Memory cache mechanism-based unmanned ship formation maneuvering control method considering intermittent communication failure

Abstract

The invention discloses an unmanned ship formation maneuvering control method considering intermittent communication failure and based on a memory cache mechanism, and belongs to the field of unmanned ship cooperative control. The method and the device are used for solving the problems of poor control robustness, insufficient formation flexibility and strong dependence on global information in a communication interruption scene in the prior art. The method comprises the following steps: establishing an unmanned ship kinematics and dynamics model; constructing a cache region and endowing historical data with different weights to realize state prediction compensation; defining a cluster communication topology and an affine transformation relationship, and dividing a leader and a follower; designing a positive minimum interval event triggering mechanism to reduce communication frequency; constructing a fully distributed state observer based on memory cache; and designing a distributed formation controller to realize flexible formation control. The method is suitable for stable control and dynamic formation in cooperative tasks of multiple unmanned ships in a limited water area or a complex environment.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of unmanned surface vehicle (USV) cooperative control, and particularly relates to USV formation cooperative maneuver control. BACKGROUND

[0002] In recent years, USVs have become an important direction of intelligent maritime system research due to their high maneuverability and low cost advantage in tasks such as maritime patrol, water quality monitoring, and target tracking. Among them, multi-USV cooperative formation control is the core key to realize its efficient operation, especially when performing tasks in restricted waters such as ports, inland rivers, and near islands, the flexibility and robustness of the formation are particularly important.

[0003] Most existing formation control methods use centralized control or distributed methods based on fixed formation. For example, some research uses consensus control protocol to achieve formation keeping, or introduces artificial potential field function to control the relative position between USVs, thereby improving the cooperative maneuverability to a certain extent. However, these schemes often assume that the system has continuous global communication link support, and lack the ability to respond to unstable communication situations.

[0004] In addition, although the fixed formation control method has a clear structure, it lacks flexibility and is difficult to adapt to complex dynamic environments, especially when there are obstacles in the task area or the path is limited, it is difficult to adjust the formation in real time to avoid obstacles or risks. For example, in a narrow channel, fixed V-shaped or columnar formation may be difficult to adapt to changes in heading due to its overall rigidity.

[0005] In terms of communication strategy, some scholars have proposed control methods based on event-triggered mechanism to reduce communication frequency and alleviate communication load. For example, some research introduces a dynamic threshold event trigger in a networked control system, and only communicates when the system state deviation reaches a certain threshold, thereby reducing resource waste. However, most of these methods ignore the state prediction or compensation mechanism during communication interruption, resulting in inaccurate observations and ineffective control instructions in the case of intermittent disconnection.

[0006] To address the state observation problem, some methods introduce extended Kalman filter (EKF), sliding mode observer, and other tools to enhance system state estimation accuracy, but the observation accuracy may still decrease during intermittent communication. At the same time, such observers often rely on global information and are difficult to deploy in fully distributed systems. In addition, frequent updates of the controller can improve response speed, but also increase the risk of system attacks, affecting long-term stability.

[0007] In summary, the prior art has the following defects: lack of robust control mechanism suitable for intermittent communication failure scenarios, insufficient flexibility of formation control, strong dependence on global information and low efficiency of communication resource utilization, which is difficult to meet the comprehensive requirements of stability, safety and flexibility of unmanned ship formation system in actual complex task environment. SUMMARY

[0008] To solve the defects of lack of robust control mechanism suitable for intermittent communication failure scenarios, insufficient flexibility of formation control, strong dependence on global information and low efficiency of communication resource utilization in the prior art, the technical scheme provided by the present application is: An unmanned ship formation maneuvering control method based on a memory buffer mechanism considering intermittent communication failure, comprising: A step of establishing the kinematics and dynamics mathematical model of the unmanned ship to provide system basic parameters for subsequent state estimation and control; A step of constructing a buffer area for storing historical state data, setting different weights according to the time sequence of the data, and using historical data for state prediction compensation during communication interruption; A step of defining the communication topology structure and affine transformation relationship of the unmanned ship cluster, dividing the unmanned ship into leaders and followers, and establishing a one-to-one correspondence between the leaders and the target configuration; A step of designing a positive minimum interval dynamic event trigger mechanism, calculating whether the communication trigger condition is met based on the state error and auxiliary variable, and determining the minimum trigger interval in combination with the buffer area state; A step of constructing a completely distributed state observer based on memory buffer, estimating the expected position and velocity by neighbor state and local cache historical data; A step of designing a distributed formation controller in combination with the affine transformation principle, generating follower control instructions according to the expected state output by the observer and the leader trajectory information, and realizing flexible formation maneuvering control of the unmanned ship cluster.

