A method and device for unmanned aerial vehicle trajectory planning based on conflict reduction

Abstract

The application discloses a kind of unmanned aerial vehicle trajectory planning method and device based on conflict reduction, obtain unmanned aerial vehicle initial trajectory information and obstacle information, based on B spline curve generation unmanned aerial vehicle next time period's trajectory information;According to the trajectory information of neighborhood unmanned aerial vehicle, first collision detection is carried out to trajectory information;When there is collision, if the priority of the first trajectory information list neighborhood unmanned aerial vehicle is less than the priority of current unmanned aerial vehicle, trajectory information and the trajectory information of neighborhood unmanned aerial vehicle in second trajectory information list are secondly collided;When there is collision, if the priority of the first trajectory information list neighborhood unmanned aerial vehicle is less than the priority of current unmanned aerial vehicle, execute trajectory information;The trajectory information of unmanned aerial vehicle can be generated by B spline curve in the application, smooth trajectory planning and control can be realized, the problem of trajectory conflict in distributed trajectory update is solved, and collision is avoided.

Description

Technical Field

[0001] This invention belongs to the field of trajectory planning technology for unmanned aerial vehicle (UAV) swarm systems, and particularly relates to a UAV trajectory planning method and apparatus based on conflict reduction. Background Technology

[0002] During flight, drone swarms may need to pass through flight scenarios with dense obstacles at any time. To ensure the physical safety of the drone swarm, the drone swarm not only needs to perceive and avoid external obstacles, but also needs to avoid collisions within the swarm, so as to generate a collision-free and kinematically feasible path for each drone in dense scenarios.

[0003] Most existing technologies employ a centralized approach to find the optimal trajectory for each drone, but this method has high time and space complexity and is unsuitable for real-time online systems. Although some distributed methods have been proposed, the trajectories they plan are usually conservative, and they may not find a solution in dense obstacle scenarios. Furthermore, drone trajectory updates are prone to conflicts, causing collisions among drones in the swarm. Summary of the Invention

[0004] The purpose of this invention is to provide a method and apparatus for UAV trajectory planning based on conflict reduction, so as to solve the problem of trajectory conflict in the distributed trajectory update process.

[0005] This invention adopts the following technical solution: a UAV trajectory planning method based on conflict reduction, comprising the following steps:

[0006] Acquire the initial trajectory information and obstacle information of the UAV, generate the trajectory information of the UAV in the next time period based on the B-spline curve, and simultaneously receive the trajectory information of neighboring UAVs and store it in the first trajectory information list.

[0007] The first collision detection is performed on the trajectory information based on the trajectory information of neighboring drones;

[0008] When a collision occurs between the trajectory information of a neighboring drone in the first trajectory information list and the trajectory information of the current drone, if the priority of the neighboring drone in the first trajectory information list is lower than the priority of the current drone, a second collision detection is performed between the trajectory information and the trajectory information of the neighboring drone in the second trajectory information list. The second trajectory information list contains the trajectory information of the neighboring drones received during the first collision detection. When a collision occurs between the trajectory information of a neighboring drone in the second trajectory information list and the trajectory information of the current drone, if the priority of the neighboring drone in the first trajectory information list is lower than the priority of the current drone, the trajectory information is executed.

[0009] Furthermore, the trajectory information of the UAV for the next time period generated based on the B-spline curve includes:

[0010] Generate the velocity sampling point set and acceleration sampling point set of the UAV for the next time period based on the initial trajectory information and obstacle information;

[0011] By filtering the velocity sampling point set and the acceleration sampling point set, the optimal velocity sampling point and the optimal acceleration sampling point are obtained;

[0012] Generate trajectory control points for the next time period based on the optimal velocity sampling point and the optimal acceleration sampling point;

[0013] Generate the drone's trajectory for the next time period based on the trajectory control points.

[0014] Furthermore, the selection of velocity sampling point sets and acceleration sampling point sets includes:

[0015] Based on the convex hull property and cost function of B-spline curves, the velocity sampling point set and the acceleration sampling point set are filtered sequentially.

