kernel, and the number of hidden layers in the LSTM module are adjusted.

[0093] The final network structure is determined, and all input training data are used as the training set to train the neural network. Training ends when the target minimum error is reached and remains relatively stable, generating a radar data correction model as the target model.

[0094] In some embodiments, the training process of the CNN-LSTM neural network further includes the following steps:

[0095] Based on the EOF ellipse orientation angle error, the average flow orientation angle error, and the EOF ellipse major and minor axis errors, evaluation indicators are determined, and the target model is validated; this step specifically involves:

[0096] A data matrix is ​​constructed, with the first and second columns representing the u and v components, respectively. Then, the velocity matrix is ​​decomposed into a first mode and a second mode through EOF analysis. The major axis of the ellipse represents the first mode, which is the eigenvalue of the maximum standard deviation of the velocity, and the minor axis represents the second mode, which is the eigenvalue of the maximum standard deviation of the velocity.

[0097] The direction angle of an ellipse is calculated using the formula. The calculation yields the feature vectors of the first and second modes, respectively:

[0098] The formulas for calculating the EOF elliptical direction angle error and the average flow direction angle error are as follows:

[0099] Δθ=|(θ1-θ2+180°)mod360°-180°|, where θ1 and θ2 represent the angles of the EOF ellipse direction or the average flow direction of the observed and corrected values, respectively, mod is the modulo operation, and Δθ is the error angle;

[0100] The formula for calculating the major and minor axis error E of the EOF ellipse is:

[0101] Among them, Lmaj radar The major axis of the EOF ellipse representing radar data; Lmaj observvation The major axis of the EOF ellipse representing the observed data; Lmin radar The minor axis of the ellipse representing the EOF of radar data; Lmin observation The minor axis of the EOF ellipse representing the observed data.

[0102] In some embodiments, the initial model established based on the CNN-LSTM neural network includes the following steps:

[0103] Features are extracted from Convolutional Neural Networks (CNNs) through convolution operations. The convolution operation extracts local features step-by-step by sliding a convolution kernel across the input data. The formula for feature extraction is as follows: Where Y(x, y) represents the value at position (x, y) in the feature matrix extracted by the integrative neural network; r is the ReLU activation function; the size of the convolution kernel is (m×n); I(x+i, y+j) is the value at position (i, j) of the convolution kernel in the input data matrix; K(i, j) is the value at position (i, j) in the convolution kernel; and b is the bias term.

[0104] The formula for calculating average pooling is as follows:

[0105]

[0106] Where O(x, y) represents the value at position (x, y) in the output feature matrix; the size of the pooling window is (m×n); s is the step size, representing the step size of each sliding of the pooling window; Y(x·s+i, y·s+j) represents the elements in the input feature map covered by the pooling window;

[0107] The formula for calculating the network parameters of an LSTM neural network is as follows:

[0108] f t =σ(W f ·[h t-1 x t ]+b f )

[0109] i t =σ(Wi ·[h t-1 x t ]+b i )

[0110]

[0111] o t =σ(W o ·[h t-1 x t ] +b o )

[0112] h t =o t *tanh(C t )

[0113] Where σ represents the sigmoid activation function; tanh represents the hyperbolic tangent activation function; · represents the dot product; * represents the Hadamard product; t is the time step; x t It is the input data at time t; h t It is the hidden state at time t; Indicates the candidate cell state; C t f is the cell state at time t; t It's a forget gate, which determines the state from cell C. t What information is discarded? t It is the input gate, which controls how much information about the candidate state at the current moment needs to be saved into the cell state; t It is an output gate, and the output gate controls the current state C. t How much information is output to the hidden state h? t W f W i W c and W o These are the weights of the corresponding gates and units; b f b i b c and b o These are the bias terms corresponding to gates and units;

[0114] The formula for calculating the loss function used during neural network training is:

[0115]

[0116] Where N is the number of samples; y i Let be the value of the observation on the i-th sample; Let be the predicted value for the i-th sample.

