Lightning electromagnetic pulse target detection method and system and electronic equipment

Abstract

The invention belongs to the field of lightning electromagnetic pulse signal propagation and detection, artificial intelligence and deep learning, particularly relates to a lightning electromagnetic pulse target detection method and system and electronic equipment, and aims to solve the problem that an existing lightning electromagnetic pulse target detection method cannot accurately recognize and detect lightning electromagnetic pulses from collected electromagnetic pulse signals, and the detection accuracy is high. And electromagnetic noise interference is easily caused. The method comprises the following steps: collecting lightning electromagnetic pulse data, and inputting a pre-trained electromagnetic pulse and lightning type mapping model to obtain a feature layer matrix and an output layer matrix; electromagnetic pulse data attention data are obtained through a lightning electromagnetic pulse target attention analysis method; performing peak indexing on the electromagnetic pulse target attention data to obtain a lightning electromagnetic pulse target detection result; and outputting a lightning electromagnetic pulse target detection result. According to the invention, the lightning electromagnetic pulse can be accurately identified and detected in a strong interference environment.

Description

Technical Field

[0001] The present invention belongs to the fields of lightning electromagnetic pulse signal propagation and detection, artificial intelligence and deep learning, and specifically relates to a lightning electromagnetic pulse target detection method, system and electronic equipment. Background Art

[0002] Natural lightning phenomena are accompanied by a large number of electromagnetic radiation signals, which propagate over long distances in the ground-ionosphere waveguide. Professional lightning electromagnetic pulse detection equipment can record lightning electromagnetic pulse signals in real time. At present, the mainstream lightning location system uses lightning electromagnetic pulse signals for positioning. The lightning location system consists of at least four lightning electromagnetic pulse detection devices and a data processing center. The positioning algorithm calculates the time difference between the lightning electromagnetic pulse signal reaching each detection device and combines the time difference variance to solve the time and location of the lightning.

[0003] The lightning electromagnetic pulse signal received by the detection equipment is usually affected by the interference signal, resulting in signal distortion, which makes the signal unable to be effectively detected. At the same time, due to the difference in lightning discharge types (such as cloud-to-ground lightning, cloud-to-cloud lightning) or the influence of the signal propagation path, the accuracy of the calculation of the arrival time of the lightning electromagnetic pulse signal will be affected. Therefore, a method is needed to efficiently detect the lightning electromagnetic pulse signal from the signal collected by the detection equipment to improve the accuracy of the lightning detection results. Summary of the invention

[0004] In order to solve the above problems in the prior art, that is, to solve the problem that the existing lightning electromagnetic pulse target detection method cannot accurately identify and detect lightning electromagnetic pulses from the collected electromagnetic pulse signals and is easily interfered by electromagnetic noise, the first aspect of the present invention proposes a lightning electromagnetic pulse target detection method, which includes:

[0005] S10, collecting lightning electromagnetic pulse data as input data;

[0006] S20, inputting the input data into a pre-trained electromagnetic pulse and lightning type mapping model to obtain a feature layer matrix and an output layer matrix;

[0007] The electromagnetic pulse and lightning type mapping model is constructed based on a deep learning model;

[0008] S30, obtaining electromagnetic pulse data attention data through a lightning electromagnetic pulse target attention analysis method based on the feature layer matrix and the output layer matrix;

[0009] S40, performing peak indexing on the electromagnetic pulse target attention data to obtain a lightning electromagnetic pulse target detection result;

[0010] S50, outputting the lightning electromagnetic pulse target detection result.

[0011] In some preferred embodiments, the input of the electromagnetic pulse and lightning type mapping model is lightning electromagnetic pulse data, and the output is lightning type; the lightning types include cloud-to-ground lightning, cloud lightning, and pocket cloud lightning.

[0012] In some preferred embodiments, the electromagnetic pulse and lightning type mapping model includes an input layer, a hidden layer, and an output layer;

[0013] The hidden layer is constructed based on M1 convolution modules, a global mean layer, M2 fully connected layers, and a sigmoid activation layer connected in sequence;

[0014] Each convolution block is constructed based on two one-dimensional convolutional layers and a maximum pooling layer connected in sequence; each one-dimensional convolutional layer is connected to a ReLU activation function layer;

[0015] Except for the last one of the M2 fully connected layers, all other fully connected layers are connected to a ReLU activation function layer.

