A method for cooperative target tracking control of cluster unmanned aerial vehicles under field of view angle constraint

Abstract

The application discloses a kind of field of view angle constraints under cluster unmanned aerial vehicle cooperative target tracking control method, and the heterogeneous role allocation algorithm based on relative position between unmanned aerial vehicles is as the basis of motion control;Considering the movement of the target for a period of time, the proportion of the current detection range of the cluster unmanned aerial vehicle to the range that the target can reach within a certain time is used to measure the target tracking effect, the heading angle controller is combined with the heterogeneous role of unmanned aerial vehicle, including the prediction of the movement of the target for a period of time, different role unmanned aerial vehicle has corresponding heading angle expected value, so that the unmanned aerial vehicle can adaptively change the tracking formation according to the target movement;In the speed controller, the deflection characteristics of the Lyapunov navigation vector field are used to generate corresponding deflection characteristic navigation vector fields for different role unmanned aerial vehicles, to guide the unmanned aerial vehicles to the fixed point around the target, which enables the cluster unmanned aerial vehicles to form a target tracking formation in order, and the safety distance between the unmanned aerial vehicles is easy to maintain.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of unmanned aerial vehicle control, and particularly relates to a cluster unmanned aerial vehicle cooperative target tracking control method under field of view angle constraint. BACKGROUND

[0002] A cluster unmanned aerial vehicle cooperative system is composed of multiple unmanned aerial vehicles to jointly complete a same combat task. Specifically, resources and information are shared between multiple unmanned aerial vehicle platforms, and advantages are complemented to maximize combat effectiveness. In the actual application of unmanned aerial vehicles, many tasks are completed with target tracking as the core, such as line patrol, forest fire prevention, and search and rescue of personnel in sea search and rescue. With the development of computer vision, some algorithms can effectively detect and track targets. However, when the target maneuvers to escape, the relative motion between the unmanned aerial vehicle and the target increases, and a single unmanned aerial vehicle may lose the tracking of the target. Cluster unmanned aerial vehicle cooperative target tracking can expand the camera detection range and reduce the probability of target loss through information interaction between unmanned aerial vehicles. Meanwhile, the cluster unmanned aerial vehicle system has better robustness and can reduce the impact of a single unmanned aerial vehicle failure on the task.

[0003] The problem of cluster unmanned aerial vehicle cooperative target tracking has existed for a long time and has been deeply researched. Early researches mainly used fixed-wing unmanned aerial vehicles to perform target tracking tasks. Since fixed-wing unmanned aerial vehicles are subject to dynamics constraints, they need to maintain a certain speed to maintain lift, and the minimum flight speed is usually faster than the target motion. Therefore, it is difficult for fixed-wing unmanned aerial vehicles to maintain the synchronization of the onboard camera tracking the target, and the unmanned aerial vehicle needs to fly in a curved trajectory around the target to maintain tracking of the target. Lawrence et al. proposed Lyapunov Vector Fields, i.e., Lyapunov navigation vector field, to generate the desired speed according to the relative position of the unmanned aerial vehicle to the target, and guide the unmanned aerial vehicle to fly on the limit cycle around the target. Based on the idea of navigation vector field, a large number of methods have emerged to solve the real-time planning problem of the trajectory in the cooperative target tracking of unmanned aerial vehicles. However, since it is difficult to realize real-time planning of the trajectory, many such methods do not consider whether the target is in the camera detection range.

[0004] With the development of multi-rotor unmanned aerial vehicle technology, the cooperative target tracking problem of multi-rotor unmanned aerial vehicle has also been widely concerned. In recent years, multi-agent swarm control technology has developed rapidly and has been applied to the field of unmanned aerial vehicles. Many control methods inspired by the hunting formation of animals in nature and the growth of cells have been proposed, such as the leader-follower method, the artificial potential field method, and the behavior-based method. The leader-follower method is simple, but if the leader fails, the task cannot continue, and the robustness cannot be guaranteed. The virtual structure method will greatly increase the communication and calculation of each individual if the number of individuals in the group increases, and it only stays in the simulation stage. The behavior-based method is complex and often lacks theoretical proof. In addition, in the research of multi-agent swarm control technology, in order to simplify the problem, the kinematics model and the perception model of unmanned aerial vehicles are often simplified, considering that unmanned aerial vehicles can perceive targets within their own perception range, which is quite different from the actual situation.

[0005] Therefore, a cooperative target tracking control method for swarm unmanned aerial vehicles under field of view angle constraints is proposed to solve the above problems. SUMMARY

[0006] The present application aims to provide a cooperative target tracking control method for swarm unmanned aerial vehicles under field of view angle constraints to solve or improve at least one of the above technical problems.

[0007] Therefore, a cooperative target tracking control method for swarm unmanned aerial vehicles under field of view angle constraints is proposed to solve the above problems.

