Unmanned aerial vehicle cluster control management method based on positioning information assistance

Abstract

The invention provides an unmanned aerial vehicle cluster control management method based on positioning information assistance, and the method comprises the steps: constructing an unmanned aerial vehicle cluster system in an indoor region, taking each unmanned aerial vehicle in the unmanned aerial vehicle cluster system as a node, and obtaining an unmanned aerial vehicle cluster network; each unmanned aerial vehicle node in the network broadcasts own key information and shares the key information to a corresponding base station; the base station processes the key information by adopting a firefly group intelligence-based cooperative positioning algorithm to obtain the position coordinates of the unmanned aerial vehicle; constructing a multi-dimensional fitness function according to the key information; according to the multi-dimensional fitness function, processing the position coordinates of the unmanned aerial vehicle by adopting firefly group intelligence-based cluster management measurement to obtain an unmanned aerial vehicle control strategy; and completing control management of the unmanned aerial vehicle cluster according to the unmanned aerial vehicle control strategy. According to the method, the interactive behaviors among firefly individuals are simulated, and the FSICM strategy realizes dynamic and efficient clustering management.

Description

Technical Field

[0001] This invention belongs to the field of UAV swarm communication and collaborative control technology, specifically relating to a UAV swarm control and management method based on positioning information assistance. Background Technology

[0002] In disaster relief and emergency monitoring scenarios, FANET can quickly build communication networks, supporting multi-UAV collaborative operations and significantly improving mission execution efficiency. However, when UAV swarms perform tasks in complex environments (such as personnel rescue), they often enter indoor environments where satellite signals are difficult to cover, leading to satellite positioning denial. In these special scenarios, high-precision location information becomes crucial to ensuring stable formation and reliable communication for UAV swarms. Existing positioning algorithms often require highly accurate initial solutions, and their performance degrades significantly under high measurement noise. Furthermore, with the assistance of positioning information, FANET faces the more challenging problem of routing optimization and cluster structure management caused by dynamic changes in network topology. UAV swarms are in a high-speed moving state, causing frequent inter-cluster transfers. Dynamic node joining and leaving triggers routing protocol rediscovery, increasing routing overhead and causing frequent CH switching. Existing CH selection mechanisms do not fully consider the link survival probability of nodes, leading to nodes with unstable communication performance being selected as CHs, further exacerbating the CH switching problem. Furthermore, the irregular high-speed movement of cluster members can cause rapid changes in the cluster topology and frequent interruptions in communication links, affecting the continuity and reliability of data transmission and becoming a key bottleneck restricting the efficient operation of FANET in complex task scenarios.

[0003] Therefore, there is an urgent need for a positioning and cluster management method with higher positioning accuracy, better robustness, longer link lifespan, and longer network lifetime to improve the communication performance of indoor FANET. Summary of the Invention

[0004] To address the problems existing in the prior art, this invention proposes a drone cluster control and management method based on positioning information. The method includes: constructing an indoor drone cluster system, treating each drone in the cluster system as a node to obtain a drone cluster network; each drone node in the network broadcasting its own key information and sharing this information with its corresponding base station; the base station processing the key information using a collaborative positioning algorithm based on firefly swarm intelligence to obtain the drone's position coordinates; constructing a multi-dimensional fitness function based on the key information; processing the drone's position coordinates using cluster management measurement based on firefly swarm intelligence according to the multi-dimensional fitness function to obtain a drone control strategy; and completing the control and management of the drone cluster according to the drone control strategy.

[0005] The beneficial effects of this invention are:

[0006] The high-precision positioning information provided by the FSICL algorithm of this invention lays a solid foundation for CH selection and cluster maintenance in the FSICM strategy. The positioning information-assisted cluster management mechanism ensures the scientific nature and stability of the cluster structure. By simulating the interaction behavior between individual fireflies, the FSICM strategy achieves dynamic and efficient cluster management. CM nodes can adjust their flight trajectories in real time based on the fitness information broadcast by the CH, effectively reducing communication overhead and cluster structure changes caused by node movement. The FSICM strategy ensures a longer LET through a dual mechanism, achieving more stable intra-cluster communication. Combined with the high-precision positioning of the FSICL method, the communication performance and network stability of indoor FANET are ensured. Attached Figure Description

[0007] Figure 1 This is a flowchart illustrating the location-information-assisted bio-inspired cluster management method of the present invention.

