A data-driven guidance situation assessment method

Abstract

The present invention discloses a guidance situation assessment method based on data-driven, comprising: in an inertial coordinate system, taking an aircraft as a mass point, establishing a reachable set and an isotimeline model for both the missile and the target during the guidance process; calculating a set of isotime points based on the reachable set and the isotimeline model for fitting, obtaining an isotime curve that meets the fitting index as sample data; establishing a neural network model and training it using the sample data, and generating dynamic isotimelines based on the performance parameters and position information of both parties under the current situation using the trained neural network model; and evaluating the guidance situation based on the generated isotimelines. The present invention uses data-driven technology, and the model can be continuously optimized through continuous learning. The model trained using supervised learning can more accurately reflect complex and changeable combat situations, thereby improving the accuracy of situation assessment and significantly enhancing the response speed and combat effectiveness of the command and control system.

Description

Technical Field

[0001] The present invention belongs to the field of guidance and control of aerospace vehicles, and relates to a guidance situation assessment method based on data driving. Background Art

[0002] An equal-time line (ETL) is a continuous curve connecting the intersection points of two aircraft flying along a Dubins trajectory. It describes the set of all possible locations where, under given speed constraints and aircraft maneuverability, a given aircraft takes the same time to reach its target location while starting from a starting point along the Dubins path. ETLs are crucial for situational assessment and the development of interception / penetration strategies, and they also provide crucial information for the design of collaborative guidance laws. In complex interception / penetration scenarios, utilizing ETLs can provide precise predictions of the areas the attacker may pass through within a certain period of time. This effectively predicts the attacker's future location and possible intersection points, enabling the defender to plan ahead and develop more effective guidance strategies, optimize interception paths, and even conduct preemptive strikes. By comparing the overlapping areas of the ETLs, the relative maneuverability advantages and disadvantages of the two combatants can be assessed, providing a scientific basis for command decision-making.

[0003] In practical applications, the calculation of isotimelines requires a high degree of real-time performance to cope with the ever-changing battlefield situation. However, current isotimeline calculations typically use numerical iteration methods, which are computationally expensive and inefficient, making them difficult to apply in modern air combat scenarios. Therefore, in-depth research and optimization of isotimelines can overcome their limitations and help improve the guidance performance of existing guidance strategies in complex geometric situations. Summary of the Invention

[0004] The purpose of the present invention is to provide a data-driven guidance situation assessment method to solve the problem that the existing timeline calculation efficiency is low and it is difficult to apply in practice.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] A data-driven guidance situation assessment method, comprising:

[0007] In the inertial coordinate system, taking the aircraft as a mass point, the reachable set and equal timeline model of both the missile and the target during the guidance process are established;

[0008] According to the reachable set and the isotime line model, the isotime point set is calculated for fitting, and the isotime curve that meets the fitting index is obtained as the sample data;

[0009] A neural network model is built and trained using sample data. Based on the performance parameters and position information of both parties under the current situation, the trained neural network model is used to generate dynamic timelines.

[0010] Based on the generated isotimeline, the guidance situation is evaluated.

[0011] Furthermore, the kinematic models of the missile and target during the guidance process are:

[0012]

[0013]

[0014]

[0015] Among them, the subscript Indicates aircraft T or aircraft M, Indicates speed, represents the heading angle, represents the normal acceleration, represents the maximum normal acceleration of the aircraft T, represents the maximum normal acceleration of the aircraft M.

[0016] Furthermore, the dynamic models of the missile and target in the guidance process are:

[0017]

[0018] in, and Represents the state vector and control input, then The state vector at the moment is:

[0019] .

[0020] Furthermore, the reachable set is

[0021]

[0022] in, represents the initial position, Indicates The state vector at time t, Indicates the time range [ ] continuous control instructions within, represents all feasible control instructions within the constraints.

[0023] Furthermore, the isotimeline model is

[0024]

[0025] in, represents the speed of the aircraft M, represents the speed of the aircraft T, represents the maximum normal acceleration of the aircraft M, represents the maximum normal acceleration of the aircraft T, express The position of aircraft M at time t, express The position of the aircraft T at time t, Indicates the earliest interception time.

