Artillery outer trajectory prediction method, computer equipment and storage medium

Abstract

The invention relates to the technical field of signal processing, in particular to an artillery outer ballistic trajectory prediction method, computer equipment and a storage medium, and the method comprises the steps: obtaining an artillery outer ballistic trajectory signal in a complex environment; generating a modal number and a penalty factor in a VMD algorithm by using a group optimization algorithm, and performing VMD decomposition on the obtained artillery outer ballistic signal to obtain a plurality of intrinsic modal components; screening the plurality of obtained intrinsic mode components to obtain effective signal components; superposing all the effective modes to obtain a de-noised outer ballistic signal; and according to the obtained de-noised outer ballistic signal, performing outer ballistic signal prediction by using an LSTM model. According to the method, group optimization and VMD are combined, and high-frequency impulse noise hidden in a complex signal is accurately separated from a low-frequency ballistic trend; a pure-data-driven LSTM network is adopted, future artillery outer ballistic trajectory signals are deduced by learning dynamic characteristics of historical trajectory data, and explicit meteorological parameter input is not needed.

Description

Technical Field

[0001] This invention belongs to the field of signal processing technology, and particularly relates to a method for predicting the external trajectory of artillery, a computer device, and a storage medium. Background Technology

[0002] Traditional methods for predicting the external trajectory of artillery primarily rely on physical mechanism modeling, namely establishing a six-degree-of-freedom (6-DOF) or point-mass external trajectory differential equation system based on Newton's laws of motion, and performing numerical integration extrapolation using a high-order Runge-Kutta algorithm. While the physical meaning of these methods is clear, their prediction accuracy is highly dependent on the accurate identification of parameters such as air drag coefficients and ballistic coefficients, as well as the acquisition of real-time meteorological data. However, under actual field conditions, real-time full-trajectory meteorological profiles are often difficult to obtain accurately, and solving complex differential equations is time-consuming, leading to errors accumulating with range. This makes them unsuitable for scenarios with extremely high real-time requirements, such as anti-missile interception.

[0003] To overcome the limitations of physical models, data-driven prediction methods have gradually become a research hotspot, especially the use of Kalman filtering (KF) and its derivative algorithms (such as EKF and UKF) to process radar data. However, traditional filtering methods typically assume that noise follows a Gaussian distribution, while in complex electromagnetic interference environments, radar echo signals often contain non-stationary and nonlinear complex noise (such as radar scintillation noise and impulse noise caused by active interference). In this case, Kalman filtering is prone to filtering lag or failure to track abrupt changes in signals. In recent years, Long Short-Term Memory (LSTM) networks have been introduced into this field due to their powerful temporal modeling capabilities; however, single neural network models have weak noise resistance, and if heavily noisy raw data is directly input, the model is prone to overfitting the noise features, leading to a decrease in generalization ability.

[0004] To address noise issues, signal decomposition techniques are used as preprocessing methods. Among them, Variational Mode Decomposition (VMD) offers better frequency domain sparsity and resistance to mode aliasing compared to Empirical Mode Decomposition (EMD). However, the decomposition effect of VMD is highly dependent on the selection of two key parameters: the number of modes and the second-order penalty factor. Currently, these two parameters are mostly determined manually or through trial and error, lacking adaptability to different battlefield environments and limiting their application in actual combat systems. Therefore, there is an urgent need for a combined method that can adaptively optimize decomposition parameters, effectively eliminate non-stationary noise, and perform high-precision prediction. Summary of the Invention

[0005] In view of this, the present invention aims to provide a method for predicting the external trajectory of artillery, which uses a swarm intelligence optimization algorithm to adaptively optimize the variational mode decomposition parameters and combines a long short-term memory network for time series prediction.

[0006] To achieve the above objectives, the technical solution created by this invention is implemented as follows: A method for predicting the external ballistics of artillery, comprising: S1: Obtain the external ballistic signal of the artillery under complex conditions; S2: Use a swarm optimization algorithm to generate the number of modes and the penalty factor in the VMD algorithm; S3: Based on the number of modes and the penalty factor obtained in step S2, perform VMD decomposition on the artillery external ballistic signal obtained in step S1 to obtain multiple intrinsic mode components. S4: Filter the multiple intrinsic mode components obtained in step S3 to obtain the effective signal components; superimpose all effective modes to obtain the denoised external ballistic signal; S5: Based on the denoised external ballistic signal obtained in step S4, use the LSTM model to predict the external ballistic signal.