[0016] Furthermore, the filtering of velocity and acceleration sampling point sets based on the convex hull property of B-spline curves includes:

[0017] ,

[0018] in, Let u represent the set of velocity sampling points after convex hull screening for the u-th UAV. This represents the trajectory control point corresponding to the velocity sampling point of the u-th UAV in the m-th time period. This represents the trajectory control point corresponding to the velocity sampling point of the u-th UAV in the (m-1)-th time period. This represents the velocity sampling point of the u-th drone. Indicates the time span, This represents the set of velocity sampling points for the u-th drone. This represents the convex hull of the u-th drone in the m-th time interval. This represents the collision model of the u-th drone relative to a static obstacle. Indicates obstacles convex hull, Represents a set of static obstacles. This represents the convex hull of the u-th and v-th drones relative to the B-spline curve in the m-th time interval. This is a relative collision model between the u-th drone and the v-th drone. Let u represent the set of acceleration sampling points after convex hull filtering for the u-th UAV. This represents the speed control point of the u-th drone in the m-th time interval. This represents the speed control point of the u-th drone in the (m-1)-th time interval. This represents the acceleration sampling point of the u-th UAV. Indicates the position of the u-th drone on the coordinate axis. The maximum speed on the top, Let u represent the set of acceleration sampling points for the u-th UAV. This represents the set of neighboring drones within the communication range of the u-th drone.

[0019] Furthermore, the cost function is:

[0020] ,

[0021] in, This represents the cumulative pricing function. This represents the trajectory control point of the u-th UAV in the i-th time period of the cumulative pricing function. This represents the heuristic cost function.

[0022] Furthermore, a neighboring drone has a higher priority than the current drone when any of the following conditions are met;

[0023] The distance between neighboring drones and the target point is less than the distance between the current drone and the target point;

[0024] The neighboring drone has already executed the trajectory information for the next time period.

[0025] Furthermore, when generating trajectory information for the drone in the next time period, the trajectory information is sent to neighboring drones.

[0026] Furthermore, after the first collision detection, if no collision occurs, the trajectory information is executed.

[0027] Furthermore, if the priority of the UAV in the neighborhood of the first trajectory information list is greater than or equal to the priority of the current UAV, the trajectory information of the UAV for the next time period is regenerated based on the B-spline curve.

[0028] Another technical solution of the present invention: a UAV trajectory planning device based on conflict reduction, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-described method.

[0029] The beneficial effects of this invention are: by generating UAV trajectory information through B-spline curves, this invention can achieve smooth trajectory planning and control, flexibly adjust the shape and path of the trajectory, and at the same time, by combining obstacle information and neighboring UAV information, it can solve the trajectory conflict problem in distributed trajectory updates and avoid collisions. Attached Figure Description

[0030] Figure 1 This is a flowchart illustrating a UAV trajectory planning method based on conflict reduction, according to an embodiment of the present invention.

[0031] Figure 2 This is a schematic diagram of a UAV trajectory planning method based on conflict reduction, according to an embodiment of the present invention. Detailed Implementation

[0032] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0033] When drone swarms fly in dense obstacle scenarios, the dense obstacles pose a significant challenge to the safety, stability, and reliability of drone trajectory planning. Therefore, this invention proposes a cooperative conflict mitigation method for complex task environments, resolving the trajectory conflict problem between drones in a distributed and asynchronous communication architecture. This enables the drone swarm to successfully complete all tasks from both physical and mission safety perspectives.

[0034] This method is implemented using asynchronous communication, making it easier to deploy in the real world. Benefiting from the favorable mathematical properties of B-spline curves (especially convex hull and local control properties), this invention uses B-spline curves to represent the UAV's trajectory and comprehensively considers UAV kinematic constraints, collision avoidance constraints, and UAV trajectory smoothness indices, modeling the trajectory planning problem as a constrained optimization problem for solution. Based on this, a cooperative conflict reduction method is designed to address potential conflicts in trajectory updates between UAVs under an asynchronous communication architecture.

