A method, device, electronic device, readable storage medium and program product for intelligent autonomous decision-making of unmanned aerial vehicles

Abstract

The embodiments of the present invention disclose a method, device, electronic device, readable storage medium, and program product for intelligent autonomous decision-making of unmanned aerial vehicles. The method includes: using a tactical behavior sample set for training to obtain a tactical behavior prediction model; obtaining the action parameter information of the enemy aircraft under a preset time sequence, and determining the characteristics of the enemy aircraft's next maneuver sequence based on the action parameter information and the tactical behavior prediction model; constructing a situation analysis model based on the spatial occupancy of our aircraft and the enemy aircraft; determining the comprehensive position advantage value and position comprehensive threat value of our aircraft relative to the enemy aircraft based on the situation analysis model, and determining the radar comprehensive advantage value and radar comprehensive threat value; making the optimal countermeasure decision for the enemy aircraft's next maneuver sequence characteristics based on the comprehensive position advantage value, position comprehensive threat value, radar comprehensive advantage value, and radar comprehensive threat value. The above technical solution can effectively improve the efficiency and accuracy of autonomous intelligent decision-making on the basis of improving the ability of autonomous intelligent decision-making.

Description

Technical Field

[0001] The present invention relates to the field of unmanned aerial vehicle (UAV) intelligent decision-making technology, and in particular to a method, device, electronic device, readable storage medium, and program product for unmanned aerial vehicle (UAV) intelligent autonomous decision-making. Background Art

[0002] Judging from current technological advancements, the demand for intelligent drones with integrated reconnaissance and strike capabilities is rapidly increasing. The combat environment is typically assessed through drone radar monitoring and user thinking. Consequently, during specific operations, manned and unmanned aircraft often collaborate to conduct and transmit combat actions, wasting valuable time. The ever-changing nature of the modern battlefield, with rapidly changing information, makes overall decision-making less than ideal. UAV systems will inevitably become a trend in complex future battlefields, possessing enhanced thinking and autonomous decision-making capabilities. Developing intelligent response strategies quickly is a critical issue, and a method for intelligent autonomous decision-making by drones is urgently needed to address these technical challenges. Summary of the Invention

[0003] In view of this, the present invention provides a method, device, electronic device, readable storage medium and program product for intelligent autonomous decision-making of unmanned aerial vehicles, which can enhance the ability to make autonomous intelligent decisions in complex battlefield environments and effectively improve the efficiency and accuracy of autonomous intelligent decision-making.

[0004] According to one aspect of the present invention, an embodiment of the present invention provides a method for intelligent autonomous decision-making of a drone, the method comprising:

[0005] Obtaining a pre-built tactical behavior sample set, and using the tactical behavior sample set to train a behavior prediction model to obtain a trained tactical behavior prediction model;

[0006] Obtaining action parameter information of the enemy aircraft in a preset time sequence, and determining the next maneuver sequence characteristics of the enemy aircraft based on the action parameter information and the tactical behavior prediction model;

[0007] Constructing a situation analysis model based on the spatial position of the friendly aircraft and the enemy aircraft; wherein the situation analysis model includes: an angle situation analysis model, a distance situation analysis model, a speed situation analysis model, and an altitude situation analysis model;

[0008] Determining a comprehensive positional advantage value and a comprehensive positional threat value of the friendly aircraft relative to the enemy aircraft based on the situation analysis model;

[0009] determining a radar integrated advantage value and a radar integrated threat value of the friendly aircraft's airborne radar relative to the enemy aircraft's airborne radar;

[0010] An optimal confrontation decision is made for the next maneuver sequence characteristics of the enemy aircraft according to the comprehensive occupation advantage value, the comprehensive occupation threat value, the comprehensive radar advantage value, and the comprehensive radar threat value.

[0011] According to another aspect of the present invention, an embodiment of the present invention further provides a drone intelligent autonomous decision-making device, the device comprising:

[0012] A data acquisition module is used to obtain a pre-built tactical behavior sample set and use the tactical behavior sample set to train the behavior prediction model to obtain a trained tactical behavior prediction model;

[0013] An enemy aircraft maneuver sequence prediction module is used to obtain action parameter information of the enemy aircraft in a preset time sequence, and determine the next maneuver sequence characteristics of the enemy aircraft based on the action parameter information and the tactical behavior prediction model;

[0014] A situation analysis establishment module is used to construct a situation analysis model based on the spatial position of the friendly aircraft and the enemy aircraft; wherein the situation analysis model includes: an angle situation analysis model, a distance situation analysis model, a speed situation analysis model, and an altitude situation analysis model;

[0015] a position comprehensive value determination module, configured to determine the position comprehensive advantage value and position comprehensive threat value of the friendly aircraft relative to the enemy aircraft based on the situation analysis model;

[0016] a radar comprehensive value determination module, configured to determine a radar comprehensive advantage value and a radar comprehensive threat value of the friendly aircraft's airborne radar relative to the enemy aircraft's airborne radar;

[0017] The confrontation decision determination module is used to make an optimal confrontation decision on the next maneuver sequence characteristics of the enemy aircraft according to the comprehensive occupation advantage value, the comprehensive occupation threat value, the comprehensive radar advantage value and the comprehensive radar threat value.

[0018] According to another aspect of the present invention, an embodiment of the present invention further provides an electronic device, comprising:

[0019] at least one processor;

[0020] and a memory communicatively connected to the at least one processor; wherein,

[0021] The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the drone intelligent autonomous decision-making method described in any embodiment of the present invention.

[0022] According to another aspect of the present invention, an embodiment of the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the drone intelligent autonomous decision-making method described in any embodiment of the present invention when executed.

[0023] The above-mentioned technical scheme of the embodiment of the present invention obtains a tactical behavior prediction model through training with a tactical behavior sample set, and determines the next maneuver sequence characteristics of the enemy aircraft based on the action parameter information of the enemy aircraft in a preset time series and the tactical behavior prediction model. On this basis, a situation analysis model is constructed based on the spatial occupancy of our aircraft and the enemy aircraft. Based on the situation analysis model, the comprehensive position advantage value and comprehensive position threat value of our aircraft relative to the enemy aircraft are determined, and the comprehensive radar advantage value and comprehensive radar threat value of our aircraft's airborne radar relative to the enemy aircraft's airborne radar are determined. Therefore, the optimal confrontation decision is made for the next maneuver sequence characteristics of the enemy aircraft based on the comprehensive position advantage value, comprehensive position threat value, comprehensive radar advantage value and comprehensive radar threat value, which can enhance the ability to make autonomous intelligent decisions in a complex battlefield environment and effectively improve the efficiency and accuracy of autonomous intelligent decision-making.

[0024] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present invention, nor is it intended to limit the scope of the present invention. Other features of the present invention will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0026] Figure 1 A flowchart of an intelligent autonomous decision-making method for a drone provided by one embodiment of the present invention;

[0027] Figure 2 A perspective diagram of a horizontal right turn action provided by one embodiment of the present invention;

[0028] Figure 3 A perspective diagram of a horizontal left turn provided by one embodiment of the present invention;

[0029] Figure 4 A perspective diagram of a large-angle dive turn provided by one embodiment of the present invention;

[0030] Figure 5 A perspective diagram of a large-angle pull-up and rotation action provided by one embodiment of the present invention;

[0031] Figure 6 A schematic diagram of a circling escape maneuver provided by one embodiment of the present invention;

[0032] Figure 7 A flowchart of constructing a situation analysis model in a method for intelligent autonomous decision-making of a UAV provided by one embodiment of the present invention;

[0033] Figure 8 A schematic diagram of an air combat position situation provided by one embodiment of the present invention;

[0034] Figure 9 A schematic diagram of a radar coordinate system of an airborne radar of a friendly aircraft and a body coordinate system of an enemy aircraft provided by one embodiment of the present invention;

[0035] Figure 10 A flowchart of another UAV intelligent autonomous decision-making method provided by one embodiment of the present invention;

[0036] Figure 11 This is a structural block diagram of an intelligent autonomous decision-making device for a drone provided by one embodiment of the present invention;

[0037] Figure 12 A schematic structural diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0038] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of the present invention.

[0039] It should be noted that the terms "first", "second", etc. in the description and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way can be interchanged where appropriate so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0040] In one embodiment, Figure 1 This is a flow chart of a method for intelligent autonomous decision-making of a drone provided by one embodiment of the present invention. This embodiment is applicable to situations where intelligent countermeasures are made against enemy aircraft in air combat. This method can be executed by a drone intelligent autonomous decision-making device, which can be implemented in the form of hardware and / or software and can be configured in an electronic device. Figure 1 As shown, the method includes:

[0041] S110 , obtaining a pre-built tactical behavior sample set, and using the tactical behavior sample set to train a behavior prediction model to obtain a trained tactical behavior prediction model.

[0042] The behavior prediction model can be a recurrent neural network model for processing sequence data. The recurrent neural network model can be a neural network architecture such as a GRU gated recurrent unit or an LSTM long short-term memory neural network. In this embodiment, the recurrent neural network model can be based on the MobileNet-V1 network architecture.

[0043] In this embodiment, after obtaining a tactical behavior sample set, the tactical behavior sample set is used as a training sample set to train a behavior prediction model, thereby obtaining a trained tactical behavior prediction model for predicting the characteristics of the enemy aircraft's next maneuver sequence. In some embodiments, data preprocessing is performed on each sample in the tactical behavior sample set, and the target sample set is input into the behavior prediction model to optimize model parameters until the model parameters are optimized or a predefined loss function is minimized, thereby obtaining a trained tactical behavior prediction model. In some embodiments, model parameters can be optimized using gradient descent, Bayesian optimization, genetic algorithms, and the like. These model parameters may include learning rates, regularization, and other factors. Of course, to reduce computational complexity, the model structure can also be modified accordingly. For example, depth-wise separable convolution is used to reduce computational complexity and parameter count. A depth-wise separable convolution structure composed of DWCONV+(1×1CONV) is used to reduce model parameters and forward reasoning.

[0044] In one embodiment, the construction of a tactical behavior sample set includes: obtaining motion trajectory information of a UAV at a historical moment, determining basic actions of the UAV based on the motion trajectory information; forming maneuvering actions from the basic actions, and forming tactical behaviors based on at least two continuous time series of maneuvering actions; wherein the tactical behavior represents the tactical intention; establishing a continuous maneuver sequence of the UAV with typical tactical behavior based on a preset aircraft dynamics model, a preset flight envelope, and tactical intention to form a tactical behavior library, and using the tactical behavior library as the tactical behavior sample set.