[0008] The first aspect of the application provides a field of view angle constraint under cluster unmanned aerial vehicle cooperative target tracking control method, adopt heterogeneous role unmanned aerial vehicle controller control cluster unmanned aerial vehicle track target, the heterogeneous role unmanned aerial vehicle controller includes heading angle controller and speed controller, the method comprises the following steps: S1, the position relationship between unmanned aerial vehicles is divided into heterogeneous roles: cluster unmanned aerial vehicles cooperatively carry out target tracking task, one of the unmanned aerial vehicles first discovers the target, the unmanned aerial vehicle is named, according to the position relationship between the current role named unmanned aerial vehicle and its adjacent unmanned aerial vehicle, until all unmanned aerial vehicles are named; S2, the establishment of cluster unmanned aerial vehicle cooperative target tracking effect measurement index: according to the target area that can be reached within a certain time and the proportion of the current detection range of cluster unmanned aerial vehicle, the tracking effect is measured; S3, the establishment of space-time coordination of heterogeneous role unmanned aerial vehicle controller: the heading angle controller changes the cluster unmanned aerial vehicle target tracking formation according to the target motion condition, expands the camera detection area of cluster unmanned aerial vehicle, optimizes the tracking effect measurement index in S2; the speed controller guides the unmanned aerial vehicle to the fixed point around the target on the navigation vector field, so that the cluster unmanned aerial vehicle can form the tracking formation of the target; S4, output the space-time coordination of cluster unmanned aerial vehicle cooperative target tracking control result: according to the output result of S1-S3, realize the space-time coordination of cluster unmanned aerial vehicle cooperative target tracking.

[0009] The application provides a field of view angle constraint under cluster unmanned aerial vehicle cooperative target tracking control method, according to the position relationship between unmanned aerial vehicles, the heterogeneous role is distributed, combined with the heterogeneous role controller of the application, the cluster unmanned aerial vehicle can orderly and effectively complete the cluster unmanned aerial vehicle target tracking task, fully play the role of different roles, respectively consider the actual situation between each unmanned aerial vehicle and the target in actual target tracking, and respectively control, effectively guarantee the formation integrity in continuous tracking and the unmanned aerial vehicles will not interfere with each other and collide;

[0010] The target motion possible to reach the area in a period of time is considered, the coverage ratio of the current detection range of unmanned aerial vehicle to the target possible to reach the area in a certain time is taken as an index, which can effectively reflect the cluster unmanned aerial vehicle cooperative target tracking effect, avoid the problem that when the target escapes, the relative motion between unmanned aerial vehicle and target increases, and a single unmanned aerial vehicle may lose the tracking of the target;

[0011] The spatiotemporal coordinated heterogeneous role unmanned aerial vehicle controller, wherein the heading angle controller is combined with the unmanned aerial vehicle heterogeneous role, the heading angle controller contains the prediction of the motion of the target in a period of time, different role unmanned aerial vehicles have corresponding heading angle expected values, so that the unmanned aerial vehicles can adaptively change the tracking formation according to the target motion, optimize the cooperative target tracking effect of the swarm unmanned aerial vehicles, and the cooperative target tracking effect is dominant when the target maneuvers; the speed controller first generates a corresponding biased characteristic navigation vector field according to the unmanned aerial vehicle heterogeneous role, and simultaneously guides the unmanned aerial vehicles to the fixed point around the target on the navigation vector field, so that the swarm unmanned aerial vehicles can form a target tracking formation in an orderly manner, and the safety distance between the unmanned aerial vehicles is easy to maintain.

[0012] In addition, the technical scheme provided by the embodiment of the application can also have the following additional technical features:

[0013] In any of the above technical schemes, the number of the unmanned aerial vehicles is six, and the step S1 specifically comprises: S101: a certain unmanned aerial vehicle first discovers the target, the role of the unmanned aerial vehicle is named as alpha (α), and the information is transmitted to all unmanned aerial vehicles within a communication range; other five unmanned aerial vehicles receive the information and transmit their position vectors to the unmanned aerial vehicle α, the unmanned aerial vehicle α calculates the positions of the five unmanned aerial vehicles relative to itself, takes the direction of the unmanned aerial vehicle α towards the target as a positive direction, assigns the leftmost one of the remaining five unmanned aerial vehicles as beta (β) l , assigns the rightmost one of the remaining four unmanned aerial vehicles as beta (β) r , and transmits to all unmanned aerial vehicles within a communication range; S102: the unmanned aerial vehicle β l assigns the leftmost one of the remaining three unmanned aerial vehicles without a role as gamma (γ) l , and transmits to all unmanned aerial vehicles within a communication range; S103: the unmanned aerial vehicle β r assigns the rightmost one of the remaining two unmanned aerial vehicles without a role as gamma (γ) r , and transmits to all unmanned aerial vehicles within a communication range; S104: the unmanned aerial vehicle β r assigns the unmanned aerial vehicle without a role as gamma (γ) m , and transmits to all unmanned aerial vehicles within a communication range.

[0014] In the technical scheme, at the initial moment of the swarm unmanned aerial vehicle cooperative target tracking task, a certain unmanned aerial vehicle first discovers the target, so the unmanned aerial vehicle has more information than other unmanned aerial vehicles, and the closer the other unmanned aerial vehicles are to the unmanned aerial vehicle, the shorter the time is for the other unmanned aerial vehicles to discover the target and perform the cooperative target tracking.

[0015] The heterogeneous role assignment result of the UAVs will guide the cluster UAVs to form a target tracking formation in which one UAV surrounds the target; the subscripts l and r of the role respectively represent that the expected position of the UAV is located on the left side and the right side of the UAV in the head direction of the UAV; the subscript m represents that the expected position of the UAV is located between the left and right UAVs; and the expected target tracking formation of the cluster UAVs with heterogeneous roles is that the UAV that first discovers the target keeps a relatively close distance from the target, and the remaining UAVs surround the target to enclose the target in the middle and track the movement of the target.