[0035] This invention discloses a method for UAV trajectory planning based on conflict reduction, such as... Figure 1 As shown, the process includes the following steps: acquiring initial trajectory information and obstacle information of the UAV; generating trajectory information for the UAV in the next time period based on B-spline curves; simultaneously receiving trajectory information of neighboring UAVs and storing it in a first trajectory information list; performing a first collision detection on the trajectory information based on the neighboring UAV trajectory information; when a collision occurs between the trajectory information of a neighboring UAV in the first trajectory information list and the current UAV trajectory information, if the priority of the neighboring UAV in the first trajectory information list is lower than the priority of the current UAV, performing a second collision detection on the trajectory information with the trajectory information of the neighboring UAV in a second trajectory information list; wherein, the second trajectory information list consists of the neighboring UAV trajectory information received during the first collision detection process; when a collision occurs between the trajectory information of a neighboring UAV in the second trajectory information list and the current UAV trajectory information, if the priority of the neighboring UAV in the second trajectory information list is lower than the priority of the current UAV, executing the trajectory information.

[0036] This invention generates UAV trajectory information using B-spline curves, enabling smooth trajectory planning and control, flexible adjustment of trajectory shape and path, and the resolution of trajectory conflict issues in distributed trajectory updates by combining obstacle information and neighboring UAV information, thus avoiding collisions.

[0037] This invention employs a B-spline curve (B-spline curve) method for representing UAV trajectories. By adjusting control points and node vectors, the shape and path of the trajectory can be flexibly adjusted to meet different application requirements. B-spline curve-based trajectory representation offers advantages such as smoothness, adjustability, high-order approximation capability, interpolation capability, and concise mathematical representation in the UAV control domain, making it an excellent trajectory planning and control method. Based on these characteristics, B-spline curves can effectively perform online trajectory replanning while satisfying constraints.

[0038] In one embodiment, generating trajectory information of the UAV for the next time period based on B-spline curves includes: generating a set of velocity sampling points and a set of acceleration sampling points for the UAV for the next time period based on initial trajectory information and obstacle information; filtering the set of velocity sampling points and the set of acceleration sampling points to obtain the optimal velocity sampling points and the optimal acceleration sampling points; generating trajectory control points for the next time period based on the optimal velocity sampling points and the optimal acceleration sampling points; and generating the trajectory of the UAV for the next time period based on the trajectory control points.

[0039] B-spline curves possess excellent properties such as convex hull and local adjustability. These properties can greatly simplify the considerations in the trajectory planning problem of UAVs. Therefore, this invention represents the trajectory using B-spline curves.

[0040] Necessarily, we will first introduce trajectory modeling based on B-splines, and then briefly introduce the properties of B-spline curves in conjunction with trajectory representation requirements. Given... Control points and Time nodes The time vector formed by these control points represents the initial trajectory points planned by the UAV. The k-th order B-spline curve can be represented as:

[0041] (1)

[0042] in, Let the k-th order B-spline basis function be denoted by the de Boer-Cox recurrence relation in equation (2):

[0043] (2)

[0044] The above-mentioned number of control points n, number of elements in the time vector m, and order of the B-spline function satisfy When the time vector is uniformly distributed, i.e., it satisfies When the B-spline curve is uniform, it is called a uniform B-spline curve; otherwise, it is called a non-uniform B-spline curve. (Half-open interval) This is called the i-th time span.

[0045] To simplify the calculation of B-spline basis functions, this invention uses uniform B-spline curves with fixed node interval sizes. The B-spline curve and its derivatives can then be expressed in matrix form as follows:

[0046] (3)

[0047] in, Let represent the normalized basis vectors, d represent the differentiation, and l represent the order of the derivative. This indicates that it is a B-spline curve of order. The uniquely determined constant matrix This represents a matrix composed of control points at corresponding locations.

