A target-oriented obstacle avoidance method for drone swarm based on Vicek model
By establishing a target-oriented UAV cluster Vicsek model in a two-dimensional plane and fitting a three-dimensional surface, the problem of high model complexity and high computational cost when the UAV cluster avoids obstacles in three-dimensional space is solved, and low error is achieved. Highly generalized obstacle avoidance effect.
Patent Information
- Application Number
- CN202411104907.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-13
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2044-08-13
AI Technical Summary
The existing drone cluster obstacle avoidance method has complex models, high computational cost, poor generalization in three-dimensional space, and the interpretability and reliability of the intelligent control system are insufficient.
Establish a target-oriented UAV cluster Vicsek model in a two-dimensional plane, fit a three-dimensional surface by collecting environmental sampling points, and obtaining the three-dimensional coordinates of the drone individual using the interpolation method to form a target-oriented UAV cluster Vicsek Three-dimensional obstacle avoidance model to perform obstacle avoidance in drone clusters.
It realizes low error, low computing cost and high generalization of obstacle avoidance of drone clusters, and can effectively control the drone cluster in three-dimensional space without relying on adaptive communication and optimal control.
Smart Images

Figure CN118625855B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the field of unmanned aerial vehicles, and in particular to a target-oriented unmanned aerial vehicle cluster obstacle avoidance method based on a Vicseek model. Background Art
[0002] At present, the obstacle avoidance methods of drone swarms mainly refer to the obstacle avoidance behavior of animal swarms, and most of the obstacle avoidance methods are still limited to the two-dimensional plane. Among them, the Vicsek model is widely used in actual production due to its advantages such as simple model, wide applicability, and simple calculation, but it is no longer applicable in three-dimensional space. When considering three-dimensional space, most swarm obstacle avoidance methods often set many complex rules, which have the disadvantages of complex models, high calculation costs, and poor generalization. The remaining swarm obstacle avoidance methods use intelligent control systems, which have poor interpretability and insufficient reliability. Summary of the invention
[0003] The purpose of the present invention is to overcome the shortcomings of the prior art and provide a target-oriented obstacle avoidance method for drone clusters based on the VicSek model, comprising the following steps:
[0004] Step 1: Establish a two-dimensional obstacle avoidance model of the drone cluster VICSEK with target orientation in a two-dimensional plane;
[0005] Step 2: When the drone cluster is detecting in the air, it collects several environmental sampling points outside the safe distance of the surrounding environment, and fits the three-dimensional surface of the environment based on the collected environmental sampling points, the location of the drone cluster and the location of the target point;
[0006] Step 3: Use the interpolation method to get the coordinates: Individual drones exist Moment Axis coordinates, then the drone individual exist Three-dimensional coordinates of time , according to the obtained three-dimensional coordinates of each UAV, a target-oriented UAV cluster VICSEK three-dimensional obstacle avoidance model is obtained;
[0007] Step 4: Obstacle avoidance of the UAV cluster is performed based on the obtained target-oriented three-dimensional obstacle avoidance model of the UAV cluster VICSEK.
[0008] Furthermore, the establishment of a target-oriented UAV cluster VICSEK two-dimensional obstacle avoidance model in a two-dimensional plane includes:
[0009] Features drone individuals, in Flying in the plane, the target point is ; Drone individual exist The direction of motion at the moment is , the coordinate position is , the velocity vector is:
[0010]
[0011] The individual drones have the same velocity at the initial moment. , randomly generated movement direction , randomly generate initial positions within a set range of non-target points:
[0012]
[0013] in, They are the left boundary horizontal coordinate, lower boundary vertical coordinate, right boundary horizontal coordinate, and upper boundary vertical coordinate of the set range respectively. is the random number corresponding to the individual coordinates of the i-th drone;
[0014] According to the drone individuals obtained exist The position, speed and direction of movement at any given moment can be sent to the individual drone. exist Position, speed and direction of movement at the moment:
[0015] set up
[0016]
[0017]
[0018] The movement direction of each drone is updated as follows:
[0019]
[0020] in is a uniformly distributed noise signal, ;
[0021] Update the formula according to the speed:
[0022]
[0023] Let the time interval be , then each individual in The position at this moment is updated as:
[0024] .
[0025] Furthermore, when the drone cluster is detecting in the air, several environmental sampling points outside the safe distance of the surrounding environment are collected, and a three-dimensional surface of the environment is fitted according to the collected environmental sampling points, the position of the drone cluster and the position of the target point, including:
[0026] When the drone cluster is conducting detection in the air, it collects several environmental sampling points outside a safe distance of the surrounding environment, and fits the three-dimensional surface of the environment through analytical or numerical methods based on the collected environmental sampling points, the location of the drone cluster and the location of the target point. The analytical method includes quadratic surface fitting and cubic surface fitting; the numerical method includes a NURBS surface generation method.
