An ai large model-based countermeasure system and a countermeasure method

Abstract

This invention discloses an AI-based large-scale model-based drone countermeasure system and method, including the following steps: detecting suspicious targets by monitoring the airspace with millimeter-wave radar; identifying and locating suspicious targets using a visual detection device, collecting signal information of suspicious targets using an radio frequency module, and collecting noise information of suspicious targets using an acoustic sensor; determining whether the suspicious target is a drone based on the image information identified by the visual detection device, the signal information collected by the radio frequency module, and the noise information collected by the acoustic sensor; if so, collecting the three-dimensional trajectory of the drone; obtaining the threat level of the drone based on the three-dimensional trajectory of the drone and the environmental information in which the drone is located; and taking corresponding countermeasures to counter the drone based on the threat level. By fusing four-dimensional data (radar, vision, radio frequency, and acoustic signature) and combining it with deep reasoning using an AI large-scale model, the ability to distinguish between disguised drones and jamming objects is significantly improved.

Description

Technical Field

[0001] This invention relates to the field of low-altitude economy, and in particular to a countermeasure system and method based on AI large model. Background Technology

[0002] The civilian drone market is experiencing explosive growth. Industry statistics show that the global number of civilian drones has exceeded ten million, widely used in aerial photography, logistics, agricultural plant protection, power line inspection, and other fields. However, the widespread adoption of this technology has also brought serious safety hazards. Unauthorized drone flights are frequent, including unauthorized flights around airports, reconnaissance of sensitive areas, and the dropping of dangerous items, seriously threatening public safety and social stability.

[0003] Currently, mainstream drone countermeasures typically use millimeter-wave radar to monitor moving targets in the airspace, but it is difficult to effectively distinguish drones from other non-cooperative aerial targets (such as birds, kites, balloons, etc.), leading to frequent false alarms. Summary of the Invention

[0004] Therefore, there is a need to provide an AI-based large-scale model-based countermeasure system and method to address the problem that existing drone countermeasure methods are unable to effectively distinguish drones from other non-cooperative aerial targets.

[0005] To achieve the above objectives, the inventors provide a method for countering unmanned aerial vehicles (UAVs), comprising the following steps:

[0006] Suspicious targets were detected by monitoring the airspace using millimeter-wave radar.

[0007] Suspicious targets are identified and located using a visual inspection device, signal information of suspicious targets is collected using a radio frequency module, and noise information of suspicious targets is collected using a voiceprint sensor.

[0008] Based on the image information identified by the visual inspection device, the signal information collected by the radio frequency module, and the noise information collected by the acoustic sensor, it is determined whether the suspicious target is a drone. If so, the three-dimensional trajectory of the drone is collected.

[0009] The threat level of a drone is determined based on its three-dimensional trajectory and the environmental information it is in.

[0010] Take appropriate countermeasures based on the threat level to counter drones.

[0011] Furthermore, when taking appropriate countermeasures against drones based on threat levels, the following steps are also included:

[0012] If the threat level is high, the drone's communication link will be disrupted through radio jamming or protocol interference.

[0013] Furthermore, when taking appropriate countermeasures against drones based on threat levels, the following steps are also included:

[0014] If the threat level is extremely high, the drone will be jammed using a laser emitter or a microwave emitter.

[0015] Furthermore, it also includes the following steps:

[0016] Determine whether the drone's communication link has been successfully interfered with or cracked. If not, interfere with the drone using a laser transmitter or microwave transmitter.

[0017] Furthermore, when determining the threat level of a drone based on its three-dimensional trajectory and the environmental information it is in, the following steps are also included:

[0018] Continuously track drones;

[0019] Determine whether the drone is flying around sensitive areas or suddenly accelerating; if so, the threat level is determined to be high or extremely high.

[0020] Furthermore, when taking appropriate countermeasures against drones based on threat levels, the following steps are also included:

[0021] If the threat level is medium, intercept aircraft to drive away the drone or launch a fishing net.

[0022] Furthermore, when determining whether a suspicious target is a drone based on image information identified by the visual inspection device, signal information collected by the radio frequency module, and noise information collected by the acoustic sensor, the following steps are also included:

[0023] By using a recognition model based on the Transformer architecture to identify the model, color, and flight status of suspicious targets from image information, signal information, and noise information, drones can be distinguished from animals / balloons.

[0024] Furthermore, when determining the threat level of a drone based on its three-dimensional trajectory and the environmental information it is in, the following steps are also included:

[0025] The Deep Q-Network-based inference model infers the drone's intent and assesses the corresponding threat level.

[0026] Furthermore, the visual inspection device is a hyperspectral photoelectric sphere machine.

[0027] To achieve the above objectives, the inventors also provide a drone countermeasure system, including a storage medium on which a computer program is stored. When the computer program is executed by a processor, it implements the drone countermeasure method as described in any of the above embodiments.