[0008] Figure 2 This is a schematic diagram of the specific process of FSICL of the present invention;

[0009] Figure 3 The noise variance of the present invention and Relationship diagram;

[0010] Figure 4 This is a schematic diagram of the specific process of FSICM of the present invention;

[0011] Figure 5 This is a schematic diagram illustrating the inter-node link survival probability of the present invention. Detailed Implementation

[0012] 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 skilled in the art without creative effort are within the scope of protection of the present invention.

[0013] A method for controlling and managing unmanned aerial vehicle (UAV) clusters based on location information is disclosed. The method includes: constructing an indoor UAV cluster system, treating each UAV in the cluster system as a node to obtain a UAV cluster network; broadcasting key information from each UAV node in the network and sharing this key information with the corresponding base station; processing the key information using a collaborative positioning algorithm based on firefly swarm intelligence to obtain the UAV's position coordinates; constructing a multi-dimensional fitness function based on the key information; processing the UAV's position coordinates using cluster management measurement based on firefly swarm intelligence based on the multi-dimensional fitness function to obtain a UAV control strategy; and completing the control and management of the UAV cluster according to the UAV control strategy.

[0014] This embodiment includes a location-information-assisted bio-inspired cluster management method applied to cluster management in an indoor FANET environment. It includes: constructing a bio-inspired cooperative localization algorithm (FSICL) based on firefly swarm intelligence and a cluster management strategy (FSICM) based on firefly swarm intelligence; using the constructed cooperative localization algorithm and cluster management strategy, by integrating the advantages of the firefly algorithm (FA) and the Chan algorithm, to achieve optimized cooperative localization among indoor UAVs; and then, based on the localization results, abstracting each UAV as a firefly individual with luminous characteristics, whose light intensity is dynamically adjusted by a fitness function composed of multi-dimensional parameters such as link survival probability, node degree difference, average neighbor distance, and remaining energy, to complete spontaneous cluster management according to the "phototaxis" rule; this embodiment obtains cluster management results by adopting a location-information-assisted bio-inspired strategy, combined with firefly swarm intelligence and multi-dimensional fitness evaluation, such as... Figure 1 As shown.

[0015] The key information is processed using a cooperative localization algorithm based on firefly swarm intelligence, including:

[0016] S101. The UAV periodically sends signals to the base station, and the base station receives and records the signal TDOA.

[0017] S102. The Chan algorithm is used to process the TDOA data to obtain the initial position of the UAV;

[0018] S103. Establish a cube search area based on the initial position of the UAV;

[0019] S104. The FA algorithm is used to perform a global search optimization on the cube search area to obtain the optimal result, where the optimal result is the precise coordinates of the UAV.

[0020] Specifically, such as Figure 2 As shown, the implementation of the FSICL cooperative localization algorithm includes:

[0021] S1. Data Acquisition and TDOA Calculation. The UAV periodically broadcasts beacon signals containing location information. The base station receives these signals and records their TDOA. The core of the TDOA algorithm for positioning lies in constructing geometric constraints by measuring the signal propagation time difference to achieve accurate target location. In a 3D indoor positioning scenario, it is assumed that... base stations ( ), base station The position is The location of the drone is The drone transmits UWB signals in a three-dimensional indoor area, and all base stations within the drone's positioning range will receive the wireless signals.