[0117] In some embodiments, using a 9×9 matrix centered at the data points to be corrected as input data for the CNN-LSTM neural network includes the following steps:

[0118] Based on the observed ocean current u and v component composite velocity, tidal current velocity u component, tidal current velocity v component, wind speed u component, wind speed v component, and altitude data, an input data matrix for the CNN module is constructed. Each batch of input data contains matrix data for 10 time steps: t, t+1, ..., t+9. After two convolution and pooling operations, the input matrix data is flattened and input into the LSTM module. The output of the LSTM module is combined with the flattened altitude data, and finally, the corrected u and v components for 10 time steps are output through a fully connected layer (FC).

[0119] The input data is normalized, and the expression for the normalization process is:

[0120]

[0121] Among them, X i Represents the normalization result; X ioriginal Represents the raw data in the input field; X max X represents the maximum value of the original data; min This represents the minimum value of the original data;

[0122] The normalized input training data is randomly divided into two parts, 80% and 20%, which are used as the training set and the test set, respectively, to train network parameters and test network performance.

[0123] In some embodiments, the step of correcting the radar data of the terrain-affected area according to the target model to obtain corrected radar data specifically involves:

[0124] The radar data of the terrain-affected area is corrected according to the target model. The input data matrix during correction includes the composite velocity of ocean current u and v components, tidal current velocity u component, tidal current velocity v component, wind speed u component, wind speed v component, and altitude data obtained by radar detection. Finally, the corrected radar data is obtained.

[0125] The specific implementation process of this invention will be described in detail below using a specific application scenario as an example:

[0126] refer to Figure 3 The specific implementation steps of the present invention include the following steps:

[0127] Step 1: Based on the accuracy differences of radar echo data, in radar detection areas near coastlines and islands, areas where the ellipse direction angle error of the empirical orthogonal function (EOF) synthesized ocean current is greater than 40° are determined to be the radar detection data range affected by topography.

[0128] Step 2: Train the CNN-LSTM neural network, refer to... Figure 4 and Figure 5 The specific steps are as follows:

[0129] Step 0: Modeling the CNN-LSTM neural network. Feature extraction in the Convolutional Neural Network (CNN) primarily relies on convolution operations. Convolution operations extract local features step-by-step by sliding a convolution kernel across the input data. The formula for feature extraction is as follows:

[0130]

[0131] In the formula, Y( x， y) represents the feature matrix extracted by the multiplicative neural network ( x， The value at position (i,j) is given by r; the ReLU activation function is given by r; the size of the convolution kernel is (m×n); I(x+i, y+j) is the value at position (i,j) of the convolution kernel in the input data matrix; K(i,j) is the value at position (i,j) of the convolution kernel; and b is the bias term.

[0132] The formula for calculating average pooling is as follows:

[0133]

[0134] In the formula, O( x， y) represents the output feature matrix ( x， The value at position y); the size of the pooling window is (m×n); s is the step size, representing the step size of each sliding of the pooling window; Y(x·s+i, y·s+j) represents the elements in the input feature map covered by the pooling window.

[0135] LSTM (Laser-Size-Meter) neural networks are a special type of recurrent neural network (RNN). To address the vanishing or exploding gradient problems of traditional RNNs when processing long-term data series, LSTM introduces two mechanisms: memory cells and gating control. Memory cells are responsible for storing and updating long-term dependencies in the time series. The gating control uses gates to manage the use of state information, comprising three gates: a forget gate, an input gate, and an output gate. These gates collectively determine the retention, updating, and final output of information. The LSTM neural network structure is shown below. Figure 6 As shown, its main calculation formula is as follows:

[0136] f t =σ(W f ·[h t-1 x t ]+b f )

[0137] i t =σ(W i ·[h t-1 x t ]+b i )

[0138]

[0139] o t =σ(W o ·[h t-1 x t ]+b o )

[0140] h t =o t *tanh(C t )

[0141] In the formula above, σ represents the sigmoid activation function; tanh represents the hyperbolic tangent activation function; · represents the dot product; * represents the Hadamard product; t is the time step; x t It is the input data at time t; h t It is the hidden state at time t; Indicates the candidate cell state; C t It is the cell state at time t; ft is the forget gate, which determines the state from cell state C. t What information is discarded? tIt is the input gate, which controls how much information about the candidate state at the current moment needs to be saved into the cell state; t It is an output gate, and the output gate controls the current state C. t How much information is output to the hidden state h? t W f W i W c and W o These are the weights of the corresponding gates and units; b f b i b c and b o These are the bias terms for the corresponding gates and units.