[0016] In any of the technical solutions described above, the target tracking formation of the UAVs is that the UAV that first discovers the target keeps a relatively close distance from the target, and the remaining five UAVs surround the target.

[0017] The position of the target is calculated by using a camera and other sensors, and the six UAVs are controlled to track the target by using a navigation vector field combined with heterogeneous roles; in the tracking process, the UAVs converge to the expected positions to form a tracking formation in which one UAV keeps a relatively close distance from the target, and the remaining UAVs surround the target.

[0018] In any of the technical solutions described above, the step S2 specifically comprises the following steps: S201, calculating an escape region of the target within a time t; and S202, calculating a real detection range of the camera of the cluster UAVs within the time t, and obtaining a tracking effect measurement index according to the calculated escape region.

[0019] In this technical solution, the cluster UAVs cooperatively track the target, and the tracking effect measurement index is given by the proportion of the current detection range of the cluster UAVs to the possible escape region of the target within a certain time, which can represent the probability that the target can escape from the detection range of the cluster UAVs after a certain time.

[0020] In any of the technical solutions described above, the step S201 specifically comprises the following steps: S2011, simplifying the movement of the target within the time t into two cases: wherein a is the acceleration of the target, ω is the angular velocity of the target, the subscript max represents the maximum value, case ① represents that the acceleration is any normal number within the acceleration range and the angular velocity is any constant within the angular velocity range, and case ② represents that the acceleration is any negative constant within the acceleration range and the angular velocity is any constant within the angular velocity range; S2012, after the time t, the position reached by the target is as follows: wherein v0 is the linear velocity of the target, θ0 is the yaw angle of the target, t0 is the initial time, t is the prediction time, and p(t0) = [x0, y0] T is the initial position of the target, and the union set of all possible positions reached by the target is the escape region S(t) of the target. e .

[0021] In the technical scheme, the motion of the target in a certain period of time is classified according to different acceleration directions of the target, and then the position of the target after the certain period of time is analyzed and calculated, so that the actual motion of the target and the actual detection area of the unmanned aerial vehicle camera are considered, and the actual situation is closer.

[0022] Case 1 indicates that the acceleration a of the target is a constant located in [0, a max ], and the angular velocity ω of the target is a constant located in [-ω max , ω max ], and case 2 is the same.

[0023] In any of the above technical solutions, the step of S202 specifically includes: calculating the cluster unmanned aerial vehicle cooperative target tracking effect measurement index as the following formula: Wherein, S(t) e is the escape area of the target in time t, S(i) is the actual detection range of the camera of the unmanned aerial vehicle i at the previous moment, and Area(S) is the area of a two-dimensional closed region S.

[0024] In the technical scheme, according to the above formula, the proportion of the union range of the possible escape area of the target in time t and the actual detection range of all unmanned aerial vehicles to the possible escape area of the target in time t is taken as the cluster unmanned aerial vehicle cooperative target tracking effect measurement index. Specifically, the actual detection range is related to the camera parameters of the unmanned aerial vehicle, and can be simplified as a fan-shaped region or a quadrilateral region.

[0025] In any of the above technical solutions, the step of S3 specifically includes: S301, the heading angle controller is combined with the heterogeneous roles of the unmanned aerial vehicle, the target tracking formation of the cluster unmanned aerial vehicle is adaptively changed according to the target motion, the detection area of the camera of the cluster unmanned aerial vehicle is expanded, and when the target is maneuvered, the cluster unmanned aerial vehicle cooperative target tracking effect measurement index established in S2 is optimized; S302, the speed controller is combined with the heterogeneous roles of the unmanned aerial vehicle, the Lyapunov navigation vector field is used to guide the unmanned aerial vehicle to a point on the limit cycle, and the navigation vector field with corresponding deflection characteristics is generated for different role unmanned aerial vehicles in combination with the heterogeneous roles of the unmanned aerial vehicle.

[0026] The unmanned aerial vehicle is guided to a point on the limit cycle of the Lyapunov navigation vector field by the speed controller, instead of guiding the unmanned aerial vehicle to fly around the target on the limit cycle, which can avoid the confusion of the tracking array when moving, and can also reduce the tracking energy consumption of the unmanned aerial vehicle.

[0027] In any of the above technical solutions, the heading angle controller is specifically the following formula: wherein, is the error of the UAV relative target heading angle; v e is the velocity error; v d is the desired velocity of the UAV given by the navigation vector field; v is the velocity of the UAV; UAV heading angle control output; is the UAV velocity control output; a table (i) the calculation formula is P angle and D angle are PID controller parameters; is the angle of the deviation of the target position from the center line of the field of view of the UAV; is the angle of the deviation of the position reached after the target time t from the center line of the field of view of the UAV; and are the differentials of the heading angle error and the velocity error, respectively.

[0028] In the technical solution, the heading angle controller is combined with the heterogeneous role of the UAV. For the UAV of the a role, the UAV heading angle error is: φ e = 0 - φ, and the controller makes the target located in the center of the camera field of view; for the UAV of other roles, the UAV heading angle error is: φ e = 0 - φ P (t), wherein φ P (t) represents the angle of the deviation of the position reached after the target time t from the center line of the field of view of the UAV, and contains the prediction of the target motion in the future period of time; the heading angle controller adaptively changes the target tracking formation of the swarm UAV according to the target motion, expands the camera detection area of the swarm UAV, and can optimize the effect measurement index of the swarm UAV cooperative target tracking established in S2 when the target maneuvers.

