A method and device for generating an anti-aircraft shell arrangement

Abstract

The application discloses a kind of generation method and device of air defense shell arrangement scheme, method includes: real-time acquisition of the detection data of incoming target, and pre-processing, generate pre-processing data;According to the trajectory of pre-generated incoming target according to pre-processing data, and the trajectory of pre-generated incoming target is corrected, and the trajectory of corrected incoming target is obtained;According to the trajectory of corrected incoming target, air defense shell arrangement planning is carried out in the square coverage area that incoming target can appear, and the air defense shell arrangement scheme is obtained;Particle swarm optimization algorithm is used to optimize air defense shell arrangement scheme, and the optimized air defense shell arrangement scheme is obtained, for intercepting incoming target.The application improves the response speed of traditional air defense system when coping with dynamic threat, and improves the interception accuracy.

Description

Technical Field

[0001] The present invention belongs to the technical field of anti-aircraft artillery shell arrangement optimization, and in particular relates to a method and device for generating an anti-aircraft artillery shell arrangement plan. Background Art

[0002] In modern warfare, battlefield environments are complex and ever-changing, and early warning systems often face the challenge of being unable to accurately track the trajectory of incoming targets. In such situations, early warning systems can only provide the range of possible target locations, but cannot pinpoint their specific locations. This places higher demands on the responsiveness of air defense systems, especially when a single air defense projectile, due to limited mobility, cannot independently cover all potential threat areas. This poses an even greater challenge to air defense interception. To effectively counter the threat of incoming targets, a coordinated interception strategy involving multiple air defense projectiles is necessary. By carefully designing the launch locations, flight paths, and attack timing of multiple air defense projectiles, they can form intersecting interception zones in space. This coordinated design strategy can expand overall air defense coverage, ensuring that targets can be intercepted promptly regardless of their location. Achieving this coordinated interception requires comprehensive consideration of multiple factors, including the projectile's speed, maneuverability, trajectory characteristics, and the target's expected motion pattern. Furthermore, advanced command and control systems and communication networks are crucial for successful coordinated interception, ensuring high-level coordination and information sharing among multiple air defense projectiles. Through this multi-level, multi-dimensional coordinated interception approach, air defense systems can more effectively respond to complex threats and increase the probability of destroying incoming targets. This paper will explore in detail the specific technical methods and tactical applications of coordinated interception of a single incoming target by multiple air defense projectiles, as well as their practical application and development prospects in modern air defense operations.

[0003] In existing technologies, area coverage optimization primarily focuses on the positioning and deployment of combat units. Specifically, each combat unit moves to the optimal position based on a specific algorithm and ultimately settles there to achieve optimal perception and coverage of the mission area. In current research, the coverage capabilities of combat units are generally considered independent of each other; the operating schedule of one combat unit does not affect the coverage capabilities of other units. Furthermore, the coverage capabilities of combat units are assumed to be constant over time. However, in certain coverage missions, the coverage capabilities of combat units or the covered mission area may change over time. Furthermore, information transfer can improve the coverage capabilities of some combat units. Therefore, it is necessary to consider the mutual influence between combat units when scheduling work schedules, which in turn drives the optimization of the number of combat units to improve overall coverage efficiency. Summary of the Invention

[0004] In order to solve the problems in the prior art, the application provides a generation method and device of an anti-aircraft shell arrangement scheme, which improves the response speed of a traditional anti-aircraft system when coping with dynamic threats and improves the interception accuracy.

[0005] To achieve the above object, the application adopts the technical scheme of:

[0006] In the first aspect, the application provides a generation method of an anti-aircraft shell arrangement scheme, which comprises the following steps: collecting detection data of an incoming target in real time and preprocessing the detection data to generate preprocessed data; pre-generating a trajectory of the incoming target according to the preprocessed data and correcting the pre-generated trajectory of the incoming target to obtain a corrected trajectory of the incoming target; performing anti-aircraft shell arrangement planning in a square coverage area where the incoming target may appear according to the corrected trajectory of the incoming target to obtain an anti-aircraft shell arrangement scheme; and optimizing the anti-aircraft shell arrangement scheme by using a particle swarm optimization algorithm to obtain an optimized anti-aircraft shell arrangement scheme for intercepting the incoming target.

[0007] Further, the real-time collection of the detection data of the incoming target and the preprocessing of the detection data comprise the following steps: synchronizing and time-aligning different types of detection data by using a timestamp technology to obtain intermediate data; and denoising, cleaning and data fusing the intermediate data to obtain the preprocessed data.

[0008] Further, the pre-generation of the trajectory of the incoming target according to the preprocessed data comprises the following steps: obtaining state estimation of the incoming target according to the preprocessed data and establishing a motion model of the incoming target; and obtaining a mean value and a covariance matrix of the motion state of the incoming target according to the motion model of the incoming target to pre-generate the trajectory of the incoming target.

