Unmanned vehicle-unmanned aerial vehicle cooperative reconnaissance strike system

Abstract

The invention discloses an unmanned vehicle-unmanned aerial vehicle cooperative reconnaissance and strike system which comprises an unmanned vehicle system, an unmanned aerial vehicle system, a central control system and a remote command system. The unmanned vehicle system and the unmanned aerial vehicle system directly communicate through the data link communication module; the unmanned vehicle system communicates with the remote command system through a satellite communication module; the unmanned vehicle system performs comprehensive perception and data acquisition on the surrounding environment through the multi-mode sensor assembly; the central control system is configured to integrate information of an unmanned vehicle, an unmanned aerial vehicle and a weapon system. The cooperative control module is used for optimizing task allocation and a cooperative process; the situation awareness module and the target recognition module are used for fusing and analyzing multi-modal sensor data in real time to construct a unified battlefield situation map; and the resource optimization module is used for managing system resources. The advantages of all the systems are exerted, and the overall combat ability is improved.

Description

Technical Field

[0001] This invention belongs to the field of fastening equipment technology, and more specifically, relates to an unmanned vehicle-drone collaborative reconnaissance and strike system. Background Technology

[0002] With the rapid development of artificial intelligence and unmanned systems technology, unmanned vehicles and drones have become important equipment in modern military operations and security monitoring. Unmanned vehicles typically undertake ground mobility, reconnaissance, and heavy weapons delivery missions, while drones are mainly responsible for aerial reconnaissance, surveillance, and light strike missions. Working together can fully leverage their respective advantages, compensate for the limitations of a single system, and achieve more efficient mission execution.

[0003] In existing technologies, there are various implementation schemes for autonomous vehicle and drone collaborative systems. For example, one scheme adopts a three-part system architecture of drone-ground station-autonomous vehicle, in which the drone first conducts reconnaissance and transmits target information to the ground station, which then directs the drone to the target location for close-range reconnaissance. Although this scheme achieves basic collaborative reconnaissance functions, the ground station, as an intermediary, is prone to information transmission delays, and the system still requires human intervention for decision-making and lacks autonomous learning capabilities.

[0004] Another approach addresses the issues of drone range and precise landing by mounting the drone on an unmanned vehicle platform, incorporating charging ports and positioning holes. It also includes a vehicle-mounted hydraulic outrigger to adapt to complex terrain and an automatic lighting system to enhance nighttime reconnaissance capabilities. However, this approach focuses heavily on hardware design, resulting in relatively simple control logic, a lack of autonomous decision-making capabilities, and insufficient coordination between the drone and weapon systems.

[0005] Another approach employs a single-operator control architecture, where the drone detects the target and transmits images and location information. The operator then makes decisions and directs the unmanned vehicle to the target location. This approach also incorporates precise positioning and automatic locking mechanisms, as well as a multi-radio communication architecture. However, this system still relies on manual decision-making, lacks self-learning capabilities, and suffers from insufficient integration between the unmanned vehicle and the weapon system.

[0006] In summary, the existing technologies mentioned above share many common problems. In terms of decision-making, existing systems largely rely on manual decision-making or preset rules, making it difficult to cope with complex and ever-changing environments. Regarding system coordination efficiency, the coordination between subsystems is not close enough, information exchange efficiency is low, and it is difficult to maximize combat effectiveness. In terms of autonomous learning capabilities, they cannot learn and optimize decision-making strategies from actual combat experience, resulting in poor adaptability. Regarding weapon system integration, most solutions focus on reconnaissance functions, with insufficient research on the coordinated control of weapon systems with chassis and UAVs. Under conditions of limited communication, the system's autonomous combat capability is limited. Therefore, there is an urgent need for an unmanned vehicle-UAV cooperative reconnaissance and strike system capable of highly autonomous decision-making, close coordinated control, continuous learning and optimization, and maintaining high combat efficiency in complex environments. Summary of the Invention

[0007] This invention aims to provide an unmanned vehicle-drone collaborative reconnaissance and strike system. By coordinating the unmanned vehicle system, drone system, central control system, and remote command system, it solves the problems of high dependence on human decision-making, low collaborative efficiency, and lack of autonomous learning ability in the existing technology, and achieves highly autonomous decision-making and close collaborative control. It is suitable for scenarios such as military reconnaissance and border patrol.

[0008] To address the aforementioned deficiencies or improvement needs of existing technologies, this invention provides a collaborative reconnaissance and strike system for unmanned vehicles and unmanned aerial vehicles, comprising:

[0009] The system is configured for unmanned vehicle systems including ground mobility, reconnaissance and heavy weapon strikes; unmanned aerial vehicle systems including aerial reconnaissance and light strikes; a central control system for decision-making and collaborative control; and a remote command system for mission assignment, monitoring and key action authorization.