[0029] In any of the above technical solutions, the step of S302 specifically comprises: S3021, establishing a velocity controller, specifically the following formula: wherein, r p is the vector of the UAV position pointing to the desired position; r d is the desired distance between the UAV and the target; is the normalized desired velocity generated by the Lyapunov navigation vector field, and the calculation formula is is the UAV position vector, which can be decomposed into the flight plane and the vertical plane of the flight plane the Lyapunov function is wherein p is the radius of the UAV around the target, is the gradient of the Lyapunov function, is a normalization factor, and I is a unit vector, The formula for calculating γ(t) is γ(t) = -k·sign(d), where k is the deflection coefficient of the navigation vector field; the formula for calculating sign(d) is... The formula for calculating μ is: The deflection term is calculated as follows: S3022, Calculate the controller output: based on the desired velocity v given by the speed controller. d The control output speed is calculated using the following formula. in, The controller outputs speed; The error of the UAV's heading angle relative to the target; v e For speed error; For UAV heading angle control output; P velocity With D velocity For PID controller parameters; and The derivative of the heading angle error and the velocity error; v is the UAV velocity; the control output given by the speed controller. Decomposed into components along the x and y axes in the UAV's own coordinate system: in, For the control components of the x-axis of the UAV's own coordinate system; For the y-axis control component; Let be the heading angle of the UAV; under the control of the controller, the speed of the UAV will converge to the desired speed given by the navigation vector field, and the position of the UAV will converge to a fixed point around the target on the navigation vector field.

[0030] In this technical solution, Lyapunov navigation vector fields with different deflection characteristics are generated for UAVs with different roles. Specifically, the UAV navigation vector field located to the left of α has a right-hand rotation characteristic, and vice versa, it has a left-hand rotation characteristic, which makes it easy to maintain a safe distance between UAVs.

[0031] The expected speed of a drone is proportional to the deviation between the drone's position and the target's position. For drones of role α, the expected distance is shorter. For drones of other roles, the expected distance between the drone and the target is the same. The principle for selecting the expected distance between the drone and the target is to make the target located in the center of the drone's camera detection range. Drones with subscripts l and r have different deflection characteristics.

[0032] The beneficial effects of this invention compared to the prior art are as follows:

[0033] By assigning heterogeneous roles according to the position relationship between unmanned aerial vehicles when a target is first discovered by a certain unmanned aerial vehicle, and combining the heterogeneous role controller, the cluster unmanned aerial vehicle can orderly and effectively complete the cluster unmanned aerial vehicle target tracking task, and fully play the role of different roles.

[0034] The cluster unmanned aerial vehicle cooperative target tracking effect measurement index provided by the application considers the area that the target can reach within a period of time, and takes the coverage ratio of the current detection range of the unmanned aerial vehicle to the area that the target can reach within a period of time as an index, which can effectively reflect the cluster unmanned aerial vehicle cooperative target tracking effect.

[0035] The spatiotemporal coordinated heterogeneous role unmanned aerial vehicle controller; the heading angle controller is combined with the unmanned aerial vehicle heterogeneous role, and includes the prediction of the movement of the target within a period of time, the different role unmanned aerial vehicles have corresponding heading angle expected values, so that the unmanned aerial vehicles can adaptively change the tracking formation according to the target movement, optimize the cluster unmanned aerial vehicle cooperative target tracking effect, and make the cooperative target tracking effect dominant when the target maneuvers; the speed controller first generates a corresponding biased characteristic navigation vector field according to the unmanned aerial vehicle heterogeneous role, and simultaneously guides the unmanned aerial vehicles to the fixed point around the target on the navigation vector field, so that the cluster unmanned aerial vehicles can orderly form a target tracking formation, and the safety distance between the unmanned aerial vehicles is easy to maintain.

[0036] Additional aspects and advantages of embodiments according to the present application will become apparent from the following description with reference to the drawings. BRIEF DESCRIPTION OF DRAWINGS

[0037] The accompanying drawings are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the principles of the application.

[0038] Figure 1 The role assignment result of step S1 of the application and the expected tracking formation schematic diagram;

[0039] Figure 2 The real imaging area schematic diagram of the unmanned aerial vehicle of the application under the field of view angle constraint;

[0040] Figure 3 The range that the target can reach within a period of time t in step S201 of the application schematic diagram;

[0041] Figure 4 The motion controller control block diagram schematic diagram in step S3 of the application;

[0042] Figure 5 The heading angle optimization schematic diagram of the application;

[0043] Figure 6 The Lyapunov navigation vector field schematic diagram of the application;

[0044] Figure 7 Fig. 1 is a schematic diagram of a navigation vector field for different roles of the present application;

[0045] Figure 8 Fig. 2 is a schematic diagram of a navigation vector field for different roles of the present application;

[0046] Figure 9 Fig. 3 is a schematic diagram of a navigation vector field for different roles of the present application;

[0047] Figure 10 Fig. 4 is a schematic diagram of a camera imaging modeling of a UAV of the present application;

[0048] Figure 11 Fig. 5 is a UAV and target trajectory in a simulation example of the present application;

[0049] Figure 12 Fig. 6 is a UAV target tracking formation at each time in a simulation example of the present application;

[0050] Figure 13 Fig. 7 is a comparison of cooperative target tracking effects with and without heading angle optimization in a simulation example of the present application. DETAILED DESCRIPTION

[0051] In order to more clearly understand the above-mentioned purposes, features and advantages of the present application, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0052] In the following description, many specific details are set forth in order to provide a thorough understanding of the present application, however, the present application can also be implemented in other ways different from those described herein, therefore, the scope of protection of the present application is not limited by the specific embodiments disclosed below.

