Unmanned vehicle cluster fixed time distributed cooperative positioning method under positioning failure

Abstract

The present application relates to the technical field of unmanned vehicle, especially to a positioning fault under unmanned vehicle cluster fixed time distributed cooperative positioning method. The present application firstly combines the kinematic model of unmanned vehicle to construct the unmanned vehicle pose estimation model, then substitutes the linear velocity and angular velocity of the unmanned vehicle measured by the sensor into the unmanned vehicle pose estimation model to deduce the pose estimation of each unmanned vehicle. The present application designs a complete distributed positioning algorithm framework through the relative pose information obtained by the communication between unmanned vehicles and the vehicle-mounted sensor measurement, effectively avoids the single point failure risk of centralized architecture, and proposes a robust distributed positioning algorithm for the measurement deviation problem caused by the decline of vehicle-mounted sensor accuracy or environmental interference, which can effectively suppress the sensor measurement error.

Description

Technical Field

[0001] This invention relates to the field of unmanned vehicle technology, and in particular to a fixed-time distributed cooperative localization method for unmanned vehicle clusters under localization failure. Background Technology

[0002] With the deep integration of artificial intelligence and automation technologies, multi-vehicle systems performing distributed tasks (such as cooperative detection, formation control, and target tracking) in dynamic environments has become a key paradigm. Traditional positioning methods rely on external infrastructure (such as GPS and UWB), but have significant limitations: GPS signals are easily blocked indoors, underground, or in dense urban areas, rendering traditional positioning methods that rely on external infrastructure largely ineffective in complex scenarios such as indoors, underground, or in dense cities; UWB deployment is costly and difficult to adapt to dynamic environmental changes. In addition, centralized information processing architectures face the risks of single points of failure and communication bandwidth bottlenecks.

[0003] In recent years, distributed cooperative positioning technology has provided a positioning solution that does not require global information by integrating relative measurements in unmanned workshops, local environmental perception data, and multi-sensor filtering and fusion, which has alleviated the failure risk of centralized architecture to some extent. However, existing methods mostly ignore the impact of sensor measurement errors (such as noise, bias, or drift), and their convergence speed is mostly asymptotically stable, making it difficult to meet the real-time requirements of highly dynamic scenarios.

[0004] Fixed-time control strategies have advantages such as preset convergence time and strong anti-interference ability, but their application in multi-vehicle positioning is still insufficient, especially lacking targeted fault-tolerant mechanisms for common faults (such as positioning sensor and measurement sensor faults). Summary of the Invention

[0005] To overcome the lack of fault-tolerant mechanisms for sensor failures in autonomous vehicle cluster cooperative localization in existing technologies, this invention proposes a fixed-time distributed cooperative localization method for autonomous vehicle clusters under localization failures. This method can achieve accurate pose estimation within a fixed time period through unmanned vehicle communication and on-site sensor measurements, and effectively address the challenges posed by measurement errors.

[0006] This invention proposes a fixed-time distributed cooperative localization method for unmanned vehicle clusters under localization failure. First, it constructs an unmanned vehicle pose estimation model by combining the kinematic model of the unmanned vehicle. Then, it substitutes the linear velocity and angular velocity of the unmanned vehicles measured by sensors into the pose estimation model to derive the pose estimation of each unmanned vehicle. );

[0007] The autonomous vehicle pose estimation model is as follows:

[0008]

[0009] in, , and These are the estimated values ​​of the x-coordinate, y-coordinate, and orientation of the autonomous vehicle i. , and They are respectively , and The first derivative; and These are the linear velocity and angular velocity measurements of the unmanned vehicle i, respectively. sgn is the symbolic function, and g is the parameter.

[0010] , and The set coefficient; The weighted cumulative error of driverless vehicle i in the x-direction is calculated based on the relative position measurements between driverless vehicles and between driverless vehicle and landmark. The weighted cumulative error of driverless vehicle i in the y direction is calculated based on the relative position measurements between driverless vehicles and between driverless vehicle and landmark. The weighted cumulative error for autonomous vehicle i in orientation is calculated based on the angular differences between autonomous vehicles and between the autonomous vehicle and the landmark.