[0009] Further, the correction of the pre-generated trajectory of the incoming target comprises the following steps:

[0010] Error model construction:

[0011] The error model comprises a system error and a measurement error, the error model is assumed to be Gaussian distribution, the system error and the measurement error are irrelevant to each other, and satisfy

[0012] , ,

[0013] wherein, represents a system noise, represents a measurement noise, represents a system noise covariance matrix, is a measurement noise covariance matrix, represents a Gaussian distribution with a mean value of 0 and a variance of represents a Gaussian distribution with a mean value of 0 and a variance of ​Gaussian distribution, the system's motion equation is expressed as

[0014] ,

[0015] in, represents the state vector of the system at time k, represents the state transition matrix, express The control input at the moment, represents the input matrix, express System noise at each moment;

[0016] The measurement model equation is

[0017] ,

[0018] in, represents the measurement value of the sensor at time k, represents the observation matrix, represents the measurement noise at time k;

[0019] Incoming target status prediction:

[0020] According to the motion model of the incoming target, the state of the incoming target at time k Prediction

[0021] ,

[0022] in, represents the state prediction value at time k, is the estimated value of the state based on the previous moment k-1, It's the last moment Control input;

[0023] Forecast error covariance matrix

[0024] ,

[0025] in, Indicates at time The error covariance matrix of the predicted state represents the magnitude of the prediction error, which is based on The estimated and observed data of the time are obtained; Indicates based on The observation data at the time The error covariance matrix at time t;

[0026] Measurement information update:

[0027] The measurement value collected by the sensor at time k and predicted value The difference between Defined as

[0028] ,

[0029] Its covariance matrix is

[0030] ,

[0031] in, It reflects the error between actual measurement and prediction and is used to correct state estimation;

[0032] Kalman gain calculation:

[0033] Kalman gain Determines the weight distribution of measured values ​​and predicted values ​​in state updates, and its calculation formula is:

[0034] ,

[0035] The larger the Kalman gain, the more the system relies on the current measurement value; the smaller the Kalman gain, the more the system relies on the predicted value;

[0036] Status Update:

[0037] The predicted state is corrected using the Kalman gain, and the updated state estimate is

[0038] ,

[0039] At the same time, the covariance matrix is ​​updated:

[0040] ,

[0041] in, Indicates based on The observation data at the time The error covariance matrix at time , Indicates based on The observation data at the time The error covariance matrix at time t;

[0042] Through this state update, the trajectory error of the incoming target is corrected, making the state estimation closer to the actual trajectory of the incoming target.

[0043] Furthermore, according to the corrected trajectory of the incoming target, the air defense shells are deployed in the square coverage area where the incoming target may appear, including: defining the center coordinates of the square coverage area where the incoming target may appear as , with a side length of L; divide the square coverage area into several grid cells, and calculate the weight of each grid cell based on the probability that the incoming target may appear in a grid cell and the area of ​​the grid cell; evenly distribute the shells in the square area, calculate the spacing of the shells to ensure that every grid cell in the square coverage area is covered, and determine the launch position of each shell using the following formula:

[0044] ,

[0045] in, Indicates the launch position of the shell, p is the index of the shell, M is the number of shells in each row, 、 The horizontal and vertical spacing of the shells, where the horizontal and vertical spacing refer to the horizontal and vertical distances between adjacent shells on the plane; the coverage effect of a grid cell is determined by calculating the distance from the center point of the grid cell to the shell launch position

[0046] ,

[0047] in, For the The center coordinates of the grid cells, is the launch position of the p-th shell, represents the coverage effect of the grid cell, The coverage effect of the grid cell refers to whether each grid cell is completely or partially covered by the damage radius of the shell in the block grid cells. The target area is decomposed into multiple small areas through grid division to evaluate and optimize the coverage effect of the shell. Finally, the total coverage rate of the entire square area is calculated to select the anti-aircraft shell layout plan.

[0048]

[0049] in, represents the total coverage of the entire square area, represents the weight of the grid cell, It represents the sum of the weights of all shells and is used to measure whether the coverage area meets the requirements.

[0050] Furthermore, the weight of each grid cell is calculated by the following formula:

[0051] ,

[0052] in, represents the probability that the target is within the grid, Represents the area of ​​a grid cell.

[0053] Furthermore, the particle swarm optimization algorithm is used to optimize the anti-aircraft artillery shell deployment scheme, including: defining the basic parameters of the particle swarm and setting the objective function :

[0054] ,

[0055] ,

[0056] in, Indicates the total number of grids;

[0057] Based on the set objective function, particle swarm optimization is performed, including: initializing the particle swarm, randomly generating the position and velocity of the initial particle, the position of the particle represents the launch position of the projectile, and the velocity represents the change in position, then updating the velocity and position of each particle according to the formula of the particle swarm optimization algorithm; for each particle, checking whether the current position is better than its historical best position, if so, updating the best position of particle i ,

[0058] ,

[0059] Update the global best position ,

[0060] ,

[0061] If the maximum number of iterations is reached Or if the change of the objective function is less than the predetermined threshold, the iteration is stopped; after the iteration is completed, the position of the optimal particle This is the optimized anti-aircraft artillery shell layout plan.

[0062] Furthermore, define the triggering conditions for trajectory changes , Indicates time The change in the target position is used to trigger the dynamic optimization process.

[0063] ,

[0064] When the trajectory changes Exceeding the preset threshold Dynamic optimization is triggered when , regenerate the shell layout plan.

[0065] Furthermore, when dynamic optimization is triggered, the objective function is first updated, assuming that the area of ​​the incoming target is represented by a rectangular grid. Indicates that the new optimization goal is to re-cover the area, and the objective function is expressed as:

[0066] ,

[0067] in, Represents particles In the new grid coverage situation, in order to speed up the optimization, the particle speed update formula is adjusted to:

[0068] ,

[0069] By adjusting the inertia weight , learning factor and , achieving rapid convergence of the particle swarm, thus generating a new optimal layout plan in a relatively short time;

[0070] After generating a new optimal layout plan, update the artillery launch instructions, where the artillery launch angle Corrected to,

[0071] ,

[0072] in, Indicates the coordinates of the incoming target, Indicates the coordinates of the launcher.