[0016] In some preferred embodiments, based on the feature layer matrix and the output layer matrix, electromagnetic pulse data attention data is obtained by a lightning electromagnetic pulse target attention analysis method, and the method is:

[0017] Taking the maximum value of the output layer matrix as the starting point of reverse gradient calculation, performing back propagation, and obtaining the gradient matrix of the feature layer matrix;

[0018] Calculate the weight of each feature channel of the gradient matrix;

[0019] The feature layer matrix is ​​multiplied and accumulated with the weights corresponding to each feature channel, and processed through an activation function to obtain electromagnetic pulse data attention data.

[0020] In some preferred embodiments, the feature layer matrix is ​​the feature matrix output by the first one-dimensional convolution layer in the first convolution block or the feature matrix output by the one-dimensional convolution layer with a dimension of 1×N×k, where N is the number of columns of the dimension of the input layer input of the electromagnetic pulse and lightning type mapping model, and k represents the number of feature channels.

[0021] In some preferred embodiments, the feature layer matrix is ​​multiplied and accumulated with the weights corresponding to each feature channel, and processed by an activation function to obtain electromagnetic pulse data attention data, and the method is as follows:

[0022] L = ReLU (∑ k w k A k )

[0023]

[0024] Among them, L represents the electromagnetic pulse data attention data, A k represents the feature layer matrix, w n represents the weight, and x represents the input of the ReLU activation function.

[0025] In some preferred embodiments, peak indexing is performed on the electromagnetic pulse target attention data to obtain lightning electromagnetic pulse target detection results, and the method is as follows:

[0026] Normalizing the electromagnetic pulse target attention data to obtain normalized data;

[0027] The index value corresponding to the peak in the normalized data is obtained as the lightning electromagnetic pulse target detection result.

[0028] A second aspect of the present invention provides a lightning electromagnetic pulse target detection system, the system comprising:

[0029] A signal acquisition module is configured to acquire lightning electromagnetic pulse data as input data;

[0030] A signal processing module is configured to input the input data into a pre-trained electromagnetic pulse and lightning type mapping model to obtain a feature layer matrix and an output layer matrix;

[0031] The electromagnetic pulse and lightning type mapping model is constructed based on a deep learning model;

[0032] Based on the feature layer matrix and the output layer matrix, obtaining electromagnetic pulse data attention data through a lightning electromagnetic pulse target attention analysis method;

[0033] Performing peak indexing on the electromagnetic pulse target attention data to obtain a lightning electromagnetic pulse target detection result;

[0034] Output the lightning electromagnetic pulse target detection result.

[0035] In some preferred embodiments, the signal acquisition module includes an antenna, a signal filtering and amplifying circuit, and an analog / digital conversion circuit;

[0036] The antenna includes a magnetic loop antenna, a flat electric field antenna or a whip antenna; the antenna is used to sense electromagnetic pulse signals;

[0037] The signal filtering and amplifying circuit is used to filter and amplify the electromagnetic pulse signal;

[0038] The analog / digital conversion circuit is used to perform analog-to-digital conversion on the amplified electromagnetic pulse signal to obtain a digital signal, namely, electromagnetic pulse data.

[0039] According to a third aspect of the present invention, an electronic device is provided, comprising:

[0040] At least one processor; and a memory communicatively connected to at least one of the processors; wherein the memory stores instructions executable by the processor, and the instructions are used to be executed by the processor to implement the above-mentioned lightning electromagnetic pulse target detection method.

[0041] Beneficial effects of the present invention:

[0042] The present invention realizes the accurate identification and detection of lightning electromagnetic pulses in a strong interference environment.

[0043] 1) The present invention combines the lightning electromagnetic pulse type mapping model with the target attention analysis method, which can realize the accurate detection of lightning electromagnetic pulse types and any multiple targets, and enhance the anti-interference ability of the method, and can adapt to various strong electromagnetic interference environments;

[0044] 2) The computational complexity of the model of the present invention is low, and a single calculation can simultaneously complete signal type and target detection, which can meet the requirements of real-time processing of electromagnetic pulse signals. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Other features, objects and advantages of the present application will become more apparent by reading the detailed description of non-limiting embodiments made with reference to the following drawings.

