WO2022077817A1 - Multiple unmanned aerial vehicle cooperative control method and system based on vision and performance constraints - Google Patents

Abstract

A multiple unmanned aerial vehicle cooperative control method and system based on vision and performance constraints. The method comprises: selecting an optimal pilot from an unmanned system of a sub-task, and a follower detecting, on the basis of vision, a mark carried by the pilot, so as to acquire a relative posture of the follower relative to the pilot (S2); designing an error transformation method that is based on a predetermined task performance specification (S4); and designing a PID control law of the follower according to a transformed error, so as to ensure that the follower follows the pilot according to predetermined task performance, and finally achieve the goal of autonomous cooperative control of multiple unmanned aerial vehicles (S5). By means of the method and system, multi-dimensional effective cooperation of time, space, tasks, etc., can be realized in an environment in which GPS is missing, thereby meeting the requirements of miniaturization, intelligentization and autonomization.

Description

A multi-UAV cooperative control method and system based on vision and performance constraints

This application claims the priority of the Chinese patent application submitted to the Chinese Patent Office on October 13, 2020, and the Chinese patent application is: application number 202011088200.9, and the name of the invention is "a vision and performance constraint-based multi-UAV collaboration Control Method and System", which is incorporated herein by reference in its entirety.

technical field

The invention relates to the technical field of unmanned aerial vehicles, in particular to a method and system for collaborative control of multiple unmanned aerial vehicles based on vision and performance constraints.

Background technique

With the rapid development of a new generation of information technology represented by artificial intelligence, military operations and industrial people's livelihood are changing from using machines to perform tasks to drones performing tasks autonomously, especially to solve high-dynamic, high-risk, multi-tasking tasks. Combat and operational requirements, autonomous unmanned systems have become an important support for military intelligence and industrial intelligence. Autonomous unmanned systems have significant application requirements in both military and civilian fields, including military reconnaissance, strike, information relay, terrain mapping, and intelligent warehousing. Because autonomous unmanned systems have the advantages of low cost, wide application, and good effect, It is a new technological commanding height in military operations and industrial development. Unmanned systems have always been a research hotspot at home and abroad, but most of the research focuses on the research of multi-UAV task allocation methods. There are few autonomous cooperative control schemes for multi-UAVs under the constraints of task performance, especially in GPS in a missing environment.

SUMMARY OF THE INVENTION

The present invention provides a multi-UAV cooperative control method and system based on vision and performance constraints, the purpose of which is to solve the technical problem of autonomous cooperative control of multiple UAVs in the background without GPS.

In order to achieve the above-mentioned purpose, a kind of multi-UAV cooperative control method based on vision and performance constraints provided by an embodiment of the present invention comprises the following steps:

Step S1: Decompose the overall target task to obtain mutually independent sub-tasks, confirm the type and quantity of UAVs according to the sub-tasks, and establish an unmanned system for the sub-tasks;

Step S2: select an optimal leader in the unmanned system of the current sub-task, and the follower detects the ArUco fiducial marker carried by the leader, thereby obtaining its relative pose relative to the leader;

Step S3: establishing an unmanned system model of the sub-task based on the leader-follower framework;

Step S4: designing an error transformation method based on the predetermined task performance specification;

Step S5: Design the PID control law of the follower according to the transformed error to ensure that the follower follows the leader according to the predetermined task performance, and finally achieves the goal of autonomous cooperative control of multiple UAVs.

Preferably, step S1 specifically includes the following steps:

Step S101: decompose the overall target task to obtain each subtask that is independent of each other;

Step S102: Select the form of aerial drone, ground drone, or combination of aerial drone and ground drone according to the requirements of the subtask, and determine the number of drones to establish an unmanned system for the subtask.

