CN-122020288-A - Unmanned aerial vehicle threat assessment method and device, computer equipment and storage medium

CN122020288ACN 122020288 ACN122020288 ACN 122020288ACN-122020288-A

Abstract

The invention relates to the technical field of computers and discloses an unmanned aerial vehicle threat assessment method, device, computer equipment and storage medium, wherein the method comprises the steps of detecting a target unmanned aerial vehicle with invasion risk in a preset area; the method comprises the steps of obtaining threat elements of a target unmanned aerial vehicle, extracting space-time characteristics of the threat elements through a time sequence mixer, wherein the threat elements at least comprise dynamic threat elements and static threat elements, mapping the space-time characteristics to a threat prototype space, and determining threat levels of the target unmanned aerial vehicle according to prototype classification preset in the threat prototype space. The method solves the problems of low threat assessment accuracy and insufficient instantaneity of the unmanned aerial vehicle in a few-sample and dynamic complex scene in the prior art.

Inventors

CHEN YUEGUANG
LIN QUANYI
LI ZHONGCHENG
GUAN PENG
DING ZHIXIN
FAN HONG
LAN TIANYUN
JIANG YUNPENG
Wang Pingqiang
QIN XIAOYU

Assignees

深圳市城市公共安全技术研究院有限公司
深圳中广核工程设计有限公司

Dates

Publication Date: 20260512
Application Date: 20260120

Claims (10)

1. A method of unmanned aerial vehicle threat assessment, the method comprising: detecting a target unmanned aerial vehicle with an invasion risk in a preset area; Acquiring threat elements of the target unmanned aerial vehicle, and extracting space-time characteristics of the threat elements through a time sequence mixer, wherein the threat elements at least comprise dynamic threat elements and static threat elements; mapping the space-time characteristics to a threat prototype space, and determining the threat level of the target unmanned aerial vehicle according to prototype classification preset in the threat prototype space.
2. The method of claim 1, wherein the acquiring the threat elements of the target drone includes: acquiring monitoring data of the target unmanned aerial vehicle in the preset area; Extracting a situation threat element and a radar threat element of the target unmanned aerial vehicle from the monitoring data, and taking the situation threat element and the radar threat element as the dynamic threat element; And identifying physical attribute elements of the target unmanned aerial vehicle from the monitoring data, and taking the physical attribute elements as the static threat elements.
3. The method of claim 2, wherein the situational threat element comprises at least one of speed, heading angle, climb angle, altitude, radial distance, horizontal distance, azimuth angle, and elevation angle, the radar threat element comprises at least one of pulse width, operating frequency, pulse repetition frequency, and radar cross-sectional area, and the physical attribute element comprises at least one of flyer type, aggressiveness, and maneuverability.
4. The method of claim 1, wherein the extracting the spatiotemporal features of the threat element by a time series mixer comprises: mapping the threat elements to a preset dimension space through a channel mixing module in a time sequence mixer to obtain a space mixing characteristic; Establishing time dependency relations among unmanned aerial vehicle dynamic characteristics in different time periods according to the spatial mixing characteristics through a time mixing module in a time sequence mixer; and taking the spatial mixing characteristic and the time dependence as the space-time characteristic.
5. The method of claim 4, wherein said establishing time-dependent relationships between unmanned aerial vehicle dynamic characteristics for different time periods based on said spatial mixing characteristics comprises: calculating a query matrix, a key matrix and a value matrix corresponding to the spatial mixing characteristics by using a pre-trained weight matrix; According to the query matrix and the key matrix, calculating the attention weight among different time step characteristics; weighting and fusing the value matrix based on the attention weight to obtain fused time characteristics; And performing multi-head splicing on the fused time features to obtain time dependency relations among unmanned aerial vehicle dynamic features in different time periods.
6. The method of claim 1, wherein mapping the spatiotemporal features to a threat prototype space, determining a threat level of the target drone from a prototype classification preset inside the threat prototype space, comprises: Acquiring a prototype classification preset in the threat prototype space; constructing a query sample by utilizing the space-time characteristics, and calculating the similarity between the query sample and each prototype classification; Taking the category prototype with the minimum similarity as a target category prototype to which the query sample belongs; And determining the target threat level corresponding to the target class prototype as the threat level of the target unmanned aerial vehicle based on the mapping relation between the prototype classification and the preset threat level.
7. The method of claim 1, wherein prior to mapping the spatiotemporal features to threat prototype space, the method further comprises: Acquiring unmanned aerial vehicle samples corresponding to a plurality of preset threat levels; extracting space-time characteristics of each unmanned aerial vehicle sample to obtain a plurality of sample feature vectors; calculating the average value of sample feature vectors belonging to the preset threat levels aiming at each preset threat level, and taking the average value as a class prototype corresponding to the preset threat level; and constructing the threat prototype space by utilizing the category prototype and the preset threat level.
8. An unmanned aerial vehicle threat assessment apparatus, the apparatus comprising: The detection module is used for detecting the target unmanned aerial vehicle with the invasion risk in the preset area; The acquisition module is used for acquiring threat elements of the target unmanned aerial vehicle, and extracting space-time characteristics of the threat elements through a time sequence mixer, wherein the threat elements at least comprise dynamic threat elements and static threat elements; and the determining module is used for mapping the space-time characteristics to a threat prototype space and determining the threat level of the target unmanned aerial vehicle according to prototype classification preset in the threat prototype space.
9. A computer device, comprising: a memory and a processor in communication with each other, the memory having stored therein computer instructions which, upon execution, cause the processor to perform the method of any of claims 1 to 7.
10. A computer readable storage medium having stored thereon computer instructions for causing a computer to perform the method of any one of claims 1 to 7.