[0010] The unmanned vehicle system and the unmanned aerial vehicle system communicate directly through the data link communication module of the communication system in the unmanned vehicle system; the unmanned vehicle system and the remote command system communicate through the satellite communication module of the communication system; the unmanned vehicle system comprehensively perceives and collects data on the surrounding environment through multimodal sensor components;

[0011] The central control system includes a collaborative control module configured to integrate information from unmanned vehicles, drones, and weapon systems, optimize task allocation, and coordinate processes; a situational awareness module and a target recognition module configured to perform real-time fusion and analysis of multimodal sensor data to construct a unified battlefield situation map; and a resource optimization module configured to manage system resources.

[0012] Furthermore, the unmanned vehicle system also includes a main vehicle platform, which adopts a modular design and allows for the replacement of functional modules according to mission requirements.

[0013] Furthermore, the unmanned vehicle system also includes a vehicle-mounted drone station, which uses a mechanically foldable launch platform to fold and hide the drone on the roof when it is in standby mode; the vehicle-mounted drone station automatically locks the drone after it lands using a mechanical locking mechanism; the vehicle-mounted drone station guides the drone to land using a positioning and recovery system; and the vehicle-mounted drone station provides energy to the drone using a fast charging system.

[0014] Furthermore, the multimodal sensor assembly includes a visual sensor configured to acquire visual images of the vehicle's surroundings, a radar sensor configured to detect moving targets within a set range, a lidar for generating a three-dimensional map, an infrared sensor for detecting heat source targets, and an acoustic sensor for capturing ambient sound; the multi-source data acquired by the multimodal sensor assembly is transmitted to the situational awareness module of the central control system.

[0015] Furthermore, the unmanned aerial vehicle system also includes a quadcopter body and foldable wings; the foldable wings are mounted on the quadcopter body and automatically unfold during the drone's cruise and retract during takeoff and landing.

[0016] Furthermore, the situational awareness module fuses various sensor data through a multi-source data fusion submodule; the multi-source data fusion submodule integrates different sensor data by fusing Kalman filtering and particle filtering methods.

[0017] Furthermore, the specific process of fusing the Kalman filter and particle filter methods is as follows:

[0018] Based on the data type and noise characteristics, determine whether the data conforms to the linear Gaussian model, and initialize the parameters of the Kalman filter and particle filter.

[0019] Time synchronization and spatial coordinate transformation of data from different sensors;

[0020] Independent prediction using both Kalman filtering and particle filtering methods;

[0021] The prediction results of Kalman filtering are used as the prior information of particle filtering. The matching degree between each particle and the Kalman predicted state is calculated by Mahalanobis distance. The weight of mismatched particles is reduced, and the particle swarm is concentrated in the high probability region, reducing the number of invalid particles.

[0022] When the particle filter detects nonlinear characteristics in the data, it feeds back to adjust the noise covariance of the Kalman filter to enhance its adaptability to sudden changes.

[0023] Kalman filtering updates the state estimate based on sensor measurements, calculates the Kalman gain, and corrects the prediction error. Particle filtering updates the weight of each particle based on the measurement model, retains high-weight particles and discards low-weight particles through resampling. The estimation errors of the two methods are calculated, and the weights are dynamically allocated according to the magnitude of the error. Finally, the fused state estimate is output by weighted summation.

[0024] Furthermore, the central control system also includes a task planning module configured to generate an optimal task execution plan based on task objectives, environmental conditions, and system status; the task planning module includes a path planning submodule, a time resource allocation submodule, and an emergency plan generation submodule.

[0025] Furthermore, the path planning submodule generates optimal paths that take into account terrain, threats, and energy consumption through an improved A method and a fast expanding random tree method;

[0026] The improved A method and the fast expanding random tree method are as follows: the fast expanding random tree method is used to sample "security critical points" near the threat area to construct a simplified topology map; the A method is used to calculate the optimal path on the topology map to reduce the number of nodes in the discrete grid.

[0027] Method A is improved by reconstructing the cost function, as detailed below:

[0028] f(n) = ω1·g terrain (n)+ω2·g threat (n)+ω3·g energy (n)+h((n)

[0029] Among them, g terrain (n) represents the terrain cost; g threat (n) represents the cost of the threat; g energy (n) represents the energy cost; ω1, ω2, ω3 are weighting coefficients, which must satisfy ω1 + ω2 + ω3 = 1 and are dynamically adjusted according to task requirements; and terrain and threats are introduced to adjust the heuristic estimation terms as follows:

[0030]

[0031] Where, x goal ,y goal These are the x and y coordinates of the target point, respectively; x n ,y n η(n) represents the horizontal and vertical coordinates of the current node n, respectively; η(n) is the environmental impact factor, which is determined based on the terrain complexity and threat level of the current node n.

[0032] As a second aspect of the present invention, a method for coordinated reconnaissance and strike operations using unmanned vehicles and unmanned aerial vehicles is also provided, comprising:

[0033] S1. After receiving the task, the remote command system (400) transmits it to the central control system (300). The central control system (300) completes self-check and resource assessment, and then activates the corresponding decision-making strategy according to the task type.