[0027] Furthermore, the coordinates obtained by using the interpolation method are Individual drones exist Moment Axis coordinates, then the drone individual exist Three-dimensional coordinates of time According to the obtained three-dimensional coordinates of each UAV, the target-oriented UAV cluster VICSEK three-dimensional obstacle avoidance model is obtained, including:
[0028] According to the projection surface of the three-dimensional model, the coordinates are obtained by substitution method: Individual drones exist Moment Axis coordinates, then the drone individual exist Three-dimensional coordinates of time According to the UAV paths formed by the three-dimensional coordinates of each UAV at any time, a target-oriented UAV cluster VICSEK three-dimensional obstacle avoidance model is obtained.
[0029] The beneficial effects of the present invention are as follows: the present method does not need to realize the obstacle avoidance of drone clusters through adaptive communication, optimal control and other technologies, but only needs to perform a small amount of numerical calculations, which not only leads to small errors, but also requires a small amount of data, low computational cost and strong generalization. At the same time, this projection method can also be combined with other two-dimensional cluster control methods to derive a cluster control model in three-dimensional space. BRIEF DESCRIPTION OF THE DRAWINGS
[0030] Figure 1 It is a flowchart of a target-oriented obstacle avoidance method for drone swarms based on the Vicek model;
[0031] Figure 2 The figure is a schematic diagram of the obstacle avoidance simulation of a target-oriented drone swarm based on the Vicek model. DETAILED DESCRIPTION
[0032] The technical solution of the present invention is further described in detail below in conjunction with the accompanying drawings, but the protection scope of the present invention is not limited to the following.
[0033] The features and performance of the present invention are further described in detail below in conjunction with the embodiments.
[0034] like Figure 1 As shown, a target-oriented obstacle avoidance method for drone clusters based on the Vicsek model includes the following steps:
[0035] Step 1: Establish a two-dimensional obstacle avoidance model of the drone cluster VICSEK with target orientation in a two-dimensional plane;
[0036] Step 2: When the drone cluster is detecting in the air, it collects several environmental sampling points outside the safe distance of the surrounding environment, and fits the three-dimensional surface of the environment based on the collected environmental sampling points, the location of the drone cluster and the location of the target point;
[0037] Step 3: Use the interpolation method to get the coordinates: Individual drones exist Moment Axis coordinates, then the drone individual exist Three-dimensional coordinates of time , according to the obtained three-dimensional coordinates of each UAV, a target-oriented UAV cluster VICSEK three-dimensional obstacle avoidance model is obtained;
[0038] Step 4: Obstacle avoidance of the UAV cluster is performed based on the obtained target-oriented three-dimensional obstacle avoidance model of the UAV cluster VICSEK.
[0039] The method of establishing a target-oriented UAV cluster VICSEK two-dimensional obstacle avoidance model in a two-dimensional plane includes:
[0040] Features drone individuals, in Flying in the plane, the target point is ; Drone individual exist The direction of motion at the moment is , the coordinate position is , the velocity vector is:
[0041]
[0042] The individual drones have the same velocity at the initial moment. , randomly generated movement direction , randomly generate initial positions within a set range of non-target points:
[0043]
[0044] in, They are the left boundary horizontal coordinate, lower boundary vertical coordinate, right boundary horizontal coordinate, and upper boundary vertical coordinate of the set range respectively. is the random number corresponding to the individual coordinates of the i-th drone;
[0045] According to the drone individuals obtained exist The position, speed and direction of movement at any given moment can be sent to the individual drone. exist Position, speed and direction of movement at the moment:
[0046] set up
[0047]
[0048]
[0049] The movement direction of each drone is updated as follows:
[0050]
[0051] in is a uniformly distributed noise signal, ;
[0052] Update the formula according to the speed:
[0053]
[0054] Let the time interval be , then each individual in The position at this moment is updated as:
[0055] .
[0056] When the drone cluster is detecting in the air, several environmental sampling points outside the safe distance of the surrounding environment are collected, and a three-dimensional surface of the environment is fitted according to the collected environmental sampling points, the position of the drone cluster and the position of the target point, including:
[0057] When the drone cluster is conducting detection in the air, it collects several environmental sampling points outside a safe distance of the surrounding environment, and fits the three-dimensional surface of the environment through analytical or numerical methods based on the collected environmental sampling points, the location of the drone cluster and the location of the target point. The analytical method includes quadratic surface fitting and cubic surface fitting; the numerical method includes a NURBS surface generation method.