[0022] Defining drones to base stations The distance is The formula is:

[0023]

[0024] This indicates the distance from the drone to base station 1 and to base station 2. The distance difference is calculated using the formula:

[0025]

[0026] in Indicates the speed of UWB radio waves. This indicates the drone's connection to base station 1 and base station 2. TDOA value, It has a mean of zero and a variance of zero. Additive white Gaussian noise. Based on the above equation, the formula is:

[0027]

[0028] in, and .

[0029] This formula can be used to obtain the location of the drone.

[0030] S2. Initial Solution Calculation: The Chan algorithm is used to preliminarily process the TDOA data to obtain an initial estimate of the UAV's position. The Chan algorithm uses Weighted Least Squares (WLS) to solve the equations twice: First, the initial nonlinear equations are transformed into linear equations containing relevant TDOA data. The first WLS provides the initial solution, and the second WLS uses the known constraints between the UAV coordinates and additional variables to provide an improved position estimate.

[0031] S3. This invention introduces Algorithm Optimization (FA) based on the Chan algorithm. First, the Chan algorithm is used to obtain an initial solution, providing a foundation for subsequent optimization. Then, a finite cubic search region is defined centered on this initial solution, and FA is applied within this region for further optimization to obtain a more accurate solution. In this process, each individual firefly in space represents a possible localization estimate. Due to the limited search region, the probability of FA getting trapped in local optima is significantly reduced, while the convergence speed of the algorithm is improved, thereby effectively increasing the localization accuracy.

[0032] Specifically, the cooperative localization algorithm includes:

[0033] Step 1: Calculate the initial solution using the Chan algorithm. And obtain the distance difference from the solution to base station 1 and to each base station. The formula is:

[0034]

[0035] Define the Chan algorithm The difference, the formula is:

[0036]

[0037] If the search area is set too small, the true solution of the UAV may be excluded from the search space; while if the search area is too large, it will significantly increase the search time and make the algorithm more likely to get trapped in local optima. Figure 3 The trend shown indicates that the distance between the estimated and actual positions of the drone is less than a threshold. .

[0038] Step 2, using the initial solution Centered on, establish a side length of A firefly swarm search is performed within the cubic search area, and the solution obtained through FA is given by the following formula:

[0039]

[0040] Step 3: Obtain the distance difference between the solution to base station 1 and to each base station. The formula is:

[0041]

[0042] Define FA The difference, the formula is:

[0043]

[0044] Step 4: When FA reaches its maximum number of iterations, perform the following judgment: If the condition is met... If the condition is met, the solution obtained by FA will be used as the final position of the UAV; otherwise, if the condition is not met, the solution obtained by Chan's algorithm will be used as the final position of the UAV.

[0045] In this embodiment, the solutions obtained through FA include:

[0046] S101. Initialize parameters, including population size, maximum attraction, light absorption coefficient, step size factor, and maximum number of iterations;

[0047] S102. Initialize the firefly population by randomly generating the initial positions of n fireflies within the cube region. The position of each firefly is represented by a three-dimensional vector X. i =(x i ,y i ,z i );

[0048] S103. Define the brightness of fireflies as the inverse mapping of the objective function value, calculate the brightness of two fireflies; calculate the Euclidean distance between the two fireflies.

[0049] The formula for calculating the brightness of fireflies is:

[0050]

[0051] in, The objective function value, This represents the position of firefly i.

[0052] The formula for calculating the objective function value is:

[0053] ;

[0054] The Euclidean distance between two fireflies is calculated as follows:

[0055]

[0056] in, The position of firefly i. This is the position of firefly j. Let x be the x-coordinate of firefly i. Let be the ordinate of firefly i. Let z be the z-coordinate of firefly i; Let x be the x-coordinate of firefly j. Let be the ordinate of firefly j. Let be the z-coordinate of firefly j.

[0057] S104. Calculate the attraction between two fireflies based on Euclidean distance;

[0058] The formula for calculating attractiveness is:

[0059]

[0060] in, For initial attraction, The square of the Euclidean distance. is the light absorption coefficient; where The larger the size, the faster the attraction diminishes. The smaller the size, the slower the attraction diminishes.