[0034] S2. Generate an initial mission plan by combining terrain, threat and resource conditions, assign unmanned vehicle and drone missions to optimize energy and reconnaissance efficiency, and formulate collaborative strategies;

[0035] S3. The unmanned vehicle proceeds to conduct reconnaissance along the planned route, and decides whether to release drones for collaborative reconnaissance operations based on the complexity of the environment;

[0036] S4. When a suspicious target is identified, the central control system (300) calculates the value of different actions based on the current status and historical experience, and selects the optimal action to execute; when it involves actions that require manual authorization, including the use of weapons, it requests confirmation from the remote authorization module (430);

[0037] S5. Based on the decision results, the strike mission is issued, the UAV assesses the results after the strike and transmits the information back, and the central control system (300) optimizes the decision strategy accordingly.

[0038] S6. After the mission is completed, the drone returns and is recycled and recharged by the unmanned vehicle and enters standby mode. At the same time, the central control system (300) generates a mission report and sends it back to the remote system.

[0039] In summary, compared with the prior art, the above-described technical solutions conceived by this invention can achieve the following beneficial effects:

[0040] 1. The unmanned vehicle-unmanned aerial vehicle (UAV) collaborative reconnaissance and strike system of the present invention, by setting up an unmanned vehicle system, an UAV system, a central control system and a remote command system, wherein the unmanned vehicle system is responsible for ground mobility, reconnaissance and heavy weapon strikes, the UAV system is responsible for aerial reconnaissance and light strikes, the central control system is responsible for decision-making and collaborative control, and the remote command system is responsible for mission issuance, monitoring and authorization of key actions, realizes a clear division of labor and collaborative cooperation among the functions of each system, enabling the system to efficiently execute reconnaissance and strike missions, give full play to the advantages of each system, and enhance the overall combat capability.

[0041] 2. The unmanned vehicle-unmanned aerial vehicle (UAV) collaborative reconnaissance and strike system of the present invention enables direct communication between the unmanned vehicle system and the UAV system via the data link communication module of the unmanned vehicle communication system, and communication between the unmanned vehicle system and the remote command system via the satellite communication module of the communication system. This eliminates the information delay problem of the ground station as an intermediate layer in the traditional architecture, ensuring the efficiency and reliability of communication, while ensuring the feasibility of remote command. It enables the system to maintain a stable communication connection in different scenarios and improves the real-time response capability of the system.

[0042] 3. The unmanned vehicle-unmanned aerial vehicle (UAV) collaborative reconnaissance and strike system of the present invention comprehensively perceives and collects data on the surrounding environment through the multimodal sensor components of the unmanned vehicle system. The collaborative control module of the central control system integrates information from the unmanned vehicle, UAV, and weapon system and optimizes task allocation and collaborative processes. The situational awareness module and target recognition module fuse and analyze multimodal sensor data in real time to construct a unified battlefield situation map. The resource optimization module is responsible for system resource management, realizing accurate acquisition and processing of environmental information, ensuring the rationality of task allocation and the efficiency of collaborative processes, improving battlefield situational awareness, and realizing the rational allocation of resources, thereby improving the system's combat effectiveness and continuous combat capability. Attached Figure Description

[0043] Figure 1 This is a schematic diagram of the structure of the unmanned vehicle-unmanned aerial vehicle cooperative reconnaissance and strike system according to an embodiment of the present invention;

[0044] Figure 2 This is a schematic diagram of the unmanned vehicle system according to an embodiment of the present invention;

[0045] Figure 3 This is a schematic diagram of the structure of an unmanned aerial vehicle system according to an embodiment of the present invention;

[0046] Figure 4 This is a schematic diagram of the functional modules of the central control system according to an embodiment of the present invention.

[0047] Figure 5 This is a flowchart of the unmanned vehicle-drone collaborative reconnaissance and strike method according to an embodiment of the present invention.

[0048] 100 - Unmanned Vehicle System, 200 - Unmanned Aerial Vehicle System, 300 - Central Control System, 400 - Remote Command System, 110 - Main Vehicle Platform, 111 - Power System, 112 - Power Supply System, 113 - Suspension System, 120 - Multimodal Sensor Components, 121 - Visual Sensor, 122 - Radar Sensor, 123 - LiDAR, 124 - Infrared Sensor, 125 - Acoustic Sensor, 130 - Vehicle-Mounted Weapon System, 131 - Main Weapon Module, 132 - Auxiliary Weapon Module, 133 - Weapon Control Unit, 140 - Vehicle-Mounted Computing Platform, 141 - High-Performance Computing Unit 142-Data storage unit, 143-Thermal management system, 150-Vehicle-mounted UAV station, 151-Mechanically foldable launch platform, 152-Precision positioning and recovery system, 153-Fast charging system, 154-Mechanical locking mechanism, 160-Communication system, 161-Data link communication module, 162-Satellite communication module, 163-Communication encryption module, 210-Quadrotor body, 211-Battery pack, 212-Power system, 213-Navigation system, 220-Foldable wing, 221-Main wing, 222-Control surfaces, 223-Mode conversion mechanism, 230-Modular mission payload 231-Electro-optical pod, 232-Electronic reconnaissance module, 233-Light weapon module, 240-Intelligent flight control system, 241-Flight controller, 242-Obstacle avoidance system, 243-Mission execution module, 310-Deep reinforcement learning decision-making module, 311-Environmental perception network, 312-Policy network, 313-Value network, 314-Online learning engine, 320-Mission planning module, 321-Path planning submodule, 322-Time resource allocation submodule, 323-Emergency plan generation submodule, 330-Situational awareness module, 331-Multi-source data fusion submodule 332-Environmental Modeling Submodule, 333-Threat Assessment Submodule, 340-Target Recognition Submodule, 341-Target Detection Submodule, 342-Target Tracking Submodule, 343-Target Attribute Analysis Submodule, 350-Collaborative Control Submodule, 351-Task Allocation Submodule, 352-Collaborative Planning Submodule, 353-Communication Management Submodule, 360-Resource Optimization Submodule, 361-Energy Management Submodule, 362-Weapon Resource Management Submodule, 363-Computing Resource Allocation Submodule, 410-Remote Human-Machine Interaction Terminal, 420-Remote Monitoring Submodule, 430-Remote Authorization Submodule. Detailed Implementation