[0058] The coordinates obtained by the interpolation method are Individual drones exist Moment Axis coordinates, then the drone individual exist Three-dimensional coordinates of time According to the obtained three-dimensional coordinates of each UAV, the target-oriented UAV cluster VICSEK three-dimensional obstacle avoidance model is obtained, including:
[0059] According to the projection surface of the three-dimensional model, the coordinates are obtained by substitution method: Individual drones exist Moment Axis coordinates, then the drone individual exist Three-dimensional coordinates of time According to the UAV paths formed by the three-dimensional coordinates of each UAV at any time, a target-oriented UAV cluster VICSEK three-dimensional obstacle avoidance model is obtained.
[0060] Specifically, this method first establishes a two-dimensional obstacle avoidance model of a target-oriented UAV cluster VICSEK in a two-dimensional plane, then establishes a projection surface in three-dimensional space according to the mission scenario and mission constraints, and finally uses an interpolation method to obtain a three-dimensional obstacle avoidance model of a target-oriented UAV cluster VICSEK.
[0061] Establishing a two-dimensional obstacle avoidance model for target-oriented UAV swarm VICSEK
[0062] In the basic Vicsek model, Autonomous individuals or particles form a discrete time system, and the initial positions and movement directions of the individuals are randomly distributed. The speed of each individual movement is constant at each moment, and the direction of movement is updated according to the vector average of the angular velocities of all its neighbors, and the direction update process is also disturbed by a noise signal with a mean of zero.
[0063] In the target-oriented UAV cluster VicSek model, the movement direction is updated by averaging the angle vectors between all neighbor positions and the target position within the sensing range. Assuming that drone individuals, in Flying in the plane, the target point is All drone individuals in the group move on a two-dimensional plane. The drone individuals have the same perception ability, and each drone individual can intelligently perceive the information of its neighbor individuals within its neighborhood. exist The direction of motion at the moment is , the coordinate position is , the velocity vector is:
[0064] .
[0065] Assume that all drone individuals have the same speed at the initial moment , randomly generated movement direction , randomly generate the initial position within a certain range of non-target points (assuming it is a rectangular area parallel to the coordinate axis)
[0066] ,in,
[0067] They are the left border horizontal coordinate, the lower border vertical coordinate, the right border horizontal coordinate, and the upper border vertical coordinate of the rectangular area. For the The random number corresponding to each individual coordinate.
[0068] Known individuals exist The position, speed and direction of movement at the moment are given below to find the individual exist A method for determining position, velocity, and direction of movement at any given moment.
[0069] set up
[0070]
[0071]
[0072] The movement direction of each individual is updated according to the following formula
[0073]
[0074] in is a uniformly distributed noise signal, . Substitute into the speed update formula
[0075]
[0076] In the example, let the time interval be , then each individual exist The position update at this moment is as follows:
[0077]
[0078] Establish a three-dimensional projection surface in three-dimensional space according to the mission scenario and task constraints. When the drone cluster is detecting in the air, several (usually few) environmental sampling points are collected outside the safe distance of the detected surrounding environment (that is, the distance at which the drone will not collide with the environment). According to these environmental sampling points, the location of the drone cluster and the location of the target point, the three-dimensional surface of the environment is fitted through analytical methods (quadratic surface fitting, cubic surface fitting, etc.) or numerical methods (NURBS surface generation method, etc.), and then the height coordinate of the surface is increased by a fixed value (determined by the collective scenario, as long as it is guaranteed not to intersect with the environment), and a surface that does not intersect with the environment is obtained as the projection surface of the three-dimensional model.
[0079] Using interpolation method to obtain the target-oriented UAV cluster VicSek obstacle avoidance model
[0080] According to the projection surface of the three-dimensional model, the coordinates are converted into Individual drones exist Moment The axis coordinates are calculated, and the individual drone is obtained. exist Three-dimensional coordinates of time . Then, a target-oriented UAV cluster VICSEK obstacle avoidance model was obtained.
[0081] Examples and simulation results
[0082] Assume there are 10 drone individuals. Flying in the plane, the target point is . Assume that the drone exist The direction of motion at the moment is , the coordinate position is , the velocity vector is
[0083]
[0084] Assume that all drone individuals have the same speed at the initial moment , randomly generated movement direction , within a certain range of non-target points Randomly generate initial position ,in, For the The random number corresponding to each individual coordinate.