[0046] Figure 1 It is a flow chart of a lightning electromagnetic pulse target detection method according to an embodiment of the present invention;

[0047] Figure 2 It is a detailed flow chart of a lightning electromagnetic pulse target detection method according to an embodiment of the present invention;

[0048] Figure 3 It is a schematic diagram of the framework of an electromagnetic pulse detection device according to an embodiment of the present invention;

[0049] Figure 4 It is a structural schematic diagram of an electromagnetic pulse and lightning type mapping model according to an embodiment of the present invention;

[0050] Figure 5 It is a schematic diagram of the analysis process of a lightning electromagnetic pulse target attention analysis method according to an embodiment of the present invention;

[0051] Figure 6 It is a schematic diagram of a lightning electromagnetic pulse target detection result according to an embodiment of the present invention. DETAILED DESCRIPTION

[0052] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution in the embodiment of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiment is a part of the embodiment of the present invention, not all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0053] The present application will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the relevant invention, rather than to limit the invention. It is also necessary to explain that, for ease of description, only the parts related to the relevant invention are shown in the accompanying drawings.

[0054] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application may be combined with each other.

[0055] A lightning electromagnetic pulse target detection method according to a first embodiment of the present invention is as follows: Figure 1 As shown, the following steps are included:

[0056] S10, collecting lightning electromagnetic pulse data as input data;

[0057] S20, inputting the input data into a pre-trained electromagnetic pulse and lightning type mapping model to obtain a feature layer matrix and an output layer matrix;

[0058] The electromagnetic pulse and lightning type mapping model is constructed based on a deep learning model;

[0059] S30, obtaining electromagnetic pulse data attention data through a lightning electromagnetic pulse target attention analysis method based on the feature layer matrix and the output layer matrix;

[0060] S40, performing peak indexing on the electromagnetic pulse target attention data to obtain a lightning electromagnetic pulse target detection result;

[0061] S50, outputting the lightning electromagnetic pulse target detection result.

[0062] In order to more clearly illustrate the lightning electromagnetic pulse target detection method of the present invention, the following is combined with the attached Figure 2 Each step in an embodiment of the method of the present invention is described in detail.

[0063] The present invention provides a lightning electromagnetic pulse target detection method, which uses a deep learning model and model reverse gradient calculation to achieve real-time target detection of lightning electromagnetic pulse signals. The specific process is as follows:

[0064] S10, collecting lightning electromagnetic pulse data as input data;

[0065] In this embodiment, the electromagnetic pulse detection device is used to convert the analog electromagnetic pulse signal collected by the front-end circuit into a digital signal, namely, lightning electromagnetic pulse data, in real time, and store it in the device memory. The specific collection process is detailed in the second embodiment and will not be repeated here.

[0066] S20, inputting the input data into a pre-trained electromagnetic pulse and lightning type mapping model to obtain a feature layer matrix and an output layer matrix; the electromagnetic pulse and lightning type mapping model is constructed based on a deep learning model;

[0067] In this embodiment, the electromagnetic pulse and lightning type mapping model is constructed based on a deep learning model; the electromagnetic pulse and lightning type mapping model includes an input layer, a hidden layer, and an output layer, such as Figure 4 As shown; the input of the electromagnetic pulse and lightning type mapping model is electromagnetic pulse data, and the input dimension is 1×N, where N is the number of columns of the dimension of the input layer input of the electromagnetic pulse and lightning type mapping model, and its size depends on the sampling rate of the electromagnetic pulse detection device and the time length of a single sampling (assuming that the input layer data dimension is 1×1000, the sampling rate of the electromagnetic pulse detection device is 1MSPS, and the time length of a single sampling is 1 millisecond), and the output is the lightning type; lightning types include cloud-to-ground lightning, cloud lightning, and pocket cloud lightning; that is, the input layer passes the lightning electromagnetic pulse data to be mapped into the model, the hidden layer extracts and fuses the input data, and the output layer maps the features calculated by the hidden layer to the lightning type;

[0068] The hidden layer is constructed based on M1 convolution modules, a global mean layer, M2 fully connected layers, and a sigmoid activation layer connected in sequence; each convolution block is constructed based on two one-dimensional convolution layers and a maximum pooling layer connected in sequence; each one-dimensional convolution layer is connected to a ReLU activation function layer; except for the last one of the M2 fully connected layers, all other fully connected layers are connected to a ReLU activation function layer.

