Miniature rotor unmanned aerial vehicle full-autonomous pursuit method suitable for unknown denial environment

Abstract

The invention belongs to the technical field of unmanned aerial vehicle chasing. The invention provides a miniature unmanned rotorcraft full-autonomous pursuit method suitable for an unknown denial environment. According to the embodiment of the invention, the grid map is constructed in real time, and the planning safety of the unmanned aerial vehicle is ensured; and according to map information and unmanned aerial vehicle positioning information which are generated in real time, a search trajectory is planned according to a search strategy, and the unmanned aerial vehicle is guided to efficiently search for a potential target in a denial environment. In the searching process, the target is recognized in real time, and the specific position of the target is calculated; a smooth and stable tracking trajectory is planned by predicting a future possible motion trajectory of a target and combining dynamics of the unmanned aerial vehicle, target visibility constraints and the like. In the tracking process, the unmanned aerial vehicle continuously monitors the environment and the target state, and when the target meets the strike condition, the unmanned aerial vehicle switches to the target strike mode to execute autonomous target strike. And based on the current unmanned aerial vehicle and the target state, planning a trajectory meeting the strike effect requirement.

Description

Technical Field

[0001] Embodiments of the present disclosure relate to the technical field of drone pursuit, and particularly to a fully autonomous pursuit method for a micro-rotor drone applicable to an unknown denial environment. Background Art

[0002] In the era of rapid technological development, drone technology is being widely applied in multiple fields such as civilian use and rescue, and its application scope continues to expand. Among them, micro-rotor drones have become a key direction of technology research and application due to their simple structure, flexible and portable characteristics. However, the application of these drones still relies on human intervention and cannot truly achieve fully autonomous operation.

[0003] The fully autonomous pursuit of a micro-rotor drone refers to the autonomous decision-making, planning, and pursuit tasks being completed through an on-board sensor and algorithm system without human intervention. This technology has important value in defense, can significantly improve combat efficiency, reduce casualties, and become an indispensable intelligent tool in the future. However, there is currently no micro-rotor drone system that can achieve fully autonomous pursuit in an unknown denial environment.

[0004] Therefore, it is necessary to improve one or more problems existing in the above-mentioned related technical solutions.

[0005] It should be noted that this part aims to provide background or context for the technical solutions of the present disclosure stated in the claims. The description herein is not admitted to be prior art merely because it is included in this part. Summary of the Invention

[0006] The purpose of the embodiments of the present disclosure is to provide a fully autonomous pursuit method for a micro-rotor drone applicable to an unknown denial environment, thereby at least to some extent overcoming one or more problems caused by the limitations and defects of the related technologies.

[0007] According to the embodiments of the present disclosure, there is provided a fully autonomous pursuit method for a micro-rotor drone applicable to an unknown denial environment, the method comprising: Fusing and positioning the IMU information and sensor information of the drone to obtain the drone positioning information in the world coordinate system; Based on the drone pose and point cloud information, combining with the Bayesian estimation method, constructing a grid map in the world coordinate system to obtain map information; According to the drone positioning information and the map information, planning a search trajectory in combination with a search strategy; Based on the search trajectory, real-time acquiring image data, and using a target detection algorithm to identify whether there is a target in the image data. If there is, calculate the target position; Predict the motion trajectory of the target according to the target position, generate a tracking trajectory according to the motion trajectory, and determine whether the hitting condition is met; If the hitting condition is met, generate a hitting trajectory according to the motion trajectory of the target and strike the target.

[0008] Furthermore, in the step of constructing a grid map in the world coordinate system based on the UAV pose and point cloud information and combining with the Bayesian estimation method to obtain map information, it includes: Extract the point cloud data of the environment around the UAV using a depth camera or lidar; Based on the UAV pose and point cloud information, combine with the Bayesian estimation method to construct a grid map in the world coordinate system to obtain map information; where the grid map includes multiple grids.

