Transferable water area target identification verification system

Abstract

The invention provides a transferable water area target identification verification system. The transferable water area target identification verification system comprises a load platform, a video acquisition module, a target identification module and a target positioning module, the load platform is used for supporting water area target identification and verification equipment; based on the load platform, acquiring water area target image information through the video acquisition module, acquiring video data through the target identification module, and detecting a target to obtain target information; and obtaining the target information through the target positioning module to obtain the position information of the target. The unmanned ship-unmanned aerial vehicle combined execution water area suspicious target identification and verification system has mobility and can be carried on an unmanned ship and unmanned aerial vehicle platform to realize three-dimensional observation of a patrol water area, and the unmanned ship can supply energy to the unmanned aerial vehicle through a wireless charging parking apron, so that the unmanned aerial vehicle can be conveniently and rapidly charged. The operation efficiency of unmanned ship-machine combined target identification and verification is improved, and the risk of artificial suspicious target identification and verification in a complex dangerous water area is reduced.

Description

Technical Field

[0001] The present invention belongs to the field of surface target identification and verification, and specifically, to a portable water target identification system and verification method. More specifically, it is a portable water target identification system and verification method, and more specifically, it is a system and method design for joint identification and verification of suspicious surface targets by using an unmanned boat-machine through a target identification system and a navigation module. The portable means that it can be carried on an unmanned boat or an unmanned aerial vehicle platform to achieve stereoscopic observation of patrol waters. Background Art

[0002] Most traditional water target identification and verification tasks require human participation, and the efficiency of target identification and verification is relatively low. However, with the rapid development of intelligent equipment, the application of cross-domain unmanned systems in water target identification and verification has received widespread attention.

[0003] Compared with traditional methods of water target identification and verification, unmanned systems can effectively reduce operating costs and quickly and accurately locate water target identification and verification targets. Among them, as an important tool for water target identification and verification, unmanned boats can perform long-term and large-scale target identification and verification in complex and dangerous sea conditions, but the unmanned boat has limited field of view information and cannot quickly and accurately obtain target identification and verification information; UAVs rely on their advantages such as three-dimensional maneuverability and wide field of view to make up for the shortcomings of the unmanned boat's visual threshold limitation; and unmanned boats can also solve the problems of short drone endurance and weak load capacity.

[0004] However, since the existing water target identification system is a single platform, there is a problem of low efficiency in target identification and verification, which leads to high verification costs and weak autonomy of unmanned patrol energy. This problem needs to be solved urgently. Summary of the invention

[0005] In view of the defects in the prior art, the purpose of the present invention is to provide a portable water area target identification and verification system.

[0006] A portable water target identification and verification system provided by the present invention includes: a load platform, a video acquisition module, a target identification module and a target positioning module;

[0007] The payload platform is used to support equipment for identifying and verifying water targets;

[0008] Based on the load platform, the image information of the water area target is obtained through the video acquisition module, and the video data is obtained and the target is detected through the target recognition module to obtain the target information;

[0009] The target information is acquired through the target positioning module to obtain the position information of the target.

[0010] Preferably, the load platform includes: an unmanned aerial vehicle flight platform, an unmanned boat and a ground monitoring platform;

[0011] The UAV flight platform is used to detect, identify and locate suspicious targets on the water surface; the suspicious targets on the water surface are boats;

[0012] The unmanned boat is a catamaran unmanned boat; the catamaran unmanned boat is used to receive the position information of the suspicious target on the water surface and plan a path;

[0013] The ground monitoring platform is used to display the status information of the UAV and issue control instructions.

[0014] Preferably, it can realize the wide-area search phase, the surface target identification and positioning phase, and the boat-aircraft joint approach identification and verification phase;

[0015] The wide area search phase includes:

[0016] The ground control platform issues instructions to the UAV and the unmanned boat to arrive at the patrol position and start patrolling; the video acquisition module obtains the image information of the water target through the UAV, and the UAV scores and sorts the target according to the target image information by a comprehensive weight sorting method to obtain the priority order of verification;

[0017] The surface target identification and positioning stage includes:

[0018] The target recognition module determines whether the UAV finds a suspicious target on the water surface; if the result is no, the UAV is ordered to return to the unmanned boat for charging after the patrol mission is completed; if the result is yes, the suspicious target on the water surface is identified and located according to the priority order of the verification to obtain target information;

[0019] The joint boat-aircraft approach identification and verification phase includes:

[0020] The target information is acquired through the target positioning module to obtain the target location information; the drone is instructed to send the target location information to the unmanned boat, and the unmanned boat is instructed to approach and verify the suspicious target on the water surface based on the target location information, and return after the verification is completed.