[0049] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0050] Example 1

[0051] Please refer to Figure 1 This embodiment 1 provides an unmanned vehicle-unmanned aerial vehicle (UAV) collaborative reconnaissance and strike system, including: an unmanned vehicle system configured for ground mobility, reconnaissance and heavy weapon strikes; an UAV system configured for aerial reconnaissance and light strikes; a central control system configured for decision-making and collaborative control; and a remote command system configured for mission issuance, monitoring and key action authorization.

[0052] (1) Unmanned vehicle system

[0053] Please refer to Figure 2 The unmanned vehicle system 100 includes: a main vehicle platform 110, a multimodal sensor assembly 120, an onboard weapon system 130, an onboard computing platform 140, an onboard unmanned aerial vehicle station 150, and a communication system 160.

[0054] The main vehicle platform 110 adopts a modular design, facilitating rapid replacement of functional modules according to mission requirements. The main vehicle platform 110 internally houses a power system 111, a power supply system 112, and a suspension system 113. The power system 111 employs a hybrid powertrain design, combining a diesel engine and an electric motor, allowing for both conventional fuel and electric drive, with a maximum range of up to 500 kilometers. The power supply system 112 includes a high-capacity lithium-ion battery pack and an intelligent power management unit, providing a continuous and stable power supply to the entire vehicle system and the drone, and featuring fast charging and intelligent power allocation capabilities. The suspension system 113 utilizes an adaptive hydraulic suspension, capable of automatically adjusting the vehicle's height and stiffness according to terrain, improving passability and platform stability.

[0055] The multimodal sensor assembly 120, acting as the system's "eyes and ears," includes a vision sensor 121, a radar sensor 122, a lidar sensor 123, an infrared sensor 124, and an acoustic sensor 125. The vision sensor 121 comprises multiple high-definition cameras, covering a 360° field of view around the vehicle, and is equipped with an image stabilization system and night vision capabilities, providing clear images under various lighting conditions. The radar sensor 122 employs phased array technology, capable of detecting moving targets within a 30-kilometer range and possessing anti-jamming capabilities. The lidar sensor 123 generates high-precision 3D maps to assist navigation and obstacle recognition, achieving centimeter-level accuracy. The infrared sensor 124 can detect heat-generating targets in all weather conditions, making it particularly suitable for nighttime or low-visibility environments. The acoustic sensor 125 captures ambient sounds and identifies specific sound characteristics such as engine noise and gunshots, providing additional dimensions of information for target identification.

[0056] The vehicle-mounted weapon system 130 includes a main weapon module 131, an auxiliary weapon module 132, and a weapon control unit 133. The main weapon module 131 is a remotely controlled, stabilized platform that can be equipped with weapons such as a 12.7mm machine gun, a 40mm automatic grenade launcher, or a light missile launcher, depending on mission requirements. It has a 360° rotation capability and a -10° to +60° elevation range. The auxiliary weapon module 132 includes a smoke generator, non-lethal weapons (such as a sonic dispersal device), and a close-in weapon system for self-defense and special mission scenarios. The weapon control unit 133 is responsible for the precise aiming and firing control of the weapon, and has automatic target tracking and firing correction functions. It can automatically adjust firing parameters based on factors such as wind speed and distance to improve accuracy.

[0057] The onboard computing platform 140 is the core processing unit of the system, including a high-performance computing unit 141, a data storage unit 142, and a thermal management system 143. The high-performance computing unit 141 adopts a multi-core CPU and GPU architecture, supports the real-time execution of complex deep learning methods, and possesses powerful parallel computing capabilities. The data storage unit 142 uses a high-speed solid-state drive, capable of storing large amounts of reconnaissance data and learning models, and features data encryption and fast access capabilities. The thermal management system 143 ensures stable operation of the computing platform in extreme environments, employing phase change materials and active heat dissipation design, and can operate normally within an ambient temperature range of -40℃ to +60℃.

