An electromagnetic interference identification method and system based on an LSTM neural network

Abstract

The application discloses an electromagnetic interference identification method and system based on an LSTM neural network, and builds a test evaluation system, including a radar seeker model of a certain project, a target signal simulation system, an electronic interference simulation system and an antenna horn. By outputting different target simulation signals, distance drag type false target interference signals, speed drag type false target interference signals, suppression type noise interference signals and various interference signals, the combat scene that the radar seeker intercepts and tracks the simulation target signal under complex electromagnetic conditions is simulated. The application is simple, can be quickly transplanted to the anti-interference model of the radar seeker, improves the accuracy of the identification of different types of electromagnetic interference, and solves the problem that the existing seeker anti-interference model does not have universality and cannot realize the mutual use of the electromagnetic interference identification models of different types of seekers.

Description

Technical Field

[0001] This invention relates to the field of electromagnetic interference identification models, and in particular to an electromagnetic interference identification method and system based on LSTM neural networks. Background Technology

[0002] Anti-jamming models for radar seekers are a significant challenge in radar seeker anti-jamming technology, and electromagnetic interference (EMI) identification models are a crucial component of these models. Currently, domestic EMI identification models lack universality, requiring specialized designs for specific radar seeker models. This necessitates substantial investment of manpower and resources in repetitive work, resulting in poor versatility and hindering interoperability between different radar seeker models. Summary of the Invention

[0003] The purpose of this invention is to provide an electromagnetic interference identification model based on LSTM neural network to solve the problem that existing electromagnetic interference identification models cannot be universally applied across different models.

[0004] The technical solution of this invention is:

[0005] In a first aspect, this invention discloses an electromagnetic interference identification method based on an LSTM neural network, the specific steps of which are as follows:

[0006] The first step is to construct an indoor field test system for radar seeker anti-jamming.

[0007] The radar seeker anti-jamming indoor test system includes: air-to-air radar seeker, antenna horn, target signal simulation system, electronic jamming simulation system, data bus host computer, and electromagnetic interference identification model based on LSTM neural network;

[0008] The air-to-air radar seeker is connected to the target signal simulation system, the air-to-air radar seeker is connected to the data bus host computer, the antenna horn is connected to the target signal simulation system, the antenna horn is connected to the electronic interference simulation system, the target signal simulation system is connected to the electronic interference simulation system, the target signal simulation system is connected to the data bus host computer, the electronic interference simulation system is connected to the data bus host computer, and the data bus host computer is connected to the electromagnetic interference identification model based on LSTM neural network.

[0009] The second step is to obtain the test data from the host computer on the data bus.

[0010] After the radar seeker anti-jamming indoor test system was set up, the target signal simulation system parameters were set and the simulated target signal was output. The air-to-air radar seeker was controlled by the host computer via the data bus to intercept and stably track the simulated target signal. Then, the electronic jamming simulation system parameters were set and the electronic jamming signal was output to simulate the battlefield electromagnetic environment under complex electromagnetic conditions. The host computer on the data bus recorded the missile-to-target distance information output by the air-to-air radar seeker in real time. The target signal simulation system recorded the information of the simulated target signal output, and the electronic jamming simulation system recorded the information of the output.

[0011] The third step involves test data processing and statistics.

[0012] During the test, the host computer on the data bus records combinations of different simulated target signals and different electronic jamming signals. The electromagnetic information received by the antenna output by the air-to-air radar seeker, the target electromagnetic information uploaded by the target signal simulation system, and the electronic jamming data information uploaded by the electronic jamming simulation system in the same test are packaged into a set of data in the dataset. After preprocessing the dataset, the preprocessed dataset is divided into a training dataset and a detection dataset.

[0013] Step 4: Training and Detection of the Electromagnetic Interference Recognition Model using an LSTM Neural Network

[0014] The training dataset obtained in step 3 is used as input data to train the LSTM neural network electromagnetic interference recognition model, and the detection dataset is used as detection data to input the LSTM neural network electromagnetic interference recognition model for detection. The detection results are obtained, and the parameters of the LSTM neural network electromagnetic interference recognition model are tuned and optimized based on the detection results. The tuned and optimized LSTM neural network electromagnetic interference recognition model is then used as a new model to repeat the above training and detection process until the detection results meet the accuracy requirements.

[0015] This completes the work on the electromagnetic interference identification method based on LSTM neural network.