[0028] Unlike existing technologies, the above-mentioned technical solution significantly improves the ability to distinguish between camouflaged drones and jamming objects by fusing four-dimensional data from radar, vision, radio frequency, and acoustic signature, combined with deep reasoning from a large AI model.

[0029] The above description of the invention is merely an overview of the technical solution of this application. In order to enable those skilled in the art to better understand the technical solution of this application and to implement it based on the description and drawings, and to make the above-mentioned objectives and other objectives, features and advantages of this application easier to understand, the following description is provided in conjunction with the specific embodiments and drawings of this application. Attached Figure Description

[0030] The accompanying drawings are only used to illustrate the principles, implementation methods, applications, features, and effects of specific embodiments of the present invention and other related contents, and should not be considered as limitations on this application.

[0031] Figure 1 This is a flowchart of the drone countermeasure method in this application;

[0032] Figure 2 This is one of the flowcharts for a specific implementation of the countermeasure method in this application;

[0033] Figure 3 The second flowchart is a specific implementation of the countermeasure method in this application;

[0034] Figure 4 The third flowchart is a specific implementation of the countermeasure method in this application;

[0035] Figure 5 This is an architecture diagram of the drone countermeasure method in this application;

[0036] Figure 6 This is a schematic diagram of the drone countermeasure system in this application;

[0037] Figure 7 This is a schematic diagram of the drone countermeasure device in this application.

[0038] Explanation of reference numerals in the attached figures:

[0039] 1. Unmanned Aerial Vehicle Countermeasure; 11. Frame; 12. Millimeter-wave Radar; 13. Visual Inspection Device; 14. Countermeasure Components; 2. Monitoring and Management Platform; 3. Network Switch; 4. Storage Server; 5. Browser Terminal; 6. Display; 7. Video Wall; 8. Radio Frequency Module; 9. Voiceprint Sensor. Detailed Implementation

[0040] To illustrate the possible application scenarios, technical principles, implementable specific solutions, and achievable objectives and effects of this application in detail, the following description, in conjunction with the listed specific embodiments and accompanying drawings, provides a detailed explanation. The embodiments described herein are merely illustrative of the technical solutions of this application and are therefore intended to limit the scope of protection of this application.

[0041] In this document, the term "embodiment" means that a specific feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The term "embodiment" appearing in various places throughout the specification does not necessarily refer to the same embodiment, nor does it specifically limit its independence or connection with other embodiments. In principle, in this application, as long as there are no technical contradictions or conflicts, the technical features mentioned in each embodiment can be combined in any way to form corresponding implementable technical solutions.

[0042] Unless otherwise defined, the technical terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the use of related terms herein is merely for the purpose of describing particular embodiments and is not intended to limit this application.

[0043] In the description of this application, the term "and / or" is used to describe the logical relationship between objects, indicating that three relationships can exist. For example, A and / or B means: A exists, B exists, and A and B exist simultaneously. Additionally, the character " / " in this document generally indicates that the preceding and following objects have an "or" logical relationship.

[0044] In this application, terms such as “first” and “second” are used only to distinguish one entity or operation from another, and do not necessarily require or imply any actual quantity, hierarchy or order relationship between these entities or operations.

[0045] Unless otherwise specified, the use of terms such as “comprising,” “including,” “having,” or other similar expressions in this application is intended to cover non-exclusive inclusion, which does not exclude the presence of additional elements in a process, method, or product that includes the stated elements, such that a process, method, or product that includes a list of elements may include not only those defined elements but also other elements not expressly listed, or elements inherent to such a process, method, or product.

[0046] As understood in the Examination Guidelines, in this application, expressions such as "greater than," "less than," and "exceeding" are understood to exclude the stated number; expressions such as "above," "below," and "within" are understood to include the stated number. Furthermore, in the description of the embodiments in this application, "multiple" means two or more (including two), and similar expressions related to "multiple" are also understood in this way, such as "multiple groups" and "multiple times," unless otherwise explicitly specified.

[0047] In the description of the embodiments of this application, the space-related expressions used, such as "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "vertical," "top," "bottom," "inner," "outer," "clockwise," "counterclockwise," "axial," "radial," and "circumferential," indicate the orientation or positional relationship based on the orientation or positional relationship shown in the specific embodiments or drawings. They are only for the purpose of describing the specific embodiments of this application or for the reader's understanding, and do not indicate or imply that the device or component referred to must have a specific position, a specific orientation, or be constructed or operated in a specific orientation. Therefore, they should not be construed as limitations on the embodiments of this application.

[0048] Unless otherwise expressly specified or limited, the terms "installation," "connection," "linking," "fixing," and "setting," as used in the description of the embodiments of this application, should be interpreted broadly. For example, "connection" can be a fixed connection, a detachable connection, or an integral setting; it can be a mechanical connection, an electrical connection, or a communication connection; it can be a direct connection or an indirect connection through an intermediate medium; it can be the internal connection of two components or the interaction between two components. For those skilled in the art to which this application pertains, the specific meaning of the above terms in the embodiments of this application can be understood according to the specific circumstances.