[0011] Preferably, a communication weight is introduced into the weighted cumulative error, and the communication weight is set as follows:

[0012] Based on the assumption that the system communication topology is strongly connected, an adjacency matrix A is constructed to represent the communication weights between autonomous vehicles; a landmark association matrix is ​​also constructed. , The communication weights between the autonomous vehicle i and the landmark k are given. This indicates that driverless car i can measure landmark k; n and m are the number of driverless cars and the number of landmarks, respectively.

[0013] Based on the following constraints, the adjacency matrix A and the landmark association matrix B are solved to obtain the communication weights between autonomous vehicles and between autonomous vehicles and landmarks;

[0014] The constraints include:

[0015] ;

[0016] It is symmetric and positive definite;

[0017] Where L is the Laplacian matrix corresponding to the adjacency matrix A, and P is the detail balance weight matrix constructed based on the assumption that the system communication topology is strongly connected and has a detail balance graph. , ; This represents the communication weight between vehicle i and j. This represents the communication weight between vehicle j and i. , , and Let be the weights of the autonomous vehicles 1, i, j, and n in P, respectively, where 1 ≤ i ≤ n and 1 ≤ j ≤ n.

[0018] Preferably, the coefficients satisfy the following constraints:

[0019]

[0020] in, The maximum error constant of the sensor used; For transition coefficients, .

[0021] Preferably, the fixed time required for the pose estimation error of each autonomous vehicle in the autonomous vehicle cluster to converge is:

[0022]

[0023] Where T is a fixed time. To set a constant, , and All are coefficients.

[0024] Preferred, ; For matrix The smallest eigenvalue.

[0025] Preferred: n is the number of driverless cars. For matrix The smallest eigenvalue.

[0026] Preferably, the convergence range of the pose estimation error of the obtained unmanned vehicle is:

[0027]

[0028]

[0029]

[0030]

[0031]

[0032] Where P is the detail balance weight matrix, L is the Laplacian matrix, and B is the landmark association moment; e, f, and g represent the estimation error vectors of n autonomous vehicles in the x, y, and orientation directions, respectively, and V1, V2, and V3 are the transition functions. and For coefficients; To set a constant, .

[0033] Preferably, the kinematic model of the i-th unmanned vehicle is:

[0034]

[0035] in, Represents a swarm of autonomous vehicles; Let x, y, and orientation be the first derivatives of the x-coordinate, y-coordinate, and orientation of the autonomous vehicle i, respectively. , and These represent the linear velocity, angular velocity, and orientation of the unmanned vehicle i.

[0036] The present invention proposes a system comprising a memory and a processor. The memory stores a computer program, and the processor is connected to the memory. The processor is used to execute the computer program to implement the fixed-time distributed cooperative positioning method for unmanned vehicle clusters under positioning failure.

[0037] The present invention proposes a storage medium storing a computer program, which, when executed, is used to implement the fixed-time distributed cooperative positioning method for unmanned vehicle clusters under positioning failure.

[0038] The advantages of this invention are:

[0039] (1) This invention aims to solve the positioning challenges caused by the lack of global information in dynamic environments (such as GPS positioning sensor failure, signal blockage or failure). To this end, this invention designs a complete distributed positioning algorithm framework by using the relative pose information obtained by unmanned workshop communication and vehicle-mounted sensor measurement, which effectively avoids the single point of failure risk of centralized architecture. In addition, for the measurement deviation problem caused by the decrease in the accuracy of vehicle-mounted sensors or environmental interference, a robust distributed positioning algorithm is proposed, which can effectively suppress sensor measurement errors.

[0040] (2) The distributed architecture proposed in this invention based on unmanned workshop communication does not require global information, thus avoiding the impact of single point of failure; it reduces the dependence on global information and enhances the practicality and scalability of the system. In practical application scenarios such as industrial automation, intelligent warehousing, and disaster relief, this method can significantly improve the positioning accuracy, collaborative efficiency, and operational reliability of unmanned vehicle clusters, providing important technical support for multi-unmanned vehicle collaborative operations in complex environments.