[0061] S105. Determine the firefly's movement direction based on its brightness and attractiveness, and update the firefly's position.

[0062] The updated formula is:

[0063]

[0064] Where rand is the random perturbation term. This is the step size factor.

[0065] S106. Determine whether the fireflies are outside the search area based on their updated positions. If they are outside, pull them back to the boundary.

[0066] S107. Repeat steps S103 to S106 until the maximum number of iterations is reached or the objective function value converges, then output the optimal position of the population.

[0067] Specifically, such as Figure 4 As shown, the implementation of cluster head selection includes:

[0068] S1. The drone swarm is randomly distributed in three-dimensional space. Each drone is assigned an initial position, velocity, and remaining energy. Each drone calculates its own fitness value based on multiple parameters such as link survival probability, node degree difference, average neighbor distance, and remaining energy.

[0069] Specifically, the calculation of fitness values ​​includes:

[0070] Step 1, Current drone In the cycle Continuous reception of neighboring drones within the cluster ( From position Move to location Two sets of Hello messages are sent at the same time, with a communication radius of R. Assume that... The relative speed and direction of the drone remain unchanged within a certain time period, and its position remains constant. express Compared to The relative position, such as Figure 5 As shown.

[0071] Drone m in position and Two sets of consecutive Hello messages are sent. For the continuously received Hello messages, the drone m sends two sets of consecutive Hello messages at a time. The formula for coordinate changes within the space is:

[0072]

[0073] in, Let m be the unit vectors for the X, Y, and Z axes, respectively. Further, the velocity of the UAV m can be obtained from the above formula:

[0074]

[0075]

[0076]

[0077]

[0078] Similarly, the movement speed of UAV n can also be obtained using the formula:

[0079]

[0080] Therefore, the relative velocity between drones m and n is given by the formula:

[0081]

[0082] The coordinates at position e are given by the formula:

[0083]

[0084] Furthermore, It can be regarded as a line segment The direction vector is given, therefore the line segments can be calculated separately. and The length of is given by the formula:

[0085]

[0086] Define the drones m and n in the periodic The formula for the change in internal distance difference is:

[0087]

[0088] Furthermore, the link survival probability between drones n and m is related to their relative speed and distance; therefore, the link survival probability is given by the formula:

[0089]

[0090] Where the normalization factor The formula is:

[0091]

[0092] Furthermore, the average link survival probability of drone n can be obtained by the formula:

[0093]

[0094] Step 2: Node n in the network obtains its degree by sending a Hello message to its neighbor m. The formula is:

[0095]

[0096] Specifically, the optimal number of drones in the network is affected by the intra-cluster communication bandwidth. Inter-cluster communication bandwidth The number of drones in the entire network, N, has a combined influence. The optimal number of drones within the cluster is calculated using the following formula:

[0097]

[0098] The node degree difference of UAV n represents the similarity between its actual number of nodes and the theoretical optimal value, and the formula is:

[0099]

[0100] Furthermore, the normalized degree difference of the n-node unmanned aerial vehicle (UAV) The formula is:

[0101]

[0102] Step 3: Calculate the distance between drone n and all neighboring nodes. The formula is:

[0103]

[0104] Furthermore, the average distance between drone n and its neighboring nodes The formula is:

[0105]

[0106] Step 4: When the UAV acts as the CM, the energy consumed per node per unit time is... When it is used as CH, the energy consumed per unit time per node degree is When it hovers, the energy consumed per unit time is Calculate UAV The remaining energy after a period of time is expressed by the formula:

[0107]

[0108] in It is a drone initial energy, It is a UAV Number of times acting as CM It is a UAV In the The node degree that acts as CM. It is the first The time during which the UAV acts as a CM. It refers to the number of times a UAV acts as a CH. It is a UAV In the The node degree that acts as CH is secondary. It is the first The time during which the UAV acts as CH. It is a UAV Hovering time, It is a UAV The weight.

