Multi-robot target tracking formation control method based on performance distance

By applying rigid graph theory and RBFNN network in vehicle formation control, the problem that the prior art cannot describe the vehicle behavior of complex dynamics and nonlinear characteristics is solved, and better transient and steady-state performance and robustness are achieved, ensuring that The stability and flexibility of vehicle formations in complex environments.

CN119024849APending Publication Date: 2024-11-26CHINA UNIV OF GEOSCIENCES (WUHAN)
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411152002.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-21
Publication Date
2024-11-26

AI Technical Summary

Technical Problem

Existing formation control methods cannot accurately describe vehicle behavior with complex dynamics and nonlinear characteristics, especially in complex environments, and are difficult to maintain stability and flexibility.

Method used

Construct a multi-robot target tracking formation control method based on rigid graph theory and RBFNN network. By establishing a vehicle dynamic model, considering actuator failure, unknown nonlinear terms and external perturbation, design performance functions and controllers to realize the target trajectory of the vehicle formation Tracking and formation control.

Benefits of technology

It improves the transient and steady-state performance of the system, enhances robustness, can effectively describe vehicle behavior in complex environments, maintains the stability and flexibility of the formation, and reduces oscillation and steady-state errors.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119024849A_ABST
    Figure CN119024849A_ABST
Patent Text Reader

Abstract

The invention provides a multi-robot target tracking formation control method based on a performance distance, and relates to the field of robot control, and the method comprises the steps: constructing a vehicle dynamics model based on an unknown nonlinear term and external disturbance; constructing a rigid graph framework of the vehicle formation based on a rigid graph theory; an RBFNN network is designed; determining a target trajectory, setting a first vehicle in the vehicle formation as a leader, and setting other vehicles as followers; an unknown nonlinear item and an offset fault item of a vehicle dynamics model are approached through an RBFNN network, and target trajectory tracking control is carried out in combination with a rigid graph frame. The built vehicle dynamics model considers the actuator fault condition of the vehicle, and the model contains unknown nonlinear terms and external disturbance terms, so that the motion of the vehicle can be better reflected.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of robot control, and in particular to a multi-robot target tracking formation control method based on performance distance. Background Art

[0002] Vehicle formation control technology is a specific mobile robot formation technology. Vehicle formation control technology is to communicate with each other among multiple mobile robots, transmit position information and speed information to each other, and after obtaining this information, use the controller to keep a fixed distance and angle between them, so that they can travel on the road in a stable formation.

[0003] According to the sensors and control variables, the existing methods can be divided into position-based, displacement-based, and distance-based formation control methods.

[0004] The position-based formation control method achieves coordinated motion by controlling the position and posture of each robot in space. It has high flexibility and can realize various formation forms and task configurations as needed, but it has high requirements on the environment. It requires accurate position information of each robot in the formation and may be disturbed and perform poorly in complex or unstable environments. In addition, frequent communication between robots is required to achieve position synchronization, which may increase communication overhead and delay, and has high requirements on perception and control.

[0005] The displacement-based formation control method achieves formation collaboration and movement by controlling the relative displacement between each robot. It is relatively simple, has relatively low environmental requirements, and does not require precise absolute position information, but has poor flexibility, is easily affected by external disturbances, and is relatively complex to control.

[0006] Compared with the above two methods, the distance-based formation control method is relatively simple and intuitive. It only needs to control the relative distance between each robot in the formation. It has low environmental requirements and is suitable for environments with weak GPS signals. Moreover, the distance is usually easier to measure and the control is relatively flexible. By changing the distance between each robot, formation switching of different forms and structures can be achieved.

[0007] Most distance-based platoon control methods consider single-integral or double-integral dynamic models, which cannot accurately describe the vehicle behavior with complex dynamics and nonlinear characteristics. Summary of the invention

[0008] The purpose of the present invention is to provide a multi-robot target tracking formation control method based on performance distance in order to solve the problem that the existing formation control method cannot accurately describe the vehicle behavior with complex dynamics and nonlinear characteristics.

