Unmanned ship consistency control method and device

Abstract

The application provides a kind of unmanned ship consistency control method and equipment, it is related to unmanned system cooperative control technical field.The method comprises: according to the real-time state of the unmanned ship and the state of the unmanned ship at last event trigger time, construct real-time local measurement error;According to the state of the unmanned ship at last event trigger time and the state of all adjacent unmanned ships of the unmanned ship, construct joint error;Based on the real-time local measurement error and joint error of each unmanned ship, judge whether event trigger condition is satisfied, when any unmanned ship satisfies event trigger condition, obtain and according to the state of the unmanned ship at event trigger condition satisfying time, the state of the adjacent unmanned ship of the unmanned ship and unmanned ship controller gain matrix, update the controller output of the unmanned ship and the adjacent unmanned ship of the unmanned ship, carry out consistency control to each unmanned ship according to controller output;The application can accurately and safely carry out multi-unmanned ship consistency control under the condition that system model is unknown.

Description

Technical Field

[0001] This invention relates to the field of unmanned system cooperative control technology, and in particular to an unmanned vessel consistency control method and device. Background Technology

[0002] With the rapid development of unmanned systems technology, multi-unmanned vessel cooperative systems have become an important research direction in the field of marine engineering. In complex marine missions such as maritime patrol, cooperative exploration, formation operations, environmental monitoring, and maritime search and rescue, multi-unmanned vessel cooperative systems have higher operational efficiency, stronger environmental adaptability, and better fault tolerance compared to single-unmanned vessel operation modes.

[0003] To achieve efficient collaborative operations among multiple unmanned vessels in complex marine environments, it is necessary to design effective collaborative control strategies to ensure that the motion states (position, speed, heading, etc.) of each unmanned vessel converge consistently. This is a key technology to ensure the stable operation of multi-unmanned vessel formations and the successful completion of collaborative tasks.

[0004] Currently, model-free control is mainly used to achieve consistent control of unmanned vessels. Existing model-free control methods mainly include system identification methods and reinforcement learning methods. However, system identification methods typically achieve control design by constructing a system model and identifying parameters. For consistent control of unmanned vessels, this method requires the construction of an intermediate dynamic model first, which increases computational complexity. Furthermore, model mismatch is prone to occur between the model and the actual system, and modeling and parameter estimation errors are inevitably introduced. Reinforcement learning methods, on the other hand, rely heavily on online interaction and trial and error, cannot perform offline learning, have high training costs, and the stability and safety of the system are difficult to strictly control. In addition, unsafe learning behaviors exist, and the safety and reliability lack rigorous theoretical support. Summary of the Invention

[0005] This invention provides a method and device for unmanned vessel consistency control to solve the problem of how to accurately and safely perform consistency control of multiple unmanned vessels under conditions where the system model is unknown.

[0006] In a first aspect, embodiments of the present invention provide a method for unmanned surface vessel consistency control, comprising: For each unmanned vessel in the multi-unmanned vessel system, a real-time local measurement error is constructed based on the real-time state of the unmanned vessel and its state at the time of the last event trigger; a joint error is constructed based on the state of the unmanned vessel at the time of the last event trigger and the states of all its neighboring unmanned vessels. Based on the real-time local measurement error and the joint error of each unmanned surface vessel (USV), it is determined whether the event triggering condition is met. When any USV meets the event triggering condition, the state of the USV at the moment the event triggering condition is met, the states of its neighboring USVs, and the USV controller gain matrix are acquired and updated. The controller outputs of the USV and its neighboring USVs are then updated, and consistency control is performed on each USV based on the controller outputs. The offline construction process of the unmanned vessel controller gain matrix is ​​as follows: conduct offline experiments on at least one unmanned vessel, construct an offline dataset based on the control input matrix and state matrix of the unmanned vessel under the offline experiment; construct linear matrix inequality constraints based on the offline dataset, and solve the controller gain matrix of the unmanned vessel under the linear matrix inequality constraints.

[0007] In one possible implementation, the step of constructing linear matrix inequality constraints based on the offline dataset and solving for the controller gain matrix of the unmanned vessel under the linear matrix inequality constraints includes: Based on the distributed communication connection topology diagram among multiple unmanned vessels, the adjacency matrix of each unmanned vessel is obtained, and the Laplace matrix of the distributed communication connection topology diagram among multiple unmanned vessels is obtained based on the adjacency matrix. By introducing a decision variable matrix, a symmetric positive definite matrix, and positive scalar parameters, a linear matrix inequality constraint is constructed to ensure that the state matrix in the offline dataset satisfies the solvability condition of the linear matrix inequality constraint; the solvability condition is that the state matrix has full row rank. Under the solvable conditions, the controller gain matrix of the unmanned vessel is solved using the control input matrix, the state data matrix, the decision variable matrix, and the Laplace matrix.

[0008] In one possible implementation, the linear matrix inequality constraint is: ; in, For the offline dataset, the first i The state data matrix of the unmanned vessel For the first i The symmetric positive definite matrix of an unmanned surface vessel. Let be the positive definite matrix. For the positive scalar parameter, This is the upper bound matrix for external perturbation.

