A flight trajectory-based interception method

Abstract

The application provides a flight trajectory-based interception method, and belongs to the field of flight trajectory interception. The method comprises the following steps: acquiring historical flight trajectory data of an incoming aircraft, extracting data points of the historical flight trajectory data at a preset sampling interval to form a sequence data set for model training and prediction; constructing a neural network model based on an encoder-decoder structure as a trajectory prediction model, training the trajectory prediction model by using the sequence data set, and using the trained trajectory prediction model to predict a predicted position of the incoming aircraft at a certain specific time in the future according to real-time trajectory data of the incoming aircraft at the present time and in the past, and controlling an interceptor to fly to the predicted position; and when the interceptor reaches or approaches the predicted position, taking a real-time actual position of the incoming aircraft as an actual target point, and controlling the interceptor to intercept the actual target point. Thus, the response capability of intelligent threat avoidance in air defense interception training is effectively improved.

Description

Technical Field

[0001] This invention belongs to the field of flight trajectory interception, and specifically relates to an interception method based on flight trajectory. Background Technology

[0002] In a virtual training environment for air defense interception, in order to construct repeatable and quantifiable adversarial scenarios, it is usually necessary to pre-set the motion patterns of incoming aircraft.

[0003] Existing training systems mostly use preset fixed routes, such as straight lines, circular arcs, or pre-programmed waypoint sequences, or plan and predict target trajectories based on simple kinematic models, such as uniform straight lines, uniform acceleration, and coordinated turns, so that trainees can practice interception procedures according to predetermined plans.

[0004] However, in real combat, aircraft often employ complex evasive maneuvers such as rapid turns, dives, and serpentine maneuvers, exhibiting strong nonlinearity and time-varying characteristics. Existing training systems rely on fixed routes or simplified models, which cannot accurately simulate the strong nonlinearity and time-varying nature of complex evasive maneuvers. This results in significant deviations between their predictions and the actual target position, especially after the target performs sudden maneuvers, where the error amplifies rapidly. Summary of the Invention

[0005] To address the problem that existing technologies struggle to accurately predict the evasive maneuvers of highly maneuverable targets, resulting in poor interception effectiveness, this invention provides an interception method based on flight trajectory.

[0006] To achieve the above objectives, the present invention provides the following technical solution: An interception method based on flight trajectory, the method comprising: Acquire real-time trajectory data of incoming aircraft in a virtual training environment; In the interception simulation application of the virtual training environment, a pre-trained neural network model is used as a trajectory prediction model. Based on the real-time trajectory data of the incoming aircraft, its predicted position at a specific time in the future is predicted, the predicted position is set as the virtual target point, and the interceptor in the virtual training environment is controlled to fly towards the virtual target point. When the interceptor arrives at or approaches the virtual target point, it switches to interception mode, using the real-time location of the attacking aircraft in the virtual training environment as the actual target point, and controls the interceptor to fly towards the actual target point to complete the interception operation in the virtual training environment.

[0007] Optionally, the neural network model is an encoder-decoder structure, where the encoder uses a recurrent neural network or a long short-term memory network structure to compress the temporal features of the historical trajectory sequence; the decoder uses a recurrent neural network or a long short-term memory network structure to generate the future trajectory sequence step by step based on the feature vector output by the encoder.

[0008] Optionally, before predicting the predicted position of the incoming aircraft at a specific time in the future, the type of the incoming aircraft is identified, and the corresponding trajectory prediction model is selected based on the identified type.

[0009] Optionally, a navigation law for a fixed target is used to control the interceptor to fly toward the predicted position, and a navigation law for a maneuvering target is used to control the interceptor to intercept the actual target point; the navigation law for the fixed target is a proportional guidance method, and the navigation law for the maneuvering target is a proportional guidance method.

[0010] Optionally, the specific moment is dynamically determined based on the interceptor's flight speed, the current relative distance to the incoming aircraft, and the maneuverability of the incoming aircraft.