Preferably, step S2 specifically includes the following steps:

Step S201: in the unmanned system of the current sub-task, the control station selects an optimal UAV as the leader of the task issued by the control station in the unmanned system of the current sub-task;

Step S202: Each UAV is equipped with an ArUco square fiducial marker of known size, and the follower detects the ArUco marker using airborne vision. The ArUco marker consists of a black border and an internal binary matrix that determines its identifier. A single marker can be Provide enough correspondence (four corners) to obtain the pose of the camera relative to the ArUco marker. According to the fixed transformation relationship between the camera coordinate system and the follower coordinate system, the ArUco marker coordinate system and the leader coordinate system, obtain the leader and follower. The relative pose ζ _lf between the drones, and the ArUco mark provides a corresponding ID for each drone to ensure reliable and effective following.

Preferably, step S3 includes establishing a leader-follower model according to the leader-follower framework:

ζ _lf = ζ _l -ζ _f

Among them, ζ _lf is the pose of the leader relative to the follower, ζ _l is the pose of the leader in the world coordinate system, and ζ _f is the pose of the follower in the world coordinate system.

Preferably, the unmanned system model is composed of n-1 above-mentioned leader-follower models, where n is the number of unmanned aerial vehicles in the unmanned system.

Preferably, the S4 specifically includes the following steps:

Step S401: define the error, specifically:

Design an error transformation method with predetermined performance indicators according to the task, and estimate the relative pose between the leader and the follower through the ArUco marker, then the error is defined as

in,

Represents the desired pose of the leader relative to the follower; if the relative pose r _lf between the leader and the follower is indirectly represented by image information according to visual kinematics, then e represents the current acquired image feature and the desired image error in characteristics;

Step S402: Define error performance, specifically:

The error performance function is defined such that the output error e _k converges to a predefined residual set along the absolute decay time function ρ _k :

Among them, _ek is expressed as the kth output error amount of the error vector e, and the parameters Υ _k and

for

ρ _k (0) respectively represent the initial maximum allowable error, so as to ensure that the absolute value of the initial error satisfies 0<||e _k (0)||<ρ _k (0), the time function ρ _k (t) of the design absolute decay is

ρ _k (t)=(ρ ₀ -ρ _∞ )e ^-lt +ρ _∞

where the parameter l>0 controls the speed of exponential convergence,

Represents the steady state level of the predetermined mission performance specification, which can be designed to be small enough to guarantee the mission performance specification;

Step S403: Set the output error function, specifically:

To achieve control that meets the task performance specification, the output error is set as:

e _k =S(ε _k )ρ _k (t)

where S(ε _k ) is a monotonically increasing continuous smooth function that satisfies the following requirements:

According to the above requirements, the designed transformation function is:

Step S404: Obtain an error transformation function with a predetermined performance specification: define χ _k =e _k /ρ _k , since S(ε _k ) is strictly increasing, its inverse function always exists, so it has a predetermined performance specification. The error transformation ε _k is described as:

Preferably, step S5 specifically includes the following steps:

Step S501: Design the PID control law of the follower according to the transformed error to ensure the convergence of the transformed error _εk . The discrete form of the control law is as follows:

Among them, _uk represents the kth control variable, k _p is the proportional coefficient, _ki is the integral coefficient, and k _d is the differential coefficient;

Step S502: According to the properties of the error transformation function, when the transformed error ε _k converges, the error transformation ε _k function with the predetermined performance specification obtained in step S404 can be used to obtain the true state error e _k converges according to the predetermined performance, and more The UAV completes the desired sub-tasks, and completes each sub-task synchronously or sequentially according to the overall task goal, and finally achieves effective multi-dimensional coordination in time, space and tasks.

An embodiment of the present invention provides a vision-based autonomous cooperative control system for multiple unmanned aerial vehicles, including a control station and multiple unmanned aerial vehicles, the control station controls multiple unmanned aerial vehicles, and a plurality of unmanned aerial vehicles The aircraft includes a plurality of aerial drones and a plurality of ground drones, and the plurality of the aerial drones and the plurality of ground drones include a leader and a follower, and the control station is used to control the sub-tasks. The navigator sets offline automatic or real-time manual tasks for the navigator, the follower maintains the desired relative pose with the navigator, and follows the designated navigator to move to achieve autonomous coordinated control of multiple drones.