[0009] In one possible implementation, under the solvable condition, the controller gain matrix of the unmanned vessel is solved using the control input matrix, the state data matrix, the decision variable matrix, and the Laplace matrix, including: pass Solve for the controller gain matrix of the unmanned surface vessel; in, The controller gain matrix for the unmanned surface vessel. For the offline dataset, the first i The control input matrix of an unmanned vessel For the offline dataset, the first i The state matrix of the unmanned vessel, For the first i The decision variable matrix of an unmanned vessel. For the Laplace matrix The non-zero smallest eigenvalue.

[0010] In one possible implementation, the determination of whether the event triggering condition is met based on the real-time local measurement error and the joint error of each unmanned vessel includes: For each unmanned vessel, the norm of the real-time local measurement error of the unmanned vessel is calculated and denoted as the first error norm, and the norm of the joint error of the unmanned vessel is calculated and denoted as the second error norm. Substitute the first error norm and the second error norm into the preset event triggering criterion function to obtain the real-time event triggering criterion function value of the unmanned vessel; The value of the real-time event triggering criterion function is compared with a preset threshold. If the value of the real-time event triggering criterion function is greater than or equal to the preset threshold, it is determined that the unmanned vessel meets the event triggering condition at the current moment. If the value of the real-time event triggering criterion function is less than the preset threshold, it is determined that the unmanned vessel does not meet the event triggering conditions at the current moment. The preset threshold can be determined by one or both of the following methods: static determination and dynamic determination.

[0011] In one possible implementation, the preset threshold obtained based on the static determination method is: ; in, Let be the time from the last time an event was triggered on the unmanned vessel to the current time. For the first i The initial trigger threshold amplitude of the unmanned vessel under the static determination method. For the first i The time decay rate coefficient of the initial trigger threshold amplitude of an unmanned vessel.

[0012] In one possible implementation, the preset threshold obtained based on the dynamic determination method is: ; in, , For the first i The proportional adjustment coefficient of the unmanned vessel under the aforementioned dynamic determination method. for t Time of the first i The preset threshold generation function for unmanned vessels under the dynamic determination method. for The derivative of Let be the time from the last time an event was triggered on the unmanned vessel to the current time. For the first i Error feedback gain coefficient of an unmanned surface vessel. For the first i The unmanned surface vessel obtains the attenuation coefficient of the preset threshold under the dynamic determination method. This is a preset event trigger criterion function. , For the first i The weighting coefficient of real-time local measurement error of an unmanned surface vessel in the event triggering criterion. Let the first error norm be... For the first i The weighting coefficient of the joint error of the most recent event triggering time of the unmanned vessels in the event triggering criterion. This is the second error norm.

[0013] In one possible implementation, the preset threshold is determined by a static determination method and a dynamic determination method. Before comparing the real-time event triggering criterion function value with a preset threshold, the method further includes: The second error norm is compared with the first consistency error switching threshold and the second consistency error switching threshold of the unmanned vessel, respectively, wherein the first consistency error switching threshold is greater than the second consistency error switching threshold, and the second consistency error switching threshold is greater than zero; The real-time event triggering criterion function value is compared with a preset threshold, including: If the second error norm is greater than the first threshold for consistency error switching, the value of the real-time event triggering criterion function is compared with the preset threshold obtained based on the static determination method. If the second error norm is less than or equal to the second threshold for switching the consistency error, the value of the real-time event triggering criterion function is compared with the preset threshold obtained based on the dynamic determination method. If the second error norm is greater than the second threshold for consistency error switching and less than or equal to the first threshold for consistency error switching, the real-time event triggering criterion function value is compared with a preset threshold obtained based on a target determination method, wherein the target determination method is the same as the method used to determine the preset threshold when the real-time event triggering criterion function value was compared with the preset threshold last time.

[0014] In one possible implementation, updating the controller outputs of the unmanned vessel and its neighboring unmanned vessels based on the state of the unmanned vessel at the moment the event trigger condition is met, the states of the unmanned vessels, and the unmanned vessel controller gain matrix includes: Calculate the difference between the state of each adjacent unmanned vessel at the moment the event triggering condition is met and the state of the unmanned vessel at the moment the event triggering condition is met, and record it as the state error of the unmanned vessel. The state error of the unmanned vessel is weighted and summed, and the product of the weighted summation result and the gain matrix of the unmanned vessel controller is calculated to obtain the controller output of the unmanned vessel. When the event trigger condition is met, the state of the unmanned vessel is updated to the state error of its neighboring unmanned vessels, and the controller output of the neighboring unmanned vessels is obtained by solving the problem.

[0015] Secondly, embodiments of the present invention provide an unmanned vessel consistency control device, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the method described in the first aspect or any possible implementation of the first aspect.