[0009] Further, in a preferred embodiment, the buffer area sets the weight value according to the distance of the data time from the current time, and the closer the time is to the current time, the higher the weight of the historical data.

[0010] Further, in a preferred embodiment, an auxiliary variable is set in the event trigger mechanism to describe the error dynamic trend, and a set of positive design parameters is set to limit the minimum event trigger interval.

[0011] Further, in a preferred embodiment, the distributed state observer receives neighbor node state information for updating when communication is available, and uses buffer area historical data for error compensation when communication is interrupted.

[0012] Further, in a preferred embodiment, the affine transformation comprises translation, rotation, scaling or any combination thereof, for dynamically adjusting the formation shape of the unmanned surface vehicle cluster.

[0013] Further, in a preferred embodiment, the number of leaders is not less than the configuration space dimension plus one.

[0014] Also provided is an unmanned surface vehicle formation maneuvering control device based on a memory buffer mechanism considering intermittent communication failure, comprising: A kinematics and dynamics mathematical model of the unmanned surface vehicle is established to provide system basic parameters for subsequent state estimation and control; A buffer area is constructed for storing historical state data, and different weights are set according to the time sequence of the data, and historical data is used for state prediction compensation during communication interruption; A communication topology structure and affine transformation relationship of the unmanned surface vehicle cluster are defined, the unmanned surface vehicles are divided into leaders and followers, and a one-to-one correspondence between the leaders and the target configuration is established; A positive minimum interval dynamic event triggering mechanism is designed, and whether the communication triggering condition is met is calculated based on the state error and the auxiliary variable, and the minimum triggering interval is determined in combination with the buffer area state; A completely distributed state observer based on a memory buffer is constructed to estimate the expected position and speed through neighbor state and local buffer historical data; A distributed formation controller is designed based on the affine transformation principle, and a follower control instruction is generated according to the expected state output by the observer and the leader trajectory information, so as to realize flexible formation maneuvering control of the unmanned surface vehicle cluster.

[0015] Also provided is a computer storage medium for storing a computer program, when the computer program is read by a computer, the computer executes the method.

[0016] Also provided is a computer comprising a processor and a storage medium, when the processor reads the computer program stored in the storage medium, the computer executes the method.

[0017] Compared with the prior art, the technical solution provided by the present application has the following advantages: By designing a buffer area in the unmanned surface vehicle system and introducing a memory-based event triggering mechanism, the problem of control instability caused by intermittent communication failure is effectively solved. Compared with the traditional method of setting the control quantity to zero or maintaining the old value during communication interruption, this scheme uses historical state information in the buffer for prediction compensation, so that the system can still maintain high control accuracy and response continuity during non-communication period, thereby improving the anti-interference ability and robustness of the system.

[0018] The proposed distributed memory observer realizes accurate estimation of the desired state under local communication conditions by fusing the leader's partial state and neighbor node information. Unlike centralized observers that rely on global information, this method enables fully distributed deployment, enhancing the scalability and practicality of the observer in large-scale unmanned ship systems, while maintaining good state observation performance in intermittent communication environments, improving the system's operational adaptability.

[0019] A positive minimum interval dynamic event triggering mechanism is adopted to set a dynamic threshold for communication triggering conditions, ensuring that inter-ship communication only occurs when necessary, effectively avoiding the system load and energy consumption problems caused by frequent communication. Compared with existing fixed threshold or periodic sampling communication methods, this method not only reduces the communication frequency, but also avoids the occurrence of Zeno phenomenon, improving the stability and execution efficiency of the system.

[0020] An affine transformation method is introduced in the formation control layer to realize dynamic reconstruction and flexible adjustment of the unmanned ship formation, enabling it to adaptively adjust the formation shape and heading according to the leader's position changes. Compared with the traditional fixed formation structure that cannot cope with complex environmental changes, this method supports various formation evolution forms including rotation, scaling, translation, etc., greatly improving the system's maneuverability and task adaptability in narrow sea areas or obstacle-dense scenarios.