[0049] Please see Figure 1 This embodiment provides a method for countering drones based on AI large-scale models, including the following steps:

[0050] Step S101: Suspicious targets are detected by monitoring the airspace using millimeter-wave radar. Millimeter-wave radar enables non-contact, long-range detection of UAVs, and is particularly suitable for stable operation under complex weather conditions such as nighttime and rain / fog. The radar host integrates a signal processing unit, a data acquisition module, and control logic for transmitting and receiving millimeter-wave / microwave signals. Its antenna emits electromagnetic waves, and the distance, speed, and azimuth information of the target are obtained by analyzing the reflected echoes.

[0051] Step S102: Identify and locate suspicious targets using a visual inspection device, collect signal information of suspicious targets using an radio frequency module, and collect noise information of suspicious targets using an acoustic sensor.

[0052] The visual inspection device can be a hyperspectral PTZ camera, integrating a high-definition visible light / infrared dual-mode camera with autofocus, zoom, and night vision enhancement capabilities. It can capture images of suspicious targets and, in conjunction with AI recognition models, assist in identifying their model, color, flight status, and other characteristics. The RF module supports wide-bandwidth scanning to capture drone remote control and image transmission signals, supporting mainstream protocols such as WiFi, OcuSync (DJI's wireless transmission technology), and Lightbridge. The acoustic sensor includes an array of multiple high-sensitivity microphones, which can, for example, collect the noise spectrum generated by the drone propellers and identify its type through acoustic features, with a long effective range.

[0053] Step S103: Based on the image information identified by the visual inspection device, the signal information collected by the radio frequency module, and the noise information collected by the acoustic sensor, determine whether the suspicious target is a drone. If yes, proceed to step S104 to collect the three-dimensional trajectory of the drone. If no, continue monitoring until the suspicious target appears.

[0054] Step S105: Obtain the threat level of the drone based on its three-dimensional trajectory and the environmental information of the drone.

[0055] Step S106: Take appropriate countermeasures to counter the drone based on the threat level.

[0056] The millimeter-wave radar continuously scans the airspace it operates in. Once it detects a moving target, it identifies it as a suspicious target and sends basic information such as the target's position, distance, speed, and trajectory to the main control circuit board. The main control circuit board then triggers other sensors to work in coordination.

[0057] The system controls the visual inspection device to turn towards the direction of the suspicious target, enabling precise identification of the target type (such as rotor structure, aircraft outline, etc.) and accurate positioning; it controls the radio frequency module to perform radio frequency spectrum scanning on the location of the suspicious target, and collect radio frequency characteristics such as its remote control signal and image transmission signal; and it controls the acoustic fingerprint sensor to pick up the unique noise spectrum generated by the target during flight (such as acoustic fingerprints such as motor speed and propeller resonant frequency).

[0058] The three types of information mentioned above—image information, radio frequency signal information, and acoustic noise information—are aggregated by the main control circuit board and uploaded to the monitoring and management platform in real time. Based on a pre-set UAV feature database or algorithm, the monitoring and management platform comprehensively analyzes this heterogeneous data. For example, if a target simultaneously possesses a multi-rotor structure (photoelectric identification), a specific ISM band frequency-hopping signal (radio frequency characteristic), and high-frequency propeller noise (acoustic noise), it can be identified as a UAV with high confidence. Birds, on the other hand, typically have no radio frequency signal, their acoustic spectrum is continuous and lacks fixed structural features, and balloons have no dynamic noise and their movement trajectory is dominated by wind, thus effectively eliminating them from the list.

[0059] Once a suspicious target is confirmed to be an illegal drone, the monitoring and management platform pre-sets the threat level, such as low, medium, high, and very high. It then sends an activation command to the corresponding countermeasure component 14 through the main control circuit board. The monitoring and management platform can execute the corresponding countermeasure strategy according to the threat level, implement targeted countermeasures, and force the drone to return, hover, or land.

[0060] In some embodiments, when taking appropriate countermeasures to counter the drone according to the threat level in step S106, the following steps are also included:

[0061] If the threat level is high, the communication link of the drone will be disrupted by radio interference or protocol. When the threat level of a drone is determined to be "high", radio interference will be initiated. This is a soft kill control method that uses non-physical destruction methods such as cutting off the drone's communication link to interfere with, deceive, or take over the target drone's communication, navigation, or control system, achieving a countermeasure effect with no damage and low collateral damage. It is suitable for areas with extremely high security requirements, such as cities, airports, and sensitive facilities.

[0062] If the threat level is extremely high, the drone will be jammed using laser or microwave transmitters. When a drone is determined to have an "extremely high" threat level, such as carrying explosives or engaging in high-speed penetration, weapons such as high-energy laser transmitters or high-power microwave transmitters will be used to physically destroy or capture it to directly eliminate the threat. This is considered hard-kill control and ensures safety.