[0048] (4) (5)

[0049] Represents the size of the node interval. The first in Line number The elements of the column are calculated as follows:

[0050] (6)

[0051] Considering that the order of the B-spline curve directly affects the shape of the curve, the higher the order, the smoother the curve, the greater the deviation from the control point, and the higher the computational complexity, this embodiment of the invention uses a cubic B-spline curve (…). =3) indicates a drone The trajectory of is in matrix form:

[0052] (7)

[0053] in, Indicates drone Similarly, given the variable or vector, its velocity trajectory can be obtained as follows:

[0054] (8)

[0055] The acceleration trajectory is:

[0056] (9)

[0057] in, , and These represent the trajectory, velocity trajectory, and acceleration trajectory of the UAV u, respectively. , and They represent drones The There are position, velocity, and acceleration control points that satisfy the following equation:

[0058] (10)

[0059] B-spline curves possess characteristics such as non-negativity, uniformity of division, segmentation, local adjustment, and convex hull properties. Moreover, the derivative of a B-spline curve is still a B-spline. This invention only elaborates on the properties that are mainly used in trajectory representation and explains the angle from which these properties are utilized in trajectory representation.

[0060] (1) Local adjustment.

[0061] according to Figure 2 The recursive formula for the B-spline basis function shown can be derived from this. In the interval The value is not zero, therefore, when changing the control point... It will only affect the defined interval. The curve segment.

[0062] In trajectory optimization, the control points of the B-spline curve are used as optimization variables. When optimizing the control points, only the curve within the corresponding domain is changed without affecting the entire curve. This property can be used to make local adjustments to the trajectory.

[0063] (2) Convex hull property.

[0064] B-spline curve It is completely surrounded by the convex hull formed by the control points in the corresponding time domain, that is, for a 3rd-order B-spline. Controlled point The convex hull formed contains this.

[0065] (3) The derivative of a spline curve is also a spline curve.

[0066] The derivative of a B-spline curve is still a spline curve, and the curve derivative control points are... Compared with the original curve control points The following relationship exists:

[0067] (11)

[0068] in, It refers to the time span. After describing the position information using B-spline curves, the velocity spline curve and acceleration spline curve can be obtained through the derivative property of spline curves, and constraints can be applied to the velocity and acceleration.

[0069] Specifically, the screening of velocity sampling point sets and acceleration sampling point sets includes: screening the velocity sampling point sets and acceleration sampling point sets sequentially based on the convex hull property and cost function of B-spline curves.

[0070] The map is discretized using a spatial grid, with each grid storing the location of the control point with the lowest cost. Typically, to balance search efficiency and trajectory quality, the cost function of the control points is... From the cumulative cost function and heuristic cost function It consists of two parts:

[0071] (12)

[0072] in, The cumulative cost function represents the trajectory control point of the u-th UAV in the i-th time period. The design primarily determines the quality of the trajectory, and the design is as follows:

[0073] (13)

[0074] in, Indicates the starting control point of the u-th UAV. To the current i-th control point European distance, A distance cost term is introduced as the weighting coefficient. This is to ensure that the value of the cumulative cost function is always positive; Defined as:

[0075] (14)

[0076] in, The third derivative of the normalized basis vectors is given by... The third derivative of the trajectory (i.e., jerk). Indicates drone In the The matrix composed of control points of the segment trajectory Indicates time integral.

[0077] Heuristic cost function The design of the heuristic cost determines the search speed of the method, but simply increasing the heuristic cost will increasingly fail to guarantee the optimality of the trajectory or even its neighborhood. In this section, the heuristic cost function is designed without considering obstacle constraints, and directly uses the Euclidean distance cost from the current node to the target node as the heuristic cost of the method:

[0078] (15)

[0079] in, is the scaling factor of the heuristic function.

[0080] During graph search, the method of this invention expands neighbor nodes according to the recursive formula for B-spline control points. Depending on the expansion method, the neighbor set of control points can be divided into two types: velocity-based and acceleration-based. and To define these two sets, the corresponding velocity sampling sets were first defined. and acceleration sampling set :

[0081] (16)

[0082] in, This represents the velocity sampling point of the u-th drone at the m-th time node (i.e., the m-th time period). This represents the acceleration sampling point of the u-th UAV at the m-th time node. and For the sampling interval of velocity and acceleration, and Extend the node sequence for velocity and acceleration. Indicates the index of the coordinate axis. and For drones On the coordinate axes The maximum velocity and maximum acceleration on the surface, at the same time When calculating the elements, the sampling rate should be taken into account, so that the corresponding acceleration control points must meet the constraints.

