A multi-aun formation tracking and obstacle avoidance control method, program, device and storage medium based on asmcp under a cooperation and competition network

Abstract

The present application belongs to the technical field of AUV formation control, and particularly relates to a multi-AUV formation tracking and obstacle avoidance control method, program, device and storage medium based on ASMPC under a cooperation and competition network. The present application is aimed at the problems of unknown time-varying disturbance, obstacle avoidance and energy consumption constraint faced by the multi-AUV two-part formation in the complex marine environment. The present application reduces the influence of the limited step length prediction interval on the dynamic terminal constraint set by designing a fast adaptive sliding mode auxiliary control law to construct a contraction stable domain. At the same time, the obstacle safety distance is converted into a hard constraint condition embedded in the optimization framework to realize formation obstacle avoidance. In order to reduce the calculation resource occupation of online optimization, the present application proposes an event-triggered adaptive sliding mode predictive control architecture. The core lies in constructing a dynamic threshold adjustment mechanism driven by state error. Through dynamic inhibition of unnecessary optimization calculation, the optimal allocation of calculation resources is realized while maintaining the asymptotic stability of the closed-loop system.

Description

Technical Field

[0001] This invention belongs to the field of AUV formation control technology, specifically relating to a method, program, device, and storage medium for multi-AUV formation tracking and obstacle avoidance control based on ASMPC under cooperative and competitive networks. Background Technology

[0002] With the development of technology, autonomous underwater vehicles (AUVs) are increasingly widely used in seabed surveying, mapping, and exploration. Compared with single-platform operations, multi-AUV collaboration has advantages such as a wider operating range, stronger fault tolerance, and higher resource utilization. Existing formation control research often assumes a purely cooperative relationship between nodes, but in actual systems, cooperation and competition coexist, leading to the development of bipartite formations. Bipartite formations divide the AUV cluster into two competing subsets, each performing tasks such as data acquisition, information transmission, and area monitoring. This improves task flexibility, reduces resource competition and functional overlap, and provides new theoretical support for seabed resource exploration and cross-domain heterogeneous joint operations. Effective control of bipartite formations can be achieved using complex graph theory theories such as multi-layer networks. Bipartite formation control under cooperative-adversarial networks is a cutting-edge direction in multi-agent collaboration, having evolved from symbolic graph modeling to dynamic system implementation, showing significant advantages in heterogeneous platform collaboration. However, bipartite formation motion is susceptible to external time-varying disturbances, obstacles, actuator saturation, and energy consumption constraints. While model predictive control can cope with multiple constraints, it still faces significant challenges in complex underwater environments due to issues such as strict power consumption requirements and limited computing resources of underwater AUVs. Summary of the Invention

[0003] The purpose of this invention is to provide a method, program, device, and storage medium for multi-AUV formation tracking and obstacle avoidance control based on ASMPC in a cooperative and competitive network.

[0004] A multi-AUV formation tracking and obstacle avoidance control method based on ASMPC under cooperative and competitive networks includes the following steps:

[0005] Based on a cooperative and competitive network, a communication topology network structure for AUV two-squad formations is set up.

[0006] An event-triggered mechanism is introduced. For the following AUV in a two-part AUV formation, the control input and predicted output sequences at the last event trigger moment, as well as the actual pose vector and velocity vector at the current moment, are obtained. The error weighted norm between the actual pose vector at the current moment and the predicted pose vector at the corresponding moment in the predicted output sequence is calculated. If the current moment reaches the prediction interval of the last event trigger moment or the error weighted norm exceeds the threshold, the current moment is recorded as the event trigger moment, and tracking and obstacle avoidance control calculations are performed. Otherwise, the control input from the last event trigger moment is executed.

[0007] The current moment is the event trigger moment. Based on the desired formation of the AUV binary formation and the pose vector of the lead AUV at the current moment, the desired pose vector of the following AUV at the current moment is obtained, and the pose tracking error, desired velocity vector, and velocity tracking error are calculated. A sliding mode predictive controller is adopted, and a new convergence constraint is constructed by building an adaptive finite-time auxiliary control law. The formation cost function is constructed, and the formation tracking control problem is transformed into a rolling time domain optimization problem. Obstacle information is introduced as a constraint condition, and the optimization problem is solved to obtain the control input and predicted output sequences.

[0008] Furthermore, the AUV duo-formation includes one lead AUV and... Taiwan follows AUV;

[0009] Based on cooperative and competitive networks, a communication topology network structure for AUV two-squad formations is set up. , For a set of nodes, , To lead the way in AUVs; Let be the set of edges. ; It is an adjacency matrix. ; , indicating the first Taiwan AUV and the Taiwan AUV has a cooperative relationship; , indicating the first Taiwan AUV and the Taiwanese AUVs are competitors.