[0016] In this embodiment of the invention, by constructing the unmanned vessel controller gain matrix offline, there is no need to establish an accurate dynamic model of the unmanned vessel, thus solving the problem of inaccurate linear dynamic model construction under unknown system model conditions. When any unmanned vessel meets the event triggering condition, the state of the unmanned vessel and its neighboring unmanned vessels at the moment the event triggering condition is met is obtained, and the controller output of the unmanned vessel and its neighboring unmanned vessels is updated based on the unmanned vessel controller gain matrix to perform consistency control. Communication and control updates are performed when the triggering condition is met, effectively reducing the system communication burden. By monitoring the real-time state of each unmanned vessel, a real-time local measurement error is constructed based on the real-time state of each unmanned vessel and the state at the last event triggering time. A joint error is constructed based on the sum of the state differences between each unmanned vessel at the last event triggering time and all neighboring unmanned vessels. This enables the system to maintain the consistency and stability of multiple unmanned vessels even in the presence of external disturbances, improving the robustness of the system. Attached Figure Description

[0017] Figure 1This is a flowchart of the offline construction process of the unmanned vessel controller gain matrix in the unmanned vessel consistency control method provided in this embodiment of the invention; Figure 2 This is a flowchart illustrating the implementation of the unmanned vessel consistency control method provided in this embodiment of the invention. Figure 3 This is a schematic diagram of the communication topology in the unmanned vessel consistency control method provided in this embodiment of the invention; Figure 4 This is a schematic diagram illustrating the change of the consistency error norm over time in the unmanned vessel consistency control method provided in this embodiment of the invention. Figure 5 This is a schematic diagram of the control input for the unmanned vessel consistency control method provided in this embodiment of the invention; Figure 6 This is a schematic diagram of the event triggering time sequence of the unmanned vessel consistency control method provided in the embodiments of the present invention; Figure 7 This is a schematic diagram of the unmanned vessel consistency control device provided in an embodiment of the present invention. Detailed Implementation

[0018] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0019] Figure 1 This is a flowchart illustrating the offline construction process of the unmanned vessel controller gain matrix in the unmanned vessel consistency control method provided in this embodiment of the invention. Figure 1 As shown, the offline construction process of the unmanned surface vessel controller gain matrix includes: In step 101, an offline experiment is conducted on at least one unmanned vessel, and an offline dataset is constructed based on the control input matrix and state matrix of the unmanned vessel under the offline experiment.

[0020] In this embodiment, step 101 includes: A known continuous excitation input signal is applied to at least one unmanned vessel. The control input data of the unmanned vessel is obtained by sampling the continuous excitation input signal at a preset sampling period. The state of the unmanned vessel under the continuous excitation input signal is collected at a preset sampling period to obtain the state data of the unmanned vessel.

[0021] An offline dataset is constructed based on status data and control input data.

[0022] The offline dataset includes a state matrix and a control input matrix; each state in the state matrix includes roll rate, yaw rate, heading angle, roll rate, and roll angle; each control input element in the control input matrix includes rudder angle.

[0023] For example, the offline dataset is: ; in, Indicates the number of the multi-unmanned vessel. , It refers to the number of unmanned vessels. For the first i The state matrix of the unmanned vessel, For the first i The control input matrix of an unmanned vessel For the preset sampling period, T The preset number of sampling periods; t The status data and control input data at each moment are as follows: ; ; in, For the first i unmanned boat t The state data at any given time, For the first i unmanned boat t Control input data at all times, For the first i unmanned boat t The roll rate at any given moment, For the first i unmanned boat t Yaw angular velocity at time t, For the first i unmanned boat t The heading angle at any moment, For the first i unmanned boat t The roll rate at any given moment, For the first i unmanned boat t The roll angle at any moment, For the first i unmanned boat t The rudder of time.

[0024] For example, the offline dataset also includes a state differential matrix and a perturbation matrix: ; in, For the first i The state differential matrix of an unmanned vessel, For the first i The perturbation matrix of an unmanned vessel. For the first i unmanned boat t State differential data at time step For the first i unmanned boat tThe perturbation data at any given time. Among them, It is unmeasurable and unknown data, used only for stability analysis.

[0025] In step 102, linear matrix inequality constraints are constructed based on the offline dataset, and the controller gain matrix of the unmanned vessel is solved under the linear matrix inequality constraints.

[0026] In this embodiment, step 102 includes: Based on the distributed communication connection topology diagram among multiple unmanned vessels, the adjacency matrix of each unmanned vessel is obtained, and the Laplace matrix of the distributed communication connection topology diagram among multiple unmanned vessels is obtained based on the adjacency matrix.

[0027] By introducing decision variable matrices, symmetric positive definite matrices, and positive scalar parameters, linear matrix inequality constraints are constructed to ensure that the state matrix in the offline dataset satisfies the solvable condition of the linear matrix inequality constraints; the solvable condition is that the state matrix has full row rank.

[0028] Under solvable conditions, the controller gain matrix of the unmanned vessel is solved by using the control input matrix, state data matrix, decision variable matrix, and Laplace matrix.