[0021] The constructed leader-follower model combined with affine localizability theory can guide the entire unmanned ship formation for efficient cooperative control through a small number of leaders. Based on maintaining the distributed structure of the system, this model realizes overall formation transformation by adjusting the leader's position, reducing control costs and avoiding dependence on centralized path planning systems, improving the system's rapid deployment capability and flexible scheduling level in actual tasks.

[0022] Suitable for stable control and flexible maneuvering scenarios in multi-unmanned ship cooperative formation tasks in restricted water areas or complex environments. BRIEF DESCRIPTION OF DRAWINGS

[0023] Figure 1 A schematic diagram of an unmanned ship formation maneuvering control method considering intermittent communication failure based on a memory caching mechanism; Figure 2 A curve graph of unmanned ship formation maneuvering trajectory; Figure 3 A curve graph of desired position observation error; Figure 4 A curve graph of desired speed observation error. DETAILED DESCRIPTION

[0024] In order to make the advantages and beneficial effects of the technical solutions provided by the present application clearer, the technical solutions provided by the present application will be described in further detail below in conjunction with the drawings, and the specific embodiments are as follows: Embodiment one, the embodiment provides a memory buffer mechanism-based unmanned surface vehicle formation maneuvering control method considering intermittent communication failure, comprising: The step of establishing the kinematics and dynamics mathematical model of the unmanned surface vehicle provides system basic parameters for subsequent state estimation and control; The step of constructing a buffer area for storing historical state data, setting different weights according to the time sequence of the data, and performing state prediction compensation with historical data during communication interruption; The step of defining the communication topology structure and affine transformation relationship of the unmanned surface vehicle cluster, dividing the unmanned surface vehicle into leaders and followers, and establishing a one-to-one correspondence between the leaders and the target configuration; The step of designing a positive minimum interval dynamic event triggering mechanism, calculating whether the communication triggering condition is met based on the state error and auxiliary variable, and determining the minimum triggering interval in combination with the buffer area state; The step of constructing a completely distributed state observer based on memory buffer, estimating the expected position and velocity through neighbor state and local cache historical data; The step of designing a distributed formation controller in combination with the affine transformation principle, generating follower control instructions according to the expected state output by the observer and the leader trajectory information, and realizing flexible formation maneuvering control of the unmanned surface vehicle cluster.

[0025] The buffer area sets the weight value according to the distance of the data time from the current time, and the closer the time is to the current time, the higher the weight of the historical data.

[0026] The auxiliary variable is set in the event triggering mechanism to describe the error dynamic trend, and a set of positive design parameters is set to limit the minimum event triggering interval.

[0027] The distributed state observer receives neighbor node state information for updating when communication is available, and uses buffer area historical data for error compensation when communication is interrupted.

[0028] Affine transformation includes translation, rotation, scaling or any combination thereof, which is used to dynamically adjust the formation mode of the unmanned surface vehicle cluster.

[0029] The number of leaders is not less than the dimension of the configuration space plus one.

[0030] Embodiment two, the embodiment is a further detailed description of the technical solutions provided by embodiment one, and the specific embodiments are as follows: An unmanned surface vehicle (USV) formation maneuvering control method based on a memory buffer mechanism considering intermittent communication failures, the overall process includes establishing a motion model, constructing a buffer mechanism, defining a formation topology relationship, designing an event trigger mechanism, constructing a distributed observer, and implementing an affine formation control, which is as follows: First, the kinematic and dynamic mathematical models of the USV are established to provide a system basis for subsequent control and observer design. The model includes the position coordinates, heading angle, longitudinal velocity, lateral velocity, and yaw angle velocity of the USV in two-dimensional space, and introduces a nonlinear coupling term to describe the motion state of the USV under the action of hydrodynamic force. By setting a set of design parameters, the system dynamics is converted into a standard form suitable for control and state estimation, providing accurate calculation basis for state feedback in formation control.

[0031] Based on the construction of the mathematical model, in view of the problem that communication interruption may cause state update interruption, a finite capacity buffer area is introduced in each USV. The buffer area is used to store historical state data, and different weights are given to data at different time points, and the closer to the current data, the higher the weight. The buffer mechanism can provide approximate state information during temporary communication interruption, effectively reducing the impact of data packet loss on system control accuracy, and providing data support for subsequent trigger mechanism and observer.