[0026] Furthermore, the fitting method of the isochronous curve is:

[0027] Taking the earliest equal-time point of both the missile and the target as the starting time, the intersection of the frontiers of the reachable sets of both the missile and the target is calculated as the equal-time point set in the guidance scenario at this time. The equal-time curve is obtained by fitting the equal-time point set through a polynomial.

[0028] The polynomial fitting includes quadratic polynomial fitting and linear polynomial fitting.

[0029] Furthermore, the neural network model is

[0030]

[0031] in, represents the sight angle of the aircraft T, represents the velocity lead angle of the aircraft T, represents the velocity lead angle of the aircraft M, Indicates the relative distance between the two aircraft. represents the speed of the aircraft T, and Represents the network structure.

[0032] Furthermore, the neural network model is a multi-layer perceptron structure, which includes an input layer, an output layer and several hidden layers.

[0033] Furthermore, each of the input layer, output layer and several hidden layers includes several neurons, each of the neurons is connected to each neuron in the next layer, and the connections between the neurons are weighted.

[0034] Furthermore, the performance parameters and position information of both parties under the current situation include the relative distance between the two aircraft, the positions of the two aircraft, the speed lead angles of the two aircraft, the line of sight angles of the two aircraft, and the heading angles of the two aircraft.

[0035] Compared with the prior art, the present invention has the following beneficial effects:

[0036] The present invention provides a data-driven guidance situation assessment method. Based on an isotimeline calculation method, data-driven technology is applied to extract features from isotimelines to obtain data samples. Supervised learning is then performed on the data samples. The nonlinear relationship between multiple characteristic parameters of the isotimeline and the performance parameters of both parties is obtained using a data-driven approach. Polynomial fitting is then performed on the geometric shape of the isotimeline based on a large amount of sample data. In practical applications, by acquiring situation information from both parties, the isotimeline can be quickly fitted using a trained model, thereby achieving rapid and accurate guidance situation assessment. Using data-driven technology, the present invention can automatically learn and extract the nonlinear relationship between multiple characteristic parameters of the isotimeline and the performance parameters of both parties from a large amount of data. As data accumulates and updates, the model can be continuously optimized through continuous learning and automatically adapt to changes in the combat situation. Using the model trained through supervised learning, in practical applications, only the situation information of both parties needs to be input to quickly fit the isotimeline through the model. This allows for more accurate reflection of complex and changing combat situations, thereby improving the accuracy of situation assessment and significantly enhancing the response speed and combat effectiveness of the command and control system. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. It should be understood that the following drawings only illustrate certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without paying any creative work.

[0038] Figure 1 It is a schematic diagram of the situation parameters of both the missile and the target during the guidance process of the present invention.

[0039] Figure 2 Schematic diagram of the reachable set of the uniform motion aircraft of the present invention.

[0040] Figure 3 This is a schematic diagram of the earliest time point of the present invention.

[0041] Figure 4 It is a schematic diagram of the isochronous timeline of the present invention.

[0042] Figure 5 A flow chart is generated for the sample of the present invention.

[0043] Figure 6 Schematic diagram of the neural network model structure of the present invention.

[0044] Figure 7This is a flow chart of the data-driven guidance situation assessment method of the present invention.

[0045] Figure 8 This is a diagram showing the fitting effects of the training set and the test set in Example 1 of the present invention.

[0046] Figure 9 Root mean square error diagram of the training set and the test set in Example 1 of the present invention.

[0047] Figure 10 This is a graph of the mean square error of the network fitting of the training set and the test set in Example 1 of the present invention.

[0048] Figure 11 This is a schematic diagram of the isochronous timeline generated in Example 1 of the present invention.

[0049] Figure 12 This is a comparison diagram of the isotimelines generated by the data-driven method and the isotimelines generated by the numerical method in Example 1 of the present invention.

