Missile interception method and device based on digital main line and storage medium

Abstract

The invention relates to the technical field of data processing, in particular to a missile interception method and device based on a digital main line and a storage medium. Comprising the following steps: acquiring flight data of a missile at different time points collected by a radar of a space satellite; according to the time sequence of different time points, taking the flight data corresponding to the first time point set as input, taking the flight data corresponding to the second time point as output, and constructing a regression model; all time points in the first time point set are arranged before the second time point in the time sequence; performing iterative optimization on the regression model by using a bionic algorithm and flight data to obtain a trajectory driving model; and intercepting the missile according to a prediction result of the trajectory driving model. According to the embodiment of the invention, the accuracy of predicting the missile trajectory can be improved, so that the hit rate of intercepting missiles is improved.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a missile interception method, device and storage medium based on a digital main line. Background Art

[0002] Missiles are important weapons in modern warfare, with extremely strong destructive power, which can pose a huge threat to people's lives and property safety. Therefore, missile interception based on the digital mainline is crucial.

[0003] The current missile interception method usually uses base radar to detect and track the missile trajectory, and then uses interceptor missiles or laser beams to destroy the missile to achieve defense. However, the current interception method still has limited hit rates when facing high-speed and complex targets. Summary of the invention

[0004] In order to solve the above technical problems, the present application provides a missile interception method, device and storage medium based on a digital main line, which can improve the accuracy of predicting the missile trajectory, thereby improving the hit rate of intercepting the missile.

[0005] In a first aspect, the present application provides a missile interception method based on a digital main line, including: obtaining flight data of a missile collected by a radar of a space satellite at different time points; constructing a regression model according to the time sequence of different time points, using the flight data corresponding to a first time point set as input and the flight data corresponding to a second time point as output; all time points in the first time point set are arranged before the second time point in the time sequence; iteratively optimizing the regression model using a bionic algorithm and flight data to obtain a trajectory-driven model; and intercepting the missile according to the prediction results of the trajectory-driven model.

[0006] In the second aspect, the present application provides a missile interception device based on a digital main line, including: an acquisition module, used to obtain flight data of the missile collected by the radar of the space satellite at different time points; a construction module, used to construct a regression model according to the time sequence of different time points, using the flight data corresponding to the first time point set as input and the flight data corresponding to the second time point as output; all time points in the first time point set are arranged before the second time point in the time sequence; a generation module, used to iteratively optimize the regression model using a bionic algorithm and flight data to obtain a trajectory-driven model; an interception module, used to intercept the missile according to the prediction results of the trajectory-driven model.

[0007] In a third aspect, the present application provides an electronic device comprising: a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein when the computer program is executed by the processor, the digital main line-based missile interception method of the first aspect is implemented.

[0008] In a fourth aspect, the present application provides a computer-readable storage medium, comprising: a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, it implements the missile interception method based on the digital main line as in the first aspect.

[0009] In a fifth aspect, the present application provides a computer program product, comprising: when the computer program product runs on a computer, the computer implements the missile interception method based on the digital main line as in the first aspect.

[0010] The technical solution provided by the present application has the following advantages over the prior art: first, the flight data of the missile collected by the radar of the space satellite at different time points is obtained, and according to the time sequence of the different time points, the flight data corresponding to the first time point set is used as input, and the flight data corresponding to the second time point is used as output to construct a regression model. Among them, all time points in the first time point set are arranged before the second time point in the time sequence. Afterwards, the regression model is iteratively optimized using the bionic algorithm and flight data to obtain a trajectory-driven model. Finally, the missile is intercepted according to the prediction results of the trajectory-driven model. In this way, on the one hand, a regression model can be constructed based on the flight data, which effectively solves the problem of missile interception error deviation of the pure dynamics model; on the other hand, the regression model can be iteratively optimized using the bionic algorithm and flight data to obtain a trajectory-driven model, which effectively improves the accuracy of the generated trajectory-driven model. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0012] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0013] Figure 1 One of the flowcharts of the missile interception method based on the digital main line provided in the embodiment of the present application;

[0014] Figure 2 A second flowchart of a missile interception method based on a digital main line provided in an embodiment of the present application;

[0015] Figure 3 One of the structural schematic diagrams of a missile interception device based on a digital main line provided in an embodiment of the present application;

[0016] Figure 4The second structural diagram of the missile interception device based on the digital main line provided in the embodiment of the present application;

[0017] Figure 5 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0018] In order to more clearly understand the above-mentioned purposes, features and advantages of the present application, the scheme of the present application will be further described below. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0019] In the following description, many specific details are set forth to facilitate a full understanding of the present application, but the present application may also be implemented in other ways different from those described herein; obviously, the embodiments in the specification are only part of the embodiments of the present application, rather than all of the embodiments.