Preferably, the UAV includes a control unit, a sensing unit, a communication unit and a power supply unit, the communication unit is used for the UAV to receive tasks issued by the control station, and the sensing unit is an onboard camera for the follower The ArUco mark carried by the navigator is detected, the control unit is an on-board CPU, used to calculate and give the control law of the drone, and the power supply unit provides electrical energy for the drone.

Preferably, the aerial drone includes an on-board camera, and the on-board camera can be rotated appropriately according to different tasks.

The technical effect that can be achieved by the invention is that the multi-dimensional effective coordination of time, space and tasks can be realized in the environment where GPS is missing, and the requirements of miniaturization, intelligence and autonomy of the unmanned system can be met. The cooperative control of the unmanned system mainly relies on perception and control technology. The visual perception method relies on the ArUco marker detected by the airborne camera to obtain the pose of the target relative to the local coordinate system. It does not rely on GPS and can be deployed in indoor or outdoor scenes. It has the advantages of small size, low cost, and rich target information. It does not rely on GPS and can be deployed in any indoor or outdoor scene. It has the advantages of small size, low cost, and rich target information. of UAVs to build a large-scale autonomous collaborative unmanned system. The controller design of UAV often needs to consider output-limited control. Common control methods include model predictive control (MPC), barrier function-based control, etc. However, the above controllers can only guarantee space constraints, and consider more general constraints. That is, for the simultaneous constraints of time and space, the control method based on the predetermined performance specification can effectively solve the dual constraints of time and space, improve the intelligence level of the UAV, and make it more accurate to complete the target task. The unmanned system based on the leader-follower framework has simple implementation and application scalability, which is conducive to the realization of distributed collaborative control of unmanned systems and further improves the autonomy of unmanned systems.

Compared with the prior art, the advantages of the present invention are:

(1) High degree of miniaturization

The invention adopts a low-cost perception tool, namely airborne vision, which has the characteristics of small size and light weight. In addition, the control law based on the performance function has strong robustness, low computational complexity, and does not depend on model information. It reduces the computing load of the UAV, so it can further reduce the size of the UAV in terms of sensing and computing hardware, which is beneficial to a large number of low-cost, miniaturized UAVs to establish a large-scale autonomous collaborative unmanned system. .

(2) Consider time constraints

The control law proposed by the invention ensures that the output error satisfies the constraints of time and space at the same time, and the output error converges to a predetermined residual error with the time function of absolute decay, so that the multi-UAV autonomous cooperative control system can be completed more accurately target task.

(3) High degree of autonomy

The unmanned system proposed by the present invention is based on the navigator-follower framework, which has the simplicity of implementation and application scalability, and can establish unmanned systems of different scales according to different tasks. The distributed cooperative control of unmanned systems is realized based on airborne vision. There is no need for communication between UAVs, and the cooperative control between UAVs is realized by relying on rich visual information, which further enhances the autonomy of unmanned systems.

Description of drawings

1 is a general block diagram of the process flow of a vision-based multi-UAV cooperative control method based on vision and performance constraints of the present invention;

2 is a schematic diagram of an embodiment of a vision-based multi-UAV cooperative control method based on vision and performance constraints according to the present invention;

FIG. 3 is a specific flowchart of a leader-follower control method of a vision-based multi-UAV cooperative control method based on vision and performance constraints of the present invention.

Detailed ways

In order to make those skilled in the art better understand the technical solutions of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings.

Aiming at the existing problems, the present invention provides a vision-based multi-UAV cooperative control method based on vision and performance constraints, as shown in Figure 1 and Figure 3, including the following steps:

Step S1: Decompose the overall target task to obtain mutually independent sub-tasks, confirm the type and quantity of UAVs according to the sub-tasks, and establish an unmanned system for the sub-tasks;

Step S2: select an optimal leader in the unmanned system of the current sub-task, and the follower detects the ArUco fiducial marker carried by the leader, thereby obtaining its relative pose relative to the leader;

Step S3: establishing an unmanned system model of the sub-task based on the leader-follower framework;

Step S4: designing an error transformation method based on the predetermined task performance specification;

Step S5: Design the PID control law of the follower according to the transformed error to ensure that the follower follows the leader according to the predetermined task performance, and finally achieves the goal of autonomous cooperative control of multiple UAVs.