[0009] The above-mentioned purpose of the present application is achieved through the following technical solutions: S1: Construct a vehicle dynamics model based on unknown nonlinear terms and external disturbances; S2: Based on the rigid graph theory, a rigid graph framework for vehicle formation is constructed; S3: Design RBFNN network; S4: Determine the target trajectory, set the first vehicle in the vehicle formation as the leader, and the remaining vehicles as followers; use the RBFNN network to approximate the unknown nonlinear terms and bias fault terms of the vehicle dynamics model, and combine it with the rigid graph framework to perform target trajectory tracking control.

[0010] Specifically, rigid graph theory is applied to realize multi-robot target tracking and distance-based formation control problems. A vehicle model is established, and the vehicle actuator failure is considered. The vehicle is modeled as a second-order nonlinear system with actuator failure control input, unknown nonlinear terms and external disturbances. The bias fault term and unknown nonlinear term in the vehicle model are approximated by the RBFNN network. Using the relevant knowledge of rigid graph theory, the stiffness matrix is ​​established through the multi-vehicle formation topology structure. Based on this, the distance error of each edge of the rigid graph is calculated, and the performance function is designed. The distance error of each edge in the rigid frame is constrained under a given target to ensure the transient and steady-state performance of the control. At the same time, in the control process, a vehicle is selected as the leader, and its position error with the target trajectory is considered and added to the designed controller to achieve multi-vehicle tracking of the target trajectory and formation control. The bias fault term and unknown nonlinear term in the vehicle model are approximated by the RBFNN network, and the second-order nonlinear model of the vehicle is obtained. Compared with the single integral or double integral model, the system has stronger robustness and can better describe the behavior of the vehicle in a complex environment. At the same time, the tracking problem of the target trajectory of the leader within the formation is taken into consideration in the design of the controller, and a performance function is added to make the control have better transient and steady-state performance.

[0011] Optionally, step S1 includes: S11: Considering the vehicle actuator failure, the single vehicle dynamics model with unknown nonlinear terms and external disturbances is described as follows:

[0012] in, Indicates The position of the vehicle, Indicates The speed of the vehicle, Indicates that the actuator is faulty Control inputs to the vehicle; represents the nonlinear unmodeled dynamic uncertainty term, represents the external disturbance term, Indicates the number of vehicles in the formation; S12: The form of actuator failure is as follows:

[0013] in, and denote the fault efficiency coefficient and bias fault, respectively, and , and is the unknown fault moment; definition ; S13: The overall vehicle dynamics model of multiple vehicle systems is as follows:

[0014]

[0015] in, and Respectively represent the overall position and speed of the vehicle in the whole system; , Represents the unknown nonlinear terms of the entire system; represents the failure efficiency of the entire system, represents the bias fault term of the entire system, represents the external disturbance term of the entire system.

[0016] Optionally, step S2 includes: Constructing an undirected graph represents the topological relationship between vehicle formations, where The vertices and Composition edge; definition and Respectively represent the number of vertices and edges of the undirected graph; the vertices of the undirected graph represent each vehicle, and the edges of the undirected graph represent the communication relationship between each vehicle; Define a pair of rigid graph frames ,in and , is a vertex in two-dimensional or three-dimensional space location; about For any order of edges in The edge function of : , as shown below:

[0017] in, represents the Euclidean norm; Rigid Graph Framework The stiffness matrix Defined as:

[0018] Assuming the desired vehicle formation shape is a regular pentagon, construct the stiffness matrix of the regular pentagon .

[0019] Optionally, the stiffness matrix As shown below:

[0020] in, ,and Represents vertices exist Position in coordinates.