[0029] For example, the distributed communication connection topology among multiple unmanned surface vessels (USVs) treats each USV as a communication node, and the set of nodes is represented as follows: An edge set is formed by any two nodes. If the node Able to receive nodes The information is called a node. With nodes There are edges The topological connection weights between them are ,otherwise, Accordingly, the adjacency matrix is ​​represented as ,in The degree matrix is ​​defined as follows: ,in The Laplace matrix of the distributed communication connection topology is represented as follows: The elements of the Laplace matrix are represented as and Furthermore, if the distributed communication connection topology between multiple unmanned vessels is... If the graph is undirected and connected, then the Laplace matrix of the distributed communication connection topology graph is... There is one zero eigenvalue, and all other eigenvalues ​​are strictly positive.

[0030] Considering that multiple unmanned surface vessel systems meet the following implementation prerequisites: Communication topology connectivity condition: The distributed communication topology between multiple unmanned vessels is an undirected connected graph, that is, there is at least one path consisting of several communication links between any two unmanned vessels, so that the system can achieve global consistency adjustment through neighbor information.

[0031] Control feasibility conditions: Under actual operating conditions, the motion state of the multi-unmanned vessel system can be stabilized and the consistency control requirements can be met by inputting the rudder angle.

[0032] Solvable conditions for offline data: The state matrix in the offline dataset satisfies the solvable condition of the linear matrix inequality constraint; the solvable condition is that the rows of the state matrix are full rank.

[0033] Bounded condition for external disturbances: External disturbances that may exist in the multi-unmanned surface vessel system (such as equivalent disturbances caused by wind, waves, currents, etc.) are unknown but bounded, that is, their amplitude does not exceed a preset upper bound; furthermore, in offline data modeling, this disturbance satisfies the bounded set constraint: there exists an upper bound matrix for external disturbances. This makes the perturbation matrix satisfy This condition is used to introduce a robust margin in controller design, ensuring that the desired control gain remains consistent and stable even when disturbances are present.

[0034] The linear matrix inequality constraint is: ; in, For the first i The symmetric positive definite matrix of an unmanned surface vessel. It is a positive definite matrix. It is a positive scalar parameter. This is the upper bound matrix for external perturbation.

[0035] Assuming all unmanned surface vessels (USVs) have the same model parameter information and the same controllers, therefore all USVs will also have the same controller gain matrix: ; in, The controller gain matrix for the unmanned surface vessel. For the offline dataset, the first i The control input matrix of an unmanned vessel For the offline dataset, the first i The state matrix of the unmanned vessel, For the first i The decision variable matrix of an unmanned vessel. Laplace matrix The non-zero smallest eigenvalue.

[0036] After obtaining the controller gain matrix of the unmanned surface vessel, see [link / reference]. Figure 2 The diagram illustrates the implementation flowchart of the unmanned vessel consistency control method provided in this embodiment of the invention, which is described in detail below: In step 201, for each unmanned vessel in the multi-unmanned vessel system, a real-time local measurement error is constructed based on the real-time state of the unmanned vessel and the state of the unmanned vessel at the time of the last event trigger; a joint error is constructed based on the state of the unmanned vessel at the time of the last event trigger and the states of all its neighboring unmanned vessels.

[0037] For example, the real-time local measurement error is: ; in, For the first i unmanned boat t Real-time local measurement error at time t. For the first i The first unmanned ship The next trigger time is the time when the previous event was triggered. For the first i The first unmanned ship +1 trigger time, which is the time of the next event trigger. No. i The first unmanned ship The state at the moment of the next trigger (the last time the event was triggered), No. i The real-time status of an unmanned vessel.

[0038] The joint error is: ; in, For the first i The first unmanned ship The joint error at the time of the next trigger (the time of the previous event trigger), For the first j The first unmanned ship The state at the moment of the next trigger (the last time the event was triggered), For the first i The unmanned ship and the first j Topology connection weights for unmanned vessels.

[0039] In step 202, it is determined whether the event triggering conditions are met based on the real-time local measurement error and joint error of each unmanned vessel.

[0040] For example, step 202 includes: For each unmanned vessel, calculate the norm of its real-time local measurement error, denoted as the first error norm, and calculate the norm of its joint error, denoted as the second error norm.

[0041] Substitute the first error norm and the second error norm into the preset event triggering criterion function to obtain the real-time event triggering criterion function value of the unmanned vessel.

[0042] The real-time event triggering criterion function value is compared with a preset threshold.

[0043] If the value of the real-time event triggering criterion function is greater than or equal to the preset threshold, the unmanned vessel is determined to meet the event triggering conditions at the current moment.

[0044] If the value of the real-time event triggering criterion function is less than the preset threshold, it is determined that the unmanned vessel does not meet the event triggering conditions at the current moment.

[0045] The preset threshold includes preset thresholds obtained by static determination and preset thresholds obtained by dynamic determination.

[0046] In one embodiment, the preset threshold obtained based on the static determination method is: .