[0032] Based on the buffer mechanism and the state model of the USV, the topology structure and formation configuration of the USV cluster are further established. The system communication topology is represented by a directed graph, each USV corresponds to a graph node, and the information exchange between adjacent USVs is defined by edges. On the basis of this topology, the USV cluster is divided into leaders and followers, the leaders are responsible for deciding the target configuration and path trajectory of the overall formation, and the followers complete local control according to neighbor information and leader state. Affine transformation is used to describe the change of the cluster formation shape, so that the overall formation has the ability to translate, rotate, and scale, laying a foundation for the deployment of subsequent formation control methods.

[0033] Subsequently, a memory-based positive minimum interval dynamic event trigger mechanism is constructed. This mechanism is based on the state data in the buffer area, and communication is triggered only when the observation error meets a certain threshold, thereby realizing sparse transmission of control information and reducing the communication burden of the system. The event trigger mechanism includes an auxiliary variable to describe the dynamic trend of the error over time, and a series of positive design parameters are set to control the minimum time interval of event triggering, avoid triggering too frequently, and ensure the stable operation of the system, providing the basis for the design of the distributed observer.

[0034] On the basis of the event-triggered mechanism, a completely distributed state observer based on memory is designed. The observer estimates the current expected position and velocity by receiving the state information uploaded by the neighbor nodes at the trigger time and combining the historical data in the local cache. Specifically, the observer continuously updates the state estimation in the communication available state and uses the cached data to make a predictive correction during the communication interruption, thereby ensuring that the system can maintain high-precision observation of the target state even when there is intermittent communication failure, providing real-time feedback for the following control of the unmanned ship.

[0035] Finally, based on the expected state information output by the above state observer, a distributed formation maneuvering control method for the unmanned ship is constructed based on affine transformation theory. This control strategy relies on the trajectory information of the leader unmanned ship and realizes the synchronous adjustment of the expected target configuration of the follower by real-time solving of the affine transformation parameters. Since the affine transformation can be defined independently of the dynamics layer, the controller design process does not need to rely on global system information, improving the distributed characteristics and implementability of the control strategy. At the same time, this scheme supports time-varying formation adjustment and adapts to the actual operation requirements in complex dynamic water environments.

[0036] Embodiment Three, in combination Figures 1-4 To illustrate the embodiments, the above-mentioned technical solutions are further described in detail through specific embodiments. (1) Establish the kinematic and dynamic mathematical models of the unmanned ship. Define to represent the position coordinates of the unmanned ship, to represent the heading angle of the unmanned ship, to represent the longitudinal and lateral velocities of the unmanned ship, to represent the yaw angle velocity of the unmanned ship. Define a new variable wherein , is a design parameter. Then, the kinematic and dynamic system of the unmanned ship can be represented as: In the formula and represent nonlinear coupling terms, and the specific forms are: represents the control input. , , ; , and represent the hydrodynamic coefficients. represents the mass of the unmanned ship. denotes the inertia matrix.

[0037] (2) Design a buffer zone.

[0038] In practical applications, communication interruption often occurs, that is, there is no information exchange between the communication links of the ships within the time interval of disconnection. Such communication failure may be caused by various adverse factors, such as network attack, communication range limitation or link failure, etc. Therefore, for the needs of actual task scenarios, it is of great significance to carry out research using the intermittent communication framework. This communication framework can be divided into two different states: communication available and communication interruption. Let the initial time be , the set of communication available intervals can be defined as , where satisfies , and the set corresponding to communication interruption can be defined as .

[0039] In order to improve the stability performance of the system, each unmanned ship is added with a buffer zone of size . Define the positive design parameters , which represent the importance of the data in the buffer zone, and satisfy the conditions and .

[0040] It is worth noting that the data closer to the current time point in the buffer zone is more important. Therefore, the selection of parameter is given priority over other parameters, and smaller parameter and larger parameter will lead to an increase in the frequency of event triggering, thereby improving the control performance. In addition, adjusting parameter can change the minimum event triggering interval, for example, larger parameter will result in a smaller minimum event triggering interval.

[0041] (3) Define the topological relationship and affine stress relationship between the unmanned ship cluster.