[0085] set up
[0086]
[0087]
[0088] The movement direction of each individual is updated according to the following formula
[0089]
[0090] in is a uniformly distributed noise signal, . Substitute into the speed update formula
[0091]
[0092] In the example, let the time interval be , then each individual exist The position update at this moment is as follows:
[0093] .
[0094] The simulation results are as follows Figure 2 As shown, according to the trajectory of the drone group, it can be seen that the obstacle avoidance of the drone cluster is achieved.
[0095] The above is only a preferred embodiment of the present invention. It should be understood that the present invention is not limited to the form disclosed herein, and should not be regarded as excluding other embodiments, but can be used in various other combinations, modifications and environments, and can be modified within the scope of the concept described herein through the above teachings or the technology or knowledge of the relevant field. The changes and modifications made by those skilled in the art shall not deviate from the spirit and scope of the present invention, and shall be within the scope of protection of the claims attached to the present invention.
Claims
1. A target-oriented unmanned swarm obstacle avoidance method based on the Vicsek model, characterized in that: The steps include: Step 1: Establish a two-dimensional obstacle avoidance model of the unmanned cluster vicsek with target orientation in a two-dimensional plane; Step 2: When the drone cluster is detecting in the air, it collects several environmental sampling points outside the safe distance of the surrounding environment, and fits the three-dimensional surface of the environment based on the collected environmental sampling points, the location of the drone cluster and the location of the target point; Step 3: Use the interpolation method to get the coordinates: Individual drones exist Moment Axis coordinates, then the drone individual exist Three-dimensional coordinates of time , according to the obtained three-dimensional coordinates of each UAV, a target-oriented unmanned cluster vicsek three-dimensional obstacle avoidance model is obtained; Step 4: Perform unmanned cluster obstacle avoidance based on the obtained target-oriented unmanned cluster VICSEK three-dimensional obstacle avoidance model.
2. The method for unmanned cluster obstacle avoidance with goal orientation based on the Vicsek model according to claim 1, characterized in that: The method of establishing a target-oriented unmanned cluster vicsek two-dimensional obstacle avoidance model in a two-dimensional plane includes: Features drone individuals, in Flying in the plane, the target point is ; Drone individual exist The direction of motion at the moment is , the coordinate position is , the velocity vector is: The individual drones have the same velocity at the initial moment. , randomly generated movement direction , randomly generate initial positions within a set range of non-target points: in, They are the left boundary horizontal coordinate, lower boundary vertical coordinate, right boundary horizontal coordinate, and upper boundary vertical coordinate of the set range respectively. is the random number corresponding to the individual coordinates of the i-th drone; According to the drone individuals obtained exist The position, speed and direction of movement at any given moment can be sent to the individual drone. exist Position, speed and direction of movement at the moment: set up The movement direction of each drone is updated as follows: in is a uniformly distributed noise signal, ; Update the formula according to the speed: Let the time interval be , then each individual in The position at this moment is updated as: 。 3. The target-oriented unmanned cluster obstacle avoidance method based on the Vicsek model according to claim 2, characterized in that: When the drone cluster is detecting in the air, several environmental sampling points outside the safe distance of the surrounding environment are collected, and a three-dimensional surface of the environment is fitted according to the collected environmental sampling points, the position of the drone cluster and the position of the target point, including: When the drone cluster is conducting detection in the air, it collects several environmental sampling points outside a safe distance of the surrounding environment, and fits the three-dimensional surface of the environment through analytical or numerical methods based on the collected environmental sampling points, the location of the drone cluster and the location of the target point. The analytical method includes quadratic surface fitting and cubic surface fitting; the numerical method includes a NURBS surface generation method.
4. The target-oriented unmanned swarm obstacle avoidance method based on the Vicsek model according to claim 3, characterized in that: The coordinates obtained by the interpolation method are Individual drones exist Moment Axis coordinates, then the drone individual exist Three-dimensional coordinates of time According to the obtained three-dimensional coordinates of each UAV, the target-oriented unmanned cluster VICSEK three-dimensional obstacle avoidance model is obtained, including: According to the projection surface of the three-dimensional model, the coordinates are obtained by substitution method: Individual drones exist Moment Axis coordinates, then the drone individual exist Three-dimensional coordinates of time According to the UAV paths formed by the three-dimensional coordinates of each UAV at any time, a target-oriented unmanned cluster VICSEK three-dimensional obstacle avoidance model is obtained.
Citation Information
Patent Citations
Heat conduction analysis method and equipment for heat exchanger based on fractal Vicsek model
CN114818413A
Associating three-dimensional coordinates with two-dimensional feature points
CN114830190A
Heuristic cluster prediction control method for multi-unmanned aerial vehicle system
CN117519292A