[0041] (3) The present invention constructs communication weight constraints based on the system communication topology, which further improves the robustness of the algorithm and ensures that the positioning error converges to a bounded region near the origin within a fixed time. It further solves the problem of pose estimation of multiple unmanned vehicles under the absence of global information and measurement errors, and improves the positioning accuracy and convergence speed.

[0042] (4) In the presence of sensor measurement errors, the robust algorithm can effectively suppress the influence of errors, ensuring that the positioning error is stable within a bounded area, and the range of the area can be adjusted by adjusting the coefficient. , and To reduce the size and enhance system reliability. (Coefficient) , and Adjustable, allowing for performance optimization in different application scenarios, such as by increasing... Or adjust To balance convergence speed and accuracy.

[0043] (5) The present invention adopts the fixed time control theory to ensure that the pose estimation converges within a predetermined time. It can accurately estimate the global pose of each unmanned vehicle within a fixed time. The convergence time can be calculated in advance and is independent of the initial conditions. It breaks through the limitations of the traditional progressive convergence method, improves the system response speed, and has the advantages of fast convergence speed, strong anti-interference ability, and wide applicability.

[0044] (6) The system and memory proposed in this invention serve as the carrier of the method of this invention, which greatly facilitates the promotion of the method. Attached Figure Description

[0045] Figure 1 This is a flowchart of the method of the present invention;

[0046] Figure 2 This is a schematic diagram illustrating the principle of the distributed cooperative positioning system for multiple unmanned vehicles in an embodiment of the present invention.

[0047] Figure 3(a) shows the desired pursuit formation of the unmanned vehicle swarm in an embodiment of the present invention;

[0048] Figure 3(b) is a communication topology diagram of the unmanned vehicle cluster in an embodiment of the present invention;

[0049] Figure 4(a) is a comparison curve of the x-direction coordinate estimation of the unmanned vehicle 1 under the two estimation models in the embodiment of the present invention;

[0050] Figure 4(b) is a comparison curve of the y-direction coordinate estimation of the unmanned vehicle 1 under the two estimation models in the embodiment of the present invention;

[0051] Figure 4(c) is a comparison curve of the orientation estimation of the unmanned vehicle 1 under the two estimation models in the embodiment of the present invention;

[0052] Figure 5 This is a target pursuit trajectory diagram of the unmanned vehicle swarm in an example of the present invention;

[0053] Figure 6(a) is a schematic diagram of convergence in the x-direction using the method of the present invention in an embodiment of the present invention;

[0054] Figure 6(b) is a schematic diagram of y-direction convergence using the method of the present invention in an embodiment of the present invention;

[0055] Figure 6(c) is a schematic diagram of the orientation convergence using the method of the present invention in an embodiment of the present invention. Detailed Implementation

[0056] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0057] A fixed-time distributed collaborative localization method for unmanned vehicle clusters under localization failure is used to estimate the pose of each unmanned vehicle, i.e., the position and orientation of the unmanned vehicle, by combining the mutual communication between the unmanned vehicle cluster and the landmark when the localization function of the unmanned vehicle fails.

[0058] In this embodiment, let the pose of the i-th unmanned vehicle in the cluster be denoted as . Where (x,y) represents the location. Indicates direction; subscript i indicates the autonomous vehicle's identity; that is: ( Let be the location coordinates of the i-th unmanned vehicle. Let be the orientation of the i-th autonomous vehicle. The linear velocity and angular velocity of the i-th autonomous vehicle are denoted as [formula missing]. and Both can be measured using sensors inside the autonomous vehicle (such as motor encoders). In this application, all distance units are meters, and all angle units are radians (rad).

[0059] The kinematic model of the i-th autonomous vehicle is:

[0060] (1)

[0061] in, Represents a swarm of autonomous vehicles; They are respectively The first derivative.

[0062] The relative pose between autonomous vehicle i and j is defined as follows:

[0063] (2)

[0064] in, Let j be the pose of autonomous vehicle j in the autonomous vehicle swarm; Let be the set of neighbors of driverless car i, that is, the set of driverless cars other than driverless car i in the driverless car cluster. This represents the relative distance between driverless vehicles i and j in the x-direction; This represents the relative distance between driverless vehicles i and j in the y-direction; This represents the angular difference in orientation between driverless vehicles i and j.