Description

Unmanned aerial vehicle threat assessment method and device, computer equipment and storage medium Technical Field The invention relates to the technical field of computers, in particular to an unmanned aerial vehicle threat assessment method, an unmanned aerial vehicle threat assessment device, computer equipment and a storage medium. Background The rapid development of unmanned aerial vehicle technology makes the unmanned aerial vehicle technology widely applied to the fields of aerial photography, logistics, agriculture and the like, but brings new challenges of low-altitude safety management and key infrastructure protection. In important places such as nuclear power plants, airports, military bases and the like, illegal unmanned aerial vehicle intrusion can cause serious security threats. The anti-unmanned aerial vehicle system generally comprises links such as detection, identification, threat assessment and interception, wherein the threat assessment is a core link and directly determines the reasonable allocation of defensive resources and the effectiveness of a response strategy. At present, unmanned aerial vehicle threat assessment mainly depends on methods such as expert experience rules, uncertainty reasoning, data driving models or multi-criterion optimization, and the like, and the methods have certain effects under the conditions of sufficient samples and stable scenes, but still have obvious limitations. The existing threat assessment method generally faces the following problems that firstly, the assessment method relies on a large number of historical samples for training, and in practice, unmanned aerial vehicle attack events have scarcity and burstiness, so that model accuracy is remarkably reduced under the condition of few samples, secondly, the traditional method is high in calculation complexity and poor in instantaneity in a dynamic complex environment, and is difficult to meet the instant response requirement on a fast-invasive unmanned aerial vehicle, thirdly, the assessment model cannot effectively fuse dynamic threat elements (such as motion states and radar features) and static threat elements (such as model types and attack capability) of the unmanned aerial vehicle, so that assessment accuracy is insufficient under a variable attack mode. Disclosure of Invention In view of the above, the embodiment of the invention provides an unmanned aerial vehicle threat assessment method, an unmanned aerial vehicle threat assessment device, computer equipment and a storage medium, so as to solve the problems of low unmanned aerial vehicle threat assessment accuracy and insufficient instantaneity in a dynamic complex scene with few samples in the prior art. In a first aspect, an embodiment of the present invention provides a method for evaluating threat of an unmanned aerial vehicle, where the method includes: detecting a target unmanned aerial vehicle with an invasion risk in a preset area; Acquiring threat elements of the target unmanned aerial vehicle, and extracting space-time characteristics of the threat elements through a time sequence mixer, wherein the threat elements at least comprise dynamic threat elements and static threat elements; mapping the space-time characteristics to a threat prototype space, and determining the threat level of the target unmanned aerial vehicle according to prototype classification preset in the threat prototype space. Further, the acquiring the threat element of the target unmanned aerial vehicle includes: acquiring monitoring data of the target unmanned aerial vehicle in the preset area; Extracting a situation threat element and a radar threat element of the target unmanned aerial vehicle from the monitoring data, and taking the situation threat element and the radar threat element as the dynamic threat element; And identifying physical attribute elements of the target unmanned aerial vehicle from the monitoring data, and taking the physical attribute elements as the static threat elements. Further, the situation threat element comprises at least one of speed, course angle, climbing angle, altitude, radial distance, horizontal distance, azimuth angle and elevation angle, the radar threat element comprises at least one of pulse width, working frequency, pulse repetition frequency and radar cross-sectional area, and the physical attribute element comprises at least one of flyer type, attack capability and maneuverability. Further, the extracting, by the time series mixer, the spatiotemporal features of the threat element includes: mapping the threat elements to a preset dimension space through a channel mixing module in a time sequence mixer to obtain a space mixing characteristic; Establishing time dependency relations among unmanned aerial vehicle dynamic characteristics in different time periods according to the spatial mixing characteristics through a time mixing module in a time sequence mixer; and taking the spatial mixing characteristic and the