[0007] Furthermore, in step S2, the Pelican optimization algorithm is used to generate the number of modes and the penalty factor; during the optimization process, the minimum envelope entropy is used as the fitness function.

[0008] Furthermore, step S4 includes: calculating the Pearson correlation coefficient between each intrinsic mode component and the artillery external ballistic signal obtained in step S1; setting a correlation coefficient threshold and retaining intrinsic mode components with correlation coefficients greater than or equal to the threshold as valid signal components; and superimposing all valid modes to reconstruct a pure denoised external ballistic signal.

[0009] Furthermore, step S5 includes: processing the denoised external ballistic signal using the sliding window method to obtain learning samples; and inputting the learning samples into the LSTM model in time sequence to obtain the predicted external ballistic signal.

[0010] Furthermore, the LSTM model in step S5 includes an input layer, multiple LSTM hidden layers, and a fully connected output layer. The input signal enters the multiple LSTM hidden layers via the input layer. These hidden layers include forget gates, input gates, and output gates, which extract features from the input signal to obtain deep features. The fully connected output layer performs a fully connected operation on these deep features to obtain the predicted external ballistic signal.

[0011] Furthermore, in step S5, the training process of the LSTM model includes: generating simulated artillery external ballistic signals under complex environments; repeating steps S2 to S4 to obtain denoised simulated artillery external ballistic signals; processing the denoised simulated artillery external ballistic signals using the sliding window method; and training the constructed LSTM network model using the processed signals.

[0012] Furthermore, the process of generating simulated artillery external ballistic signals under complex environments includes: using the fourth-order Runge-Kutta method to solve the differential equations of the external ballistics of a particle to generate noiseless ideal artillery external ballistic signals; introducing a Gaussian mixture model to introduce non-stationary noise into the ideal artillery external ballistic signals to obtain simulated artillery external ballistic signals.

[0013] Furthermore, the Gaussian mixture model is a two-component Gaussian mixture model, and the non-stationary noise includes background noise used to simulate conventional radar ranging errors, and anomalous impulse noise used to simulate strong electromagnetic interference or radar scintillation effects.

[0014] Compared with the prior art, the present invention can achieve the following beneficial effects: The artillery external ballistic prediction method described in this invention utilizes a swarm optimization algorithm to automatically search for the globally optimal mode number and penalty factor without manual intervention. This allows it to adapt to data from different caliber artillery, different firing environments, and different radar sampling rates, thus adapting to the ever-changing battlefield environment. By leveraging the inherent frequency domain separation characteristics of the VMD algorithm, combined with the parameters of swarm optimization, it can accurately separate high-frequency pulse noise hidden in complex signals from low-frequency ballistic trends, effectively removing "glitch" and outliers from radar data and providing a clean data source for subsequent predictions. Furthermore, by employing a purely data-driven LSTM network, it learns the dynamic characteristics of historical trajectory data to predict the future without requiring explicit meteorological parameter input. Under the same noise interference, it can meet the precision strike requirements of modern artillery's "first-strike coverage." Attached Figure Description

[0015] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments and descriptions of the invention are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings: Figure 1 A flowchart illustrating the artillery external ballistic prediction method described in an embodiment of the present invention; Figure 2 A flowchart illustrating the artillery external ballistic prediction method described in the embodiments of the present invention; Figure 3 A schematic diagram of the computer device described in an embodiment of the present invention.

[0016] Explanation of reference numerals in the attached figures: 1. Electronic device; 2. External device; 3. Processing unit; 4. Bus; 5. Network adapter; 6. Display; 7. (I / O) interface; 8. System memory; 9. Random access memory; 10. Cache memory; 11. Storage system; 12. Utility; 13. Program module. Detailed Implementation

[0017] 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 specific embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and do not constitute a limitation thereof.