[0065] The relative pose between the autonomous vehicle i and the landmark k is defined as follows:

[0066] (3)

[0067] in, This represents the relative distance between the autonomous vehicle i and the landmark k in the x-direction; This represents the relative distance between the autonomous vehicle i and the landmark k in the y-direction; This represents the angular difference in orientation between the autonomous vehicle i and the landmark k; ) represents the pose of a known landmark k; that is, ( () represents the location coordinates of landmark k. Let k be the orientation of the landmark. The pose of the landmark is known. Specifically, landmarks can be divided into static landmarks and dynamic landmarks. The pose of static landmarks is a set value, while the pose of dynamic landmarks is uploaded in real time.

[0068] Based on the principle of bounded error, the definition is as follows:

[0069] (4)

[0070] in, and Let represent the linear velocity and angular velocity measurements of the i-th unmanned vehicle, respectively. and These represent the linear velocity and angular velocity of the i-th unmanned vehicle, respectively, corresponding to their actual values; Let be the speed measurement error of the i-th unmanned vehicle, i.e. Because sensor measurements have inherent errors, and In principle, it is unknown; but It is bounded.

[0071] and Let i and j represent the measured and actual relative distances of autonomous vehicles i and j in the x-direction, respectively. and These represent the measured and actual relative distances between driverless vehicles i and j in the y-direction, respectively. and These represent the measured and actual angle difference values ​​for autonomous vehicles i and j in terms of orientation, respectively. Let be the relative pose measurement error between vehicles i and j, i.e.: .

[0072] and Let i and k represent the measured and actual relative distances between the autonomous vehicle i and the landmark k in the x-direction, respectively. and Let i and k represent the measured and actual relative distances between the autonomous vehicle i and the landmark k in the y-direction, respectively. and These represent the measured and actual angular difference between the autonomous vehicle i and the landmark k in terms of orientation, respectively. Let be the relative pose measurement error between the unmanned vehicle i and the landmark k, i.e.:

[0073] .

[0074] Relative position and relative orientation can be detected using a vision platform, either mounted on the autonomous vehicle or other vision platforms set up on-site. , , , , and Provided by a visual inspection platform. , , , , and It is unknown in principle.

[0075] Introducing the maximum error constant It is the known measurement accuracy of autonomous vehicle sensors, namely:

[0076]

[0077] Then we have: , , .

[0078] In this embodiment, based on the principle of bounded error, a robust algorithm is used to construct the autonomous vehicle pose estimation model in the cluster as follows:

[0079] (5)

[0080] in, , and They are respectively , and The corresponding estimated value, i.e. ( ) represents the pose estimation of driverless vehicle i; , and They are respectively , and The first derivative; and , respectively, are the linear velocity and angular velocity measurements of the unmanned vehicle i; sig is a function, sgn is the sign function;

[0081] , and The coefficients are set; the three satisfy the following constraints:

[0082] (6)

[0083] Where n and m are the number of driverless cars and the number of landmarks, respectively. and This is the transition coefficient.

[0084] , and These are the weighted cumulative errors of driverless vehicle i in the x-direction, y-direction, and orientation, respectively.

[0085] (7)

[0086] in, and Let represent the relative distance measurements between driverless vehicles i and j in the x and y directions, respectively. This indicates the angular deviation in orientation between driverless vehicles i and j; and Let represent the relative distance measurements between the autonomous vehicle i and the landmark k in the x and y directions, respectively. This indicates the angular deviation in orientation between the autonomous vehicle i and the landmark k; For the neighbor set of driverless car i, A collection of landmarks; The communication weights for vehicles i and j are determined based on the system communication topology. The communication weights between autonomous vehicle i and landmark k are determined based on the degree of association between the landmark and the autonomous vehicle. For pose estimation of autonomous vehicle j, The position of landmark k.

[0087] Definition: Adjacency matrix Indicates the connection weight of the unmanned factory. Let i be the communication weights for autonomous vehicles i and j;

[0088] Laplace matrix ,in It is a degree matrix;

[0089] Landmark Association Matrix , Indicates driverless cars to landmarks Measurement relationship, This indicates that driverless car i can measure landmark k; in a driverless car swarm system, at least one driverless car satisfies This means that at least one unmanned vehicle has the ability to measure landmarks; diag represents a diagonal matrix.