[0083] Preferably, the filtering of velocity and acceleration sampling point sets based on the convex hull property of B-spline curves (i.e., the corresponding neighbor set can be recursively obtained from the elements in the sampling set) includes:

[0084] (17)

[0085] in, Let u represent the set of velocity sampling points after convex hull screening for the u-th UAV. This represents the trajectory control point corresponding to the velocity sampling point of the u-th UAV in the m-th time period. This represents the trajectory control point corresponding to the velocity sampling point of the u-th UAV in the (m-1)-th time period. This represents the velocity sampling point of the u-th drone. Indicates the time span, This represents the set of velocity sampling points for the u-th drone. This represents the convex hull of the u-th drone in the m-th time interval. This represents the collision model of the u-th drone relative to a static obstacle. Indicates obstacles convex hull, Represents a set of static obstacles. This represents the convex hull of the u-th and v-th drones relative to the B-spline curve in the m-th time interval. This is a relative collision model between the u-th drone and the v-th drone. Let u represent the set of acceleration sampling points after convex hull filtering for the u-th UAV. This represents the speed control point of the u-th drone in the m-th time interval. This represents the speed control point of the u-th drone in the (m-1)-th time interval. This represents the acceleration sampling point of the u-th UAV. Indicates the position of the u-th drone on the coordinate axis. The maximum speed on the top, Let u represent the set of acceleration sampling points for the u-th UAV. This represents the set of neighboring drones within the communication range of the u-th drone.

[0086] Based on the convex hull property of B-spline curves, the B-spline curve strictly lies within the range of the convex hull formed by the corresponding control points. Now, we define the UAV... No. The convex hull of the segment trajectory is:

[0087] (18)

[0088] in, This represents the Mincowsky sum. Indicates obstacles convex hull, It is the convex hull of the m-th segment of the drone's trajectory. This represents the convex hull of the trajectory of UAVs u and v in the m-th segment relative to the B-spline curve; It is a collision model of the drone u relative to a static obstacle. This is a relative collision model between drones u and v; o is a single static obstacle. It is a collection of static obstacles. This indicates that drone v is within the communication range of drone u.

[0089] The neighbor set calculation also requires constraints on the new control point and its predecessor. The safety of the trajectory is ensured by the safety of the convex hull composed of control points. In addition, for the neighbor set using acceleration sampling, it should be ensured that the velocity control points derived therefrom are within the corresponding constraint range.

[0090] Furthermore, since the planning process for a drone is typically not instantaneous but always involves a time period ΔT, under an asynchronous communication architecture, it cannot be guaranteed that the drone will always receive updated trajectories from other drones in its neighborhood at the start of its planning phase. In particular, when a drone receives new trajectories from other drones during the planning process that conflict with its own planning results, the safety of its own trajectory cannot be guaranteed. Therefore, this invention designs a cooperative optimization-checking method to resolve the aforementioned conflicts and introduces a priority avoidance rule to address potential deadlock phenomena in confined spaces.

[0091] For ease of description, "this machine" refers to the currently planned drone "u", and "other drones" refers to other drones within the neighboring region of this machine. .

[0092] To introduce conflict mitigation strategies, the local planning process can be divided into two phases: optimization and checking. Within each planning cycle, the time allocation is as follows: Figure 2 As shown.

[0093] The optimization phase is the main stage for executing the trajectory planning method. At the beginning of the optimization phase, the local machine first stores the trajectories published by other drones. Then, under the constraints of these time-parameterized trajectories and its own perceived obstacles, it generates a collision-free trajectory based on the trajectory optimization problem described above. During this process, the local machine also needs to store the trajectories updated by other drones, storing the drone number and its trajectory in the list NewTraj_1, which is then used as input for subsequent algorithm checks. Finally, at the end of the optimization phase, the local machine needs to publish the optimization result to other drones in the neighborhood; that is, when generating the trajectory information for the next time period, the trajectory information is sent to neighboring drones, but the result is not directly executed.