[0018] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0019] like Figures 1 to 2 As shown, the artillery external ballistic prediction method described in the embodiments of the present invention includes: S1: Obtain the external ballistic signal of the artillery under complex conditions. The external ballistic signal is a three-dimensional trajectory, which can be represented as s(t) = [x(t), y(t), z(t)]. T .

[0020] S2: Utilize a swarm optimization algorithm to generate the modality number and penalty factor in the VMD algorithm. In some embodiments, the pelican optimization algorithm is used to generate the modality number and penalty factor.

[0021] To address the difficulty in determining the number of modes K and the penalty factor α in the VMD algorithm, this invention introduces the Pelican Optimization Algorithm (POA) for global optimization. POA is a swarm intelligence algorithm that mimics the predatory behavior of pelicans, characterized by fast convergence speed and high optimization accuracy. Specifically, the optimization process of the Pelican Optimization Algorithm consists of two stages: finding the optimal number of modes K_best and the penalty factor α_best. The first stage is the exploration stage (approaching the prey): the process by which individual pelicans simulate discovering prey (the potential region of the optimal parameter solution) and moving toward it. In this stage, the population is guided to move toward regions with better fitness by randomly updating the position vector. This ensures a strong global search capability and avoids getting trapped in local optima. The second stage is the development stage (surface predation): after the pelican reaches above the prey, its behavior of circling and hunting on the water surface is simulated. In this stage, the algorithm linearly decreases the search step size as the number of iterations increases, and performs a refined search in the neighborhood of the optimal solution, thereby significantly improving the accuracy of parameter solving.

[0022] In some embodiments, minimum envelope entropy is used as the fitness function in the optimization process. Envelope entropy reflects the sparsity of a signal, and its physical property—lower entropy values ​​indicate clearer features and less noise—makes it an indicator for evaluating VMD decomposition. For ballistic signals, effective physical trajectories exhibit strong regularity (low entropy), while noise signals exhibit high disorder (high entropy). Therefore, finding the parameter combination that minimizes the envelope entropy of the VMD decomposition components means finding the optimal decomposition strategy that can maximally separate noise from effective signals.

[0023] Specifically, for any set of candidate parameters (K, α), VMD decomposes the gun's external ballistic signal s(t) under complex conditions obtained in step S1 into K eigenmode components. For each mode μ k (t) Construct its analytic signal z k (t): z k (t)=μ k (t)+jH{μ k (t)}; Where H represents the Hilbert transform, and j represents the imaginary part; Obtain the analytic signal z k The instantaneous envelope A of (t) k (t): ; Instantaneous envelope A k Discrete sampling is performed on (t) to obtain the envelope sequence. For envelope sequences Normalized to probability distribution p k,j : ; Therefore, the envelope entropy E of the k-th mode is defined. k (K,α): ; Since noise-dominated components tend to exhibit more "random" envelope fluctuations, resulting in higher entropy values, while components containing the main effective structural / impact features have more "concentrated" envelopes and relatively lower entropy values, the minimum envelope entropy is defined as the evaluation index J(K,α) under this set of parameters: ; And using this evaluation index J(K,α), we construct the optimization objective for POA:

[0024] The objective function described above ensures that the influence of different (K,α) parameters on modal morphology can be effectively reflected, thus providing a clear evaluation basis for the global search of POA and ultimately obtaining the optimal VMD parameters (K). best ,α best ).

[0025] S3: Based on the mode number and penalty factor obtained in step S2, perform VMD decomposition on the artillery external ballistic signal obtained in step S1 to obtain multiple intrinsic mode components. It is understandable that the optimal mode number K obtained in step S2... best and penalty factor α best The VMD algorithm can be determined, which decomposes the complex external ballistic signal of the artillery into K... best (i.e., mode number) intrinsic mode components (IMFs) with different center frequencies.

[0026] S4: Filter the multiple intrinsic mode components obtained in step S3 to obtain the effective signal components; superimpose all effective modes to obtain the denoised external ballistic signal.

[0027] In some embodiments, step S4 includes: calculating the Pearson correlation coefficient (CC) between each intrinsic mode component and the artillery external ballistic signal obtained in step S1; setting a correlation coefficient threshold and retaining intrinsic mode components with correlation coefficients greater than or equal to the threshold as valid signal components; and superimposing all valid modes to reconstruct a clean, denoised external ballistic signal.