[0021] Optionally, step S3 includes: Assuming that the unknown nonlinear term and the bias fault term are smooth, then in the compact set , which is approximately:

[0022] in, represents the unknown nonlinear term; represents the paranoid fault term; and is the weight of the RBFNN neural network; and represents the approximation error; ,and and is a known smooth function, represents the number of neurons; Estimated value and Given by:

[0023] in, and is the weight matrix of the RBFNN neural network, the activation function The choices are as follows:

[0024] in, is the center of the activation function, is the width of the Gaussian function.

[0025] Optionally, step S4 includes: The tracking error between the leader and target trajectories is defined as follows:

[0026] in, , weight and Respectively, leaders and Tracking error in direction, vector and denote the target trajectory and the leader's position, respectively, and is the reference velocity of the target trajectory; Defining vehicles in a platoon and vehicles The distance error between them is as follows:

[0027] in, ; Indicates vehicle With vehicle The distance between Indicates vehicle and vehicles The expected distance between The dynamics of the distance error are as follows:

[0028] in, express The transpose of Indicates vehicle speed; Indicates vehicle speed; Indicates The speed of a vehicle.

[0029] Optionally, step S4 further includes: The control goal of the target trajectory tracking control is to design a decentralized control protocol so that the leader can follow the target trajectory in the event of a vehicle actuator failure, while the distance error between vehicles in the vehicle formation satisfies the constraint:

[0030]

[0031] in, and is satisfied and The decreasing performance limit of represents the lower bound of the distance error; represents the upper bound of the distance error; because and is decreasing, select and , obtain the bounds of the distance error and ensure the infinitesimal stiffness of the entire vehicle formation; and Respectively represent the initial values ​​of the upper and lower bounds of the distance error; According to rigid graph theory, if is the minimum rigidity, is a small enough positive number that satisfies ,but It is also always the least rigid; when , ,and When is a sufficiently small positive constant. To ensure the minimum stiffness of the rigid graph, set ; represents the initial distance error; Indicates distance error; represents a compact set; express A real matrix of ; Defining the squared distance error , as shown below:

[0032] By introducing an exponential performance function: , among which and , transforming the control objective into the following constraints:

[0033] in, ; , , , , They represent a positive constant respectively; A positive constant representing the lower bound of the control target; A positive constant representing the lower bound of the control target; According to the rigid graph theory, the upper and lower bounds of the control objective are calculated. ,as follows:

[0034] Using the error transformation technique, the original dynamic error system with constraints is transformed into an equivalent unconstrained system, considering the following modulation distance squared error:

[0035] The error transformation corresponding to each edge in the rigid graph framework is introduced as follows:

[0036] in Indicates that it corresponds to and The transformation error of Represents a transformation relationship.

[0037] Optionally, step S4 further includes: Define a Lyapunov function that satisfies the stability condition ,have ; Indicates vehicle and The energy function of Indicates vehicle and The control gain of Represents the total energy function of the vehicle formation; represents the total conversion error of the vehicle formation; represents the total control gain; Through Seek guidance, guarantee ,in is a positive constant, express The derivative of Find the expected speed of each vehicle under the given performance boundary conditions while ensuring the rigid vehicle distance error , Included in: Tracking error between leader and target trajectory and the desired velocity of the target trajectory ; According to the expected speed , defining the velocity tracking error , Represents the total velocity matrix of the entire system; Define the energy function , for the energy function To perform derivation; Indicates the speed tracking error; represents the energy function; In the process of derivation, Cauchy-Schwarz inequality and Young inequality are applied to perform scaling; Considering the weight error of RBFNN network approximation , ,in , ,and and They are and An estimated value of represents the ideal approximate parameter vector of unknown weights in the neural network used to approximate the unknown nonlinear terms, i.e., the true value; represents the ideal approximate parameter vector of the unknown weights in the neural network used to approximate the paranoid fault term, i.e., the true value; Define the energy function: ,in , , ; Represents an energy function containing speed error and weight error; represents the weight error matrix of the neural network used to approximate the unknown nonlinear terms; represents the weight error matrix of the neural network used to approximate the paranoid fault term; represents a positive definite matrix; express A matrix where all elements are 1; Designing adaptive law for weight update of RBFNN network , ,in Indicates vehicle The velocity tracking error matrix; Represents the vehicle in the adaptive law for updating the weights of the neural network used to approximate the unknown nonlinear terms The control gain of represents the vehicle in the adaptive law for updating the weights of the neural network to approximate the bias fault term The control gain of represents the vehicle used to approximate the unknown nonlinear term Adaptive law for updating neural network weights; represents the vehicle used to approximate the bias fault term Adaptive law for updating neural network weights At the same time To derive Derivative ,in , get the control input of the formation ; Through the control input , control the entire vehicle formation and track the target trajectory while maintaining the desired shape within the vehicle formation.