[0011] Optionally, when the interceptor arrives at or approaches the predicted position, using the real-time actual position of the attacking aircraft as the actual target point, controlling the interceptor to intercept the actual target point includes: When the distance between the interceptor and the predicted location is less than a preset threshold, or when the interceptor's flight time reaches the preset flight duration for intercepting the predicted location, the interceptor is controlled to intercept the actual target point.

[0012] Optionally, after acquiring the historical flight trajectory data of the incoming aircraft, the historical flight trajectory data is further cleaned and normalized to obtain preprocessed trajectory data; and the data is then used based on the preprocessed trajectory data. The data cleaning process includes detecting and correcting missing and outlier values ​​in historical flight trajectory data, and the normalization process includes mapping the position coordinates in the trajectory data to a uniform numerical range according to a preset range.

[0013] An interception device based on flight trajectory, the device comprising: The acquisition module is used to acquire real-time trajectory data of incoming aircraft in the virtual training environment; The first interception module is used in the interception simulation application of the virtual training environment to use a pre-trained neural network model as a trajectory prediction model, predict the predicted position of the incoming aircraft at a specific time in the future based on the real-time trajectory data of the incoming aircraft, set the predicted position as a virtual target point, and control the interceptor in the virtual training environment to fly towards the virtual target point. The second interception module is used to switch the interception mode when the interceptor arrives at or approaches the virtual target point, and control the interceptor to fly towards the actual target point based on the real-time location of the attacking aircraft in the virtual training environment, thereby completing the interception operation in the virtual training environment.

[0014] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the aforementioned interception method based on flight trajectory.

[0015] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the aforementioned interception method based on flight trajectory.

[0016] The interception method based on flight trajectory provided by this invention has the following beneficial effects: This invention targets highly maneuverable targets in virtual training environments and employs a two-stage navigation strategy of predicting virtual point interception and correcting for real targets. By constructing a trajectory prediction model, it autonomously learns the nonlinear motion laws of highly maneuverable targets from historical trajectories. During the mid-flight phase, it uses the predicted information to adjust the course in advance, achieving accurate prediction of the future trajectory. In the terminal phase, it switches to real-time target tracking, effectively overcoming the shortcomings of traditional fixed routes or simple motion models in dealing with complex evasive maneuvers, and significantly improving the prediction accuracy and interception success rate of highly maneuverable targets. Attached Figure Description

[0017] To more clearly illustrate the embodiments and design schemes of the present invention, the accompanying drawings required for this embodiment will be briefly described below. The drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart illustrating an interception method based on flight trajectory provided by the present invention according to an exemplary embodiment.

[0019] Figure 2 This is a block diagram of an interception device based on a flight trajectory according to an exemplary embodiment of the present invention. Detailed Implementation

[0020] To enable those skilled in the art to better understand and implement the technical solutions of the present invention, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention and should not be construed as limiting the scope of protection of the present invention.

[0021] The technical solutions provided by the various embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0022] First, this invention provides an interception method based on flight trajectory, specifically as follows: Figure 1 As shown, it includes the following steps: S101. Obtain real-time trajectory data of incoming aircraft in the virtual training environment.

[0023] Data points of the real-time flight trajectory data are extracted at preset sampling intervals to form a sequence dataset for model prediction.

[0024] After acquiring the historical flight trajectory data of the incoming aircraft, the historical flight trajectory data is cleaned and normalized to obtain preprocessed trajectory data. The data is then used based on the preprocessed trajectory data. The data cleaning includes detecting and correcting missing and outlier values ​​in the historical flight trajectory data, and the normalization includes mapping the position coordinates in the trajectory data to a uniform numerical range according to a preset range.