[0028] Understandably, after completing the signal decomposition in step S3, the Pearson correlation coefficient (CC) between each IMF and the original observed signal (i.e., the external ballistic signal obtained in step S1) is calculated. Since noise is typically distributed in high-frequency and cluttered modes, it has a low correlation with the overall ballistic trend but a high correlation with low-frequency modes containing the true motion patterns. This invention sets a correlation coefficient threshold, retaining modes with correlation coefficients greater than this threshold as valid signal components and discarding noise modes with correlation coefficients lower than the threshold. Finally, all valid modes are superimposed to reconstruct a clean ballistic trajectory sequence.

[0029] In this embodiment of the invention, step S4 specifically includes: For discrete raw observation signals and discrete each IMF The Pearson correlation coefficient between the two is calculated to be CC. k : ; in, , As can be seen from the definition, When |CC kThe closer |CC is to 1, the stronger the correlation between the IMF and the original observation signal; when |CC k If the value is small, the IMF may correspond to high-frequency noise or disturbance components unrelated to the effective structure, which can be removed during denoising and reconstruction.

[0030] Since the external ballistic signal of the artillery is a three-dimensional trajectory, i.e., s(t)=[x(t),y(t),z(t)] T The IMF corresponding to the k-th mode is μ. k (t)=[μ k,x (t),μ k,y (t),μ k,z (t)] T Based on the above formula, calculate the correlation coefficients CC of the x-axis, y-axis, and z-axis respectively. k,x CC k,y and CC k,z The correlation coefficients of the three axes are fused using the following formula to obtain the overall correlation index between the original observation signal and each IMF: ; Where, ω x ω y and ω z ω represents the weight of the correlation coefficients of the three axes in the overall correlation index. x +ω y +ω z =1. In this embodiment of the invention, let ω = 1 when there is no prior preference. x =ω y =ω z =1 / 3, thus comprehensively reflecting the correlation on each coordinate axis.

[0031] Filtering rules can be set based on correlation coefficient results. Thresholds can be set. , retain the condition that satisfies |CC k |≥τ|IMF. Therefore, the preserved mode set is defined as: S={k||CC k |≥τ}; Finally, the denoised external ballistic signal s is obtained. clean (t) is: .

[0032] In step S4 of this invention, by suppressing the noise-dominant component, a smoother trajectory sequence with more concentrated information is obtained, thereby providing high-quality input for subsequent LSTM time series modeling and prediction.

[0033] S5: Based on the denoised external ballistic signal obtained in step S4, use the LSTM model to predict the external ballistic signal.

[0034] In some embodiments, step S5 includes: processing the denoised external ballistic signal using a sliding window method to obtain learning samples; and inputting the learning samples into the LSTM model in time sequence to obtain the predicted external ballistic signal.

[0035] In some embodiments, the LSTM model includes an input layer, multiple LSTM hidden layers, and a fully connected output layer. The input signal enters the multiple LSTM hidden layers through the input layer. These hidden layers include a forget gate, an input gate, and an output gate to extract features from the input signal, obtaining deep features. The fully connected output layer performs a fully connected operation on the deep features to obtain the predicted external ballistic signal. The LSTM network model effectively solves the gradient vanishing problem inherent in traditional RNNs through the forget gate, input gate, and output gate, enabling it to capture long-distance temporal dependencies and making it highly suitable for processing sequential data with strong temporal correlation, such as ballistic signals. In this embodiment, the LSTM model also includes a Dropout layer to prevent overfitting. Furthermore, the loss function used to train the LSTM model is the mean squared error loss function.

[0036] In some embodiments, the training process of the LSTM model includes: generating simulated artillery external ballistic signals under complex environments; repeating steps S2 to S4 to obtain denoised simulated artillery external ballistic signals; processing the denoised simulated artillery external ballistic signals using the sliding window method; and training the constructed LSTM network model using the processed signals.

[0037] In this embodiment of the invention, the denoised simulated artillery external ballistic signal is normalized, and the normalized signal is used to train the constructed LSTM network model; in addition, the signal output by the current LSTM network model is inversely normalized to obtain the current output predicted external ballistic signal.