[0021] Preferably, in the wide area search stage, the comprehensive weight sorting method is to represent the evaluation value of the attribute by triangular fuzzy numbers, and then perform priority sorting according to the evaluation value;

[0022] The expression of the triangular fuzzy number is:

[0023]

[0024] in, is a triangular fuzzy number, a i is the minimum value, b i is the most likely value, c i is the maximum value;

[0025] For the maximization objective, the linear membership function of the triangular fuzzy number is expressed as:

[0026]

[0027] For the minimization objective, the linear membership function of the triangular fuzzy number is expressed as:

[0028]

[0029] For dynamic characteristics, the expression of the linear membership function is:

[0030]

[0031] in, is the linear membership function of the dynamic characteristics;

[0032] In the comprehensive weight ranking method, the weight of each attribute is determined, that is, w I 、w T 、w D With w C , reflects the relative importance of each attribute in the verification task, and the mathematical expression is:

[0033] w I +w T +w D +w C =1(5)

[0034] Among them, w I is the weight of importance; w T is the weight of threat level; w D is the weight of dynamic characteristics; w C is the weight of the verification difficulty;

[0035] Based on the w I 、w T 、w D With w C , for each target i, calculate the comprehensive score, the mathematical expression is:

[0036]

[0037] Among them, S i is a comprehensive rating. is the degree of membership of target i in importance, is the membership of target i in terms of threat level, is the membership degree of target i in dynamic characteristics, is the membership of target i in the verification difficulty; the symbol · is the product;

[0038] All targets are arranged in descending order according to the comprehensive score; the higher the score of the comprehensive score, the higher the priority of verification.

[0039] Preferably, in the surface target recognition and positioning stage, the unmanned boat adjusts its own position information according to the position information of the drone; the position error between the unmanned boat and the drone is expressed mathematically as:

[0040]

[0041] Where ΔX iav-usv is the position error between the unmanned boat and the drone in the X-axis direction; ΔY iav-usv is the position error between the unmanned boat and the UAV in the Y-axis direction; ΔZ uav-usv is the position error between the unmanned boat and the drone in the Z-axis direction; X a , Y a With Z a They are the position information of the drone on the X axis, Y axis and Z axis respectively; s , Y s With Z s They are the position information of the unmanned boat on the X-axis, the Y-axis and the Z-axis respectively;

[0042] Determine whether the unmanned boat meets the position error requirement, if yes, do not adjust the position of the unmanned boat; if no, adjust the position of the unmanned boat;

[0043] The position error requirement is:

[0044] ΔX uav-usv =0

[0045] ΔY uav-usv =0

[0046] ΔZ uav-usv =Z a

[0047] Among them, Z a It is the position information of the drone on the Z axis.

[0048] Preferably, the unmanned boat and the accompanying unmanned aerial vehicle maintain a relative posture through the process control target cost function;

[0049] The process control objective cost function is expressed as follows:

[0050]

[0051] Wherein, min J is the process control objective cost function; Q represents the weight of the situation keeping error; superscript T is the transposition identifier; X is the position of the unmanned boat; is the control amount of the unmanned boat; R is the weight of the control amount difference; ΔU a is the target value of the control variable of the unmanned boat; X uav The reference trajectory given to the UAV.

[0052] Preferably, in the boat-aircraft joint approach identification and verification stage, the unmanned boat * The global path planning algorithm obtains and verifies the target based on the approach path;

[0053] In the wide area search stage, the water area target image information is obtained through the YOLOv8 model;

[0054] Preferably, the UAV maintains fixed-point hovering and sends the location information of the target to the unmanned boat.

[0055] Preferably, the drone executes a "bow"-shaped patrol path.

[0056] Preferably, the drone collects 25 frames of image data per second, the flight altitude is 50m, and the patrol speed is 5m / s.

[0057] Compared with the prior art, the present invention has the following beneficial effects:

[0058] 1. The unmanned boat-aircraft joint suspicious target identification and verification system adopted by the present invention is portable and can be carried on unmanned boats and unmanned aerial vehicle platforms to achieve three-dimensional observation of patrol waters. In addition, the unmanned boat can supply energy to the unmanned aerial vehicle through the wireless charging pad, thereby improving the operational efficiency of the unmanned boat-aircraft joint target identification and verification and reducing the risk of manual identification and verification of suspicious targets in complex and dangerous waters.