[0050] Figure 13 This is a schematic diagram of isochronous timelines generated in a two-to-one interception scenario in Example 1 of the present invention.

[0051] Figure 14 This is a diagram of the interception effect when flying at an initial heading angle of 95° in Example 1 of the present invention.

[0052] Figure 15 This is a diagram of the interception effect when flying at an initial heading angle of 90° in Example 1 of the present invention.

[0053] Figure 16 Schematic diagram of the structure of a data-driven guidance situation assessment system according to a preferred embodiment of the present invention.

[0054] Figure 17 This is a schematic diagram of the structure of an electronic device according to a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0055] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions of the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings herein can be arranged and designed in various different configurations.

[0056] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention as claimed, but rather merely represents selected embodiments of the present invention. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of the present invention without creative effort are intended to fall within the scope of protection of the present invention.

[0057] The following description of exemplary embodiments of the present application is made in conjunction with the accompanying drawings, including various details of the embodiments of the present application to facilitate understanding. These details should be considered as merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present application. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0058] Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making any creative work shall fall within the scope of protection of this application.

[0059] It should be noted that the terminals involved in the embodiments of the present application may include but are not limited to mobile phones, personal digital assistants (PDAs), wireless handheld devices, tablet computers, personal computers (PCs), MP3 players, MP4 players, wearable devices (for example, smart glasses, smart watches, smart bracelets, etc.), smart home appliances and other smart devices.

[0060] In this document, the term "and / or" simply describes a relationship between related objects, indicating that three possible relationships exist. For example, "A and / or B" can represent: A exists alone, A and B exist simultaneously, or B exists alone. Furthermore, the character " / " in this document generally indicates that the related objects are in an "or" relationship.

[0061] The present invention is described in further detail below with reference to the accompanying drawings:

[0062] See also Figure 1 The present invention provides a data-driven guidance situation assessment method, comprising the following steps:

[0063] Step 1: Assume that in the inertial coordinate system In the figure, aircraft M and aircraft T are point masses, and their reachable sets and equal timeline models are established.

[0064] During the guidance process, the situation parameters of both the missile and the target are as follows: Figure 1 shown.

[0065] The kinematic model of both parties is shown as follows:

[0066]

[0067]

[0068]

[0069] Among them, the subscript Indicates aircraft T or aircraft M, Indicates speed, represents the heading angle, represents the normal acceleration, and They represent the maximum normal acceleration of aircraft T and aircraft M respectively. This value determines the minimum turning radius. The speed and maneuverability of aircraft T are better than those of aircraft M.

[0070] The dynamic model of both parties can be summarized as:

[0071]

[0072] in, and Represents the state vector and control input, then The state vector at the moment is:

[0073]

[0074] According to the kinematic equation, a certain aircraft M or aircraft T is The reachable set at time is defined as the initial position Next, in The set of all possible locations that can be reached at a moment:

[0075]

[0076] in, Indicates the time range [ ] continuous control instructions within, represents all feasible control instructions within the constraints.

[0077] Analyzing reachable sets using Dubins shortest path trajectories, Dubins shortest paths can be divided into three categories: straight-line paths, minimum-turn-radius paths, and minimum-turn-radius paths combined with straight-line paths. The shortest path trajectory is equivalent to the shortest-time trajectory. If both parties use Dubins shortest path trajectories to move, their guidance strategy can be defined as the minimum-time guidance strategy (MTG).

[0078] The reachable set front (RSF) is a curve consisting of the farthest distance points that can be reached within a period of time. The reachable set of a uniformly moving aircraft can be obtained by the involute of its left and right turning radius circles, such as Figure 2 shown.

[0079] The Equal-time Line (ETL) is a continuous boundary that both encountering aircraft can reach simultaneously when using the MTG guidance strategy. Points on the Equal-time Line are called Equal-time Points (ETPs), which are obtained by calculating the intersection of the frontiers of the reachable sets of both parties. When the flight time of both parties increases from zero, there is a moment when the frontiers of the reachable sets of both parties are tangent, such as Figure 3 As shown, the earliest intersection point is the earliest equal time point.