Specifically, step S1 specifically includes the following steps:

Step S101: decompose the overall target task to obtain each subtask that is independent of each other;

Step S102: Select the form of aerial drone, ground drone, or combination of aerial drone and ground drone according to the requirements of the subtask, and determine the number of drones to establish an unmanned system for the subtask.

Step S2 specifically includes the following steps:

Step S201: in the unmanned system of the current sub-task, the control station selects an optimal UAV as the leader of the task issued by the control station in the unmanned system of the current sub-task;

Step S202: Each UAV is equipped with an ArUco square fiducial marker of known size, and the follower detects the ArUco marker using airborne vision. The ArUco marker consists of a black border and an internal binary matrix that determines its identifier. A single marker can be Provide enough correspondence (four corners) to obtain the pose of the camera relative to the ArUco marker. According to the fixed transformation relationship between the camera coordinate system and the follower coordinate system, the ArUco marker coordinate system and the leader coordinate system, obtain the leader and follower. The relative pose ζ _lf between the drones, and the ArUco mark provides a corresponding ID for each drone to ensure reliable and effective following.

Step S3 includes establishing a leader-follower model according to the leader-follower framework:

ζ _lf = ζ _l -ζ _f

Among them, ζ _lf is the pose of the leader relative to the follower, ζ _l is the pose of the leader in the world coordinate system, and ζ _f is the pose of the follower in the world coordinate system.

The unmanned system model consists of n-1 above-mentioned leader-follower models, where n is the number of UAVs in the unmanned system.

The S4 specifically includes the following steps:

Step S401: define the error, specifically:

Design an error transformation method with predetermined performance indicators according to the task, and estimate the relative pose between the leader and the follower through the ArUco marker, then the error is defined as

in,

Represents the desired pose of the leader relative to the follower; if the relative pose r _lf between the leader and the follower is indirectly represented by image information according to visual kinematics, then e represents the current acquired image feature and the desired image error in characteristics;

Step S402: Define error performance, specifically:

The error performance function is defined such that the output error e _k converges to a predefined residual set along the absolute decay time function ρ _k :

Among them, _ek is expressed as the kth output error amount of the error vector e, and the parameters Υ _k and

for

ρ _k (0) respectively represent the initial maximum allowable error, so as to ensure that the absolute value of the initial error satisfies 0<||e _k (0)||<ρ _k (0), the time function ρ _k (t) of the design absolute decay is

ρ _k (t)=(ρ ₀ -ρ _∞ )e ^-lt +ρ _∞

where the parameter l>0 controls the speed of exponential convergence,

Represents the steady state level of the predetermined mission performance specification, which can be designed to be small enough to guarantee the mission performance specification;

Step S403: Set the output error function, specifically:

To achieve control that meets the task performance specification, the output error is set as:

e _k =S(ε _k )ρ _k (t)

where S(ε _k ) is a monotonically increasing continuous smooth function that satisfies the following requirements:

According to the above requirements, the designed transformation function is:

Step S404: Obtain an error transformation function with a predetermined performance specification: define χ _k =e _k /ρ _k , since S(ε _k ) is strictly increasing, its inverse function always exists, so it has a predetermined performance specification. The error transformation ε _k is described as:

Step S5 specifically includes the following steps:

Step S501: Design the PID control law of the follower according to the transformed error to ensure the convergence of the transformed error _εk . The discrete form of the control law is as follows:

Among them, _uk represents the kth control variable, k _p is the proportional coefficient, _ki is the integral coefficient, and k _d is the differential coefficient;

Step S502: According to the properties of the error transformation function, when the transformed error ε _k converges, from the error transformation ε _k function with a predetermined performance specification obtained in step S404, it can be obtained that the real state error e _k converges according to the predetermined performance, and there is no Human-machine completes the desired sub-tasks, and completes each sub-task synchronously or sequentially according to the overall task objective, and finally achieves multi-dimensional effective coordination in time, space, and tasks.