[0059] 2. The unmanned boat-drone combined identification and verification of suspicious targets in waters provided by the present invention is composed of a combination of phased identification and verification strategies. Compared with traditional identification and verification methods, the identification and verification method of the present invention is targeted at typical waters to be patrolled. First, suspicious targets are extracted through wide-area search, and a comprehensive weight ranking method is proposed to prioritize the approach confirmation and verification of suspicious targets; further, suspicious targets are identified and located through the high-altitude perspective of the drone, and the positioning information of the suspicious targets is uploaded to the unmanned boat end, and the unmanned boat and the drone work together to approach the vicinity of the suspicious targets to carry out identification and verification.

[0060] 3. The present invention provides a boat-aircraft collaborative search model to meet the needs of unmanned autonomous target identification and verification. Compared with traditional target identification and verification methods, the unmanned boat-aircraft joint target identification and verification has high efficiency and low identification and verification cost, and the unmanned autonomous patrol has strong energy autonomy. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings:

[0062] Figure 1 A schematic diagram of the system composition of the load platform provided by the present invention;

[0063] Figure 2 A schematic diagram of the working principle provided by the present invention;

[0064] Figure 3 A flow chart of the identification and verification of suspicious targets on the water surface provided by the unmanned boat-aircraft combination of the present invention;

[0065] Figure 4 A first flow chart of the main working modes of the unmanned boat provided by the present invention;

[0066] Figure 5 This is a second flow chart of the main working mode of the drone provided by the present invention. DETAILED DESCRIPTION

[0067] The present invention is described in detail below in conjunction with specific embodiments. The following embodiments will help those skilled in the art to further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those of ordinary skill in the art, several changes and improvements can also be made without departing from the concept of the present invention. These all belong to the protection scope of the present invention.

[0068] The water target identification system and verification method provided by the present invention, compared with the current state of the art, focuses on the research of "cross-domain joint, unmanned autonomous identification and verification of ground, air and sea".

[0069] This embodiment provides a suspicious target identification system and verification method in water areas. The identification system is portable and can be carried on platforms such as unmanned boats and drones.

[0070] A portable water target recognition system provided by the present invention includes: a video acquisition module, a target recognition module and a target positioning module;

[0071] The video acquisition module includes: a binocular camera, a video decoding module and a video stream output module; the video acquisition module can obtain water area target image information in real time through the binocular camera;

[0072] The target recognition module is used to acquire video data in real time and detect and identify features of suspicious targets in water areas through an improved YOLOv8 model;

[0073] The target positioning module is used to obtain the target information identified by the target recognition module in real time, by extracting the pixel coordinates of the target in the window and performing coordinate conversion. Further, the target positioning module can receive the positioning information of the load platform in real time, and finally convert the position information of the target in the geodetic coordinates;

[0074] Specifically, this embodiment uses a payload platform equipped with the water area identification system proposed in the present invention to illustrate its working principle and method, wherein the payload platform includes: a four-rotor UAV flight platform, a catamaran unmanned boat and a ground monitoring platform; the method is divided into three stages: wide-area search for suspicious targets on the water surface, identification and positioning of suspicious targets on the water surface, and unmanned boat-aircraft joint approach identification and verification.

[0075] The quad-rotor UAV flight platform includes: a control module, a navigation and positioning module, a communication module, a wireless charging module, and a power system; the UAV flight platform can be equipped with a water target identification system to detect, identify and locate suspicious targets on the water surface; the UAV control module is used for UAV flight attitude control; the UAV navigation and positioning module is used to realize UAV route planning and accurately locate suspicious targets on the water surface; the UAV communication module is used to receive the command data from the ground monitoring platform and send the positioning data of suspicious targets on the water surface to the catamaran unmanned boat, and transmit the suspicious target data back to the ground monitoring platform; the wireless charging module is used to replenish energy when the UAV is docked at the wireless charging apron, and the UAV power system is used to realize the UAV's rapid maneuvering search for suspicious targets on the water surface;

[0076] The ground monitoring platform includes: a communication module, monitoring software and a remote control terminal; the communication module of the ground monitoring platform is connected to the wireless data image transmission module, and is used to receive the drone status information, video information and the identified suspicious target location information sent back by the drone communication module; the monitoring software is used to display the drone status information, video information and the suspicious target location information, and is used to receive the accusation information from the ground remote control terminal; the remote control terminal of the ground monitoring platform is used to issue drone control instructions, and is used to control the flight speed and steering of the drone;

[0077] Figure 1 Schematic diagram of the system composition of the load platform of the present invention, as shown in Figure 1 As shown, the quad-rotor UAV flight platform includes: a target recognition system, a differential positioning unit, a wireless digital image transmission unit and a flight control unit;