[0047] Accordingly, the preset threshold obtained based on the static determination method is used to determine whether the event triggering condition (i.e., the static triggering condition) is met: ; in, For the first i The most recent incident involving the unmanned vessel was triggered at that time. For the first i The current event triggered on the unmanned vessel. For the first i Real-time local measurement error of an unmanned surface vessel. For the first i The joint error of the most recent event trigger time for the unmanned vessels. , For the first i The weighting coefficient of real-time local measurement error of an unmanned surface vessel in the event triggering criterion. For the first i The weighting coefficient of the joint error of the most recent event triggering time of the unmanned vessels in the event triggering criterion. For the first i The time decay rate coefficient of the initial trigger threshold amplitude of the unmanned vessel. For the first i The initial trigger threshold amplitude of the unmanned vessel event triggering mechanism. inf{} This represents the smallest time operator that satisfies the triggering condition. This is the event trigger criterion function. The first error norm, This is the second error norm.

[0048] In one embodiment, the preset threshold obtained based on the dynamic determination method is: .

[0049] Accordingly, the preset threshold obtained based on the dynamic determination method is used to determine whether the event triggering condition (i.e., the dynamic triggering condition) is met: ; in, For the first i The most recent incident involving the unmanned vessel was triggered at that time. For the first i The current event triggered on the unmanned vessel. For the first i Real-time local measurement error of an unmanned surface vessel. For the first i The joint error of the most recent event trigger time for the unmanned vessels. , For the first i The weighting coefficient of real-time local measurement error of an unmanned surface vessel in the event triggering criterion. For the first i The weighting coefficient of the joint error of the most recent event triggering time of the unmanned vessels in the event triggering criterion. For the first i The proportional adjustment coefficient for the dynamic trigger threshold of unmanned vessels. for t Time of the first i An unmanned surface vessel is used to generate internal state variables for dynamically triggering thresholds. For the first i Error feedback gain coefficient of unmanned surface vessel For the first i The decay coefficient of the dynamic trigger threshold for unmanned vessels. inf{} This represents the minimum time operator that satisfies the event triggering conditions. This is the event trigger criterion function. The first error norm, This is the second error norm.

[0050] In another embodiment, a preset threshold is obtained based on both a static and dynamic determination method. Furthermore, before comparing the real-time event triggering criterion function value with the preset threshold, the method further includes: The second error norm is compared with the first threshold for consistency error switching and the second threshold for consistency error switching of the unmanned vessel, respectively, wherein the first threshold for consistency error switching is greater than the second threshold for consistency error switching, and the second threshold for consistency error switching is greater than zero.

[0051] The real-time event triggering criterion function value is compared with a preset threshold, including: If the second error norm is greater than the consistency error switching first threshold, the real-time event triggering criterion function value is compared with the preset threshold obtained based on the static determination method.

[0052] If the second error norm is less than or equal to the consistency error switching second threshold, the real-time event triggering criterion function value is compared with the preset threshold obtained based on the dynamic determination method.

[0053] If the second error norm is greater than the second threshold for consistency error switching and less than or equal to the first threshold for consistency error switching, the real-time event triggering criterion function value is compared with the preset threshold obtained based on the target determination method. The target determination method is the same as the method used to determine the preset threshold when the real-time event triggering criterion function value was compared with the preset threshold last time.

[0054] Specifically: Definition of the first i The consistency error switching first threshold of the unmanned ships Switching the second threshold for consistency error ,and ; When the i The second error norm of the unmanned ship satisfies When using static event triggering conditions: ; When the i The second error norm of the unmanned ship When using dynamic event triggering conditions: ; When the i The consensus joint error norm of the unmanned ships is between and During this period, the current event triggering mode remains unchanged.

[0055] It is important to note the moment when the event triggering mode switches from static triggering conditions to dynamic triggering conditions. For dynamic variables Initialize it to satisfy: ; This ensures that the trigger threshold remains continuous during the switching process.

[0056] In step 203, when any unmanned vessel meets the event triggering condition, the state of the unmanned vessel at the time the event triggering condition is met, the state of the unmanned vessel's neighboring unmanned vessels, and the unmanned vessel controller gain matrix are obtained and the controller output of the unmanned vessel and its neighboring unmanned vessels is updated, and the unmanned vessels are controlled in a consistent manner according to the controller output.

[0057] In this embodiment, step 203 includes: Calculate the difference between the state of each adjacent unmanned vessel at the moment the event triggering condition is met and the state of the unmanned vessel at the moment the event triggering condition is met, and record it as the state error of the unmanned vessel.

[0058] The state error of the unmanned vessel is weighted and summed, and the product of the weighted sum and the unmanned vessel controller gain matrix is ​​calculated to obtain the controller output of the unmanned vessel.

[0059] The state of the unmanned vessel at the event trigger condition is updated to the state error of the adjacent unmanned vessels, and the controller output of the adjacent unmanned vessels is obtained by solving.

[0060] For example, the controller output of the unmanned vessel is: ; in, For the first i The controller output of the unmanned vessel (the unmanned vessel) is in the period of the [number missing]. i The unmanned vessel is active from the moment this event is triggered until the moment the next event is triggered.