[0020] Hypersonic missiles refer to missile weapons and equipment that fly in the atmosphere at speeds exceeding Mach 5. They have the characteristics of high speed, long range, and strong maneuverability. Among them, Mach number refers to the ratio of the speed of a hypersonic missile to the speed of sound.

[0021] Hypersonic missile defense has its own unique characteristics: First, the trajectory of hypersonic missiles is changeable and the position is difficult to predict. Hypersonic missiles fly in the atmosphere and can use aerodynamic forces to change their flight trajectories. For example, air-breathing hypersonic cruise missiles can use their engines to perform large-scale lateral maneuvers. Therefore, the traditional method of relying on ballistic calculations to predict missile trajectories is no longer applicable, and its position can only be determined by continuous detection and tracking. Second, the orbit of hypersonic missiles is low and the detectable distance is short. The orbit of hypersonic missiles greatly limits the detection distance of ground-based radars. At present, the maximum detection distance of early warning radars for hypersonic targets is only about 1,000 km, which can only provide about 11 minutes of early warning time.

[0022] Therefore, hypersonic missiles have the characteristics of long range, high speed, strong mobility and penetration capability, which pose a full threat to the existing ballistic missile defense system. In a large country with a vast territory, it is difficult to achieve full coverage of the terminal defense system, and the mid-course defense system is required to defend against incoming missiles as extensively as possible. The existing missile interception system usually chooses to directly collide and destroy the incoming hypersonic missiles in the air, but the high maneuverability and high speed of hypersonic missiles make it difficult for traditional interceptor missiles to accurately hit them.

[0023] In view of the above problems, the embodiments of the present application provide a missile interception method, device and storage medium based on digital main line, the method includes: obtaining the flight data of the missile collected by the radar of the space satellite at different time points; according to the time sequence of different time points, the flight data corresponding to the first time point set is used as input, and the flight data corresponding to the second time point is used as output to construct a regression model; all time points in the first time point set are arranged before the second time point in the time sequence; the regression model is iteratively optimized using the bionic algorithm and flight data to obtain a trajectory driving model; the missile is intercepted according to the prediction results of the trajectory driving model. The regression model can be constructed according to the flight data, which effectively solves the problem of missile interception error deviation of the pure dynamics model; the regression model can also be iteratively optimized using the bionic algorithm and flight data to obtain a trajectory driving model, which effectively improves the accuracy of the generated trajectory driving model.

[0024] The missile interception method based on the digital main line provided in the embodiment of the present application can be executed by a missile interception device based on the digital main line, and the missile interception device based on the digital main line can be hardware or software. When the missile interception device based on the digital main line is hardware, it can be various electronic devices with the function of running missile interception based on the digital main line, including but not limited to computers, computers, tablet computers, etc. When the missile interception device based on the digital main line is software, it can be installed in the electronic devices listed above. It can be implemented as multiple software or software modules, or it can be implemented as a single software or software module. No specific limitation is made here.

[0025] Figure 1 A flowchart of a missile interception method based on a digital main line provided in an embodiment of the present application is shown in FIG. Figure 1 As shown, the missile interception method based on the digital main line may include the following steps.

[0026] S11. Obtain the flight data of the missile at different time points collected by the space satellite radar.

[0027] Specifically, when the radar of the space satellite detects a missile, it continuously detects the missile through the radar, obtains the flight data of the missile at different time points, and sends the warning information and flight data to the missile interception device based on the digital main line through the low-orbit transmission layer satellite. Among them, the digital main line refers to the digital data flow built using modeling and simulation tools, covering the entire life cycle and value chain of the product, from basic materials, design, process, manufacturing to use and maintenance, integrating and driving the product design, manufacturing and guarantee with a unified model as the core.

[0028] In some embodiments, the flight data may be collected by ordinary space satellites, by dedicated space satellites in a missile defense system, or by both ordinary space satellites and dedicated space satellites.

[0029] For example, when the radar on an ordinary space satellite detects a missile, it transmits detailed measurement data to the dedicated space satellite in the missile defense system. After the dedicated space satellite successfully captures the missile based on the measurement data, it continues to track the missile using its onboard detection equipment to collect its flight data, and sends the early warning information and flight data to the missile interception device based on the digital main line through the transport layer satellite network in low Earth orbit.

[0030] In some embodiments, the flight data includes at least the position state of the missile and the time interval between the current position state and the previous position state, and the position state includes longitude, latitude, speed, heading, and altitude.

[0031] In some embodiments, the missile is a hypersonic missile.

[0032] S12. According to the time sequence of different time points, the flight data corresponding to the first time point set is used as input, and the flight data corresponding to the second time point is used as output to construct a regression model.

[0033] All time points in the first time point set are arranged before the second time point in the time sequence. The first time point set may include multiple time points or one time point, for example, the first time point set includes three time points.