[0078] The target recognition system is used to search for wide-area targets in the target waters, exclude inherent targets on the water surface, identify and locate suspicious targets on the water surface in combination with the improved YOLOv8 model, calculate the relative position of the suspicious target and convert it in combination with the differential positioning data of the UAV, and transmit the converted position information of the suspicious target on the water surface to the unmanned boat using wireless data transmission. The suspicious target identification and positioning process is controlled by the flight control unit to control the attitude and height of the UAV;

[0079] The catamaran unmanned boat includes: a target identification system, a mission planning unit, a propulsion drive unit, and a wireless data transmission unit; the catamaran unmanned boat is mainly used to receive the coordinate position information of the suspicious target on the water surface from the drone, and use the improved A* algorithm to perform global path planning, and in combination with the target identification system carried by the catamaran unmanned boat, obtain the suspicious target on the water surface in real time and replan the path to the vicinity of the suspicious target;

[0080] The ground monitoring platform mainly uses the communication module to exchange data with the onboard flight control unit. The monitoring software is used to display the status information of the UAV in real time and issue control commands. The remote control terminal is used to manually control the UAV and has the function of switching between autonomous and manual modes.

[0081] This embodiment provides a method for identifying and verifying suspicious targets on the water surface by combining an unmanned boat and an aircraft. The method mainly realizes the functions of wide-area target search by unmanned boats, identification and positioning of suspicious targets on the water surface, and autonomous route planning by unmanned boats to approach suspicious targets on the water surface. The specific working principle of the system is as follows: Figure 2 As shown;

[0082] Take a certain reservoir to be patrolled as an example, assuming that the total area of ​​the water area is 3.7km 2 , there are fixed targets such as buoys and islands on the water surface. The unmanned boat-aircraft joint identification and verification platform is permanently moored at the command department dock. After receiving the patrol mission instruction, the unmanned boat carries the drone to the mission starting point and waits. The drone then rises to the patrol altitude to patrol suspicious targets on the water surface; the suspicious targets are boats with suspicious tracks; the drone is set to be equipped with a 150° wide-angle camera, a flight altitude of 50m, a patrol speed of 5m / s, and collects 25 frames of image data per second to complete 1km 2 The water area inspection task takes 10 minutes.

[0083] Therefore, while ensuring the efficiency of the drone search, two drones are used to work in turns. When the drone battery is low, it can return to the unmanned boat for wireless charging, and efficiently complete the water target identification and verification work. Among them, the target identification and verification of the unmanned boat-aircraft joint identification and verification platform is divided into three stages: wide-area search stage, surface target identification and positioning stage, and boat-aircraft joint approach identification and verification stage.

[0084] Specifically, the steps of performing the wide area search phase are as follows:

[0085] Step 101: The unmanned boat-aircraft joint identification and verification platform is stationed at a dock in the accident-prone waters. After receiving the water patrol mission instruction issued by the command department, it quickly sails from the dock to the patrol starting point;

[0086] Step 102: After the unmanned boat-aircraft joint identification and verification platform arrives at the patrol starting point, the unmanned boat takes off for 50 meters and uses the 150° wide-angle camera on board to conduct a wide-area search of the patrol waters. The onboard high-performance visual edge processing computer quickly detects whether there are any suspicious targets in the waters within the field of view.

[0087] Step 103: If a suspected target is detected and verified, the drone will process the image coordinates of the suspected target and perform preliminary positioning by combining the high-precision single-antenna RTK positioning data on the aircraft. The preliminary positioning information of the suspected target is compared with the marked long-term target position on the water surface. If the position information of the two is within the set error range r, it is determined to be a long-term target on the water surface. Otherwise, the drone will approach further for detection and positioning. If no suspected target is detected initially, that is, a suspicious target on the water surface, the drone will autonomously execute a "bow"-shaped patrol path, with the field of view of the drone-mounted camera as the "bow"-shaped path spacing to ensure search efficiency;

[0088] Step 104: The drone sorts the identified suspected targets, mainly using a comprehensive weight sorting method. The attributes of each target include importance, threat level, dynamic characteristics, and verification difficulty. The evaluation value of each attribute is represented by a triangular fuzzy number. For the attribute x of target i, its triangular fuzzy number can be expressed as:

[0089]

[0090] in, is a triangular fuzzy number, a i is the minimum value, b i is the most likely value, c i is the maximum value.

