KR-102963614-B1 - LEARNING DEVICE, WARHEAD IDENTIFICATION APPARATUS USING SAME, AND WARHEAD IDENTIFICATION METHOD

KR102963614B1KR 102963614 B1KR102963614 B1KR 102963614B1KR-102963614-B1

Abstract

The present invention relates to a learning device for training a warhead identification device to distinguish between a warhead and fragments of a missile detected by radar, comprising: a collection unit for collecting prior detection information in which the radar has previously detected the warhead and fragments of the missile; a generation unit for generating learning data by matching the prior detection information with a corresponding object among the warhead or fragments; and a learning unit for training the identification unit by inputting the learning data into the identification unit provided in the warhead identification device so as to improve the accuracy of identifying the warhead, and can perform learning to distinguish between the warhead and fragments of a missile detected by radar.

Inventors

노영민

Assignees

한화시스템 주식회사

Dates

Publication Date: 20260513
Application Date: 20250225

Claims (16)

A learning device for training a warhead identification device to distinguish between the warhead and fragments of a missile detected by radar, A collection unit for collecting prior detection information in which the radar detects the missile's warhead and fragments prior to performing the above learning; A generation unit for determining whether the above-mentioned prior detection information is prior detection information of a warhead or prior detection information of a fragment, and for generating training data by matching the prior detection information with the identified object among the warhead or fragment; and A learning unit for inputting the above-mentioned training data into an identification unit provided in the above-mentioned warhead identification device, and comparing the identification result obtained by the identification unit analyzing prior detection information included in the above-mentioned training data to identify an object with the object information included in the input training data, and training the identification unit so that the identification result of the identification unit becomes identical to the object information included in the training data; The above prior detection information is, It includes at least one of velocity information regarding the speed at which an object flies and acceleration information regarding the amount of change in acceleration of the object, and reflection information regarding the radar reflection area of the object. The above identification unit is, A learning device that first classifies an object as a fragment or a suspected warhead object based on at least one of the object's velocity information and acceleration information, and secondarily classifies the suspected warhead object as a warhead or fragment based on reflection information if there is an object classified as a suspected warhead object.
In claim 1, The above-mentioned warhead identification device is equipped with a plurality of identification parts, and The above learning unit is, The above-mentioned prior detection information is further included in an information classification unit for classifying the prior detection information into a first prior detection information detected in the ascent section where the missile is ascending, and a second prior detection information detected in the descent section where the missile is descending. The above generating unit is, First training data is generated using the first prior detection information above, and second training data is generated using the second prior detection information above, and The above learning unit is, A learning device that inputs the first learning data into one of the plurality of identification units and inputs the second learning data into the other.
In claim 1, The above collection unit is, If the missile detected by the above radar separates into multiple objects and the warhead identification device identifies the multiple objects, the identification result and detection information of the object detected by the radar are additionally collected, and The above generating unit is, A learning device that generates learning data usable for future learning by matching the identification results additionally collected by the above-mentioned collection unit with the detection information of the object.
As a warhead identification device for distinguishing between the warhead and fragments of a missile detected by radar, When a missile detected by the above radar separates into multiple objects, an information acquisition unit for receiving detection information from the radar that detects each of the multiple objects; and A warhead identification device comprising: an identification unit learned by a learning device of any one of claims 1 to 3 to identify a warhead among the plurality of objects by analyzing the detection information above.
In claim 4, The above identification unit is, An extraction unit for extracting speed information regarding the flight speed of each of the plurality of objects and reflection information regarding the radar reflection area of each of the plurality of objects from the detection information; A first object classification unit for classifying each of the above plurality of objects into fragments and suspected warhead objects according to the velocity information; and A warhead identification device comprising: a second object classification unit for classifying each object classified as a suspected warhead object into a suspected fragment object or a warhead according to the reflection information.
In claim 4, The above identification unit is, An extraction unit for extracting acceleration information regarding the change in acceleration of each of the plurality of objects and reflection information regarding the radar reflection area of each of the plurality of objects from the detection information; A first object classification unit for classifying each of the plurality of objects as a suspected fragment or warhead object according to the acceleration information; and A warhead identification device comprising: a second object classification unit for classifying each object classified as a suspected warhead object into a suspected fragment object or a warhead according to the reflection information.