[0042] The information interaction topology of the unmanned ship cluster system is represented by a directed graph , where denotes the node set, denotes the edge set. Considering that there are unmanned ships in the -dimensional Euclidean space , the Euclidean coordinates of the unmanned ships are defined as , where . It is assumed that the position of each unmanned ship corresponds to each node of the graph , and the communication relationship between adjacent unmanned ships corresponds to each edge of the graph .

[0043] Configuration yes indivual A finite set of points in space is represented by ,Right now The configuration specifies the position relationship of each individual in the unmanned boat swarm system in space, especially the relative position and orientation between them. Directed graph in space and configuration composed of, expressed as .

[0044] definition The unmanned boat is the leader, defining the rest The unmanned boat is the follower. Under this premise, and are used to represent the set of leaders and followers respectively. It can be decomposed into leader configuration and follower configuration, namely and Denote the leader and follower configurations respectively. Assume that each leader Neither can access information from other UAVs and can be based on the desired trajectory Navigation, define the leader's expected configuration as .

[0045] Affine span: a set of points Affine span of Defined as: According to the above definition, the affine span of two different points is a one-dimensional linear space passing through these two points; the affine span of three non-collinear points is a two-dimensional plane passing through these three points; the affine span of four non-coplanar points is a three-dimensional space containing these four points. is restricted to be non-negative, then the affine span degenerates into the convex hull.

[0046] Given any affine span, it is always possible to translate it to a position containing the origin, thereby obtaining a linear space. The dimension of the obtained linear space is defined as the dimension of the affine span. If the dimension of the affine span is , then the affine tensor of these points becomes .

[0047] If there is a set of scalars that are not all zero Make the point set satisfy: and , then the point set is affine dependent, otherwise the point set is affine independent. Define a configuration matrix and an augmented matrix They are: , If and only if The rows of are linearly dependent, that is, there exists Make , then the point set is affine dependent; similarly, if and only if The rows of are linearly independent, then the point set are affine independent. Because have Column, so in space The most common The points are affine independent.

[0048] For the framework , stress is a set of scalar quantities ,in With edge Related, when the node With node When there is attraction between ; When the node With node When there is repulsive force between ; Other situations, If the stress satisfies the following equation, it is called equilibrium stress: Means that the node set Applied to the node The forces on are balanced. It is worth noting that the equilibrium stress can only be determined to a scalar factor. This means that if is the equilibrium stress, then for any , It is also a balanced stress.

[0049] It can be further expressed in matrix form as follows: Where, For the framework The stress matrix satisfies: The stress matrix has a similar structure to the graph Laplacian matrix. The difference is that the weights of the edges in the stress matrix can be positive, negative, or zero, while the edge weights in the graph Laplacian operator are usually positive.

[0050] Consider a configuration in space The affine image of this configuration is defined as: where denotes an affine transformation. is a matrix that can be used to realize rotation, scaling, and shearing with respect to the nominal configuration is a vector that is used to realize translation.

[0051] It is worth noting that any matrix can be decomposed by singular value decomposition into where and are orthogonal matrices, is a diagonal matrix. This shows that any frame in can be obtained by an affine motion of , i.e., rotation , scaling along different axes, another rotation , and finally translation .

[0052] When any point in will correspond to a unique pair of . When for any , there are infinitely many that satisfy .

[0053] Next, the definition of the target affine formation is given: Target affine formation: The target frame of the time-varying target affine formation is: where and are continuous with respect to time and can be constant or time-varying. The desired position of node in the target affine formation is .

[0054] It is worth noting that if each unmanned surface vehicle needs to know , and ​​So that they can all track their respective desired trajectories, i.e., requiring all times to be specified in advance . , and , and store them on each USV, is impractical. In this way, the USV swarm system cannot dynamically respond to the sailing environment such as restricted waters.

[0055] To achieve the distributed control of the target swarm, a leader-follower strategy will be adopted. In this framework, the required formation maneuvers are taken care of by a small number of USVs called leaders, while the other followers coordinate according to the actions of the leaders. As mentioned before, the position of a leader has a one-to-one correspondence with an affine transformation . Therefore, the affine transformation of the USV swarm system is achieved by controlling the positions of the leaders. Since the number of leaders is usually small, it is assumed that they can be effectively controlled, and the leaders in the actual task scenario can be managed by human operators or intelligent decision-making programs. In this setting, it is assumed that the position of each leader is equivalent to the desired position in the target formation, i.e., . Therefore, the control objective of this chapter is to enable the followers to accurately track the desired trajectory based on the leader position information when the leaders reach the desired positions, i.e., when , .

