KR102494978B1 - Method and Apparatus for Fire Control of Close-In Weapon System Using Deep Learning - Google Patents

Abstract

Disclosed is a fire control method of a proximity defense system using deep learning and a device therefor.
A fire control method of a proximity defense system according to an embodiment of the present invention stores target maneuver data obtained by detecting and tracking a target, and uses a target maneuver analysis model for acquiring target information using the stored target maneuver data. generating a target maneuver analysis step; a fire control analysis step of storing shooting data for a target and generating a fire control analysis model for obtaining a trajectory and an impact point of a bullet using the stored shooting data; a proximity defense analysis step of calculating an interception result on whether or not the bullet is shot down using a proximity defense analysis model generated based on the target maneuver analysis model and the fire control analysis model, and deep learning-processing the interception result; and an interception setting control step of performing an interception setting for fire control based on a learning result of the interception result.

Description

Method and Apparatus for Fire Control of Close-In Weapon System Using Deep Learning

The present invention relates to a method for controlling fire of a proximity defense system using deep learning and an apparatus therefor.

The contents described in this section simply provide background information on the embodiments of the present invention and do not constitute prior art.

The close-in defense weapon system is operated at a very close distance from the target as the last defense method in self-ship defense. In order to check the performance of the system, it is necessary to conduct a test of intercepting a high-speed target by flying it, but it is difficult to perform multiple tests due to many limitations such as cost / safety.

Existing countermeasures shoot using the currently acquired detection result of the radar/optical device for the target. In this way, it is not easy to increase the accuracy of a high-mobility target through a fire control algorithm.

The present invention provides deep learning that performs fire control to improve the performance (accuracy rate, kill rate, etc.) of a close-in weapon system (CIWS) through M&S (modeling & simulation) analysis process and deep learning learning. The main purpose is to provide a fire control method of the used proximity defense system and a device therefor.

Other non-specified objects of the present invention may be additionally considered within the scope that can be easily inferred from the following detailed description and effects thereof.

According to one aspect of the present invention, a fire control method of a proximity defense system for achieving the above object includes storing target maneuver data obtained by detecting and tracking a target, and acquiring target information using the stored target maneuver data. A target maneuver analysis step of generating a target maneuver analysis model for; a fire control analysis step of storing shooting data for a target and generating a fire control analysis model for obtaining a trajectory and an impact point of a bullet using the stored shooting data; a proximity defense analysis step of calculating an interception result on whether or not the bullet is shot down using a proximity defense analysis model generated based on the target maneuver analysis model and the fire control analysis model, and deep learning-processing the interception result; and an interception setting control step of performing an interception setting for fire control based on a learning result of the interception result.

In addition, according to another aspect of the present invention, a fire control device for achieving the above object stores maneuver data obtained by detecting and tracking a target, and maneuvers the target for acquiring maneuver characteristics using the stored maneuver data. a target maneuver analysis unit generating an analysis model; a fire control analysis unit that stores shooting data for a target and generates a fire control analysis model for obtaining a trajectory and an impact point of a bullet using the stored shooting data; a proximity defense analysis unit calculating an interception result on whether or not the bullet was shot down based on the target maneuver analysis model and the fire control analysis model, and processing the interception result through deep learning; and an interception setting controller configured to perform interception setting for fire control based on a learning result of the interception result.

As described above, the present invention has the effect of reducing cost and securing safety by quantifying and analyzing data of an actual weapon system and increasing reliability of the system by applying a deep learning algorithm.

In addition, the present invention can predict and analyze the performance of a melee defense weapon system by quantifying and analyzing actual weapon system data and applying a deep learning algorithm, and improves interception performance by controlling fire control of the melee defense weapon system. There are effects that can be done.

In addition, the present invention has an effect of increasing reliability by continuously updating through analysis results and actual system performance tests by processing deep learning learning using actual measurement data of a modeling & simulation (M&S) analysis process.