[0038] In some embodiments, the process of generating simulated artillery external ballistic signals under complex environments includes: Based on the differential equations of the external ballistics of a point mass, a noiseless ideal artillery external ballistic signal is generated by solving the fourth-order Runge-Kutta method. By introducing a Gaussian mixture model, non-stationary noise is introduced into the ideal artillery external ballistic signal to obtain the simulated artillery external ballistic signal.

[0039] In this embodiment of the invention, the process of generating the ideal artillery external ballistic signal specifically includes: Treating the artillery projectile as an ideal point mass, neglecting its rotation around its center of mass and the asymmetry of its structure. With the center of the gun muzzle as the origin O, and the horizontal line in the firing plane as O... x The axis, vertically upward, is O. y The axis is determined by the right-hand rule.z Establish a ground-fixed coordinate system with the axis as an axis, where the gravitational acceleration g is constant. During flight, the artillery projectile is primarily affected by gravity G and air resistance. The effect of air resistance The direction of the velocity vector of the artillery projectile relative to the air is always opposite to that of the projectile's velocity vector. According to Newton's second law, the equation of motion for the center of mass of the artillery projectile is: ; in, This represents the position vector of an artillery projectile. Let m represent the velocity vector of the artillery projectile, and m represent the mass of the artillery projectile. This represents the gravitational acceleration vector, where gravitational acceleration g = -9.8 m / s². 2 .

[0040] air resistance for: ; Where S represents the characteristic area of ​​the artillery projectile, C D The drag coefficient is a function of the Mach number Ma. This represents the velocity vector of an artillery projectile relative to the air. Represents the wind speed vector, v s Represents the relative airspeed, and ρ represents the atmospheric density function as a function of altitude y, specifically: ; Where ρ0 represents the atmospheric density at sea level, M represents the molar mass of air, M=0.029kg / mol, R represents the universal gas constant, R=8.314J / (mol×K), and T represents the atmospheric thermodynamic temperature, T=288K.

[0041] Projecting the above vector equations onto a ground-fixed coordinate system, let the drag acceleration factor... Furthermore, since they are completely consistent in the three-axis terms, we can obtain the following system of six first-order ordinary differential equations: ; The above equation is solved numerically using the fourth-order Runge-Kutta method: Define the state vector at time t Using the above formula to analyze the state vector Differentiation, simplified as The state update formula from time t to time t+1 is: ; Where h represents the integration step size, and K1~K4 represent the four slope terms, as follows: ; In each iterative calculation, the speed of sound *a* and atmospheric density *ρ* are updated in real time based on the current geometric height *y*, thereby updating the Mach number *Ma* = *v*. s / a and drag coefficient C D (Ma), thereby accurately simulating the external ballistic trajectory of artillery in complex environments.

[0042] In some embodiments, the Gaussian mixture model is a two-component Gaussian mixture model. The non-stationary noise includes background noise used to simulate conventional radar ranging errors, and anomalous impulse noise used to simulate strong electromagnetic interference or radar scintillation effects. The background noise has a smaller variance, while the anomalous impulse noise has a larger variance. This bimodal distribution of non-stationary noise accurately reflects the intermittent "outliers" and data mutations that occur when the radar tracks artillery.

[0043] In this embodiment of the invention, a two-component Gaussian mixture model is introduced to introduce non-stationary noise into the ideal artillery external ballistic signal, thereby obtaining a simulated artillery external ballistic signal. The process specifically includes: Suppose that the radar's observation noise μ(t) of the target at time t follows the following probability distribution: p(μ(t))=(1-λ)N(0,σ n 2 )+λN(0,σ g 2 ); Where p represents the probability distribution, λ represents the noise weight, and N(0,σ) n 2 () represents the background noise under normal observation conditions, and the standard deviation σ is set. n The value is 5, corresponding to the ranging accuracy of conventional radar, N(0,σ) g 2 ) represents abnormal impulse noise, σ g It represents the standard deviation of abnormal impulse noise, used to simulate large-amplitude measurement jumps caused by radar target scintillation or strong electromagnetic interference.