[0056] Next, how to determine the leaders is introduced. In order to manipulate the entire system through the leaders, it is necessary to select suitable nodes as leaders. If a configuration is in , then the position of the leader can uniquely determine the position of the follower. Next, the necessary and sufficient conditions for affine localizability are given.

[0057] Leader condition for affine localizability: A set of points is affine localizable if and only if the affine span of is the standard frame .

[0058] In the standard frame, any USV in the affine span of can be selected as a leader to ensure affine localizability. Since the affine span of requires at least d+1 points, the minimum number of leaders is d+1. For example, at least three leaders are needed in , and at least four leaders are needed in . When there are exactly d+1 leaders, the leaders affine span is There is a one-to-one correspondence between the position of a leader and the affine transformation (A, b), given any leader position p l , i.e., there is always a solution (A, b) to get the follower position p fWhen there are more than d+1 leaders, the positions of leaders must be dependent on each other.

[0059] (4) A positive minimum interval dynamic event-triggered mechanism is proposed.

[0060] Definition and are the estimated values of and respectively. and respectively represent the estimated value of at time , and represents the triggering time of the nth unmanned surface vehicle. The measurement errors and are defined as: Define auxiliary variables and to represent the error dynamics, and the specific expression is: Combining the above formula, the following relationship can be obtained: In order to facilitate the design and analysis of the control method, it is expressed in a compact form: Based on the buffer area, a memory-based distributed positive minimum interval dynamic event-triggered mechanism is as follows: and , where , , and represent the positive design parameters , ; , , represent the positive design preset time parameters represent the positive design parameters.

[0061] According to the above analysis, the triggering time sequence of the unmanned surface vehicle can be determined as: where represents the positive design parameter.

[0062] (5) A distributed observer based on the memory event-triggered mechanism is proposed.

[0063] To observe the expected position of the follower under the intermittent communication mechanism, a memory event triggered fully distributed observer is designed as: When , we have: When , we have: In compact form, we have: When , we have: When , we have: In which, , , , , , .

[0064] The affine transformation used in the design of the memory event triggered fully distributed observer gives the USV swarm system significant maneuverability. This makes it more sensitive to respond to changes in the environment encountered. It is worth noting that the above observer is designed based on the motion layer, so that the affine transformation is independent of the complex dynamics system. This independence enhances the feasibility of the proposed control method. In addition, the above observer ensures that the error system converges within a predetermined time.

[0065] The technical effect is that: A novel affine transformation algorithm is adopted to solve the maneuvering control problem of the USV swarm system. Previous formation control schemes are all based on fixed formation, and the present invention considers affine transformation, including translation, rotation, scaling and combination of these transformations. Due to the inherent flexibility of affine formation configuration, the proposed formation maneuvering control scheme can perform formation tasks in complex environments, including crossing narrow sea areas, avoiding obstacles, etc. And only the leader USV needs to know the position of the obstacle to generate a feasible motion trajectory for formation navigation with obstacle avoidance performance.

[0066] A distributed event-triggered memory observer is proposed. A buffer is embedded in the state observer to store previous states. Thus, the accuracy of the observer in intermittent communication is improved. Unlike the method of setting the control signal to zero during communication interruption, the impact of packet loss caused by communication interruption is mitigated due to the existence of the buffer. Therefore, it provides more reasonable control signals, thereby improving the reliability. The proposed fully distributed observer based on memory event triggering only needs to know a subset of the leader state, and by combining it with the information of the neighbors, the expected state of itself can be observed in real time. In addition, the proposed observer eliminates the dependence on global information, thereby enhancing the practicality of the proposed observer.

[0067] A dynamic event-triggered mechanism based on memory is proposed. The USV only transmits the observed state to the neighbor when the trigger condition is met. This strategy effectively avoids continuous communication, thereby greatly reducing the complexity of communication and the possibility of system failure. It is worth emphasizing that the trigger mechanism can be conceptualized as a timer, and the trigger interval corresponds to From the time required to reduce to zero, the decreasing rate is jointly determined by the system dynamics and the design parameters of the USV. There is a positive lower limit for the event-triggering interval of each USV, and the lower limit is controllable, thereby excluding the occurrence of Zeno phenomenon.