[0080] Figure 3 middle and are the centers of the minimum turning radius of aircraft M and aircraft T respectively. The length of the fastest interception trajectory corresponding to the earliest equal time point is:

[0081]

[0082] in, and Indicates aircraft and aircraft The minimum turning radius is equal to and , and Indicates the center angle of the turning trajectory corresponding to the minimum turning radius circle, represents the length of the common tangent of the two turning circles, The value of varies depending on whether the two circles are inscribed or circumscribed:

[0083]

[0084] in,( , )and( , ) indicates a point and The coordinates of and Indicates the corresponding turning radius. The earliest interception time is recorded as , given by the total trajectory length and the sum of the velocities of the two encountering parties Calculated, and the earliest equal time point is when both parties The reachable set tangent point corresponding to the moment.

[0085] Starting from the earliest equal time point, the equal time line can be established by the intersection of the frontiers of the equal time reachable sets of both parties, as shown in the following formula. The calculated equal time line example is Figure 4 shown.

[0086]

[0087] Step 2: Obtain training samples and parameterize the situation information.

[0088] There are many characteristic parameters that can be selected to describe the battle scene of generating equal time lines, but for the sake of simplicity and from the perspective of describing the collision triangle, the relative distance is selected. , the velocity lead angle of the aircraft T , the velocity lead angle of aircraft M , the sight angle of the aircraft T Four characteristic parameters.

[0089] By setting different parameters in the interception scenario, calculating the set of equal time points, and then fitting according to the calculated equal time points, we can get the equal time curve that meets the fitting index as the sample data. The interception scenario generated by the sample is in a two-dimensional inertial coordinate system. The range of the initial missile-target distance R is set according to the requirements, and the sight angle range is set according to the interception scenario. rad, aircraft speed in M Set to a constant value according to requirements, aircraft T speed It can be set to a certain range, where the heading angle coincides with the sight angle. Since the premise of generating samples of equal time lines is to form a collision triangle, and the leading angle range meets the requirements of forming a collision triangle, the sample leading angle range is

[0090] .

[0091] The earliest equal time point for both the missile and the target is taken as the starting time, and the intersection of the frontiers of the reachable sets of both the missile and the target is calculated and recorded as the set of equal time points in the guidance scenario at this time. The set of equal time points is fitted with a polynomial to obtain the equal time line curve. In order to conform to the shape change law of the equal time line, the fitting of the equal time line is divided into two sections. The equal time curve is fitted with a quadratic polynomial when the distance is far, and a linear polynomial when the distance is close. The equal time curve is fitted with a polynomial, and the fitting index requires the fitting degree to be higher than a certain set value, such as 0.99, so the geometric characteristic parameters of the equal time line are the quadratic polynomial coefficients. The sample generation process is as follows: Figure 5 shown.

[0092] Step 3: Build a neural network model and set the training method.

[0093] The four parameters of sight angle, aircraft T lead angle, aircraft M lead angle, relative distance, and aircraft T speed are used as input variables, and the fitting coefficient of the isochronous curve is used as the output variable. The network model is established as follows:

[0094]

[0095] The network adopts a multilayer perceptron (MLP) structure. Multilayer perceptron (MLP) is an artificial neural network and the most basic feedforward neural network. The network structure is shown in the figure below. Figure 6 As shown in Figure 2, an MLP consists of an input layer, one or more hidden layers, and an output layer. Each layer contains multiple neurons (also called nodes or units). Each neuron in an MLP is connected to every neuron in the next layer, and each connection has a weight.

[0096] Figure 6 middle, The network has 5 layers, 5 input parameters, 3 output parameters, and the hidden layer is divided into 3 layers. Hidden layer 1 has 16 neurons, hidden layer 2 has 24 neurons, and hidden layer 3 has 16 neurons. Network structure and The network is similar and only two network parameters are output.

[0097] The network training method adopts the Levenberg-Marquardt (LM) algorithm. After normalizing the output data, the network weights are initialized using a random method, and hyperparameter adjustment and model optimization are performed as needed.