[0044] In this embodiment of the invention, before processing the denoised external ballistic signal using the sliding window method in step S5, the method further includes: performing data normalization processing on the denoised external ballistic signal, that is, processing the normalized signal using the sliding window method to obtain learning samples. Furthermore, the output signal of the LSTM model is inversely normalized to obtain the predicted external ballistic signal.

[0045] Understandably, step S5 includes: Perform data normalization processing on the denoised external ballistic signal; The normalized and denoised external ballistic signal is processed using the sliding window method to obtain the learning samples {X(tw), X(tw-1), ..., X(t-1), X(t)}, where t represents time and w represents the window width; The training samples {X(tw), X(tw-1), ..., X(t-1), X(t)} are input into the trained LSTM model LSTM_Train for time series modeling to obtain the output signal; The output signal is inversely normalized to obtain the predicted external ballistic signal. Accordingly, according to embodiments of the present invention, the present invention also provides a computer device, a readable storage medium, and a computer program product.

[0046] Figure 3 This is a schematic diagram of the structure of an electronic device 1 provided in an embodiment of the present invention. Figure 3 A block diagram is shown of an exemplary electronic device 1 suitable for implementing embodiments of the present invention. Figure 3 The electronic device 1 shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of the present invention.

[0047] like Figure 3 As shown, electronic device 1 is represented in the form of a general-purpose computing device. Electronic device 1 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. Electronic device 1 can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0048] The components of electronic device 1 may include, but are not limited to: one or more processors or processing units 3, system memory 8, and bus 4 connecting different system components (including system memory 8 and processing unit 3).

[0049] Bus 4 represents one or more of several bus architectures, including memory buses or memory controllers, peripheral buses, graphics acceleration ports, processors, or local buses using any of the various bus architectures. Examples of these architectures include, but are not limited to, the Industry Standard Architecture (ISA) bus, the Micro Channel Architecture (MAC) bus, the Enhanced ISA bus, the Video Electronics Standards Association (VESA) local bus, and the Peripheral Component Interconnect (PCI) bus.

[0050] Electronic device 1 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by electronic device 1, including volatile and non-volatile media, removable and non-removable media.

[0051] System memory 8 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) 9 and / or cache memory 10. Electronic device 1 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, storage system 11 may be used to read and write non-removable, non-volatile magnetic media (…). Figure 3 Not shown; usually referred to as a "hard drive"). Although Figure 3 Not shown, a disk drive for reading and writing to a removable non-volatile disk (e.g., a "floppy disk") and an optical disk drive for reading and writing to a removable non-volatile optical disk (e.g., a CD-ROM, DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to bus 4 via one or more data media interfaces. Storage system 11 may include at least one program product having a set (e.g., at least one) of program modules configured to perform the functions of the embodiments of the present invention.

[0052] A program / utility 12 having a set (at least one) of program modules 13 may be stored in, for example, a storage system 11. Such program modules 13 include, but are not limited to, an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include an implementation of a network environment. Program modules 13 typically perform the functions and / or methods described in the embodiments of the present invention.

[0053] Electronic device 1 can also communicate with one or more external devices 2 (e.g., keyboard, pointing device, display 6, etc.), and with one or more devices that enable a user to interact with electronic device 1, and / or with any device that enables electronic device 1 to communicate with one or more other computing devices (e.g., network card, modem, etc.). This communication can be performed via input / output (I / O) interface 7. Furthermore, electronic device 1 can also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public networks, such as the Internet) via network adapter 5. As shown, network adapter 5 communicates with other modules of electronic device 1 via bus 4. It should be understood that, although not shown in the figure, other hardware and / or software modules can be used in conjunction with electronic device 1, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0054] The processing unit 3 executes various functional applications and data processing by running programs stored in the system memory 8, such as implementing the artillery external ballistic prediction method provided in the embodiments of the present invention.

[0055] This invention also provides a non-transitory computer-readable storage medium storing computer instructions, on which a computer program is stored, wherein the program, when executed by a processor, is the artillery external ballistic prediction method provided in all embodiments of this application.