[0090] In practical implementation, the system communication topology can be assumed to be a strongly connected and detail-balanced graph. Strong connectivity requires that there is a path (whether unidirectional or bidirectional) between any two autonomous vehicle nodes in the communication graph, ensuring that information can be disseminated throughout the entire system through multi-hop transmission, avoiding information silos or system segmentation due to local link failures. Detail balance refers to the existence of a set of positive weights. (usually meets) ), such that the communication weight matrix satisfies (in This represents the communication weight between vehicle i and j. (representing the communication weights from vehicle j to i), this symmetry guarantees the Laplace matrix. With weight matrix The matrix formed after combination It is symmetric and positive definite. , , and Let be the weights of the autonomous vehicles 1, i, j, and n in P, respectively, where 1 ≤ i ≤ n and 1 ≤ j ≤ n.

[0091] Thus, it can be determined according to the following constraints: "and" It is a symmetric positive definite method that calculates the adjacency matrix A and the landmark association matrix B to obtain the communication weights between autonomous vehicles. And the communication weight between autonomous vehicles and landmarks .

[0092] Thus, given the adjacency matrix A, the landmark association matrix B, and the poses of each landmark, substituting these into formulas (5)-(7) allows for the calculation of the pose estimates for each unmanned vehicle. ).

[0093] As can be seen from formula (6), in this embodiment, the maximum error constant is... Given the known measurement accuracy of autonomous vehicle sensors, It is a constant.

[0094] Using this method, the fixed time required for the pose estimation error of autonomous vehicles in a swarm to converge is:

[0095] (8)

[0096] (9)

[0097] (10)

[0098]

[0099] Where T is a fixed time. and All are coefficients. To set a constant, For matrix The smallest eigenvalue.

[0100] The pose estimation obtained by this method converges to the following error:

[0101] (11)

[0102] (12)

[0103] in, The coefficients are: e, f, and g, which represent the estimation error vectors of n unmanned vehicles in the x-direction, y-direction, and orientation directions, respectively. i and e n Let represent the estimation errors of autonomous vehicles 1, i, and n in the x-axis direction, respectively. f 1. f i and f n The estimation errors of unmanned vehicles 1, i, and n in the y-axis direction, g1 and g2, respectively. i and g n Let $\mathbf$ and $\mathbf$ represent the estimation errors of autonomous vehicles $1$, $i$, and $n$ in terms of orientation, respectively.

[0104] (13)

[0105] (14)

[0106] (15)

[0107] (16)

[0108] (17)

[0109] V1, V2, and V3 are functions, L is the Laplacian matrix, B is the landmark association matrix, and P is the weight matrix. .

[0110] The following specific embodiments verify the above-mentioned fixed-time distributed cooperative localization method for unmanned vehicle clusters under localization failure (hereinafter referred to as the method of the present invention).

[0111] In this embodiment, the number of autonomous vehicles in the autonomous vehicle cluster is set to 3, and the pose estimation is denoted as... The target being captured is a dynamic landmark, denoted as... .

[0112] In this embodiment, two methods are used to obtain the pose estimation of each unmanned vehicle in the unmanned vehicle cluster, namely the method of this invention and the linear positioning estimation algorithm.

[0113] The autonomous vehicle pose estimation model used in this invention is as follows:

[0114] (18)

[0115] The parameters are set as follows: ;

[0116] The estimation model used in the linear positioning estimation algorithm is:

[0117] (19)

[0118] The parameters are set as follows: . In the above formulas (1) (2), , and These are the weighted cumulative errors of driverless vehicle i in the x-direction, y-direction, and orientation, respectively.

[0119] Relative pose between the autonomous vehicle and the target being pursued:

[0120] (20)

[0121] Therefore, the target acquisition error of each unmanned vehicle can be obtained as follows:

[0122] (twenty one)

[0123] in, The deviation between the unmanned vehicle i in the desired pursuit formation and the target is the distance between the unmanned vehicle i in the x and y directions and the angular deviation in the orientation between the unmanned vehicle i in the desired pursuit formation and the target. Position for tracking the target. For pose estimation of autonomous vehicle i, Let be the error between the pose estimation of autonomous vehicle i and the pose of autonomous vehicle i in the desired pursuit formation.