[0038] The beneficial effects of the technical solution provided by this application are: 1. The established vehicle dynamics model takes into account the actuator failure of the vehicle, and includes unknown nonlinear terms and external disturbance terms in the model, which can better reflect the movement of the vehicle. Based on the rigid graph theory, a rigid graph framework between vehicles is established to ensure that the formation is always at the minimum rigidity and the formation does not deform. At the same time, in the process of formation control design, the formation's tracking of the target trajectory is also considered, and the formation control and trajectory tracking problems are considered comprehensively.

[0039] 2. Considering the transient and steady-state performance requirements of the system, an exponential performance function is used in the design of the controller, which has advantages in the transient and steady-state performance of the control. The controller is designed with a performance function. In terms of transient performance, it can achieve rapid response of the system, effectively suppress the oscillation phenomenon of the system, reduce overshoot and oscillation, and make the system more stable in the transient process; in terms of steady-state performance, it can ensure that the system has good stability and robustness, and effectively reduce the steady-state error. According to the initial value of the distance error of each edge in the rigid graph, the upper and lower bounds of the performance function are determined. At the same time, the error conversion technology is used to convert the time-varying constrained control problem into an equivalent unconstrained problem, and the virtual control rate and adaptive law are designed. The RBFNN network is used to approximate the paranoid fault terms and unknown nonlinear terms in the vehicle model to ensure system stability and eliminate uncertain effects. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The present application will be further described below with reference to the accompanying drawings and embodiments, in which: Figure 1 is a step diagram in an embodiment of the present application; Figure 2 is the expected formation topology diagram in the embodiment of the present application; Figure 3 is a schematic diagram of target trajectory tracking and formation control targets in an embodiment of the present application; Figure 4 is a formation simulation result diagram in an embodiment of the present application; Figure 5 is a diagram of the squared error and constraint range of the distance between vehicles 1 and 2 in an embodiment of the present application; Figure 6 It is a vehicle formation distance error diagram in an embodiment of the present application. DETAILED DESCRIPTION

[0041] In order to have a clearer understanding of the technical features, purposes and effects of the present application, the specific implementation methods of the present application are now described in detail with reference to the accompanying drawings.

[0042] An embodiment of the present application provides a multi-robot target tracking formation control method based on performance distance.

[0043] Please refer to Figure 1 , Figure 1 : is a step diagram of a multi-robot target tracking formation control method based on performance distance in an embodiment of the present application, comprising: S1: Construct a vehicle dynamics model based on unknown nonlinear terms and external disturbances; Step S1 includes: S11: Considering the vehicle actuator failure, the single vehicle dynamics model with unknown nonlinear terms and external disturbances is described as follows:

[0044] in, Indicates The position of the vehicle, Indicates The speed of the vehicle, Indicates that the actuator is faulty Control inputs to the vehicle; represents the nonlinear unmodeled dynamic uncertainty term, represents the external disturbance term, Indicates the number of vehicles in the formation; S12: The form of actuator failure is as follows:

[0045] in, and denote the fault efficiency coefficient and bias fault, respectively, and , and is the unknown fault moment; definition ; S13: The overall vehicle dynamics model of multiple vehicle systems is as follows:

[0046]

[0047] in, and Respectively represent the overall position and speed of the vehicle in the whole system; , Represents the unknown nonlinear terms of the entire system; represents the failure efficiency of the entire system, represents the bias fault term of the entire system, represents the external disturbance term of the entire system.