[0034] Exemplarily, when there are three time points in the first time point set, the process of the regression model can be described by the following formula: s (t+1)=f(x s (t), x s (t-1), x s (t-2)), where x s (t) is used to represent the flight data at time t, f(x s (t), x s (t-1), x s (t-2)) is a nonlinear mapping function, x s (t-1) is used to represent the flight data at the moment before time t. s (t+1) is used to represent the flight data at the next moment after time t, x s (t-2) is used to represent the flight data at the moment before (t-1).

[0035] Considering that missile motion involves multiple motion states, the missile is regarded as a motion model with six degrees of freedom. The flight data at time t includes: xs (t) = {X t , Y t , V t , R t , H t , ΔT}. Among them, X t Used to indicate the longitude of the missile at time t, Y t Used to indicate the missile latitude at time t, V t Used to represent the speed of the missile at time t, R t Used to indicate the heading of the missile at time t, H t It is used to indicate the altitude of the missile at time t, and ΔT is the time interval between the current position state and the previous position state.

[0036] In some embodiments, the ability to build data-driven models provided by the artificial intelligence engine in the digital main line platform can be used to construct a regression model based on the flight data transmitted by the radar of the space satellite.

[0037] In some embodiments, according to the time sequence of different time points, the flight data corresponding to the first time point set is used as input, and the flight data corresponding to the second time point is used as output. The way to construct the regression model can also be to use the flight data corresponding to the first time point set as input, and the flight data corresponding to the second time point as output, based on the polynomial chaos expansion algorithm, use the N-order truncation method to perform finite truncation approximation to construct the regression model. Among them, N in the N-order truncation is preset, for example, a default value, or a value set by relevant personnel according to actual conditions. For example, N = 2, or N = 3.

[0038] Specifically, Polynomial Chaos Expansion (PCE) is the most widely used meta-model for propagating uncertainty. It originates from the homogeneous chaos theory proposed by Wiener in 1938 and is used to reflect any random process with a finite second-order distance. For a function y=f(x) containing independent random variables, according to the polynomial chaos expansion theory, f(x) can be approximated by the following orthogonal polynomial chaos expansion forms corresponding to the input variables.

[0039]

[0040] Where i=[i1i2…i N ] indicates a multi-order subscript, satisfying |i|=i1+i2+…+i N ; x is multidimensional input data (i.e. flight data), x = [x (1) ,x (2) ,…,x (N) ] T ; a i isi (x) corresponds to the expansion coefficient to be determined; ψ i (x) is a multivariate polynomial basis function satisfying the following equation: ∫ψ α (x)ψ β (x)dx=E(ψ α (x)ψ β (x)) = δ αβ . Where; E(·) is the expectation function, when α=β, δ αβ =1; when α≠β, δ αβ =0.

[0041] In practical applications, the expression is truncated to d terms. The general truncation criterion is to retain polynomials with an order not higher than d. For example, it can be expressed as:

[0042] Among them, α=[a0,α1,…,a d-1 ] T is the coefficient vector, the multivariate polynomial basis function ψ i (x) can be decomposed into the tensor product of the univariate basis functions, ψ a (x) = ψ a1 (x1)ψ a2 (x2)…ψ ai (x j )…ψ ad (x N ), ψ ai (x j ) is the variable x j The i-th order orthogonal polynomial

[0043] When using the above model for approximation, a finite truncation approximation can be obtained. The number of expansion terms P of the model is:

[0044] For the regression model, it can be expressed as the following linear equation form:

[0045] Among them, the prediction factor ψ i is a multivariate orthogonal polynomial in the input variables, u is the coefficient of pce, and Y is the corresponding response output.

[0046] The orthogonal polynomial corresponding to each input variable depends on the distribution form of the input variable. For uniformly distributed variables with values ​​between [-1,1], the optimal basis function is the Lengendre orthogonal polynomial. The larger the maximum expansion order N, the more accurate the polynomial expansion, but the number of expansion terms will also increase accordingly. For general engineering problems, N of 2 to 3 can ensure higher accuracy. For example, when N is 2, the expression of the regression model is as follows:

[0047] Where ψ1(x i ) is the variable x i The corresponding first-order orthogonal polynomial; ψ2(x i ,x j ) is the variable x i and x j The corresponding 2nd order orthogonal polynomial.

[0048] S13. Iteratively optimize the regression model using the bionic algorithm and flight data to obtain a trajectory driving model.

[0049] Specifically, in the process of building a regression model, the hyperparameters (such as regularization parameters) inside the regression model play a key role between the empirical risk and structural risk of the model. In order to solve this hyperparameter selection problem, the embodiment of the present application uses the Salamander Swarm Algorithm (SSA) to optimize the hyperparameters inside the regression model during the cross-validation stage, and determines the regression model corresponding to the optimal hyperparameters as the trajectory-driven model.

[0050] S14. Intercept the missile according to the prediction results of the trajectory driving model.

[0051] Specifically, after obtaining the trajectory-driven model, the latest flight data is input into the trajectory-driven model to predict the missile's flight path and impact point. Guided by this information, the optimal interception plan is calculated and executed to complete the missile interception based on the digital main line.