[0056] The computer storage medium of this invention can be any combination of one or more computer-readable media. The computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. More specific examples (a non-exhaustive list) of computer-readable storage media include: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this document, a computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0057] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, capable of sending, propagating, or transmitting programs for use by or in connection with an instruction execution system, apparatus, or device.

[0058] The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to wireless, wired, optical fiber, RF, etc., or any suitable combination thereof. The computer program code for performing the operations of this invention can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages—such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer can be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0059] This invention also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method for predicting the external ballistic trajectory of artillery.

[0060] It should be understood that the various forms of processes shown above can be used to reorder, add, or delete steps. For example, the steps described in this invention disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this invention can be achieved, and this is not limited herein.

[0061] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for predicting the external trajectory of artillery, characterized in that, include: S1: Obtain the external ballistic signal of the artillery under complex conditions; S2: Use a swarm optimization algorithm to generate the number of modes and the penalty factor in the VMD algorithm; S3: Based on the number of modes and the penalty factor obtained in step S2, perform VMD decomposition on the artillery external ballistic signal obtained in step S1 to obtain multiple intrinsic mode components. S4: Filter the multiple intrinsic mode components obtained in step S3 to obtain the effective signal components; superimpose all effective modes to obtain the denoised external ballistic signal; S5: Based on the denoised external ballistic signal obtained in step S4, use the LSTM model to predict the external ballistic signal.

2. The method for predicting the external ballistic trajectory of artillery according to claim 1, characterized in that, In step S2, the Pelican optimization algorithm is used to generate the number of modes and the penalty factor; during the optimization process, the minimum envelope entropy is used as the fitness function.

3. The method for predicting the external ballistic trajectory of artillery according to claim 1, characterized in that, Step S4 includes: Calculate the Pearson correlation coefficient between each intrinsic mode component and the gun external ballistic signal obtained in step S1; Set a correlation coefficient threshold and retain intrinsic mode components with a correlation coefficient greater than or equal to the threshold as valid signal components; By superimposing all effective modes, a clean, denoised external ballistic signal is reconstructed.

4. The method for predicting the external ballistic trajectory of artillery according to claim 1, characterized in that, Step S5 includes: The sliding window method is used to process the denoised external ballistic signal to obtain learning samples; The training samples are input into the LSTM model in time sequence to obtain the predicted external ballistic signal.

5. The method for predicting the external ballistic trajectory of artillery according to claim 1, characterized in that, The LSTM model in step S5 includes an input layer, multiple LSTM hidden layers, and a fully connected output layer. The input signal enters the multiple LSTM hidden layers through the input layer. The multiple LSTM hidden layers include a forget gate, an input gate, and an output gate to extract features from the input signal and obtain deep features. The fully connected output layer performs a fully connected operation on the deep features to obtain the predicted external ballistic signal.

6. The method for predicting the external ballistic trajectory of artillery according to claim 1, characterized in that, In step S5, the training process of the LSTM model includes: Generate simulated artillery external ballistic signals under complex environments; Repeat steps S2 to S4 to obtain the denoised simulated artillery external ballistic signal. The denoised simulated artillery external ballistic signal was processed using the sliding window method. The processed signals are used to train the constructed LSTM network model.

7. The method for predicting the external ballistic trajectory of artillery according to claim 6, characterized in that, The process of generating simulated artillery external ballistic signals under complex environments includes: Based on the differential equations of the external ballistics of a point mass, a noiseless ideal artillery external ballistic signal is generated by solving the fourth-order Runge-Kutta method. By introducing a Gaussian mixture model, non-stationary noise is introduced into the ideal artillery external ballistic signal to obtain the simulated artillery external ballistic signal.

8. The method for predicting the external ballistic trajectory of artillery according to claim 7, characterized in that, The Gaussian mixture model is a two-component Gaussian mixture model. The non-stationary noise includes background noise used to simulate conventional radar ranging errors, and anomalous impulse noise used to simulate strong electromagnetic interference or radar scintillation effects.

9. A computer device, characterized in that, include: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the artillery external ballistic prediction method according to any one of claims 1 to 8.

10. A non-transitory computer-readable storage medium storing computer instructions, characterized in that, The computer instructions are used to cause the computer to execute the artillery external ballistic prediction method according to any one of claims 1 to 8.

Images