[0073] In a second aspect, a device for generating an air defense artillery shell deployment plan is provided, comprising: a data acquisition module for acquiring detection data of an incoming target in real time, and performing preprocessing to generate preprocessed data; a trajectory generation module for pregenerating a trajectory of the incoming target based on the preprocessed data, and correcting the pregenerated trajectory of the incoming target to obtain a corrected trajectory of the incoming target; a plan generation module for planning the deployment of air defense artillery shells within a square coverage area where the incoming target may appear based on the corrected trajectory of the incoming target, to obtain an air defense artillery shell deployment plan; and a plan optimization module for optimizing the air defense artillery shell deployment plan using a particle swarm optimization algorithm to obtain an optimized air defense artillery shell deployment plan for intercepting the incoming target.

[0074] Compared with the prior art, the present invention has the following beneficial effects:

[0075] (1) The fixed deployment schemes in the prior art are usually unable to effectively cope with the maneuvering or trajectory changes of the target. However, the present invention can adjust the projectile deployment scheme in real time according to the dynamic changes of the target trajectory. The deployment is continuously optimized through the particle swarm algorithm to ensure the interception success rate. Even if the target changes during flight, the coverage area of ​​the projectile can be quickly adjusted.

[0076] (2) Compared with traditional artillery shell placement algorithms, the present invention uses a particle swarm optimization (PSO) algorithm for optimization. The PSO algorithm, with its efficient global search capability, can quickly approximate the optimal solution in complex, multi-objective optimization problems. Therefore, the present invention can generate the optimal artillery shell placement scheme more quickly and efficiently, maximize the coverage of the target area, and improve the system's response speed and interception accuracy.

[0077] (3) Traditional air defense systems are difficult to respond efficiently when faced with changing air threats. The present invention can not only respond to changes in a single target, but also rationally plan the deployment of artillery shells through rapid iteration of algorithms in a multi-target, multi-threat environment, so that the system can maintain efficient defense capabilities in a complex combat environment and adapt to complex and changing threat environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] Figure 1 Schematic diagram of the implementation steps of a method for generating an air defense artillery shell deployment plan provided by an embodiment of the present invention;

[0079] Figure 2 It is the artillery shell layout plan diagram within the square coverage area before the anti-aircraft artillery shell layout plan is optimized;

[0080] Figure 3 It is the artillery shell layout plan within the square coverage area after the anti-aircraft artillery shell layout plan is optimized. DETAILED DESCRIPTION

[0081] The present invention will be further described below in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention and are not intended to limit the scope of protection of the present invention.

[0082] Example 1

[0083] like Figure 1 As shown, a method for generating an air defense artillery shell deployment plan includes: collecting detection data of an incoming target in real time and preprocessing it to generate preprocessed data; pregenerating a trajectory of the incoming target based on the preprocessed data, and correcting the pregenerated trajectory of the incoming target to obtain a corrected trajectory of the incoming target; planning the deployment of air defense artillery shells within a square coverage area where the incoming target may appear based on the corrected trajectory of the incoming target to obtain an air defense artillery shell deployment plan; and optimizing the air defense artillery shell deployment plan using a particle swarm optimization algorithm to obtain an optimized air defense artillery shell deployment plan for intercepting the incoming target.

[0084] S1. Collect detection data of incoming targets in real time, perform preprocessing, and generate preprocessed data.

[0085] S11. Detection data collection

[0086] In actual air defense systems, a variety of sensor devices, such as radar, infrared detectors, optoelectronic detection equipment, satellite monitoring systems, or drone reconnaissance equipment, are used to monitor aerial targets in real time. Different devices have different characteristics. For example, radar can provide precise distance and speed information, infrared equipment can capture the target's heat signature, optoelectronic equipment can provide high-resolution visual imagery, and satellite monitoring and drone reconnaissance can cover a wide area. These multiple sources of equipment work together to effectively reduce the limitations of a single sensor and ensure the integrity and accuracy of incoming target data.

[0087] To better capture data, multi-source sensor fusion technology can be employed. In this technique, multiple sensors operate simultaneously and transmit their data to a central processing system. Through fusion, the data from these multiple sensors can be combined to obtain a more comprehensive image and information about the target.

[0088] S12. Data synchronization and time alignment

[0089] Due to the differences in sensor types, data refresh rates, latency, and acquisition cycles may vary. Therefore, before processing this data, all data must be synchronized and time-aligned. This invention uses timestamp technology to add precise time information to each set of data, ensuring that data from each sensor is consistent at the same point in time.

[0090] The key to time synchronization is ensuring that all sensor systems use a unified clock signal. For example, using the Global Positioning System (GPS) to synchronize the time across all devices can prevent data bias caused by time differences. This approach allows air defense systems to obtain a complete picture of the target at the same point in time, ensuring accurate subsequent analysis.

[0091] S13. Data denoising and cleaning

[0092] Due to the complex detection environment, sensor equipment may be affected by meteorological conditions, electromagnetic interference, or other external noise, resulting in noise or outliers in the data. Therefore, the collected raw data must be denoised and cleaned. Common denoising techniques include low-pass filters and Kalman filters.

[0093] Low-pass filters can effectively remove high-frequency noise, filtering out sudden anomalies in detection data while preserving the target's smooth trajectory. Kalman filters, a classic data processing technique, can optimize estimates of target position, velocity, and other information across multiple measurements, reducing the impact of random errors on the data.

[0094] Through denoising and cleaning processing, more accurate and stable target data can be obtained, thereby providing high-quality data input for subsequent prediction and decision-making.