[0058] The vehicle-mounted drone station 150 serves as the "home" for the drone and includes: a mechanically foldable launch platform 151, a precise positioning and recovery system 152, a fast charging system 153, and a mechanical locking mechanism 154. The mechanically foldable launch platform 151 folds and hides in the roof of the vehicle in standby mode, automatically unfolding into a takeoff platform when needed, reducing the exposed area of ​​the system and improving concealment. The precise positioning and recovery system 152 uses radar and optical positioning technology to guide the drone to a precise landing, enabling safe recovery even in complex environments or when the vehicle is moving. The fast charging system 153 employs wireless charging technology, fully charging the drone's battery within a preset time. The mechanical locking mechanism 154 automatically locks the drone after landing, preventing it from moving or being damaged during vehicle movement, while also providing waterproof and dustproof protection.

[0059] The communication system 160 includes a data link communication module 161, a satellite communication module 162, and a communication encryption module 163. The data link communication module 161 supports high-bandwidth, low-latency communication with the UAV, with a maximum effective communication range of 30 kilometers, and employs frequency hopping and spread spectrum technologies to resist interference. The satellite communication module 162 is used to communicate with a remote command system, enabling beyond-line-of-sight command and control, and supporting mission execution globally. The communication encryption module 163 employs quantum-level encryption technology to ensure communication security and prevent interception or interference by the enemy.

[0060] (2) Unmanned Aerial Vehicle System

[0061] Please refer to Figure 3 The unmanned aerial vehicle system 200 includes: a quadcopter body 210, foldable wings 220, a modular mission payload interface 230, and an intelligent flight control system 240.

[0062] The quadcopter body 210 is made of carbon fiber composite material, which is lightweight and high-strength. It integrates a battery pack 211, a power system 212, and a navigation system 213. The battery pack 211 is a high-energy-density lithium polymer battery, which can support 2 hours of flight on a single charge. The battery management system has overcharge protection and thermal management functions. The power system 212 includes a high-efficiency brushless motor and a high-precision electronic speed controller, with a redundant design, allowing flight to continue even if one motor fails. The navigation system 213 integrates multiple technologies including GPS, BeiDou, inertial navigation, and visual navigation, ensuring stable flight even in environments where GPS signals are interfered with, with a positioning accuracy better than 0.5 meters.

[0063] The foldable wing 220 is an innovative design of this system, adding fixed-wing flight capability to the quadcopter architecture, enabling dual-mode flight of vertical takeoff and landing and high-speed cruise. The foldable wing 220 includes a main wing 221, control surfaces 222, and a mode-switching mechanism 223. During takeoff and landing, it operates in quadcopter mode with the wings folded up; during cruise, the wings automatically unfold, switching to fixed-wing mode, achieving a maximum cruise speed of 120 km / h, increasing cruise efficiency by 40%, and significantly extending the combat radius. The mode-switching mechanism 223 can complete the transition within 5 seconds, ensuring a smooth transition.

[0064] The modular mission payload interface 230 includes: an electro-optical pod 231, an electronic reconnaissance module 232, and a lightweight weapon module 233. The electro-optical pod 231 integrates a high-definition visible light camera, an infrared thermal imager, and a laser rangefinder, supporting 30x optical zoom for all-weather reconnaissance. Its image stabilization system ensures clear images during flight. The electronic reconnaissance module 232 can detect and locate radio signal sources for electronic intelligence gathering, supporting signal analysis in the 2MHz-6GHz frequency band. The lightweight weapon module 233 can be quickly replaced according to mission requirements, supporting small guided bombs or micro-missiles with a maximum payload of 2 kg, possessing precision strike capability.

[0065] The intelligent flight control system 240 includes a flight controller 241, an obstacle avoidance system 242, and a mission execution module 243. The flight controller 241 employs a multi-core processor and redundancy design to ensure flight safety and supports autonomous flight, formation flight, and handling of abnormal situations. The obstacle avoidance system 242 utilizes visual and laser sensors to achieve 360° obstacle avoidance without blind spots, with a reaction time of less than 0.1 seconds, enabling safe flight in complex environments. The mission execution module 243 is responsible for parsing and executing mission instructions from the central control system, performing functions such as preset flight routes, target tracking, and autonomous return.

[0066] (3) Central control system

[0067] Please refer to Figure 4 The central control system 300 includes: a deep reinforcement learning decision-making module 310, a task planning module 320, a situational awareness module 330, a target recognition module 340, a collaborative control module 350, and a resource optimization module 360.

[0068] The deep reinforcement learning decision-making module 310 is the "brain" of the system, comprising: an environment perception network 311, a policy network 312, a value network 313, and an online learning engine 314. The environment perception network 311 employs a deep convolutional neural network structure, fusing heterogeneous data (images, radar, thermal imaging, etc.) from multiple sensors into a unified state representation, supporting multi-resolution feature extraction and temporal correlation analysis. The policy network 312 is responsible for generating action distributions based on the current state, guiding system behavior, and adopts an Actor-Critic architecture, supporting both continuous and discrete action spaces. The value network 313 evaluates the value of the current state and the merits of each possible action, providing quantitative basis for decision-making, and employs a double-Q learning method to reduce overestimation problems. The online learning engine 314 continuously optimizes network parameters based on practical experience, improving decision-making performance, and employs priority experience replay and target network techniques to improve learning efficiency and stability.