[0009] Furthermore, the state of each grid in the grid map includes: Unknown state, free state, and occupied state; where The unknown state means that the grid has not been scanned by the sensor, the free state means that it has been scanned but there is no obstacle, and the occupied state means that it has been scanned and there is an obstacle.

[0010] Furthermore, in the step of planning a search trajectory according to the UAV positioning information and map information and combining with a search strategy, it includes: Limit the sensing range of the depth camera or lidar to be the same as that of the RGB camera; According to the UAV positioning information and map information, use an exploration algorithm to generate the planned target points of the UAV, According to the planned target points of the UAV, the dynamics and safety constraints of the UAV, use a numerical optimization method to generate a search trajectory.

[0011] Furthermore, in the step of, based on the search trajectory, obtaining image data in real time and using a target detection algorithm to identify whether there is a target in the image data, if there is, calculate the target position, it includes: During the search process, the UAV searches according to the search trajectory and obtains image data using the RGB camera; If a target is recognized, generate the target box position information in the image coordinate system; According to the target box position information in the image coordinate system, the internal and external parameters of the RGB camera, and the UAV positioning information, convert the target box position information from the image coordinate system to the world coordinate system to obtain the target position.

[0012] Furthermore, if no target is recognized, the UAV continues to search according to the search trajectory.

[0013] Further, in the step of predicting the movement trajectory of the target according to the target position, generating a tracking trajectory according to the movement trajectory, and determining whether the strike condition is met, it includes: Predict the movement trajectory of the target according to the target position; Using a numerical optimization method, combining the movement trajectory of the target, the dynamics of the UAV, and the target visibility constraint, generate the tracking trajectory of the UAV; Combining the map information and the real-time movement information of the target, determine whether the target meets the strike condition.

[0014] Further, if the strike condition is met, in the step of generating a strike trajectory according to the movement trajectory of the target and striking the target, it includes: If the strike condition is met, generate a strike trajectory according to the movement trajectory of the target, the dynamic characteristics of the UAV, and the safety constraint, so as to ensure that the target can be struck in the best attitude and angle.

[0015] The technical solutions provided by the embodiments of the present disclosure may include the following beneficial effects: In the embodiments of the present disclosure, through the above-mentioned fully autonomous pursuit method of a micro-rotor UAV applicable to unknown denial environments, on the one hand, a grid map is constructed in real time to ensure the safety of UAV planning; and according to the real-time generated map information and UAV positioning information, a search trajectory is planned according to the search strategy to guide the UAV to efficiently search for potential targets in the denial environment. During the search process, the target is identified in real time and the specific position of the target is calculated; once the enemy target is detected, the specific position and relevant feature information of the target will be returned, and the UAV state will be switched to the target tracking mode. By predicting the possible future movement trajectory of the target, a smooth and stable tracking trajectory is planned in combination with the dynamics of the UAV, target visibility constraints, etc. During the tracking process, the UAV will continuously monitor the environment and the target state. When the target meets the strike condition, the UAV will switch to the target strike mode and perform autonomous target strikes. Based on the current UAV and target states, a trajectory that meets the strike effect requirements is planned. On the other hand, after the UAV takes off, it can automatically complete the search and strike of the target without human intervention and can operate in a denial environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The accompanying drawings herein are incorporated into the specification and constitute a part of this specification, showing embodiments consistent with the present disclosure, and are used together with the specification to explain the principles of the present disclosure. Obviously, the accompanying drawings in the following description are only some embodiments of the present disclosure, and those of ordinary skill in the art can obtain other drawings based on these drawings without creative efforts.

[0017] Figure 1A step diagram of a fully autonomous pursuit method for a micro-rotor UAV applicable to an unknown denial environment in an exemplary embodiment of the present disclosure; Figure 2 A schematic diagram of a rotor UAV with autonomous pursuit capabilities in an exemplary embodiment of the present disclosure. Detailed implementation manners

[0018] Example embodiments will now be described more fully with reference to the accompanying drawings. However, the example embodiments can be implemented in various forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that this disclosure will be more thorough and complete, and will fully convey the concept of the example embodiments to those skilled in the art. The features, structures, or characteristics described may be combined in any suitable manner in one or more embodiments.

