A packet loss detection method for unmanned surface vessel formations based on adaptive fuzzy membership filtering

Abstract

This invention discloses a packet loss detection method for unmanned surface vessel (USV) formations based on adaptive fuzzy membership filtering. The method includes: establishing state equations describing the motion of both the leader and follower USVs; linearizing these nonlinear equations to simplify the complex curvilinear motion relationships into linear mathematical expressions, thus obtaining simplified state equations for the entire formation system; introducing an adjustable measurement noise figure based on the simplified formation model, dynamically adjusting the filtering intensity according to the current measurement noise level, thereby more accurately estimating the real-time state of the USVs; and simultaneously detecting packet loss during state updates, combining preset communication quality indicators and real-time calculated noise levels to determine the reliability of received data. This invention achieves good detection results for packet loss in noisy channels and can be widely applied in the field of data detection.

Description

Technical Field

[0001] This invention relates to the field of data detection, and in particular to a method for detecting packet loss in unmanned surface vessel formations based on adaptive fuzzy set-membership filtering. Background Technology

[0002] Unmanned surface vessels (USVs) can perform exploration and patrol operations in water and have wide applications in military and other fields. Therefore, state estimation and attack detection are necessary to determine whether USVs can perform their missions normally. Since USV formations often operate in the same sea area, all sensors are exposed to a set of strong common disturbance sources, such as waves, wind loads, and changes in electromagnetic and acoustic propagation conditions at the sea surface. In this situation, noise from different sensors is often correlated, and the measurement noise and process noise of the same sensor are also correlated. Furthermore, the statistical characteristics of this noise are difficult to analyze, rendering traditional Kalman filtering inapplicable. Considering that USVs operate in water, their communication environment is harsh, channels are often subject to suboptimal noise, and packet loss is common during communication. Existing detection methods do not consider these factors, making it difficult to determine whether packet loss has occurred in noisy channels. Summary of the Invention

[0003] In view of this, in order to solve the technical problem of low packet loss detection accuracy of existing detection methods in noisy channels, this invention proposes a packet loss detection method for unmanned surface vessel formations based on adaptive fuzzy membership filtering. The method includes the following steps: First, state equations describing the motion of both the leader and follower UVs are established. Then, these nonlinear equations are linearized, simplifying the complex curvilinear motion relationships into linear mathematical expressions, thus obtaining the simplified state equations for the entire formation system. Based on the simplified formation model, an adjustable measurement noise figure is introduced to address the issue of inaccurate sensor measurements. This figure acts like an "adaptive switch," dynamically adjusting the filtering intensity according to the current measurement noise level, thereby more accurately estimating the real-time state of the UVs. During state updates, data packet loss detection is performed simultaneously. Specifically, the reliability of received data is determined by combining preset communication quality indicators and real-time calculated noise levels. Once a data packet is detected as lost or severely interfered with (determined as abnormal data), the system automatically removes it to prevent erroneous information from affecting formation control.

[0004] Based on the above scheme, this invention provides a packet loss detection method for unmanned surface vessel (USV) formations based on adaptive fuzzy ensemble filtering. To mitigate the decrease in estimation accuracy caused by overly conservative noise bound values ​​in traditional ensemble filtering, an online adaptive learning mechanism is introduced. This mechanism dynamically adjusts the scale and shape of the ellipsoidal boundary between process noise and measurement noise, achieving a tighter ensemble envelope while maintaining the noise boundedness assumption. This improves the accuracy and robustness of asynchronous multi-sensor fusion estimation. Furthermore, by adding a scalar channel and an amplification factor, this method achieves better detection results for packet loss in noisy channels. It also reduces the impact of erroneous data on the system state. Attached Figure Description

[0005] Figure 1 This is a flowchart of the steps of an unmanned surface vessel (USV) swarm packet loss detection method based on adaptive fuzzy set-member filtering according to the present invention. Figure 2 This is a structural block diagram of an unmanned surface vessel (USV) swarm packet loss detection system based on adaptive fuzzy set-membership filtering according to the present invention. Detailed Implementation

[0006] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0007] It should be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described herein can be combined with each other.