[0063] If the threat level is medium, interceptor aircraft can be used to drive away drones or fishing nets can be launched. Multiple countermeasure resources (fixed stations, mobile platforms, interceptor aircraft, etc.) can be coordinated to achieve multi-node, multi-method spatiotemporal collaborative interception, enhancing the ability to respond to high-speed, clustered, and slow-moving small targets.

[0064] If the threat level is low, for example, if the drone is flying slowly outside the protected area, has not entered a sensitive area, has no abnormal maneuvering behavior, and its signal characteristics match those of a legitimate consumer-grade drone, then a non-interventionist approach, primarily based on monitoring and recording, will be adopted.

[0065] The system supports a closed-loop feedback mechanism for execution results, transmitting countermeasure status (success / failure / escape) in real time for strategy iteration and model optimization.

[0066] In some embodiments, the drone countermeasure method further includes the following steps:

[0067] Determine if the drone's communication link has been successfully interfered with or compromised. If not, interfere with the drone using a laser or microwave transmitter. If soft kill fails, further countermeasures can be taken using hard kill control.

[0068] In some embodiments, when obtaining the threat level of the UAV based on the UAV's three-dimensional trajectory and the environmental information of the UAV in step S105, the following steps are also included:

[0069] Continuously track drones;

[0070] Determine whether the drone is flying around sensitive areas or suddenly accelerating; if so, the threat level is determined to be high or extremely high.

[0071] After initially obtaining the drone's three-dimensional trajectory, the monitoring and management platform does not immediately fix its threat level. Instead, it initiates a continuous tracking mechanism, updating its position, speed, heading, and acceleration data in real time. During this process, if the platform detects that the drone is circling around a sensitive area, making multiple turns, or lingering for extended periods (i.e., "circling"), or if its speed suddenly increases beyond a preset threshold (i.e., sudden acceleration), the monitoring and management platform raises the threat level of the target to "high" or "extremely high".

[0072] Please see Figure 5 In some embodiments, the monitoring and management platform is configured with a countermeasure execution and control unit. This unit translates decision-making instructions into concrete physical actions and ensures effective execution through closed-loop feedback. It primarily consists of three modules: multi-means coordinated countermeasures, precise energy management, and coordinated interception scheduling. This enables soft-kill and hard-kill control, dynamically adjusting and controlling the power of energy-intensive devices such as lasers and microwaves through closed-loop control. It also intelligently schedules the takeoff of countermeasure drones to perform tasks such as escorting, driving away, and launching capture nets. Sub-modules include: soft-kill control, hard-kill control, coordinated interception, and energy management.

[0073] In some embodiments, the drone countermeasure method further includes an energy management step:

[0074] Fine-grained power scheduling and energy allocation are performed on high-energy-consuming countermeasure components 14 (such as laser emitters, microwave emitters, and interceptor aircraft) to ensure the stability and sustainability of the system during continuous operations.

[0075] In some embodiments, when determining whether a suspicious target is a drone based on the image information identified by the visual inspection device, the signal information collected by the radio frequency module, and the noise information collected by the acoustic sensor in step S103, the following steps are also included:

[0076] By using a recognition model based on the Transformer architecture to identify the model, color, and flight status of suspicious targets from image information, signal information, and noise information, drones can be distinguished from animals / balloons.

[0077] Please see Figure 5The monitoring and management platform is equipped with an AI-enhanced recognition and behavior analysis unit. After acquiring radar point cloud, image information, radio frequency signal information, and acoustic noise information, it inputs a multimodal recognition model based on the Transformer architecture. The recognition model outputs fine-grained discrimination results for suspicious targets by leveraging the correlation features between different modalities. The Transformer architecture recognition model performs deep mining of perceived data, achieving a leap from "seeing the target" to "understanding the target." High-precision target recognition can accurately classify low-altitude, slow-moving small targets such as drones, birds, kites, and balloons, effectively reducing false alarm rates. It improves anti-camouflage and anti-interference recognition capabilities. Dynamic behavior modeling and tactical intent prediction enable trajectory tracking and prediction, swarm behavior analysis, and other functions, effectively improving the prediction of drone behavior and advance action deployment. The recognition model includes sub-modules such as target classification, aircraft intent prediction, and swarm flight behavior modeling.

[0078] 1) Target Classification: The target classification submodule is responsible for high-precision identification and classification of low-altitude, slow-moving small targets (such as drones, birds, kites, balloons, etc.), which is the basic capability of the system to achieve "clear vision and accurate classification".

[0079] 2) Intent prediction of drones: This module models the historical trajectory, flight status, environmental context and other information of individual aircraft to predict their future behavioral intentions (such as reconnaissance, penetration, hovering, return to base, attack, etc.), providing early warning for defense decisions.

[0080] 3) Swarm flight behavior modeling: For swarms of multiple UAVs (such as bee swarms or formations), this module understands their organizational logic and collective intent through swarm dynamics analysis, cooperative pattern recognition, and topology inference, supporting the early identification of large-scale cooperative threats and the generation of countermeasure strategies.