[0069] The task planning module 320 is responsible for generating an optimal task execution plan based on task objectives, environmental conditions, and system status. The task planning module 320 includes a path planning submodule 321, a time resource allocation submodule 322, and an emergency plan generation submodule 323. The path planning submodule 321 uses an improved A-method and a Rapid Expanding Random Tree (RRT) method to generate an optimal path that considers terrain, threats, and energy consumption. The time resource allocation submodule 322 optimizes task execution timing and resource usage to ensure efficient system operation. The emergency plan generation submodule 323 pre-determines contingency plans for possible abnormal situations to improve system robustness.

[0070] In a preferred embodiment, the improved A method and the fast expanding random tree method specifically involve: using the fast expanding random tree method to sample "security critical points" near the threat area to construct a simplified topology map; and using the A method to calculate the optimal path on the topology map to reduce the number of nodes in the discrete grid.

[0071] Method A is improved by reconstructing the cost function, as detailed below:

[0072] f(n) = ω1·g terrain (n)+ω2·g threat (n)+ω3·g energy (n)+h((n)

[0073] Among them, g terrain (n) represents the terrain cost; g threat (n) represents the cost of the threat; g energy (n) represents the energy cost; ω1, ω2, ω3 are weighting coefficients, which must satisfy ω1 + ω2 + ω3 = 1 and are dynamically adjusted according to task requirements; and terrain and threats are introduced to adjust the heuristic estimation terms as follows:

[0074]

[0075] Where, x goal ,y goal These are the x and y coordinates of the target point, respectively; x n ,y n η(n) represents the horizontal and vertical coordinates of the current node n, respectively; η(n) is the environmental impact factor, which is determined based on the terrain complexity and threat level of the current node n.

[0076] The situational awareness module 330 fuses various sensor data in real time to construct a battlefield situation map, assess threat levels and environmental complexity, and provide a comprehensive environmental awareness foundation for decision-making. The situational awareness module 330 includes: a multi-source data fusion submodule 331, an environmental modeling submodule 332, and a threat assessment submodule 333. The multi-source data fusion submodule 331 uses Kalman filtering and particle filtering methods to integrate data from different sensors, improving perception accuracy and reliability. The environmental modeling submodule 332 constructs a high-precision three-dimensional environmental model to support navigation and mission planning. The threat assessment submodule 333 analyzes potential threats, calculates risk levels, and provides security assurance for the system.

[0077] In a preferred embodiment, the specific process of the multi-source data fusion submodule 331 fusing the Kalman filter and particle filter methods is as follows:

[0078] Based on the data type and noise characteristics, determine whether the data conforms to the linear Gaussian model, and initialize the parameters of the Kalman filter and particle filter.

[0079] Time synchronization and spatial coordinate transformation are performed on data from different sensors (e.g., radar distance, visual images);

[0080] Independent prediction using both Kalman filtering and particle filtering methods;

[0081] The prediction results (mean and covariance) of Kalman filtering are used as the prior information of particle filtering. The matching degree between each particle and the Kalman predicted state is calculated by Mahalanobis distance. The weight of mismatched particles is reduced, and the particle swarm is concentrated in the high probability region, reducing the number of invalid particles.

[0082] When the particle filter detects nonlinear characteristics in the data (such as a sudden change in target direction), it feeds back to adjust the noise covariance of the Kalman filter to enhance its adaptability to sudden changes.

[0083] Kalman filtering updates the state estimate based on sensor measurements, calculates the Kalman gain, and corrects the prediction error. Particle filtering updates the weight of each particle based on the measurement model, retains high-weight particles and discards low-weight particles through resampling. The estimation errors of the two methods are calculated, and the weights are dynamically allocated according to the magnitude of the error. Finally, the fused state estimate is output by weighted summation.

[0084] The target recognition module 340 employs advanced deep learning methods to identify and classify various military targets. The target recognition module 340 includes: a target detection submodule 341, a target tracking submodule 342, and a target attribute analysis submodule 343. The target detection submodule 341 uses a method-based approach to detect targets within the field of view in real time, supporting the identification of small targets and partially occluded targets. The target tracking submodule 342 uses a multi-target tracking method, maintaining target IDs and trajectory information, and supports re-identification after target occlusion. The target attribute analysis submodule 343 evaluates target type, threat level, and value, providing a basis for decision-making.

[0085] The collaborative control module 350 is responsible for coordinating the actions of the unmanned vehicles and drones to ensure effective collaborative combat capabilities. The collaborative control module 350 includes a task allocation submodule 351, a collaborative planning submodule 352, and a communication management submodule 353. The task allocation submodule 351 dynamically allocates reconnaissance and strike tasks based on mission requirements and system capabilities. The collaborative planning submodule 352 generates collaborative action plans for the unmanned vehicles and drones to maximize collaborative benefits. The communication management submodule 353 optimizes the use of communication resources, ensures timely transmission of critical information, and adjusts the operating mode when communication is limited.