[0053] Referring to Figures 1-13 The first aspect of the present application provides a field of view angle constrained cluster UAV cooperative target tracking control method, comprising the following steps:

[0054] S1: dividing the heterogeneous roles of the UAVs through the position relationship between the UAVs:

[0055] At the initial moment of the cluster UAV cooperative target tracking task, a UAV first discovers the target, so this UAV has more information than other UAVs; at this time, the closer the other UAVs are to this UAV, the shorter the time they can discover the target and perform cooperative target tracking; based on this, the heterogeneous role allocation algorithm includes the following steps:

[0056] S101: A certain unmanned aerial vehicle first discovers a target, its role is named as α and transmits this information to all unmanned aerial vehicles within the communication range, other 5 unmanned aerial vehicles receive this information and transmit their position vectors to the α unmanned aerial vehicle, α calculates the positions of the 5 unmanned aerial vehicles relative to itself, taking the direction of the target as the positive direction, α assigns the leftmost one of the remaining 5 unmanned aerial vehicles as β l , assigns the rightmost one of the remaining 4 unmanned aerial vehicles as β r , and transmits to all unmanned aerial vehicles within the communication range;

[0057] S102: β l assigns the leftmost one of the remaining 3 unmanned aerial vehicles as γ l , and transmits to all unmanned aerial vehicles within the communication range;

[0058] S103: β r assigns the rightmost one of the remaining 2 unmanned aerial vehicles as γ r , and transmits to all unmanned aerial vehicles within the communication range;

[0059] S104: β r assigns an unmanned aerial vehicle without a role as γ m , and transmits to all unmanned aerial vehicles within the communication range;

[0060] The heterogeneous roles of unmanned aerial vehicles are the basis for target tracking motion control, and the cluster unmanned aerial vehicles form a tracking formation according to the heterogeneous roles;

[0061] The heterogeneous role allocation result is shown in Figure 1 , the subscripts l and r of the role respectively represent that the unmanned aerial vehicle is located on the left and right sides of the α unmanned aerial vehicle in the head direction, the subscript m represents the role of the remaining one unmanned aerial vehicle, and the heterogeneous role cluster unmanned aerial vehicle expects a target tracking formation;

[0062] S2: Establish a cluster unmanned aerial vehicle cooperative target tracking effect measurement index:

[0063] The present application measures the effect of cluster unmanned aerial vehicle target tracking by the proportion of the area covered by the target within a certain time within the current detection range of the cluster unmanned aerial vehicle; considering the actual motion of the target and the actual detection area of the unmanned aerial vehicle camera, it is closer to the actual situation and includes the following steps:

[0064] S201: Calculate the target possible escape area within time t:

[0065] If the unmanned aerial vehicle is located at p(t0)=[x0,y0] T at t0, the linear velocity is v0, and the yaw angle is θ0, the motion of the target within time t is simplified into two cases, specifically formula (1):

[0066]

[0067] Case ① means that the target acceleration a is a constant located in [0, a max ] interval, the target angular velocity is a constant located in [-ω max ,ω max ] interval, case ② is the same;

[0068] Then the target reaches the position after time t as formula (2):

[0069]

[0070] Figure 3 The figure shows the possible target reaching area within a certain time under cases ① and ②: the possible target reaching area within 1.6s when the initial target speed is 10m / s, the possible target reaching area within 1.6s when the initial target speed is 0; it can be seen that the target speed has a great influence on the possible target escape area

[0071] S202: Calculate the cluster UAV cooperative target tracking effect measurement index:

[0072] The possible target escape area in S201 within time t is defined as S(t) e , where t represents the prediction time; define Where S(i) represents the real detection range of the camera of UAV i at the current time; further calculate the cluster UAV cooperative target tracking effect measurement index as formula (3):

[0073]

[0074] Where the function Area(S) represents the area of a two-dimensional closed region S; Figure 2 The figure shows the real imaging area of UAVs 1-6, and the area represented by S(t) e is shown in the figure Figure 3 , from which the cluster UAV cooperative target tracking effect can be calculated;

[0075] S3: Establish a spatiotemporal coordinated heterogeneous role UAV controller:

[0076] The connotation of spatiotemporal coordination is that the heterogeneous role UAV controller controls the cluster UAV to form a surround according to the heterogeneous role to the target, and optimizes the cluster UAV cooperative target tracking effect, while maintaining a safe distance between the cluster UAVs in space; the heterogeneous role UAV controller is divided into a heading angle controller and a speed controller, and the controller control block diagram is shown in the figure Figure 4 , which includes the following steps:

[0077] S301: Establish a heading angle controller:

[0078] Since the target may make large maneuvers, a fast response from the heading angle controller is required, so PD control is selected; combined with the heterogeneous role UAV, the heading angle controller is as shown in equation (4):

[0079]

[0080] Where α table The definition is as shown in equation (5):

[0081]

[0082] like Figure 5 As shown, the heading angle controller is combined with the heterogeneous role of the UAV. For the UAV with role α, the heading angle error is: φ e =0-φ, the controller positions the target in the center of the camera's field of view; for drones with other roles, the drone's heading angle error is: φ e =0-φ P (t), where φ P (t) represents the angle of deviation of the target's position after time t from the centerline of the UAV's field of view, which includes a prediction of the target's movement over a period of time in the future; the heading angle controller adaptively changes the target tracking formation of the swarm UAVs according to the target's movement, expands the detection area of ​​the swarm UAV cameras, and can optimize the swarm UAV cooperative target tracking performance evaluation index established in S2 when the target is maneuvering.