[0094] The purpose of the inspection phase is to check whether other drones have published new trajectories during the local optimization phase and to mitigate their conflicts. To ensure the effectiveness of the inspection process, this invention designs a dual inspection mechanism, and the local machine still stores the trajectories updated by other drones during the inspection process. In the first inspection, the local machine first checks whether NewTraj_1 is empty. If it is empty, it skips the remaining steps and starts executing the current planning result; otherwise, it performs collision detection and calculates the priority of drones that are colliding with each other. At this point, drone u has a lower priority than drone v if and only if either of the following two conditions is met: :

[0095] (1) The distance between UAV u and the target point is less than the distance between UAV v and the target point;

[0096] (2) The drone v has passed the inspection phase and has begun to execute its next time period planning results (i.e. trajectory information).

[0097] When there are other drones with higher priority than this one, this drone needs to abandon the current planning result and start the next planning cycle to replan the trajectory, that is, regenerate the trajectory information of the drone for the next time period based on the B-spline curve.

[0098] Conversely, if no other drone has a higher priority than this one, the check result is true, and this one begins to execute the optimized result, that is, after the first collision detection, the trajectory information is executed when there is no collision.

[0099] Furthermore, during the initial check, it's necessary to store the updated trajectories of other drones and store their serial numbers and trajectories in the list NewTraj_2 as input for the secondary check. During the secondary check, if NewTraj_2 is empty, the process is skipped; otherwise, the stored drone serial numbers are processed according to the procedure for conflicting trajectories observed in the initial check. Priority is calculated directly, and the optimization result is abandoned based on whether a drone with a higher priority exists. Since the initial drone check process is much shorter than the optimization process, it typically doesn't enter the secondary check process. Even if it does, this process involves only a few Boolean operations, and the time interval is negligible. Therefore, this dual-check mechanism is sufficient to mitigate conflicts.

[0100] Furthermore, it is worth noting that the method proposed in this invention can guarantee the safety of UAVs in a distributed and asynchronous communication architecture under the following conditions: When the local drone fails during the optimization or checking phase, it will continue to maintain the trajectory planned at the previous moment. This guarantees the safety of the local drone because the trajectory successfully planned at the previous moment is always safe, and other UAVs will also consider the local drone's trajectory at the previous moment when updating their own trajectories. Moreover, during local planning, the planning boundary of the local drone is always kept in a hovering state. In this way, when the local drone continuously fails to obtain a new solution, it will always hover at the end of the trajectory at the previous moment to ensure that other UAVs can effectively avoid it.

[0101] Communication radius and planning radius Must meet This prevents drones outside the communication radius from entering the planning radius within a planning cycle, thus avoiding the potential threat posed by other drones entering the communication range during the planning process. In reality, this is easily guaranteed because the communication radius is usually much larger than the planning radius.

[0102] In this embodiment of the invention, deadlock is a common problem in distributed cluster trajectory planning methods, especially in confined spaces. It is understood that currently there is no proven method to resolve deadlock under these conditions; most methods employ heuristics for mitigation. Similarly, this invention designs a heuristic method based on a priority avoidance rule, comprising two steps: deadlock detection and deadlock reduction.

[0103] Deadlock detection is based on the state of the local drone and other drones at the time of deadlock. Deadlock occurs when neither the local drone nor other drones continue flying and the local drone has not yet reached the target point. Deadlock mitigation is achieved by adding new target points to low-priority drones. When deadlock occurs, low-priority drones iteratively insert new target points along the normal of the line connecting the highest-priority drone in their neighborhood to the target point, until the highest-priority drone begins to move.

[0104] The present invention also discloses a UAV trajectory planning device based on conflict reduction, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-described method.

[0105] The present invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0106] The present invention also provides a computer program product that, when run on a data storage device, enables the data storage device to implement the steps in the above-described method embodiments.

[0107] If the integrated unit module is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above embodiments of the present invention can be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include at least: any entity or device capable of carrying computer program code to a storage device, a recording medium, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium. Examples include USB flash drives, portable hard drives, magnetic disks, or optical disks. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electrical carrier signals or telecommunication signals.

[0108] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0109] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed in this invention can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0110] In the embodiments provided by this invention, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of modules or units is merely a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed.