[0098] Step 4: Generate dynamic isotimes based on the trained model and evaluate the guidance situation based on the generated isotimelines.

[0099] Taking the performance parameters and position information of both parties under the current situation as input, the trained network model function is called to generate the timeline.

[0100] The function inputs include: aircraft T sight angle, aircraft T lead angle, aircraft M lead angle, relative distance, aircraft T position, aircraft M position and aircraft T heading angle.

[0101] The output of the function includes: equal time point coordinates and equal time line fitting polynomial coefficients.

[0102] By connecting all equal time points, an equal time line can be obtained, and the generated equal time line can be used to evaluate the guidance situation.

[0103] The present invention is described in further detail below with reference to specific embodiments:

[0104] Example 1:

[0105] This embodiment uses one-on-one and two-on-one interception scenarios at a constant speed with the attacking party having a performance advantage as examples to verify the effectiveness and superiority of the equal timeline generation method proposed in the present invention.

[0106] First, training samples need to be obtained and trained to facilitate the subsequent dynamic generation of reachable sets.

[0107] Given different situation parameters in a combat scenario, we calculate the isochronous points and fit them into isochronous lines. The isochronous curves that meet the fitting criteria are the sample data. Based on the commonly used aircraft performance parameters and environmental parameters in the interception scenario, the relative distance range of the sample is set to:

[0108]

[0109] The viewing angle range is set to:

[0110]

[0111] To satisfy the formation of the collision triangle, the lead angles and lead angle ranges of the sample's aircraft M and aircraft T are:

[0112]

[0113] The earliest equal-time point between the missile and the target is used as the starting time. The intersection of the frontiers of the reachable sets of both the missile and the target is calculated and recorded as the set of equal-time points in the guidance scenario at that time. This set of equal-time points is then fitted using a polynomial to obtain an equal-time line curve. The equal-time line fitting is divided into two stages: a quadratic polynomial is used for long distances, and a linear polynomial is used for short distances. The required fit index is 0.99.

[0114] After obtaining the sample, define the network model:

[0115]

[0116] Multilayer Perceptron (MLP) is used for training. According to the defined network model, the network parameters are set as follows:

[0117] 、 The network consists of 5 layers, namely input layer, hidden layer 1, hidden layer 2, hidden layer 3 and output layer. The number of neurons or input and output numbers in each layer are shown in Table 1 below:

[0118] Table 1 Network parameters

[0119]

[0120] The LM algorithm is used for training. After normalizing the output data, the network weights are initialized using a random method. Hyperparameters are adjusted and the model is optimized as needed. The training strategy uses the Adam optimizer, with a minimum batch size of 32 and a learning rate of 0.001. The trained network is tested using the sample set to verify the training effect of the neural network. After 3000 iterations, the fitting effect graphs, root mean square error, and network fitting mean square error graphs of the training set and test set are as follows: Figure 8 、 Figure 9 、 Figure 10 As shown in the figure, the training effect is good, and the best training effect is achieved after the 46th round of training.

[0121] After training is completed, verify the accuracy of the dynamically generated timeline in a one-to-one scenario.

[0122] In this interception scenario, since aircraft T is faster, aircraft M cannot intercept it and cannot generate equal timelines. Therefore, it is set that both aircraft fly towards each other. The performance parameters of both aircraft are set as follows:

[0123] Table 2 Performance parameters of both parties in one-to-one scenario

[0124]

[0125] First, import the performance parameters and position information of both parties under the current situation, and then call the trained network model function to generate the isochronous timeline. At a certain position, the input and output values ​​of the function are shown in Table 3 and Table 4, and the isochronous timeline is generated as follows: Figure 11 shown.

[0126] Table 3 Input parameters of equal timeline function

[0127]

[0128] Table 4 Input and output parameters of equal timeline function

[0129]

[0130] The isotimelines generated by the data-driven approach are compared with those generated by the numerical approach.