[0083] S302: Establish speed controller:

[0084] The speed controller, combined with the heterogeneous role of the UAV, is a typical example of Lyapunov navigation vector fields, such as... Figure 6 As shown, this invention utilizes Lyapunov navigation vector fields to guide a UAV to a point on a limit cycle, rather than guiding the UAV to fly around a target on the limit cycle; it combines the heterogeneous roles of UAVs to generate navigation vector fields with corresponding deflection characteristics for different role UAVs, including the following steps:

[0085] S3021: Heterogeneous Role Lyapunov Navigation Vector Field:

[0086] The Lyapunov navigation vector field proposed by Lawrence et al. is shown in equation (6):

[0087]

[0088] in, For the deflection term, only the deflection characteristics of the navigation vector field are changed, without changing the magnitude of the navigation vector field; as shown in equation (7), the deflection characteristics are determined by the parameter γ:

[0089]

[0090] For heterogeneous roles, the parameter μ is generated according to the index l, r, m of the heterogeneous roles of the UAV in S1:

[0091]

[0092] For different roles, the parameter γ algorithm of the Lyapunov navigation vector field is generated as formula (9):

[0093]

[0094] Thus, the Lyapunov navigation vector field with different deflection characteristics is generated for different role UAVs. Specifically, the UAV navigation vector field located on the left side of α has right rotation characteristics, and vice versa, which makes it easy to maintain the safe distance between UAVs. The length of the arrow in the navigation vector field, that is, the expected speed size, is as formula (10):

[0095]

[0096] Wherein is the normalized expected speed generated by the Lyapunov navigation vector field, r p is the vector of the UAV position pointing to the expected position, r d is the expected distance between the UAV and the target, that is, the expected speed of the UAV is proportional to the deviation between the UAV position and the target position; the navigation vector field generated for different roles is as shown in Figures 7-9 Figure 7 The left side of the figure is the navigation vector field of the α UAV, and the right side of the figure is the navigation vector field of the γ m UAV, Figure 8 The left side of the figure is the navigation vector field of the β l UAV, and the right side of the figure is the navigation vector field of the β r UAV, Figure 9 The left side of the figure is the navigation vector field of the γ l UAV, and the right side of the figure is the navigation vector field of the γ r UAV), and the asterisk in the figure indicates the position of the target, and the radius r d of the circular ring is the expected distance between the UAV and the target. For the α role UAV, the expected distance is shorter; for the UAVs of other roles, the expected distance between the UAV and the target is the same. The principle for selecting the expected distance between the UAV and the target is to make the target located in the central detection range of the UAV camera; the UAVs with index l, r have different deflection characteristics.

[0097] S3022: The calculation controller controls the output:

[0098] ​The Lyapunov navigation vector field for heterogeneous roles provides desired velocities for different roles of the UAVs at any position in space, and a controller is designed to control the aircraft velocity to track the desired velocity given by the navigation vector field as in equation (11):

[0099]

[0100] where v d is the desired velocity generated for heterogeneous roles. The control output is the control quantity in the world coordinate system, and since the heading angle of the UAV is controlled by the heading angle controller, the UAV needs to re-distribute power according to the current heading angle at each control moment. The control quantity is rotated and transformed into the body coordinate system using equation (12) to decompose the control input of the x and y axes of the UAV;

[0101]

[0102] S4: Output the spatiotemporally coordinated swarm UAV cooperative target tracking control result:

[0103] Given the kinematic model of the UAV and the target, the initial state of the UAV and the target, and the field of view angle constraint received by the camera of the UAV to determine the task scenario of the swarm UAV cooperative target tracking:

[0104] At the beginning of the swarm UAV cooperative target tracking task, i.e., at the moment when a UAV first discovers the target, the heterogeneous roles of the swarm UAVs are assigned according to the heterogeneous role assignment algorithm established in S1. At each control moment thereafter: the UAV first calculates the output of the angular velocity controller of S3-1. The angular velocity controller introduces the optimization of the heading angle of the heterogeneous role, which includes the prediction of the target motion within a period of time t, and optimizes the swarm UAV cooperative target tracking effect defined in S2. Secondly, the output of the speed controller of S3-2 is calculated. The Lyapunov navigation vector field with corresponding deflection characteristics is generated for the UAVs according to the heterogeneous roles. The speed controller controls the UAV speed to track the desired speed of the navigation vector field. Then, the spatiotemporally coordinated swarm UAV cooperative target tracking is realized, and the cooperative target tracking effect is superior when the target maneuvers.