[0124] The desired pursuit formation and communication topology among the unmanned vehicles are shown in Figures 3(a) and 3(b), where, d represents the expected distance between the autonomous vehicle and the dynamic landmark. The angle represents the distribution of the unmanned vehicles. In the communication topology shown in Figure 3(b), only unmanned vehicle 1 can measure dynamic landmarks, with a communication strength of 3; unmanned vehicle 2 communicates with unmanned vehicles 1 and 3 respectively, with a communication strength of 1 for each.

[0125] Based on the target pursuit error shown in the above formula (21), the desired pursuit formation of the unmanned vehicle can be designed as shown in the following formula (22), and then the command ( , The driverless car is driven to move in pursuit of the target;

[0126] (twenty two)

[0127] in, All are coefficients, and, All are greater than 0. , ; and These are the linear velocity and angular velocity of the autonomous vehicle i, respectively. and These are the linear velocity and angular velocity of the target being pursued, respectively. for The first derivative; , Both g and g refer to parameters.

[0128] In this embodiment, the parameters that can be set for formula (22) are as follows: , , , , , .

[0129] The initial poses of the unmanned vehicle and the target, as well as the linear and angular velocities of the target's motion, are set as follows:

[0130]

[0131] in, and These represent the initial poses of unmanned vehicles 1, 2, and 3, and the target being pursued, respectively.

[0132] In this embodiment, the pursuit trajectory of unmanned vehicle 1 in the Matlab simulation environment is shown by the red lines in Figures 4(a), 4(b), and 4(c); when formulas (18) and (22) are used to drive the unmanned vehicle, the tracking path of unmanned vehicle 1 is shown by the dashed lines (FO) in Figures 4(a), 4(b), and 4(c); when formulas (19) and (22) are used to drive the unmanned vehicle, the tracking path of unmanned vehicle 1 is shown by the lines (LO) in Figures 4(a), 4(b), and 4(c). It can be seen that the method of the present invention has a faster convergence speed than the linear positioning estimation algorithm, and the motion trajectory of unmanned vehicle 1 is more closely aligned with the target being pursued.

[0133] In the Matlab simulation environment, using formulas (18) and (22) to drive the unmanned vehicle, the simulation results of the pursuit trajectory of each unmanned vehicle are as follows: Figure 5 As shown, each unmanned vehicle closely follows and pursues its target.

[0134] Figures 6(a), 6(b), and 6(c) show the pursuit trajectories of each unmanned vehicle obtained by the method of the present invention (using formulas (18) and (22) in combination with the driving unmanned vehicle). Figure 5 (As shown) plus the corresponding expected deviation The degree of consistency between the tracking trajectory and the target's movement trajectory is observed. It can be seen that after the implementation of the method of the present invention, the position tracking of the unmanned vehicle converges in about 5 seconds, the orientation converges in about 2 seconds, and the desired pursuit formation is maintained after convergence.

[0135] from Figures 4(a) to 6(c) As can be seen, under measurement error, the global positioning estimation error of the method of the present invention can converge to the region near the origin within a fixed time. Compared with the linear positioning estimation algorithm (LO), the method of the present invention (FO) has better estimation effect, faster response speed, and the entire unmanned vehicle system can form the desired formation while the pursuit error can be effectively converged.

[0136] Of course, those skilled in the art will recognize that the present invention is not limited to the details of the exemplary embodiments described above, but also includes the same or similar structures that can be implemented in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered illustrative and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0137] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

[0138] The technologies, shapes, and structures not described in detail in this invention are all known technologies.