[0010] Furthermore, for the second AUV in a two-part formation... Follow AUV to get the time of the last event trigger. control input and predicted output sequence and at the present moment Actual pose vector With velocity vector Calculate the actual pose vector at the current moment. With the predicted output sequence The predicted pose vector corresponding to the current time step. Error weighted norm ;

[0011]

[0012] in, This is the weight matrix.

[0013] Furthermore, the calculation of pose tracking error specifically involves:

[0014]

[0015]

[0016] in, For the first Follow the AUV at all times The pose tracking error; For the first The rotation matrix that follows the AUV; For the first Follow the AUV at all times The desired pose vector.

[0017] Furthermore, the calculation of the desired velocity vector specifically involves:

[0018]

[0019] in, and All are positive definite matrices. , ,and and The elements in satisfy , ;

[0020] Then calculate the first Follow the AUV at all times Speed ​​tracking error .

[0021] Furthermore, the sliding mode predictive controller is employed to construct new convergence constraints by building an adaptive finite-time auxiliary control law, specifically as follows:

[0022] Using a sliding mode predictive controller, the first... Follow the AUV at all times Fast integral sliding surface ;

[0023]

[0024] in, It is a positive definite matrix; and It is a positive odd number and satisfies ;

[0025] Construct the first Follow the AUV at all times Adaptive finite-time auxiliary control law :

[0026]

[0027]

[0028]

[0029] in:

[0030]

[0031]

[0032]

[0033]

[0034] in, For the first The inertia matrix following the AUV; For the first Coriolis force-added matrix following AUV; For the first Damping matrix following the AUV; For adaptive gain.

[0035] Furthermore, the formation construction cost function is as follows:

[0036]

[0037] in, For the first Follow the AUV at all times Predicted time pose vector; For the first Follow the AUV at all times The predicted first Taiwan AUV at all times The pose vector, based on the formation of the AUV binary formation and Solve this problem; For the first Follow the AUV at all times Predicted time Control input; This is the weight matrix; , ; For the prediction interval;

[0038] The constraints are:

[0039] (1) In the AUV bi-group formation, each following AUV is within the prediction interval The pose vector change within the AUV model must satisfy the discrete state-space equations of the AUV model;

[0040] (2) Obstacle status information is introduced as a constraint condition, and the distance between each following AUV and the obstacle in the AUV platoon is within a safe distance;

[0041] (3) ; ; , The maximum allowed control input; For control intervals;

[0042] (4) Satisfies stability constraints;

[0043] The formation tracking control problem is transformed into a rolling time-domain optimization problem. By solving the optimization problem, the optimal predictive control sequence that minimizes the corresponding formation cost function is obtained. and predicted output sequence ;

[0044] The optimal predictive control sequence is obtained by solving the optimization problem. and predicted output sequence ;

[0045]

[0046]

[0047] make As the first Follow the AUV at all times Optimal control input

[0048] A computer device includes a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the above-described multi-AUV formation tracking and obstacle avoidance control method based on ASMPC under cooperative and competitive networks.

[0049] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the aforementioned multi-AUV formation tracking and obstacle avoidance control method based on ASMPC under cooperative and competitive networks.

[0050] A computer program product includes computer instructions that, when executed by a processor, implement the steps of the aforementioned multi-AUV formation tracking and obstacle avoidance control method based on ASMPC under cooperative and competitive networks.

[0051] The beneficial effects of this invention are as follows:

[0052] To address the challenges of unknown time-varying disturbances, obstacle avoidance, and energy consumption constraints faced by multi-AUV convoys in complex marine environments, this invention constructs a contraction stability domain by designing a fast adaptive sliding mode auxiliary control law, thereby reducing the impact of finite step-size prediction intervals on the dynamic terminal constraint set. Simultaneously, obstacle safety distances are transformed into hard constraints and embedded into the optimization framework to achieve convoy obstacle avoidance. To reduce the computational resource consumption of online optimization, this invention proposes an event-triggered adaptive sliding mode predictive control architecture. Its core lies in constructing a state error-driven dynamic threshold adjustment mechanism, which dynamically suppresses unnecessary optimization calculations, achieving optimal allocation of computational resources while maintaining the asymptotic stability of the closed-loop system. Attached Figure Description

[0053] Figure 1 This is a schematic diagram of the AUV in the inertial coordinate system and the carrier coordinate system.