[0045] Among them, the maneuver sequence characteristics of typical tactical behaviors include at least: no tactical maneuver, dive turn, circling maneuver, small radius somersault, serpentine maneuver, bell maneuver, Herbst maneuver, Immelmann maneuver, cobra maneuver, vertical serpentine maneuver and jump turn; each maneuver sequence characteristic corresponds to a corresponding maneuver parameter value; the maneuver parameter value includes at least: altitude value range, speed value range, pitch angle value, yaw angle value, and roll angle value.

[0046] In this embodiment, the preset aircraft dynamics model is a nonlinear simulation model of the aircraft. This model can be a simulation model of a specific drone model selected by the user. It includes the aircraft's structure and parameters, including but not limited to mass, reference wingspan, pressure altitude, moment of inertia, and so on. This model can be written using Matlab and S-functions. The S-function uses C language and must be compiled before running. Simply enter "mex" followed by the file name in the command line.

[0047] The preset flight envelope in this embodiment refers to a closed geometric figure that represents the aircraft's flight range and operational constraints, using parameters such as flight speed, altitude, overload, and ambient temperature as coordinates. This flight envelope can include a level flight speed envelope, a speed overload envelope, and a gust overload envelope. In this embodiment, motion trajectory information refers to trajectory points, which are the original particle trajectories obtained by radar. Each radar detection point can be considered a particle, and its motion trajectory is represented by a time series. These trajectory points are motion trajectory points in three-dimensional space.

[0048] In this embodiment, the basic actions of the drone can be determined based on the motion trajectory information, thereby forming maneuvering actions from the basic actions, and tactical behaviors can be composed based on at least two continuous time series of maneuvering actions. The tactical behavior can represent the tactical intention, so as to establish a continuous maneuver sequence of the drone with typical tactical behavior based on the preset aircraft dynamics model, the preset flight envelope and the tactical intention to form a tactical behavior library, and the tactical behavior library is used as a tactical behavior sample set. In this embodiment, the basic maneuvers mainly include level flight, pull-up, dive, left turn, and right turn. Combined with the combination of these basic maneuvers, the maneuvers that need to be identified are 11 types: level flight, high-angle dive, high-angle pull-up, general dive, dive left turn, dive right turn, pull-up left turn, pull-up right turn, horizontal right turn, horizontal left turn, and general pull-up. Thus, a tactical behavior is composed of at least two continuous time series of maneuvers. Tactical behavior represents tactical intent. It can be understood that a fighter's tactical behavior can be decomposed into a maneuver sequence composed of several basic maneuver units, each of which reflects a tactical intent. Based on the aircraft dynamics model, flight envelope, and tactical intent, a continuous aircraft maneuver sequence for typical tactical behaviors is established to form a tactical behavior library. In this embodiment, for tactical behavior identification and intent understanding, the aircraft's basic behavior type can be determined by observing the fighter's flight maneuvers in space and continuous maneuvers within a certain time period. At the same time, enemy tactical behavior can be determined based on factors such as enemy distance and position, and the enemy's intention can be inferred.

[0049] It should be noted that after the basic action forms a maneuver, it is necessary to perform action perspective on the maneuver, which can be understood as visualizing the performance of the action in three-dimensional space, so as to help understand the spatial relationship and execution effect of the action, so as to more intuitively analyze and evaluate the rationality and effect of the action, and assist in decision-making and optimization. For example, in order to better understand the action diagram after the basic action forms a maneuver, Figure 2 A perspective diagram of a horizontal right turn action provided by one embodiment of the present invention; Figure 3 A perspective diagram of a horizontal left turn provided by one embodiment of the present invention; Figure 4 A perspective diagram of a large-angle dive turn provided by one embodiment of the present invention; Figure 5 This is a perspective diagram of a large-angle pull-up and turn action provided by one embodiment of the present invention. Similarly, after forming a tactical behavior based on at least two continuous time series of maneuvers, before establishing a continuous maneuver sequence of drones with typical tactical behaviors to form a tactical behavior library, it is also necessary to perspective the sample tactical actions, similarly presenting the performance of the action in three-dimensional space in a visual manner. In order to better understand the perspective of the sample tactical actions, Figure 6 A schematic diagram of a circling escape maneuver provided by one embodiment of the present invention.

[0050] In this embodiment, the maneuver sequence characteristics of typical tactical behaviors and the corresponding maneuver parameter values are stored in a maneuver action library as knowledge. A typical maneuver sequence characteristic table is constructed based on the tactical requirements of beyond-visual-range and medium-range air combat. The temporal process of the maneuver sequence is described using altitude, speed, and pitch, yaw, and roll angles as key characteristics. Each basic maneuver can be achieved by combining simpler atomic actions (such as left pull-up, altitude hold, speed hold, and heading hold) in a specific time sequence. This embodiment uses the definition of a maneuver sequence for a fighter aircraft in beyond-visual-range air combat as an example. The definition of this maneuver sequence for a fighter aircraft in beyond-visual-range air combat is shown in Table 1. In this embodiment, the typical maneuver sequence characteristics may include a typical maneuver name, a basic description, an altitude description, a speed description, and descriptions of pitch, yaw, and roll angles. For example, the typical maneuver name is a circling escape maneuver; the basic description is a rapid change of course to escape a tailgate collision; the altitude is 3000-5000 meters; the speed is 0.4-0.6 Ma; the pitch, yaw, and roll angles are a sharp 360° circle with a minimum radius; for another example, the typical maneuver name is a bell maneuver; the basic description is a rapid pull-up, with the pitch approaching 90 degrees, maintained for 4 seconds, followed by a pitching of the aircraft with its back facing downward, and finally a turn to exit the maneuver and resume level flight (a 360° circle); the altitude is below 3000 meters; the speed is 0.2-0.6 Ma; the pitch, yaw, and roll angles are 80°±10°, followed by a continuous increase in pitch angle, without yaw or roll, and a vertical circle.

[0051] Table 1: Definition of the maneuver sequence of fighters in beyond visual range air combat

[0052]

[0053] S120 , obtaining action parameter information of the enemy aircraft in a preset time sequence, and determining the next maneuver sequence characteristics of the enemy aircraft based on the action parameter information and the tactical behavior prediction model.

[0054] The action parameter information in the preset time sequence may be the action parameters corresponding to each time in a continuous time. The action parameter information includes at least altitude, speed, pitch angle, yaw angle, roll angle, longitude, and latitude. The preset time sequence is a continuous time sequence.

[0055] In this embodiment, the next maneuver sequence feature can be understood as the maneuver sequence feature that the enemy aircraft may perform at the next moment. The maneuver sequence feature is the form of a maneuver, and the maneuver corresponds to a corresponding maneuver category. The maneuver sequence feature may include, but is not limited to, various types such as a jump, a sharp turn, and a circling escape maneuver. The next maneuver sequence feature corresponds to a corresponding maneuver parameter value, which may include an altitude value, a speed value, a pitch angle value, a yaw angle value, and a roll angle value.

[0056] In this embodiment, the enemy aircraft's action parameter information over a preset time series can be used to determine the characteristics of the enemy aircraft's next maneuver sequence based on this action parameter information and a tactical behavior prediction model. In some embodiments, the action parameter information can be preprocessed and then directly input into the tactical behavior prediction model to predict the characteristics of the enemy aircraft's next maneuver sequence. In other embodiments, a reinforcement learning model can be used to simulate the enemy aircraft pilot's decision-making process and generate an optimal maneuver based on a reward function (such as "occupying an advantageous position" or "avoiding a missile"). In other embodiments, a game theory model can be used to predict the enemy aircraft's optimal response action, assuming that the enemy aircraft adopts a Nash equilibrium strategy. This embodiment is not limited to this.

[0057] S130. Construct a situation analysis model based on the spatial positions of the friendly aircraft and the enemy aircraft; wherein the situation analysis model includes: an angle situation analysis model, a distance situation analysis model, a speed situation analysis model, and an altitude situation analysis model.

[0058] Spatial positioning, also known as air combat positioning, can be understood as the positioning of both our aircraft and the enemy during a confrontation. Through maneuvers, we can control or predict key positions of both sides in three-dimensional space to gain tactical advantages (such as attack windows, defensive barriers, or energy advantages). This means that both sides must consider their position during a confrontation. For example, in an air combat, the higher the position, the safer it is.

[0059] In this embodiment, the enemy aircraft is detected by the onboard radar of the friendly aircraft to determine its spatial position relative to the friendly aircraft. Based on the spatial position, the attitude angle of the enemy aircraft relative to the incident beam of the onboard radar of the friendly aircraft at each moment is determined. The radar cross section (RCS) value of the enemy aircraft exposed to the onboard radar of the friendly aircraft at each moment is then determined based on the attitude angle. Based on this, the spatial situation is analyzed based on the RCS value of the enemy aircraft to construct a situation analysis model. The situation analysis model includes an angle situation analysis model, a distance situation analysis model, a speed situation analysis model, and an altitude situation analysis model. Specifically, the situation analysis model is expressed as follows: ; among them, among them, Expressed as an angle situation analysis model, Expressed as a speed situation analysis model, Expressed as a distance situation analysis model, Expressed as a high-level situation analysis model; 、 、 and They are respectively expressed as the weight coefficients corresponding to the angle situation analysis model, speed situation analysis model, distance situation analysis model and height situation analysis model.

[0060] In other embodiments, a drone air combat confrontation motion model can be constructed based on the position, speed, and posture of the drones of both sides in the air combat. Then, the drone air combat confrontation motion model can be combined with the air combat situation elements to establish an air combat situation assessment model based on the advantage function method; in other embodiments, the combat advantage index parameters of our drone relative to the enemy drone, the threat situation index parameters and target value index parameters of the enemy drone relative to our drone can be calculated first, and then the comprehensive situation function can be calculated based on the combat advantage index parameters, threat situation index parameters and target value index parameters. This embodiment does not limit this.

[0061] S140. Determine the comprehensive positional advantage value and comprehensive positional threat value of our aircraft relative to the enemy aircraft based on the situation analysis model.

[0062] Among them, the comprehensive position advantage value can be understood as the advantage value of our aircraft's spatial position relative to the enemy aircraft, which can represent the probability of our aircraft's advantage in spatial position. This advantage can include angle advantage, distance advantage, height advantage and speed advantage; the comprehensive position threat value can be understood as the threat value of our aircraft's spatial position relative to the enemy aircraft, which can represent the probability of the enemy aircraft's threat to our aircraft in spatial position. Similarly, this threat can include angle threat, distance threat, height threat and speed threat.