[0008] It should be understood that the terms "system," "apparatus," "unit," and / or "module" used in this application are a method of distinguishing different components, elements, parts, sections, or assemblies at different levels. However, if other terms can achieve the same purpose, they may be replaced by other expressions.

[0009] As indicated in this application and claims, unless the context clearly indicates otherwise, the words "a," "an," "a," and / or "the" are not specifically singular and may include the plural. Generally, the terms "comprising" and "including" only indicate the inclusion of expressly identified steps and elements, which do not constitute an exclusive list, and the method or apparatus may also include other steps or elements. An element defined by the phrase "comprising an..." does not exclude the presence of other identical elements in the process, method, product, or apparatus that includes the element.

[0010] In the description of the embodiments of this application, "a plurality of" refers to two or more. The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature.

[0011] Furthermore, flowcharts are used in this application to illustrate the operations performed by the system according to embodiments of this application. It should be understood that the preceding or following operations are not necessarily performed precisely in sequence. Instead, the steps can be processed in reverse order or simultaneously. Additionally, other operations can be added to these processes, or one or more steps can be removed from them.

[0012] Reference Figure 1 This is a flowchart illustrating an optional example of the packet loss detection method for unmanned surface vessel formations based on adaptive fuzzy membership filtering proposed in this invention. This method can be applied to computer equipment, and the detection method proposed in this embodiment may include, but is not limited to, the following steps: Step S1: Define the state equations of the leader unmanned surface vessel (USV) and the follower USV, and linearize them to obtain the transformed state equations of the USV formation. Step S2: Based on the transformed state equation of the unmanned surface vessel formation, introduce a measurement noise boundary scaling factor to perform adaptive noise adjustment; Step S3: Receive information from other unmanned surface vessels (USVs), perform packet loss detection by combining scalar and noise data, obtain the detection results, and update the state of the USVs. Step S4: Based on the detection results, discard abnormal data and update the status of the unmanned surface vessel.

[0013] In some feasible embodiments, before step S1, a system model is defined: Unmanned surface vessel model: The kinematic and dynamic model of the unmanned surface vessel can be given by the following state equations: (1) in The coordinates and yaw angle of the unmanned surface vessel are shown. This represents the velocity vector and yaw rate of the unmanned surface vessel. This indicates external noise. This indicates the controller input. It is the inertial matrix composed of the mass of the unmanned surface vessel and the added mass of the hull. It is the rotation matrix between the coordinates of the unmanned surface vessel and the geodetic coordinates, and , and These are the Coriolis centripetal force matrix and the water surface damping matrix, respectively. The specific forms of these matrices are shown below: in , , This represents the inertial parameters of the unmanned surface vessel's own weight and various added masses. , , This represents the hydrodynamic damping coefficient.

[0014] Model discretization: Assume the system sampling time is The model of equation (1) can be transformed into the following form through discrete sampling. (2) in, , , , , ; In some feasible embodiments, step S1 specifically includes: For a leader-follower unmanned surface vessel (USV) swarm, where one USV holds a leadership position and coordinates and controls the other USVs to perform target activities, the state equation of the leader USV is as follows: (3) in, Indicates that the leader's unmanned surface vessel is The state at any given moment, It is a nonlinear function. The leader UAV typically acts as a command generator, providing a reference state for the followers. We can assume that the leader UAV's motion is unaffected by process noise and is deterministic for the follower UAVs, as they receive information from the leader UAV.

[0015] Considering equation (2), the state equation of the follower unmanned surface vessel can be written in the following form: (4) in, Indicates the first A follower drone The state at any given moment, It is a non-linear function. , , This represents the corresponding system matrix. Indicates the first A follower drone Time-based control input, Indicates the first A follower drone System noise at any given moment Indicates the first A follower drone Sensor measurements at any given time Indicates the first A follower drone The state at any given moment, Indicates the first A follower drone Measurement noise at time, , Indicates the serial number of the follower unmanned surface vessel. This represents the number of follower unmanned surface vessels (USVs) in a USV swarm.