[0142] The formula for calculating the loss function used during neural network training is:

[0143]

[0144] Where N is the number of samples, i.e. the total number of data points; yi is the value of the observation on the i-th sample; Let be the predicted value for the i-th sample.

[0145] Step 1: The input data for the neural network is a 9×9 matrix centered at the data point to be corrected. The input data matrix for the CNN module includes the synthesized velocities of observed ocean current u and v components, tidal current u and v components, wind speed u and v components, and elevation data. Each batch of input data contains matrix data for 10 time steps: t, t+1, ..., t+9. After two convolution and pooling operations, the input matrix data is flattened and input into the LSTM module. The output of the LSTM module is combined with the flattened elevation data, and finally, the corrected u and v components for 10 time steps are output through a fully connected layer (FC). To speed up the process, the data can be normalized.

[0146]

[0147] Among them, X i Represents the normalization result; X ioriginal Represents the raw data in the input field; X max X represents the maximum value of the original data; min This represents the minimum value of the original data. The normalized input training data is randomly divided into two parts, 80% and 20%, which are used as the training set and the test set, respectively, to train the network parameters and test the network performance.

[0148] Step 2: The training cycle of the CNN-LSTM network model is set to 200 rounds. The weight parameters of the network model are updated using the Adam optimization algorithm. The initial learning rate is set to 0.001. In order to prevent the model from overfitting and enhance the generalization ability of the model, L2 regularization is used during the model training process, and the weight decay coefficient is set to 0.00001.

[0149] Step 3: During the training of the model, the number of convolutional layers, the size of the convolutional kernel, and the number of hidden layers in the LSTM module are adjusted. The radar data correction effect of the model under different parameters is compared to minimize the error.

[0150] Step 4: After the above steps, the final network structure is determined. All input training data are used as the training set to train the neural network. Training ends when the target minimum error is reached and remains relatively stable, generating the radar data correction model.

[0151] Step 5: Evaluation metrics: EOF ellipse orientation angle error, average flow orientation angle error, and EOF ellipse major and minor axis errors.

[0152] The calculation method for the EOF ellipse is as follows: First, a data matrix is ​​established, with the first and second columns representing the u and v components, respectively. Then, the velocity matrix is ​​decomposed into a first mode and a second mode through EOF analysis. The major axis of the ellipse represents the first mode, which is the eigenvalue of the maximum standard deviation of the flow velocity, and the minor axis represents the second mode, which is the eigenvalue of the maximum standard deviation of the flow velocity. The direction angle of the ellipse is calculated using the following formula, where v1 and v2 are the eigenvectors of the first and second modes, respectively.

[0153]

[0154] The formulas for calculating the EOF elliptical direction angle error and the average flow direction angle error are as follows:

[0155] Δθ=|(θ1-θ2+180°)mod360°-180°|

[0156] In the above formula, θ1 and θ2 represent the angles of the EOF ellipse direction or the average flow direction of the observed and corrected values, respectively, mod is the modulo operation, and △θ is the error angle.

[0157] The formula for calculating the major and minor axis errors of the EOF ellipse is as follows:

[0158] Among them, Lmaj radar The major axis of the EOF ellipse representing radar data; Lmaj observation The major axis of the EOF ellipse representing the observed data; Lmin radar The minor axis of the ellipse representing the EOF of radar data; Lmin observationThe minor axis of the EOF ellipse representing the observed data. Step 6: Save the model: Save the trained and validated model.

[0159] Step 3: The model trained in Step 2 is used for radar data correction in areas affected by terrain. The input data matrix during correction includes the composite velocity of ocean current u and v components, tidal current velocity u component, tidal current velocity v component, wind speed u component, wind speed v component, and altitude data obtained by radar detection. Finally, the corrected radar data is obtained.