[0063] In this embodiment, a weight assignment method based on intercriteria correlation (CRITIC) can be used to first solve the variable weight coefficients corresponding to altitude, angle, distance, and speed in the situation analysis model. Based on these variable weight coefficients, the comprehensive position advantage value and comprehensive position threat value of the friendly aircraft relative to the enemy aircraft can be obtained. This can be understood as assigning corresponding weights to altitude, angle, distance, and speed, which can be adjusted according to tactical requirements. The comprehensive position advantage value and comprehensive position threat value of the friendly aircraft relative to the enemy aircraft are obtained through weighted summation.

[0064] S150. Determine the radar integrated advantage value and radar integrated threat value of the friendly aircraft's airborne radar relative to the enemy aircraft's airborne radar.

[0065] The radar comprehensive advantage can be understood as the advantage value of our aircraft's radar relative to the enemy aircraft's radar; the radar comprehensive threat value can be understood as the threat value of our aircraft relative to the enemy aircraft's radar. In this embodiment, the main factors affecting the radar comprehensive threat value and radar comprehensive advantage value in air-to-air combat are: the radar's maximum detection range (antenna gain, transmit power), the radar detection capability (signal-to-noise ratio), and the reflection cross-sectional area.

[0066] In this embodiment, the onboard radar of our aircraft is used to detect enemy aircraft to determine the spatial position of the enemy aircraft relative to our aircraft. The attitude angle of the enemy aircraft relative to the incident beam of the onboard radar of our aircraft at each moment is determined based on the spatial position. The RCS value of the enemy aircraft is determined based on the attitude angle. On this basis, the signal-to-noise ratio is determined based on the power received by the radar antenna and the noise power generated to characterize the radar monitoring capability. The maximum detection range of the onboard radar of our aircraft is determined based on the signal-to-noise ratio and the RCS value of the enemy aircraft. Simulation software is then used to determine empirical value results corresponding to the maximum detection range, signal-to-noise ratio, and RCS value, respectively. Each empirical value result is used as a corresponding weight coefficient. Finally, the weighted value is calculated to obtain the radar comprehensive advantage value and radar comprehensive threat value of the onboard radar of our aircraft relative to the onboard radar of the enemy aircraft.

[0067] S160. Make an optimal countermeasure decision based on the enemy aircraft's next maneuver sequence characteristics based on the comprehensive position advantage value, comprehensive position threat value, comprehensive radar advantage value, and comprehensive radar threat value.

[0068] Among them, the optimal confrontation decision can be understood as the tactical action of the confrontation decision finally executed by our machine.

[0069] In this embodiment, the optimal countermeasure decision is made based on the space occupation comprehensive advantage value and the space occupation comprehensive threat value, as well as the radar comprehensive advantage value and the radar comprehensive threat value. In some embodiments, the comprehensive advantage value is determined by the space occupation comprehensive advantage value and the radar comprehensive advantage value, and the comprehensive threat value is determined based on the space occupation comprehensive threat value and the radar comprehensive threat value. Then, corresponding thresholds are pre-set for the comprehensive threat value and the comprehensive advantage value, so that the final countermeasure action is performed based on the set thresholds. Of course, in addition to setting thresholds, the comprehensive threat value and the comprehensive advantage value can also be used to determine whether there is interference. If interference exists, the optimal countermeasure action can be set as a level flight action. In other embodiments, drone air combat confrontation can also be carried out by establishing a drone air combat confrontation strategy model, which is not limited in this embodiment.

[0070] The above-mentioned technical scheme of the embodiment of the present invention obtains a tactical behavior prediction model through training with a tactical behavior sample set, and determines the next maneuver sequence characteristics of the enemy aircraft based on the action parameter information of the enemy aircraft in a preset time series and the tactical behavior prediction model. On this basis, a situation analysis model is constructed based on the spatial occupancy of our aircraft and the enemy aircraft. Based on the situation analysis model, the comprehensive position advantage value and comprehensive position threat value of our aircraft relative to the enemy aircraft are determined, and the comprehensive radar advantage value and comprehensive radar threat value of our aircraft's airborne radar relative to the enemy aircraft's airborne radar are determined. Therefore, the optimal confrontation decision is made for the next maneuver sequence characteristics of the enemy aircraft based on the comprehensive position advantage value, comprehensive position threat value, comprehensive radar advantage value and comprehensive radar threat value, which can enhance the ability to make autonomous intelligent decisions in a complex battlefield environment and effectively improve the efficiency and accuracy of autonomous intelligent decision-making.

[0071] In one embodiment, Figure 7 A flowchart of constructing a situation analysis model in an intelligent autonomous decision-making method for a drone provided in one embodiment of the present invention. Based on the above embodiments, this embodiment further refines the method by using a tactical behavior sample set to train the behavior prediction model to obtain a trained tactical behavior prediction model, determining the next maneuver sequence characteristics of the enemy aircraft based on action parameter information and the tactical behavior prediction model, and constructing a situation analysis model based on the spatial occupancy of our aircraft and the enemy aircraft.

[0072] like Figure 7 As shown, the situation analysis model is constructed in the UAV intelligent autonomous decision-making method in this embodiment, which may specifically include the following steps:

[0073] S710. Obtain a pre-constructed tactical behavior sample set, and perform a zero-padding operation on the time series of each sample in the tactical behavior sample set to make the time series length of each sample consistent to obtain a processed sample; wherein the tactical behavior sample set includes maneuver sequence features of typical tactical behaviors and maneuver parameter values corresponding to each maneuver sequence feature.

[0074] In this embodiment, the flight time of each tactical action is not uniform. For example, 0.5s is defined as a time series. According to statistics, the total action time of most samples, 60% of the samples, is around 40s, while a very small number are around 20s and 120s. Since the samples input to the model need to be time series of fixed length, it is necessary to perform zero-padding operation on the time series of each sample in the tactical behavior sample set to make the time series length of each sample consistent to obtain the processed samples. It can be understood that the samples are padded with zeros to the maximum time series. For example, the maximum time series is 150s. For 300 sequences, for a 120s sample, it is equivalent to padded to a 150s action, in which all the data are filled with 0, so that the fixed input requirement of the model can be met.

[0075] S720: Perform data enhancement on the processed samples to obtain data-enhanced samples.

[0076] In this example, due to operational habits or environmental variations in real-world scenarios, each flight maneuver will experience some deviation in bank angle and angle of attack. Without this historical information during model training, the model's predictions will be inaccurate. This issue primarily requires improving the model's generalization capabilities, but excessive generalization can lead to insufficient model memory, requiring a balance. Therefore, data augmentation was used in actual training to add Gaussian white noise. By adding small perturbations, the model's training sample library was expanded, improving generalization.

[0077] S730: For the samples after data enhancement, insert a level flight action after each sample action is completed to initialize the next sample action, thereby obtaining a target sample set.

[0078] In this example, for data-augmented samples, the drone maneuvers to be recognized are continuous data, requiring the model to identify the start and end points of the maneuvers for continuous prediction. The model cannot access this flag or information, so a level flight is inserted after each tactical maneuver in the training samples. By identifying these continuous level flights, recognition of the next tactical maneuver is initialized.

[0079] S740: Input the target sample set into the behavior prediction model to optimize the model parameters until the model parameters are optimal or the predefined loss function is minimized, thereby obtaining a trained tactical behavior prediction model.

[0080] Among them, the behavior prediction model is a recurrent neural network model; the recurrent neural network model is based on the MobileNet-V1 network architecture; the convolutional layer of the recurrent neural network model consists of a deep separable convolution structure composed of DWCONV+ (1×1CONV), which reduces model parameters and forward reasoning.

[0081] In this embodiment, the activation function in the behavior prediction model is RELU, which can improve the forward reasoning calculation speed. In addition, the SDCA module, DWCONV+1*1CONV+, can greatly reduce the amount of calculation after improving the structure of the model. The attention module can ensure the model learning and memory capabilities of the low-computation structure, reduce the network width, and add Attention+Shortcut after the depth-wise separable convolution structure to avoid gradient dissipation. When using RELU, for negative input, the gradient of ReLU is 0. Due to the existence of the shortcut, there is no need to do more protection mechanisms, and the priority is to improve the calculation speed.

[0082] In this embodiment, when training the behavior prediction model, the target sample set is input into the improved behavior prediction model to optimize the model parameters until the model parameters are optimized or a predefined loss function is minimized, resulting in a trained tactical behavior prediction model. For example, after initializing the model parameters (random normal distribution), the first training phase begins with 1000 iterations of the current batch of samples to obtain a baseline effect model. This baseline model is loaded, and iterations are stopped when the threshold model parameters are optimal. The current neural network model is trained until the loss is less than 10, and iterations are stopped when the loss is less than 5. This model is then combined with the decision model. Based on simulation data results, model issues are identified, and additional samples with these issues are added to continue training the optimized model.

[0083] S750, obtaining the action parameter information of the enemy aircraft in a preset time sequence, performing data preprocessing on the action parameter information, and inputting the preprocessed action parameter information into the tactical behavior prediction model to predict the next maneuver sequence characteristics of the enemy aircraft.

[0084] In this embodiment, the manner of performing data preprocessing on the action parameter information may include but is not limited to performing a zero-padding operation, a sample enhancement operation, inserting a level flight action, and the like.

[0085] In this embodiment, data preprocessing is performed on the action parameter information of the enemy aircraft under a preset time sequence, and the preprocessed action parameter information is input into the tactical behavior prediction model to predict the next maneuver sequence characteristics of the enemy aircraft, so that our aircraft can make the next intelligent decision by judging the behavior, the spatial positioning of our aircraft and the enemy aircraft, the radar advantage, etc.

[0086] S760. Use your aircraft's airborne radar to detect enemy aircraft to determine their spatial position relative to your aircraft.