1 is a schematic block diagram of a fire control device according to an embodiment of the present invention.
2 is a diagram for explaining a fire control method according to an embodiment of the present invention.
3 is an exemplary view showing the result of interception performance according to the fire control operation of the fire control device according to an embodiment of the present invention.
4 is an exemplary diagram for explaining an operation of updating a fire control database according to an embodiment of the present invention.
5 is an exemplary view showing a fire control M&S result according to an embodiment of the present invention.
6 is an exemplary view showing a target maneuver M&S result according to an embodiment of the present invention.
7 is an exemplary view showing a fire control result according to intercept setting control according to an embodiment of the present invention.
8 is a flowchart illustrating a fire control method according to an embodiment of the present invention.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. Advantages and features of the present invention, and methods of achieving them, will become clear with reference to the detailed description of the following embodiments taken in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below and may be implemented in various different forms, only the present embodiments make the disclosure of the present invention complete, and common knowledge in the art to which the present invention belongs. It is provided to completely inform the person who has the scope of the invention, and the present invention is only defined by the scope of the claims. Like reference numbers designate like elements throughout the specification.

Unless otherwise defined, all terms (including technical and scientific terms) used in this specification may be used in a meaning commonly understood by those of ordinary skill in the art to which the present invention belongs. In addition, terms defined in commonly used dictionaries are not interpreted ideally or excessively unless explicitly specifically defined.

In this specification, terms such as "first" and "second" are used to distinguish one component from another component, and the scope of rights should not be limited by these terms. For example, a first element may be termed a second element, and similarly, a second element may be termed a first element.

In this specification, identification codes (eg, a, b, c, etc.) for each step are used for convenience of explanation, and identification codes do not describe the order of each step, and each step is clearly Unless a specific order is specified, it may occur in a different order from the specified order. That is, each step may occur in the same order as specified, may be performed substantially simultaneously, or may be performed in the reverse order.

In this specification, expressions such as "has", "may have", "includes" or "may include" indicate the existence of a corresponding feature (eg, numerical value, function, operation, or component such as a part). indicated, and does not preclude the presence of additional features.

In addition, the term '~unit' described in this specification means software or a hardware component such as a field-programmable gate array (FPGA) or ASIC, and '~unit' performs certain roles. However, '~ part' is not limited to software or hardware. '~bu' may be configured to be in an addressable storage medium and may be configured to reproduce one or more processors. Therefore, as an example, '~unit' refers to components such as software components, object-oriented software components, class components, and task components, processes, functions, properties, and procedures. , subroutines, segments of program code, drivers, firmware, microcode, circuitry, data structures and variables. Functions provided within components and '~units' may be combined into smaller numbers of components and '~units' or further separated into additional components and '~units'.

Proximity defense system is the last defense system to neutralize the target approaching the trap. Targets include anti-aircraft, anti-ship, and missiles, especially anti-ship missiles such as the Sea-Skimmer. In the case of missiles, it is difficult to test them due to their high speed and difficulty in developing a weapon system targeting them.

The present invention can improve the intercept performance (accuracy rate, shooting down rate, etc.) of the melee defense system through an M&S (modeling & simulation) analysis process and a deep learning learning process in order to overcome the difficulties of analysis and testing of the melee defense system. . The M&S analysis process may consist of an M&S analysis process for target maneuver, an M&S analysis process for fire control, and an M&S analysis process for interception results. In addition, various types of learning methods may be applied to the deep learning learning process to produce an optimal interception result.

Hereinafter, a fire control method of a proximity defense system using deep learning proposed in the present invention and a device therefor will be described in detail with reference to the drawings.

1 is a schematic block diagram of a fire control device according to an embodiment of the present invention.

The fire control apparatus 100 of the proximity defense system according to the present embodiment includes a target maneuver analysis unit 110, a fire control analysis unit 120, a proximity defense analysis unit 130, and an interception setting control unit 140. The fire control device 100 of FIG. 1 is according to an embodiment, and all blocks shown in FIG. 1 are not essential components, and some blocks included in the fire control device 100 in another embodiment are added or changed. or can be deleted.

The fire control device 100 performs fire control to improve the performance (hit rate, shoot down rate, etc.) of a close-in weapon system (CIWS) through M&S (modeling & simulation) analysis process and deep learning learning. do.