[0091] Furthermore, in order to convert the fuzzy evaluation into a specific value, it is necessary to define a linear membership function

[0092] For the maximization objectives, such as importance I and threat level T, the mathematical expression of the linear membership function is:

[0093]

[0094] For the minimization objective, such as verification difficulty, i.e., C, the mathematical expression of the linear membership function is:

[0095]

[0096] For the dynamic characteristics, i.e. D, an appropriate membership function can be selected according to specific needs, such as:

[0097]

[0098] in, is the linear membership function of the dynamic characteristics;

[0099] Further, determine the weight w of each attribute I ,w T ,w D ,w C ,The above weights reflect the relative importance of each attribute in the ,verification task. The weights can be scored by experts and ,analytic hierarchy process, i.e., AHP, meet the conditions;

[0100] In other words, in the comprehensive weight ranking method, the weight of each attribute is determined, that is, w I 、w T 、w D With w C , reflects the relative importance of each attribute in the verification task, and the mathematical expression is:

[0101] w I +w T +w D +w C =1(5)

[0102] Among them, w I is the weight of importance; w T is the weight of threat level; w D is the weight of dynamic characteristics; w C is the weight of the verification difficulty;

[0103] Further, the comprehensive score is calculated. For each target i, its comprehensive score S is calculated. i :

[0104]

[0105] in, is the degree of membership of target i in importance, is the membership of target i in terms of threat level, is the membership degree of target i in dynamic characteristics, is the membership of target i in the verification difficulty; the symbol · is the product;

[0106] Further, according to the calculated comprehensive score S i Arrange all targets in descending order. The higher the score, the higher the priority of verification.

[0107] Specifically, the steps of the surface target recognition and positioning stage are as follows:

[0108] Step 201: If the UAV does not detect a suspected identification and verification target on the water surface during the patrol, the UAV continues to execute the "bow"-shaped patrol path until the mission is completed, and returns to the UAV for wireless charging;

[0109] Step 202: If the UAV finds multiple suspected identification and verification targets during the patrol process, the autonomous patrol mode is stopped, the multiple suspected identification and verification targets observed are identified, and the identification and verification targets are preliminarily located according to the target positioning algorithm.

[0110] Step 203: Prioritize the suspicious targets, and the UAV approaches to identify and locate the suspicious targets according to their priorities. If the target is a false identification and verification target, the UAV continues to perform the "bow"-shaped autonomous patrol mode;

[0111] Step 204: The unmanned boat implements an escort mechanism for the UAV according to the minimum return requirement of the UAV, mainly the remaining endurance and communication quality constraints;

[0112] Furthermore, when the UAV approaches the target for precise positioning, the unmanned boat needs to ensure that the UAV returns to the destination with low battery. Therefore, the unmanned boat needs to accompany the UAV. Based on the above constraints, the problem of trajectory tracking is transformed. The UAV sends its own position information, that is, the reference trajectory given by the UAV, which is X uav =[X a ,Y a ,Z a ] T After receiving the position information of the UAV, the unmanned boat adjusts its own position information to achieve the purpose of relative situation maintenance. The position error between the unmanned boat and the UAV can be expressed as:

[0113]

[0114] Where ΔX uav-usv is the position error between the unmanned boat and the drone in the X-axis direction; ΔY uav-usv is the position error between the unmanned boat and the UAV in the Y-axis direction; ΔZ uav-usv is the position error in the Z-axis direction between the unmanned boat and the drone; a , Y a With Z a They are the position information of the drone on the X axis, Y axis and Z axis respectively; s , Y s With Z sThey are respectively the position information of the unmanned boat on the X-axis, the Y-axis and the Z-axis.

[0115] In order to ensure that the unmanned boat can accurately maintain the situation with the drone, the position error in the horizontal plane should satisfy ΔX uav-usv =0,ΔY uav-usv = 0. When the unmanned boat is moving on the water surface, the altitude is set to 0, so ΔZ uav-usv =Z a ;

[0116] The model predictive control algorithm realizes the unmanned boat accompanying the UAV in a rolling optimization manner, so that the predicted output of the unmanned boat at the future time P can approach the situation expectation of the UAV, and the process control target cost function J of the unmanned boat accompanying the UAV to maintain the relative situation is expressed as:

[0117]

[0118] Wherein, min J is the process control objective cost function; Q represents the weight of the situation keeping error; superscript T is the transposition identifier; X is the position of the unmanned boat; is the control quantity of the unmanned boat; ΔU a is the target value of the control variable of the unmanned boat;