[0054] Figure 2 This is a block diagram of the overall architecture of the present invention.

[0055] Figure 3 This is a topology diagram of the communication network for an AUV two-team formation system.

[0056] Figure 4 This is a schematic diagram illustrating the event triggering time during operation in one embodiment of the present invention.

[0057] Figure 5 This is a schematic diagram of the motion trajectory during operation in one embodiment of the present invention. Detailed Implementation

[0058] The present invention will now be further described with reference to the accompanying drawings.

[0059] like Figure 1 As shown, this invention considers multiple AUV formations navigating in still water at a fixed depth, and satisfies the following assumptions: 1) All AUV nodes in the formation are rigid bodies with uniform mass and symmetry about three axes, and their deformation is ignored; 2) During the AUV's travel, the first and second derivatives of its path trajectory parameters exist and are continuous.

[0060] This invention addresses the bipartite formation control problem under complex marine constraints such as unknown time-varying disturbances, collision and obstacle avoidance, and energy consumption limitations. It proposes a multi-AUV formation tracking and obstacle avoidance control method based on ASMPC within a cooperative and competitive network. The method transforms obstacle safety distances into hard constraints embedded in the optimization framework to construct a formation obstacle avoidance mechanism. Furthermore, it dynamically adjusts the trigger threshold based on the error between the system's actual and optimal states, effectively reducing the solution frequency of the rolling time-domain optimization problem. The specific block diagram is shown below. Figure 2 As shown.

[0061] This invention sets the desired formation of the AUV bipartite formation and obtains the desired pose vector of each following AUV in the formation at the current moment based on the virtual trajectory of the lead AUV. If the following AUVs are in a cooperative relationship, they are consistent with the virtual trajectory of the lead AUV. If the following AUVs are in a competitive relationship, they are symmetrically consistent with the trajectory of the competing AUVs.

[0062] Based on the expected pose vector and actual pose vector of each following AUV in the AUV bi-segment formation at the current moment, a sliding mode predictive controller is designed. By constructing an auxiliary control law, a new convergence constraint condition is built, and the optimization problem in the rolling time domain is solved to obtain the corresponding control input. Then, the control input is fed back to each following AUV in the AUV bi-segment formation to realize the tracking control task at the current moment.

[0063] Based on the above-designed sliding mode predictive controller, the state information of obstacles is incorporated into the rolling time-domain optimization problem of the predictive controller as a constraint. Combined with the event triggering mechanism, a fast adaptive sliding mode predictive control algorithm with obstacle avoidance function is proposed. If there is a following AUV in the AUV bipartite formation that satisfies the event triggering function, the current event trigger count and triggering time are recorded. According to the state-space model under the cooperative-competitive network, the current state vector of each following AUV in the AUV bipartite formation is estimated. If none of the following AUVs in the AUV bipartite formation meet the triggering condition of the event triggering function, each following AUV normally collects and obtains its own state vector through the communication link.

[0064] Solve the optimization problem and obtain the control input, then feed the control input back to each following AUV in the AUV formation to complete the tracking control at the current moment.

[0065] Based on cooperative and competitive networks, a communication topology network structure for AUV two-squad formations is set up. ;

[0066] The AUV convoy consists of one lead AUV and... Taiwan follows AUV; , For a set of nodes, , For the lead AUV, N is the total number of AUVs in the two-part AUV formation; Let be the set of edges. ; It is an adjacency matrix. ;

[0067] , , indicating the first Taiwan AUV and the Taiwan AUV has a cooperative relationship;

[0068] , , indicating the first Taiwan AUV and the Taiwanese AUVs are competitors;

[0069] For the first in the two-part AUV formation Follow the AUV to obtain the last time the event was triggered, i.e. Control input output at all times and predicted output sequence ;

[0070] Get the Following AUV in the present Actual pose vector at time t , ; In inertial coordinate system Time of the first Follow the position of the AUV, for Time of the first Follow the AUV's heading angle;

[0071] Calculate the actual pose vector at the current moment. With the predicted output sequence The predicted pose vector corresponding to the current time step. Error weighted norm ;

[0072]

[0073] in, This is the weight matrix;

[0074] If the current time reaches the prediction interval of the previous event trigger time, that is... , For the prediction interval, or the error-weighted norm If the threshold is exceeded, record the current value. The timeframe is the event trigger moment, at which tracking and obstacle avoidance control calculations are performed; otherwise, the process continues from the previous event trigger moment. Time control input until At any given moment, a new assessment will be made.

