Unmanned aerial vehicle cluster collaborative navigation host distribution optimization method based on genetic algorithm

Abstract

The invention provides an unmanned aerial vehicle cluster collaborative navigation host distribution optimization method based on a genetic algorithm, and the method comprises the steps: an unmanned aerial vehicle cluster comprises not less than three hosts and not less than one slave, the hosts employ a satellite receiver to obtain the position information of the hosts, and the slave employs the position of the hosts as a reference; calculating own position information by using the relative distance between the host and the slave provided by the data link; the Fisher information amount is used for representing the navigation positioning precision of the slave machines, and the total Fisher information amount of all the slave machines is used as a to-be-optimized target function; and recording the positions of all hosts as a host position sequence, iteratively optimizing the host position sequence by using a genetic algorithm, and taking the host position sequence corresponding to the minimum value of the target function as the optimal distribution of the hosts. According to the method, the Fisher information amount is used for representing the positioning precision of a single slave, the total Fisher information amount of all slaves is used as a target function, and the host distribution is iteratively optimized through the genetic algorithm, so that the overall navigation positioning precision of the cluster is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of cooperative navigation of UAV swarms, and particularly relates to an optimization method for the distribution of master nodes in cooperative navigation of UAV swarms based on genetic algorithms. Background Art

[0002] With the booming development of the UAV manufacturing field, small UAVs have gradually taken on important responsibilities in various mission fields. However, single small UAVs have limitations such as small payloads and poor anti-risk capabilities. In response to the increasingly harsh working environments and diverse mission requirements, the working mode of swarm UAVs has gradually become the focus of research.

[0003] Inertial / satellite integrated navigation is a commonly used navigation method for UAVs, and related algorithms are very mature. Among them, the microelectromechanical inertial navigation system (MEMS) has the characteristics of small volume, low price, continuous output, and independence, but its accuracy is relatively low and the error accumulates over time, requiring other navigation information for auxiliary correction; satellite navigation systems have the characteristics of all-weather, high precision, and non-accumulating errors over time, but the cost is relatively high. Considering the overall positioning performance of UAV swarms, a data link ranging system can be introduced to replace part of the satellite navigation system of the airframe to provide auxiliary reference information, and an inertial / satellite / data link integrated navigation scheme can be constructed. By utilizing the high-efficiency short-distance data transmission ability of the data link system, both navigation performance and deployment cost can be taken into account.

[0004] In the cooperative navigation of UAV swarms based on data link ranging, the swarm is divided into two categories: master nodes and slave nodes. A small number of master nodes are equipped with inertial navigation systems, data link systems, and satellite navigation systems, and can obtain their own absolute position information in real time. A large number of slave nodes are only equipped with inertial navigation systems and data link systems, and use the position information of the master nodes and relative ranging information transmitted by the data link as references to assist in their own navigation and positioning.

[0005] In the swarm, the relative position relationship between each master node and each slave node will affect the observability of the position of the slave node. Currently, most related research focuses on the optimal distribution of multiple master nodes relative to a single slave node. Considering that the master nodes need to provide absolute position information references for all slave nodes, how to design the distribution of the master nodes relative to the entire UAV swarm to achieve the optimal global cooperative navigation and positioning performance is an urgent problem to be solved. Summary of the Invention

[0006] The object of the present invention is to provide an optimization method for the distribution of master drones in cooperative navigation of a drone swarm based on a genetic algorithm. The master drones in the swarm use satellite receivers to obtain their own position information. The slave drones in the swarm use the positions of the master drones as references and calculate their own positions using the ranging information of the data link. The Fisher information quantity is used to characterize the positioning accuracy of a single slave drone, and the total Fisher information quantity of all slave drones is used as the objective function. The position sequence of the master drones is iteratively optimized through the genetic algorithm to obtain the optimal distribution of the master drones, so as to achieve the highest overall navigation and positioning accuracy of the swarm.