[0081] 4) Digital Twin Pre-simulation Module: Simulates the countermeasure process in a virtual environment, optimizes laser path, interference parameters and interception strategies, and supports real-time simulation and deduction.

[0082] In some embodiments, when obtaining the threat level of the UAV based on the UAV's three-dimensional trajectory and the environmental information of the UAV in step S105, the following steps are also included:

[0083] The Deep Q-Network-based inference model infers the drone's intent and assesses the corresponding threat level.

[0084] After acquiring the UAV's 3D trajectory (including position, velocity, acceleration, and heading angle sequence) and its environmental information (such as proximity to sensitive areas or no-fly zones), the monitoring and management platform inputs this time-series data into an intent reasoning model based on Deep Q-Network (DQN). This model, trained offline, has learned multiple typical tactical behavior patterns and can identify the most likely attack intent type corresponding to the UAV's current behavior. For example, low-altitude hovering corresponds to "reconnaissance," high-speed straight-line approach corresponds to "impact," hovering and then descending corresponds to "payload dropping," and multi-UAV coordinated approach corresponds to "swarm attack." The model outputs the probability distribution of this intent and maps it to a preset four-level threat level (low, medium, high, and extremely high). For example, "reconnaissance" intent corresponds to "medium" threat, and "impact" corresponds to "extremely high" threat.

[0085] Please see Figure 5 In some embodiments, the monitoring and management platform is configured with a core control unit, which includes the following core functional modules:

[0086] 1) AI-enhanced target recognition module: It adopts a cross-modal attention mechanism to fuse radar point cloud, visual image, radio frequency spectrum and voiceprint data, and inputs them into the AI ​​large model (based on Transformer architecture) for end-to-end recognition, supporting the recognition of disguised drones and the differentiation of birds / balloons.

[0087] 2) Drone behavior analysis module: continuously tracks the target's three-dimensional trajectory, identifies multiple flight modes such as hovering, circling, and serpentine maneuvers, detects abnormal behaviors (such as flying around sensitive areas, sudden acceleration), and has a response time of ≤1s.

[0088] 3) Drone Intent Prediction Module: Combining payload estimation, flight path and tactical mode, it uses a deep reinforcement learning model (DQN) to infer attack intent (reconnaissance, throwing, impact, swarm attack, etc.) and outputs a four-level threat level (low / medium / high / extremely high).

[0089] In particular, the AI-enhanced recognition and behavior analysis module can leverage large AI models to deeply mine perceived data, achieving a leap from "seeing the target" to "understanding the target." Through high-precision target recognition, it can accurately classify low-altitude, slow-moving small targets such as drones, birds, kites, and balloons, effectively reducing false alarm rates. It improves anti-camouflage and anti-interference recognition capabilities; dynamic behavior modeling and tactical intent prediction enable trajectory tracking and prediction, swarm behavior analysis, and other functions, effectively improving the prediction of drone behavior and advance action deployment. This includes sub-modules such as target classification, aircraft intent prediction, and modeling of swarm flight behavior.

[0090] Please see Figure 5In some embodiments, the monitoring and management platform is configured with a countermeasure strategy generation and decision-making unit. This unit performs command functions and can intelligently generate optimal countermeasure plans based on the nature of the threat and environmental constraints, achieving precise, efficient, and compliant handling. Dynamic threat assessment determines the threat level by constructing a multi-factor assessment model. A built-in countermeasure strategy knowledge base and a decision engine combining the Analytic Hierarchy Process (AHP) and Reinforcement Learning (RL) are used to generate combined strategies. Simultaneously, through digital twin pre-simulation and optimization, the generated strategies are simulated and extrapolated in a virtual environment in real time. This includes sub-modules such as threat assessment model, countermeasure matching, digital twin pre-simulation, and dynamic strategy updates.

[0091] 1) Threat Assessment Model: This module is responsible for performing multi-dimensional quantitative analysis of the input threat intelligence and outputting a structured threat level score to provide a basis for subsequent strategy generation.

[0092] 2) Countermeasure matching: Based on the threat assessment results, the system intelligently matches and combines the optimal response plan from the built-in countermeasure strategy knowledge base.

[0093] 3) Digital twin simulation: Conduct high-fidelity simulation and deduction of candidate countermeasure strategies in a virtual environment to verify their effectiveness, side effects and resource feasibility.

[0094] 4) Dynamic strategy update: Based on simulation results, actual execution feedback and environmental changes, continuously optimize the countermeasure strategy library and decision logic to achieve closed-loop self-adaptation.

[0095] Please see Figures 6 to 7 This embodiment also provides an AI-based large-scale model drone countermeasure system, including a storage medium on which a computer program is stored. When the computer program is executed by a processor, it implements the drone countermeasure method as described in any of the above embodiments.