[0075] All AUVs in the AUV bi-group shall perform the above determination respectively;

[0076] For the first in the two-part AUV formation Following the AUV, performing tracking and obstacle avoidance control calculations includes the following steps:

[0077] Step 1: Based on the desired formation of the AUV bi-group and the current... The constant-time navigation AUV pose vector is used to obtain the first AUV in the binary formation. Following AUV in the present The expected pose vector at time 1 ;

[0078] Get the Following AUV in the present The actual velocity vector at time t. Calculate pose tracking error ;

[0079]

[0080]

[0081] in, respectively in the carrier coordinate system Time of the first Follow the AUV's pitch speed, sway speed, and yaw rate; For the first The rotation matrix that follows the AUV;

[0082] Then calculate the first Following the AUV Expected velocity vector at time t :

[0083]

[0084] in, and It is a positive definite matrix. , ,and and The elements in satisfy , ;

[0085] According to the Following the AUV Desired velocity vector at any moment With the actual velocity vector Calculate speed tracking error :

[0086] Step 2: Calculate the fast integral sliding surface ;

[0087]

[0088] in, It is a positive definite matrix. ; and It is a positive odd number and satisfies ;

[0089] Then calculate the first Adaptive finite-time auxiliary control law for following AUV ;

[0090]

[0091]

[0092]

[0093] in:

[0094]

[0095]

[0096]

[0097]

[0098] in, For the first The inertia matrix following the AUV; For the first Coriolis force-added matrix following AUV; For the first Damping matrix following the AUV; For adaptive gain;

[0099] Step 3: Construct the formation cost function, transforming the formation tracking control problem into a rolling time-domain optimization problem. Introduce obstacle state information as constraints, solve the optimization problem, and obtain the control input. and predicted output sequence ;

[0100] Constructing the discrete state-space equations of the AUV model:

[0101]

[0102]

[0103] in, , , ;

[0104] Set the prediction range and control range Set predictive control sequence and predicted output sequence :

[0105]

[0106]

[0107] , and satisfy:

[0108]

[0109] in:

[0110]

[0111]

[0112] Formation cost function:

[0113]

[0114] in, The first in the AUV two-part formation Following the AUV Time prediction The pose vector at time t; The first in the AUV two-part formation Following the AUV The first time predicted Taiwanese AUVs The pose vector at time t, based on the formation of the AUV binary formation and Solve this problem; For the first Follow the AUV at all times Predicted time Control input; This is the weight matrix; , ;

[0115] Transform the formation tracking control problem into a rolling time-domain optimization problem:

[0116]

[0117] The constraints are:

[0118] (1) In the AUV bi-group formation, each following AUV is within the prediction interval The pose vector change within the AUV model must satisfy the discrete state-space equations of the AUV model;

[0119] (2) Obstacle status information is introduced as a constraint condition, and the distance between each following AUV and the obstacle in the AUV platoon is within a safe distance;

[0120] (3) ; ; , The maximum allowed control input;

[0121] (4) Satisfies stability constraints;

[0122] The optimal predictive control sequence is obtained by solving the optimization problem. and predicted output sequence ;

[0123]

[0124] Pick The first item in the list, namely, let That is, to obtain the first The optimal control input for the AUV to follow.

[0125] This invention enables AUV bi-group formations to construct a contraction stability region under complex marine constraints such as unknown time-varying disturbances, collision and obstacle avoidance, and energy consumption limitations, thereby reducing the impact of finite step-size prediction intervals on the dynamic terminal constraint set. This invention employs an event-triggered adaptive sliding mode predictive control architecture, based on which a state error-driven dynamic threshold adjustment mechanism is constructed. By dynamically suppressing unnecessary optimization calculations, it achieves optimal allocation of computational resources while maintaining the asymptotic stability of the closed-loop system.

[0126] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A multi-AUV formation tracking and obstacle avoidance control method based on ASMPC under cooperative and competitive networks, characterized in that: Based on a cooperative and competitive network, a communication topology network structure for AUV two-squad formations is set up. An event-triggered mechanism is introduced. For the following AUV in a two-part AUV formation, the control input and predicted output sequences at the last event trigger moment, as well as the actual pose vector and velocity vector at the current moment, are obtained. The error weighted norm between the actual pose vector at the current moment and the predicted pose vector at the corresponding moment in the predicted output sequence is calculated. If the current moment reaches the prediction interval of the last event trigger moment or the error weighted norm exceeds the threshold, the current moment is recorded as the event trigger moment, and tracking and obstacle avoidance control calculations are performed. Otherwise, the control input from the last event trigger moment is executed. The current moment is the event trigger moment. Based on the desired formation of the AUV binary formation and the pose vector of the lead AUV at the current moment, the desired pose vector of the following AUV at the current moment is obtained, and the pose tracking error, desired velocity vector, and velocity tracking error are calculated. A sliding mode predictive controller is adopted, and a new convergence constraint is constructed by building an adaptive finite-time auxiliary control law. The formation cost function is constructed, and the formation tracking control problem is transformed into a rolling time domain optimization problem. Obstacle information is introduced as a constraint condition, and the optimization problem is solved to obtain the control input and predicted output sequences.