Specifically, the fire control apparatus 100 utilizes actual data of the proximity defense system and reflects it in the M&S analysis process. The M&S analysis process, in which the preset theory and actual data of the system included in the proximity defense system are reflected, has high reliability and is updated through continuous data update. The fire control device 100 is A lot of data stored through the M&S analysis process is applied to the deep learning learning process. In addition, the fire control apparatus 100 may generate new virtual data by itself in the M&S analysis process and perform deep learning learning when a certain amount of data corresponding to a predetermined criterion is accumulated.

The target maneuver analysis unit 110 performs data acquisition and analysis on target maneuvers.

Specifically, the target maneuver analysis unit 110 stores target maneuver data obtained by detecting and tracking a target, and generates a target maneuver analysis model for acquiring target information using the stored target maneuver data.

The target maneuver analysis unit 110 according to this embodiment includes a target maneuver result storage unit 112 and a target maneuver M&S analyzer 114 .

The target maneuvering result storage unit 112 detects and tracks a target based on data actually measured through a monitoring system, and acquires and stores target maneuvering data for at least one target.

The target maneuver M&S analysis unit 114 generates a target maneuver analysis model based on a thrust model according to the maneuver characteristics of the target, a mass change model according to fuel consumption, and a DOF motion model reflecting target maneuver data, and the target maneuver analysis model Obtain target information through

Specifically, the target maneuvering M&S analyzer 114 may calculate a detection probability according to radar performance and target characteristics and determine whether to acquire a target when a target maneuvering during operation according to predetermined search beam scheduling enters the beam. The target maneuvering M&S analysis unit 114 may start tracking by handing over the target information captured in the search to the tracking radar. At this time, the target maneuvering M&S analysis unit 114 may perform M&S analysis on the sensor system using [Equation 1] below.

Here, P _d represents the probability of detection. N represents a noncoherent integration number. P _fa represents the false positive rate. SNR represents the signal-to-noise ratio.

Meanwhile, the target maneuver analysis unit 110 constructs a target maneuver analysis model by utilizing the actually measured maneuver results by utilizing data of a weapon system monitoring system (radar, etc.). This target maneuver analysis model has high reliability because it is based on actual measurement data including systematic information (theory) and errors. In the target maneuver analysis model, the coefficients are updated and updated using the actual measured data of the target in [Equation 2] and [Equation 3] composed of the following process and equation/theory.

The target maneuver analysis unit 110 first performs data acquisition by calculating the aerodynamic coefficient according to the target's diameter, length, physical shape such as control rudder shape, altitude, and speed. Then, the target maneuver analysis unit 110 reflects the characteristics of the target propulsion engine through [Equation 2] to build a thrust model and simulate mass change according to fuel consumption.

Here, I _sp is the specific thrust coefficient, m is the mass, t _b is the burning time, and t _e is the required injection time.

Thereafter, the target maneuver analysis unit 110 updates the simulation of the 6 degree of freedom motion model reflecting the actual target maneuver measurement data through [Equation 3] to the target maneuver analysis model to generate a similar trajectory (trajectory for a new target) can do.

Here, U, V, and W denote target velocity, P, Q, and R denote target angular velocity, I, J denote inertia, and F and M denote external forces and moments acting on the target.

The fire control analyzer 120 acquires and analyzes fire control data.

The fire control analyzer 120 stores shooting data for a target and creates a fire control analysis model for acquiring the trajectory and impact point of a bullet using the stored shooting data.

The fire control analysis unit 120 according to the present embodiment includes a shooting test result storage unit 122 and a fire control M&S analysis unit 124.

The shooting test result storage unit 122 acquires and stores the shooting data for the shooting test of the shooting system of the proximity defense system.

The fire control M&S analysis unit 124 updates the newly acquired new shooting data to the previously stored shooting data to create a fire control analysis model using the updated shooting data, and the fire control analysis model to determine the trajectory of the bullet and obtain an impact point.

Specifically, the fire control M&S analysis unit 124 may perform M&S analysis on fire control. That is, the fire control M&S analysis unit 124 may calculate an azimuth drive command and an elevation drive command using target information detected and tracked through the sensor system M&S analyzer 110 .

And, the fire control M&S analysis unit 124 may calculate a predicted hitting point (PHP) as shown in FIG. 3 using target information detected and tracked through the sensor system M&S analysis unit 110. .