[0016] The nonlinear function in the state update equations for the leader and followers is expressed as: That is, the system state is now affected by the nonlinear terms in this nonlinear function, which mainly include three categories: the first being... and The attitude transformation term is used to describe the mapping relationship between the velocity in the hull coordinate system and the rate of change of position in the inertial coordinate system; the second is... , and The constant velocity product term is used to describe the inertial coupling effect between the unmanned surface vessel's pitch, sway, and yaw; the third is... , and The second-order damping term is used to describe the suppressive effect of hydrodynamic nonlinear damping on longitudinal, lateral, and yaw motions.

[0017] Linearization using the TS fuzzy model: Nonlinear terms exist in equations (3) and (4), respectively. and Therefore, we need to linearize the unmanned surface vessel (USV) before performing state estimation. Common linearization methods, such as Taylor expansion, require calculating the gradient matrix of the nonlinear terms, which is computationally intensive and complex. Therefore, we use fuzzy control... The rules are linearized, making this method more robust to noise, easier to implement, less computationally intensive, and easier to use. For simplicity, Gaussian noise in the communication channels between unmanned surface vessels is not considered.

[0018] : yes , yes ... yes ; in For the number of rules, It is a fuzzy set. This refers to the antecedent variable in the fuzzy rule, which is the nonlinear variable in the unmanned surface vessel's state equation. , , and It is a system matrix of appropriate dimensions. Therefore, the state equation of the leader-follower unmanned surface vessel formation can be transformed into the following linear form: (5) in It is the normalized weight of each fuzzy rule. and ,in Antecedent variable The corresponding fuzzy set. It is the error after fuzzy linearization and satisfies: in and It is a known matrix. Unknown but bounded, and .

[0019] Therefore, the nonlinear terms in (3) and (4) can be linearized using the above fuzzy rules to approximate the original equations.

[0020] In some feasible embodiments, step S2 specifically includes: Since unmanned surface vessel (USV) formations are subject to external noise interference when performing tasks on the water surface, and since the statistical characteristics of most of this noise are unavailable, we use quorum estimation for filtering. In quorum estimation, system noise is considered unknown but bounded; therefore, we assume that in equation (4)... and In the following ellipsoidal set.

[0021] Due to process noise Measurement noise The actual amplitude varies with factors such as sea state and speed. Using a fixed ellipsoid boundary would lead to overly conservative filtering results or inconsistencies. To address this, this paper introduces an online learning mechanism for the noise boundary, enabling the shape matrix of the noise ellipsoid to adaptively shrink or expand based on real-time observation data. This allows for a tighter error envelope while maintaining the bounded noise assumption.

[0022] To facilitate online learning, the noise ellipsoid boundary is represented in the form of "reference ellipsoid + scaling factor": in , which is a pre-defined reference noise boundary. This is the scaling factor obtained from online learning. The larger the scaling factor, the looser the noise boundary; the smaller the scaling factor, the tighter the noise boundary.

[0023] At any moment The measurement residual (new information) is defined as follows: .in For the predicted state. To characterize whether the residuals are covered by the current noise ellipsoid boundary, a normalized residual index is introduced: .when At this point, it is considered that the current measurement error amplitude exceeds the reference limit, and the ellipsoid needs to be appropriately enlarged. This indicates that the noise boundary is too conservative and can be gradually reduced. Based on this, the scaling factor for the measured noise boundary adopts the following adaptive update rule: Thus, the noise boundary of the time-varying process is obtained: .in, Indicates the scaling factor for the measurement noise boundary. This represents the scaling factor for the maximum system noise boundary. Represents the scaling factor for the system noise boundary. This represents the normalized residual index. , This represents the threshold constant for adaptive adjustment.