[0048] S2: Based on the rigid graph theory, a rigid graph framework for vehicle formation is constructed; Step S2 includes: Constructing an undirected graph represents the topological relationship between vehicle formations, where The vertices and Composition edge; definition and Respectively represent the number of vertices and edges of the undirected graph; the vertices of the undirected graph represent each vehicle, and the edges of the undirected graph represent the communication relationship between each vehicle; Define a pair of rigid graph frames ,in and , is a vertex in two-dimensional or three-dimensional space location; about For any order of edges in The edge function of : , as shown below:

[0049] in, represents the Euclidean norm; Rigid Graph Framework The stiffness matrix Defined as:

[0050] Assuming the desired vehicle formation shape is a regular pentagon, construct the stiffness matrix of the regular pentagon .

[0051] The stiffness matrix As shown below:

[0052] in, ,and Represents vertices exist Position in coordinates.

[0053] S3: Design RBFNN network; Step S3 includes: Assuming that the unknown nonlinear term and the bias fault term are smooth, then in the compact set , which is approximately:

[0054] in, represents the unknown nonlinear term; represents the paranoid fault term; and is the weight of the RBFNN neural network; and represents the approximation error; ,and and is a known smooth function, represents the number of neurons; Estimated value and Given by:

[0055] in, and is the weight matrix of the RBFNN neural network, the activation function The choices are as follows:

[0056] in, is the center of the activation function, is the width of the Gaussian function.

[0057] S4: Determine the target trajectory, set the first vehicle in the vehicle formation as the leader, and the rest of the vehicles as followers; approximate the unknown nonlinear terms and bias fault terms of the vehicle dynamics model through the RBFNN network, and combine it with the rigid graph framework to perform target trajectory tracking control.

[0058] Specifically, for tracking the target trajectory, the first vehicle is taken as the leader of the entire formation, which tracks the target trajectory, and the remaining vehicles follow the leader and maintain the desired formation formation, such as Figure 3 shown.

[0059] Step S4 includes: The tracking error between the leader and target trajectories is defined as follows:

[0060] in, , weight and Respectively, leaders and Tracking error in direction, vector and denote the target trajectory and the leader's position, respectively, and is the reference velocity of the target trajectory; Defining vehicles in a platoon and vehicles The distance error between them is as follows:

[0061] in, ; Indicates vehicle With vehicle The distance between Indicates vehicle and vehicles The expected distance between The dynamics of the distance error are as follows:

[0062] in, express The transpose of Indicates vehicle speed; Indicates vehicle speed; Indicates The speed of a vehicle.

[0063] Step S4 also includes: The control goal of the target trajectory tracking control is to design a decentralized control protocol so that the leader can follow the target trajectory in the event of a vehicle actuator failure, while the distance error between vehicles in the vehicle formation satisfies the constraint:

[0064]

[0065] in, and is satisfied and The decreasing performance limit of represents the lower bound of the distance error; represents the upper bound of the distance error; because and is decreasing, select and , obtain the bounds of the distance error and ensure the infinitesimal stiffness of the entire vehicle formation; and Respectively represent the initial values ​​of the upper and lower bounds of the distance error; According to rigid graph theory, if is the minimum rigidity, is a small enough positive number that satisfies ,but It is also always the least rigid; when , ,and When is a sufficiently small positive constant. To ensure the minimum stiffness of the rigid graph, set ; represents the initial distance error; Indicates distance error; represents a compact set; express A real matrix of ; Defining the squared distance error , as shown below:

[0066] By introducing an exponential performance function: , among which and , transforming the control objective into the following constraints:

[0067] in, ; , , , , They represent a positive constant respectively; A positive constant representing the lower bound of the control target; A positive constant representing the lower bound of the control target; According to the rigid graph theory, the upper and lower bounds of the control objective are calculated. ,as follows:

[0068] Using the error transformation technique, the original dynamic error system with constraints is transformed into an equivalent unconstrained system, considering the following modulation distance squared error:

[0069] The error transformation corresponding to each edge in the rigid graph framework is introduced as follows:

[0070] in Indicates that it corresponds to and The transformation error of Represents a transformation relationship.