[0025] S102. In the interception simulation application of the virtual training environment, a pre-trained neural network model is used as a trajectory prediction model. Based on the real-time trajectory data of the incoming aircraft, its predicted position at a specific time in the future is predicted, the predicted position is set as a virtual target point, and the interceptor in the virtual training environment is controlled to fly towards the virtual target point.

[0026] In one embodiment, the neural network model is an encoder-decoder structure. The encoder uses a recurrent neural network or a long short-term memory network structure to compress the temporal features of the historical trajectory sequence. The decoder uses a recurrent neural network or a long short-term memory network structure to generate the future trajectory sequence step by step based on the feature vector output by the encoder, and maps the decoder output to the specific target position prediction value through a linear layer.

[0027] During the training phase, the trajectory prediction model uses the error between the predicted trajectory and the actual trajectory as the model optimization metric. This error is evaluated by calculating the average Euclidean distance between the endpoints of the predicted and actual trajectories. Since the trajectories of different aircraft vary significantly, separate training sessions are conducted using corresponding flight trajectory data for different aircraft types. Furthermore, before predicting the location of an incoming aircraft at a specific future moment, the type of the incoming aircraft is identified, and the corresponding trajectory prediction model is selected based on the identified type.

[0028] The specific time point is dynamically determined based on the interceptor's flight speed, its current relative distance to the incoming aircraft, and the incoming aircraft's maneuverability. In the virtual training environment, there is a certain spatial distance between the interceptor and the incoming aircraft, requiring the determination of a suitable future time point as the time reference for predicting the target point. If the incoming aircraft has weak maneuverability and its trajectory is relatively gentle, the prediction time can be set relatively long, allowing the interceptor ample flight time. If the incoming aircraft has strong maneuverability, the uncertainty of its position increases rapidly over time, and an excessively long prediction time will lead to a sharp amplification of the model's extrapolation error. In this case, the prediction time should be appropriately shortened to bring the virtual target point closer to the target's current position. Although this reduces the time available for the interceptor to adjust its course, it significantly improves the reliability of the predicted position. Furthermore, when the interceptor has a high speed and strong approach capability, the prediction time can be shortened accordingly to obtain higher accuracy prediction results; when the interceptor's speed is limited, the prediction time needs to be appropriately extended to ensure that the interceptor has sufficient time to reach the predicted point. Therefore, the prediction time can be dynamically adjusted to a reasonable range under different confrontation situations, achieving a balance between prediction accuracy and interception feasibility, thereby ensuring the effective implementation of the two-stage interception strategy.

[0029] S103. When the interceptor arrives at or approaches the virtual target point, switch the interception mode and use the real-time location of the attacking aircraft in the virtual training environment as the actual target point. Control the interceptor to fly towards the actual target point and complete the interception operation in the virtual training environment.

[0030] In this step, once the distance between the interceptor and the predicted location is less than a preset threshold, or the interceptor's flight time reaches the preset flight duration for intercepting the predicted location, the interceptor is controlled to intercept the actual target point.

[0031] In the above steps, the interceptor is controlled to fly toward the predicted position using a navigation law for a fixed target, and the interceptor is controlled to intercept the actual target point using a navigation law for a maneuvering target; the navigation law for the fixed target is a proportional guidance method, and the navigation law for the maneuvering target is a proportional guidance method.

[0032] Using the above method, a two-stage navigation strategy of predicting virtual point interception and real target correction is adopted for highly maneuverable targets in a virtual training environment. By constructing a trajectory prediction model, the nonlinear motion law of highly maneuverable targets is autonomously learned from historical trajectories. In the mid-flight stage, the course is adjusted in advance using the prediction information to achieve accurate prediction of the future trajectory. In the terminal stage, it switches to real-time target tracking, which effectively overcomes the shortcomings of traditional fixed routes or simple motion models in dealing with complex evasive maneuvers, and significantly improves the prediction accuracy and interception success rate of highly maneuverable targets.

[0033] Based on the above method, the present invention also provides an embodiment for illustration.