[0024] Unmanned surface vessel status update: Leader-follower system state prediction: Based on the above fuzzy rules, a fuzzified one-step prediction equation similar to equation (5) can be obtained: (6) in, Indicates the first A follower drone Predicted value at time, It is the normalized weight of each fuzzy rule. It is a matrix of appropriate dimensions. Indicates the first A follower drone State estimate at time 10:00 Indicates the first A follower drone Control signals at specific times. The system's control is based on the leader-follower control protocol. As shown below: (7) in It is a constant matrix that we need to design. Represents the first A group of unmanned surface vessels (USVs) are adjacent to other USVs, and when the USV... Unmanned surface vessels can be received. When receiving information, ,otherwise Because the entire follower unmanned surface vessel system is controlled by the leader unmanned surface vessel, therefore, when Unmanned surface vessel that follows at all times Upon receiving information from the leader's unmanned surface vessel, ,otherwise .

[0025] Leader-follower system status updates: The system update equation is: (8) in, Indicates the first A follower drone State estimate at time 10:00 Indicates preset parameters; Indicates the first A follower drone The measured value at time, Indicates the first A follower drone Predicted measurements at any given time.

[0026] In some feasible embodiments, step S3 specifically includes: When unmanned surface vessels (USVs) are performing missions, the followers need to receive information from the leader and other followers. If packet loss occurs in this process, the followers may directly treat the noise in the channel as information from the leader and other followers and update their status. This may cause the USVs to malfunction. At best, it will reduce the performance of the USVs and make it unable to complete the mission. At worst, it may cause the USVs to break away from the formation or even collide with other USVs.

[0027] To address this problem, this embodiment proposes a method of adding a scalar channel to help followers determine whether the received information is normal. For ease of representation, the information received by the follower will be uniformly represented as... After adding a scalar channel, Gaussian noise exists in the channel: The actual information received is ,in This is a scalar added to the additional channel. If packet loss occurs, only noise remains in the channel; otherwise, it is... The sum of the noise term. If no packet loss occurs, the scalar channel satisfies... By using the likelihood function, we can distinguish which distribution the received signal comes from, thus determining whether packet loss has occurred. When the detection rate is sufficient, the two distributions can be effectively distinguished, thus achieving a high detection rate. This prevents unmanned surface vessels (USVs) from mistaking noise in the channel for data from other USVs.

[0028] In some feasible embodiments, step S4 specifically includes: When the unmanned surface vessel is detected by the above attack detection method... To mitigate the impact of an attack on subsequent state estimations and ensure the continued normal operation of the unmanned surface vessel (USV), it discards data deemed abnormal after receiving information from the leader and other followers, i.e., it considers the data otherwise. , This mitigates the impact of packet loss on subsequent system state estimation, thereby ensuring the normal operation of the unmanned surface vessel.

[0029] like Figure 2 As shown, an unmanned surface vessel (USV) swarm packet loss detection system based on adaptive fuzzy membership filtering includes: A state equation construction module is used to execute step S1; The status update module is used to execute step S2; The packet loss detection module is used to execute step S3; The data discard module is used to execute step S4.

[0030] The content of the above method embodiments is applicable to this system embodiment. The specific functions implemented in this system embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments.

[0031] A packet loss detection device for unmanned surface vessel formations based on adaptive fuzzy membership filtering: At least one processor; At least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements the packet loss detection method for unmanned surface vessel formation based on adaptive fuzzy set-membership filtering as described above.

[0032] The content of the above method embodiments is applicable to the device embodiments. The specific functions implemented by the device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0033] A storage medium storing processor-executable instructions, which, when executed by a processor, are used to implement the packet loss detection method for unmanned surface vessel formations based on adaptive fuzzy set-membership filtering as described above.

[0034] The content of the above method embodiments is applicable to this storage medium embodiment. The specific functions implemented in this storage medium embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments.

[0035] The above is a detailed description of the preferred embodiments of the present invention. However, the present invention is not limited to the embodiments described. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the present invention. All such equivalent modifications or substitutions are included within the scope defined by the claims of this application.