[0052] In some embodiments, after intercepting the missile according to the prediction results of the trajectory-driven model, the missile interception method based on the digital main line also includes, when it is determined that the missile has not been intercepted successfully, returning to execute steps S12-S14 until the missile is intercepted successfully. In this way, defensive combat operations can be effectively organized and implemented, which not only enhances the ability to respond immediately to potential threats, but also greatly improves the effectiveness and interception success rate of the overall system, thereby ensuring safety and stability in a complex and changing battlefield environment.

[0053] In the above scheme, first, the flight data of the missile collected by the radar of the space satellite at different time points are obtained, and according to the time sequence of different time points, the flight data corresponding to the first time point set is used as input, and the flight data corresponding to the second time point is used as output to construct a regression model. Among them, all time points in the first time point set are arranged before the second time point in the time sequence. Afterwards, the regression model is iteratively optimized using the bionic algorithm and flight data to obtain a trajectory-driven model. Finally, the missile is intercepted according to the prediction results of the trajectory-driven model. In this way, on the one hand, the regression model can be constructed according to the flight data, which effectively solves the problem of missile interception error deviation of the pure dynamic model; on the other hand, the regression model can be iteratively optimized using the bionic algorithm and flight data to obtain a trajectory-driven model, which effectively improves the accuracy of the generated trajectory-driven model. On the other hand, the flight data of the missile collected by the radar of the space satellite can be used to construct a trajectory-driven model, and the flight trajectory of the missile can be predicted to intercept the missile in the middle stage. Even if the initial interception fails, the secondary interception can be initiated, which improves the fault tolerance of the interception process and indirectly improves the success rate of the final interception result by increasing the number of interceptions.

[0054] In some embodiments, the method of iteratively optimizing the regression model using the bionic algorithm and flight data to obtain the trajectory driving model can be: first, the projection of the regression model in the multidimensional target space is calculated according to the hyperparameters of the regression model, and the projection is determined as the loss function of the regression model. Then, the bionic algorithm is used to iteratively optimize the regression model in the multidimensional target space according to the flight data and the loss function to obtain the trajectory driving model. The multidimensional target space is a square integrable function space with a weight function, also known as Space. The dimensions of the multidimensional target space are preset, for example, default values, or values ​​set by relevant personnel according to actual conditions. For example, the dimensions of the multidimensional target space may be 4 dimensions.

[0055] First, the projection of the regression model in the multidimensional target space is calculated according to the hyperparameters of the regression model, and the projection is determined as the loss function of the regression model.

[0056] The polynomial chaos coefficient determines the accuracy of the regression model output estimation. The polynomial chaos expansion coefficient a can be determined by two methods: the embedded method (intrusive method) and the non-embedded method (non-intrusive method). In the process of calculating the polynomial chaos coefficient, the embedded method needs to improve and adjust the original model, while the non-embedded method does not need to change the original model. And f(x) can be obtained directly from the digital main line platform, and there is no need to improve and adjust the control equation of the missile. Therefore, the embodiment of the present application mainly adopts the non-embedded method to solve the polynomial chaos expansion coefficient of f(x).

[0057] Specifically, according to the orthogonal polynomial rule, for any random input, an orthogonal polynomial is constructed. Consider the approximation of f(x) in space. Usually in space The random variables considered above are all square integrable and can be expressed as follows: Where ρ(x) is the probability density function. The inner product and norm in the space are as follows:<u,v> =∫u(x)v(x)ρ(x)dx,||u|| 2 =<u,v> .

[0058] In this way, The subspace V in the matrix finds the best approximation of f(x), that is, the orthogonal projection of f(x) in V, which is also the loss function of the regression model, and the expression is as follows: Among them, v is the predicted flight data output by the regression model, and f is the actual flight data corresponding to the predicted flight data; the dimension P of the multidimensional target space is usually expanded to dimension 4.

[0059] Afterwards, the bionic algorithm is used to iteratively optimize the regression model in the multidimensional target space according to the flight data and the loss function to obtain the trajectory driving model.

[0060] Specifically, Figure 2 As shown, the method of iteratively optimizing the regression model using the bionic algorithm and the flight data to obtain the trajectory driving model may include the following steps:

[0061] S131. Generate multiple groups of candidate solutions of the regression model by using the flight data in the regression model.

[0062] The candidate solutions include the input flight data input into the regression model, the predicted flight data of the regression model, and the actual flight data corresponding to the predicted flight data. The input flight data refers to the flight data input into the regression model; the predicted flight data refers to the data output by the regression model after the flight data is input into the regression model.

[0063] Specifically, firstly, according to the data input and output characteristics of the constructed regression model, the flight data is divided into a plurality of candidate data sets, each candidate data set includes the input flight data of the regression model and the actual flight data.