[0061] With the i The unmanned vessel (the unmanned vessel) adjacent to the first j The controller output of the unmanned vessel is: ; in, For the first j The unmanned ship (the first) i The controller output of a neighboring unmanned vessel (UVV). In order to be with the first j The adjacent unmanned ship l Unmanned ships, including the updated [number] i The status of one unmanned vessel is updated, but the status of other unmanned vessels is not updated; the time period is [number missing]. i The unmanned vessel is active from the moment this event is triggered until the moment the next event is triggered.

[0062] The following specific embodiments illustrate the unmanned vessel consistency control method provided by the present invention: The simulation system consists of six unmanned surface vessels (USVs), numbered USV1 to USV6. The communication topology between the USVs is as follows: Figure 3 The graph shown is an undirected connected graph. All unmanned vessels use the same dynamic parameters: ; Among them, subscript Indicates the number of the multi-unmanned vessel. ; express t Time of the first i The roll rate of the unmanned vessel express t Time of the first i The yaw rate of the unmanned vessel express t Time of the first i The heading angle of the unmanned vessel express t Time of the first i The roll rate of the unmanned vessel express t Time of the first i The roll angle of an unmanned vessel express t Time of the first i The rudder of an unmanned boat, express t Environmental disturbances at any time affect the first i The equivalent effects of the heading motion of an unmanned vessel. express t Environmental disturbances at any time affect the first i The equivalent effects of the roll motion of an unmanned surface vessel; the dynamic response time constant of the lateral velocity. The dynamic response time constant of yaw rate The coupling coefficient between lateral velocity and yaw motion The effect of rudder angle on yaw rate gain The influence coefficient of rudder angle on roll motion The influence coefficient of rudder angle on lateral velocity The excitation coefficient of lateral velocity on roll motion Undamped natural frequency Damping coefficient .

[0063] Assuming all unmanned vessels have the same model information, based on this assumption, the first... The dynamic models of unmanned vessels are organized into a unified state-space form: ; The matrices A and B are constructed as follows: , ; During the offline data acquisition phase, an unmanned vessel was selected as a representative, and excitation input signals (control inputs) were applied to it. Collect its status data State differential data and control input A total of 200 sets of data were collected (T=200), with the sampling period set as follows: ; Bounded noise superimposed during data acquisition ,satisfy , The value is the average of the bounded noise. Construct the state matrix. State differential matrix Control input matrix Subsequently, based on the data-driven control gain design method proposed in this invention, the linear matrix inequality is solved to obtain a unified feedback gain. and positive definite matrix as follows: ; Set the static event trigger parameters as follows: , , , At the same time, set , The initial state of each unmanned vessel is set to Under the influence of external bounded disturbances, closed-loop simulation is performed using the data-driven event-triggered consensus control protocol proposed in this invention (i.e., the controller output of the unmanned vessel in the unmanned vessel consensus control method). The simulation results are as follows: Figure 4 — Figure 6 As shown.

[0064] Figure 4 This demonstrates the consistency error. How the norm changes over time: Among them, consistency error for: ; It can be seen that the error converges to the neighborhood of the origin, satisfying the definition of consistency.

[0065] Figure 5 The control inputs of each unmanned vessel were demonstrated. The curve indicates that the control signal is bounded and smooth.

[0066] Figure 6 It shows the sequence of triggering moments under the event triggering mechanism.

[0067] Simulation results show that the data-driven unmanned vessel event-triggered consistency control method proposed in this invention can achieve consistent convergence of multiple unmanned vessel states under conditions of unknown system model and external bounded disturbances. At the same time, the event-triggered mechanism significantly reduces the communication frequency, verifying the effectiveness, robustness and engineering applicability of the method.

[0068] To verify the feasibility of the event-triggered control mechanism in engineering implementation, it is necessary to prove that the Zeno phenomenon, i.e., the situation where an infinite number of events are triggered within a finite time, will not occur during system operation.

[0069] Consider any first An unmanned vessel, its event trigger time sequence is denoted as According to the definition of measurement error, in any interval Inside, there is a real-time local measurement error derivative. Therefore, the upper right derivative of the real-time local measurement error norm satisfies: ; in, The upper right derivative of the local measurement error norm. For a composite norm term, .

[0070] Under the designed assumptions of bounded control inputs and disturbances, the system state variables, auxiliary variables, and external disturbances all remain bounded. Therefore, there exist positive constants. , making Therefore, the upper bound of the error norm can be obtained as follows: According to the event triggering conditions, the time interval between two consecutive event triggers must meet the following requirements: ; From the above inequality, we know that there exists a positive minimum time interval such that: ; in, , It is a positive integer threshold, the existence of which is guaranteed by the definition of Zeno behavior: if Zeno behavior exists in the system, then the event triggering time sequence... There must be a limit point. And for any given There always exists a positive integer. , so that all satisfy The event trigger times all fall within the interval Therefore, within any finite time interval, the number of event triggers is finite, and the Zeno phenomenon will not occur in the system. The above provides a proof that static event triggers exclude Zeno behavior; the proof that dynamic event triggers exclude Zeno behavior is similar and will not be repeated.