[0095] S14, Data Fusion and Model Establishment

[0096] After completing data cleaning and normalization, the next step is data fusion. Using data fusion algorithms such as weighted average method or Bayesian estimation, the system can integrate the data of various sensors to generate a more accurate and reliable target state estimation. For example, radar data provides accurate distance information, and infrared data provides thermal signal characteristics. By fusing these data, the system can generate complete information of the target, including position, speed, direction, and characteristic attributes.

[0097] Using the radar and other sensor data obtained by the above steps to obtain real-time position, speed and acceleration data of the target, and then by analyzing the historical motion trajectory and known characteristics (such as speed, acceleration, maneuverability) of the target, the motion model of the target is established. Assuming that the target has the following motion model

[0098]

[0099] wherein, represents the motion state of the incoming target, represents its derivative, represents the control input of the incoming target, represents the measurement information of the detection system, represents a nonlinear function.

[0100] When the motion model of the incoming target presents nonlinear characteristics, its motion equation may not be solved by analytical method. At this time, numerical calculation method can be used to estimate the position of the target. Specifically, the motion model of the target can be discretized, and the discrete time step is set as , thus the discrete motion model of the incoming target is obtained:

[0101]

[0102] wherein, , respectively represent the state of the target at discrete time and , represents the dynamic function of the target, which is used to describe the change rule of the state over time, represents the control input of the target at time.

[0103] S15, Obtain the Mean and Covariance Matrix of the Current Motion State of the Incoming Target

[0104] Assume that the input of the incoming target and the motion state at the initial moment obey Gaussian distribution, where the mean of the control input is , the variance is , the mean motion state at the initial moment is , the covariance matrix is In this setting, let the control input be , through iterative calculation of the discrete motion model, the predicted hit time can be obtained The mean of the target motion state at each moment . The covariance matrix of the target motion state at the moment It can be calculated by iterative calculation using the following equation

[0105]

[0106] in, , Denote the state transfer matrix and input control matrix, respectively, reflecting the dynamic changes of the system within discrete time steps. Through this iterative approach, the covariance matrix of the target can be continuously updated and predicted, thus providing an accurate state estimate at the time of impact.

[0107] S2. Pre-generate a trajectory of the incoming target according to the pre-processed data, and correct the pre-generated trajectory of the incoming target to obtain a corrected trajectory of the incoming target.

[0108] S21. Error model construction

[0109] The construction of the error model is the basis for trajectory correction. The following two types of errors are usually used:

[0110] (1) System error: caused by the error of the system itself (such as detection equipment), usually expressed as the system noise covariance matrix express;

[0111] (2) Measurement error: The random error generated during the sensor measurement process is expressed as the measurement noise covariance matrix express.

[0112] The error model usually assumes that the noise is Gaussian distributed and the system noise and measurement noise are uncorrelated, satisfying:

[0113] ,

[0114] in, represents the system noise, represents the measurement noise, The mean is 0 and the variance is Gaussian distribution, The mean is 0 and the variance is Gaussian distribution.

[0115] The system's equation of motion is usually expressed as:

[0116]

[0117] in, represents the state vector (such as speed, position) of the system at time k, represents the state transition matrix, express The control input at the moment, represents the input matrix, express System noise at any moment.

[0118] The measurement model equation is:

[0119]

[0120] in, yes The sensor's measurement value at that moment, is the observation matrix, yes The measurement noise at the moment.

[0121] S22. Incoming target status prediction

[0122] According to the motion model, the state of the target at time k The prediction is:

[0123]

[0124] in, represents the state prediction value at time k, is the estimated value of the state based on the previous moment k-1, It's the last moment control input.

[0125] Forecast error covariance matrix:

[0126]

[0127] in, Indicates at time The error covariance matrix of the predicted state represents the magnitude of the prediction error, which is based on The estimated time and the observed data are obtained, Indicates based on The observation data at the time The error covariance matrix at time t.

[0128] S23, measurement information update

[0129] The measurement value collected by the sensor at time k and predicted value The difference between is defined as:

[0130]

[0131] Its covariance matrix is:

[0132]

[0133] in, It reflects the error between actual measurement and prediction and is used to correct state estimation.

[0134] S24, Kalman gain calculation

[0135] Kalman gain Determines the weight distribution of measured values ​​and predicted values ​​in state updates, and its calculation formula is:

[0136]

[0137] The larger the Kalman gain, the more the system relies on the current measurement value; the smaller the gain, the more the system relies on the predicted value.

[0138] S25. Status Update

[0139] The predicted state is corrected using the Kalman gain, and the updated state estimate is

[0140]

[0141] At the same time, the covariance matrix is ​​updated:

[0142]

[0143] in, Indicates based on The observation data at the time The error covariance matrix at time , Indicates based on The observation data at the time The error covariance matrix at time t;

[0144] Through this state update, the error of the incoming target trajectory can be effectively corrected, making the state estimation closer to the actual trajectory of the incoming target.

[0145] S26. Obtain the area where the incoming target may appear

[0146] According to the update method in S21-S25, the mean and covariance matrix mentioned in S1 are updated to obtain the predicted hit area at the time of prediction hit.

[0147]

[0148] in, The mean is , the variance is Gaussian distribution.

[0149] S3. Based on the corrected trajectory of the incoming target, air defense artillery shells are deployed within the square coverage area where the incoming target may appear, to obtain an air defense artillery shells deployment plan.

[0150] S31. Define a square coverage area;

[0151] First determine the parameters of the square area: the center position is the center coordinate of the square area, the side length L is the side length of the square area, and the corner coordinates of the square area can be calculated as:

[0152]

[0153] The square coverage area is then divided into grid cells so that the coverage performance of each grid cell can be evaluated. The gridding steps are as follows:

[0154] Select the grid size: Select the appropriate grid size d, which is determined based on the effective coverage radius of the projectile.