[0111] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

Claims

1. A method for UAV trajectory planning based on conflict reduction, characterized in that, Includes the following steps: Acquire the initial trajectory information and obstacle information of the UAV, generate the trajectory information of the UAV in the next time period based on the B-spline curve, and simultaneously receive the trajectory information of neighboring UAVs and store it in the first trajectory information list. The trajectory information is subjected to a first collision detection based on the neighboring UAV trajectory information; when there is a collision between the neighboring UAV trajectory information in the first trajectory information list and the current UAV trajectory information, if the priority of the neighboring UAV in the first trajectory information list is lower than the priority of the current UAV, the trajectory information is subjected to a second collision detection with the neighboring UAV trajectory information in the second trajectory information list; wherein, the second trajectory information list is the neighboring UAV trajectory information received during the first collision detection process; When there is a collision between the trajectory information of a neighboring UAV in the second trajectory information list and the trajectory information of the current UAV, if the priority of the neighboring UAV in the second trajectory information list is lower than the priority of the current UAV, the trajectory information is executed. The trajectory information of the UAV for the next time period generated based on B-spline curves includes: Generate the velocity sampling point set and acceleration sampling point set of the UAV for the next time period based on the initial trajectory information and the obstacle information; By filtering the set of velocity sampling points and the set of acceleration sampling points, the optimal velocity sampling points and the optimal acceleration sampling points are obtained. The trajectory control points for the next time period are generated based on the optimal velocity sampling points and the optimal acceleration sampling points. Generate the drone's trajectory for the next time period based on the trajectory control points; Filtering the velocity sampling point set and acceleration sampling point set includes: Based on the convex hull property and cost function of B-spline curves, the velocity sampling point set and acceleration sampling point set are sequentially filtered, including: ， in, Let u represent the set of velocity sampling points after convex hull screening for the u-th UAV. This represents the trajectory control point corresponding to the velocity sampling point of the u-th UAV in the m-th time period. This represents the trajectory control point corresponding to the velocity sampling point of the u-th UAV in the (m-1)-th time period. This represents the velocity sampling point of the u-th drone. Indicates the time span, This represents the set of velocity sampling points for the u-th drone. This represents the convex hull of the u-th drone in the m-th time interval. This represents the collision model of the u-th drone relative to a static obstacle. Indicates obstacles convex hull, Represents a set of static obstacles. This represents the convex hull of the u-th and v-th drones relative to the B-spline curve in the m-th time interval. This is a relative collision model between the u-th drone and the v-th drone. Let u represent the set of acceleration sampling points after convex hull filtering for the u-th UAV. This represents the speed control point of the u-th drone in the m-th time interval. This represents the speed control point of the u-th drone in the (m-1)-th time interval. This represents the acceleration sampling point of the u-th UAV. Indicates the position of the u-th drone on the coordinate axis. The maximum speed on the top, Let u represent the set of acceleration sampling points for the u-th UAV. This represents the set of neighboring drones within the communication range of the u-th drone; The cost function is: ， in, This represents the cumulative pricing function. This represents the trajectory control point of the u-th UAV in the i-th time period of the cumulative pricing function. Represents the heuristic cost function; The neighboring drone has a higher priority than the current drone when any of the following conditions are met; The distance between neighboring drones and the target point is less than the distance between the current drone and the target point; The neighboring drone has already executed the trajectory information for the next time period.

2. The UAV trajectory planning method based on conflict reduction as described in claim 1, characterized in that, When generating trajectory information for the next time period of the drone, the trajectory information is sent to neighboring drones.

3. The UAV trajectory planning method based on conflict reduction as described in claim 2, characterized in that, After the first collision detection, if no collision occurs, the trajectory information is executed.

4. The UAV trajectory planning method based on conflict reduction as described in claim 3, characterized in that, If the priority of the drone in the neighborhood of the first trajectory information list is greater than or equal to the priority of the current drone, the trajectory information of the drone for the next time period is regenerated based on the B-spline curve.

5. A UAV trajectory planning device based on conflict reduction, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method as described in any one of claims 1-4.

Images