[0016] In one specific embodiment, the air-to-air radar seeker functions at least as follows: receiving electromagnetic signals fed by the antenna horn and providing excitation signals for the target signal simulation system;

[0017] The antenna horn function includes at least: signal output for air-fed target signal simulation system and electronic interference simulation system;

[0018] The target signal simulation system has at least the following functions: outputting simulated target signals to the antenna horn and providing excitation signals for the electronic interference simulation system.

[0019] In one specific embodiment, the electronic interference simulation system functions at least as follows: outputting different types of interference signals to the antenna horn;

[0020] The data bus host computer functions include at least: monitoring and recording electromagnetic signal data uploaded by the air-to-air radar seeker, controlling the air-to-air radar seeker, controlling the target signal simulation system, and controlling the electronic jamming simulation system.

[0021] In one specific implementation, the data mentioned in the third step is:

[0022] The horizontal axis of each set of data represents time, with the time allowed for the air-to-air radar seeker to acquire the target as the zero point of the horizontal axis; the vertical axis of each set of data represents the electromagnetic information received by the antenna output by the air-to-air radar seeker, the target electromagnetic information uploaded by the target signal simulation system, and the electronic interference data information uploaded by the electronic interference simulation system.

[0023] In one specific implementation, after preprocessing the dataset, the preprocessed dataset is divided into a training dataset and a detection dataset, specifically including:

[0024] After obtaining the dataset through numerous experiments, the dataset is preprocessed, including data completion, standardization, normalization, and fuzzification.

[0025] The preprocessed dataset is divided into a training dataset and a detection dataset, with a ratio of 4:1 between the number of data groups in the training dataset and the detection dataset. The data groups in the preprocessed dataset are randomly assigned to the training dataset and the detection dataset.

[0026] Secondly, this invention discloses an electromagnetic interference identification system based on an LSTM neural network, the system comprising: an air-to-air radar seeker, an antenna horn, a target signal simulation system, an electronic interference simulation system, a data bus host computer, and an electromagnetic interference identification model based on an LSTM neural network.

[0027] The air-to-air radar seeker is connected to the target signal simulation system, the air-to-air radar seeker is connected to the data bus host computer, the antenna horn is connected to the target signal simulation system, the antenna horn is connected to the electronic jamming simulation system, the target signal simulation system is connected to the electronic jamming simulation system, the target signal simulation system is connected to the data bus host computer, the electronic jamming simulation system is connected to the data bus host computer, and the data bus host computer is connected to the electromagnetic interference identification model based on LSTM neural network.

[0028] In one specific embodiment, the air-to-air radar seeker functions at least as follows: receiving electromagnetic signals fed by the antenna horn and providing excitation signals for the target signal simulation system;

[0029] The antenna horn function includes at least: signal output for air-fed target signal simulation system and electronic interference simulation system;

[0030] The target signal simulation system has at least the following functions: outputting simulated target signals to the antenna horn and providing excitation signals for the electronic interference simulation system.

[0031] In one specific embodiment, the electronic interference simulation system functions at least as follows: outputting different types of interference signals to the antenna horn;

[0032] The data bus host computer functions include at least: monitoring and recording electromagnetic signal data uploaded by the air-to-air radar seeker, controlling the air-to-air radar seeker, controlling the target signal simulation system, and controlling the electronic jamming simulation system.

[0033] Thirdly, a computing device is provided, comprising at least one processor and at least one memory, wherein the memory stores a computer program, and the processor is configured to read the computer program from the memory and execute any step of the method described in the first aspect.

[0034] Fourthly, a computer-readable storage medium is provided, the computer-readable storage medium storing computer-executable instructions for causing a computer to perform any step of the method described in the first aspect.

[0035] The beneficial technical effects of this invention are:

[0036] This invention significantly improves the versatility of electromagnetic interference (EMI) identification models. Different EMI identification models can be universally applied across different radar seeker models simply by retraining the LSTM neural network-based model using a large amount of field-tested data. This EMI identification model is practical and effective, greatly improving the versatility of radar seeker EMI identification models and saving time and economic costs. Attached Figure Description

[0037] Figure 1 Block diagram of an electromagnetic interference identification system based on LSTM neural network.

[0038] 1. Air-to-air radar seeker 2. Antenna horn 3. Target signal simulation system 4. Electronic jamming simulation system

[0039] 5. Data bus host computer 6. Electromagnetic interference identification model based on LSTM neural network Detailed Implementation

[0040] To address the problem that existing electromagnetic interference identification models cannot be universally applied across different models, this invention provides an electromagnetic interference identification method and system based on an LSTM neural network.