[0019] In addition, the accompanying drawings are only schematic illustrations of the embodiments of the present disclosure and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and thus their repeated description will be omitted. Some of the block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities.

[0020] In this example embodiment, a fully autonomous pursuit method for a micro-rotor UAV applicable to an unknown denial environment is provided. Referring to Figure 1 as shown, the fully autonomous pursuit method for a micro-rotor UAV applicable to an unknown denial environment may include: Step S101 to Step S106.

[0021] Step S101: Fuse and locate the IMU information and sensor information of the UAV to obtain the UAV positioning information in the world coordinate system; Step S102: Based on the UAV pose and point cloud information, combine with the Bayesian estimation method to construct a grid map in the world coordinate system to obtain the map information; Step S103: According to the UAV positioning information and the map information, combine with a search strategy to plan a search trajectory; Step S104: Based on the search trajectory, obtain image data in real time, and use an object detection algorithm to identify whether there is an object in the image data. If there is, calculate the object position; Step S105: Predict the movement trajectory of the object according to the object position, generate a tracking trajectory according to the movement trajectory, and determine whether the strike condition is met; Step S106: If the strike condition is met, generate a strike trajectory according to the movement trajectory of the object and strike the object.

[0022] Through the above-mentioned fully autonomous pursuit method of the micro-rotor UAV applicable to unknown denied environments, on the one hand, a grid map is constructed in real time to ensure the safety of the UAV's planning; and according to the real-time generated map information and the UAV's positioning information, a search trajectory is planned according to the search strategy to guide the UAV to efficiently search for potential targets in the denied environment. During the search process, the target is identified in real time and the specific position of the target is calculated; once the enemy target is detected, the specific position and relevant feature information of the target will be returned, and the UAV state will be switched to the target tracking mode. By predicting the possible future movement trajectory of the target and combining the dynamics of the UAV, target visibility constraints, etc., a smooth and stable tracking trajectory is planned. During the tracking process, the UAV will continuously monitor the environment and the target state. When the target meets the strike conditions, the UAV will switch to the target strike mode and perform autonomous target strikes. Based on the current UAV and target states, a trajectory that meets the requirements of the strike effect is planned. On the other hand, after the UAV takes off, it can automatically complete the search and strike of the target without human intervention and can operate in a denied environment.

[0023] Next, reference will be made to Figures 1 to 2 to describe each step of the above-mentioned fully autonomous pursuit method of the micro-rotor UAV applicable to unknown denied environments in the present exemplary embodiment in more detail.

[0024] In step S101, the IMU information and sensor information of the UAV are fused for positioning to obtain the UAV positioning information in the world coordinate system.

[0025] Specifically, autonomous positioning and navigation. By fusing the IMU (Inertial Measurement Unit) information with the sensor information of the depth camera or lidar, multi-sensor fusion positioning is realized, and the pose of the UAV in the world coordinate system is accurately calculated, providing an accurate positioning basis for subsequent tasks.

[0026] In step S102, based on the UAV pose and point cloud information, combined with the Bayesian estimation method, a grid map in the world coordinate system is constructed to obtain the map information Specifically, autonomous perception and mapping. According to the pose information (i.e., the UAV pose) calculated by the UAV in real time, the depth camera or lidar is used to extract the point cloud data of the surrounding environment. By removing the point cloud information around the target, the environmental modeling effect is further optimized. Combined with the Bayesian estimation method, a grid map in the world coordinate system is constructed to provide the UAV with the ability of global environmental perception for target search, tracking and strike.

[0027] In steps S103 and S104, according to the UAV positioning information and the map information, a search trajectory is planned in combination with the search strategy; based on the search trajectory, image data is obtained in real time, and the target detection algorithm is used to identify whether there is a target in the image data. If so, the target position is calculated.