[0105] Embodiment 1

[0106] As shown in Figures 10-13 , the method specifically comprises the following steps:

[0107] 1. Consider the kinematic model and kinematic constraints of the UAV:

[0108] Suppose that there are six UAVs performing the cooperative target tracking task, then the set of UAVs can be represented as: V = {i | i = 1, 2, …, 6}. The Ox e y eGround coordinate system; the body coordinate system is established with the UAV mass point as the origin, x b axis points to the UAV nose direction, y b axis is the x axis rotated 90° counterclockwise. The UAV state vector is represented as, and the UAV kinematics model is:

[0109]

[0110] where [x i ,y i ] T represents the position of the UAV i in the Ox e y e world coordinate system. represents the heading angle and linear velocity of the UAV i, u i = [ω i ,a i ] T is the control input, where the heading angle is 0 in the Ox e axis direction and the positive direction is counterclockwise. [ω i ,a i ] T is the angular velocity and linear acceleration, which is the control input of the UAV i state vector. The UAV is subject to maximum flight speed constraints, maximum angular velocity, and maximum acceleration constraints as shown in equation (14).

[0111]

[0112] 2. Consider the UAV camera model subject to field of view angle constraints:

[0113] Assume that the UAV on-board camera model is KS12A884, equipped with a SONY IMX377 photosensitive chip. The camera is equipped with an 8mm lens, with a monitoring field of view angle of about 40°, an optimal illumination distance of 3-12m, and a maximum detection distance of 24m. The camera is fixed to the UAV platform, and the optical axis coincides with the direction of the UAV nose. Project the camera detection area onto the Ox e y e coordinate plane to obtain the actual detection area of the camera fixed to the UAV nose x b direction, which is related to factors such as the camera field of view angle, as shown in Figure 10 . The transformation relationship from a point on the UAV camera CMOS to a point on the real imaging plane is as shown in equation (15):

[0114]

[0115] 3. Set the simulation parameters required for simulation:

[0116] Let S2 predict the UAV's motion at time t = 0.9s, and let the navigation vector field parameter k = 0.5 in S3-2-2. Let the UAV's kinematic constraints be ω∈[-50° / s, 50° / s], a∈[-5m / s]. 2 5m / s 2 Drone camera parameters l best =12m, α=45°, β=20°, sampling and control frequency is 20Hz. The target is programmed to accelerate and decelerate to escape the target. The initial states of the UAV and the target are given in Table 1:

[0117] Table 1 Initial State of the UAV and Target

[0118]

[0119] Based on the technical solution in this invention, a MATLAB-based simulation environment was built. The basic process for achieving cooperative target tracking of swarm UAVs under field-of-view constraints is shown in Table 2:

[0120] Table 2 Simulation Framework for Collaborative Target Tracking by Clustered UAVs

[0121]

[0122]

[0123] 4. Output and analyze the simulation results:

[0124] During the simulation, the trajectories of the UAV and the target are as follows: Figure 11 As shown, the formations of the UAV and target tracking at each moment are as follows: Figure 12 As shown, the cooperative target tracking performance of swarm drones with and without heading angle optimization is compared to... Figure 13 As shown.

[0125] Without heading angle optimization, after the target accelerates, the area the target may reach within a certain time increases, while the tracking formation remains unchanged, resulting in a decrease in the cooperative target tracking performance of the swarm drones. With heading angle optimization for heterogeneous roles, the cooperative target tracking performance of the swarm drones decreases the instant the target's speed increases, subsequently changing from... Figure 12 The tracking formation at 17s shows an adjustment towards 51s, increasing the adaptability of the drone camera coverage area and restoring the collaborative target tracking performance of the swarm drones to its optimal level. Figure 13 As shown, the swarm UAV cooperative target tracking control method proposed in this invention has superior tracking performance when the target is maneuvering.

[0126] In the description of the present application, it needs to be understood that the terms "longitudinal", "transverse", "upper", "lower", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer" and the like indicate the orientation or positional relationship shown in the drawings, which are only for the convenience of describing the present application, and do not indicate or imply that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation on the present application.

[0127] The above-described embodiments are only to describe the preferred modes of the present application, and not to limit the scope of the present application. Without departing from the design spirit of the present application, various modifications and improvements to the technical solutions of the present application made by those skilled in the art shall fall within the protection scope determined by the claims of the present application.

Claims

1. A method for cooperative target tracking control of swarmed UAVs under field-of-view constraints, characterized in that, A heterogeneous role-based UAV controller is used to control a cluster of UAVs to track a target. The heterogeneous role-based UAV controller includes a heading angle controller and a speed controller. The method includes the following steps: S1, the heterogeneous roles of drones are divided according to the positional relationship between drones: the drones in the swarm cooperate to carry out target tracking tasks. One of the drones first discovers the target and assigns a role to the drone. Based on the positional relationship between the drone currently assigned a role and its neighboring drones, the neighboring drones are named until all drones are assigned roles. S2, Establish a performance evaluation index for collaborative target tracking by swarm drones: measure the tracking effectiveness based on the ratio of the area that the target can reach within a certain time to the area covered by the current detection range of the swarm drones; S3, establish a spatiotemporally coordinated heterogeneous role UAV controller: the heading angle controller changes the target tracking formation of the swarm UAVs according to the target motion status, expands the camera detection area of ​​the swarm UAVs, and optimizes the tracking effect measurement indicators in S2. The speed controller guides the drones to a fixed point around the target in the navigation vector field, enabling the swarm of drones to form a target-tracking formation. S4 outputs the spatiotemporally coordinated target tracking control results of the swarm of UAVs: Based on the output results of S1-S3, spatiotemporally coordinated swarm of UAVs achieves cooperative target tracking.