[0131] It can be seen that the data-driven generated isotimeline fitting effect is good. The simulation time and maximum error under the experimental equipment are shown in the following table.

[0132] Table 5 Input and output parameters of equal timeline function

[0133]

[0134] From the table analysis, it can be seen that the isotimeline generation method proposed in the present invention is faster in simulation time and has a smaller error range than the method calculated by numerical iteration, which reflects the superiority of the present invention.

[0135] The following is an application example of dynamic generation of equal timelines in a two-to-one interception scenario.

[0136] Import the performance parameters and position information of both parties under the current situation, and then call the trained network model function to generate the isotimeline. At a certain position, the performance parameters and situation information of both parties are shown in Table 6, and the isotimeline is generated as shown in Table 6. Figure 13 shown.

[0137] Table 6 Performance parameters and situation information of both sides

[0138]

[0139] like Figure 13 As shown in the figure, there is a gap between the equal time lines generated by aircraft M1 and aircraft M2 for the incoming missile. When aircraft T flies at initial heading angles of 95° and 90° respectively, aircraft M1 and aircraft M2 use proportional guidance law for interception with a guidance law coefficient of 4. The interception effect is as follows: Figure 14 、 Figure 15 shown.

[0140] When aircraft T flew at a heading angle of 95°, its flight path fell within the coverage of the isotimeline, and aircraft T was ultimately intercepted by aircraft M1. However, when aircraft T flew at a heading angle of 90°, its flight path crossed the gap in the isotimeline, falling outside its coverage. During the interception, the minimum relative distance between aircraft T and aircraft M was 25.58 meters, exceeding the kill radius required for successful interception by aircraft M, allowing aircraft T to successfully penetrate. This demonstrates that interception / penetration strategies in air combat scenarios can be formulated based on real-time generated isotimelines, thereby improving the success rate of interception / penetration in complex geometric situations.

[0141] Example 2:

[0142] The embodiment 2 provided by the present invention is an embodiment of the guidance situation assessment system based on data drive provided by the present invention, such as Figure 16 As shown, an embodiment of the system includes: a modeling module, a fitting module, a training module and an evaluation module.

[0143] A modeling module, wherein the modeling module is used to establish a reachable set and an equal timeline model of both the missile and the target during the guidance process in an inertial coordinate system with the aircraft as a mass point;

[0144] A fitting module is used to calculate and fit a set of equal time points based on a reachable set and an equal time line model, and obtain an equal time curve that meets the fitting index as sample data;

[0145] A training module, which is used to establish a neural network model and train it using sample data, and to generate dynamic isotimelines using the trained neural network model based on the performance parameters and position information of both parties under the current situation;

[0146] An evaluation module is used to evaluate the guidance situation according to the generated isotimeline.

[0147] It can be understood that the data-driven guidance situation assessment system provided by the present invention corresponds to the data-driven guidance situation assessment method provided in the aforementioned embodiments. The relevant technical features of the data-driven guidance situation assessment system can refer to the relevant technical features of the data-driven guidance situation assessment method, which will not be repeated here.

[0148] Another object of the present invention is to provide an electronic device, such as Figure 17 As shown, it includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor executes the steps of the data-driven guidance situation assessment method.

[0149] The data-driven guidance situation assessment method comprises the following steps:

[0150] In the inertial coordinate system, taking the aircraft as a mass point, the reachable set and equal timeline model of both the missile and the target during the guidance process are established;

[0151] According to the reachable set and the isotime line model, the isotime point set is calculated for fitting, and the isotime curve that meets the fitting index is obtained as the sample data;

[0152] A neural network model is built and trained using sample data. Based on the performance parameters and position information of both parties under the current situation, the trained neural network model is used to generate dynamic timelines.

[0153] Based on the generated isotimeline, the guidance situation is evaluated.

[0154] A fourth object of the present invention is to provide a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the data-driven guidance situation assessment method are implemented.