[0087] In this embodiment, the airborne radar of our aircraft will detect the position of the enemy aircraft in real time. The detected spatial position relationship may include: our aircraft is above the enemy aircraft, our aircraft is below the enemy aircraft, and our aircraft and the enemy aircraft are on the same horizontal line. In this embodiment, in order to better understand the air combat position situation between our aircraft and the enemy, Figure 8 This is a schematic diagram of an air combat position situation provided by one embodiment of the present invention. Figure 8 As shown, A is our fighter plane, T is the target plane in the airspace (that is, the enemy plane), and are the velocity vectors corresponding to our machine and the opponent's machine respectively, is the heading angle. It can be seen that when our aircraft and the enemy aircraft are in the same horizontal plane, the attitude angle of our aircraft exposed to the enemy aircraft's radar is equal to the azimuth angle. In air combat, the fighter first faces the threat of detection by the enemy aircraft's radar. How to choose the appropriate approach based on the situation between the two sides, ensure that the aircraft is exposed to the enemy radar threat in a good posture, and reduce the probability of enemy detection is the main means of achieving stealth. On the other hand, the static RCS characteristics of a fighter are a relatively complex variable, requiring measurement in a real environment or a microwave anechoic chamber, or simulation calculation. Under air combat confrontation conditions, the attitude angle of the enemy aircraft exposed to our aircraft's airborne radar changes in real time, and the RCS is highly dynamic.

[0088] S770. Determine the attitude angle of the enemy aircraft relative to the incident beam of the friendly aircraft's airborne radar at each moment based on the spatial position; wherein the attitude angle includes: azimuth and pitch angle.

[0089] In this embodiment, the attitude angle of the moving target relative to the airborne radar incident wave is solved according to the spatial position of the airborne radar and the target. Specifically, Figure 8 Take the example to establish the coordinate system. Figure 9 A schematic diagram of a radar coordinate system of an onboard radar of an own aircraft and a body coordinate system of an enemy aircraft is provided in one embodiment of the present invention, as shown in FIG. Figure 9 As shown, The coordinate origin of our aircraft's airborne radar; is the radar coordinate system of our aircraft (taken as the spherical rectangular coordinate system where the aircraft is located, with the x-axis pointing due east, the y-axis pointing due north, and the z-axis vertically upward), and r is the origin of the radar. The straight-line distance from the enemy aircraft. The position coordinates of the enemy aircraft in our aircraft's airborne radar coordinate system are marked as ( ). is the coordinate origin in the enemy aircraft's body coordinate system, The enemy aircraft's body coordinate system. is the projection of the enemy aircraft on the xy plane of our aircraft coordinate system, which can be used to calibrate the angle. During the movement of the enemy aircraft, it is exposed to the airborne radar beam in a certain posture, so it is necessary to solve the azimuth and pitch angle of our aircraft radar line of sight in the enemy aircraft body coordinate system. Given any point (x, y, z) in the airborne radar coordinate system of our aircraft, its coordinates in the enemy aircraft body coordinate system are recorded as , then the conversion relationship between the airborne radar coordinate system and the enemy aircraft body coordinate system is: , where The transformation matrix from the enemy aircraft body coordinate system to the airborne radar coordinate system. When the position coordinates of the detection carrier in the enemy aircraft body coordinate system are ,in, , then the position of the detection aircraft in the enemy aircraft body coordinate system and coordinate form is expressed as: , where 、 Azimuth, The above variables all change dynamically over time. By calculating the azimuth and pitch angles of the radar incident beam at each moment, the RCS value of the enemy aircraft can be obtained. This RCS sequence represents the RCS value exposed to the airborne radar at each moment when the target moves according to a certain trajectory and attitude.

[0090] S780. Determine the RCS value of the enemy aircraft based on the attitude angle; wherein the RCS value represents the RCS value of the enemy aircraft when it is exposed to the airborne radar of the friendly aircraft at each moment.

[0091] In this embodiment, the RCS value, also known as the radar signature, is a measure of the degree to which a radar can detect an object. A larger RCS value indicates a more detectable object. The RCS value of an enemy aircraft exposed to the onboard radar of a friendly aircraft at any given moment can be determined based on its attitude angle.

[0092] S790. Analyze the spatial situation based on the RCS value of the enemy aircraft to construct a situation analysis model.

[0093] In this embodiment, the RCS value of the enemy aircraft is used to analyze the spatial situation, thereby constructing a situation analysis model. Specifically, the spatial situation refers to determining the superiority and inferiority of the enemy aircraft based on the motion state information and RCS value between the friendly and enemy aircraft. The key to air combat is to keep the enemy aircraft within the friendly aircraft's attack zone while preventing the friendly aircraft from entering the enemy's attack zone. The four factors that significantly influence the attack zone range are angle, speed, distance, and altitude. Based on these factors, a spatial situation advantage function is established.

[0094] In this embodiment, the situation analysis model is expressed as follows: ;in, Expressed as an angle situation analysis model, Expressed as a speed situation analysis model, Expressed as a distance situation analysis model, Expressed as a high-level situation analysis model; 、 、 and They are respectively expressed as the weight coefficients corresponding to the angle situation analysis model, speed situation analysis model, distance situation analysis model and height situation analysis model.

[0095] Among them, the angle situation analysis model includes: azimuth advantage and entry angle advantage , respectively expressed as: Where, Indicates azimuth advantage; represents the entry angle advantage; 、 denote the weights of azimuth and entry angle respectively, ;in,

[0096] , Where, Indicates azimuth advantage; represents the entry angle advantage; represents the entry angle, represents the azimuth, represents the maximum search azimuth, represents the maximum off-axis emission angle, represents the inescapable cone angle, where ; The speed situation analysis model is expressed as: hour, ;when When , the speed situation analysis model is expressed as: Where, Indicates the speed of our aircraft; Indicates the speed of the enemy aircraft; It is expressed as the optimal air combat speed of our aircraft; the distance situation analysis model is expressed as: , where Indicates distance advantage; Indicates the distance between our aircraft and the enemy aircraft; Indicates the maximum detection distance of the radar. Indicates the maximum attack distance. Indicates the maximum distance from which one cannot escape. Indicates the minimum distance from which one cannot escape. represents the minimum attack distance, where ; The height situation analysis model is expressed as follows: Where, Indicates high advantage; Indicates the altitude of our aircraft; Indicates the enemy aircraft's altitude; It indicates the optimal positioning height for our aircraft.

[0097] In this embodiment, there is a coupling relationship between the azimuth angle and the entry angle in affecting the air combat situation. The main influencing factor of the angle situation function is the target azimuth angle. and target entry angle , , , The smaller, The larger it is, the larger the air-to-air missile attack zone will be and the better the angle situation will be.

[0098] In one embodiment, Figure 10A flowchart of another unmanned aerial vehicle intelligent autonomous decision-making method provided in one embodiment of the present invention. Based on the above embodiments, this embodiment further refines the method by using a tactical behavior sample set to train the behavior prediction model to obtain a trained tactical behavior prediction model, determining the next maneuver sequence characteristics of the enemy aircraft based on the action parameter information and the tactical behavior prediction model, and constructing a situation analysis model based on the spatial occupancy of our aircraft and the enemy aircraft.

[0099] like Figure 10 As shown, the situation analysis model is constructed in the UAV intelligent autonomous decision-making method in this embodiment, which may specifically include the following steps:

[0100] S1010. Perform reverse engineering based on the preset CRITIC method to obtain the first variable weight coefficient corresponding to each first indicator in the situation analysis model, and determine the sum of the products of each first variable weight coefficient and the corresponding first indicator to obtain the comprehensive position threat value of our aircraft relative to the enemy aircraft; wherein the first indicator includes: height, angle, distance and speed; wherein the first variable weight coefficient is the position threat coefficient.

[0101] The pre-set CRITIC method is an objective weighting method that determines the weight of each indicator by analyzing the contrast strength (standard deviation) and conflict (correlation) between indicators. This can be understood as determining weights by combining the impact of conflict between indicators and data variability on indicator weights. In this embodiment, the weights of the comprehensive advantage value and comprehensive threat value can be dynamically adjusted in air combat position assessment. Contrast strength: A larger standard deviation of an indicator indicates greater data dispersion and should be assigned a higher weight. Conflict: A lower correlation between indicators indicates less information overlap and should therefore be assigned a higher weight. Comprehensive weighting: This combines contrast strength and conflict to avoid subjective bias.

[0102] In this embodiment, the reverse transformation can be understood as the smaller the value of the first indicator, the better. For example, the forward transformation indicator 1: ; Forwarding can be understood as the larger the value of the first indicator, the better. For example, reverse index 1: .

[0103] In this embodiment, in the process of air combat between the two sides from far to near, the weight values of various threat and advantage factors change accordingly according to changes in the positions, maneuvers and other conditions of the two sides, so it is necessary to realize dynamic weight changes. In this embodiment, since the preset CRITIC method is used to solve the first variable weight coefficient corresponding to each first indicator in the situation analysis model, and the second weight coefficient are the same, it should be noted that although the preset CRITIC method is used for solving, when calculating the threat value, the inverse processing is adopted, that is, the smaller the indicator is hoped, the better; when calculating the advantage value, the forward processing is adopted, that is, the larger the indicator is hoped, the better.

[0104] In this embodiment, a preset CRITIC method is used to solve the first variable weight coefficient corresponding to the height, angle, distance, and speed in the situation analysis model. The specific steps may include: 1) constructing a situation assessment matrix; wherein the situation assessment matrix includes the index value of the jth index at the i-th moment, and j is 1, 2, 3, or 4, wherein 1 represents the angle, 2 represents the speed, 3 represents the distance, and 4 represents the height; 2) normalizing the situation assessment matrix to eliminate the influence of the index type and dimension, that is, when calculating the threat value, using the reverse processing, that is, the smaller the index is, the better; when calculating the advantage value, using the forward processing, that is, the larger the index is, the better. For example, the forward factor 1: ; Reversal factor 1: . 3) For the normalized situation assessment matrix, the conflict between indicators and the contrast intensity between data are represented by the correlation coefficient and standard deviation, and the comprehensive information content of the indicators is evaluated; 4) By evaluating the comprehensive information content of the indicators, the constant weight vector between the indicators can be determined, and the state variable weight vector can be constructed. The variable weight vector is obtained by Hadamard analysis of the state variable weight vector and the constant weight vector.