[0064] For example, the flight data includes x s (t-2), x s (t-1), x s (t), x s (t+1), x s(t+2), and the input of the regression model is the flight data of the first three moments, and the output is the flight data of the next three moments, the above flight data can be divided into the first candidate data set (x s (t-2), x s (t-1), x s (t), x s (t+1)) and the second candidate dataset (x s (t-1), x s (t), x s (t+1), x s (t+2)). The input flight data in the first candidate data set includes x s (t-2), x s (t-1), x s (t), the actual flight data is x s (t+1); the input flight data in the second candidate data set includes x s (t-1), x s (t), x s (t+1), the actual flight data is x s (t+2).

[0065] Secondly, the input flight data in each candidate dataset is input into the regression model to generate the predicted flight data of the candidate dataset.

[0066] Finally, the input flight data of each set of regression models, the predicted flight data corresponding to the input flight data, and the actual flight data corresponding to the predicted flight data are determined as a set of candidate solutions.

[0067] For example, the input flight data x in the first candidate dataset above s (t-2), x s (t-1), x s (t), input into the regression model, and obtain the predicted flight data f(x1); the input flight data x in the second candidate data set is s (t-1), x s (t), x s (t+1), input into the regression model, and the predicted flight data is f(x2). Then the candidate solution corresponding to the first candidate data set is (x s (t-2), x s (t-1), x s (t), f(x1), x s (t+1)); the candidate solution corresponding to the second candidate data set is (x s (t-1), x s (t), x s (t+1), f(x2), xs (t+2)).

[0068] S132. Calculate the fitness of each group of candidate solutions according to the predicted flight data and the actual flight data of each group of candidate solutions.

[0069] The fitness is used to indicate the similarity between the predicted flight data and the actual flight data. The higher the fitness is, the more similar the predicted flight data is to the actual flight data.

[0070] In some embodiments, the fitness of each group of candidate solutions can be calculated by a similarity algorithm, a reverse mean square error algorithm, a reverse absolute error algorithm, a reverse cross entropy algorithm, or the like.

[0071] In some embodiments, the fitness can be calculated based on the value of the loss function in the multi-dimensional target space.

[0072] S133: Determine a candidate solution whose fitness is greater than a fitness threshold as a leader, and determine a candidate solution whose fitness is less than or equal to the fitness threshold as a follower.

[0073] In some embodiments, the fitness threshold is a preset value, for example, a default value, or a value set by relevant personnel according to actual conditions.

[0074] In some embodiments, the fitness threshold may also be a calculated average fitness. Specifically, before determining a candidate solution with a fitness greater than the fitness threshold as a leader, the average fitness of all candidate solutions is calculated based on the predicted flight data and actual flight data of all candidate solutions, and the average fitness is determined as the fitness threshold.

[0075] In the above scheme, when determining the fitness threshold, the average fitness of all candidate solutions is determined as the fitness threshold, so that when the leader and follower are determined later, they can be determined in half, that is, half of the candidate solutions are leaders and the other half are followers. This avoids the local optimal trap and slow convergence speed that the salp optimization algorithm may encounter when dealing with high-dimensional problems as much as possible, and improves the accuracy of determining the trajectory driving model.

[0076] S134 . Use the first movement algorithm to update the candidate solution of the leader, and use the second movement algorithm to update the candidate solution of the follower.

[0077] First, the salp optimization algorithm can be used to update the candidate solutions.

[0078] Specifically, the solution space of the optimization problem can be represented as the foraging space of a group of individuals, and the individuals in the population are divided into two categories: leaders and followers. The leader group independently explores the solution space, simulating the active foraging and self-directed behavior of salps in the natural environment. The followers move according to a strict hierarchical structure, and their movement direction and speed are only affected by the salps immediately in front, forming a chain search mechanism similar to the chain arrangement of salps, and gradually move towards the area where the optimal solution is located. Compared with other swarm intelligence optimization algorithms, the influence of the leader in the salp optimization algorithm is limited and can only directly affect the follower individuals following it. After that, each follower's influence on its successor is transmitted in turn, forming a step-by-step weakening effect, ensuring that the entire salp group maintains moderate diversity during the position update process, avoiding the premature convergence problem that may be caused by the guidance of a single leader.

[0079] Assume that the search space is a D×N Euclidean space, N is the number of salps, and D is the dimension of the search space. The position of the salps in the solution space is represented by X n =[X n1 ,X n2 ,…,X nD ] T , the location of the food is indicated by F n =[F n1 ,F n2 ,…,F nD ] T The upper bound of the search space is μb, and the lower bound is lb. The individual positions of salps are initialized as follows: D×N =rand(D,N)·(μb(D,N)-lb(D,N))+lb(D,N).

[0080] Second, the leader's candidate solution is updated using the first-move algorithm.