[0086] The resource optimization module 360 ​​is responsible for managing system resources, including energy, ammunition, and computing resources, ensuring efficient resource utilization and extending the system's operational duration. The resource optimization module 360 ​​includes: an energy management submodule 361, a weapon resource management submodule 362, and a computing resource allocation submodule 363. The energy management submodule 361 monitors and predicts energy consumption, optimizes energy usage strategies, and extends operational time. The weapon resource management submodule 362 monitors ammunition status, optimizes weapon usage, and ensures effective engagement of key targets. The computing resource allocation submodule 363 dynamically adjusts computing resource allocation to ensure sufficient computational support for critical tasks (such as target identification and decision-making).

[0087] (5) Remote command system

[0088] The remote command system 400 includes: a remote human-machine interface terminal 410, a remote monitoring module 420, and a remote authorization module 430. The remote human-machine interface terminal 410 provides an intuitive user interface, facilitating operators to issue task commands and monitor task execution, and supports multiple interaction methods including touch, voice, and gestures. The remote monitoring module 420 displays the status, location, and reconnaissance results of unmanned vehicles and drones in real time, supporting multi-view switching and situational map display. The remote authorization module 430 is responsible for the manual confirmation of critical actions (especially weapon use), ensuring the system operates safely under human supervision, and recording all authorized activities for traceability.

[0089] Example 2

[0090] Please refer to Figure 5 This embodiment 2 provides a method for coordinated reconnaissance and strike operations using unmanned vehicles and unmanned aerial vehicles, including:

[0091] S1. After receiving the task, the remote command system (400) transmits it to the central control system (300). The central control system (300) completes self-check and resource assessment, and then activates the corresponding decision-making strategy according to the task type.

[0092] S2. Generate an initial mission plan by combining terrain, threat and resource conditions, assign unmanned vehicle and drone missions to optimize energy and reconnaissance efficiency, and formulate collaborative strategies;

[0093] S3. The unmanned vehicle proceeds to conduct reconnaissance along the planned route, and decides whether to release drones for collaborative reconnaissance operations based on the complexity of the environment;

[0094] S4. When a suspicious target is identified, the central control system (300) calculates the value of different actions based on the current status and historical experience, and selects the optimal action to execute; when it involves actions that require manual authorization, including the use of weapons, it requests confirmation from the remote authorization module (430);

[0095] S5. Based on the decision results, the strike mission is issued, the UAV assesses the results after the strike and transmits the information back, and the central control system (300) optimizes the decision strategy accordingly.

[0096] S6. After the mission is completed, the drone returns and is recycled and recharged by the unmanned vehicle and enters standby mode. At the same time, the central control system (300) generates a mission report and sends it back to the remote system.

[0097] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A collaborative reconnaissance and strike system of unmanned vehicle and unmanned aerial vehicle, characterized in that, include: The system includes an unmanned vehicle system (100) configured for ground mobility, reconnaissance and heavy weapon strikes; an unmanned aerial vehicle system (200) configured for aerial reconnaissance and light strikes; a central control system (300) configured for decision-making and collaborative control; and a remote command system (400) configured for mission assignment, monitoring and key action authorization. The unmanned vehicle system (100) and the unmanned aerial vehicle system (200) communicate directly through the data link communication module (161) of the communication system (160) in the unmanned vehicle system (100); the unmanned vehicle system (100) and the remote command system (400) communicate through the satellite communication module (162) of the communication system (160); the unmanned vehicle system (100) uses a multimodal sensor assembly (120) to comprehensively perceive and collect data about the surrounding environment; The central control system (300) includes a collaborative control module (350) configured to integrate information from unmanned vehicles, drones and weapon systems, optimize task allocation and collaborative processes; a situational awareness module (330) and a target identification module (340) configured to perform real-time fusion and analysis of multimodal sensor data to construct a unified battlefield situation map; and a resource optimization module (360) configured to be responsible for system resource management.

2. The unmanned vehicle-unmanned aerial vehicle cooperative reconnaissance and strike system according to claim 1, characterized in that, The unmanned vehicle system (100) also includes a main vehicle platform (110), which adopts a modular design and allows for the replacement of functional modules according to task requirements.

3. The unmanned vehicle-unmanned aerial vehicle cooperative reconnaissance and strike system according to claim 1, characterized in that, The unmanned vehicle system (100) also includes a vehicle-mounted drone station (150), which folds and hides the drone on the roof when it is in standby mode via a mechanically foldable launch platform (151); the vehicle-mounted drone station (150) automatically locks the drone after it lands via a mechanical locking mechanism (154); the vehicle-mounted drone station (150) guides the drone to land via a positioning and recovery system (152); and the vehicle-mounted drone station (150) provides energy to the drone via a fast charging system (153).

4. The unmanned vehicle-unmanned aerial vehicle cooperative reconnaissance and strike system according to claim 1, characterized in that, The multimodal sensor assembly (120) includes a visual sensor configured to acquire visual images of the vehicle's surroundings, a radar sensor (122) configured to detect moving targets within a set range, a lidar (123) configured to generate a three-dimensional map, an infrared sensor (124) configured to detect heat source targets, and an acoustic sensor (125) configured to capture ambient sound. The multi-source data acquired by the multimodal sensor assembly (120) is transmitted to the situational awareness module (330) of the central control system (300).