[0034] The purpose of this invention is to provide an interception method based on flight trajectory. This method utilizes neural networks to learn the historical maneuvering characteristics of a target, predicts its future position, and designs a two-stage navigation law to improve the success rate and training value of protecting highly maneuverable targets in a virtual simulation environment. The method includes the following steps: Data Input and Preprocessing: Input the historical flight trajectory data of the incoming aircraft and perform preprocessing operations such as data cleaning and normalization.

[0035] Trajectory sampling: An interval sampling strategy is adopted for the preprocessed trajectory data, for example, one point is taken every a1 data points to form a sequence dataset for model training and prediction, with a corresponding time step of a2 seconds.

[0036] Trajectory Prediction Model Construction and Training: A neural network model based on an encoder-decoder structure is built as the aircraft trajectory prediction model. The encoder is used to compress and understand the features of historical trajectory sequences, and the decoder uses these features to predict trajectory sequences within a future time period. At the end of the model, a linear layer maps the features output by the decoder to specific, observable target position predictions.

[0037] Model Evaluation: During the training phase, the error between the predicted trajectory and the true trajectory is used as the model optimization metric. One specific evaluation method is to calculate the average Euclidean distance between the predicted trajectory's endpoint position and the true endpoint position.

[0038] Two-stage predictive interception: In the interception simulation application, the following steps are performed: The first stage: Predictive interception. Using a trained model, the position of the incoming aircraft at a specific future moment is predicted based on its current and historical real-time trajectories, and this predicted position is treated as a virtual stationary target. The interceptor is then guided to this predicted point using a navigation law for fixed targets (such as proportional guidance).

[0039] The second stage: Real target interception. Once the interceptor arrives at or approaches the predicted virtual target location, it immediately switches modes. At this point, the real-time location of the attacking aircraft is used as the target, and navigation laws such as proportional guidance are employed again to provide final protection.

[0040] Intelligent prediction: This invention utilizes a neural network model to learn the nonlinear motion patterns of highly maneuverable targets from historical data. Compared with traditional kinematic models, it can predict future trajectories more accurately and adapt to complex maneuvers.

[0041] Protection strategy optimization: The proposed two-stage strategy of "predictive virtual point protection + real-time target correction" uses predictive information to correct the course in the middle of the interceptor's flight, and ensures accurate locking of the real target in the terminal stage, which improves the rationality and success rate of protecting against highly maneuverable targets.

[0042] High realism: This method provides a highly realistic adversary model and adversarial environment for air defense interception, enabling personnel to practice dealing with threat targets with intelligent evasion capabilities, greatly improving the quality and level of training.

[0043] Secondly, the present invention also provides an interception device based on flight trajectory, such as... Figure 2 As shown, it includes: The acquisition module 201 is used to acquire real-time trajectory data of incoming aircraft in the virtual training environment.

[0044] The first interception module 202 is used in the interception simulation application of the virtual training environment to use a pre-trained neural network model as a trajectory prediction model, predict the predicted position of the incoming aircraft at a specific time in the future based on the real-time trajectory data of the incoming aircraft, set the predicted position as a virtual target point, and control the interceptor in the virtual training environment to fly towards the virtual target point.

[0045] The second interception module 203 is used to switch the interception mode when the interceptor arrives at or approaches the virtual target point, and control the interceptor to fly towards the actual target point based on the real-time position of the attacking aircraft in the virtual training environment, so as to complete the interception operation in the virtual training environment.

[0046] The present invention also provides a computer-readable storage medium storing a computer program that can be used to execute the above-described... Figure 1 The steps of a flight trajectory-based interception method are provided.

[0047] This invention also provides a computer device. At the hardware level, the computer device includes a processor, an internal bus, a network interface, memory, and non-volatile memory, and may also include other hardware required for various operations. The processor reads the corresponding computer program from the non-volatile memory into memory and then executes it to achieve the above-mentioned functions. Figure 1 The steps of a flight trajectory-based interception method are provided.