[0069] The training process of the electromagnetic pulse and lightning type mapping model is as follows:

[0070] A data set is established, and the collected electromagnetic pulse data is labeled using a manual labeling method; the labeling includes information such as the type of lightning electromagnetic pulse data, arrival time, and location of the collection device; the established data set is then used to train and evaluate the model parameters, and the model structure is adjusted and trained based on the evaluation results, and ultimately a better model is obtained (i.e., the collected electromagnetic pulse data is input into the pre-built electromagnetic pulse and lightning type mapping model to obtain the lightning type, and combined with the label, the loss is calculated, the model parameters are adjusted based on the loss, and then the model is trained in a loop until a trained model is obtained).

[0071] Finally, based on the trained electromagnetic pulse and lightning type mapping model, the feature layer matrix and output layer matrix are obtained; among which, the feature layer matrix is ​​the feature matrix output by the first one-dimensional convolutional layer in the first convolutional block or the feature matrix output by the one-dimensional convolutional layer with a dimension of 1×N×k, where k represents the number of feature channels.

[0072] S30, obtaining electromagnetic pulse data attention data through a lightning electromagnetic pulse target attention analysis method based on the feature layer matrix and the output layer matrix;

[0073] In this embodiment, the electromagnetic pulse data attention data is obtained by the lightning electromagnetic pulse target attention analysis method, and the specific process is as follows:

[0074] The model is forward inferred to obtain the feature layer matrix A k And the model output layer matrix y, where k represents different feature channels; Figure 5 As shown, assuming the electromagnetic pulse and lightning type mapping model in this embodiment, the output dimension of the first convolutional layer is a feature layer of 1×1000×16, the number of feature layer channels is 16, and the output layer matrix y={y1, y2, y3}, wherein y1, y2, y3 correspond to the confidences of three types of cloud-to-ground flash, cloud flash, and pocket cloud flash, respectively, and the confidence value interval is [0, 1]. The closer the value is to 1, the higher the confidence of the result. Preferably, if the maximum value of y detected is less than 0.5, it means that noise is detected and there is no need to perform S40-S50 calculations.

[0075] The model reverse gradient calculation selects the maximum value of matrix y as the starting point for reverse gradient calculation, and back propagates to obtain the matrix A for the feature layer k The gradient matrix A k ′;

[0076] Calculate the gradient matrix A k ′The weight of each channel w k , which is calculated as:

[0077] w n =Average(A n ′), n∈k

[0078] The class activation thermal matrix is ​​calculated to obtain the electromagnetic pulse target attention data L, which is calculated as follows:

[0079] L = ReLU (∑ k w k A k )

[0080] Among them, ReLU(·) is a linear activation function, and its output for input x is:

[0081]

[0082] S40, performing peak indexing on the electromagnetic pulse target attention data to obtain a lightning electromagnetic pulse target detection result;

[0083] In this embodiment, the peak value index is performed on the electromagnetic pulse target attention data, specifically:

[0084] The electromagnetic pulse target attention data is normalized to obtain the normalized data L′:

[0085] L′=L / max(L)

[0086] Get the index value corresponding to the peak in the normalized data as the lightning electromagnetic pulse target detection result. Figure 6 As shown, the target position of the lightning electromagnetic pulse is represented by a five-pointed star. It can be clearly seen from the result diagram that, for a signal containing strong interfering background electromagnetic noise, a lightning electromagnetic pulse target detection method of the present invention can accurately identify any multiple lightning electromagnetic pulse targets contained in the signal and output the target position information.

[0087] S50, outputting the lightning electromagnetic pulse target detection result.

[0088] In this embodiment, the lightning target detection result is output through the network port or the serial port.