5. The unmanned vehicle-unmanned aerial vehicle cooperative reconnaissance and strike system according to claim 1, characterized in that, The unmanned aerial vehicle system (200) also includes a quadcopter body (210) and foldable wings (220); the foldable wings (220) are mounted on the quadcopter body (210), and automatically unfold during the drone's cruise and retract during takeoff and landing.

6. The unmanned vehicle-unmanned aerial vehicle cooperative reconnaissance and strike system according to claim 1, characterized in that, The situational awareness module (330) fuses various sensor data through the multi-source data fusion submodule (331); the multi-source data fusion submodule (331) integrates different sensor data by fusing Kalman filtering and particle filtering methods.

7. The unmanned vehicle-unmanned aerial vehicle cooperative reconnaissance and strike system according to claim 6, characterized in that, The specific process of fusing the Kalman filter and particle filter methods is as follows: Based on the data type and noise characteristics, determine whether the data conforms to the linear Gaussian model, and initialize the parameters of the Kalman filter and particle filter. Time synchronization and spatial coordinate transformation of data from different sensors; Independent prediction using both Kalman filtering and particle filtering methods; The prediction results of Kalman filtering are used as the prior information of particle filtering. The matching degree between each particle and the Kalman predicted state is calculated by Mahalanobis distance. The weight of mismatched particles is reduced, and the particle swarm is concentrated in the high probability region, reducing the number of invalid particles. When the particle filter detects nonlinear characteristics in the data, it feeds back to adjust the noise covariance of the Kalman filter to enhance its adaptability to sudden changes. Kalman filtering updates the state estimate based on sensor measurements, calculates the Kalman gain, and corrects prediction errors. Particle filtering updates the weight of each particle based on the measurement model, retaining high-weight particles and discarding low-weight particles through resampling; The estimation errors of the two methods are calculated, and the weights are dynamically allocated according to the magnitude of the errors. Finally, the fused state estimate is output by weighted summation.

8. The unmanned vehicle-unmanned aerial vehicle cooperative reconnaissance and strike system according to claim 1, characterized in that, The central control system (300) also includes a task planning module (320) configured to generate an optimal task execution plan based on task objectives, environmental conditions and system status; the task planning module (320) includes a path planning submodule (321), a time resource allocation submodule (322) and an emergency plan generation submodule (323).

9. The unmanned vehicle-unmanned aerial vehicle cooperative reconnaissance and strike system according to claim 8, characterized in that, The path planning submodule (321) generates an optimal path that takes into account terrain, threats and energy consumption through an improved A method and a fast expanding random tree method; The improved A method and the fast expanding random tree method are as follows: the fast expanding random tree method is used to sample "security critical points" near the threat area to construct a simplified topology graph; the A method is used to calculate the optimal path on the topology graph to reduce the number of nodes in the discrete grid. Method A is improved by reconstructing the cost function, as detailed below: f(n)=ω1·g terrain (n)+ω2·g threat (n)+ω3·g energy (n)+h((n) Among them, g terrain (n) represents the terrain cost; g threat (n) represents the cost of the threat; g energy (n) represents the energy cost; ω1, ω2, ω3 are weighting coefficients, which must satisfy ω1 + ω2 + ω3 = 1 and are dynamically adjusted according to task requirements; and terrain and threats are introduced to adjust the heuristic estimation terms as follows: Where, x goal ,y goal These are the x and y coordinates of the target point, respectively; x n ,y n η(n) represents the horizontal and vertical coordinates of the current node n, respectively; η(n) is the environmental impact factor, which is determined based on the terrain complexity and threat level of the current node n.

10. A method for coordinated reconnaissance and strike operations using unmanned vehicles and drones, characterized in that: include: S1. After receiving the task, the remote command system (400) transmits it to the central control system (300). The central control system (300) completes self-check and resource assessment, and then activates the corresponding decision-making strategy according to the task type. S2. Generate an initial mission plan by combining terrain, threat and resource conditions, assign unmanned vehicle and drone missions to optimize energy and reconnaissance efficiency, and formulate collaborative strategies; S3. The unmanned vehicle proceeds to conduct reconnaissance along the planned route, and decides whether to release drones for collaborative reconnaissance operations based on the complexity of the environment; S4. When a suspicious target is identified, the central control system (300) calculates the value of different actions based on the current status and historical experience, and selects the optimal action to execute; when it involves actions that require manual authorization, including the use of weapons, it requests confirmation from the remote authorization module (430); S5. Based on the decision results, the strike mission is issued, the UAV assesses the results after the strike and transmits the information back, and the central control system (300) optimizes the decision strategy accordingly. S6. After the mission is completed, the drone returns and is recycled and recharged by the unmanned vehicle and enters standby mode. At the same time, the central control system (300) generates a mission report and sends it back to the remote system.

Images