In claim 5, The extraction unit further extracts frequency information regarding the frequency pattern that each of the plurality of objects has due to Doppler shift from the detection information, and The above identification unit further comprises a third object classification unit for classifying each object classified as a suspected fragment object into a fragment or a warhead according to the frequency information.
In claim 4, The above identification unit is provided in multiple units, one of which is learned from prior detection information detected during the ascent phase in which the missile ascends, and the other is learned from prior detection information detected during the descent phase in which the missile descends. A warhead identification device further comprising a selection unit for selecting an identification unit to identify a warhead among a plurality of identification units by confirming whether the detection information received by the information acquisition unit is detection information detected in an ascending section or a descending section.
In claim 4, A warhead identification device further comprising a transmission unit for transmitting the identification result of the identification unit to the radar so that the radar tracks objects classified as fragments, excluding objects classified as fragments.
As a warhead identification method for distinguishing between the warhead and fragments of a missile detected by radar, A process in which a learning device trains an identification unit equipped in a warhead identification device so as to improve the identification accuracy of identifying warheads; A process in which, when a missile detected by the radar is separated into multiple objects and a single object is detected as multiple objects by the radar, the radar transmits detection information of each of the multiple objects to the identification unit; and The above identification unit analyzes the detection information to identify a warhead among the plurality of objects; the process includes, The process of training the above identification unit is, The process of collecting prior detection information in which the above radar detects the missile's warhead and fragments before performing learning, A process of determining whether the above-mentioned prior detection information is prior detection information of a warhead or prior detection information of a fragment, and generating training data by matching the prior detection information with the identified object among the warhead or fragment, and The process includes inputting the above-mentioned training data into an identification unit provided in the warhead identification device, and comparing the identification result obtained by the identification unit analyzing prior detection information included in the training data to identify an object with the object information included in the input training data, thereby performing training of the identification unit so that the identification result of the identification unit becomes identical to the object information included in the input training data. The above detection information is, It includes at least one of velocity information regarding the speed at which an object flies and acceleration information regarding the amount of change in acceleration of the object, and reflection information regarding the radar reflection area of the object. The process of identifying the above-mentioned warhead is, A warhead identification method that first classifies an object as a fragment or a suspected warhead based on at least one of the velocity information and acceleration information of an object, and secondarily classifies the suspected warhead object as a warhead or fragment based on reflection information if there is an object classified as a suspected warhead.
delete
In claim 10, The process of identifying the above-mentioned warhead is, A process of extracting speed information regarding the flight speed of each of the plurality of objects and reflection information regarding the radar reflection area of each of the plurality of objects from the detection information; A process of classifying each of the above plurality of objects as a fragment or suspected warhead object according to the above velocity information; and A method for identifying a warhead, comprising the process of classifying each object classified as a suspected warhead object into a suspected fragment object or a warhead according to the reflection information.
In claim 10, The process of identifying the above-mentioned warhead is, A process of extracting acceleration information regarding the change in acceleration of each of the plurality of objects and reflection information regarding the radar reflection area of each of the plurality of objects from the detection information; A process of classifying each of the above plurality of objects as a fragment or suspected warhead object according to the acceleration information; and A method for identifying a warhead, comprising the process of classifying each object classified as a suspected warhead object into a suspected fragment object or a warhead according to the reflection information.
In claim 12 or claim 13, The process of identifying the above-mentioned warhead is, A process of extracting frequency information regarding the frequency pattern that each of the plurality of objects has due to Doppler shift from the detection information; and A warhead identification method further comprising the process of classifying each object classified as a suspected fragment object into a fragment or a warhead according to the frequency information.
In claim 10, After identifying the above warhead, A process of additionally collecting the identification result in which the above identification unit identifies an object and detection information of the object; and A warhead identification method further comprising the process of matching additionally collected identification results with detection information of the object to generate training data usable for future learning.
In any one of claims 10, 12, and 13, After identifying the above warhead, A warhead identification method further comprising the process of transmitting the identification result of the identification unit to the radar so that the radar tracks objects classified as fragments, excluding objects classified as fragments.