Also, the fire control M&S analyzer 124 may fire a bullet based on the aiming point and obtain a trajectory of the bullet based on the aiming point.

Also, the fire control M&S analyzer 124 may acquire the point of impact of the bullet (see FIG. 5) based on the trajectory of the bullet. Bullets launched along the projectile trajectory reflect characteristics such as the shape and material of the projectile and fly, and have inherent dispersion in azimuth and elevation directions, through which the final impact point of the projectile can be determined.

At this time, the fire control M&S analysis unit 124 may perform M&S analysis for fire control using [Equation 4] below.

이다. g는 중력 가속도를 나타낸다.Here, t _f represents the elapsed time. γ represents the launch angle. x and y represent the target location. t represents the elapsed time after the shot is fired. r represents the target distance. R _target represents the target distance. V ₀ represents the initial velocity of the bullet. θ represents the target elevation. β represents the drag coefficient (0.087). v is

am. g represents the gravitational acceleration.

Meanwhile, the fire control analysis unit 120 uses actual data to process a fire control analysis model constructed based on theory based on shooting data of a close-in defense weapon system in actual operation or the same/similar shooting system. When a shooting test is performed under certain conditions, it is measured by a monitoring system and made into a database. Data stored in a database includes muzzle shake, warhead trajectory, etc. in azimuth and elevation under conditions of firing toward one target.

Referring to FIG. 4 , the fire control analyzer 120 acquires shooting data, and the shooting data has a peak distribution. The fire control analyzer 120 updates and updates new shooting data (Fig. 4(b)) to the old (existing) database (Fig. 4(a)), and updates the updated database (Fig. 4(c) )) is used in the fire control analysis model. That is, the fire control analysis model utilizes bullet trajectory (azimuth and elevation) data databased through the conditions of muzzle fired toward the target during simulation.

The proximity defense analysis unit 130 calculates an interception result on whether or not a bullet is shot down using the proximity defense analysis model generated based on the target maneuver analysis model and the fire control analysis model, and deep learning processes the interception result.

The proximity defense analysis unit 130 according to the present embodiment includes a proximity defense M&S analysis unit 132 and a model learning unit 134.

The proximity defense M&S analysis unit 132 calculates an interception result by determining whether or not the bullet is shot down based on the impact point and the area where the shot can be shot down.

The model learning unit 134 processes deep learning learning with target information, impact point, and intercept result as inputs, and outputs a final learning result in which target information and impact point are optimized to increase the probability of interception.

The model learning unit 134 may calculate different learning results for one target and output a final learning result for the final impact point through the calculation. For example, the model learning unit 134 extracts an optimal first impact point among at least one impact point associated with target information of the first target based on the interception result calculated by the proximity defense M&S analysis unit 132, 1 Calculate learning outcomes. Thereafter, the model learning unit 134 calculates a second learning result by extracting an optimal second impact point from among at least one impact point associated with the trajectory of the first target and target information of a second target located within a preset range. Thereafter, the model learning unit 134 assigns preset different weights to each of the first learning result and the second learning result, calculates the final impact point for the first target based on the first impact point and the second impact point, The final learning result can be output.

The melee defense M&S analysis unit 132 may perform M&S analysis on hit/kill.

That is, the proximity defense M&S analyzer 132 may determine an interception result as to whether or not the bullet is shot down based on the impact point and the shootable area obtained through the fire control M&S analyzer 124 .

More specifically, the proximity defense M&S analyzer 132 may obtain a miss distance between the point of impact and the target. For example, the proximity defense M&S analysis unit 132 calculates the error distance between the expected impact location of each bullet in the impact point and the target for each bullet, and the average value of the error distance calculated for each bullet is the error distance between the impact point and the target. can decide

In addition, the proximity defense M&S analysis unit 132 may acquire a projection area from which the target is viewed from the gun that fired the bullet as a possible area to shoot down. Here, the area capable of shooting down may be obtained based on the first projection area, the second projection area, the third projection area, and the fourth projection area. That is, the proximity defense M&S analysis unit 132 may obtain an intersection area of the third projection area and the fourth projection area, and obtain a union area of the first projection area, the second projection area, and the intersection area as a possible shooting down area. there is.