[0071] Based on Lyapunov stability theory, the proposed data-driven distributed event-triggered control method is verified to still ensure consistency in multi-unmanned vessel systems under conditions of unknown system model and the presence of disturbances.

[0072] S1. Based on the controller and event triggering mechanism, construct stability constraints that include system state error, event triggering error, and disturbance terms. By jointly designing the constraints, ensure that the constraints satisfy the semi-negative definite condition, thereby guaranteeing the stability of the system under event-triggered control.

[0073]

[0074]

[0075] in, It is a positive definite matrix (where ), , , ,parameter , It is a matrix The largest eigenvalue, It is an equation The solution.

[0076] Under the condition of satisfying the stability constraints, the consensus error of the multi-agent system decays exponentially with time and eventually converges to a bounded neighborhood centered at the zero point. The size of the neighborhood is related to the event triggering parameters and the upper bound of the external disturbance.

[0077] S2. To simplify the analysis of the dynamic characteristics of multiple unmanned vessels, a global state variable is defined: ; Selecting candidate Lyapunov functions According to the dynamic equation We can obtain: ; According to the second matrix inequality condition in S1 The following results can be obtained: ; According to Young's inequality, the inequality is bounded as follows: ; S3. Combining data-driven control gain, event triggering conditions, and the assumption of nonlinear boundedness of the system, the variation of the Lyapunov function is analyzed, showing that the consistency error eventually converges to the zero neighborhood, thereby achieving consistent motion of the multi-unmanned vessel system.

[0078] ; in, , ,here Let represent the smallest eigenvalue of matrix Q and the largest eigenvalue of matrix P, respectively. Therefore, the consistency error... The exponential convergence to the region near zero Furthermore, in the absence of interference, It disappears from the inequality, leading to consistency error. It converges asymptotically to zero.

[0079] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0080] like Figure 7 As shown, this embodiment of the invention also provides an unmanned vessel consistency control device 7, including a memory 71 and a processor 70. The memory 71 stores a computer program 72, and the processor 70 executes the computer program 72 to implement the method described in the above method embodiment.

[0081] In the above embodiments, the descriptions of each embodiment have their own emphasis. Parts not detailed or described in a particular embodiment can be referred to in the relevant descriptions of other embodiments. Unless otherwise specified or in conflict with logic, the terminology and / or descriptions between different embodiments are consistent and can be referenced interchangeably. Technical features in different embodiments can be combined to form new embodiments based on their inherent logical relationships.

[0082] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for consistency control of unmanned surface vessels, characterized in that, include: For each unmanned vessel in the multi-unmanned vessel system, a real-time local measurement error is constructed based on the real-time state of the unmanned vessel and its state at the time of the last event trigger; a joint error is constructed based on the state of the unmanned vessel at the time of the last event trigger and the states of all its neighboring unmanned vessels. Based on the real-time local measurement error and the joint error of each unmanned vessel, it is determined whether the event triggering condition is met. When any unmanned vessel meets the event triggering condition, the state of the unmanned vessel at the time the event triggering condition is met, the state of the unmanned vessel's neighboring unmanned vessels, and the unmanned vessel controller gain matrix are obtained and the controller output of the unmanned vessel and its neighboring unmanned vessels is updated. Consistent control is performed on each unmanned vessel based on the controller output. The offline construction process of the unmanned vessel controller gain matrix is ​​as follows: conduct offline experiments on at least one unmanned vessel, construct an offline dataset based on the control input matrix and state matrix of the unmanned vessel under the offline experiment; construct linear matrix inequality constraints based on the offline dataset, and solve the controller gain matrix of the unmanned vessel under the linear matrix inequality constraints.

2. The unmanned vessel consistency control method according to claim 1, characterized in that, The step of constructing linear matrix inequality constraints based on the offline dataset and solving for the controller gain matrix of the unmanned vessel under these constraints includes: Based on the distributed communication connection topology diagram among multiple unmanned vessels, the adjacency matrix of each unmanned vessel is obtained, and the Laplace matrix of the distributed communication connection topology diagram among multiple unmanned vessels is obtained based on the adjacency matrix. By introducing a decision variable matrix, a symmetric positive definite matrix, and positive scalar parameters, a linear matrix inequality constraint is constructed to ensure that the state matrix in the offline dataset satisfies the solvability condition of the linear matrix inequality constraint; the solvability condition is that the state matrix has full row rank. Under the solvable conditions, the controller gain matrix of the unmanned vessel is solved using the control input matrix, the state data matrix, the decision variable matrix, and the Laplace matrix.

3. The unmanned vessel consistency control method according to claim 2, characterized in that, The linear matrix inequality constraint is: ； in, For the offline dataset, the first i The state data matrix of the unmanned vessel For the first i The symmetric positive definite matrix of an unmanned surface vessel. Let be the positive definite matrix. For the positive scalar parameter, This is the upper bound matrix for external perturbation.