[0155] Generate grid: Generate several grids of size The coordinate range of the grid cell is

[0156]

[0157] Where i and j are the row and column indices of the grid cell.

[0158] Finally, the weight of the grid cell is calculated. The weight is calculated for each grid cell. The weight can be based on factors such as the probability of the target appearing in the grid and the area of ​​the grid cell:

[0159]

[0160] in For the The weight of the grid cells, is the probability that the target is within the grid, is the area of ​​the grid cell.

[0161] S32. Develop a shell deployment plan

[0162] First, set the layout constraints. The projectile range G represents the effective range of each projectile. The layout quantity It represents the number of available shells, and the overlapping effect represents the overlapping effect of the shell's detonation coverage area.

[0163] Then generate the placement plan. Generate the shell placement plan based on the gridded coverage area data. The following methods can be used:

[0164] (1) Uniform distribution: The shells are evenly distributed in the square area. The spacing of the shells is calculated to ensure that every grid cell in the coverage area is covered. The launch position of each shell can be determined by the following formula:

[0165]

[0166] Among them, p is the index of the shell, M is the number of shells in each row, 、 The horizontal and vertical spacing of the shells, where the horizontal and vertical spacing refer to the horizontal and vertical distances between adjacent shells on the plane;

[0167] (2) Coverage evaluation: Evaluate the coverage effect of each layout scheme. The coverage effect can be evaluated by calculating the coverage degree of each grid cell. Assuming that the detonation area of ​​the shell is a circle with a radius of R, then a grid cell The coverage effect can be judged by calculating the distance from the center of the grid cell to the launch position of the projectile:

[0168]

[0169] in, For the The center coordinates of the grid cells, is the launch position of the pth shell, represents the coverage effect of the grid cell, Represents the radius of the shell's detonation zone. The coverage of a grid cell typically refers to whether each cell in the grid is fully or partially covered by the shell's coverage area (i.e., damage radius). Gridding is used to break down a larger target area into smaller areas, making it easier to evaluate and optimize shell coverage.

[0170] Finally, the air defense artillery shell layout plan is selected by calculating the total coverage of the entire square area.

[0171]

[0172] in, represents the total coverage of the entire square area, represents the weight of the grid cell, It represents the sum of the weights of all shells, which can be used to measure whether the effect on the coverage area meets the requirements.

[0173] Finally, the selected optimal solution is verified and adjusted to ensure its effectiveness in actual application. The effectiveness of the solution can be verified through simulation tests or actual drills.

[0174] S4. Use the particle swarm optimization algorithm to optimize the air defense artillery shell deployment plan to obtain the optimized air defense artillery shell deployment plan for intercepting incoming targets.

[0175] S41. Define the parameters of the particle swarm algorithm

[0176] The basic parameters of the particle swarm are as follows:

[0177] Number of particles : The number of particles in the particle swarm. The number of particles affects the comprehensiveness of the search and the computational complexity. Maximum number of iterations : Maximum number of iterations of the particle swarm algorithm.

[0178] Inertia Weight : Controls the degree to which the particle's current velocity affects its position.

[0179] Learning Factor : They are the learning factors of particle self-cognition and global cognition, respectively, used to adjust the update of particle position.

[0180] Then define the position and velocity of the particle:

[0181] Location : The current position of the i-th particle, where d is the dimension (here the launch position of the cannonball).

[0182] speed : The velocity of the ith particle.

[0183] S42. Setting the objective function

[0184] First, define the objective function, which is used to evaluate the quality of each particle placement scheme. For the shell placement, the objective function can be defined as the sum of the coverage effects or the minimization of the error.

[0185] The coverage effect is the total coverage effect of the shells in the area, and the calculation formula is as follows:

[0186]

[0187] in, is the coverage effect of the i-th particle arrangement scheme on the j-th grid cell (1 means coverage, 0 means no coverage), Indicates the total number of grids.

[0188] The objective function f is defined as follows

[0189]

[0190] The goal is to maximize coverage.

[0191] S43. Execute particle swarm optimization

[0192] The first step is to initialize the particle swarm and randomly generate the position and velocity of the initial particles. Indicates the launch position and speed of the projectile Indicates the change in position.

[0193]

[0194]

[0195] in, To generate random positions or parameter values ​​within a specified range, To generate a random speed value within a given speed range

[0196] Then, update the position and velocity of the particle. Update the velocity and position of each particle according to the formula of the particle swarm optimization algorithm. The velocity update formula is:

[0197]

[0198] in, is a random number, is the best historical position of particle i, is the global best position.

[0199] The position update formula is

[0200]

[0201] in, represents the position of the i-th particle at time t+1, represents the velocity of the i-th particle at time t+1, represents the position of the i-th particle at time t.

[0202] Next, update the optimal position of the particle. For each particle, check whether the current position is better than its historical best position. If it is better, update the optimal position of particle i:

[0203]

[0204] Update the global best position

[0205]

[0206] in, represents the position of the i-th particle at time t+1, and f is an objective function used to calculate the specific variables (such as ) related values ​​to evaluate the pros and cons of a certain solution

[0207] If the maximum number of iterations is reached Or if the change of the objective function is less than a predetermined threshold, the iteration is stopped.

[0208] S44. Generate the optimal shell placement plan

[0209] First, the optimal arrangement solution is obtained. After the iteration, the position of the optimal particle This is the optimal shell placement scheme, which can achieve the best coverage effect in the target area.

[0210] Then formulate an implementation plan and formulate specific artillery firing instructions based on the optimal layout plan, including the launch position, launch time and direction of each artillery shell.