[0068] The above further describes the technical solutions provided by the present application in several specific embodiments, in order to highlight the advantages and benefits of the technical solutions provided by the present application. However, the above several specific embodiments are not used as a limitation on the present application, and any reasonable modification and improvement of the present application, combination and equivalent replacement of the embodiments, etc. based on the spirit and principles of the present application should be included in the protection scope of the present application.

Claims

1. A method for maneuvering control of an unmanned watercraft formation based on a memory cache mechanism taking into account intermittent communication failures, characterized in that: include: Establishing the kinematic and dynamic mathematical model of the unmanned boat to provide basic system parameters for subsequent state estimation and control; Constructing a buffer area for storing historical status data, setting different weights according to the time sequence of the data, and using historical data for status prediction and compensation during communication interruption; The steps of defining the communication topology and affine transformation relationship of the unmanned boat cluster, dividing the unmanned boats into leaders and followers, and establishing a one-to-one correspondence between the leader and the target configuration; Design a positive minimum interval dynamic event trigger mechanism, calculate whether the communication trigger condition is met based on the state error and auxiliary variables, and determine the minimum trigger interval step based on the buffer state; Build a fully distributed state observer based on memory cache to estimate the expected position and velocity using neighbor states and local cached historical data; A distributed formation controller is designed based on the affine transformation principle. Follower control instructions are generated according to the desired state output by the observer and the leader trajectory information, realizing the steps of flexible formation maneuver control of the unmanned boat cluster.

2. The method for controlling the maneuverability of an unmanned watercraft formation based on a memory cache mechanism taking into account intermittent communication failures according to claim 1, characterized in that: The cache area sets the weight value according to how close the data time is to the current moment. The closer the time is to the current historical data, the higher the weight.

3. The method for controlling the maneuverability of an unmanned watercraft formation based on a memory cache mechanism taking into account intermittent communication failures according to claim 1 is characterized in that: Auxiliary variables are set in the event trigger mechanism to describe the dynamic trend of the error, and a set of positive design parameters are set to limit the minimum event trigger interval.

4. The method for controlling the maneuverability of an unmanned watercraft formation based on a memory cache mechanism taking into account intermittent communication failures according to claim 1, characterized in that: The distributed state observer receives state information from neighboring nodes for updates when communication is available, and uses historical data in the cache for error compensation when communication is interrupted.

5. The method for controlling the maneuverability of an unmanned watercraft formation based on a memory cache mechanism taking into account intermittent communication failures according to claim 1, characterized in that: Affine transformation includes translation, rotation, scaling or any combination thereof, which is used to dynamically adjust the formation shape of the unmanned boat cluster.

6. The method for controlling the maneuverability of an unmanned watercraft formation based on a memory cache mechanism taking into account intermittent communication failures according to claim 1, characterized in that: The number of leaders is no less than the dimension of the configuration space plus one.

7. A maneuvering control device for an unmanned boat formation based on a memory cache mechanism taking into account intermittent communication failures, characterized in that: include: Establish the kinematic and dynamic mathematical model of the unmanned boat, and provide the basic system parameters for subsequent state estimation and control; Build a buffer area to store historical status data, set different weights according to the time sequence of the data, and use historical data to perform status prediction and compensation during communication interruptions; A module that defines the communication topology and affine transformation relationship of the UAV cluster, divides the UAVs into leaders and followers, and establishes a one-to-one correspondence between the leader and the target configuration; Design a positive minimum interval dynamic event trigger mechanism, calculate whether the communication trigger conditions are met based on the state error and auxiliary variables, and determine the module of the minimum trigger interval based on the buffer state; Build a fully distributed state observer based on memory cache, a module that estimates the expected position and velocity by using neighbor states and local cached historical data; A distributed formation controller is designed based on the affine transformation principle. Follower control instructions are generated according to the desired state output by the observer and the leader's trajectory information, realizing a module for flexible formation maneuver control of unmanned boat clusters.

8. A computer storage medium for storing a computer program, characterized in that When the computer program is read by a computer, the computer executes the method according to claim 1 .

9. A computer comprising a processor and a storage medium, characterized in that When the processor reads the computer program stored in the storage medium, the computer executes the method according to claim 1 .

Images