[0105] More specifically, 1) construct a situation assessment matrix, which consists of a real-time assessment state information set at t moments The situation assessment indicators and p constitute the indicator set The j-th index value at the i-th moment is , then the situation assessment matrix is expressed as: 2) Standardized evaluation matrix: Standardize the benefit-based, cost-based, and fixed indicators to eliminate the influence of indicator type and dimension. Benefit indicators: , cost-type indicators: , fixed indicators: , where: is the best value of the fixed indicator. After normalization, the situation matrix is It should be noted that when calculating the threat value, the reverse process is used, that is, the smaller the index is, the better. 3) Calculate the comprehensive information between the evaluation indicators: the conflict between the indicators and the comparison strength between the data are calculated by the correlation coefficient. and standard deviation To express it, the comprehensive information between the evaluation indicators is: , where: , is the average value of the j-th indicator, ,in, is the kth indicator, is the jth indicator. 4) Determine the constant weight of the evaluation indicators: The constant weight vector between the indicators can be determined by the comprehensive information of the evaluation indicators: , where: , the corresponding variable weight vector W(X) can be expressed as: , where The main method of changing weight is to construct a state-changing weight vector to achieve weighted adjustment of the state and avoid state imbalance. The state-changing weight vector can be expressed as a mapping relationship: ,in, , variable weight vector The weight vector can be changed by the state The Hadamard product with the constant weight vector w gives: The core of the variable weight solution lies in constructing an equilibrium function based on the influence of angle, speed, distance, and altitude on the air combat situation, and calculating the state variable weight vector by taking the derivative of the equilibrium function. The calculation formula is as follows: , where B(x) represents the equilibrium function of the situation assessment index x. As the engagement distance between the two sides approaches, the influence of angle, speed, and altitude on the situation becomes increasingly greater. From medium to close range, both sides are within the detection range of the other side, requiring fighters to maneuver and take position to form attack conditions. Using a positively correlated nonlinear function for incentive variable weighting, the equilibrium function is expressed as: , j=1, 2, 3, 4, where is a variable weight factor, j=1, 2, 3, 4, For incentive variable weight, is a punitive variable weight, and j in the formula represents various threat factors, with 1 being angle, 2 being speed, 3 being distance, and 4 being altitude. On the other hand, the impact of distance and combat capability on the situation decreases as distance decreases. The key factor affecting the outcome of an air battle lies in the priority of attack conditions. Using a nonlinear function for punitive variable weighting, the equilibrium function is expressed as: , the tactical intention situation is a dynamic process, which reflects the fighter's positioning process and potential trend. Based on this analysis, the comprehensive equilibrium function is expressed as: , the state variable weight vector of each indicator is expressed as: ,To sum up, the variable weight of each indicator is expressed as: .

[0106] In this embodiment, after obtaining the first variable weight coefficients corresponding to the distance, altitude, angle, and speed, respectively, the weighted sums of the distance, altitude, angle, and speed and the corresponding first variable weight coefficients are calculated to obtain the comprehensive threat value of the position of our aircraft relative to the enemy aircraft.

[0107] S1020. Perform forward solution based on the preset CRITIC method to obtain the second variable weight coefficient corresponding to each first indicator in the situation analysis model, and determine the sum of the products of each second variable weight coefficient and the corresponding first indicator to obtain the comprehensive position advantage value of our aircraft relative to the enemy aircraft; wherein the second variable weight coefficient is the position advantage coefficient.

[0108] In this embodiment, the method for solving the second variable weight coefficient corresponding to each first indicator in the situation analysis model using the preset CRITIC method is the same as the method for solving the first weight coefficient described above, so this embodiment will not be further described. In this embodiment, after obtaining the second variable weight coefficients corresponding to the distance, altitude, angle, and speed, the weighted sum of the distance, altitude, angle, and speed and the corresponding second variable weight coefficients is calculated to obtain the comprehensive position advantage value of the friendly aircraft relative to the enemy aircraft.

[0109] S1030. Determine a signal-to-noise ratio based on the power received by the radar antenna of the friendly aircraft's airborne radar and the noise power generated; wherein the signal-to-noise ratio represents the radar monitoring capability.

[0110] In this embodiment, the signal-to-noise ratio (SNR) can be determined based on the power received by the radar antenna and the noise power generated. This SNR represents the radar's detection capability. Specifically, the power of the incident wave intercepted by the target is determined, and then the power radiated per unit solid angle is determined when the target radiates uniformly throughout space. The power received by the radar antenna is then determined based on the received echo and the power radiated per unit solid angle. Ultimately, the SNR is determined based on the power received by the radar antenna and the noise power generated.

[0111] S1040. Obtain the RCS value of the enemy aircraft, and determine the maximum detection range of the airborne radar of our aircraft based on the signal-to-noise ratio and the RCS value of the enemy aircraft.

[0112] In this embodiment, the maximum detection range of the onboard radar of the friendly aircraft is determined based on the signal-to-noise ratio and the RCS value of the enemy aircraft. In this embodiment, the maximum detection range is calculated in the same manner as in the prior art and will not be described in detail in this embodiment.

[0113] S1050. Use the maximum detection range, signal-to-noise ratio, and RCS value as second indicators, and calculate the first empirical value results and second empirical value results corresponding to each second indicator according to the preset digital combat simulator DCS software, and use the first empirical value results as the corresponding first weight coefficients, and the second empirical value results as the corresponding second weight coefficients.

[0114] Among them, the first weight coefficient is the radar threat coefficient corresponding to each second indicator; the second weight coefficient is the radar advantage coefficient corresponding to each second indicator.

[0115] Among them, the preset digital combat simulator DCS software is a simulation platform in the existing technology that focuses on the simulation of fighter jets, helicopters, etc., and can obtain the first empirical value results and second empirical value results corresponding to the maximum detection distance, signal-to-noise ratio and RCS value through simulation. The first empirical value result and the second empirical value result are used as the corresponding first weight coefficient and second weight coefficient respectively to calculate the radar threat value and radar advantage value; it can be understood that the first empirical value result and the second empirical value result are the radar threat empirical value results and radar advantage empirical value results obtained in the simulation process of the battle between our aircraft and the enemy aircraft, and the radar threat empirical value result is used as the first weight coefficient, and the radar advantage empirical value result is used as the second weight coefficient.

[0116] In this embodiment, the empirical results for the radar's maximum detection range, signal-to-noise ratio (SNR), and RCS values in the DCS are affected by balance. In this embodiment, the friendly aircraft's airborne radar is used to detect enemy aircraft to determine the enemy aircraft's spatial position relative to the friendly aircraft. Based on this spatial position, the azimuth and elevation angles of the enemy aircraft's radar incident beam relative to the friendly aircraft at each moment are determined. The enemy aircraft's RCS value is then determined based on the azimuth and elevation angles. The maximum detection range, SNR, and RCS value are used as second indicators. The preset digital combat simulator DCS software is used to simulate and calculate the first and second empirical results corresponding to each second indicator. The first empirical results are used as the corresponding first weighting coefficients, and the second empirical results are used as the corresponding second weighting coefficients.

[0117] S1060. Determine the sum of the first products of the first weight coefficients and the corresponding second indicators to obtain the radar comprehensive advantage value of the airborne radar of the friendly aircraft relative to the airborne radar of the enemy aircraft, and determine the sum of the second products of the second weight coefficients and the corresponding second indicators to obtain the radar comprehensive threat value of the airborne radar of the friendly aircraft relative to the airborne radar of the enemy aircraft.

[0118] In this embodiment, after obtaining the weight coefficients for the maximum detection range, signal-to-noise ratio (SNR), and RCS value, the sum of the first products of the first weight coefficients and the corresponding second indicators is calculated to obtain the integrated radar advantage value of the friendly aircraft's airborne radar relative to the enemy aircraft's airborne radar, and the sum of the second products of the second weight coefficients and the corresponding second indicators is calculated to obtain the integrated radar threat value of the friendly aircraft's airborne radar relative to the enemy aircraft's airborne radar.

[0119] S1070: Determine a comprehensive advantage value based on the position comprehensive advantage value and the radar comprehensive advantage value, and determine a comprehensive threat value based on the position comprehensive threat value and the radar comprehensive threat value.

[0120] In this embodiment, the sum of the occupancy comprehensive advantage value and the radar comprehensive advantage value can be solved for the average value, and the average value can be used as the comprehensive advantage value. Similarly, the sum of the occupancy comprehensive threat value and the radar comprehensive threat value can be solved for the average value to obtain the comprehensive threat value.

[0121] S1080: When the comprehensive threat value is greater than or equal to the preset first threat threshold and the comprehensive advantage value is less than the preset first advantage threshold, the confrontation decision is a first type decision.

[0122] The preset first threat threshold and the preset first advantage threshold are both in the range of 0-1 and are thresholds set by the user based on experience or needs. The first type of decision is an escape decision. For example, the preset first threat threshold is 0.7, and the preset first advantage threshold is 0.6.

[0123] In this embodiment, when the comprehensive threat value is greater than or equal to a preset first threat threshold and the comprehensive advantage value is less than the preset first advantage threshold, the corresponding confrontation decision is determined to be an escape decision. This escape decision can be selected from a library of preset tactical behaviors, which may include dives, circling maneuvers, small-radius loops, serpentine maneuvers, bell-shaped maneuvers, and other tactical behaviors. Each behavior corresponds to a corresponding maneuver parameter value, which may include but is not limited to pitch, yaw, and roll angle values, altitude values, and speed values.

[0124] S1090: When the comprehensive threat value is greater than or equal to the preset second threat threshold and less than the preset first threat threshold, and the comprehensive advantage value is less than the preset second advantage threshold, the confrontation decision is a second type decision.

[0125] The preset first threat threshold is greater than the preset second threat threshold, and is also within the range of 0-1, similarly defined by the user based on experience or needs. The preset first advantage threshold is greater than the preset second advantage threshold. For example, the preset second threat threshold is 0.5, and the preset second advantage threshold is 0.5.

[0126] In this embodiment, when the comprehensive threat value is greater than or equal to the preset second threat threshold, less than the preset first threat threshold, and the comprehensive advantage value is less than the preset second advantage threshold, the confrontation decision is a second type of decision, which is a defensive decision and can also be selected from the preset tactical behavior library.

[0127] S10100: When the comprehensive threat value is less than the preset third threat threshold and the comprehensive advantage value is less than the preset first advantage threshold, the confrontation decision is a third type decision.

[0128] The second threat threshold is greater than the third threat threshold, and the third type of decision is an attack decision. For example, the third threat threshold is 0.3.

[0129] In this embodiment, when the comprehensive threat value is less than the preset third threat threshold and the comprehensive advantage value is less than the preset first advantage threshold, the confrontation decision is a third type decision, which can also be selected from the preset tactical behavior library.

[0130] S10110. Select the optimal adversarial decision from the first type of decision, the second type of decision, or the third type of decision.