[0211] Finally, the layout plan is executed and verified in the actual environment to ensure that the generated layout plan can achieve the expected coverage effect.

[0212] S45. Set dynamic optimization trigger conditions

[0213] When the target trajectory deviates significantly, the system needs to make timely adjustments. In order to trigger the dynamic optimization process, the trigger conditions for trajectory changes are defined. , indicating that in a short time Change in internal target position:

[0214]

[0215] When the trajectory changes Exceeding the preset threshold When , the system will trigger dynamic optimization, namely:

[0216]

[0217] This trigger condition is used to determine whether it is necessary to regenerate the shell deployment plan.

[0218] In order to improve the response speed, dynamic optimization is performed in a fixed time window. The target trajectory is monitored internally. Whenever a trajectory change exceeding a threshold is detected, the system will immediately start the optimization process.

[0219] S46, Real-time re-optimization based on particle swarm optimization

[0220] When the target trajectory changes, the particle swarm algorithm is re-invoked to adjust the placement scheme. Each particle represents a possible shell placement method, and the optimization goal is to maximize the coverage of all shells and overlap them with the target area.

[0221] First, update the objective function. During the dynamic optimization process, the objective function needs to be adjusted according to the latest incoming target position and motion state. Assume that the area of ​​the incoming target is represented by a rectangular grid. Indicates that the new optimization goal is to re-cover the area. The objective function can be expressed as:

[0222]

[0223] in, Represents particles Performance under new grid coverage conditions. During each iteration, the algorithm gradually optimizes the placement of the projectiles based on the coverage of the particles within the grid.

[0224] Then, adjust the particle speed. In order to speed up the optimization process, you can adjust the particle speed update formula. The particle speed update formula can be adjusted as follows:

[0225]

[0226] By adjusting the parameters (Inertia weight), (learning factor), which can achieve rapid convergence of the particle swarm and thus generate a new optimal layout plan in a shorter time.

[0227] S47, Real-time adjustment of shell placement plan

[0228] First, the firing instructions are updated. After obtaining the optimized shell placement plan, the firing instructions for each launcher need to be updated immediately, including parameters such as launch angle, time, and location. The shells must be fired synchronously to ensure the maximum coverage area.

[0229] Coordination and execution are then carried out. Coordination between multiple launchers is very important, especially in dynamic optimization, which requires synchronous adjustment based on the positions of different launchers and changes in targets to avoid repeated coverage or missed areas.

[0230] Finally, the parameters are adjusted. The key parameters of artillery launch include the launch angle , initial velocity and explosion time According to the optimized layout plan, these parameters will be updated in real time during each adjustment. The calculation of the emission angle can be corrected by the following formula:

[0231]

[0232] Correct the launch angle according to the trajectory of the incoming target to ensure that the shell explodes accurately in the target area. Indicates the coordinates of the incoming target, Indicates the coordinates of the launcher.

[0233] The present invention first pre-processes the target trajectory by integrating real-time detection data from multiple sensors to improve data accuracy and reliability. Data pre-processing involves noise filtering and error correction on the raw detection data to obtain a precise target trajectory. Next, the incoming target's trajectory is corrected in real time based on a target error model to ensure the accuracy of the target motion data. Secondly, based on the target's real-time position and motion state, a potential threat zone is generated and projectile placement is planned within this zone. Projectile placement is optimized using a particle swarm algorithm. By simulating the movement of multiple particles and iteratively updating, the optimal projectile placement configuration is found to ensure maximum coverage of the threat zone and improve interception success rate. During this process, the system fully considers changes in the target's trajectory and threat zone to provide a precise projectile placement plan. Finally, the system incorporates real-time monitoring and dynamic optimization capabilities, enabling rapid adjustment of the projectile placement plan when the target trajectory changes. By continuously monitoring real-time data, changes in target position can be detected promptly and projectile placement can be dynamically optimized based on the new data, thereby improving the system's responsiveness and interception accuracy. Simulation experiments show that the method designed in the present invention can effectively improve the combat effectiveness of the air defense system in a complex threat environment, and provide a more flexible and accurate solution for modern air defense systems.

[0234] The present invention uses MATLAB 2022b as simulation software to simulate how to ensure coverage of possible target areas with the least number of artillery shells for air defense systems or ground artillery systems. The method proposed in the present invention can achieve coverage of areas where incoming targets may appear with multiple artillery shells, thereby achieving precise strikes on incoming targets.

[0235] In the simulation environment, it is assumed that the target area is a 100m*100m square area. The boundary of the square area is shown by a black rectangular frame. The effective damage radius of each shell is 10m. The coverage area of ​​each shell is drawn in the form of a circular outline. The radius of the circular coverage area is equal to the effective coverage range of the shell.

[0236] The code calculates and demonstrates a layout plan for covering an area in a square area with the least number of circular shells. The shells are added gradually by using a shell layout planning method based on the square coverage area of ​​the possible areas, ensuring that the shell positions selected at each step maximize the coverage of the uncovered area. Figure 2 This is the layout planning diagram (before optimization) obtained by using the artillery shell layout planning method based on the square coverage area of ​​the possible occurrence area. It can be concluded that 80 artillery shells are required to fully cover the square area using this method.

[0237] In the same context, Figure 3 This is the layout planning diagram (after optimization) obtained by using the particle swarm algorithm-based artillery shell layout plan generation method. It can be concluded that 56 artillery shells are required to fully cover the square area using this method, which can effectively reduce the number of artillery shells covered compared to the above method.

[0238] It can be clearly seen from the above simulation results that the present invention can achieve the goal of multiple shells covering the area where the incoming target may appear, thereby achieving precise strikes on the incoming target.