[0007] The technical solution adopted by the present invention to achieve the above object is as follows:

[0008] The present invention provides an optimization method for the distribution of master drones in cooperative navigation of a drone swarm based on a genetic algorithm, including the following steps:

[0009] The drone swarm includes at least 3 master drones and at least 1 slave drone. The master drones use satellite receivers to obtain their own position information. The slave drones use the positions of the master drones as references and calculate their own position information using the relative distances between the master drones and the slave drones provided by the data link;

[0010] The Fisher information quantity is used to characterize the navigation and positioning accuracy of the slave drones, and the total Fisher information quantity of all slave drones is used as the objective function to be optimized;

[0011] The positions of all the master drones are recorded as the master drone position sequence, and the genetic algorithm is used to iteratively optimize the master drone position sequence. The master drone optimal distribution is the master drone position sequence corresponding to the minimum value of the objective function.

[0012] Further, the method for the slave drones to use the positions of the master drones as references and calculate their own position information using the relative distances between the master drones and the slave drones provided by the data link includes the following steps:

[0013] The slave drone communicates with at least 3 of the master drones to obtain the relative distances between the slave drone and at least 3 of the master drones;

[0014] According to the obtained q relative distances, a first set of equations is constructed:

[0015]

[0016] where X = [x, y, z] is the position of the slave drone to be solved, and [x i , y i , z i represents the position of the i-th master drone; ρ i is the relative distance between the slave drone and the i-th master drone;

[0017] Using the Newton iteration method to solve the first set of equations is equivalent to solving the following linear equations:

[0018] G·ΔX k = b

[0019] where G is the Jacobian matrix formed by taking the partial derivative of f(x, y, z, x i , y i , z i ) with respect to (x, y, z) at X k-1 ; b is the estimated residual matrix formed by ρ i - f(x, y, z, x i , y i , z i ); X k-1 represents the estimated value of the slave aircraft position obtained in the (k - 1)th iteration, and ΔX k is the update amount in the kth iteration;

[0020] Solve for the update amount ΔX using the least squares method k , and update the estimated value of the slave aircraft position X k :

[0021] ΔX k = (G T G) -1 G T b

[0022] X k = X k-1 + ΔX k

[0023] Iterate until the set accuracy, and output the position of the slave aircraft.

[0024] Further, the method for characterizing the navigation and positioning accuracy of the slave aircraft using the Fisher information amount is as follows:

[0025] The Fisher information amount of the jth slave aircraft is expressed as:

[0026]

[0027] where det(FIM) j is the Fisher information amount of the jth slave aircraft, det() represents the determinant, and q is the number of master aircraft communicating with the jth slave aircraft; H i represents the observation matrix of the position of the ith master aircraft relative to the slave aircraft;

[0028] The objective function is:

[0029]

[0030] where n is the number of slave aircraft in the UAV cluster.

[0031] Further, H i == [cosθ i cosα i sinθ i -cosθ i sinα i , where θ i and α i respectively represent the pitch angle and azimuth angle of the i-th master relative to the slave.

[0032] Further, recording the positions of all masters as the master position sequence, and using the genetic algorithm to iteratively optimize the master position sequence, with the minimum value of the objective function corresponding to the master position sequence as the optimal distribution of the masters, specifically including the following steps:

[0033] S001. Randomly generate N groups of initial master position sequences, and each group of master position sequences contains m masters;

[0034] S002. Calculate the objective function and fitness function of each group of master position sequences;

[0035] S003. Keep l groups of master position sequences with the smallest objective function intact for the next iteration;

[0036] S004. Recombine the N - l groups of master position sequences. The probability that the position of the i-th master in each recombined group of master position sequences is selected from the corresponding position of the s-th group of master position sequences is ω k , i = 1, 2, 3,.., m, s = 1, 2, 3,..., N;

[0037] S005. Add random perturbations to the recombined master position sequences;

[0038] S006. Repeat steps S002 - S005 until the number of iterations reaches the set value, and take the master position sequence corresponding to the minimum value of the objective function as the optimal distribution of the masters.

[0039] Further, the fitness function of the s-th group of master position sequences is:

[0040]

[0041] where Loss s and are the objective functions of the s-th and i-th 0 groups of master position sequences.

[0042] Further, the random perturbation is white noise.