[0119] Where, X uav is the reference trajectory given by the UAV; Q represents the weight of the situation keeping error. The larger its value is, the smaller the difference between the predicted trajectory and the actual reference trajectory is expected to be, that is, the more accurate the tracking is; R represents the weight of the control amount difference. The larger its value is, the smoother the change of the control increment is expected to be;

[0120] subject to E mim ≤E a

[0121] C min ≤C uav-usv

[0122] Z min ≤Z a ≤Z max

[0123] ΔU min ≤ΔU a ≤ΔU max

[0124] Among them, E a Indicates the endurance of the drone, which requires that the endurance of the drone should not be less than the power required for return. Subject to is the control target, E min Minimum power required for return flight; C uav-usvIndicates the communication packet loss rate between the UAV and the unmanned boat. The packet loss rate is not less than the set minimum packet loss rate C min ; Z min ≤Z a ≤Z max Defines the safe range of drone flight; Z a is the flight altitude of the UAV; ΔU min ≤ΔU a ≤ΔU max Indicates that the control amount change of the unmanned boat should be kept in a stable range; ΔU min is the minimum value of the control variable of the unmanned boat, ΔU max is the maximum value of the control variable of the unmanned boat, ΔU a Target value of control variable of unmanned boat;

[0125] Step 205: If the UAV locates the real target, the positioning information is sent to the UAV.

[0126] Specifically, the steps for the joint boat-aircraft approach identification and verification phase are as follows:

[0127] Step 301: The unmanned boat receives the identification and verification target positioning information and uses the improved A * The global path planning algorithm autonomously approaches, identifies and verifies the target;

[0128] Step 302: The UAV locates the surface to identify and verify the target, quickly approaches the target to identify and verify the target, and lowers the flight altitude while approaching to ensure the accuracy of identification. If the UAV reaches the requirements of identification accuracy and positioning accuracy, it will hover at a fixed point and continuously send the target positioning information to the unmanned boat;

[0129] Step 303: The unmanned boat quickly approaches the target for identification and verification. If the unmanned boat accurately approaches the identification and verification range R, it slows down and stops. Finally, the drone lands on the unmanned boat landing pad to complete the target identification and verification task. The unmanned boat-drone joint identification and verification platform returns to the permanent dock.

[0130] In other words, the unmanned boat-aircraft joint water target verification method is divided into three stages: wide-area search for suspicious surface targets, identification and positioning of suspicious surface targets, and unmanned boat-aircraft joint approach identification and verification.

[0131] The wide-area search phase for suspicious surface targets includes:

[0132] Step 101: rasterizing the waters to be patrolled to determine the geodetic coordinates of long-standing water surface obstacles;

[0133] Step 102: The drone flies away from the wireless charging pad of the unmanned boat, and carries a water target identification system to detect suspicious targets on the water surface in real time;

[0134] Step 103: The UAV is launched from a fixed point, removes inherent obstacles on the water surface, and searches for suspected suspicious targets;

[0135] Step 104: The unmanned boat remains on standby at the patrol starting point;

[0136] The stage of identifying and locating suspicious targets on the water surface includes:

[0137] Step 201: The UAV performs an autonomous patrol mode to search for suspicious targets on the water surface;

[0138] Step 202: The UAV carries a water target recognition system to detect suspicious targets on the water surface in real time and calculate the relative position, and controls the movement of the UAV according to the relative deviation between the suspicious target on the water surface and the center of the window. At that time, the UAV forwards the position information of the suspicious target on the water surface to the unmanned boat. The water target recognition system is used to collect video data and perform real-time recognition through an improved YOLOv8 model, and the improved YOLOv8 model has completed pre-training of the suspicious target data set;

[0139] Step 203: the unmanned boat maintains an accompanying state according to constraints such as the heading and distance between the unmanned boat and the drone;

[0140] Step 204: The present invention proposes a comprehensive weight ranking method to verify the multiple suspicious targets identified, and designs target attributes to determine the priority of approaching identification and verification of suspicious targets. The target attributes include four aspects: importance, threat level, dynamic characteristics, and verification difficulty, and triangular fuzzy numbers are used to represent the evaluation value of each attribute;

[0141] The unmanned boat-aircraft joint approach identification and verification phase includes:

[0142] Step 301: The UAV finds the target, approaches it quickly, locates it, and uploads the target location to the UAV;

[0143] Step 302: The unmanned boat receives the position information of the surface target in real time and replans the global path in combination with the improved A* algorithm, and quickly and autonomously approaches the vicinity of the suspicious target for verification;

[0144] Step 302: When the unmanned boat and the unmanned aerial vehicle platform approach the suspicious target, the unmanned boat obtains the position information, heading information and power information of the unmanned aerial vehicle in real time, establishes a boat-aircraft collaborative search model, and ensures that the unmanned aerial vehicle returns to the destination with low power;

[0145] Step 303: The unmanned boat-aircraft simultaneously sends the suspicious target status information to the ground station for display, and the shore-side operator performs identification and verification decisions;

[0146] Step 304: After the suspicious target is successfully identified and verified, the unmanned boat carries the suspicious target back to the ground monitoring end, and the drone returns to the ground monitoring end autonomously. If the drone has insufficient power to return to the ground monitoring end, it will autonomously land on the wireless charging pad at the unmanned boat end and return to the ground monitoring end together.