Description

Learning device, warhead identification apparatus, and warhead identification method The present invention relates to a learning device, a warhead identification device, and a warhead identification method, and more specifically, to a learning device, a warhead identification device, and a warhead identification method for distinguishing between a warhead and fragments of a missile detected by radar. Generally, a radar system is a device designed to detect objects in the surroundings. Therefore, a radar system can be used to detect missiles fired by enemy forces toward friendly territory. At this time, stage separation can occur during the launch and ascent of a ballistic missile, causing the warhead and propellant to separate. Therefore, when a ballistic missile is detected by a radar system, fragments such as debris from the propellant are detected along with the warhead. However, conventionally, it was impossible to distinguish between the warhead and fragments of a ballistic missile. Consequently, problems could arise where fragments are mistaken for the warhead, leading to a failure to properly respond to the actual warhead and resulting in significant damage. FIG. 1 is a diagram showing the configuration of a learning device and a warhead identification device according to one embodiment of the present invention. FIG. 2 is a diagram showing the configuration of a learning device and a warhead identification device according to another embodiment of the present invention. FIG. 3 is a flowchart illustrating a warhead identification method according to an embodiment of the present invention. Hereinafter, embodiments of the present invention will be described in more detail with reference to the attached drawings. However, the present invention is not limited to the embodiments disclosed below but may be implemented in various different forms, and these embodiments are provided merely to ensure that the disclosure of the present invention is complete and to fully inform those skilled in the art of the scope of the invention. To describe the invention in detail, the drawings may be exaggerated, and like reference numerals in the drawings refer to like elements. FIG. 1 is a diagram showing the configuration of a learning device and a warhead identification device according to an embodiment of the present invention. Below, a learning device and a warhead identification device according to an embodiment of the present invention will be described. The learning device is a learning device that trains a warhead identification device to distinguish between the warhead and fragments of a missile detected by radar. Referring to FIG. 1, the learning device (100) includes a collection unit (110), a generation unit (120), and a learning unit (130). The collection unit (110) can be connected to receive information from the radar (50). Accordingly, the collection unit (110) can collect prior detection information in which the radar (50) has previously detected the warhead and fragments of a missile. The collection unit (110) may be equipped with a storage medium for storing information. Thus, the collection unit (110) can store the collected prior detection information. The generating unit (120) can be connected to receive information from the collecting unit (110). Accordingly, the generating unit (120) can check the prior detection information stored in the collecting unit (110) and generate training data by matching each piece of prior detection information with a corresponding object, such as a warhead or a fragment. For example, a user can check the prior detection information to determine whether the corresponding object is a warhead or a fragment and input it into the generating unit (120). Thus, the generating unit (120) can generate a dataset of training data by matching the prior detection information with the corresponding object. The learning unit (130) is connected to receive learning data from the generation unit (120) and can be connected to input the learning data into the identification unit (220) provided in the warhead identification device (200). Accordingly, the learning unit (130) can input the learning data received from the generation unit (120) into the identification unit (220) to train (or machine learn) the identification unit (220). That is, prior detection information included in the learning data is input into the identification unit (220), and the identification result identified by the identification unit (220) analyzing the prior detection information can be compared with information about the object included in the learning data. Therefore, the learning unit (130) can perform learning so that the identification result identified by the identification unit (220) analyzing the prior detection information becomes identical to the information about the object included in the learning data. At this time, the collection unit (110) can additionally collect the identification result and the de