[0043] The beneficial effects of the present invention compared with the prior art:

[0044] Aiming at the problem of unclear host distribution strategy in the cooperative navigation method of UAV clusters based on data link ranging, an optimization method for host distribution of UAV cluster cooperative navigation based on genetic algorithm provided by the present invention divides the UAV cluster into two categories: host and slave. Among them, the host uses a satellite receiver to obtain its own high-precision position information, and the slave takes the host position as an absolute reference and calculates its own position using the ranging information between the host and the slave provided by the data link. The Fisher information is used to characterize the navigation and positioning accuracy of a single slave. The larger the Fisher information, the higher the positioning accuracy. And the total Fisher information of all slaves is used as the objective function for optimization. The genetic algorithm is used to iterate the host position sequence to obtain the optimal host distribution, so as to improve the overall navigation performance of the cluster without additional increasing the number of hosts, so as to achieve the highest overall navigation and positioning accuracy of the cluster. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] The accompanying drawings included are used to provide a further understanding of the embodiments of the present invention, which form a part of the specification, are used to illustrate the embodiments of the present invention, and together with the written description are used to explain the principles of the present invention. Obviously, the drawings in the following description are only some embodiments of the present invention, and those of ordinary skill in the art can also obtain other drawings based on these drawings without creative efforts.

[0046] Figure 1 It is a schematic diagram of the host distribution of UAV cluster cooperative navigation provided for the specific embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0047] The specific embodiments of the present invention will be described in detail below. In the following description, for the purpose of explanation rather than limitation, specific details are set forth in order to provide a thorough understanding of the present invention. However, it will be apparent to those skilled in the art that the present invention may be practiced in other embodiments without these specific details.

[0048] It should be noted here that in order to avoid obscuring the present invention with unnecessary details, only the device structures and / or processing steps closely related to the solution of the present invention are shown in the drawings, and other details less related to the present invention are omitted.

[0049] The optimization method for host distribution of UAV cluster cooperative navigation based on genetic algorithm proposed by the present invention mainly aims at the problem of unclear optimal distribution strategy of the host relative to the overall cluster in the UAV cluster cooperative navigation system based on data link ranging when only some hosts are equipped with satellite navigation systems.

[0050] The optimization method for host distribution of UAV cluster cooperative navigation based on genetic algorithm proposed by the present invention mainly includes:

[0051] First, the UAV swarm is divided into two categories: the master UAV and the slave UAV. The master UAV uses a satellite receiver to obtain its own high-precision position information. The slave UAV takes the position of the master UAV as the absolute reference and calculates its own position by using the ranging information between the master and slave UAVs provided by the data link.

[0052] Then, the Fisher information quantity is used to characterize the navigation and positioning accuracy of the slave UAV, which is equivalent to the quality of the spatial distribution of the master UAV relative to a single slave UAV. The total Fisher information quantity of all slave UAVs is used as the objective function to be optimized.

[0053] Finally, the position information of all master UAVs is integrated into a parameter sequence to be optimized. The genetic algorithm is used to iteratively optimize the position sequence of the master UAVs to obtain the optimal distribution of the master UAVs, so as to achieve the highest overall navigation and positioning accuracy of the swarm.

[0054] Next, a specific embodiment is combined to describe the technical solution of the present invention in detail.

[0055] A method for optimizing the distribution of master UAVs in cooperative navigation of UAV swarms based on genetic algorithm provided in this embodiment is as Figure 1 shown, and specifically includes the following steps:

[0056] 1. Cooperative navigation and positioning method

[0057] The UAV swarm is divided into master UAVs i (i = 1, 2, 3,..., m) and slave UAVs j (j = 1, 2, 3,..., n). The master UAVs are equipped with satellite navigation systems and can obtain their own three-dimensional absolute position information [x i , y i , z i . There is data link communication between the master and slave UAVs, and the real-time relative distance ρ ij can be obtained. When a slave UAV can communicate with at least 3 master UAVs, its own position information can be calculated by the following method:

[0058] (1) Denote the position of the slave UAV to be solved as X = [x, y, z], which can communicate with q (q ≤ m) master UAVs, and q is greater than or equal to 3. The q measured distance information forms a system of equations:

[0059]

[0060] In the formula, f(x, y, z, x i , y i , z i ) is an intermediate variable, which characterizes the distance between the position of the slave UAV to be solved and the known position of the master UAV, and ρ i is the relative distance between the slave UAV and the i-th master UAV provided by the data link.