[0109] S2. The system compares the fitness values ​​of all drones and selects the drone with the highest fitness value as the CH (Chain Leader). The "light intensity" of a CH node is determined by its fitness value; CHs with higher light intensity are more attractive. The fitness function is calculated as follows:

[0110]

[0111]

[0112] in, , , , These are the weights of the four parameters. Let n be the average link survival probability of drone n. Let n be the node degree difference of the unmanned aerial vehicle (UAV). Let n be the average distance of the drone. Let n be the remaining energy of the drone.

[0113] The processing of UAV position coordinates using cluster management measurement based on firefly swarm intelligence includes:

[0114] S201. Initialize the drone cluster parameters, which include time t, drone number n and neighboring drone numbers m, target drone number N, maximum neighboring drone number M and time threshold.

[0115] S202. Determine the difference between the current time t and the time threshold. If the t is greater than the time threshold, proceed to step S203; otherwise, proceed to step S204.

[0116] S203, Output the UAV control strategy;

[0117] S204. Determine the size of the current drone number n and the target drone number N. If n is greater than N, proceed to step S205; otherwise, proceed to step S206.

[0118] S205. Increment time t by 1 and return to step S202;

[0119] S206. Calculate the adaptive function value In of UAV number n using a multi-dimensional fitness function;

[0120] S207. Determine the size of the domain drone number m and the maximum number of neighboring drones M. If m is greater than M, proceed to step S208; otherwise, proceed to step S209.

[0121] S208. Increment the current drone number n by 1, and return to step S204;

[0122] S209. Calculate the adaptive function value Im for the neighboring UAV numbered m using a multi-dimensional fitness function;

[0123] S210. Compare In with Im. If In is less than Im, proceed to step S211; otherwise, proceed to step S212.

[0124] S211. Select drone n as CM node, and drone n follows CH movement, and execute step S213;

[0125] S212. Select drone n as CH node, and drone n follows the movement of the human, and execute step S213.

[0126] S213, increment the maximum number of neighboring drones M by 1, and return to step S207.

[0127] Specifically, the implementation of cluster maintenance includes:

[0128] S1, cluster node following.

[0129] When a CM node fails to receive the guidance signal carrying light intensity (fitness) information broadcast by the CH within T seconds, or when a CH node cannot obtain the status feedback message sent by the CM, the system, based on a two-way communication verification mechanism, accurately determines that the CM has left the effective sensing range of the CH, and then triggers the "delete node within cluster" step to avoid energy waste caused by invalid communication. This signal interaction-based sensing mechanism, analogous to a firefly's perception and response to the light signals of its companions, can quickly capture changes in the connection status between nodes, ensuring efficient utilization of network resources.

[0130] Furthermore, when the CM node successfully receives the guidance signal, it will initiate a trajectory correction program based on biomimetic optimization. This program simulates the behavior of fireflies adjusting their movement direction according to the light intensity of their companions during foraging or migration, establishing a quantitative relationship between fitness value and flight direction adjustment. Specifically, the CM node uses the fitness value of CH as a "light intensity" guide, and iteratively calculates and adjusts its own speed and heading in real time, continuously moving towards the direction of CH. Compared to traditional adjustment strategies based on distance or fixed rules, this biomimetic algorithm can dynamically adapt to the high-speed movement characteristics of the UAV, fully consider the relative motion state between nodes and network load, and maintain the compactness and stability of the cluster topology while reducing communication overhead.

[0131] S2, Delete a node within the cluster.

[0132] When a CM node cannot receive a Hello message with CH ID information, or a CH node cannot receive any Hello messages periodically broadcast by a CM node, it indicates that the CM node has exceeded the maximum communication range of the CH and its information needs to be removed from the cluster.

[0133] S3, Add a new node.

[0134] Essentially, new nodes include nodes added to the network and nodes removed from other clusters. When a new node attempts to join a cluster, it needs to establish a relationship through a "handshake" process and recalculate its own weights for CH selection.

[0135] S4, replace CH.