[0155] The data-driven guidance situation assessment method comprises the following steps:

[0156] In the inertial coordinate system, taking the aircraft as a mass point, the reachable set and equal timeline model of both the missile and the target during the guidance process are established;

[0157] According to the reachable set and the isotime line model, the isotime point set is calculated for fitting, and the isotime curve that meets the fitting index is obtained as the sample data;

[0158] A neural network model is built and trained using sample data. Based on the performance parameters and position information of both parties under the current situation, the trained neural network model is used to generate dynamic timelines.

[0159] Based on the generated isotimeline, the guidance situation is evaluated.

[0160] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROMs, optical storage, etc.) containing computer-usable program code.

[0161] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0162] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0163] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0164] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered by the scope of protection of the claims of the present invention.

Claims

1. A data-driven guidance situation assessment method, characterized in that: include: In the inertial coordinate system, taking the aircraft as a mass point, the reachable set and equal timeline model of both the missile and the target during the guidance process are established; According to the reachable set and the isotime line model, the isotime point set is calculated for fitting, and the isotime curve that meets the fitting index is obtained as the sample data; A neural network model is built and trained using sample data. Based on the performance parameters and position information of both parties under the current situation, the trained neural network model is used to generate dynamic timelines. Evaluate the guidance situation based on the generated isotimeline; The relative distance between the two aircraft, the velocity lead angle of the two aircraft, and the line of sight angle of the two aircraft are selected as characteristic parameters. By setting different parameters in the interception scenario, a set of equal time points is calculated. Then, fitting is performed based on the calculated equal time points to obtain an equal time curve that meets the fitting index as sample data; The neural network model is in, represents the sight angle of the aircraft T, represents the velocity lead angle of the aircraft T, represents the velocity lead angle of the aircraft M, Indicates the relative distance between the two aircraft. represents the speed of the aircraft T, and Represents the network structure; The performance parameters and position information of both parties under the current situation include the relative distance between the two aircraft, the positions of the two aircraft, the velocity lead angles of the two aircraft, the line of sight angles of the two aircraft, and the heading angles of the two aircraft; The output of the neural network model includes isotime point coordinates and isotime line fitting polynomial coefficients.

2. The data-driven guidance situation assessment method according to claim 1, wherein: The kinematic models of the missile and target during the guidance process are: Among them, the subscript Indicates aircraft T or aircraft M, Indicates speed, represents the heading angle, represents the normal acceleration, represents the maximum normal acceleration of the aircraft T, represents the maximum normal acceleration of the aircraft M.

3. The data-driven guidance situation assessment method according to claim 1, wherein: The dynamic models of both the missile and the target during the guidance process are: in, and Represents the state vector and control input, then The state vector at the moment is: 。 4. The data-driven guidance situation assessment method according to claim 1, wherein: The reachable set is in, represents the initial position, Indicates The state vector at time t, Indicates the time range [ ] continuous control instructions within, represents all feasible control instructions within the constraints.

5. The data-driven guidance situation assessment method according to claim 1, characterized in that: The isotimeline model is in, represents the speed of the aircraft M, represents the speed of the aircraft T, represents the maximum normal acceleration of the aircraft M, represents the maximum normal acceleration of the aircraft T, express The position of aircraft M at time t, express The position of the aircraft T at time t, Indicates the earliest interception time.

6. The data-driven guidance situation assessment method according to claim 1, characterized in that: The fitting method of the isochronous curve is: Taking the earliest equal-time point of both the missile and the target as the starting time, the intersection of the frontiers of the reachable sets of both the missile and the target is calculated as the equal-time point set in the guidance scenario at this time. The equal-time curve is obtained by fitting the equal-time point set through a polynomial. The polynomial fitting includes quadratic polynomial fitting and linear polynomial fitting.

7. The data-driven guidance situation assessment method according to claim 1, characterized in that: The neural network model is a multi-layer perceptron structure, which includes an input layer, an output layer and several hidden layers.

8. The data-driven guidance situation assessment method according to claim 7, characterized in that: Each of the input layer, output layer and several hidden layers includes several neurons, each of the neurons is connected to each neuron in the next layer, and the connections between the neurons have weights.

Images