[0071] Step S4 also includes: Define a Lyapunov function that satisfies the stability condition ,have ; Indicates vehicle and The energy function of Indicates vehicle and The control gain of Represents the total energy function of the vehicle formation; represents the total conversion error of the vehicle formation; represents the total control gain; Through Seek guidance, guarantee ,in is a positive constant, express The derivative of Find the expected speed of each vehicle under the given performance boundary conditions while ensuring the rigid vehicle distance error , Included in: Tracking error between leader and target trajectory and the desired velocity of the target trajectory ; According to the expected speed , defining the velocity tracking error , Represents the total velocity matrix of the entire system; Define the energy function , for the energy function To perform derivation; Indicates the speed tracking error; represents the energy function; In the process of derivation, Cauchy-Schwarz inequality and Young inequality are applied to perform scaling; Considering the weight error of RBFNN network approximation , ,in , ,and and They are and An estimated value of represents the ideal approximate parameter vector of unknown weights in the neural network used to approximate the unknown nonlinear terms, i.e., the true value; represents the ideal approximate parameter vector of the unknown weights in the neural network used to approximate the paranoid fault term, i.e., the true value; Define the energy function: ,in , , ; Represents an energy function containing speed error and weight error; represents the weight error matrix of the neural network used to approximate the unknown nonlinear terms; represents the weight error matrix of the neural network used to approximate the paranoid fault term; represents a positive definite matrix; express A matrix where all elements are 1; Designing adaptive law for weight update of RBFNN network , ,in Indicates vehicle The velocity tracking error matrix; Represents the vehicle in the adaptive law for updating the weights of the neural network used to approximate the unknown nonlinear terms The control gain of represents the vehicle in the adaptive law for updating the weights of the neural network to approximate the bias fault term The control gain of represents the vehicle used to approximate the unknown nonlinear term Adaptive law for updating neural network weights; represents the vehicle used to approximate the bias fault term Adaptive law for updating neural network weights At the same time To derive Derivative ,in , get the control input of the formation ; Through the control input , control the entire vehicle formation and track the target trajectory while maintaining the desired shape within the vehicle formation.

[0072] Specifically, the simulation effect of the multi-robot target tracking formation control method of the present application is as follows: The controller to verify the platoon control effect under the condition of vehicle actuator failure is assumed to have actuator failure during the operation of the entire platoon, that is, =0, =0, In Matlab software, select the following parameters: The initial position of the vehicle: , , , , ; Expected distance between edges in a rigid graph:

[0073]

[0074] RBFNN network initial weight value: , , , , , , , , , .

[0075] Unknown nonlinear terms: ,in The values ​​are shown in Table 1.

[0076] External disturbance term:

[0077]

[0078]

[0079]

[0080]

[0081] Table 1

[0082] Simulation results of formation Figure 4 As shown, Figure 4 It can be seen that the entire system formation always maintains the desired regular pentagonal shape; Figure 5 It can be seen that during the system change process, the vehicle's distance square error always remains within the given performance constraint range and eventually converges to near 0; Figure 6 It can be seen that the distance errors of the edges in the rigid graph of the entire system gradually converge to 0.

[0083] The above are only exemplary embodiments of the present disclosure and cannot be used to limit the scope of the present disclosure. That is, any equivalent changes and modifications made according to the teachings of the present disclosure are still within the scope of the present disclosure. After considering the disclosure of the specification and the truth of practice, those skilled in the art will easily think of other embodiments of the present disclosure.