[0136] When a CH (Chain Leader) fails to receive any Hello messages with CM ID information from within the cluster, or leaves the network for other reasons, it can be considered to have relinquished its leadership position. In this case, the CM nodes within the cluster will return to their initial state and perform a clustering process to re-elect a CH. It is worth noting that when a CH can receive at least one HELLO message from a CM, it indicates that it is still fulfilling its CH responsibilities; a decrease in the number of CMs may simply mean that they have moved to another cluster or shut down.

[0137] S5, Cluster Merging.

[0138] When the Leaders (CHs) of two clusters become one-hop neighbors, a cluster merge problem inevitably arises. At this point, the Let's say the CH and the Communicator (CM) within the same cluster are compared: if the Let's say ...

[0139] This invention systematically solves the inefficiency problem of indoor high-dynamic FANET cluster management by designing the FSICL algorithm and FSICM strategy through a bio-inspired cluster management method assisted by location information. The former achieves high-precision positioning through multi-algorithm fusion, while the latter ensures dynamic stability of the cluster based on a biomimetic mechanism. Together, they constitute an efficient and robust indoor UAV ad hoc network cluster management mechanism.

[0140] The above-described embodiments further illustrate the purpose, technical solution, and advantages of the present invention. It should be understood that the above-described embodiments are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made to the present invention within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for controlling and managing unmanned aerial vehicle (UAV) clusters based on positioning information, characterized in that, include: Construct an indoor drone swarm system, using each drone in the drone swarm system as a node to obtain a drone swarm network; Each drone node in the network broadcasts its own key information and shares it with the corresponding base station. The base station uses a collaborative localization algorithm based on firefly swarm intelligence to process the key information and obtain the drone's position coordinates. A multi-dimensional fitness function is constructed based on the key information. Based on the multi-dimensional fitness function, cluster management measurement based on firefly swarm intelligence is used to process the drone's position coordinates and obtain the drone control strategy. The drone control strategy is then used to complete the control and management of the drone cluster.

2. The method for controlling and managing unmanned aerial vehicle (UAV) clusters based on positioning information according to claim 1, characterized in that, Key information includes each drone's link survival probability, node degree difference, average neighbor distance, remaining energy, initial position, and speed.

3. The method for controlling and managing unmanned aerial vehicle (UAV) clusters based on positioning information according to claim 1, characterized in that, The key information is processed using a cooperative localization algorithm based on firefly swarm intelligence, including: S101. The UAV periodically sends signals to the base station, and the base station receives and records the signal TDOA. S102. The Chan algorithm is used to process the TDOA data to obtain the initial position of the UAV; S103. Establish a cube search area based on the initial position of the UAV; S104. The FA algorithm is used to perform a global search optimization on the cube search area to obtain the optimal result, where the optimal result is the precise coordinates of the UAV.

4. The method for controlling and managing unmanned aerial vehicle (UAV) clusters based on positioning information according to claim 3, characterized in that, The Chan algorithm is used to process TDOA data, which includes: converting the initial nonlinear equation into a linear equation containing the relevant TDOA data; performing a preliminary solution to the linear equation; and performing a second solution to the linear equation based on the initial solution and the constraints between the UAV coordinates and variables to obtain the initial position of the UAV.

5. The method for controlling and managing unmanned aerial vehicle (UAV) clusters based on positioning information according to claim 3, characterized in that, The FA algorithm is used to optimize the global search of the cube search region, including: S101. Initialize parameters, including population size, maximum attraction, light absorption coefficient, step size factor, and maximum number of iterations; S102. Initialize the firefly population by randomly generating the initial positions of n fireflies within the cube region. The position of each firefly is represented by a three-dimensional vector X. i =(x i ,y i ,z i ); S103. Define the brightness of fireflies as the inverse mapping of the objective function value, calculate the brightness of two fireflies; calculate the Euclidean distance between the two fireflies. S104. Calculate the attraction between two fireflies based on Euclidean distance; S105. Determine the firefly's movement direction based on its brightness and attractiveness, and update the firefly's position. S106. Determine whether the fireflies are outside the search area based on their updated positions. If they are outside, pull them back to the boundary. S107. Repeat steps S103 to S106 until the maximum number of iterations is reached or the objective function value converges, then output the optimal position of the population.