The first projected area may indicate a projected area on the back side of the warhead of the bullet. The second projected area may indicate a projected area of the front side of the warhead of the bullet. The third projected area may represent a projected area of a side surface of the warhead of the bullet. The fourth projected area may indicate a projected area of the other side surface of the warhead of the bullet.

At this time, the proximity defense M&S analysis unit 132 determines the first projected area based on the diameter r of the warhead of the bullet, the length H of the warhead of the projectile, the azimuth angle φ between the target and the proximity defense system, and the angle of incidence θ in the elevation direction of the projectile, A second projected area, a third projected area, and a fourth projected area may be obtained. That is, the proximity defense M&S analysis unit 132 obtains the first projected area S ₁ , the second projected area S ₂ , the third projected area S ₃ and the fourth projected area S ₄ using Equation 5 below. can do.

Also, the proximity defense M&S analysis unit 132 may determine an interception result regarding whether or not the bullet is shot down based on the error distance and the shootdown possible area.

At this time, the proximity defense M&S analysis unit 132 may determine that the bullet is shot down by the target when the position of the bullet using the error distance is located within the shootable area. For example, if an area having an error distance as a radius centered on an expected impact position of each bullet within an impact point is included in the area capable of being shot down, it may be determined that the bullet is shot down by the target. Of course, if the area whose radius is the error distance centered on the projected impact location of each bullet within the impact point is included in the shootable area at least a preset threshold (50%, etc.), the bullet can be determined to be shot down by the target. there is. In addition, the proximity defense M&S analyzer 132 may calculate the kill rate based on the result of interception of whether each bullet is shot down.

Meanwhile, the proximity defense analysis unit 130 integrates the target maneuver analysis model and the fire control analysis model to generate a proximity defense analysis model capable of analyzing the performance of the proximity defense system. Because the proximity defense analysis model simulates real data and system operation algorithms, deep learning can be processed through many simulations.

The proximity defense analysis unit 130 may continuously update the model by comparing the result of interception calculated using the deep learning-processed proximity defense analysis model with the actual test result.

The interception setting control unit 140 performs interception setting for fire control based on the result of learning about the interception result.

Specifically, the interception setting control unit 140 updates optimal target information according to a final learning result to target maneuver data, and updates an impact point to shooting data to perform interception settings for a target.

When the interception setting is performed, the interception setting control unit 140 may derive an interception result through optimized fire control as shown in FIG. 3 . That is, when a target maneuvers and approaches its own ship, it shoots and intercepts based on fire control learned through deep learning.

FIG. 2 is a view for explaining a fire control method according to an embodiment of the present invention, and FIG. 3 is an exemplary view showing the result of interception performance according to the fire control operation of the fire control apparatus according to an embodiment of the present invention.

2 is a detailed flowchart of the fire control device 100 of the proximity defense system configured by deep learning of actual shooting and target maneuvering data, and FIG. 3 shows interception results accordingly.

The fire control apparatus 100 first stores actual shooting and target maneuvering data in a database (S210, S212, S220, S222), and then constructs a target maneuver analysis model and a fire control analysis model representing the databased data, respectively. (S214, S224).

The fire control apparatus 100 configures a target maneuver analysis model and a fire control analysis model as a proximity defense analysis model of a proximity defense system that simulates a system algorithm (S230). The melee defense analysis model has high reliability because it is constructed based on measurement data (including the operation of each subsystem and errors) and theory.

The fire control apparatus 100 continuously simulates various target scenarios and fire control algorithms through a proximity defense analysis model, and learns the results through deep learning (S240).

The fire control device 100 may intercept by performing optimized shooting according to target maneuver characteristics through deep learning data based on the proximity defense analysis model (S250). In addition, the fire control apparatus 100 repeatedly performs tests according to target maneuver characteristics and continuously updates the model through comparative analysis of test results and existing data.

6 is an exemplary view showing the target maneuvering M&S result according to an embodiment of the present invention, and FIG. 7 is an example view showing a fire control result according to intercept setting control according to an embodiment of the present invention.

Figure 6 shows the modeled target maneuvering results. The fire control device 100 provides maneuvering characteristics including errors, including actual measurement data of the monitoring system based on the theory.