[0160] In summary, the embodiments of the present invention obtain the nonlinear correlation between ocean current field and radar echo, topography and marine environmental information based on deep learning methods, correct radar detection data affected by topography such as coastlines and islands, and improve the accuracy of ocean current direction in radar detection data synthesis.

[0161] The embodiments of the present invention have the following characteristics:

[0162] 1. To improve the inversion calculation method of ocean currents, this technology uses both ground wave radar echo data and actual marine environmental data.

[0163] 2. Unlike traditional methods that use first-order peak information from ground wave radar to invert ocean currents, this technology employs deep learning.

[0164] 3. This technology uses the deep learning method of CNN-LSTM neural network to obtain the optimal correlation coefficient between ocean current direction and radar data and marine environmental data.

[0165] Compared with the prior art, the present invention has the following advantages:

[0166] 1. High Accuracy: The most direct physical processes affecting ocean current fields are ocean dynamic factors, such as tides, wind fields, runoff, and topographic distribution. Traditional ocean current inversion methods are based on indirect calculations using ground-wave radar echo signals. This invention combines radar echo signals and ocean environmental information through a CNN-LSTM neural network, integrating the advantages of CNN in extracting spatial features and LSTM in capturing temporal features, thereby improving the accuracy of radar in detecting ocean current directions.

[0167] 2. Excellent generalization ability: When correcting radar data in areas affected by terrain, it is not necessary to build a separate neural network model for each radar data point; a single CNN-LSTM neural network model can achieve good correction results. Furthermore, as long as the training data contains a sufficiently long time frame, the trained model can be used to correct data from any time period in that area, eliminating the need to retrain the model for each correction of data from different time periods. This significantly reduces time costs and enhances its practical application value.

[0168] Another aspect of this invention provides a ground wave radar data quality control system based on a CNN-LSTM model, comprising:

[0169] The first module is used to determine the range of radar detection data affected by terrain based on the accuracy differences of radar echo data.

[0170] The second module is used to train the CNN-LSTM neural network based on the radar detection data range to obtain the target model;

[0171] The third module is used to correct the radar data of the terrain-affected area according to the target model, so as to obtain the corrected radar data.

[0172] It is understood that the content of the above method embodiments is applicable to this system embodiment. The specific functions implemented in this system embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments.

[0173] This invention also provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the aforementioned ground wave radar data quality control method based on the CNN-LSTM model. This electronic device can be any smart terminal, including tablet computers, in-vehicle computers, etc.

[0174] It is understood that the content of the above method embodiments is applicable to this device embodiment. The specific functions implemented by this device embodiment are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0175] Please see Figure 7 , Figure 7 The hardware structure of an electronic device according to another embodiment is illustrated. The electronic device includes:

[0176] The processor 701 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of the present invention.

[0177] The memory 702 can be implemented as a read-only memory (ROM), static storage device, dynamic storage device, or random access memory (RAM). The memory 702 can store the operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 702 and is called and executed by the processor 701 to execute the ground wave radar data quality control method based on the CNN-LSTM model of the embodiments of the present invention.

[0178] The input / output interface 703 is used to implement information input and output;

[0179] The communication interface 704 is used to enable communication and interaction between this device and other devices. Communication can be achieved through wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.).

[0180] Bus 705 transmits information between various components of the device (e.g., processor 701, memory 702, input / output interface 703, and communication interface 704);

[0181] The processor 701, memory 702, input / output interface 703, and communication interface 704 are connected to each other within the device via bus 705.

[0182] This invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described ground wave radar data quality control method based on a CNN-LSTM model.

[0183] It is understood that the content of the above method embodiments is applicable to this storage medium embodiment. The specific functions implemented in this storage medium embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments.