[0089] A lightning electromagnetic pulse target detection system according to a second embodiment of the present invention, namely an electromagnetic pulse detection device, comprises:

[0090] A signal acquisition module configured to acquire electromagnetic pulse data as input data;

[0091] In this embodiment, the signal acquisition module includes an antenna, a signal filtering and amplifying circuit, and an analog / digital conversion circuit. Figure 3 As shown;

[0092] The antenna includes a magnetic loop antenna, a flat electric field antenna or a whip antenna; the antenna is used to sense electromagnetic pulse signals; in the present invention, it is preferred to receive electromagnetic field signals in the frequency band of 3kHz to 400kHz;

[0093] A signal filtering and amplifying circuit, used for filtering and amplifying the electromagnetic pulse signal;

[0094] The analog / digital conversion circuit is used to perform analog-to-digital conversion on the amplified electromagnetic pulse signal to obtain a digital signal, namely, electromagnetic pulse data;

[0095] The signal processing module includes an electromagnetic pulse signal type mapping module and a lightning electromagnetic pulse signal real-time detection module;

[0096] an electromagnetic pulse signal type mapping module, configured to input the input data into a pre-trained electromagnetic pulse and lightning type mapping model to obtain a feature layer matrix and an output layer matrix;

[0097] The electromagnetic pulse and lightning type mapping model is constructed based on a deep learning model;

[0098] The lightning electromagnetic pulse signal real-time detection module is configured to obtain electromagnetic pulse data attention data based on the feature layer matrix and the output layer matrix through a lightning electromagnetic pulse target attention analysis method; perform peak indexing on the electromagnetic pulse target attention data to obtain a lightning electromagnetic pulse target detection result; and test the lightning electromagnetic pulse target detection result to output a lightning target detection result.

[0099] A lightning electromagnetic pulse target detection system also includes a power supply module and a data transmission module; the data transmission module is configured to transmit the real-time target detection result of the lightning electromagnetic pulse signal to the outside through a network port or a serial port. The power supply module is used to power the electromagnetic pulse detection equipment, such as the above-mentioned antenna.

[0100] It should be noted that the lightning electromagnetic pulse target detection system provided in the above embodiment is only illustrated by the division of the above functional modules. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the modules or steps in the embodiments of the present invention can be decomposed or combined. For example, the modules in the above embodiments can be combined into one module, or further divided into multiple sub-modules to complete all or part of the functions described above. The names of the modules and steps involved in the embodiments of the present invention are only for distinguishing the modules or steps, and are not regarded as improper limitations of the present invention.

[0101] An electronic device according to the third embodiment of the present invention comprises: at least one processor; and a memory communicatively connected to at least one of the processors; wherein the memory stores instructions executable by the processor, and the instructions are used to be executed by the processor to implement the above-mentioned lightning electromagnetic pulse target detection method.

[0102] A computer-readable storage medium according to a fourth embodiment of the present invention stores computer instructions, and the computer instructions are used to be executed by the computer to implement the above-mentioned lightning electromagnetic pulse target detection method.

[0103] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process and related instructions of the electronic device and computer-readable storage medium described above can refer to the corresponding process in the aforementioned method example and will not be repeated here.

[0104] Those skilled in the art should be able to appreciate that the modules and method steps of each example described in conjunction with the embodiments disclosed herein can be implemented with electronic hardware, computer software or a combination of the two, and the programs corresponding to the software modules and method steps can be placed in random access memory (RAM), internal memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disks, removable disks, CD-ROMs, or any other form of storage medium known in the technical field. In order to clearly illustrate the interchangeability of electronic hardware and software, the composition and steps of each example have been generally described in the above description according to the function. Whether these functions are performed in electronic hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present invention.

[0105] The terms "first", "second", etc. are used to distinguish similar objects rather than to describe or indicate a particular order or sequence.

[0106] The term "comprise" or any other similar term is intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus / device that includes a list of elements includes not only those elements but also other elements not expressly listed, or also includes elements inherent to such process, method, article, or apparatus / device.

[0107] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

Claims

1. A lightning electromagnetic pulse target detection method, characterized in that: The method comprises the following steps: S10, collecting lightning electromagnetic pulse data as input data; S20, inputting the input data into a pre-trained electromagnetic pulse and lightning type mapping model to obtain a feature layer matrix and an output layer matrix; The electromagnetic pulse and lightning type mapping model is constructed based on a deep learning model; S30, obtaining electromagnetic pulse data attention data through a lightning electromagnetic pulse target attention analysis method based on the feature layer matrix and the output layer matrix; S40, performing peak indexing on the electromagnetic pulse target attention data to obtain a lightning electromagnetic pulse target detection result; S50, outputting the lightning electromagnetic pulse target detection result.