[0028] Specifically, for target detection and recognition. Image data obtained by an RGB camera is used to identify target objects (i.e., targets) through a target detection algorithm. If a target is recognized, the position information of the target box in the image coordinate system is generated. Combining the internal and external parameters of the RGB camera and the real-time pose of the drone, the target position is accurately transformed from the image coordinate system to the world coordinate system to achieve the spatial positioning of the target. If no target is recognized, no target information is generated.

[0029] Autonomous target search. Optimize the perception range, limit the perception range of the depth camera or lidar to be the same as that of the RGB camera, so as to transform the autonomous target search problem into an autonomous exploration problem. Based on the real-time generated environmental map (i.e., map information) and the positioning information of the drone, using frontier-based or sampling-based exploration algorithms, generate the planned target points of the drone, and combine the dynamics of the drone and safety constraints, etc., to generate a smooth planned trajectory (i.e., search trajectory) through numerical optimization methods to guide the drone to efficiently search for potential targets in an unknown environment. During the search process, continuously analyze the environmental information, detect the target object in real time and return the specific position and relevant feature data of the target. When the target is successfully detected, the system will immediately switch to the target tracking mode.

[0030] In step S105, predict the movement trajectory of the target according to the target position, generate a tracking trajectory according to the movement trajectory, and determine whether it meets the strike conditions.

[0031] Specifically, for autonomous target tracking. According to the position information of the target, predict the future movement trajectory of the target. Also use numerical optimization methods to plan a smooth and stable tracking trajectory in combination with the dynamics of the drone, target visibility constraints, etc., so as to ensure that the drone can accurately and efficiently follow the target in a dynamic environment. At the same time, the drone will combine the map environment and the real-time movement information of the target to determine whether the target meets the strike conditions. Once the target meets the strike conditions, the system will immediately switch to the target strike mode. In step S106, if the strike conditions are met, generate a strike trajectory according to the movement trajectory of the target and strike the target.

[0032] Specifically, for autonomous target strike. After receiving the strike command, the drone will comprehensively consider the dynamic characteristics and safety constraints of the drone and plan an accurate strike trajectory in real time. By predicting the future movement trajectory of the target, the planned terminal trajectory highly coincides with the target prediction trajectory to ensure that the drone can complete the strike mission with the best flight attitude, time and angle. When the target enters the best strike range and the fuse is triggered, the warhead will be detonated to achieve the strike target.

[0033] In addition, the flight controller of the UAV is responsible for calculating the motion trajectories generated during the target search, tracking, and striking phases and converting them into specific control commands. The controller precisely controls the attitude and power output of the UAV to ensure that the UAV flies smoothly along the predetermined trajectory.

[0034] In a specific embodiment, the hardware platform carried by the UAV includes a battery, an electronic speed controller, a warhead, a fuse, a depth camera or lidar, an RGB camera, an on-board computer, and a flight controller. At the algorithm level, it includes an autonomous positioning and navigation module, an autonomous perception and mapping module, a target detection and recognition module, an autonomous target search module, a tracking module, a striking module, and a flight control module.

[0035] The autonomous positioning and navigation module ensures the accurate positioning and stable navigation of the UAV in a denied environment; the autonomous perception and mapping module is responsible for constructing the environmental map in real time to ensure the safety of the UAV's planning; the target detection and recognition module identifies the target and calculates the specific position of the target through methods such as machine learning. The autonomous target search module plans the search trajectory according to the real-time generated map information and the UAV's positioning information and guides the UAV to efficiently search for potential targets in a denied environment. During the search process, once the target detection and recognition module detects an enemy target, it will return the specific position and relevant feature information of the target and switch the UAV's state to the target tracking mode. In the autonomous target tracking module, a smooth and stable tracking trajectory is planned by predicting the possible future motion trajectory of the target and combining the dynamics of the UAV, target visibility constraints, etc. During the tracking process, the system continuously monitors the environment and the target state. When the target meets the striking conditions, the UAV will switch to the target striking mode and execute the autonomous target striking algorithm. After receiving the striking command, the autonomous target striking module plans a trajectory that meets the requirements of the striking effect based on the current UAV and target states. The entire process relies on the highly integrated and real-time response capabilities of the UAV's autonomous algorithm system to achieve full autonomous flight without human intervention.