[0147] The advantages of unmanned boats such as long endurance and strong carrying capacity can make up for the shortcomings of drones such as short endurance and weak carrying capacity. Therefore, when drones are performing water patrol target identification and verification tasks, if the drones have insufficient endurance, they can maneuver and land on unmanned boats to ensure the safety of the unmanned boat-machine joint identification and verification platform operation. The identification and verification task execution process is as follows: Figure 3 As shown;

[0148] Furthermore, during the execution of the UAV-UAV collaborative search mission, the main working mode of the UAV is divided into a complete process of autonomously arriving at the cruise starting point, accompanying the UAV, and autonomously arriving at the vicinity of the identification and verification target through global path planning to carry out identification and verification work. The specific work flow diagram is shown in the figure below. Figure 4 As shown, the MPC is a model predictive controller used for unmanned boats to accompany unmanned aerial vehicles;

[0149] Furthermore, in the execution of the UAV-UAV collaborative search mission, the main working modes of the UAV are wide-area target search, autonomous path planning patrol, target identification and verification, target identification and positioning, and autonomous return. The working processes of the UAV at different stages are as follows: Figure 5 As shown, the RTK is a high-precision differential positioning module.

[0150] During the process of identifying and verifying targets in the waters, drones can transmit information from the accident scene over long distances by being equipped with high-definition cameras and long-distance video transmission equipment. At the same time, unmanned boats can provide data relay and energy supply services for drones to ensure that information can be effectively transmitted back, allowing the command department to promptly understand the situation of suspicious targets in the waters, thereby effectively improving the efficiency and success rate of target identification and verification.

[0151] Based on this, the design of a water target identification system and verification method for a cross-domain unmanned system consisting of an unmanned boat and aircraft can effectively improve the efficiency of identification and verification and reduce costs.

[0152] Those skilled in the art know that, in addition to realizing the system and its various devices, modules, and units provided by the present invention in a purely computer-readable program code, it is entirely possible to realize the same functions in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers by logically programming the method steps. Therefore, the system and its various devices, modules, and units provided by the present invention can be considered as a hardware component, and the devices, modules, and units included therein for realizing various functions can also be regarded as structures within the hardware component; the devices, modules, and units for realizing various functions can also be regarded as both software modules for realizing the method and structures within the hardware component.

[0153] In the description of the present application, it should be understood that the terms "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating orientations or positional relationships, are based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present application.

[0154] The above describes the specific embodiments of the present invention. It should be understood that the present invention is not limited to the above specific embodiments, and those skilled in the art can make various changes or modifications within the scope of the claims, which does not affect the essence of the present invention. In the absence of conflict, the embodiments of the present application and the features in the embodiments can be combined with each other arbitrarily.

Claims

1. A portable water target identification and verification system, characterized in that: include: Payload platform, video acquisition module, target recognition module and target positioning module; The payload platform is used to support equipment for identifying and verifying water targets; Based on the load platform, the image information of the water area target is obtained through the video acquisition module, and the video data is obtained and the target is detected through the target recognition module to obtain the target information; The target information is acquired through the target positioning module to obtain the position information of the target.

2. The portable water area target identification and verification system according to claim 1 is characterized in that: The payload platform includes: an unmanned aerial vehicle flight platform, an unmanned boat and a ground monitoring platform; The UAV flight platform is used to detect, identify and locate suspicious targets on the water surface; the suspicious targets on the water surface are boats; The unmanned boat is a catamaran unmanned boat; the catamaran unmanned boat is used to receive the position information of the suspicious target on the water surface and plan a path; The ground monitoring platform is used to display the status information of the UAV and issue control instructions.