As shown in FIG. 2 , a system for applying the multi-UAV cooperative control method based on vision and performance constraints provided by the present invention includes a control station and a plurality of UAVs, and the control station controls a plurality of UAVs. A plurality of the drones, the plurality of the drones include a plurality of aerial drones and a plurality of ground drones, and the plurality of the aerial drones and the plurality of ground drones include a leader and a follower Or, the control station is used to control the leader in the sub-tasks, set offline automatic or real-time manual tasks for the leader, the follower and the leader maintain the desired relative pose, and follow the designated leader It can realize the autonomous cooperative control of multiple UAVs.

The multi-UAV cooperative control system can realize three sub-tasks of air-to-air, air-to-ground and ground-to-ground. The leader-follower framework builds a model, the follower detects the ArUco fiducial marker carried by the leader, and obtains its relative pose relative to the leader, uses a predetermined performance function to do error transformation and designs the follower's controller, so that Follower 1-1, follower 1-2 and leader 1 maintain the desired relative pose, and follow the designated leader 1 to move. Similarly, in the air-to-ground and ground-to-ground sub-missions, the control station selects the optimal leader 2, and then builds a model according to the leader-follower framework, using only the perception and control method of airborne vision, and finally completes the The aerial drones land on the ground unmanned mobile platform and the ground unmanned mobile platform formation and other coordinated tasks. It should be noted that, no matter it is an air-to-air, air-to-ground, or ground-to-ground subtask, only one leader is selected, but the number of followers can be multiple.

The UAV includes a control unit, a sensing unit, a communication unit and a power supply unit. The communication unit is used for the drone to receive the tasks issued by the control station, the sensing unit is an onboard camera, and is used for the follower to detect the ArUco mark carried by the leader, and the control unit is the onboard CPU, which is used for computing. And the control law of the drone is given, and the power supply unit provides power for the drone.

The sensing unit uses airborne vision to enable the follower to detect the ArUco mark on the leader, the control unit estimates and controls the state of the UAV, and the power supply unit provides electrical energy for the UAV.

The aerial drone includes an onboard camera, which can be rotated appropriately according to different tasks. When performing an air-to-air mission, the optical axis of the airborne camera of the aerial drone faces forward, so that it can better capture the visual characteristics of the air leader in front. The camera's optical axis is down, making it better to capture the visual features of the pilot on the ground.

Adopting a multi-UAV cooperative control method based on vision and performance constraints provided by the present invention, its technical advantages are as follows:

It can realize multi-dimensional effective coordination of time, space and tasks in the absence of GPS, and meet the needs of miniaturization, intelligence and autonomy of unmanned systems. The cooperative control of the unmanned system mainly relies on perception and control technology. The visual perception method relies on the ArUco marker detected by the airborne camera to obtain the pose of the target relative to the local coordinate system. It does not rely on GPS and can be deployed in indoor or outdoor scenes. It has the advantages of small size, low cost, and rich target information, which is beneficial to a large number of low-cost, miniaturized UAVs to establish a large-scale autonomous and collaborative unmanned system. The controller design of UAV often needs to consider output-limited control. Common control methods include model predictive control (MPC), barrier function-based control, etc. However, the above controllers can only guarantee space constraints, and consider more general constraints. That is, for the simultaneous constraints of time and space, the control method based on the predetermined performance specification can effectively solve the dual constraints of time and space, improve the intelligence level of the UAV, and make it more accurate to complete the target task. The unmanned system based on the leader-follower framework has simple implementation and application scalability, which is conducive to the realization of distributed collaborative control of unmanned systems and further improves the autonomy of unmanned systems.

Compared with the prior art, the advantages of the present invention are:

(1) High degree of miniaturization

The invention adopts a low-cost perception tool, namely airborne vision, which has the characteristics of small size and light weight. In addition, the control law based on the performance function has strong robustness, low computational complexity, and does not depend on model information. It reduces the computing load of the UAV, so it can further reduce the size of the UAV in terms of sensing and computing hardware, which is beneficial to a large number of low-cost, miniaturized UAVs to establish a large-scale autonomous collaborative unmanned system. .