[0036] In a specific embodiment, as Figure 2 shown, it is a schematic diagram of a rotary-wing UAV with autonomous pursuit capabilities. The rotary-wing UAV mainly includes a depth camera, an RGB camera, an on-board computer, motors, and a flight controller; the main sensing devices of the rotary-wing UAV include the depth camera, the RGB camera, and the IMU in the flight controller, where the depth camera and the IMU are used for autonomous positioning and navigation and autonomous perception and mapping, and the RGB camera is used for target detection and recognition; the on-board computer is used to run all algorithms except the flight control module; the flight controller receives the planning instructions calculated by the on-board computer and calculates the control instructions to control the rotation of the motors to make the UAV fly according to the plan; the fuse and the warhead are used for killing and are omitted in the figure.

[0037] The specific process of this fully autonomous pursuit method for drones is as follows: 1. Autonomous positioning and navigation. By fusing the information of the IMU (Inertial Measurement Unit) with the sensor information of the depth camera or lidar, multi-sensor fusion positioning is achieved, and the pose of the drone in the world coordinate system is accurately calculated, providing an accurate positioning basis for subsequent tasks.

[0038] 2. Autonomous perception and mapping. According to the pose information of the drone calculated in real time, the depth camera or lidar is used to extract the point cloud data of the surrounding environment. By removing the point cloud information around the target, the environmental modeling effect is further optimized. Combining with the Bayesian estimation method, a grid map in the world coordinate system is constructed to provide the drone with the ability of global environmental perception for target search, tracking and striking. The task map can be expressed as , where represents the length, width and height of the global map area. Each grid in the grid map can be divided into three states, namely "unknown", "free" and "occupied". Among them, "unknown" means that the grid has not been scanned by the sensor, "free" means that it has been scanned but there is no obstacle, and "occupied" means that it has been scanned and there is an obstacle. The grid information in the area is expressed as .

[0039] 3. Target detection and recognition. Using the image data obtained by the RGB camera, the target object is recognized through the target detection algorithm. If the target is recognized, the position information of the target box in the image coordinate system is generated. Combining the internal and external parameters of the RGB camera and the real-time pose of the drone, the target position is converted from the image coordinate system to the world coordinate system to achieve the spatial positioning of the target. If the target is not recognized, no target information is generated. The sensing range of the RGB camera is expressed as , where represents the detection distance, represents the horizontal field of view angle, represents the vertical field of view angle.

[0040] 4. Autonomous target search. Optimize the sensing range, limit the sensing range of the depth camera or lidar to be consistent with the sensing range of the RGB camera, so as to transform the autonomous target search problem into an autonomous exploration problem. In this system, and respectively represent the position trajectory and yaw angle trajectory of the drone during the search process, represents the position trajectory of the target during the search process. The autonomous exploration problem of the drone refers to solving the optimal position trajectory and yaw angle trajectory so that the drone can move safely in the three-dimensional environment and can reach the target in the shortest time A moment when the sensor can cover the target position , and the coverage condition is expressed as . The specific task can be abstracted into the following optimization problem:

[0041] where respectively represent the target positions in the target search stage. is the set of inequality constraints during the movement of the UAV, such as trajectory smoothness, safety, and dynamic feasibility constraints; is the set of equality constraints during the movement of the UAV, such as trajectory initial and terminal state constraints. The UAV solves the trajectory and after discovering the target position at , it enters the target tracking stage. The present invention processes the autonomous target search problem in segments, that is, the front-end UAV target point decision-making and the back-end search-oriented motion planning. Based on the real-time generated environmental map and the UAV's positioning information, using frontier-based or sampling-based exploration decision algorithms, an observation point set of the UAV is generated, and then the planned target point of the UAV is obtained by using a utility function or operations research optimization method. In the back-end motion planning, first, a path search algorithm is used to search for the corresponding path in the grid map, and then the path is fitted to a polynomial curve. Based on differential flatness, various constraints of the UAV are transformed into objective functions, and a smooth planned trajectory is generated through numerical optimization methods to guide the UAV to efficiently search for potential targets in an unknown environment. The trajectory optimization problem can be expressed as:

[0042] where represents the parameters of the polynomial curve, where represents the objective function corresponding to the search trajectory, represents the energy objective function, is the safety objective function, is the dynamic feasibility objective function, is the shortest time objective function, are the corresponding weight coefficients respectively. During the search process, the target detection and recognition module continuously analyzes the RGB image information, real-time detects the target object and returns the specific position and relevant feature data of the target. When the target is successfully detected, that is, when the coverage condition is satisfied, the system will immediately switch to the target tracking mode.

[0043] 5. Autonomous target tracking. First, based on the position information of the target, the future motion trajectory of the target is predicted by means of filtering, curve fitting or deep learning. The main purpose of target tracking is to make the UAV continuously approach the target until the strike condition is met, so as to improve the hit rate in the strike phase. In this process, and respectively represent the position trajectory and yaw angle trajectory of the UAV during the tracking process, represents the position trajectory of the target during the tracking process, and the optimal position trajectory needs to be solved in this task stage and yaw angle such that the UAV can meet the strike condition at the shortest time moment, and the specific task can be abstracted into the following optimization problem:

[0044] where respectively represent the target positions in the target tracking stage, is the set of inequality constraints of the UAV during the movement, such as trajectory smoothness, safety, dynamic feasibility, target visibility and distance constraints, etc.; is the set of equality constraints of the UAV during the movement, such as trajectory initial and end point state constraints. The UAV enters the target strike stage after meeting the strike condition at the moment . The present invention also processes the autonomous target tracking problem in two parts, namely, motion planning for target tracking and strike condition judgment. For motion planning, similar to the motion planning method for autonomous search, numerical optimization methods are used to plan a smooth and stable tracking trajectory in combination with the dynamics of the UAV, target visibility constraints, etc., so as to ensure that the UAV can accurately and efficiently follow the target in a dynamic environment. The trajectory optimization problem can be expressed as:

[0045] where represents the parameters of the polynomial curve, where represents the objective function corresponding to the tracking trajectory, represents the energy objective function, is the safety objective function, is the dynamic feasibility objective function, is the shortest time objective function, and respectively represent the visibility objective function and the distance objective function. are the corresponding weight coefficients respectively. For strike condition judgment, the UAV will combine the map environment and the real-time motion information of the target to judge whether the target meets the strike condition , the strike condition can be jointly determined by the environmental complexity, the target movement speed, and the distance between the UAV and the target. Once the target meets the strike condition, the system will immediately switch to the target strike mode.

[0046] 6. Autonomous target strike. Similar to the target tracking problem, the purpose of target strike is to make the UAV continuously approach the target until the damage condition is met at a certain moment, that is, to solve the optimal trajectory and , so that the UAV can meet the damage condition at the shortest time . During this process, and represent the position trajectory and yaw angle trajectory of the UAV during the strike process respectively, represents the position trajectory of the target during the strike process. The specific task can be abstracted into the following optimization problem:

[0047] where represent the target positions in the target strike stage respectively. is the set of inequality constraints of the UAV during movement, such as trajectory smoothness, safety, and dynamic feasibility constraints, etc.; is the set of equality constraints of the UAV during movement, such as initial and terminal state constraints. The UAV meets the damage condition and then can detonate the warhead to terminate the mission. The present invention also processes the autonomous target strike problem in two parts in segments, that is, the motion planning for target strike and the damage condition judgment. For the motion planning, the numerical optimization method is also used to plan the accurate strike trajectory in real time.