[0061] (2) Solve the three unknowns of the above ternary non - linear equations using the Newton - Raphson method. Denote the update amount at the k - th iteration as ΔX k , which is equivalent to solving the linear equations:

[0062] G·ΔX k =b

[0063] where G is the Jacobian matrix formed by taking the partial derivatives of the slave with respect to (x, y, z, x i , y i , z i ) at X k-1 , with a size of q×3; b is the estimated residual matrix formed by ρ i - f(x, y, z, x i , y i , z i ), with a size of q×1; X k-1 represents the estimated value of the slave position obtained at the (k - 1)-th iteration.

[0064] Solve for the update amount ΔX by least squares k and update the estimated value of the slave position X k :

[0065] ΔX k =(G T G) -1 G T b

[0066] X k =X k-1 +ΔX k

[0067] (3) Repeat the above steps, iterate until the accuracy is sufficient, and output the slave position information.

[0068] 2. Characterization of positioning accuracy

[0069] Use the Fisher information matrix to characterize the positioning accuracy of a single slave. The evaluation function is chosen as det. It is considered that the ranging accuracy σ of the data link is independent of the relative distance, and the measurement error matrix is simplified to the product of a constant and the identity matrix. Ignoring the coefficient related to the data - link ranging accuracy σ, use det(FIM) to characterize the Fisher information provided by a measurement combination, and there is:

[0070]

[0071] where H i represents the observation matrix of the position of the i - th master relative to the slave, and there is:

[0072]

[0073] Where, θ i and α i respectively represent the pitch angle and azimuth angle of the i-th host relative to the slave.

[0074] Define the overall positioning accuracy index of the UAV cluster as the sum of the Fisher information of all slaves, and design the objective function to be optimized accordingly as:

[0075]

[0076] Where, n is the total number of slaves in the cluster.

[0077] 3. Optimization of the host distribution strategy

[0078] Use the genetic algorithm to iteratively optimize the positions of the host sequences in the UAV cluster until the requirements are met:

[0079] (1) Randomly generate N groups of initial host position sequences, each group containing m hosts;

[0080] (2) Calculate the objective function Loss s and the fitness function ω s , and the definition of the fitness function is:

[0081] (3) Keep l groups of host position sequences with the smallest objective function intact for the next iteration;

[0082] (4) Recombine the N - l groups of host position sequences. The probability that the position of the i-th (i = 1, 2, 3,.., m) host in each sequence is selected from the corresponding position of the s-th (s = 1, 2, 3,..., N) group of host position sequences is ω k ;

[0083] (5) Add random perturbations to the position sequences generated in step (4), and use randomization to find local better solutions;

[0084] (6) Repeat steps (2) to (5) until the number of iterations reaches the set value. At this time, the host position sequence corresponding to min{Loss s} is the finally obtained host distribution strategy.

[0085] In this embodiment, white noise can be used as the random perturbation.

[0086] The features described and / or illustrated for one embodiment above can be used in the same or similar manner in one or more other embodiments, and / or combined with the features in other embodiments or replace the features in other embodiments.

[0087] It should be emphasized that the term "comprising / including" as used herein refers to the presence of features, whole units, steps or components, but does not exclude the presence or addition of one or more other features, whole units, steps, components or combinations thereof.

[0088] Many features and advantages of these embodiments will be apparent from this detailed description, and thus the appended claims are intended to cover all such features and advantages of these embodiments that fall within their true spirit and scope. Further, since many modifications and variations will be readily apparent to those skilled in the art, the embodiments of the present invention are not to be limited to the exact construction and operation illustrated and described, but may cover all suitable modifications and equivalents that fall within their scope.