[0096] Please see Figures 6 to 7The drone countermeasure system includes at least one drone countermeasure device and a monitoring and management platform. The drone countermeasure device includes a frame 11 and a millimeter-wave radar 12, a visual detection device 13, an radio frequency module 8, an acoustic sensor 9, and countermeasure components 14 mounted on the frame 11. The millimeter-wave radar is used to monitor the airspace and detect suspicious targets. The visual detection device is used to identify and locate suspicious targets. The radio frequency module is used to collect signal information of suspicious targets. The acoustic sensor is used to collect noise information of suspicious targets. The main control circuit board sends the above information to the monitoring and management platform. The monitoring and management platform is used to determine whether a suspicious target is a drone based on the image information identified by the visual detection device, the signal information collected by the radio frequency module, and the noise information collected by the acoustic sensor. If so, it collects the three-dimensional trajectory of the drone, obtains the threat level of the drone based on the three-dimensional trajectory and the environmental information of the drone, and takes corresponding countermeasures to counter the drone based on the threat level through the countermeasure components.

[0097] It should be noted that the main control circuit board includes a central processing unit and a power management module, responsible for data processing, sensor coordination control, and power scheduling of the UAV countermeasures unit; the network switching module is used to realize high-speed data communication between various functional units; the power supply unit supports AC 220V mains input and is equipped with a backup battery interface to connect to an emergency DC power supply, ensuring that the system can still maintain critical functions when the main power is interrupted. The network switching module is equipped with communication antennas, which are distributed on each telescopic arm, for establishing a two-way communication connection with a remote monitoring and management platform. The system supports wired transmission (via RJ45 network port) and wireless transmission (including Wi-Fi, 4G, or 5G), and can upload detection data, hyperspectral video streams, radio frequency signal characteristics, and acoustic fingerprint information to the monitoring and management platform in real time, and simultaneously receive control commands from the platform to realize remote coordinated scheduling of the countermeasures components and various sensors.

[0098] Please see Figure 6In some embodiments, the drone countermeasure system also includes a network switch 3, a storage server 4, and a display 6. The network switch is connected to the main control circuit board, the storage server, the monitoring and management platform 2, the display, and the main control circuit board, respectively. The network switch is the communication hub of the entire system, connecting all terminals, including: multiple drone countermeasures, the monitoring and management platform, the storage server, and an IE browser terminal 5 (for remote access). It provides a high-speed, stable data transmission channel, supports the gigabit Ethernet standard, and ensures smooth processing of high-concurrency data streams. The monitoring and management platform is deployed on a dedicated computer and is responsible for receiving real-time data from various sensors (such as target location, flight trajectory, signal strength, etc.) and performing comprehensive analysis and decision-making. It has task scheduling capabilities and can remotely issue commands to designated countermeasures to activate or deactivate interference functions. It supports multi-user permission management to ensure operational security. The storage server is responsible for long-term storage of system operation logs, detection records, interference event recordings, and related alarm information. It supports retrieval by time, location, type, and other conditions to meet post-event traceability and auditing needs. It is usually configured with a RAID array and backup mechanism to ensure data security and reliability. The display is used by local operators to view system status, alarm information, and the control interface. Multiple monitors 6 can form a video wall 7, centrally displaying the working images, map status, and real-time video streams of multiple countermeasures, facilitating overall monitoring and command and dispatch. Administrators can access the monitoring and management platform via a web browser or client application for remote monitoring and operation. Administrators can log in to the system from different locations, improving response efficiency and flexibility.

[0099] This system is a network-based drone countermeasure management system. Through the collaborative operation of multiple drone countermeasure devices and a central monitoring and management platform, it effectively identifies, locates, and interferes with illegal or unauthorized drones. The system adopts a combination of centralized management and distributed deployment, and features real-time monitoring, data storage, remote control, and visualization capabilities.

[0100] Each sensor (millimeter-wave radar, hyperspectral photoelectric PTZ camera, RF module, acoustic sensor, etc.) transmits data with the central monitoring and management platform through the main control circuit board. It supports wired transmission, 4G / 5G, WiFi or Mesh private network, and adopts lightweight IoT protocols such as MQTT / CoAP to ensure low latency and high reliability data transmission.

[0101] The central monitoring and management platform's system management and human-machine interaction unit provides users with an intuitive and convenient operating interface and comprehensive operation and maintenance support, ensuring stable and reliable system operation. Through an integrated command and control console, it enables comprehensive system monitoring and complete data management and auditing, greatly improving the collaborative nature of human-machine interaction. It also allows for flexible remote operation and maintenance, facilitating further management of UAVs. Sub-modules include: human-machine collaboration interface, logs and auditing, system status monitoring, and remote configuration and auditing.

[0102] 1) Human-machine interface: This module provides an intuitive, user-friendly and responsive operating interface, enabling operators to interact efficiently with the countermeasures system and achieve rapid identification, assessment and handling of drone threats.

[0103] 2) Logs and Audits: Comprehensively record all kinds of operational behaviors, event triggers, equipment status changes and security events during system operation, providing a basis for post-event analysis, accountability and compliance review.

[0104] 3) System status monitoring: Real-time collection and display of the operating status of each hardware and software component of the countermeasure system, realizing dynamic perception of the overall health and fault early warning.