[0048]

[0049] where represents the objective function corresponding to the strike trajectory, represents the energy objective function, is the safety objective function, is the dynamic feasibility objective function, is the shortest time objective function, is the distance objective function for target strike, are the corresponding weight coefficients respectively. Its purpose is to make the end segment trajectory of the plan highly coincide with the target prediction trajectory, ensuring that the UAV can complete the strike mission with the best flight attitude, time, and angle.

[0050] When the target enters the best strike range and the fuse is triggered, the damage condition is met , the warhead will be detonated to achieve the strike target.

[0051] 7. Flight control. The flight controller of the UAV is responsible for calculating the motion trajectories generated during the target search, tracking, and strike phases and converting them into specific control commands. The controller precisely controls the attitude and power output of the UAV to ensure that the UAV flies smoothly along the predetermined trajectory.

[0052] Through the above-mentioned fully autonomous pursuit method of the micro-rotor UAV applicable to unknown denial environments, on the one hand, a grid map is constructed in real time to ensure the safety of the UAV's planning; and according to the real-time generated map information and the UAV's positioning information, the search trajectory is planned according to the search strategy to guide the UAV to efficiently search for potential targets in the denial environment. During the search process, the target is identified in real time and the specific position of the target is calculated; once the enemy target is detected, the specific position and relevant feature information of the target will be returned, and the UAV state will be switched to the target tracking mode. By predicting the possible future motion trajectory of the target and combining the dynamics of the UAV, target visibility constraints, etc., a smooth and stable tracking trajectory is planned. During the tracking process, the UAV will continuously monitor the environment and the target state. When the target meets the strike conditions, the UAV will switch to the target strike mode and perform autonomous target strikes. Based on the current UAV and target states, a trajectory that meets the strike effect requirements is planned. On the other hand, after the UAV takes off, it can automatically complete the search and strike of the target without human intervention and can operate in a denial environment.

[0053] It should be understood that the orientation or positional relationship indicated by the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise", etc. in the above description is based on the orientation or positional relationship shown in the drawings, and is only for the convenience of describing the embodiments of the present disclosure and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be construed as a limitation on the embodiments of the present disclosure.

[0054] In addition, the terms "first" and "second" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include one or more of such features. In the description of the embodiments of the present disclosure, "a plurality" means two or more unless otherwise specifically defined.

[0055] In the embodiments of the present disclosure, unless otherwise clearly defined or limited, terms such as "install", "connect", "couple", "fix", etc. shall be understood in a broad sense. For example, it may be a fixed connection, a detachable connection, or an integral body; it may be a mechanical connection or an electrical connection; it may be directly connected, or indirectly connected through an intermediate medium, and it may be the communication inside two elements or the interaction relationship between two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present disclosure can be understood according to specific circumstances.

[0056] In the embodiments of the present disclosure, unless otherwise clearly defined or limited, the first feature being "on" or "under" the second feature may include the direct contact between the first and second features, or may include the situation where the first and second features are not in direct contact but in contact through other features therebetween. Moreover, the first feature being "above", "over" and "on top of" the second feature includes that the first feature is directly above and obliquely above the second feature, or merely means that the horizontal height of the first feature is higher than that of the second feature. The first feature being "under", "below" and "beneath" the second feature includes that the first feature is directly below and obliquely below the second feature, or merely means that the horizontal height of the first feature is lower than that of the second feature.

[0057] In the description of this specification, the descriptions referring to terms such as "one embodiment", "some embodiments", "example", "specific example" or "some examples", etc. mean that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present disclosure. In this specification, the schematic descriptions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in a suitable manner in any one or more embodiments or examples. In addition, those skilled in the art can combine and mix the different embodiments or examples described in this specification.

[0058] After considering the specification and practicing the invention disclosed herein, those skilled in the art will readily conceive of other implementations of the present disclosure. This application is intended to cover any variations, uses, or adaptations of the present disclosure, which follow the general principles of the present disclosure and include the known common general knowledge or conventional technical means in the technical field not disclosed in the present disclosure. The specification and the embodiments are only regarded as exemplary, and the true scope and spirit of the present disclosure are pointed out by the appended claims.