[0081] Specifically, according to the formula Update the leader’s candidate solution; where f lead Used to represent the candidate solution of the updated leader, f pos Used to represent the candidate solution of the leader before updating, It is used to represent the attenuation function, c2 is used to represent the moving step size, c3 is used to represent the moving direction, l is used to represent the number of updates, and L is used to represent the maximum number of updates.

[0082] Finally, the candidate solutions of the followers are updated using the second move algorithm.

[0083] Specifically, according to the formula Update the candidate solutions of the followers; where f Folli It is used to represent the candidate solution of the updated follower, fFolli-1 Used to represent candidate solutions for followers before updating.

[0084] S135, recalculate the updated fitness of each candidate solution after the update, and determine whether the preset termination condition is met; when it is determined that the preset termination condition is not met, return to execute steps S133-S135; when it is determined that the preset termination condition is met, execute step S136.

[0085] Among them, the termination condition includes at least one of the following: the number of updates is greater than the iteration threshold; the loss function converges; the maximum update fitness after the update is greater than a preset threshold.

[0086] S136 , using the hyperparameters of the candidate solution corresponding to the maximum fitness in the regression model as the model hyperparameters of the trajectory-driven model, and generating the trajectory-driven model.

[0087] In the above scheme, multiple groups of candidate solutions can be generated according to the regression model first, and then the bionic algorithm is used to update and calculate the multiple groups of candidate solutions to determine the optimal candidate solution of the regression model as the trajectory-driven model. In this way, a bionic algorithm that is more suitable for searching for the optimal solution in multidimensional space is provided, that is, an improved salp optimization algorithm is used to determine the model hyperparameters of the trajectory-driven model, which effectively improves the accuracy of the generated trajectory-driven model. In addition, during the entire search process, the behavior pattern of the salp swarm reflects a unique global and local search strategy: the leader is responsible for a comprehensive search in the solution space, while the follower focuses on the search in the local area. Such a movement strategy ensures that the algorithm effectively prevents falling into the local optimal solution while ensuring the convergence accuracy, so that the salp chain has both powerful global search capabilities and meticulous local optimization capabilities.

[0088] In some embodiments, before constructing the regression model, the missile interception method based on the digital main line also includes: dividing the flight data into a training set, a verification set, and a test set; using the training set to construct and train the regression model; using the verification set to optimize the regression model and generate a trajectory-driven model; using a test machine to verify the trajectory-driven model to ensure the accuracy of the trajectory-driven model.

[0089] The embodiment of the present application can divide the functional modules of the missile interception device based on the digital main line according to the above method example. For example, each functional module can be divided according to each function, or two or more functions can be integrated into one processing unit. The above integrated module can be implemented in the form of hardware or in the form of software functional modules. It should be noted that the division of modules in the embodiment of the present application is schematic and is only a logical functional division. There may be other division methods in actual implementation.

[0090] like Figure 3As shown, it is a structural schematic diagram of a missile interception device based on a digital main line provided in an embodiment of the present application, which is applied to a security operation center. The missile interception device based on a digital main line includes an acquisition module 51, a construction module 52, a generation module 53, and an interception module 54.

[0091] The acquisition module 51 is used to obtain the flight data of the missile collected by the radar of the space satellite at different time points; the construction module 52 is used to construct a regression model according to the time sequence of different time points, using the flight data corresponding to the first time point set as input and the flight data corresponding to the second time point as output; all time points in the first time point set are arranged before the second time point in the time sequence; the generation module 53 is used to iteratively optimize the regression model using the bionic algorithm and the flight data to obtain the trajectory driving model; the interception module 54 is used to intercept the missile according to the prediction result of the trajectory driving model.

[0092] In some embodiments, the construction module 52 is specifically used to take the flight data corresponding to the first time point set as input and the flight data corresponding to the second time point as output, and construct a regression model based on the polynomial chaos expansion algorithm and the finite truncation approximation using the N-order truncation method.

[0093] In some embodiments, the generation module 53 is specifically used to calculate the projection of the regression model in the multidimensional target space according to the hyperparameters of the regression model, and determine the projection as the loss function of the regression model; the multidimensional target space is a square integrable function space with a weight function; and the bionic algorithm is used to iteratively optimize the regression model in the multidimensional target space according to the flight data and the loss function to obtain a trajectory driving model.