[0184] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0185] It should be noted that in various specific embodiments of the present invention, when processing data related to user identity or characteristics, such as user information, user behavior data, user historical data, and user location information, user permission or consent is obtained first. Furthermore, the collection, use, and processing of this data comply with relevant laws, regulations, and standards. In addition, when embodiments of the present invention require access to sensitive personal information of users, separate permission or consent from the user is obtained through pop-ups or redirection to a confirmation page. Only after obtaining the user's separate permission or consent is the necessary user-related data for the normal operation of the embodiments of the present invention acquired.

[0186] The embodiments described in this invention are for the purpose of more clearly illustrating the technical solutions of the embodiments of this invention, and do not constitute a limitation on the technical solutions provided by the embodiments of this invention. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of this invention are also applicable to similar technical problems.

[0187] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of the present invention, and may include more or fewer steps than shown, or combine certain steps, or different steps.

[0188] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0189] Those skilled in the art will understand that all or some of the steps in the methods disclosed above, as well as the functional modules / units in the systems and devices, can be implemented as software, firmware, hardware, or suitable combinations thereof.

[0190] The terms "first," "second," "third," "fourth," etc. (if present) in the specification and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0191] It should be understood that in this invention, "at least one (item)" refers to one or more, and "more than one" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: 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 single or multiple.

[0192] In the several embodiments provided by this invention, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of the units described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0193] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0194] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0195] If the integrated unit is implemented as 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 solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes multiple instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing programs, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0196] The preferred embodiments of the present invention have been described above with reference to the accompanying drawings, but this does not limit the scope of the claims of the present invention. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and spirit of the present invention should be within the scope of the claims of the present invention.

Claims

1. A method for quality control of ground wave radar data based on a CNN-LSTM model, characterized in that, Includes the following steps: Based on the differences in the accuracy of radar echo data, the range of radar detection data affected by terrain is determined; Based on the radar detection data range, a CNN-LSTM neural network is trained to obtain the target model; The radar data of the terrain-affected area is corrected according to the target model to obtain the corrected radar data. The step of training a CNN-LSTM neural network to obtain a target model based on the radar detection data range includes the following steps: An initial model is built based on the CNN-LSTM neural network; A 9×9 matrix centered at the data points to be corrected is used as the input data for the CNN-LSTM neural network; The training cycle of the CNN-LSTM neural network was set to 200 rounds, and the weight parameters of the network model were updated using the Adam optimization algorithm. The initial learning rate was set to 0.

001. During the training of the network model, the number of convolutional layers, the size of the convolutional kernel, and the number of hidden layers in the LSTM module are adjusted. Determine the final network structure, use all input training data as the training set, train the neural network, and end the training when the target minimum error is reached and remains relatively stable, generating a radar data correction model as the target model. The process of using a 9×9 matrix centered at the data points to be corrected as input data for the CNN-LSTM neural network includes the following steps: Based on the observed ocean current u and v component composite velocities, tidal current u component, tidal current v component, wind speed u component, wind speed v component, and altitude data, an input data matrix for the CNN module is constructed. Each batch of input data contains matrix data for 10 time steps: t, t+1, ..., t+9. After two convolution and pooling operations, the input matrix data is flattened and input into the LSTM module. The output of the LSTM module is combined with the flattened altitude data, and finally, the corrected u and v components for 10 time steps are output through a fully connected layer (FC). The input data is normalized, and the expression for the normalization process is: in, Represents the normalization result; Represents the raw data in the input field; Represents the maximum value of the original data; This represents the minimum value of the original data; The normalized input training data is randomly divided into two parts, 80% and 20%, which are used as the training set and the test set, respectively, to train network parameters and test network performance.

2. The method for controlling the quality of ground wave radar data based on a CNN-LSTM model according to claim 1, characterized in that, Determining the radar detection data range affected by terrain based on the accuracy differences of radar echo data includes the following steps: Based on the differences in the accuracy of radar echo data, data is detected in radar detection areas near coastlines and islands; Regions where the elliptic direction angle error of the empirical orthogonal function of the synthetic ocean current is greater than 40° are identified as radar detection data ranges affected by terrain.