Claims

1. A fully autonomous pursuit method for a micro-rotor drone in an unknown denied environment, characterized in that: The method includes: The IMU information and sensor information of the drone are integrated to obtain the drone positioning information in the world coordinate system; Based on the UAV position and point cloud information, combined with the Bayesian estimation method, a grid map in the world coordinate system is constructed to obtain map information; Plan the search trajectory based on the drone positioning information and map information combined with the search strategy; Based on the search trajectory, image data is acquired in real time, and the target detection algorithm is used to identify whether there is a target in the image data. If so, the target position is calculated; Predict the target's motion trajectory based on the target's position, generate a tracking trajectory based on the motion trajectory, and determine whether the strike conditions are met; If the strike conditions are met, a strike trajectory is generated according to the target's motion trajectory, and the target is struck.

2. The fully autonomous pursuit method for a micro-rotor drone in an unknown denied environment according to claim 1 is characterized in that: Based on the UAV posture and point cloud information, combined with the Bayesian estimation method, a grid map in the world coordinate system is constructed to obtain map information, including: Use depth cameras or lidar to extract point cloud data of the drone’s surroundings; Based on the UAV posture and point cloud information, combined with the Bayesian estimation method, a grid map in the world coordinate system is constructed to obtain map information; wherein the grid map includes multiple grids.

3. The fully autonomous pursuit method for a micro-rotor drone in an unknown denied environment according to claim 2 is characterized in that: The status of each grid in the grid map includes: unknown state, free state and occupied state; among them, The unknown state means that the grid has not been scanned by the sensor, the free state means that it has been scanned but there are no obstacles, and the occupied state means that it has been scanned and there are obstacles.

4. The fully autonomous pursuit method for a micro-rotor drone in an unknown denied environment according to claim 2 is characterized in that: The steps of planning the search trajectory based on the drone positioning information and map information and combined with the search strategy include: Limit the perception range of the depth camera or lidar to that of the RGB camera; Based on the drone positioning information and map information, the exploration algorithm is used to generate the planned target points of the drone. According to the planned target point of the UAV, the dynamics of the UAV and the safety constraints, the search trajectory is generated using numerical optimization methods.

5. The fully autonomous pursuit method for a micro-rotor drone in an unknown denied environment according to claim 4 is characterized in that: Based on the search trajectory, image data is acquired in real time, and a target detection algorithm is used to identify whether a target exists in the image data. If so, the step of calculating the target position includes: During the search process, the drone searches according to the search trajectory and uses the RGB camera to obtain image data; If the target is recognized, the target frame position information in the image coordinate system is generated; According to the target frame position information in the image coordinate system, the internal and external parameters of the RGB camera and the drone positioning information, the target frame position information is converted from the image coordinate system to the world coordinate system to obtain the target position.

6. The fully autonomous pursuit method for a micro-rotor drone in an unknown denied environment according to claim 5 is characterized in that: If the target is not identified, the drone continues to search according to the search trajectory.

7. The fully autonomous pursuit method for a micro-rotor drone in an unknown denied environment according to claim 5 is characterized in that: The steps of predicting the target's motion trajectory according to the target position, generating a tracking trajectory according to the motion trajectory, and determining whether the strike condition is met include: Predict the target's trajectory based on its position; The tracking trajectory of the UAV is generated by using numerical optimization methods, combining the target's motion trajectory, the UAV's dynamics and the target visibility constraints; Combine map information with the target's real-time movement information to determine whether the target meets the strike conditions.

8. The fully autonomous pursuit method for a micro-rotor drone in an unknown denied environment according to claim 7 is characterized in that: If the strike condition is met, the step of generating a strike trajectory according to the target's motion trajectory and striking the target includes: If the strike conditions are met, a strike trajectory is generated based on the target's motion trajectory, the UAV's dynamic characteristics and safety constraints to ensure that the target can be struck at the best posture and angle.

Images