3. The portable water area target identification and verification system according to claim 2 is characterized in that: It can realize the wide-area search phase, the surface target identification and positioning phase, and the boat-aircraft joint approach identification and verification phase; The wide area search phase includes: The ground control platform issues instructions to the UAV and the unmanned boat to arrive at the patrol position and start patrolling; the video acquisition module obtains the image information of the water target through the UAV, and the UAV scores and sorts the target according to the target image information by a comprehensive weight sorting method to obtain the priority order of verification; The surface target identification and positioning stage includes: The target recognition module determines whether the UAV finds a suspicious target on the water surface; if the result is no, the UAV is ordered to return to the unmanned boat for charging after the patrol mission is completed; if the result is yes, the suspicious target on the water surface is identified and located according to the priority order of the verification to obtain target information; The joint boat-aircraft approach identification and verification phase includes: The target information is acquired through the target positioning module to obtain the target location information; the drone is instructed to send the target location information to the unmanned boat, and the unmanned boat is instructed to approach and verify the suspicious target on the water surface based on the target location information, and return after the verification is completed.

4. The portable water area target identification and verification system according to claim 2 is characterized in that: In the wide area search stage, the comprehensive weight sorting method, i.e., representing the evaluation value of the attribute by triangular fuzzy numbers, and then performing priority sorting according to the evaluation value; The expression of the triangular fuzzy number is: in, is a triangular fuzzy number, a i is the minimum value, b i is the most likely value, c i is the maximum value; For the maximization objective, the linear membership function of the triangular fuzzy number is expressed as: For the minimization objective, the linear membership function of the triangular fuzzy number is expressed as: For dynamic characteristics, the expression of the linear membership function is: in, is the linear membership function of the dynamic characteristics; In the comprehensive weight ranking method, the weight of each attribute is determined, that is, w I 、w T 、w D With w C , reflects the relative importance of each attribute in the verification task, and the mathematical expression is: In I +in T +in D +in C =1(5) Among them, w I is the weight of importance; w T is the weight of the threat level; w D is the weight of dynamic characteristics; w C is the weight of the verification difficulty; Based on the w I 、w T 、w D With w C , for each target i, calculate the comprehensive score, the mathematical expression is: Among them, S i is a comprehensive rating. is the degree of membership of target i in importance, is the membership of target i in terms of threat level, is the membership degree of target i in dynamic characteristics, is the membership of target i in the verification difficulty; the symbol · is the product; All targets are arranged in descending order according to the comprehensive score; the higher the score of the comprehensive score, the higher the priority of verification.

5. The portable water area target identification and verification system according to claim 3 is characterized in that: In the surface target recognition and positioning stage, the unmanned boat adjusts its own position information according to the position information of the drone; the position error between the unmanned boat and the drone is expressed as: Where ΔX uav-usv is the position error between the unmanned boat and the drone in the X-axis direction; ΔY uav-usv is the position error between the unmanned boat and the UAV in the Y-axis direction; ΔZ uav-usv is the position error between the unmanned boat and the drone in the Z-axis direction; X a , Y a With Z a They are the position information of the drone on the X axis, Y axis and Z axis respectively; s , Y s With Z s They are the position information of the unmanned boat on the X-axis, the Y-axis and the Z-axis respectively; Determine whether the unmanned boat meets the position error requirement, if yes, do not adjust the position of the unmanned boat; if no, adjust the position of the unmanned boat; The position error requirement is: ΔX uav-usv =0 ΔY uav-usv =0 ΔZ uav-usv =Z a Among them, Z a It is the position information of the drone on the Z axis.

6. The portable water area target identification and verification system according to claim 5 is characterized in that: Through the process control target cost function, the unmanned boat and the accompanying UAV maintain relative posture; The process control objective cost function is expressed as follows: Wherein, min J is the process control objective cost function; Q represents the weight of the situation keeping error; superscript T is the transposition identifier; X is the position of the unmanned boat; is the control amount of the unmanned boat; R is the weight of the control amount difference; ΔU a is the target value of the control variable of the unmanned boat; X uav The reference trajectory given to the UAV.

7. The portable water area target identification and verification system according to claim 6 is characterized in that: During the joint approach identification and verification phase, the unmanned boat * The global path planning algorithm obtains and verifies the target based on the approach path; In the wide area search stage, water area target image information is acquired through the YOLOv8 model.

8. The portable water area target identification and verification system according to claim 3 is characterized in that: The UAV keeps hovering at a fixed point and sends the location information of the target to the unmanned boat.

9. The portable water area target identification and verification system according to claim 3 is characterized in that: The drone executes a "bow"-shaped patrol path.

10. The portable water area target identification and verification system according to claim 3 is characterized in that: The drone collects 25 frames of image data per second, with a flight altitude of 50m and a patrol speed of 5m / s.

Images