[0084] This application is intended to cover any variation, use or adaptation of the present disclosure, which follows the general principles of the present disclosure and includes common knowledge or customary technical means in the art not described in the present disclosure. The description and examples are to be regarded as exemplary only, and the scope and spirit of the present disclosure are defined by the claims.

Claims

1. A multi-robot target tracking formation control method based on performance distance, It is characterized in that The method comprises the following steps: S1: Construct a vehicle dynamics model based on unknown nonlinear terms and external disturbances; S2: Based on the rigid graph theory, a rigid graph framework for vehicle formation is constructed; S3: Design RBFNN network; S4: Determine the target trajectory, set the first vehicle in the vehicle formation as the leader, and the rest of the vehicles as followers; approximate the unknown nonlinear terms and bias fault terms of the vehicle dynamics model through the RBFNN network, and combine it with the rigid graph framework to perform target trajectory tracking control.

2. A multi-robot target tracking formation control method based on performance distance as claimed in claim 1, It is characterized in that Step S1 includes: S11: Considering the vehicle actuator failure, the single vehicle dynamics model with unknown nonlinear terms and external disturbances is described as follows: in, Indicates The position of the vehicle, Indicates The speed of the vehicle, Indicates that the actuator is faulty Control inputs to the vehicle; represents the nonlinear unmodeled dynamic uncertainty term, represents the external disturbance term, Indicates the number of vehicles in the formation; S12: The form of actuator failure is as follows: in, and denote the fault efficiency coefficient and bias fault, respectively, and , and is the unknown fault moment; definition ; S13: The overall vehicle dynamics model of multiple vehicle systems is as follows: in, and Respectively represent the overall position and speed of the vehicle in the whole system; , Represents the unknown nonlinear terms of the entire system; represents the failure efficiency of the entire system, represents the bias fault term of the entire system, represents the external disturbance term of the entire system.

3. A multi-robot target tracking formation control method based on performance distance as claimed in claim 2, It is characterized in that Step S2 includes: Constructing an undirected graph represents the topological relationship between vehicle formations, where The vertices and Composition edge; definition and Respectively represent the number of vertices and edges of the undirected graph; the vertices of the undirected graph represent each vehicle, and the edges of the undirected graph represent the communication relationship between each vehicle; Define a pair of rigid graph frames ,in and , is a vertex in two-dimensional or three-dimensional space location; about For any order of edges in The edge function of : , as shown below: in, represents the Euclidean norm; Rigid Graph Framework The stiffness matrix Defined as: Assuming the desired vehicle formation shape is a regular pentagon, construct the stiffness matrix of the regular pentagon .

4. A multi-robot target tracking formation control method based on performance distance as claimed in claim 3, It is characterized in that The stiffness matrix As shown below: in, ,and Represents vertices exist Position in coordinates.

5. A multi-robot target tracking formation control method based on performance distance as claimed in claim 4, It is characterized in that Step S3 includes: Assuming that the unknown nonlinear term and the bias fault term are smooth, then in the compact set , which is approximately: in, represents the unknown nonlinear term; represents the paranoid fault term; and is the weight of the RBFNN neural network; and represents the approximation error; ,and and is a known smooth function, represents the number of neurons; Estimated value and Given by: in, and is the weight matrix of the RBFNN neural network, the activation function The choices are as follows: in, is the center of the activation function, is the width of the Gaussian function.

6. A multi-robot target tracking formation control method based on performance distance as claimed in claim 5, It is characterized in that Step S4 includes: The tracking error between the leader and target trajectories is defined as follows: in, , weight and Respectively, leaders and Tracking error in direction, vector and denote the target trajectory and the leader's position, respectively, and is the reference velocity of the target trajectory; Defining vehicles in a platoon and vehicles The distance error between them is as follows: in, ; Indicates vehicle With vehicle The distance between Indicates vehicle and vehicles The expected distance between The dynamics of the distance error is as follows: in, express The transpose of Indicates vehicle speed; Indicates vehicle speed; Indicates The speed of a vehicle.