[0041] The terms "first," "second," etc., used in the specification, claims, 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 the embodiments described herein can be implemented in a sequence other than that illustrated or described herein.

[0042] In this article, "multiple or several" 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: A alone, A and B simultaneously, or B alone. The character " / " generally indicates that the preceding and following related objects have an "or" relationship.

[0043] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention. Furthermore, the embodiments and features in the embodiments of the present invention can be combined with each other without conflict.

[0044] Example 1

[0045] like Figure 1 The diagram shown is a block diagram of an electromagnetic interference identification system based on an LSTM neural network. LSTM stands for Long Short-Term Memory.

[0046] An electromagnetic interference identification method based on LSTM neural network, the specific steps of which are as follows:

[0047] The first step is to construct an indoor field test system for radar seeker anti-jamming.

[0048] The radar seeker anti-jamming indoor test system includes: air-to-air radar seeker 1, antenna horn 2, target signal simulation system 3, electronic jamming simulation system 4, data bus host computer 5, and electromagnetic interference identification model based on LSTM neural network 6.

[0049] The air-to-air radar seeker 1 is connected to the target signal simulation system 3, the air-to-air radar seeker 1 is connected to the data bus host computer 5, the antenna horn 2 is connected to the target signal simulation system 3, the antenna horn 2 is connected to the electronic jamming simulation system 4, the target signal simulation system 3 is connected to the electronic jamming simulation system 4, the target signal simulation system 3 is connected to the data bus host computer 5, the electronic jamming simulation system 4 is connected to the data bus host computer 5, and the data bus host computer 5 is connected to the electromagnetic interference identification model 6 based on the LSTM neural network.

[0050] The functions of the air-to-air radar seeker 1 are: to receive electromagnetic signals fed by the antenna horn 2 and to provide excitation signals for the target signal simulation system 3; the functions of the antenna horn 2 are: to feed the signal output of the target signal simulation system 3 and the electronic jamming simulation system 4; the functions of the target signal simulation system 3 are: to output simulated target signals to the antenna horn 2 and to provide excitation signals for the electronic jamming simulation system 4; the functions of the electronic jamming simulation system 4 are: to output different types of jamming signals to the antenna horn 2; the functions of the data bus host computer 5 are: to monitor and record the electromagnetic signal data uploaded by the air-to-air radar seeker 1, to control the air-to-air radar seeker 1, to control the target signal simulation system 3, and to control the electronic jamming simulation system 4.

[0051] The second step is to acquire the test data from the host computer 5 on the data bus.

[0052] After the indoor test and evaluation system is set up, the parameters of the target signal simulation system 3 are set and the simulated target signal is output. The air-to-air radar seeker 1 is controlled by the data bus host computer 5 to intercept and stably track the simulated target signal. Then, the parameters of the electronic jamming simulation system 4 are set and the electronic jamming signal is output to simulate the battlefield electromagnetic environment under complex electromagnetic conditions. The data bus host computer 5 records the missile-to-target distance information output by the air-to-air radar seeker 1 in real time. The target signal simulation system 3 records the information of the output simulated target signal, and the electronic jamming simulation system 4 records the output information.

[0053] The third step involves test data processing and statistics.

[0054] During the test, the host computer 5 on the data bus records combinations of different simulated target signals and different electronic interference signals. The electromagnetic information received by the antenna output by the air-to-air radar seeker 1, the target electromagnetic information uploaded by the target signal simulation system 3, and the electronic interference data information uploaded by the electronic interference simulation system 4 in the same test are packaged into a set of data in the dataset.

[0055] Specifically, the horizontal axis of each set of data represents time, with the allowed target acquisition time of the air-to-air radar seeker 1 as the zero point of the horizontal axis. The vertical axis of each set of data represents the electromagnetic information received by the antenna output by the air-to-air radar seeker 1, the target electromagnetic information uploaded by the target signal simulation system 3, and the electronic interference data information uploaded by the electronic interference simulation system 4.

[0056] Specifically, after obtaining the dataset through numerous experiments, the dataset undergoes data preprocessing, including data completion, standardization, normalization, and fuzzing. The preprocessed dataset is then divided into a training dataset and a detection dataset, with a data group ratio of 4:1. The data groups in the preprocessed dataset are randomly assigned to the training and detection datasets.