As a result, as shown in FIG. 7 , the fire control device 100 controls shooting according to target maneuvering scenarios through M&S (modeling & simulation) analysis process and deep learning learning to calculate shooting conditions having an optimized interception probability. can do.

Through the present invention, it is possible to develop an M&S analysis process for a system that is difficult to actually test, such as a close-in defense weapon system, and perform highly reliable simulation through deep learning technology. In addition, through this method, the fire control device 100 controls the fire control of the melee defense weapon system to have an optimized interception probability according to the target scenario.

8 is a flowchart illustrating a fire control method according to an embodiment of the present invention.

The fire control device 100 acquires and analyzes data on target maneuvers (S810).

Specifically, the fire control apparatus 100 stores target maneuvering data obtained by detecting and tracking a target, and creates a target maneuvering analysis model for obtaining target information using the stored target maneuvering data.

The fire control apparatus 100 acquires and analyzes fire control data (S820).

Specifically, the fire control analyzer 120 stores shooting data for a target and uses the stored shooting data to create a fire control analysis model for obtaining a trajectory and impact point of a bullet.

The fire control apparatus 100 calculates an interception result on whether or not a bullet is shot down using a target maneuver analysis model and a proximity defense analysis model generated based on the fire control analysis model, and deep learning processes the interception result (S830). , S840).

The fire control apparatus 100 performs interception setting for fire control based on the learning result of the interception result (S850).

Specifically, the fire control apparatus 100 updates the optimal target information according to the final learning result to the target maneuver data, and updates the point of impact to the shooting data to set the target to be intercepted.

In FIG. 8, it is described that each step is sequentially executed, but is not necessarily limited thereto. In other words, since it will be applicable to changing and executing the steps described in FIG. 8 or executing one or more steps in parallel, FIG. 8 is not limited to a time-sequential order.

The shooting control method according to the present embodiment described in FIG. 8 may be implemented as an application (or program) and recorded on a recording medium readable by a terminal device (or computer). A recording medium in which an application (or program) for implementing the fire control method according to the present embodiment is recorded and which can be read by a terminal device (or computer) is any type of recording device storing data that can be read by a computing system or includes media

The above description is only illustrative of the technical idea of the embodiment of the present invention, and those skilled in the art to which the embodiment of the present invention pertains may make various modifications and modifications within the scope not departing from the essential characteristics of the embodiment of the present invention. transformation will be possible. Therefore, the embodiments of the present invention are not intended to limit the technical idea of the embodiment of the present invention, but to explain, and the scope of the technical idea of the embodiment of the present invention is not limited by these examples. The protection scope of the embodiments of the present invention should be construed according to the claims below, and all technical ideas within the equivalent range should be construed as being included in the scope of the embodiments of the present invention.

100: fire control device 110: target maneuver analysis unit
120: fire control analysis unit 130: close defense analysis unit
140: interception setting control unit

Claims (13)