7. A multi-robot target tracking formation control method based on performance distance as claimed in claim 6, It is characterized in that Step S4 also includes: The control goal of the target trajectory tracking control is to design a decentralized control protocol so that the leader can follow the target trajectory in the event of a vehicle actuator failure, while the distance error between vehicles in the vehicle formation satisfies the constraint: in, and is satisfied and The decreasing performance limit of represents the lower bound of the distance error; represents the upper bound of the distance error; because and is decreasing, select and , obtain the bounds of the distance error and ensure the infinitesimal stiffness of the entire vehicle formation; and Respectively represent the initial values ​​of the upper and lower bounds of the distance error; According to rigid graph theory, if is the minimum rigidity, is a small enough positive number that satisfies ,but It is also always the least rigid; when , ,and When is a sufficiently small positive constant. To ensure the minimum stiffness of the rigid graph, set ; represents the initial distance error; Indicates distance error; represents a compact set; express A real matrix of ; Defining the squared distance error , as shown below: By introducing an exponential performance function: , among which and , transforming the control objective into the following constraints: in, ; , , , , They represent a positive constant respectively; A positive constant representing the lower bound of the control target; A positive constant representing the lower bound of the control target; According to the rigid graph theory, the upper and lower bounds of the control objective are calculated. ,as follows: Using the error transformation technique, the original dynamic error system with constraints is transformed into an equivalent unconstrained system, considering the following modulation distance squared error: The error transformation corresponding to each edge in the rigid graph framework is introduced as follows: in Indicates that it corresponds to and The transformation error of Represents a transformation relationship.

8. A multi-robot target tracking formation control method based on performance distance as claimed in claim 7, It is characterized in that Step S4 also includes: Define a Lyapunov function that satisfies the stability condition ,have ; Indicates vehicle and The energy function of Indicates vehicle and The control gain of Represents the total energy function of the vehicle formation; represents the total conversion error of the vehicle formation; represents the total control gain; Through Seek guidance, guarantee ,in is a normal number, express The derivative of Find the expected speed of each vehicle under the given performance boundary conditions while ensuring the rigid vehicle distance error , Included in: Tracking error between leader and target trajectory and the desired velocity of the target trajectory ; According to the expected speed , defining the velocity tracking error , Represents the total velocity matrix of the entire system; Define the energy function , for the energy function To perform derivation; Indicates the speed tracking error; represents the energy function; In the process of derivation, Cauchy-Schwarz inequality and Young inequality are applied to perform scaling; Considering the weight error of RBFNN network approximation , ,in , ,and and They are and An estimated value of represents the ideal approximate parameter vector of unknown weights in the neural network used to approximate the unknown nonlinear terms, i.e., the true value; represents the ideal approximate parameter vector of the unknown weights in the neural network used to approximate the paranoid fault term, i.e., the true value; Define the energy function: ,in , , ; Represents an energy function containing speed error and weight error; represents the weight error matrix of the neural network used to approximate the unknown nonlinear terms; represents the weight error matrix of the neural network used to approximate the paranoid fault term; represents a positive definite matrix; express A matrix where all elements are 1; Designing adaptive law for weight update of RBFNN network , ,in Indicates vehicle The velocity tracking error matrix; Represents the vehicle in the adaptive law for updating the weights of the neural network used to approximate the unknown nonlinear terms The control gain of represents the vehicle in the adaptive law for updating the weights of the neural network to approximate the bias fault term The control gain of represents the vehicle used to approximate the unknown nonlinear term Adaptive law for updating neural network weights; represents the vehicle used to approximate the bias fault term Adaptive law for updating neural network weights At the same time To derive Derivative ,in , get the control input of the formation ; Through the control input , control the entire vehicle formation and track the target trajectory while maintaining the desired shape within the vehicle formation.