[0057] Step 4: Training and Detection of the Electromagnetic Interference Recognition Model using an LSTM Neural Network

[0058] The training dataset obtained in the previous step is used as input data to train the LSTM neural network electromagnetic interference recognition model 6. The detection dataset is then used as detection data to input the LSTM neural network electromagnetic interference recognition model 6 for detection. The detection results are obtained, and the parameters of the LSTM neural network electromagnetic interference recognition model 6 are tuned and optimized based on the detection results. The tuned and optimized LSTM neural network electromagnetic interference recognition model 6 is then used as a new model, and the above training and detection process is repeated until the detection results meet the accuracy requirements.

[0059] This completes the entire process of creating an electromagnetic interference identification model based on an LSTM neural network.

[0060] This embodiment significantly improves the versatility of the electromagnetic interference (EMI) identification model. Different EMI identification models can be universally applied across different radar seeker models simply by retraining the LSTM neural network-based model using a large amount of field measurement data. This EMI identification model is practical and effective, greatly improving the versatility of radar seeker EMI identification models and saving time and economic costs.

[0061] Example 2

[0062] like Figure 1 As shown, it is a block diagram of an electromagnetic interference identification system based on an LSTM neural network.

[0063] An indoor test and evaluation system for radar seeker anti-drag interference includes: an air-to-air radar seeker 1, an antenna horn 2, a target signal simulation system 3, an electronic interference simulation system 4, a data bus host computer 5, and an electromagnetic interference identification model 6 based on an LSTM neural network; the air-to-air radar seeker 1 is connected to the target signal simulation system 3, the air-to-air radar seeker 1 is connected to the data bus host computer 5, the antenna horn 2 is connected to the target signal simulation system 3, the antenna horn 2 is connected to the electronic interference simulation system 4, the target signal simulation system 3 is connected to the electronic interference simulation system 4, the target signal simulation system 3 is connected to the data bus host computer 5, the electronic interference simulation system 4 is connected to the data bus host computer 5, and the data bus host computer 5 is connected to the electromagnetic interference identification model 6 based on an LSTM neural network.

[0064] The functions of the air-to-air radar seeker 1 are: to receive electromagnetic signals fed by the antenna horn 2 and to provide excitation signals for the target signal simulation system 3; the functions of the antenna horn 2 are: to feed the signal output of the target signal simulation system 3 and the electronic jamming simulation system 4; the functions of the target signal simulation system 3 are: to output simulated target signals to the antenna horn 2 and to provide excitation signals for the electronic jamming simulation system 4; the functions of the electronic jamming simulation system 4 are: to output different types of jamming signals to the antenna horn 2; the functions of the data bus host computer 5 are: to monitor and record the electromagnetic signal data uploaded by the air-to-air radar seeker 1, to control the air-to-air radar seeker 1, to control the target signal simulation system 3, and to control the electronic jamming simulation system 4.

[0065] This completes the entire process of developing an electromagnetic interference identification system based on an LSTM neural network.

[0066] This invention discloses an electromagnetic interference (EMI) identification method and system based on an LSTM neural network. An experimental evaluation system is built, including a radar seeker model, a target signal simulation system, an electronic interference simulation system, and an antenna horn. By outputting various interference signals, such as simulated signals from different targets, range-towed decoy interference signals, velocity-towed decoy interference signals, and suppression noise interference signals, the system simulates a combat scenario where a radar seeker intercepts and tracks simulated target signals under complex electromagnetic conditions. Through the construction of this system, electromagnetic signal data of different targets under different interference conditions are obtained. This data is then used to train and validate the constructed LSTM neural network model. This model is universal, simple to implement, and can be quickly ported to radar seeker anti-jamming models, improving the accuracy of identifying different types of EMI. It solves the problem that existing seeker anti-jamming models lack versatility and cannot achieve mutual compatibility of EMI identification models between different seeker models.

[0067] For ease of description, the above sections are divided into modules (or units) according to their functional modules and described separately. Of course, in implementing this invention, the functions of each module (or unit) can be implemented in one or more software or hardware components.

[0068] Based on the same technical concept, the present invention provides a computing device, including at least one processor and at least one memory, wherein the memory stores a computer program, and the processor is used to read the computer program in the memory and execute an electromagnetic interference identification method based on an LSTM neural network.

[0069] Based on the same technical concept, the present invention provides a computer-readable storage medium storing computer-executable instructions for causing a computer to execute an electromagnetic interference identification method based on an LSTM neural network.