[0048] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0049] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, as well as combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0050] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0051] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0052] It should be noted that the specific embodiments described above enable those skilled in the art to more fully understand the present invention, but do not limit the present invention in any way. Therefore, although the present invention has been described in detail in this specification, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the present invention; and all technical solutions and improvements that do not depart from the spirit and scope of the present invention are covered within the protection scope of the patent of the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A method for interception based on flight trajectory, characterized in that, The method includes: Acquire real-time trajectory data of incoming aircraft in a virtual training environment; In the interception simulation application of the virtual training environment, a pre-trained neural network model is used as a trajectory prediction model. Based on the real-time trajectory data of the incoming aircraft, its predicted position at a specific time in the future is predicted, the predicted position is set as the virtual target point, and the interceptor in the virtual training environment is controlled to fly towards the virtual target point. When the interceptor arrives at or approaches the virtual target point, it switches to interception mode, using the real-time location of the attacking aircraft in the virtual training environment as the actual target point, and controls the interceptor to fly towards the actual target point to complete the interception operation in the virtual training environment.

2. The method according to claim 1, characterized in that, The neural network model is an encoder-decoder structure. The encoder adopts a recurrent neural network or a long short-term memory network structure to compress the temporal features of historical trajectory sequences. The decoder employs a recurrent neural network or long short-term memory network structure to generate future trajectory sequences step-by-step based on the feature vectors output by the encoder.

3. The method according to claim 1, characterized in that, Before predicting the location of an incoming aircraft at a specific point in the future, the type of the incoming aircraft is identified, and the corresponding trajectory prediction model is selected based on the identified type.

4. The method according to claim 1, characterized in that, The interceptor is controlled to fly towards the predicted location using a navigation law for a fixed target, and to intercept the actual target point using a navigation law for a maneuvering target; the navigation law for the fixed target is a proportional guidance method, and the navigation law for the maneuvering target is a proportional guidance method.

5. The method according to claim 1, characterized in that, The specific moment is dynamically determined based on the interceptor's flight speed, its current relative distance to the incoming aircraft, and the incoming aircraft's maneuverability.

6. The method according to claim 1, characterized in that, Once the interceptor arrives at or approaches the predicted position, using the real-time actual position of the attacking aircraft as the actual target point, controlling the interceptor to intercept the actual target point includes: When the distance between the interceptor and the predicted location is less than a preset threshold, or when the interceptor's flight time reaches the preset flight duration for intercepting the predicted location, the interceptor is controlled to intercept the actual target point.

7. The method according to claim 1, characterized in that, After acquiring the historical flight trajectory data of the incoming aircraft, the historical flight trajectory data is further cleaned and normalized to obtain preprocessed trajectory data; and the data is then used based on the preprocessed trajectory data. The data cleaning process includes detecting and correcting missing and outlier values ​​in historical flight trajectory data, and the normalization process includes mapping the position coordinates in the trajectory data to a uniform numerical range according to a preset range.

8. An interception device based on flight trajectory, characterized in that, The device includes: The acquisition module is used to acquire real-time trajectory data of incoming aircraft in the virtual training environment; The first interception module is used in the interception simulation application of the virtual training environment to use a pre-trained neural network model as a trajectory prediction model, predict the predicted position of the incoming aircraft at a specific time in the future based on the real-time trajectory data of the incoming aircraft, set the predicted position as a virtual target point, and control the interceptor in the virtual training environment to fly towards the virtual target point. The second interception module is used to switch the interception mode when the interceptor arrives at or approaches the virtual target point, and control the interceptor to fly towards the actual target point based on the real-time location of the attacking aircraft in the virtual training environment, thereby completing the interception operation in the virtual training environment.

9. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the method described in any one of claims 1 to 7.

10. A computer device, characterized in that, The method includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method described in any one of claims 1 to 7.

Images