[0131] In this embodiment, the optimal countermeasure decision can be selected from the first, second, or third types of decisions based on the maneuver parameters corresponding to the enemy aircraft's next maneuver sequence characteristics. This can be understood as follows: based on the angle, altitude, and speed values corresponding to the enemy aircraft's next maneuver sequence characteristics, one can select one of the broad categories of attack, defense, and escape. Then, based on the current situation, the broad category is determined and the optimal sub-category of maneuvers is selected, such as: attack: high-angle pull-up; defense: vertical serpentine; escape: circling to escape. It should be noted that in addition to the aforementioned broad categories of attack, defense, and escape, interference situations can also be included. In these cases, level flight is selected for countermeasures. Specifically, the selection criteria can be based on the defined tactical maneuver altitude, speed, and spatial position achieved per unit time. For example, the bell-shaped maneuver and the jump-and-turn both occur at altitudes below 3,000 meters. Initially, the altitude is increased to gain spatial advantage. However, the bell-shaped maneuver is defensive, while the jump-and-turn is escapable, representing different countermeasure decisions.

[0132] The above-mentioned technical scheme of the embodiment of the present invention obtains a tactical behavior prediction model through training with a tactical behavior sample set, and determines the next maneuver sequence characteristics of the enemy aircraft based on the action parameter information of the enemy aircraft in a preset time series and the tactical behavior prediction model. On this basis, a situation analysis model is constructed based on the spatial occupancy of our aircraft and the enemy aircraft. Based on the situation analysis model, the comprehensive position advantage value and comprehensive position threat value of our aircraft relative to the enemy aircraft are determined, and the comprehensive radar advantage value and comprehensive radar threat value of our aircraft's airborne radar relative to the enemy aircraft's airborne radar are determined. Therefore, the optimal confrontation decision is made for the next maneuver sequence characteristics of the enemy aircraft based on the comprehensive position advantage value, comprehensive position threat value, comprehensive radar advantage value and comprehensive radar threat value, which can enhance the ability to make autonomous intelligent decisions in a complex battlefield environment and effectively improve the efficiency and accuracy of autonomous intelligent decision-making.

[0133] In one embodiment, Figure 11This is a block diagram of a UAV intelligent autonomous decision-making device provided by one embodiment of the present invention. The device is suitable for training a model for DC arc detection. The device can be implemented by hardware / software. It can be configured in an electronic device to implement a UAV intelligent autonomous decision-making method in an embodiment of the present invention. Figure 11 As shown, the device includes: a data acquisition module 1110, an enemy aircraft maneuver sequence prediction module 1120, a situation analysis establishment module 1130, an occupancy comprehensive value determination module 1140, a radar comprehensive value determination module 1150 and a confrontation decision determination module 1160.

[0134] The data acquisition module 1110 is used to acquire a pre-built tactical behavior sample set and use the tactical behavior sample set to train the behavior prediction model to obtain a trained tactical behavior prediction model;

[0135] The enemy aircraft maneuver sequence prediction module 1120 is configured to obtain the action parameter information of the enemy aircraft in a preset time sequence, and determine the next maneuver sequence characteristics of the enemy aircraft based on the action parameter information and the tactical behavior prediction model;

[0136] A situation analysis building module 1130 is configured to build a situation analysis model based on the spatial position of the friendly aircraft and the enemy aircraft; wherein the situation analysis model includes an angle situation analysis model, a distance situation analysis model, a speed situation analysis model, and an altitude situation analysis model;

[0137] A comprehensive position value determination module 1140 is configured to determine a comprehensive position advantage value and a comprehensive position threat value of the friendly aircraft relative to the enemy aircraft based on the situation analysis model;

[0138] A radar comprehensive value determination module 1150 is configured to determine a radar comprehensive advantage value and a radar comprehensive threat value of the friendly aircraft's airborne radar relative to the enemy aircraft's airborne radar;

[0139] The confrontation decision determination module 1160 is used to make an optimal confrontation decision on the next maneuver sequence characteristics of the enemy aircraft based on the occupation comprehensive advantage value, the occupation comprehensive threat value, the radar comprehensive advantage value and the radar comprehensive threat value.

[0140] In an embodiment of the present invention, an enemy aircraft maneuver sequence prediction module obtains a tactical behavior prediction model through training with a tactical behavior sample set, and determines the next maneuver sequence characteristics of the enemy aircraft based on the action parameter information of the enemy aircraft in a preset time sequence and the tactical behavior prediction model. On this basis, a situation analysis establishment module constructs a situation analysis model based on the spatial occupancy of our aircraft and the enemy aircraft. The occupancy comprehensive value determination module determines the occupancy comprehensive advantage value and occupancy comprehensive threat value of our aircraft relative to the enemy aircraft based on the situation analysis model, and determines the radar comprehensive advantage value and radar comprehensive threat value of our aircraft's airborne radar relative to the enemy aircraft's airborne radar through the radar comprehensive value determination module. Thus, the confrontation decision determination module makes the optimal confrontation decision on the next maneuver sequence characteristics of the enemy aircraft based on the occupancy comprehensive advantage value, the occupancy comprehensive threat value, the radar comprehensive advantage value and the radar comprehensive threat value, which can enhance the ability to make autonomous intelligent decisions in a complex battlefield environment and effectively improve the efficiency and accuracy of autonomous intelligent decision-making.

[0141] In one embodiment, the construction of the tactical behavior sample set includes:

[0142] Obtaining the motion trajectory information of the UAV at a historical moment, and determining the basic action of the UAV based on the motion trajectory information;

[0143] Forming a maneuver action from the basic action, and forming a tactical behavior based on at least two consecutive time series of maneuvers; wherein the tactical behavior represents a tactical intention;

[0144] Establishing a continuous maneuver sequence of a UAV with typical tactical behaviors based on a preset aircraft dynamics model, a preset flight envelope, and the tactical intent to form a tactical behavior library, and using the tactical behavior library as the tactical behavior sample set;

[0145] Among them, the maneuver sequence characteristics of the typical tactical behavior include at least: no tactical maneuver, dive turn, circling escape maneuver, small radius somersault, serpentine maneuver, bell maneuver, Herbst maneuver, Immelmann maneuver, cobra maneuver, vertical serpentine maneuver and jump turn; each maneuver sequence characteristic corresponds to a corresponding maneuver parameter value; the maneuver parameter value includes at least: altitude value range, speed value range, pitch angle value, yaw angle value, and roll angle value.

[0146] In one embodiment, the behavior prediction model is a recurrent neural network model; the recurrent neural network model is based on the MobileNet-V1 network architecture; the convolution layer of the recurrent neural network model is a depthwise separable convolution structure composed of DWCONV+(1×1CONV); the activation function of the recurrent neural network model is RELU;

[0147] Accordingly, the data acquisition module 1110 includes:

[0148] performing a zero-padding operation on a sample processing unit for each sample in the tactical behavior sample set to make the time series lengths of the samples consistent to obtain processed samples; wherein the tactical behavior sample set includes maneuver sequence features of typical tactical behaviors and maneuver parameter values corresponding to each maneuver sequence feature;

[0149] A sample enhancement unit, configured to perform data enhancement on the processed sample to obtain a data-enhanced sample;

[0150] a target sample set determination unit, configured to insert a level flight action after each sample action is completed for the data-enhanced samples to initialize the next sample action, thereby obtaining a target sample set;

[0151] The model training unit is used to input the target sample set into the behavior prediction model to optimize the model parameters until the model parameters reach the optimal value or the predefined loss function reaches the minimum, thereby obtaining a trained tactical behavior prediction model.

[0152] In one embodiment, the maneuver parameter information includes at least altitude, speed, pitch angle, yaw angle, roll angle, longitude, and latitude. Accordingly, the enemy aircraft maneuver sequence prediction module 1120 includes:

[0153] The prediction unit is used to perform data preprocessing on the action parameter information and input the preprocessed action parameter information into the tactical behavior prediction model to predict the next maneuver sequence characteristics of the enemy aircraft.

[0154] In one embodiment, the situation analysis establishment module 1130 includes:

[0155] a position determination unit, configured to detect an enemy aircraft using an onboard radar of the friendly aircraft to determine a spatial position of the enemy aircraft relative to the friendly aircraft;

[0156] an attitude angle determination unit, configured to determine, at each moment, the attitude angle of the enemy aircraft relative to the incident beam of the airborne radar of the friendly aircraft based on the spatial position; wherein the attitude angle includes: an azimuth angle and a pitch angle;

[0157] an RCS value determining unit, configured to determine an RCS value of the enemy aircraft based on the attitude angle; wherein the RCS value represents the RCS value of the enemy aircraft when exposed to the airborne radar of the friendly aircraft at each moment;

[0158] The situation model building unit is used to analyze the spatial situation according to the RCS value of the enemy aircraft to build a situation analysis model.

[0159] In one embodiment, the situation analysis model is expressed as follows: ;in, Expressed as an angle situation analysis model, Expressed as a speed situation analysis model, Expressed as a distance situation analysis model, Expressed as a high-level situation analysis model; 、 、 and They are respectively represented as the weight coefficients corresponding to the angle situation analysis model, speed situation analysis model, distance situation analysis model and height situation analysis model;

[0160] Wherein, the angle situation analysis model includes: azimuth advantage and entry angle advantage , respectively expressed as: Where, Indicates azimuth advantage; represents the entry angle advantage; 、 denote the weights of azimuth and entry angle respectively, ;in,

[0161] , Where, Indicates azimuth advantage; represents the entry angle advantage; represents the entry angle, represents the azimuth, represents the maximum search azimuth, represents the maximum off-axis emission angle, represents the inescapable cone angle, where ;

[0162] The speed situation analysis model is expressed as: hour, ;when When , the speed situation analysis model is expressed as: Where, Indicates the speed of our aircraft; Indicates the speed of the enemy aircraft; It is expressed as the best air combat speed of our aircraft;

[0163] The distance situation analysis model is expressed as: , where Indicates distance advantage; Indicates the distance between our aircraft and the enemy aircraft; Indicates the maximum detection distance of the radar. Indicates the maximum attack distance. Indicates the maximum distance from which one cannot escape. Indicates the minimum distance from which one cannot escape. represents the minimum attack distance, where ;

[0164] The height situation analysis model is expressed as follows: Where, Indicates high advantage; Indicates the altitude of our aircraft; Indicates the enemy aircraft's altitude; It indicates the optimal positioning height for our aircraft.