6. The method for controlling and managing unmanned aerial vehicle (UAV) clusters based on positioning information according to claim 5, characterized in that, The brightness of fireflies is: in, The objective function value, This represents the position of firefly i.

7. The method for controlling and managing unmanned aerial vehicle (UAV) clusters based on positioning information according to claim 3, characterized in that, Establishing a cube search region involves: using the initial solution obtained from the Chan algorithm as the center, establishing a cube with side lengths of... The cube search region, in which The difference between the distance difference from the initial solution to each base station and the distance difference from the UAV to each base station obtained based on the TDOA value is given.

8. The method for controlling and managing unmanned aerial vehicle (UAV) clusters based on positioning information according to claim 1, characterized in that, Constructing a multi-dimensional fitness function includes: The drone continuously receives two sets of Hello messages sent by neighboring drones within the cluster when they move from position b to position c within a period T. Calculate the coordinate changes of the UAV within time T based on the two sets of Hello messages; calculate the changes in the relative speed and position distance difference between the two UAVs based on the coordinate changes; calculate the link survival probability of the two UAVs based on the changes in relative speed and position distance difference; calculate the average link survival probability of the UAVs based on all link survival probabilities. Node degree is obtained based on two sets of Hello messages; the optimal number of drones in the cluster is calculated; the node degree difference representation of drones is calculated based on the node degree and the optimal number of drones, and the node degree difference representation is normalized to obtain the drone node degree difference. Calculate the distance between the current drone and all neighboring nodes, and average all distances to obtain the average distance between the drone and its neighboring nodes; Calculate the remaining energy of the drone after it has traveled a certain distance; A fitness function is constructed based on the average link survival probability of the drone, the degree difference of the drone node, the average distance between the drone and its neighboring nodes, and the remaining energy.

9. A method for controlling and managing unmanned aerial vehicle (UAV) clusters based on positioning information as described in claim 8, characterized in that, The fitness function is: in, , , , These are the weights of the four parameters. Let n be the average link survival probability of drone n. Let n be the node degree difference of the unmanned aerial vehicle (UAV). Let n be the average distance of the drone. Let n be the remaining energy of the drone.

10. The method for controlling and managing unmanned aerial vehicle (UAV) clusters based on positioning information according to claim 1, characterized in that, The processing of UAV position coordinates using cluster management measurement based on firefly swarm intelligence includes: S201. Initialize the drone cluster parameters, which include time t, drone number n and neighboring drone numbers m, target drone number N, maximum neighboring drone number M and time threshold. S202. Determine the difference between the current time t and the time threshold. If the t is greater than the time threshold, proceed to step S203; otherwise, proceed to step S204. S203, Output the UAV control strategy; S204. Determine the size of the current drone number n and the target drone number N. If n is greater than N, proceed to step S205; otherwise, proceed to step S206. S205. Increment time t by 1 and return to step S202; S206. Calculate the adaptive function value In of UAV number n using a multi-dimensional fitness function; S207. Determine the size of the domain drone number m and the maximum number of neighboring drones M. If m is greater than M, proceed to step S208; otherwise, proceed to step S209. S208. Increment the current drone number n by 1, and return to step S204; S209. Calculate the adaptive function value Im for the neighboring UAV numbered m using a multi-dimensional fitness function; S210. Compare In with Im. If In is less than Im, proceed to step S211; otherwise, proceed to step S212. S211. Select drone n as CM node, and drone n follows CH movement, and execute step S213; S212. Select drone n as CH node, and drone n follows the movement of the human, and execute step S213. S213, increment the maximum number of neighboring drones M by 1, and return to step S207.

Images