(2) Consider time constraints

The control law proposed by the invention ensures that the output error satisfies the constraints of time and space at the same time, and the output error converges to a predefined residual error with the time function of absolute decay, so that the multi-UAV autonomous cooperative control system can be completed more accurately target task.

(3) High degree of autonomy

The unmanned system proposed by the present invention is based on the navigator-follower framework, which has the simplicity of implementation and application scalability, and can establish unmanned systems of different scales according to different tasks. The distributed cooperative control of unmanned systems is realized based on airborne vision. There is no need for communication between UAVs, and the cooperative control between UAVs is realized by relying on rich visual information, which further enhances the autonomy of unmanned systems.

A vision-based multi-UAV cooperative control method and system based on vision and performance constraints provided by the present invention are described above in detail. The principles and implementations of the present invention are described herein by using specific examples, and the descriptions of the above embodiments are only used to help understand the core idea of the present invention. It should be pointed out that for those skilled in the art, without departing from the principle of the present invention, several improvements and modifications can also be made to the present invention, and these improvements and modifications also fall within the protection scope of the claims of the present invention.

Claims (10)

1. A multi-unmanned aerial vehicle cooperative control system based on vision and performance constraints, characterized in that it includes the following steps:

   Step S1: Decompose the overall target task to obtain mutually independent sub-tasks, confirm the type and quantity of UAVs according to the sub-tasks, and establish an unmanned system for the sub-tasks;

   Step S2: select an optimal leader in the unmanned system of the current sub-task, and the follower detects the mark carried by the leader, thereby obtaining its relative pose relative to the leader;

   Step S3: establishing an unmanned system model of the sub-task based on the leader-follower framework;

   Step S4: designing an error transformation method based on the predetermined task performance specification;

   Step S5: Design the PID control law of the follower according to the transformed error to ensure that the follower follows the leader according to the predetermined task performance, and finally achieves the goal of autonomous cooperative control of multiple UAVs.
2. A multi-UAV cooperative control method based on vision and performance constraints according to claim 1, wherein step S1 specifically includes the following steps:

   Step S101: decompose the overall target task to obtain each subtask that is independent of each other;

   Step S102: Select the form of aerial drone, ground drone, or combination of aerial drone and ground drone according to the requirements of the subtask, and determine the number of drones to establish an unmanned system for the subtask.
3. The method and system for multi-UAV cooperative control based on vision and performance constraints according to claim 1, wherein step S2 specifically includes the following steps:

   Step S201: in the unmanned system of the current sub-task, the control station selects an optimal UAV as the leader of the task issued by the control station in the unmanned system of the current sub-task;

   Step S202: Each UAV is equipped with an ArUco square fiducial marker of known size, and the follower detects the ArUco marker using airborne vision. The ArUco marker consists of a black border and an internal binary matrix that determines its identifier. A single marker can be Provide enough correspondence (four corners) to obtain the pose of the camera relative to the ArUco marker. According to the fixed transformation relationship between the camera coordinate system and the follower coordinate system, the ArUco marker coordinate system and the leader coordinate system, obtain the leader and follower. The relative pose ζ _lf between the drones, and the ArUco mark provides a corresponding ID for each drone to ensure reliable and effective following.
4. A multi-UAV cooperative control method based on vision and performance constraints according to claim 1, wherein step S3 comprises establishing a leader-follower model according to a leader-follower framework:

   ζ _lf = ζ _l -ζ _f

   Among them, ζ _lf is the pose of the leader relative to the follower, ζ _l is the pose of the leader in the world coordinate system, and ζ _f is the pose of the follower in the world coordinate system.
5. A multi-UAV cooperative control method based on vision and performance constraints according to claim 4, wherein the unmanned system model is composed of n-1 above-mentioned leader-follower models, wherein n is the number of UAVs in the unmanned system.
6. A multi-UAV cooperative control method based on vision and performance constraints according to claim 1, wherein the S4 specifically includes the following steps:

   Step S401: define the error, specifically:

   Design an error transformation method with predetermined performance indicators according to the task, and estimate the relative pose between the leader and the follower through the ArUco marker, then the error is defined as

   

   in,

   

   Represents the desired pose of the leader relative to the follower; if the relative pose r _lf between the leader and the follower is indirectly represented by image information according to visual kinematics, then e represents the current acquired image feature and the desired image error in characteristics;

   Step S402: Define error performance, specifically:

   The error performance function is defined such that the output error e _k converges to a predefined residual set along the absolute decay time function ρ _k :

   

   Among them, _ek is expressed as the kth output error amount of the error vector e, and the parameters Υ _k and

   

   for

   

   ρ _k (0) respectively represent the initial maximum allowable error, so as to ensure that the absolute value of the initial error satisfies 0<||e _k (0)||<ρ _k (0), the time function ρ _k (t) of the design absolute decay is

   ρ _k (t)=(ρ ₀ -ρ _∞ )e ^-lt +ρ _∞

   where the parameter l>0 controls the speed of exponential convergence,

   

   Represents the steady state level of the predetermined mission performance specification, which can be designed to be small enough to guarantee the mission performance specification;

   Step S403: Set the output error function, specifically:

   To achieve control that meets the task performance specification, the output error is set as:

   e _k =S(ε _k )ρ _k (t)

   where S(ε _k ) is a monotonically increasing continuous smooth function that satisfies the following requirements:

   

   According to the above requirements, the designed transformation function is:

   

   Step S404: Obtain an error transformation function with a predetermined performance specification: define χ _k =e _k /ρ _k , since S(ε _k ) is strictly increasing, its inverse function always exists, so it has a predetermined performance specification. The error transformation ε _k is described as:

   
7. A multi-UAV cooperative control method based on vision and performance constraints according to claim 6, wherein step S5 specifically includes the following steps:

   Step S501: Design the PID control law of the follower according to the transformed error to ensure the convergence of the transformed error _εk . The discrete form of the control law is as follows:

   

   Among them, _uk represents the kth control variable, k _p is the proportional coefficient, _ki is the integral coefficient, and k _d is the differential coefficient;

   Step S502: According to the properties of the error transformation function, when the transformed error ε _k converges, the error transformation ε _k function with the predetermined performance specification obtained in step S404 can be used to obtain the true state error e _k converges according to the predetermined performance, and more The UAV completes the desired sub-tasks, and completes each sub-task synchronously or sequentially according to the overall task goal, and finally achieves effective multi-dimensional coordination in time, space and tasks.
8. A multi-unmanned aerial vehicle autonomous cooperative control system applying the visual and performance constraint-based multi-unmanned aerial vehicle cooperative control method according to any one of claims 1 to 7, characterized in that it comprises a control station and a plurality of UAVs, the control station controls a plurality of the UAVs, and the plurality of the UAVs includes a plurality of aerial UAVs and a plurality of ground UAVs, a plurality of the aerial UAVs and a plurality of The ground drones all include a leader and a follower. The control station is used to control the leader in the sub-tasks, and set offline automatic or real-time manual tasks for the leader. The follower and the leader keep the desired Relative pose, and follow the movement of the designated leader to achieve autonomous coordinated control of multiple UAVs.
9. The multi-unmanned aerial vehicle autonomous cooperative control system according to claim 8, wherein the unmanned aerial vehicle comprises a control unit, a sensing unit, a communication unit and a power supply unit, and the communication unit is used for the unmanned aerial vehicle to receive The task issued by the control station, the perception unit is an onboard camera for the follower to detect the ArUco mark carried by the leader, the control unit is the onboard CPU, which is used to calculate and give the control of the drone Law, the power supply unit provides electrical energy for the drone.
10. The multi-UAV autonomous cooperative control system according to claim 8, wherein the aerial UAV includes an onboard camera, and the onboard camera can be rotated appropriately according to different tasks.

Images