[0105] 4) Remote configuration and auditing: Allows authorized users to adjust parameters, update policies and upgrade functions of countermeasures devices deployed remotely through a secure channel, while ensuring that all remote operations can be fully recorded and audited.

[0106] This drone countermeasure system can be deployed around sensitive areas, such as airport no-fly zones, over nuclear power plants, government and military controlled areas, and communication hub buildings. After system startup, the sensors in the perception layer continuously scan the airspace. When the millimeter-wave radar detects a moving target (with a certain level of confidence), it immediately triggers the photoelectric PTZ camera to turn and position, the radio frequency module to scan signals, and the acoustic fingerprint array to collect noise, achieving multi-source collaborative observation. The collected data is transmitted to the core control unit via the communication layer. The AI ​​large model integrates multimodal data for end-to-end identification. If it is determined to be a "suspected drone," it enters the behavior analysis module to continuously track its three-dimensional trajectory. The intent prediction module combines flight path and environmental information, using a DQN (deep learning and reinforcement learning network model) model to infer its attack intent and assess the threat level. If the threat level is "high" or "extremely high," the countermeasure strategy module calls the strategy knowledge base, combining weights and reinforcement learning to output the optimal countermeasure combination, such as "radio jamming + laser strike." After the strategy is confirmed by the human-machine interface, the countermeasure execution module is activated: first, a soft kill is implemented to cut off communication; if this is ineffective, a laser is activated to implement a hard kill.

[0107] This application aims to address the following core issues existing in current drone countermeasure systems:

[0108] 1) Limited perception capabilities and high false alarm rate: Existing systems often rely on a single sensor (such as radar), which makes it difficult to effectively distinguish between drones and non-threat targets such as birds and balloons, resulting in frequent false alarms and affecting system availability.

[0109] 2) Insufficient identification accuracy, making it difficult to deal with advanced threats: There is a lack of effective identification methods for disguised drones and signal jamming drones, making it impossible to accurately determine the target type and potential threat level.

[0110] 3) Low level of intelligence and lagging decision-making: There is a lack of in-depth analysis of the behavior patterns and attack intentions of drones. Countermeasures rely on preset rules and cannot be dynamically adjusted according to the real-time situation, making it difficult to cope with complex attack scenarios.

[0111] 4) High response latency, which cannot meet the needs of actual combat: The entire process from target detection to countermeasure execution usually has a latency in the second range, making it difficult to quickly intercept high-speed moving targets.

[0112] 5) Fragmented countermeasures and lack of coordinated optimization: Soft kill (jamming, protocol cracking) and hard kill (laser, microwave, capture net) methods operate independently, lacking unified scheduling and strategy combination, resulting in low countermeasure efficiency.

[0113] To address the above problems, this invention provides a drone countermeasure system and method, achieving the following technical objectives:

[0114] 1) Achieve high-precision detection and identification of low-altitude drones, effectively distinguishing drones from interfering objects such as birds and balloons;

[0115] 2) Based on AI large-scale models, realize drone behavior analysis and attack intent prediction, and support multi-target trajectory prediction and tactical intent reasoning;

[0116] 3) Construct a dynamic countermeasure strategy generation engine, combining reinforcement learning and game theory to generate the optimal soft and hard countermeasure solution;

[0117] 4) Supports the coordinated execution of multiple methods, including radio interference, protocol cracking, high-energy laser, microwave suppression, capture net launch, and counter-drone interception;

[0118] 5) Achieve edge intelligent computing and millisecond-level response to ensure low latency throughout the entire process from detection to countermeasure;

[0119] 6) Meets national radio management and airspace regulation compliance requirements and has commercial deployment capabilities.

[0120] Please see Figure 2 , Figure 3 and Figure 4 The following is a specific implementation process for a countermeasure method:

[0121] Step 1: System Startup and Monitoring

[0122] The sensors in the perception layer (millimeter-wave radar, photoelectric PTZ camera, radio frequency module, acoustic pattern array) began to continuously scan the airspace.

[0123] Step 2: The millimeter-wave radar detects a moving target and the confidence level meets the standard.

[0124] Step 3: Trigger multi-source collaborative observation (photoelectric PTZ camera turning and positioning, radio frequency module scanning signal, acoustic pattern array collecting noise).

[0125] Step 4: Judgment and Tracking

[0126] If identified as a "suspected drone," the system will proceed to the behavior analysis module to continuously track its 3D trajectory.

[0127] Step 5: Intent prediction and threat assessment.

[0128] By using the DQN (Deep Q-Network) model to combine flight path and environmental information to infer attack intent and assess threat level, the threat level can be assessed.

[0129] Step 6: High Threat Response

[0130] If the threat level is "high" or "extremely high", the countermeasures module calls the strategy knowledge base and outputs the optimal countermeasures combination (such as radio jamming + laser strike).

[0131] Step 7: Human-Machine Collaboration Confirmation

[0132] The strategy is executed after being confirmed by the human-machine collaboration interface.