[0094] In some embodiments, Figure 4As shown, the generation module 53 includes a generation submodule, a calculation submodule, an update submodule, and a determination submodule; the generation submodule is used to generate multiple groups of candidate solutions of the regression model by using the flight data in the regression model; the candidate solutions include the input flight data of the input regression model, the predicted flight data of the regression model, and the actual flight data corresponding to the predicted flight data; the calculation submodule is used to calculate the fitness of each group of candidate solutions according to the predicted flight data and the actual flight data of each group of candidate solutions; the determination submodule is used to determine the candidate solutions with fitness greater than the fitness threshold as leaders, and the candidate solutions with fitness less than or equal to the fitness threshold as followers; the update submodule is used to The method comprises the following steps: using a first moving algorithm to update the candidate solution of the leader, and using a second moving algorithm to update the candidate solution of the follower; generating a submodule, which is also used to recalculate the updated fitness of each candidate solution after the update, and when it is determined that the preset termination condition has not been met, returning to execute the steps of determining the leader, the follower, updating the candidate solution, and recalculating the fitness; until the termination condition is met, the hyperparameter of the candidate solution corresponding to the maximum fitness in the regression model is used as the model hyperparameter of the trajectory-driven model to generate the trajectory-driven model; wherein the termination condition includes at least one of the following: the number of updates is greater than the iteration threshold; the loss function converges; the maximum updated fitness after the update is greater than the preset threshold.

[0095] In some embodiments, the calculation submodule is also used to calculate the average fitness of all candidate solutions based on the predicted flight data and actual flight data of all candidate solutions before determining the candidate solution with a fitness greater than the fitness threshold as the leader; the determination submodule is also used to determine the average fitness as the fitness threshold.

[0096] In some embodiments, the updating submodule is specifically used to: Update the leader’s candidate solution; where f lead Used to represent the candidate solution of the updated leader, f pos Used to represent the candidate solution of the leader before updating, is used to represent the attenuation function, c2 is used to represent the moving step, c3 is used to represent the moving direction, l is used to represent the number of updates, and L is used to represent the maximum number of updates; according to the formula Update the candidate solutions of the followers; where f Folli It is used to represent the candidate solution of the updated follower, f Folli-1 Used to represent candidate solutions for followers before updating.

[0097] In some embodiments, the flight data includes at least the position state of the missile and the time interval between the current position state and the previous position state, and the position state includes longitude, latitude, speed, heading, and altitude.

[0098] In some embodiments, the missile is a hypersonic missile, and the space satellite is a common space satellite and / or a dedicated space satellite in a missile defense system.

[0099] The missile interception device based on digital main line provided in this embodiment can execute the missile interception method based on digital main line provided in the above method embodiment. Its implementation principle and technical effect are similar to those of the above method and will not be repeated here.

[0100] Figure 5 An electronic device according to an exemplary embodiment may include a processor 802, and the processor 802 is used to execute application code to implement the missile interception method based on digital main line in the present application.

[0101] The processor 802 may be a central processing unit (CPU), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits for controlling the execution of the program of the present application.

[0102] like Figure 5 As shown, the electronic device may further include a memory 803. The memory 803 is used to store application program codes for executing the solution of the present application, and the execution is controlled by the processor 802.

[0103] The memory 803 may be a read-only memory (ROM) or other types of static storage devices that can store static information and instructions, a random access memory (RAM) or other types of dynamic storage devices that can store information and instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical disc, laser disc, optical disc, digital versatile disc, Blu-ray disc, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory 803 may exist independently and be connected to the processor 802 via the bus 804. The memory 803 may also be integrated with the processor 802.

[0104] like Figure 5As shown, the electronic device may further include a communication interface 801, wherein the communication interface 801, the processor 802, and the memory 803 may be coupled to each other, for example, via a bus 804. The communication interface 801 is used to exchange information with other devices, for example, to support information exchange between the electronic device and other devices.

[0105] It should be pointed out that Figure 5 The device structure shown in the figure does not constitute a limitation on the electronic device, except Figure 5 In addition to the components shown, the electronic device may include more or fewer components than shown, or combine certain components, or arrange the components differently. The electronic device provided in this embodiment can execute the missile interception method based on the digital main line provided in the above method embodiment, and its implementation principle and technical effect are similar to the above method, which will not be repeated here.

[0106] An embodiment of the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, each process of the missile interception method based on a digital main line in the above method embodiment is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0107] The computer-readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0108] An embodiment of the present application provides a computer program product, which stores a computer program. When the computer program is executed by a processor, each process of the missile interception method based on the digital main line in the above method embodiment is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0109] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media that include computer-usable program code.

[0110] In the present application, the processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0111] In this application, memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

[0112] In this application, computer-readable media includes permanent and non-permanent, removable and non-removable storage media. Storage media can be implemented by any method or technology to store information, and the information can be computer-readable instructions, data structures, modules of programs or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, read-only compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. According to the definition in this article, computer-readable media does not include temporary computer-readable media (transitory media), such as modulated data and carrier waves.

[0113] It should be noted that, in this article, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0114] The above description is only a specific implementation of the present application, so that those skilled in the art can understand or implement the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments described herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A missile interception method based on digital main line, characterized in that: include: Obtain flight data of missiles at different time points collected by space satellite radars; According to the time sequence of the different time points, the flight data corresponding to the first time point set is used as input, and the flight data corresponding to the second time point is used as output to construct a regression model; all time points in the first time point set are arranged before the second time point in the time sequence; Iteratively optimizing the regression model using a bionic algorithm and the flight data to obtain a trajectory driving model; The missile is intercepted according to the prediction result of the trajectory driving model.