2. The multi-AUV formation tracking and obstacle avoidance control method based on ASMPC in a cooperative and competitive network according to claim 1, characterized in that: The AUV convoy consists of one lead AUV and... Taiwan follows AUV; Based on cooperative and competitive networks, a communication topology network structure for AUV two-squad formations is set up. , For a set of nodes, , To lead the way in AUVs; Let be the set of edges. ; It is an adjacency matrix. ; , indicating the first Taiwan AUV and the Taiwan AUV has a cooperative relationship; , indicating the first Taiwan AUV and the Taiwanese AUVs are competitors.

3. The multi-AUV formation tracking and obstacle avoidance control method based on ASMPC in a cooperative and competitive network according to claim 2, characterized in that: For the first in the two-part AUV formation Follow AUV to get the time of the last event trigger. control input and predicted output sequence and at the present moment Actual pose vector With velocity vector Calculate the actual pose vector at the current moment. With the predicted output sequence The predicted pose vector corresponding to the current time step. Error weighted norm ; in, This is the weight matrix.

4. The multi-AUV formation tracking and obstacle avoidance control method based on ASMPC in a cooperative and competitive network according to claim 3, characterized in that: The calculation of pose tracking error is specifically as follows: in, For the first Follow the AUV at all times The pose tracking error; For the first The rotation matrix that follows the AUV; For the first Follow the AUV at all times The desired pose vector.

5. The multi-AUV formation tracking and obstacle avoidance control method based on ASMPC in a cooperative and competitive network according to claim 4, characterized in that: The calculation of the desired velocity vector is specifically as follows: in, and All are positive definite matrices. , ,and and The elements in satisfy , ; Then calculate the first Follow the AUV at all times Speed ​​tracking error .

6. The multi-AUV formation tracking and obstacle avoidance control method based on ASMPC in a cooperative and competitive network according to claim 5, characterized in that: The sliding mode predictive controller is used to construct new convergence constraints by building an adaptive finite-time auxiliary control law, specifically as follows: Using a sliding mode predictive controller, the first... Follow the AUV at all times Fast integral sliding surface ; in, It is a positive definite matrix; and It is a positive odd number and satisfies ; Construct the first Follow the AUV at all times Adaptive finite-time auxiliary control law : in: in, For the first The inertia matrix following the AUV; For the first Coriolis force-added matrix following AUV; For the first Damping matrix following the AUV; For adaptive gain.

7. The multi-AUV formation tracking and obstacle avoidance control method based on ASMPC in a cooperative and competitive network according to claim 6, characterized in that: The formation cost function is as follows: in, For the first Follow the AUV at all times Predicted time pose vector; For the first Follow the AUV at all times The predicted first Taiwan AUV at all times The pose vector, based on the formation of the AUV binary formation and Solve this problem; For the first Follow the AUV at all times Predicted time Control input; This is the weight matrix; , ; For the prediction interval; The constraints are: (1) In the AUV bi-group formation, each following AUV is within the prediction interval The pose vector change within the AUV model must satisfy the discrete state-space equations of the AUV model; (2) Obstacle status information is introduced as a constraint condition, and the distance between each following AUV and the obstacle in the AUV platoon is within a safe distance; (3) ; ; , The maximum allowed control input; For control intervals; (4) Satisfies stability constraints; The formation tracking control problem is transformed into a rolling time-domain optimization problem. By solving the optimization problem, the optimal predictive control sequence that minimizes the corresponding formation cost function is obtained. and predicted output sequence ; The optimal predictive control sequence is obtained by solving the optimization problem. and predicted output sequence ; make As the first Follow the AUV at all times The optimal control input.

8. A computer device, comprising a memory, a processor, and a computer program stored in the memory, characterized in that: The processor executes the computer program to implement the steps of the method according to any one of claims 1 to 7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When executed by a processor, the computer program implements the steps of the method according to any one of claims 1 to 7.

10. A computer program product comprising computer instructions, characterized in that: When executed by a processor, the computer instructions implement the steps of the method according to any one of claims 1 to 7.

Images