[0165] In one embodiment, the placeholder comprehensive value determination module 1140 includes:

[0166] a position threat value determination unit, configured to perform reverse engineering based on a preset CRITIC method to solve first variable weight coefficients corresponding to respective first indicators in the situation analysis model, and determine the sum of the products of the first variable weight coefficients and the corresponding first indicators to obtain a comprehensive position threat value of the friendly aircraft relative to the enemy aircraft; wherein the first indicators include: altitude, angle, distance, and speed; and wherein the first variable weight coefficient is a position threat coefficient;

[0167] A position advantage value determination unit is used to perform a forward solution based on the preset CRITIC method to obtain the second variable weight coefficient corresponding to each first indicator in the situation analysis model, and determine the sum of the products of each second variable weight coefficient and the corresponding first indicator to obtain the comprehensive position advantage value of the friendly aircraft relative to the enemy aircraft; wherein the second variable weight coefficient is the position advantage coefficient.

[0168] In one embodiment, the radar integrated value determination module 1150 includes:

[0169] a signal-to-noise ratio determining unit, configured to determine a signal-to-noise ratio based on the power received by the radar antenna of the airborne radar of the friendly aircraft and the noise power generated; wherein the signal-to-noise ratio represents the radar monitoring capability;

[0170] a maximum detection range determination unit, configured to obtain an RCS value of an enemy aircraft and determine a maximum detection range of the airborne radar of the friendly aircraft based on the signal-to-noise ratio and the RCS value of the enemy aircraft;

[0171] a weight coefficient determination unit, configured to use the maximum detection range, the signal-to-noise ratio, and the RCS value as second indicators, and to simulate and calculate, based on preset digital combat simulator (DCS) software, first and second empirical value results corresponding to each of the second indicators, and use the first and second empirical value results as corresponding first and second weight coefficients, respectively; wherein the first weight coefficient is a radar threat coefficient corresponding to each of the second indicators; and the second weight coefficient is a radar advantage coefficient corresponding to each of the second indicators;

[0172] The radar comprehensive value determination unit is used to determine the sum of the first products of each first weight coefficient and the corresponding second indicators to obtain the radar comprehensive advantage value of the airborne radar of the friendly aircraft relative to the airborne radar of the enemy aircraft, and to determine the sum of the second products of each second weight coefficient and the corresponding second indicators to obtain the radar comprehensive threat value of the airborne radar of the friendly aircraft relative to the airborne radar of the enemy aircraft.

[0173] In one embodiment, the confrontation decision determination module 1160 includes:

[0174] a comprehensive value determination unit, configured to determine a comprehensive advantage value based on the occupancy comprehensive advantage value and the radar comprehensive advantage value, and to determine a comprehensive threat value based on the occupancy comprehensive threat value and the radar comprehensive threat value;

[0175] a first decision-making unit, configured to, when the comprehensive threat value is greater than or equal to a preset first threat threshold and the comprehensive advantage value is less than the preset first advantage threshold, determine that the confrontation decision is a first type decision;

[0176] a second decision-making unit, configured to, when the comprehensive threat value is greater than or equal to a preset second threat threshold and less than the preset first threat threshold, and the comprehensive advantage value is less than the preset second advantage threshold, determine that the confrontation decision is a second type decision;

[0177] a third decision-making unit, configured to, when the comprehensive threat value is less than a preset third threat threshold and the comprehensive advantage value is less than a preset first advantage threshold, determine that the confrontation decision is a third type decision;

[0178] an optimal decision selection unit, configured to select an optimal adversarial decision from the first type of decision, the second type of decision, or the third type of decision;

[0179] Wherein, the preset first threat threshold is greater than the preset second threat threshold; the preset second threat threshold is greater than the preset third threat threshold; the preset first advantage threshold is greater than the preset second advantage threshold;

[0180] Among them, the first type of decision is an escape decision; the second type of decision is a defense decision; and the third type of decision is an attack decision.

[0181] The drone intelligent autonomous decision-making device provided in the embodiment of the present invention can execute the drone intelligent autonomous decision-making method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0182] In one embodiment, Figure 12 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. The electronic device 10 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. The electronic device may also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices (such as helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are provided for example only and are not intended to limit the implementation of the present invention described and / or claimed herein.

[0183] like Figure 12 As shown, electronic device 10 includes at least one processor 11 and memory, such as ROM 12 and RAM 13, communicatively connected to at least one processor 11. The memory stores computer programs executable by the at least one processor, and processor 11 can perform various appropriate actions and processes based on the computer programs stored in ROM 12 or loaded from storage unit 18 into RAM 13. RAM 13 can also store various programs and data required for the operation of electronic device 10. Processor 11, ROM 12, and RAM 13 are interconnected via bus 14. An I / O interface 15 is also connected to bus 14.

[0184] Multiple components in the electronic device 10 are connected to the I / O interface 15, including an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a magnetic disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0185] Processor 11 can be any general-purpose and / or specialized processing component with processing and computing capabilities. Examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various specialized artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSPs), and any other suitable processor, controller, microcontroller, etc. Processor 11 executes the various methods and processes described above, such as the drone intelligent autonomous decision-making method.

[0186] In some embodiments, the drone intelligent autonomous decision-making method can be implemented as a computer program tangibly embodied in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the drone intelligent autonomous decision-making method described above can be performed. Alternatively, in other embodiments, processor 11 can be configured to execute the drone intelligent autonomous decision-making method via any other suitable means (e.g., via firmware).

[0187] Various embodiments of the systems and techniques described above can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), system-on-chip systems (SOCs), programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include being implemented in one or more computer programs that are executable and / or interpreted on a programmable system that includes at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0188] Computer programs for implementing the methods of the present invention can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable unmanned aerial vehicle intelligent autonomous decision-making device, so that when executed by the processor, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The computer programs can be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0189] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, device, or apparatus. A computer-readable storage medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or apparatus, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media may include an electrical connection based on one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0190] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device that has: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0191] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0192] A computing system may include clients and servers. The clients and servers are typically remote from each other and typically interact via a communication network. This client-server relationship arises through computer programs running on the respective computers, creating a client-server relationship. The server may be a cloud server, also known as a cloud computing server or cloud host. This server is a hosting product within the cloud computing service ecosystem that addresses the management difficulties and limited scalability of traditional physical hosting and VPS services.

[0193] It should be understood that the various forms of the processes shown above can be used to reorder, add, or delete steps. For example, the steps described in the present invention can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solution of the present invention can be achieved. This is not limited herein.

[0194] The above specific embodiments do not limit the scope of protection of the present invention. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention are intended to be included within the scope of protection of the present invention.

Claims

1. A method for intelligent autonomous decision-making of a UAV, characterized in that: The method comprises: Obtaining a pre-built tactical behavior sample set, and using the tactical behavior sample set to train a behavior prediction model to obtain a trained tactical behavior prediction model; Obtaining action parameter information of the enemy aircraft in a preset time sequence, and determining the next maneuver sequence characteristics of the enemy aircraft based on the action parameter information and the tactical behavior prediction model; Constructing a situation analysis model based on the spatial position of the friendly aircraft and the enemy aircraft; wherein the situation analysis model includes: an angle situation analysis model, a distance situation analysis model, a speed situation analysis model, and an altitude situation analysis model; Determining a comprehensive positional advantage value and a comprehensive positional threat value of the friendly aircraft relative to the enemy aircraft based on the situation analysis model; determining a radar integrated advantage value and a radar integrated threat value of the friendly aircraft's airborne radar relative to the enemy aircraft's airborne radar; making an optimal countermeasure decision for the next maneuver sequence characteristics of the enemy aircraft based on the comprehensive position advantage value, the comprehensive position threat value, the comprehensive radar advantage value, and the comprehensive radar threat value; The behavior prediction model is a recurrent neural network model; the recurrent neural network model is based on the MobileNet-V1 network architecture; the convolutional layer of the recurrent neural network model is a depth-separable convolution structure composed of DWCONV+(1×1CONV); the activation function of the recurrent neural network model is RELU; Accordingly, the use of the tactical behavior sample set to train the behavior prediction model to obtain a trained tactical behavior prediction model includes: Performing a zero-padding operation on the time series of each sample in the tactical behavior sample set to make the time series lengths of each sample consistent to obtain processed samples; wherein the tactical behavior sample set includes maneuver sequence features of typical tactical behaviors and maneuver parameter values corresponding to each maneuver sequence feature; Performing data enhancement on the processed samples to obtain data-enhanced samples; For the samples after data augmentation, a level flight action is inserted after each sample action to initialize the next sample action, thereby obtaining a target sample set; Inputting the target sample set into the behavior prediction model to optimize model parameters until the model parameters are optimal or a predefined loss function is minimized, thereby obtaining a trained tactical behavior prediction model; The determining of the comprehensive positional advantage value and comprehensive positional threat value of the friendly aircraft relative to the enemy aircraft based on the situation analysis model includes: Performing reverse engineering based on a preset CRITIC method to solve first variable weight coefficients corresponding to respective first indicators in the situation analysis model, and determining the sum of the products of the first variable weight coefficients and the corresponding first indicators to obtain a comprehensive positional threat value of the friendly aircraft relative to the enemy aircraft; wherein the first indicators include: altitude, angle, distance, and speed; wherein the first variable weight coefficient is a positional threat coefficient; Performing a forward solution based on the preset CRITIC method to obtain the second variable weight coefficient corresponding to each first indicator in the situation analysis model, and determining the sum of the products of each second variable weight coefficient and the corresponding first indicator to obtain a comprehensive position advantage value of the friendly aircraft relative to the enemy aircraft; wherein the second variable weight coefficient is a position advantage coefficient; The determining of the radar integrated advantage value and radar integrated threat value of the friendly aircraft's airborne radar relative to the enemy aircraft's airborne radar includes: determining a signal-to-noise ratio based on the power received by the radar antenna of the airborne radar of the friendly aircraft and the noise power generated; wherein the signal-to-noise ratio represents the radar monitoring capability; Obtaining an RCS value of the enemy aircraft, and determining a maximum detection range of the airborne radar of the friendly aircraft based on the signal-to-noise ratio and the RCS value of the enemy aircraft; The maximum detection range, the signal-to-noise ratio, and the RCS value are used as second indicators, and first and second empirical value results corresponding to each of the second indicators are calculated using a preset digital combat simulator (DCS) software. The first and second empirical value results are used as corresponding first and second weight coefficients, respectively. The first and second empirical value results are used as corresponding second weight coefficients, respectively. The first and second weight coefficients are the radar threat coefficients corresponding to each of the second indicators, and the second and second weight coefficients are the radar advantage coefficients corresponding to each of the second indicators. Determine the sum of the first products of each of the first weight coefficients and the corresponding second indicators to obtain a comprehensive radar advantage value of the airborne radar of the friendly aircraft relative to the airborne radar of the enemy aircraft, and determine the sum of the second products of each of the second weight coefficients and the corresponding second indicators to obtain a comprehensive radar threat value of the airborne radar of the friendly aircraft relative to the airborne radar of the enemy aircraft.