[0239] Example 2

[0240] Based on the method for generating an air defense artillery shell deployment plan described in the first embodiment, this embodiment provides a device for generating an air defense artillery shell deployment plan, including:

[0241] The data acquisition module is used to collect detection data of incoming targets in real time and perform preprocessing to generate preprocessed data;

[0242] a trajectory generation module, configured to pre-generate a trajectory of an incoming target based on pre-processed data, and to correct the pre-generated trajectory of the incoming target to obtain a corrected trajectory of the incoming target;

[0243] A plan generation module is used to plan the deployment of air defense artillery shells within the square coverage area where the incoming target may appear based on the corrected trajectory of the incoming target, and obtain an air defense artillery shell deployment plan;

[0244] The scheme optimization module is used to optimize the air defense artillery shell deployment scheme using the particle swarm optimization algorithm to obtain the optimized air defense artillery shell deployment scheme for intercepting incoming targets.

[0245] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for generating an air defense artillery shell deployment plan, characterized in that: include: Collect detection data of incoming targets in real time, perform preprocessing, and generate preprocessed data; Pre-generating a trajectory of the incoming target according to the pre-processed data, and correcting the pre-generated trajectory of the incoming target to obtain a corrected trajectory of the incoming target; According to the corrected trajectory of the incoming target, the air defense artillery shells are deployed within the square coverage area where the incoming target appears, and an air defense artillery shells deployment plan is obtained; The particle swarm optimization algorithm is used to optimize the air defense artillery shell layout plan, and the optimized air defense artillery shell layout plan is obtained to intercept incoming targets; Among them, the trajectory of the incoming target is pre-generated based on the pre-processed data, including: Obtaining a state estimate of the incoming target based on preprocessed data and establishing a motion model of the incoming target; According to the motion model of the incoming target, the mean and covariance matrix of the incoming target's motion state are obtained, and then the trajectory of the incoming target is pre-generated; Correct the pre-generated trajectory of incoming targets, including: Error model construction: The error model includes systematic error and measurement error. It is assumed that the error model is Gaussian distributed, and the systematic error and measurement error are uncorrelated and satisfy ， ， in, represents the system noise, represents the measurement noise, represents the system noise covariance matrix, is the measurement noise covariance matrix, The mean is 0 and the variance is Gaussian distribution, The mean is 0 and the variance is Gaussian distribution, the system's motion equation is expressed as ， in, represents the state vector of the system at time k, represents the state transition matrix, express The control input at the moment, represents the input matrix, express System noise at each moment; The measurement model equation is ， in, represents the measurement value of the sensor at time k, represents the observation matrix, represents the measurement noise at time k; Incoming target status prediction: According to the motion model of the incoming target, the state of the incoming target at time k Prediction ， in, represents the state prediction value at time k, is the estimated value of the state based on the previous moment k-1, It's the last moment Control input; Forecast error covariance matrix ， in, Indicates at time The error covariance matrix of the predicted state represents the magnitude of the prediction error, which is based on The estimated and observed data of the time are obtained; Indicates based on The observation data at the time The error covariance matrix at time t; Measurement information update: The measurement value collected by the sensor at time k and predicted value The difference between Defined as ， Its covariance matrix is ， in, It reflects the error between actual measurement and prediction and is used to correct state estimation; Kalman gain calculation: Kalman gain Determines the weight distribution of measured values ​​and predicted values ​​in state updates, and its calculation formula is: ， The larger the Kalman gain, the more the system relies on the current measurement value; the smaller the Kalman gain, the more the system relies on the predicted value; Status Update: The predicted state is corrected using the Kalman gain, and the updated state estimate is ， At the same time, the covariance matrix is ​​updated: ， in, Indicates based on The observation data at the time The error covariance matrix at time , Indicates based on The observation data at the time The error covariance matrix at time t; Through this state update, the trajectory error of the incoming target is corrected, making the state estimation closer to the actual trajectory of the incoming target.

2. The method for generating an air defense artillery shell arrangement plan according to claim 1, characterized in that: Real-time acquisition of incoming target detection data and pre-processing, including: Through timestamp technology, different types of detection data are synchronized and time-aligned to obtain intermediate data; The intermediate data is denoised, cleaned and fused to obtain preprocessed data.

3. The method for generating an air defense shell arrangement plan according to claim 1, wherein: Based on the corrected trajectory of the incoming target, anti-aircraft artillery shell deployment planning is carried out within the square coverage area where the incoming target appears, including: The center coordinates of the square coverage area where the incoming target appears are defined as , with a side length of L; The square coverage area is divided into several grid cells, and the weight of each grid cell is calculated based on the probability of the incoming target appearing in a certain grid cell and the area of ​​the grid cell; The projectiles are evenly distributed within the square area. The spacing between the projectiles is calculated to ensure that every grid cell within the square coverage area is covered. The launch position of each projectile is determined by the following formula: ， in, Indicates the launch position of the shell, p is the index of the shell, M is the number of shells in each row, 、 The horizontal and vertical spacing of the shells, where the horizontal and vertical spacing refer to the horizontal and vertical distances between adjacent shells on the plane; Determine the coverage effect of a grid cell by calculating the distance from the center point of the grid cell to the launch position of the projectile ， in, For the The center coordinates of the grid cells, is the launch position of the p-th shell, represents the coverage effect of the i-th particle arrangement scheme on the j-th grid unit, Represents the radius of the shell detonation area. The coverage effect of the grid cell refers to whether each grid cell in the block is fully or partially covered by the shell's damage radius. Grid division decomposes the target area into multiple small areas to evaluate and optimize the shell's coverage effect. Finally, the air defense artillery shell layout plan is selected by calculating the total coverage of the entire square area. , in, represents the total coverage of the entire square area, represents the weight of the grid cell, It represents the sum of the weights of all shells and is used to measure whether the coverage area meets the requirements.