In the method of controlling fire in a fire control device,
a target maneuver analysis step of storing target maneuver data obtained by detecting and tracking a target and generating a target maneuver analysis model for obtaining target information using the stored target maneuver data;
a fire control analysis step of storing shooting data for a target and generating a fire control analysis model for obtaining a trajectory and an impact point of a bullet using the stored shooting data;
a proximity defense analysis step of calculating an interception result on whether or not the bullet is shot down using a proximity defense analysis model generated based on the target maneuver analysis model and the fire control analysis model, and deep learning-processing the interception result; and
An interception setting control step of performing interception setting for fire control based on the learning result of the interception result,
The target maneuver analysis step may include a target maneuver result storage step of detecting and tracking a target based on data actually measured through a monitoring system, and acquiring and storing the target maneuver data for at least one target; and generating the target maneuver analysis model based on a thrust model according to maneuver characteristics of the target, a mass change model according to fuel consumption, and a DOF motion model reflecting the target maneuver data, and generating the target maneuver analysis model through the target maneuver analysis model. A fire control method comprising a target maneuver M&S analysis step of obtaining information. delete According to claim 1,
The fire control analysis step,
a shooting test storage step of obtaining and storing the shooting data for a shooting test of a shooting system of a proximity defense system; and
Update the newly acquired new shooting data with the previously stored shooting data to create the fire control analysis model using the updated shooting data, and obtain the trajectory of the bullet and the impact point through the fire control analysis model Fire Control M&S Analysis Steps
Fire control method comprising a. According to claim 3,
The fire control analysis step,
Characterized in that a predicted hitting point (PHP) is calculated using detected and tracked target information, a trajectory of a projectile is obtained based on the aimed point, and a hitting point of the projectile is obtained based on the trajectory of the projectile fire control method. According to claim 1,
The melee defense analysis step,
an interception result calculation step of calculating the interception result by determining whether or not the bullet is shot down based on the impact point and the area where the bullet can be shot down; and
A deep learning learning step of processing the deep learning learning with the target information, the impact point, and the interception result as inputs, and outputting a final learning result obtained by optimizing the target information and the impact point for increasing the probability of interception.
Fire control method comprising a. According to claim 5,
The melee defense analysis step,
A miss distance between the impact point and the target is obtained, a projection area viewed from the gun that fired the bullet at the target is acquired as the possible shoot down area, and the miss distance and the possible shoot down area are obtained. A fire control method characterized in that it is determined whether or not the bullet is shot down based on. According to claim 5,
The deep learning learning step,
Extracting an optimal first impact point among at least one impact point associated with target information of a first target based on the interception result to calculate a first learning result;
Extracting an optimal second impact point among at least one impact point associated with target information of a second target located within a trajectory of the first target and a predetermined range to calculate a second learning result;
A preset different weight is assigned to each of the first learning result and the second learning result, and a final impact point for the first target is calculated based on the first impact point and the second impact point, thereby obtaining the final learning result. A fire control method characterized by outputting. According to claim 5,
In the intercept setting control step,
and performing the intercept setting for the target by updating the optimal target information according to the final learning result in the target maneuvering data and updating the impact point in the shooting data. In the device for fire control,
a target maneuver analysis unit configured to store maneuver data obtained by detecting and tracking a target, and to generate a target maneuver analysis model for acquiring maneuver characteristics using the stored maneuver data;
a fire control analysis unit that stores shooting data for a target and generates a fire control analysis model for obtaining a trajectory and an impact point of a bullet using the stored shooting data;
a proximity defense analysis unit calculating an interception result on whether or not the bullet was shot down based on the target maneuver analysis model and the fire control analysis model, and processing the interception result through deep learning; and
Including an interception setting control unit that performs interception setting for fire control based on the learning result of the interception result,
The target maneuver analysis unit detects and tracks a target based on data actually measured through a monitoring system, and acquires and stores the target maneuver data for at least one target; and generating the target maneuver analysis model based on a thrust model according to maneuver characteristics of the target, a mass change model according to fuel consumption, and a DOF motion model reflecting the target maneuver data, and target information through the target maneuver analysis model. A fire control device comprising a target maneuvering M&S analysis unit for obtaining a. delete According to claim 9,
The fire control analysis unit,
a shooting test storage unit for obtaining and storing the shooting data for a shooting test of a shooting system of a proximity defense system; and
Update the newly acquired new shooting data with the previously stored shooting data to create the fire control analysis model using the updated shooting data, and obtain the trajectory of the bullet and the impact point through the fire control analysis model Fire Control M&S Analysis Department
Fire control device comprising a. According to claim 9,
The melee defense analysis unit,
a proximity defense M&S analysis unit determining whether or not the bullet is shot down based on the impact point and the area where the shot can be shot down, and calculating the result of the interception; and
A model learning unit that processes the deep learning learning with target information, the impact point, and the intercept result as inputs, and outputs a final learning result obtained by optimizing the target information and the impact point for increasing the probability of interception.
Fire control device comprising a. According to claim 12,
The model learning unit,
Extracting an optimal first impact point among at least one impact point associated with target information of a first target based on the interception result to calculate a first learning result;
Extracting an optimal second impact point among at least one impact point associated with target information of a second target located within a trajectory of the first target and a predetermined range to calculate a second learning result;
A preset different weight is assigned to each of the first learning result and the second learning result, and a final impact point for the first target is calculated based on the first impact point and the second impact point, thereby obtaining the final learning result. A fire control device characterized in that it outputs.

Images