4. The unmanned vessel consistency control method according to claim 2, characterized in that, Under the solvable condition, the controller gain matrix of the unmanned vessel is solved using the control input matrix, the state data matrix, the decision variable matrix, and the Laplace matrix, including: pass Solve for the controller gain matrix of the unmanned surface vessel; in, The controller gain matrix of the unmanned vessel. For the offline dataset, the first i The control input matrix of an unmanned vessel For the offline dataset, the first i The state matrix of the unmanned vessel, For the first i The decision variable matrix of an unmanned vessel. For the Laplace matrix The non-zero smallest eigenvalue.

5. The unmanned vessel consistency control method according to claim 1, characterized in that, The determination of whether the event triggering condition is met based on the real-time local measurement error and the joint error of each unmanned vessel includes: For each unmanned vessel, the norm of the real-time local measurement error of the unmanned vessel is calculated and denoted as the first error norm, and the norm of the joint error of the unmanned vessel is calculated and denoted as the second error norm. Substitute the first error norm and the second error norm into the preset event triggering criterion function to obtain the real-time event triggering criterion function value of the unmanned vessel; The value of the real-time event triggering criterion function is compared with a preset threshold. If the value of the real-time event triggering criterion function is greater than or equal to the preset threshold, it is determined that the unmanned vessel meets the event triggering condition at the current moment. If the value of the real-time event triggering criterion function is less than the preset threshold, it is determined that the unmanned vessel does not meet the event triggering conditions at the current moment. The preset threshold can be determined by one or both of the following methods: static determination and dynamic determination.

6. The unmanned vessel consistency control method according to claim 5, characterized in that, The preset threshold obtained based on the static determination method is: ; in, Let be the time from the last time an event was triggered on the unmanned vessel to the current time. For the first i The initial trigger threshold amplitude of the unmanned vessel under the static determination method. For the first i The time decay rate coefficient of the initial trigger threshold amplitude of an unmanned vessel.

7. The unmanned vessel consistency control method according to claim 6, characterized in that, The preset threshold obtained based on the dynamic determination method is: ; in, , For the first i The proportional adjustment coefficient of the unmanned vessel under the aforementioned dynamic determination method. for t Time of the first i The preset threshold generation function for unmanned vessels under the dynamic determination method. for The derivative of Let be the time from the last time an event was triggered on the unmanned vessel to the current time. For the first i Error feedback gain coefficient of an unmanned surface vessel. For the first i The unmanned surface vessel obtains the attenuation coefficient of the preset threshold under the dynamic determination method. This is a preset event trigger criterion function. , For the first i The weighting coefficient of real-time local measurement error of an unmanned surface vessel in the event triggering criterion. Let the first error norm be... For the first i The weighting coefficient of the joint error of the most recent event triggering time of the unmanned vessels in the event triggering criterion. This is the second error norm.

8. The unmanned vessel consistency control method according to claim 7, characterized in that, The preset threshold can be determined by a static method and a dynamic method. Before comparing the real-time event triggering criterion function value with a preset threshold, the method further includes: The second error norm is compared with the first consistency error switching threshold and the second consistency error switching threshold of the unmanned vessel, respectively, wherein the first consistency error switching threshold is greater than the second consistency error switching threshold, and the second consistency error switching threshold is greater than zero; The real-time event triggering criterion function value is compared with a preset threshold, including: If the second error norm is greater than the first threshold for consistency error switching, the value of the real-time event triggering criterion function is compared with the preset threshold obtained based on the static determination method. If the second error norm is less than or equal to the second threshold for switching the consistency error, the value of the real-time event triggering criterion function is compared with the preset threshold obtained based on the dynamic determination method. If the second error norm is greater than the second threshold for consistency error switching and less than or equal to the first threshold for consistency error switching, the real-time event triggering criterion function value is compared with a preset threshold obtained based on a target determination method, wherein the target determination method is the same as the method used to determine the preset threshold when the real-time event triggering criterion function value was compared with the preset threshold last time.

9. The unmanned vessel consistency control method according to claim 1, characterized in that, The step of updating the controller outputs of the unmanned vessel and its neighboring unmanned vessels based on the state of the unmanned vessel at the moment the event trigger condition is met, the states of the unmanned vessels, and the unmanned vessel controller gain matrix includes: Calculate the difference between the state of each adjacent unmanned vessel at the moment the event triggering condition is met and the state of the unmanned vessel at the moment the event triggering condition is met, and record it as the state error of the unmanned vessel. The state error of the unmanned vessel is weighted and summed, and the product of the weighted summation result and the gain matrix of the unmanned vessel controller is calculated to obtain the controller output of the unmanned vessel. When the event trigger condition is met, the state of the unmanned vessel is updated to the state error of its neighboring unmanned vessels, and the controller output of the neighboring unmanned vessels is obtained by solving the problem.

10. An unmanned vessel consistency control device, characterized in that, It includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the method as described in any one of claims 1 to 9.

Images