Claims

1. A fixed-time distributed cooperative localization method for an unmanned vehicle cluster under localization failure, characterized in that, First, an autonomous vehicle pose estimation model is constructed by combining the kinematic model of the autonomous vehicle. Then, the linear velocity and angular velocity of the autonomous vehicle measured by the sensors are substituted into the autonomous vehicle pose estimation model to derive the pose estimation of each autonomous vehicle. ); The autonomous vehicle pose estimation model is as follows: in, , and These are the estimated values ​​of the x-coordinate, y-coordinate, and orientation of the autonomous vehicle i. , and They are respectively , and The first derivative; and These are the linear velocity and angular velocity measurements of the unmanned vehicle i, respectively. sgn is the symbolic function, and g is the parameter. , and The set coefficient; The weighted cumulative error of driverless vehicle i in the x-direction is calculated based on the relative position measurements between driverless vehicles and between driverless vehicle and landmark. The weighted cumulative error of driverless vehicle i in the y direction is calculated based on the relative position measurements between driverless vehicles and between driverless vehicle and landmark. The weighted cumulative error for autonomous vehicle i in orientation is calculated based on the angular differences between autonomous vehicles and between the autonomous vehicle and the landmark. The weighted cumulative error introduces communication weights, which are set as follows: Based on the assumption that the system communication topology is strongly connected, an adjacency matrix A is constructed to represent the communication weights between autonomous vehicles; a landmark association matrix is ​​also constructed. , The communication weights between the autonomous vehicle i and the landmark k are given. This indicates that driverless car i can measure landmark k; n and m are the number of driverless cars and the number of landmarks, respectively. Based on the following constraints, the adjacency matrix A and the landmark association matrix B are solved to obtain the communication weights between autonomous vehicles and between autonomous vehicles and landmarks; The constraints include: ； It is symmetric and positive definite; Where L is the Laplacian matrix corresponding to the adjacency matrix A, and P is the detail balance weight matrix constructed based on the assumption that the system communication topology is strongly connected and has a detail balance graph. , ; This represents the communication weight between vehicle i and j. This represents the communication weight between vehicle j and i. , , and Let be the weights of the autonomous vehicles 1, i, j, and n in P, respectively, where 1 ≤ i ≤ n and 1 ≤ j ≤ n.

2. The fixed-time distributed cooperative localization method for unmanned vehicle clusters under localization failure as described in claim 1, characterized in that, The coefficients satisfy the following constraints: in, The maximum error constant of the sensor used; For transition coefficients, .

3. The fixed-time distributed cooperative localization method for unmanned vehicle clusters under localization failure as described in claim 2, characterized in that, The fixed time required for the pose estimation error of each autonomous vehicle in a swarm to converge is: Where T is a fixed time. To set a constant, , and All are coefficients.

4. The fixed-time distributed cooperative localization method for unmanned vehicle clusters under localization failure as described in claim 3, characterized in that, ; For matrix The smallest eigenvalue.

5. The fixed-time distributed cooperative localization method for unmanned vehicle clusters under localization failure as described in claim 3, characterized in that: n is the number of driverless cars. For matrix The smallest eigenvalue.

6. The fixed-time distributed cooperative localization method for unmanned vehicle clusters under localization failure as described in claim 5, characterized in that, The convergence range of the pose estimation error of the obtained autonomous vehicle is: Where P is the detail balance weight matrix, L is the Laplacian matrix, and B is the landmark association moment; e, f, and g represent the estimation error vectors of n autonomous vehicles in the x, y, and orientation directions, respectively, and V1, V2, and V3 are the transition functions. and For coefficients; To set a constant, .

7. The fixed-time distributed cooperative localization method for unmanned vehicle clusters under localization failure as described in claim 1, characterized in that, The kinematic model of the i-th autonomous vehicle is: in, Represents a swarm of autonomous vehicles; Let x, y, and orientation be the first derivatives of the x-coordinate, y-coordinate, and orientation of the autonomous vehicle i, respectively. , and These represent the linear velocity, angular velocity, and orientation of the unmanned vehicle i.

8. A fixed-time distributed cooperative positioning system for an unmanned vehicle cluster under positioning failure, characterized in that, It includes a memory and a processor. The memory stores a computer program, and the processor is connected to the memory. The processor is used to execute the computer program to implement the fixed-time distributed cooperative positioning method for unmanned vehicle clusters under positioning failure as described in any one of claims 1-7.

9. A storage medium, characterized in that, The system contains a computer program that, when executed, implements the fixed-time distributed cooperative localization method for unmanned vehicle clusters under localization failure as described in any one of claims 1-7.

Images