[0070] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. An electromagnetic interference identification method based on LSTM neural network, characterized in that, The specific steps to achieve this are as follows: The first step is to construct an indoor field test system for radar seeker anti-jamming. The radar seeker anti-jamming indoor test system includes: air-to-air radar seeker (1), antenna horn (2), target signal simulation system (3), electronic jamming simulation system (4), data bus host computer (5), and electromagnetic interference identification model based on LSTM neural network (6). The air-to-air radar seeker (1) is connected to the target signal simulation system (3), the air-to-air radar seeker (1) is connected to the data bus host computer (5), the antenna horn (2) is connected to the target signal simulation system (3), the antenna horn (2) is connected to the electronic interference simulation system (4), the target signal simulation system (3) is connected to the electronic interference simulation system (4), the target signal simulation system (3) is connected to the data bus host computer (5), the electronic interference simulation system (4) is connected to the data bus host computer (5), and the data bus host computer (5) is connected to the electromagnetic interference identification model (6) based on the LSTM neural network. The second step is to obtain the test data from the host computer (5) on the data bus. After the radar seeker anti-jamming indoor test system is set up, the target signal simulation system (3) parameters are set and the simulated target signal is output. The air-to-air radar seeker (1) is controlled by the data bus host computer (5) to intercept and stably track the simulated target signal. Then, the electronic jamming simulation system (4) parameters are set and the electronic jamming signal is output to simulate the battlefield electromagnetic environment under complex electromagnetic conditions. The data bus host computer (5) records the missile-to-target distance information output by the air-to-air radar seeker (1) in real time. The target signal simulation system (3) records the information of the simulated target signal output, and the electronic jamming simulation system (4) records the information of the output. The third step involves test data processing and statistics. During the test, the host computer (5) on the data bus records the combination of different simulated target signals and different electronic interference signals. The electromagnetic information received by the antenna output by the air-to-air radar seeker (1) in the same test, the target electromagnetic information uploaded by the target signal simulation system (3), and the electronic interference data information uploaded by the electronic interference simulation system (4) are packaged into a set of data in the dataset. After preprocessing the dataset, the preprocessed dataset is divided into a training dataset and a detection dataset. The data mentioned in step three specifically refers to: The horizontal axis of each set of data is time, with the time when the air-to-air radar seeker (1) is allowed to intercept the target as the zero point of the horizontal axis; the vertical axis of each set of data is the electromagnetic information received by the antenna output by the air-to-air radar seeker (1), the target electromagnetic information uploaded by the target signal simulation system (3), and the electronic interference data information uploaded by the electronic interference simulation system (4). After preprocessing the dataset, the preprocessed dataset is divided into a training dataset and a detection dataset, specifically including: After obtaining the dataset through numerous experiments, the dataset is preprocessed, including data completion, standardization, normalization, and fuzzification. The preprocessed dataset is divided into a training dataset and a detection dataset, with a ratio of 4:1 between the number of data groups in the training dataset and the detection dataset. The data groups in the preprocessed dataset are randomly assigned to the training dataset and the detection dataset. Step 4: Training and Detection of the Electromagnetic Interference Recognition Model using an LSTM Neural Network The training dataset obtained in the third step is used as input data to train the electromagnetic interference recognition model (6) of the LSTM neural network, and the detection dataset is used as detection data to input the electromagnetic interference recognition model (6) of the LSTM neural network for detection. The detection results are obtained, and the parameters of the electromagnetic interference recognition model (6) of the LSTM neural network are tuned and optimized based on the detection results. The electromagnetic interference recognition model (6) of the LSTM neural network after parameter tuning and optimization is used as a new model to repeat the above training and detection process until the detection results meet the accuracy requirements. This completes all the work on the electromagnetic interference identification method based on LSTM neural network; The functions of the air-to-air radar seeker (1) include at least: receiving electromagnetic signals fed by the antenna horn (2) and providing excitation signals for the target signal simulation system (3); The antenna horn (2) has at least the following functions: signal output for the air-feed target signal simulation system (3) and the electronic interference simulation system (4); The target signal simulation system (3) has at least the following functions: outputting simulated target signals to the antenna horn (2) and providing excitation signals for the electronic interference simulation system (4); The electronic interference simulation system (4) has at least the following functions: outputting different types of interference signals to the antenna horn (2); The data bus host computer (5) has at least the following functions: monitoring and recording electromagnetic signal data uploaded by the air-to-air radar seeker (1), controlling the air-to-air radar seeker (1), controlling the target signal simulation system (3), and controlling the electronic interference simulation system (4).

Images