2. The method according to claim 1, characterized in that The construction of the tactical behavior sample set includes: Obtaining the motion trajectory information of the UAV at a historical moment, and determining the basic action of the UAV based on the motion trajectory information; Forming a maneuver action from the basic action, and forming a tactical behavior based on at least two consecutive time series of maneuvers; wherein the tactical behavior represents a tactical intention; Establishing a continuous maneuver sequence of a UAV with typical tactical behaviors based on a preset aircraft dynamics model, a preset flight envelope, and the tactical intent to form a tactical behavior library, and using the tactical behavior library as the tactical behavior sample set; The maneuver sequence characteristics of the typical tactical behavior include at least: no tactical maneuver, dive turn, circling maneuver, small radius somersault, serpentine maneuver, bell maneuver, Herbst maneuver, Immelmann maneuver and jump turn; each maneuver sequence characteristic corresponds to a corresponding maneuver parameter value; the maneuver parameter value includes at least: altitude value range, speed value range, pitch angle value, yaw angle value, and roll angle value.

3. The method according to claim 1, characterized in that The motion parameter information includes at least altitude, speed, pitch angle, yaw angle, roll angle, longitude and latitude; Accordingly, determining the next maneuver sequence characteristics of the enemy aircraft based on the action parameter information and the tactical behavior prediction model includes: The action parameter information is subjected to data preprocessing, and the preprocessed action parameter information is input into the tactical behavior prediction model to predict the next maneuver sequence characteristics of the enemy aircraft.

4. The method according to claim 1, wherein The constructing of a situation analysis model based on the spatial positions of the enemy aircraft and our aircraft includes: Using the onboard radar of the friendly aircraft to detect the enemy aircraft to determine the spatial position of the enemy aircraft relative to the friendly aircraft; Determining the attitude angle of the enemy aircraft relative to the incident beam of the airborne radar of the friendly aircraft at each moment based on the spatial position; wherein the attitude angle includes: azimuth and pitch angle; Determining a radar cross section (RCS) value of the enemy aircraft based on the attitude angle; wherein the RCS value represents the RCS value of the enemy aircraft when exposed to the airborne radar of the friendly aircraft at each moment; The spatial situation is analyzed according to the RCS value of the enemy aircraft to construct a situation analysis model.

5. The method according to any one of claims 1 or 4, characterized in that: The situation analysis model is expressed as follows: ;in, Expressed as an angle situation analysis model, Expressed as a speed situation analysis model, Expressed as a distance situation analysis model, Expressed as a high-level situation analysis model; 、 、 and They are respectively represented as the weight coefficients corresponding to the angle situation analysis model, speed situation analysis model, distance situation analysis model and height situation analysis model; Wherein, the angle situation analysis model includes: azimuth advantage and entry angle advantage , respectively expressed as: Where, Indicates azimuth advantage; represents the entry angle advantage; 、 denote the weights of azimuth and entry angle respectively, ;in, , Where, Indicates azimuth advantage; represents the entry angle advantage; represents the entry angle, represents the azimuth, represents the maximum search azimuth, represents the maximum off-axis emission angle, represents the inescapable cone angle, where ; The speed situation analysis model is expressed as: hour, ;when When , the speed situation analysis model is expressed as: Where, Indicates the speed of our aircraft; Indicates the speed of the enemy aircraft; It is expressed as the best air combat speed of our aircraft; The distance situation analysis model is expressed as: , where Indicates distance advantage, Indicates the distance between our aircraft and the enemy aircraft; Indicates the maximum detection distance of the radar. Indicates the maximum attack distance. Indicates the maximum distance from which one cannot escape. Indicates the minimum distance from which one cannot escape. represents the minimum attack distance, where ; The height situation analysis model is expressed as follows: Where, Indicates high advantage; Indicates the altitude of our aircraft; Indicates the enemy aircraft's altitude; It indicates the optimal positioning height for our aircraft.

6. The method according to claim 1, characterized in that The making of an optimal countermeasure decision for the next maneuver sequence characteristics of the enemy aircraft based on the comprehensive occupation advantage value, the comprehensive occupation threat value, the comprehensive radar advantage value, and the comprehensive radar threat value includes: Determining a comprehensive advantage value based on the occupancy comprehensive advantage value and the radar comprehensive advantage value, and determining a comprehensive threat value based on the occupancy comprehensive threat value and the radar comprehensive threat value; When the comprehensive threat value is greater than or equal to the preset first threat threshold, and the comprehensive advantage value is less than the preset first advantage threshold, the confrontation decision is a first type decision; When the comprehensive threat value is greater than or equal to the preset second threat threshold and less than the preset first threat threshold, and the comprehensive advantage value is less than the preset second advantage threshold, the confrontation decision is a second type decision; When the comprehensive threat value is less than the preset third threat threshold, and the comprehensive advantage value is less than the preset first advantage threshold, the confrontation decision is a third type decision; Selecting an optimal adversarial decision from the first type of decision, the second type of decision, or the third type of decision; Wherein, the preset first threat threshold is greater than the preset second threat threshold; the preset second threat threshold is greater than the preset third threat threshold; the preset first advantage threshold is greater than the preset second advantage threshold; Among them, the first type of decision is an escape decision; the second type of decision is a defense decision; and the third type of decision is an attack decision.

7. An intelligent autonomous decision-making device for a drone, characterized in that: The device comprises: A data acquisition module is used to obtain a pre-built tactical behavior sample set and use the tactical behavior sample set to train the behavior prediction model to obtain a trained tactical behavior prediction model; An enemy aircraft maneuver sequence prediction module is used to obtain action parameter information of the enemy aircraft in a preset time sequence, and determine the next maneuver sequence characteristics of the enemy aircraft based on the action parameter information and the tactical behavior prediction model; A situation analysis establishment module is used to construct a situation analysis model based on the spatial position of the friendly aircraft and the enemy aircraft; wherein the situation analysis model includes: an angle situation analysis model, a distance situation analysis model, a speed situation analysis model, and an altitude situation analysis model; a position comprehensive value determination module, configured to determine the position comprehensive advantage value and position comprehensive threat value of the friendly aircraft relative to the enemy aircraft based on the situation analysis model; a radar comprehensive value determination module, configured to determine a radar comprehensive advantage value and a radar comprehensive threat value of the friendly aircraft's airborne radar relative to the enemy aircraft's airborne radar; a countermeasure decision determination module, configured to make an optimal countermeasure decision for the next maneuver sequence characteristics of the enemy aircraft based on the comprehensive position advantage value, the comprehensive position threat value, the comprehensive radar advantage value, and the comprehensive radar threat value; The behavior prediction model is a recurrent neural network model; the recurrent neural network model is based on the MobileNet-V1 network architecture; the convolutional layer of the recurrent neural network model is a depth-separable convolution structure composed of DWCONV+(1×1CONV); the activation function of the recurrent neural network model is RELU; Correspondingly, the data acquisition module includes: performing a zero-padding operation on a sample processing unit for each sample in the tactical behavior sample set to make the time series lengths of the samples consistent to obtain processed samples; wherein the tactical behavior sample set includes maneuver sequence features of typical tactical behaviors and maneuver parameter values corresponding to each maneuver sequence feature; A sample enhancement unit, configured to perform data enhancement on the processed sample to obtain a data-enhanced sample; a target sample set determination unit, configured to insert a level flight action after each sample action is completed for the data-enhanced samples to initialize the next sample action, thereby obtaining a target sample set; A model training unit is used to input the target sample set into the behavior prediction model to optimize the model parameters until the model parameters are optimized or the predefined loss function is minimized, thereby obtaining a trained tactical behavior prediction model; The module for determining the comprehensive placeholder value includes: a position threat value determination unit, configured to perform reverse engineering based on a preset CRITIC method to solve first variable weight coefficients corresponding to respective first indicators in the situation analysis model, and determine the sum of the products of the first variable weight coefficients and the corresponding first indicators to obtain a comprehensive position threat value of the friendly aircraft relative to the enemy aircraft; wherein the first indicators include: altitude, angle, distance, and speed; and wherein the first variable weight coefficient is a position threat coefficient; a position advantage value determination unit, configured to perform a forward solution based on the preset CRITIC method to obtain the second variable weight coefficient corresponding to each first indicator in the situation analysis model, and determine the sum of the products of each second variable weight coefficient and each corresponding first indicator to obtain a comprehensive position advantage value of the friendly aircraft relative to the enemy aircraft; wherein the second variable weight coefficient is the position advantage coefficient; Among them, the radar comprehensive value determination module includes: a signal-to-noise ratio determining unit, configured to determine a signal-to-noise ratio based on the power received by the radar antenna of the airborne radar of the friendly aircraft and the noise power generated; wherein the signal-to-noise ratio represents the radar monitoring capability; a maximum detection range determination unit, configured to obtain an RCS value of an enemy aircraft and determine a maximum detection range of the airborne radar of the friendly aircraft based on the signal-to-noise ratio and the RCS value of the enemy aircraft; a weight coefficient determination unit, configured to use the maximum detection range, the signal-to-noise ratio, and the RCS value as second indicators, and to simulate and calculate, based on preset digital combat simulator (DCS) software, first and second empirical value results corresponding to each of the second indicators, and use the first and second empirical value results as corresponding first and second weight coefficients, respectively; wherein the first weight coefficient is a radar threat coefficient corresponding to each of the second indicators; and the second weight coefficient is a radar advantage coefficient corresponding to each of the second indicators; The radar comprehensive value determination unit is used to determine the sum of the first products of each first weight coefficient and the corresponding second indicators to obtain the radar comprehensive advantage value of the airborne radar of the friendly aircraft relative to the airborne radar of the enemy aircraft, and to determine the sum of the second products of each second weight coefficient and the corresponding second indicators to obtain the radar comprehensive threat value of the airborne radar of the friendly aircraft relative to the airborne radar of the enemy aircraft.

8. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the UAV intelligent autonomous decision-making method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the UAV intelligent autonomous decision-making method according to any one of claims 1 to 6 when executed.

10. A computer program product, characterized in that The computer program product comprises a computer program, which, when executed by a processor, implements the intelligent autonomous decision-making method for a drone according to any one of claims 1 to 6.

Images