2. The interception method according to claim 1, characterized in that: The step of constructing a regression model by taking the flight data corresponding to the first time point set as input and the flight data corresponding to the second time point set as output according to the time sequence of the different time points includes: The flight data corresponding to the first time point set is used as input, and the flight data corresponding to the second time point is used as output. Based on the polynomial chaos expansion algorithm, a finite truncation approximation is performed using an N-order truncation method to construct a regression model.

3. The interception method according to claim 1, characterized in that: The iterative optimization of the regression model using the bionic algorithm and the flight data to obtain a trajectory driving model includes: Calculating the projection of the regression model in a multidimensional target space according to the hyperparameters of the regression model, and determining the projection as the loss function of the regression model; the multidimensional target space is a square integrable function space with a weight function; The regression model is iteratively optimized in the multi-dimensional target space according to the flight data and the loss function using a bionic algorithm to obtain a trajectory driving model.

4. The interception method according to claim 3, characterized in that: The regression model is iteratively optimized in the multi-dimensional target space according to the flight data and the loss function using a bionic algorithm to obtain a trajectory driving model, including: Using the flight data in the regression model, multiple groups of candidate solutions of the regression model are generated; the candidate solutions include input flight data input into the regression model, predicted flight data of the regression model, and actual flight data corresponding to the predicted flight data; Calculating the fitness of each group of candidate solutions according to the predicted flight data and the actual flight data of each group of candidate solutions; the fitness is determined according to the value of the loss function in the multidimensional target space; Determine the candidate solution whose fitness is greater than the fitness threshold as the leader, and determine the candidate solution whose fitness is less than or equal to the fitness threshold as the follower; updating the candidate solution of the leader using a first movement algorithm, and updating the candidate solution of the follower using a second movement algorithm; Recalculate the updated fitness of each candidate solution after the update, and when it is determined that the preset termination condition is not met, return to the steps of determining the leader, the follower, updating the candidate solution, and recalculating the fitness; until the termination condition is met, use the hyperparameters of the candidate solution corresponding to the maximum fitness in the regression model as the model hyperparameters of the trajectory-driven model to generate the trajectory-driven model; wherein the termination condition includes at least one of the following: the number of updates is greater than the iteration threshold; the loss function converges; the maximum updated fitness after the update is greater than the preset threshold.

5. The interception method according to claim 4, characterized in that: Before determining the candidate solution having a fitness greater than a fitness threshold as a leader, the method further includes: Calculate the average fitness of all candidate solutions based on their predicted flight data and actual flight data; The average fitness is determined as a fitness threshold.

6. The interception method according to claim 4, characterized in that: The method of updating the candidate solution of the leader using the first movement algorithm comprises: According to the formula Updating the candidate solution of the leader; Among them, f lead It is used to represent the candidate solution of the leader after the update, f pos It is used to represent the candidate solution of the leader before updating. It is used to represent the attenuation function, c2 is used to represent the moving step size, c3 is used to represent the moving direction, l is used to represent the number of updates, and L is used to represent the maximum number of updates; The method of updating the candidate solution of the follower by using the second movement algorithm comprises: According to the formula Update the candidate solution of the follower; where f Folli It is used to represent the candidate solution of the follower after the update, f Folli-1 Used to represent the candidate solution of the follower before updating.

7. The interception method according to any one of claims 1 to 5, characterized in that: The flight data at least includes the position state of the missile and the time interval between the current position state and the previous position state. The position state includes longitude, latitude, speed, heading, and altitude.

8. The interception method according to any one of claims 1 to 5, characterized in that: The missile is a hypersonic missile, and the space satellite is an ordinary space satellite and / or a special space satellite in a missile defense system.

9. A missile interception device based on a digital main line, characterized in that: include: An acquisition module is used to obtain the flight data of the missile at different time points collected by the radar of the space satellite; a construction module, configured to construct a regression model according to the time sequence of the different time points, using the flight data corresponding to the first time point set as input and the flight data corresponding to the second time point as output; all time points in the first time point set are arranged before the second time point in the time sequence; A generation module, used for iteratively optimizing the regression model using a bionic algorithm and the flight data to obtain a trajectory driving model; An interception module is used to intercept the missile according to the prediction result of the trajectory driving model.

10. An electronic device, characterized in that: include: A processor, a memory, and a computer program stored in the memory and executable on the processor, wherein when the computer program is executed by the processor, the digital main line-based missile interception method according to any one of claims 1 to 7 is implemented.

11. A computer-readable storage medium, characterized in that: include: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the digital main line-based missile interception method according to any one of claims 1 to 8 is implemented.

12. A computer program product, characterized in that When the computer program product is executed on a computer, the computer is enabled to implement the missile interception method based on a digital main line according to any one of claims 1 to 8.

Images