[0133] Step 8: Countermeasure Execution

[0134] Soft kill: Cut off communication; if ineffective, hard kill: Activate laser strike.

[0135] Step 9: Results Feedback and Optimization

[0136] Step 10: Mission complete

[0137] The execution results are fed back in real time for strategy iteration and model optimization.

[0138] Compared with existing technologies, the countermeasures method has the following significant advantages:

[0139] 1) Multimodal fusion perception with high recognition accuracy: By fusing four-dimensional data from radar, vision, radio frequency and voiceprint, and combining it with deep reasoning from a large AI model, the ability to distinguish between camouflaged drones and interference objects is significantly improved.

[0140] 2) AI is deeply embedded in the decision-making chain: For the first time, large AI models are applied to the entire process of behavior analysis, intent prediction and countermeasure strategy generation, realizing a leap from "passive response" to "proactive prediction".

[0141] 3) Dynamic intelligent decision-making: Based on the strategy generation mechanism of reinforcement learning and game theory, it supports adaptive countermeasures in multi-objective scenarios.

[0142] 4) Coordinated execution of soft and hard attacks: Integrates five types of countermeasures, supports combined strategies and dynamic resource scheduling, and improves the success rate and flexibility of countermeasures.

[0143] 5) Edge intelligence and low latency: After quantization pruning, AI models can be deployed on edge devices, reducing inference latency and meeting practical needs.

[0144] 6) Strong scalability and compliance: The modular design supports functional expansion, the interface is standardized, it complies with national radio management and airspace supervision requirements, and has the potential for large-scale commercial use.

[0145] Finally, it should be noted that although the above embodiments have been described in the text and drawings of this application, this should not limit the scope of patent protection of this application. Any technical solutions that are based on the essential concept of this application and utilize the content described in the text and drawings of this application, resulting in equivalent structural or procedural substitutions or modifications, as well as the direct or indirect application of the technical solutions of the above embodiments to other related technical fields, are all included within the scope of patent protection of this application.

Claims

1. A drone countermeasure system and method based on an AI large-scale model, characterized in that, Includes the following steps: Suspicious targets were detected by monitoring the airspace using millimeter-wave radar. Suspicious targets are identified and located using a visual inspection device, signal information of suspicious targets is collected using a radio frequency module, and noise information of suspicious targets is collected using a voiceprint sensor. Based on the image information identified by the visual inspection device, the signal information collected by the radio frequency module, and the noise information collected by the acoustic sensor, it is determined whether the suspicious target is a drone. If so, the three-dimensional trajectory of the drone is collected. The threat level of a drone is determined based on its three-dimensional trajectory and the environmental information it is in. Take appropriate countermeasures based on the threat level to counter drones.

2. The method for countering unmanned aerial vehicles according to claim 1, characterized in that, When taking appropriate countermeasures against drones based on threat levels, the following steps are also included: If the threat level is high, the drone's communication link will be disrupted through radio jamming or protocol interference.

3. The method for countering unmanned aerial vehicles according to claim 1, characterized in that, When taking appropriate countermeasures against drones based on threat levels, the following steps are also included: If the threat level is extremely high, the drone will be jammed using a laser emitter or a microwave emitter.

4. The method for countering unmanned aerial vehicles according to claim 2, characterized in that, It also includes the following steps: Determine whether the drone's communication link has been successfully interfered with or cracked. If not, interfere with the drone using a laser transmitter or microwave transmitter.

5. The method for countering unmanned aerial vehicles according to claim 2 or 3, characterized in that, When determining the threat level of a drone based on its three-dimensional trajectory and the environmental information it is in, the following steps are also included: Continuously track drones; Determine whether the drone is flying around sensitive areas or suddenly accelerating; if so, the threat level is determined to be high or extremely high.

6. The method for countering unmanned aerial vehicles according to claim 1, 2, or 3, characterized in that, When taking appropriate countermeasures against drones based on threat levels, the following steps are also included: If the threat level is medium, intercept aircraft to drive away the drone or launch a fishing net.

7. The method for countering unmanned aerial vehicles according to claim 1, characterized in that, When determining whether a suspicious target is a drone based on image information identified by the visual inspection device, signal information collected by the radio frequency module, and noise information collected by the acoustic sensor, the following steps are also included: By using a recognition model based on the Transformer architecture to identify the model, color, and flight status of suspicious targets from image information, signal information, and noise information, drones can be distinguished from animals / balloons.

8. The method for countering unmanned aerial vehicles according to claim 1, characterized in that, When determining the threat level of a drone based on its three-dimensional trajectory and the environmental information it is in, the following steps are also included: The Deep Q-Network-based inference model infers the drone's intent and assesses the corresponding threat level.

9. The method for countering unmanned aerial vehicles according to claim 1, characterized in that, The visual inspection device is a hyperspectral photoelectric sphere machine.

10. A large-scale AI-based drone countermeasure system, comprising a storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the drone countermeasure method as described in any one of claims 1 to 9.

Images