Claims

1. A method for packet loss detection in unmanned surface vessel (USV) formations based on adaptive fuzzy membership filtering, characterized in that, Includes the following steps: Define the state equations for the leader UAV and the follower UAVs, and linearize them to obtain the transformed state equations for the UAV formation. Based on the transformed state equation of the unmanned surface vessel formation, a measurement noise boundary scaling factor is introduced for adaptive noise adjustment. It receives information from other unmanned surface vessels (USVs), performs packet loss detection by combining scalar and noise data, obtains the detection results, and updates the status of the USVs.

2. The method for packet loss detection in unmanned surface vessel formations based on adaptive fuzzy membership filtering according to claim 1, characterized in that, Also includes: Based on the detection results, discard any abnormal data.

3. The method for packet loss detection in unmanned surface vessel formations based on adaptive fuzzy membership filtering according to claim 1, characterized in that: The state equation of the leader unmanned surface vessel is as follows: in, Indicates that the leader's unmanned surface vessel is The state at any given moment, Indicates that the leader is Time-based control input, It is a non-linear function; The state equations of the follower unmanned surface vessel are as follows: in, Indicates the first A follower drone The state at any given moment, It is a non-linear function. , , This represents the corresponding system matrix. Indicates the first A follower drone Time-based control input, Indicates the first A follower drone System noise at any given moment Indicates the first A follower drone Sensor measurements at any given time Indicates the first A follower drone The state at any given moment, Indicates the first A follower drone Measurement noise at time, , Indicates the serial number of the follower unmanned surface vessel. This represents the number of follower unmanned surface vessels (USVs) in a USV swarm.

4. The method for packet loss detection in unmanned surface vessel formations based on adaptive fuzzy membership filtering according to claim 2, characterized in that, The state equations of the transformed unmanned surface vessel formation are expressed as follows: in, Indicates that the leader's unmanned surface vessel is The state at any given moment, Represents the number of rules. This represents the maximum number of rules. This represents the normalization weight of each fuzzy rule. Indicates that the leader's unmanned surface vessel is The state at any given moment, This represents the error after fuzzy linearization; This represents the corresponding system matrix; Indicates the first A follower drone Time-based control input; Indicates the first A follower drone System noise at any given moment.

5. The method for packet loss detection in unmanned surface vessel formations based on adaptive fuzzy membership filtering according to claim 4, characterized in that, The scaling factor for the measurement noise boundary is expressed as follows: in, Indicates the scaling factor for the measurement noise boundary. This represents the scaling factor for the maximum system noise boundary. This represents the scaling factor for the system noise boundary. Represents the normalized residual index. , This represents the threshold constant for adaptive adjustment.

6. The method for packet loss detection in unmanned surface vessel formations based on adaptive fuzzy membership filtering according to claim 5, characterized in that, The update equation for the step of updating the state of the unmanned surface vessel is as follows: in, Indicates the first A follower drone State estimate at time 10:00 Indicates preset parameters; Indicates the first A follower drone The measured value at time, Indicates the first A follower drone Predicted measurements at any given time.

7. The method for packet loss detection in unmanned surface vessel formations based on adaptive fuzzy membership filtering according to claim 6, characterized in that, The step of combining scalar and noise to perform packet loss detection and obtain the detection result specifically includes: Based on the received information, the data within the scalar channel is analyzed; If the data in the scalar channel includes a preset scalar and noise term, it is determined that no packet loss has occurred, and the detection result is obtained.

8. A packet loss detection system for unmanned surface vessel formations based on adaptive fuzzy membership filtering, characterized in that, include: The state equation construction module is used to define the state equations of the leader UAV and the follower UAVs, and linearize them to obtain the transformed state equations of the UAV formation. The state update module, based on the transformed state equation of the unmanned surface vessel formation, introduces a measurement noise boundary scaling factor for adaptive noise adjustment; The packet loss detection module receives information from other unmanned surface vessels (USVs), combines scalar and noise data to perform packet loss detection, obtains the detection results, and updates the status of the USVs.

9. A packet loss detection device for unmanned surface vessel formations based on adaptive fuzzy membership filtering, characterized in that, include: At least one processor; At least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements the packet loss detection method for unmanned surface vessel formation based on adaptive fuzzy set-membership filtering as described in any one of claims 1-7.

Images