[0089] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

[0090] The parts of the present invention not described in detail are well-known techniques to those skilled in the art.

Claims

1. A method for optimizing the distribution of drone cluster collaborative navigation hosts based on genetic algorithm, characterized in that: The steps include: The drone cluster includes no less than 3 hosts and no less than 1 slave. The host uses a satellite receiver to obtain its own position information, and the slave uses the host position as a reference and the relative distance between the host and the slave provided by the data link to calculate its own position information; Fisher information is used to characterize the navigation and positioning accuracy of the slaves, and the total Fisher information of all slaves is used as the objective function to be optimized; The positions of all hosts are recorded as the host position sequence, and the host position sequence is iteratively optimized using a genetic algorithm. The host position sequence corresponding to the minimum value of the objective function is the optimal distribution of the hosts.

2. The method according to claim 1, characterized in that: The method in which the slave uses the master position as a reference and uses the relative distance between the master and the slave provided by the data link to calculate its own position information comprises the following steps: The slave machine maintains communication with at least three of the hosts to obtain relative distances between the slave machine and the at least three of the hosts; According to the q relative distances obtained, the first set of equations is constructed: in, X=[x,y,z] is the position of the slave to be solved, [x i ,y i ,z i ] represents the position of the i-th host; ρ i is the relative distance between the slave and the i-th host; Using Newton's iteration method to solve the first set of equations is equivalent to solving the following linear equations: G·ΔX k =b Where G is the f(x,y,z,x) of the slave relative to the i-th host. i ,y i ,z i ) in X k-1 The Jacobian matrix formed by taking partial derivatives of (x, y, z); b is ρ i -f(x,y,z,x i ,y i ,z i ) is the estimated residual matrix; X k-1 represents the estimated position of the slave obtained at the k-1th iteration, ΔX k is the update amount of the kth iteration; Use least squares to solve the update amount ΔX k , update the slave position estimate X k : ΔX k =(G T G) -1 G T b X k =X k-1 +ΔX k Iterate to the set accuracy and output the slave position.

3. The method according to claim 1, characterized in that The method of using Fisher information to characterize the navigation positioning accuracy of the slave is as follows: The Fisher information of the j-th slave is expressed as: Among them, det(FIM) j is the Fisher information of the j-th slave, det() represents the determinant, q is the number of hosts communicating with the j-th slave; H i Represents the observation matrix of the position of the i-th host relative to the slave; The objective function is: Where n is the number of slaves in the drone cluster.

4. The method according to claim 3, characterized in that The observation matrix is H i ==[cosθ i cosα i sinθ i -cosθ i sinα i ] Among them, θ i , α i They respectively represent the pitch angle and azimuth angle of the i-th host relative to the slave.

5. The method according to claim 1, characterized in that: The positions of all hosts are recorded as a host position sequence, and the host position sequence is iteratively optimized using a genetic algorithm, and the host position sequence corresponding to the minimum value of the objective function is the optimal distribution of the hosts, which specifically includes the following steps: S001, randomly generate N groups of initial host position sequences, each group of host position sequences contains m hosts; S002, calculating the objective function and fitness function of each group of host position sequences; S003, retaining l sets of host position sequences with the smallest objective function and proceeding to the next iteration; S004, reorganize Nl groups of host position sequences, and the probability that the i-th host position in each group of reorganized host position sequences is selected from the corresponding position in the s-th group of host position sequences is ω k , i=1,2,3,...,m, s=1,2,3,...,N; S005, adding random perturbations to the recombined host position sequence; S006. Repeat steps S002 to S005 until the number of iterations reaches the set value, and the host position sequence corresponding to the minimum value of the objective function is the optimal distribution of the hosts.

6. The method according to claim 5, characterized in that The fitness function of the host position sequence of the sth group is: Among them, Loss s , is the objective function of the sth, i0th group of host position sequences.

7. The method according to claim 1, characterized in that The random disturbance is white noise.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, wherein the computer program implements the method described in any one of claims 1 to 7 when executed by a processor.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: The processor executes the computer program to implement the method described in any one of claims 1 to 7.

Images