3. The method for controlling the quality of ground wave radar data based on a CNN-LSTM model according to claim 1, characterized in that, The training process for a CNN-LSTM neural network also includes the following steps: Based on the EOF ellipse orientation angle error, the average flow orientation angle error, and the EOF ellipse major and minor axis errors, evaluation indicators are determined, and the target model is validated; this step specifically involves: A data matrix is ​​constructed, with the first and second columns representing the u and v components, respectively. Then, the velocity matrix is ​​decomposed into a first mode and a second mode through EOF analysis. The major axis of the ellipse represents the first mode, which is the eigenvalue of the maximum standard deviation of the velocity, and the minor axis represents the second mode, which is the eigenvalue of the maximum standard deviation of the velocity. The direction angle of an ellipse is calculated using the formula. Calculated, where and These are the feature vectors of the first mode and the second mode, respectively: The formulas for calculating the EOF elliptical direction angle error and the average flow direction angle error are as follows: ,in, and The angles representing the EOF ellipse direction or mean flow direction for the observed and corrected values, respectively. For modulo operation, This is the error angle; EOF Ellipse Major and Minor Axis Errors The calculation formula is: ,in, The major axis of the EOF ellipse representing radar data; The major axis of the EOF ellipse representing the observed data; The minor axis of the EOF ellipse representing radar data; The minor axis of the EOF ellipse representing the observed data.

4. The method for controlling the quality of ground wave radar data based on a CNN-LSTM model according to claim 1, characterized in that, The initial model established based on the CNN-LSTM neural network includes the following steps: Features are extracted from Convolutional Neural Networks (CNNs) through convolution operations. The convolution operation extracts local features step-by-step by sliding a convolution kernel across the input data. The formula for feature extraction is as follows: ,in, Indicating the feature matrix extracted by the multiplicative neural network The value of the position; The ReLU activation function is used; the kernel size is... ; The corresponding convolution kernel position in the input data matrix The value; In the convolution kernel The value of the position; For bias terms; The formula for calculating average pooling is as follows: in, Indicating the output feature matrix The value of the position; the size of the pooling window is ; It is the step size, representing the step size of each slide of the pooling window; This represents the elements in the input feature map covered by the pooling window; The formula for calculating the network parameters of an LSTM neural network is as follows: in, This represents the sigmoid activation function; Represents the hyperbolic tangent activation function; Represents the dot product; Represents the Hadama product; It is a time step; It is the first Input data at any given time; It is the first The hidden state at any given moment; Indicates the candidate cell status; It is the cell state at time 𝑡; It's the forget gate, which determines the state from the unit. What information is discarded? It is the input gate, which controls how much information of the candidate state at the current moment needs to be saved into the cell state; It is an output gate; the output gate controls the current state. How much information is output to the hidden state? ; , , and These are the weights of the corresponding gates and units; , , and These are the bias terms corresponding to gates and units; The formula for calculating the loss function used during neural network training is: in, The number of samples; For the observed value at the th The values ​​on each sample; For the predicted value in the th The values ​​on each sample.

5. The method for controlling the quality of ground wave radar data based on a CNN-LSTM model according to claim 1, characterized in that, The step of correcting the radar data of the terrain-affected area based on the target model to obtain the corrected radar data is as follows: The radar data of the terrain-affected area is corrected according to the target model. The input data matrix during correction includes the composite velocity of ocean current u and v components, tidal current velocity u component, tidal current velocity v component, wind speed u component, wind speed v component, and altitude data obtained by radar detection. Finally, the corrected radar data is obtained.

6. A system for implementing the ground wave radar data quality control method based on the CNN-LSTM model as described in any one of claims 1-5, characterized in that, include: The first module is used to determine the range of radar detection data affected by terrain based on the accuracy differences of radar echo data; The second module is used to train the CNN-LSTM neural network based on the radar detection data range to obtain the target model; The third module is used to correct the radar data of the terrain-affected area according to the target model, so as to obtain the corrected radar data.

7. An electronic device, characterized in that, Including the processor and memory; The memory is used to store programs; The processor executes the program to implement the method as described in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, The storage medium stores a program that is executed by a processor to implement the method as described in any one of claims 1 to 5.

Images