4. The method for generating an air defense artillery shell arrangement plan according to claim 3, characterized in that: The weight of each grid cell is calculated by the following formula: ， in, represents the probability that the target is within the grid, Represents the area of ​​a grid cell.

5. The method for generating an air defense artillery shell arrangement plan according to claim 3, characterized in that: The particle swarm optimization algorithm is used to optimize the anti-aircraft artillery shell layout scheme, including: Define the basic parameters of the particle swarm and set the objective function : ， ， in, Indicates the total number of grids; Based on the set objective function, perform particle swarm optimization, including: Initialize the particle swarm and randomly generate the position and velocity of the initial particles. The position of the particles represents the launch position of the projectile, and the velocity represents the change in position. Then, the velocity and position of each particle are updated according to the formula of the particle swarm optimization algorithm; For each particle, check whether the current position is better than its historical best position. If so, update the historical best position of particle i. , ， Update the global best position , ， If the maximum number of iterations is reached Or if the change of the objective function is less than a predetermined threshold, the iteration is stopped; After the iteration, the position of the optimal particle This is the optimized anti-aircraft artillery shell layout plan.

6. The method for generating an air defense artillery shell arrangement plan according to claim 5, characterized in that: Define the trigger conditions for trajectory changes , Indicates time The change in the target position is used to trigger the dynamic optimization process. ， When the trajectory changes Exceeding the preset threshold Dynamic optimization is triggered when , regenerates the shell placement plan.

7. The method for generating an air defense artillery shell arrangement plan according to claim 6, characterized in that: When dynamic optimization is triggered, the objective function is updated first. Assume that the target area is represented by a rectangular grid. Indicates that the new optimization goal is to re-cover the area, and the objective function is expressed as: ， in, Represents particles In the new grid coverage situation, in order to speed up the optimization, the particle speed update formula is adjusted to: ， By adjusting the inertia weight , learning factor and , to achieve rapid convergence of the particle swarm, thus generating a new optimal layout plan in a relatively short time, r 1 、r 2 is a random number; After generating a new optimal layout plan, update the artillery launch instructions, where the artillery launch angle Corrected to, ， in, Indicates the coordinates of the incoming target, Indicates the coordinates of the launcher.

8. A device for generating an air defense artillery shell deployment plan, characterized in that: include: The data acquisition module is used to collect detection data of incoming targets in real time and perform preprocessing to generate preprocessed data; a trajectory generation module, configured to pre-generate a trajectory of an incoming target based on pre-processed data, and to correct the pre-generated trajectory of the incoming target to obtain a corrected trajectory of the incoming target; A plan generation module is used to plan the deployment of air defense artillery shells within the square coverage area where the incoming target appears based on the corrected trajectory of the incoming target, and obtain an air defense artillery shell deployment plan; The scheme optimization module is used to optimize the air defense artillery shell deployment scheme using the particle swarm optimization algorithm to obtain the optimized air defense artillery shell deployment scheme for intercepting incoming targets; Among them, the trajectory of the incoming target is pre-generated based on the pre-processed data, including: Obtaining a state estimate of the incoming target based on preprocessed data and establishing a motion model of the incoming target; According to the motion model of the incoming target, the mean and covariance matrix of the incoming target's motion state are obtained, and then the trajectory of the incoming target is pre-generated; Correct the pre-generated trajectory of incoming targets, including: Error model construction: The error model includes systematic error and measurement error. It is assumed that the error model is Gaussian distributed, and the systematic error and measurement error are uncorrelated and satisfy ， ， in, represents the system noise, represents the measurement noise, represents the system noise covariance matrix, is the measurement noise covariance matrix, The mean is 0 and the variance is Gaussian distribution, The mean is 0 and the variance is Gaussian distribution, the system's motion equation is expressed as ， in, represents the state vector of the system at time k, represents the state transition matrix, express The control input at the moment, represents the input matrix, express System noise at each moment; The measurement model equation is ， in, represents the measurement value of the sensor at time k, represents the observation matrix, represents the measurement noise at time k; Incoming target status prediction: According to the motion model of the incoming target, the state of the incoming target at time k Prediction ， in, represents the state prediction value at time k, is the estimated value of the state based on the previous moment k-1, It's the last moment Control input; Forecast error covariance matrix ， in, Indicates at time The error covariance matrix of the predicted state represents the magnitude of the prediction error, which is based on The estimated and observed data of the time are obtained; Indicates based on The observation data at the time The error covariance matrix at time t; Measurement information update: The measurement value collected by the sensor at time k and predicted value The difference between Defined as ， Its covariance matrix is ， in, It reflects the error between actual measurement and prediction and is used to correct state estimation; Kalman gain calculation: Kalman gain Determines the weight distribution of measured values ​​and predicted values ​​in state updates, and its calculation formula is: ， The larger the Kalman gain, the more the system relies on the current measurement value; the smaller the Kalman gain, the more the system relies on the predicted value; Status Update: The predicted state is corrected using the Kalman gain, and the updated state estimate is ， At the same time, the covariance matrix is ​​updated: ， in, Indicates based on The observation data at the time The error covariance matrix at time , Indicates based on The observation data at the time The error covariance matrix at time t; Through this state update, the trajectory error of the incoming target is corrected, making the state estimation closer to the actual trajectory of the incoming target.

Images