2. The method for cooperative target tracking control of swarmed UAVs under field-of-view constraints according to claim 1, characterized in that, The number of drones is selected to be six. Step S1 specifically includes: S101: A drone first detects the target, is designated α, and transmits this information to all drones within communication range. The other 5 drones receive this information and transmit their own position vectors to drone α. α calculates the positions of the 5 drones relative to itself, taking the direction α is facing towards the target as the positive direction. α assigns the leftmost drone among the remaining 5 drones to be β. l The rightmost of the remaining four drones was assigned as β. r And transmit to all drones within communication range; S102:β l Assign the drone that currently has no role and is located on the far left of the remaining 3 drones to become γ. l And transmit to all drones within communication range; S103:β r Assign the drone that currently has no role and is located on the far right of the remaining two drones to become γ. r And transmit to all drones within communication range; S104:β r Assigning a drone that currently has no role to become γ m It transmits to all drones within communication range.

3. The method for cooperative target tracking control of swarm UAVs under field-of-view constraints according to claim 2, characterized in that, Step S2 specifically includes: S201: Calculate the escape zone of the target within time t; S202: Calculate the actual detection range of the cameras of the swarm drones within time t, and obtain the tracking performance metrics based on the calculated escape area.

4. The method for cooperative target tracking control of swarmed UAVs under field-of-view constraints according to claim 3, characterized in that, The steps in S201 specifically include: S2011 simplifies the motion within the target time t into two cases: Where a is the target acceleration, ω is the target angular velocity, and the subscript max represents the maximum value. Case ① indicates that the acceleration is any positive constant within the acceleration range and the angular velocity is any constant within the angular velocity range. Case ② indicates that the acceleration is any negative constant within the acceleration range and the angular velocity is any constant within the angular velocity range. S2012, after time t, the target's position is determined by the following formula: Where v0 is the target linear velocity, θ0 is the target yaw angle, t0 is the initial time, t is the prediction time, and p(t0) = [x0, y0]. T Let S be the initial position of the target. The union of all possible positions that the target can reach is the target's escape region S(t). e .

5. The method for cooperative target tracking control of swarmed UAVs under field-of-view constraints according to claim 3, characterized in that, Step S202 specifically includes: The performance metrics for collaborative target tracking by swarmed UAVs are as follows: Where, S(t) e Let S be the escape region of the target within time t, S(i) be the actual detection range of the camera of the UAV i at the previous time step, and Area(S) be the area of ​​a two-dimensional closed region S.

6. The method for cooperative target tracking control of swarm UAVs under field-of-view constraints according to claim 2, characterized in that, Step S3 specifically includes: S301 combines the heading angle controller with the heterogeneous role of the UAV to adaptively change the target tracking formation of the swarm UAVs according to the target movement, expand the detection area of ​​the swarm UAV cameras, and optimize the swarm UAV cooperative target tracking effect measurement index established in S2 when the target is maneuvering. The S302 speed controller, combined with the heterogeneous roles of the UAV, uses the Lyapunov navigation vector field to guide the UAV to a point on the extreme ring. It generates navigation vector fields with corresponding deflection characteristics for different UAV roles by combining the heterogeneous roles of the UAV.

7. The method for cooperative target tracking control of swarmed UAVs under field-of-view constraints according to claim 6, characterized in that, The heading angle controller is specifically defined by the following formula: in, The error of the UAV's heading angle relative to the target; v e For speed error; v d The desired velocity of the UAV is given by the navigation vector field; UAV heading angle control output; For drone speed control output; α table (i) The calculation formula is P angle With D angle For PID controller parameters; The angle of deviation of the target position from the centerline of the UAV's field of view; The angle representing the deviation of the target's position from the centerline of the UAV's field of view after time t. and These are the differentials of the heading angle error and the velocity error, respectively.

8. The method for cooperative target tracking control of swarmed UAVs under field-of-view constraints according to claim 6, characterized in that, The steps in S302 specifically include: S3021, Establish the speed controller, specifically using the following formula: Where, r p r is the vector pointing from the UAV's current position to the desired position. d The desired distance between the drone and the target; The normalized expected velocity generated by the Lyapunov navigation vector field is calculated as follows: The UAV's position vector can be decomposed into its position vector within the flight plane and its position vector perpendicular to the flight plane. The Lyapunov function is Where ρ is the radius of the UAV around the target. The gradient of the Lyapunov function, Let I be the normalization factor, and let I be the unit vector. The formula for calculating γ(t) is γ(t) = -k·sign(d), where k is the deflection coefficient of the navigation vector field; the formula for calculating sign(d) is... The formula for calculating μ is: The deflection term is calculated as follows: It is a unit vector perpendicular to the plane of flight; S3022, Calculate the controller output: The expected velocity v of the UAV is given by the navigation vector field. d The speed control output of the UAV is calculated using the following formula. in, Output for drone speed control; The error of the UAV's heading angle relative to the target; v e For speed error; For UAV heading angle control output; P velocity With D velocity For PID controller parameters; and denoted by , where is the derivative of the heading angle error and the velocity error; v is the UAV velocity. Output of drone speed control Decomposed into components along the x and y axes in the UAV's own coordinate system: in, For the control components of the x-axis of the UAV's own coordinate system; For the y-axis control component; Here is the heading angle of the UAV; under the control of the controller, the speed of the UAV will converge to the desired speed given by the navigation vector field, and the position of the UAV will converge to a fixed point around the target on the navigation vector field.

Images