2. A lightning electromagnetic pulse target detection method according to claim 1, characterized in that: The input of the electromagnetic pulse and lightning type mapping model is lightning electromagnetic pulse data, and the output is lightning type; the lightning types include cloud-to-ground lightning, cloud lightning, and pocket cloud lightning.

3. A lightning electromagnetic pulse target detection method according to claim 2, characterized in that: The electromagnetic pulse and lightning type mapping model includes an input layer, a hidden layer, and an output layer; The hidden layer is constructed based on M1 convolution modules, a global mean layer, M2 fully connected layers, and a sigmoid activation layer connected in sequence; Each convolution block is constructed based on two one-dimensional convolutional layers and a maximum pooling layer connected in sequence; each one-dimensional convolutional layer is connected to a ReLU activation function layer; Except for the last one of the M2 fully connected layers, all other fully connected layers are connected to a ReLU activation function layer.

4. A lightning electromagnetic pulse target detection method according to claim 1, characterized in that: Based on the feature layer matrix and the output layer matrix, the electromagnetic pulse data attention data is obtained by a lightning electromagnetic pulse target attention analysis method, and the method is as follows: Taking the maximum value of the output layer matrix as the starting point of reverse gradient calculation, performing back propagation, and obtaining the gradient matrix of the feature layer matrix; Calculate the weight of each feature channel of the gradient matrix; The feature layer matrix is ​​multiplied and accumulated with the weights corresponding to each feature channel, and processed through an activation function to obtain electromagnetic pulse data attention data.

5. A lightning electromagnetic pulse target detection method according to claim 4, characterized in that: The feature layer matrix is ​​the feature matrix output by the first one-dimensional convolution layer in the first convolution block or the feature matrix output by the one-dimensional convolution layer with a dimension of 1×N×k, where N is the number of columns of the dimension of the input layer input of the electromagnetic pulse and lightning type mapping model, and k represents the number of feature channels.

6. A lightning electromagnetic pulse target detection method according to claim 5, characterized in that: The characteristic layer matrix is ​​multiplied and accumulated with the weights corresponding to each characteristic channel, and processed by an activation function to obtain electromagnetic pulse data attention data, and the method is as follows: Among them, L represents the electromagnetic pulse data attention data, A k represents the feature layer matrix, w n represents the weight, and x represents the input of the ReLU activation function.

7. A lightning electromagnetic pulse target detection method according to claim 6, characterized in that: The peak value index is performed on the electromagnetic pulse target attention data to obtain the lightning electromagnetic pulse target detection result, and the method is as follows: Normalizing the electromagnetic pulse target attention data to obtain normalized data; The index value corresponding to the peak in the normalized data is obtained as the lightning electromagnetic pulse target detection result.

8. A lightning electromagnetic pulse target detection system, characterized in that: The system includes: A signal acquisition module is configured to acquire lightning electromagnetic pulse data as input data; A signal processing module is configured to input the input data into a pre-trained electromagnetic pulse and lightning type mapping model to obtain a feature layer matrix and an output layer matrix; The electromagnetic pulse and lightning type mapping model is constructed based on a deep learning model; Based on the feature layer matrix and the output layer matrix, obtaining electromagnetic pulse data attention data through a lightning electromagnetic pulse target attention analysis method; Performing peak indexing on the electromagnetic pulse target attention data to obtain a lightning electromagnetic pulse target detection result; Output the lightning electromagnetic pulse target detection result.

9. A lightning electromagnetic pulse target detection system according to claim 8, characterized in that: The signal acquisition module includes an antenna, a signal filtering and amplifying circuit, and an analog / digital conversion circuit; The antenna includes a magnetic loop antenna, a flat electric field antenna or a whip antenna; the antenna is used to sense electromagnetic pulse signals; The signal filtering and amplifying circuit is used to filter and amplify the electromagnetic pulse signal; The analog / digital conversion circuit is used to perform analog-to-digital conversion on the amplified electromagnetic pulse signal to obtain a digital signal, namely, electromagnetic pulse data.

10. An electronic device, characterized in that: include: at least one processor; and a memory communicatively coupled to at least one of the processors; The memory stores instructions executable by the processor, and the instructions are used to be executed by the processor to implement a lightning electromagnetic pulse target detection method as described in any one of claims 1-7.

Images