An unmanned aerial vehicle cluster ad hoc network method and system for high mountain power grid inspection
By constructing a bidirectional mapping unit for the physical-virtual layer and a dynamic spectrum allocation algorithm, combined with energy-sensing topology optimization and biomimetic path planning, the problems of poor communication capability and inaccurate path planning of UAV self-organizing networks in high-altitude power grid inspections were solved, realizing efficient and stable UAV swarm self-organizing networks and collaborative operations.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- YUNNAN POWER GRID CO LTD
- Filing Date
- 2025-11-11
- Publication Date
- 2026-05-29
AI Technical Summary
During high-altitude power grid inspections, drones face challenges such as poor communication capabilities, low spectrum utilization, unstable communication quality, and inaccurate path planning during self-organizing networks. In particular, effective spectrum allocation and obstacle avoidance strategies are difficult to implement in complex terrain.
By constructing a two-way mapping unit for the physical-virtual layer, a three-dimensional terrain matrix is built using lidar point cloud data, a terrain influence function is established, and dynamic spectrum allocation algorithm, energy-sensing topology optimization algorithm, and biomimetic path planning algorithm are combined to optimize the spectrum strategy and flight path of UAVs, thereby realizing the self-organizing network of UAV swarms.
It improves the communication stability and autonomy of the drone swarm, enhances the robustness and adaptability of the system, extends the mission execution time and service life, and improves inspection efficiency and safety.
Smart Images
Figure CN121099337B_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of unmanned aerial vehicle (UAV) technology, specifically relating to a method and system for UAV swarm self-organizing network for high-altitude power grid inspection. Background Technology
[0002] With the development of drone technology, its application in various complex environments is becoming increasingly widespread, especially in high-precision and high-efficiency mission scenarios such as high-altitude power grid inspection. However, in these specific application scenarios, the networking process of drones is often accompanied by problems such as the effective use of spectrum resources, energy management and allocation, and the impact of complex terrain on communication quality.
[0003] In the process of UAV self-organizing network, the reasonable allocation of spectrum resources is one of the key factors to ensure network performance. Since wireless channels have dynamic characteristics, and traditional spectrum allocation generally adopts a static allocation method, this method lacks flexibility and is therefore difficult to adapt to rapidly changing environmental conditions, resulting in low spectrum utilization and even spectrum conflicts.
[0004] In addition, in complex terrain environments such as mountainous areas, terrain obstacles not only affect the flight path planning of UAVs, but also significantly interfere with the propagation of wireless signals, reducing communication quality and network coverage. Static allocation methods also have the drawback of insufficient utilization of terrain information, making it difficult to achieve accurate three-dimensional topology reconstruction and effective obstacle avoidance strategies, thus limiting the operational efficiency of UAVs in complex terrain.
[0005] Existing technologies disclose a method and system for UAV trajectory planning and obstacle avoidance based on spectrum maps. In this method, the UAV pre-plans a path by constructing a three-dimensional spectrum map, and then dynamically avoids obstacles in real time based on the local spectrum map of the domain. When interference is detected, the UAV uses a perturbation fluid algorithm for dynamic obstacle avoidance. Although this method can effectively avoid obstacles in high-altitude power grid inspection, it lacks the ability to network between UAVs, resulting in poor communication capabilities among multiple UAVs in high-altitude power grid inspection.
[0006] Therefore, we propose a self-organizing network method that is suitable for high-altitude inspection operations and provides stable communication. Summary of the Invention
[0007] In view of the above-mentioned problems, the present invention is proposed.
[0008] Therefore, the purpose of this invention is to provide a method for self-organizing a drone swarm for high-altitude power grid inspection, which can solve the problem of poor communication capabilities between multiple drones.
[0009] To address the aforementioned technical issues, a method for self-organizing UAV swarms for high-altitude power grid inspection is proposed, including:
[0010] A bidirectional mapping unit for the physical-virtual layer is constructed. The physical layer of the bidirectional mapping unit consists of N UAV nodes, and the virtual layer creates a digital twin for each UAV node. The bidirectional mapping unit constructs a three-dimensional terrain matrix based on the lidar point cloud data of the UAV nodes and establishes a terrain influence function to reflect the impact of terrain on communication quality. Based on the terrain influence function, a dynamic spectrum allocation algorithm is used to obtain the optimal spectrum strategy for all UAV nodes. An energy-aware topology optimization algorithm is used to dynamically adjust the network structure according to the energy state of the UAVs. A biomimetic path planning algorithm is used to plan the optimal flight route for each UAV.
[0011] As a preferred embodiment of the UAV swarm self-organizing network method for high-altitude power grid inspection described in this invention, the bidirectional mapping unit includes:
[0012] The physical layer consists of the following drones: Each drone is equipped with a lidar and a millimeter-wave communication module, where N is the number of drones and i is the variable index. , and These are the three-dimensional coordinates of the i-th drone. For the i-th drone;
[0013] The set of digital twins for the virtual layer is as follows: And synchronize the energy state of physical layer nodes in real time. Signal-to-interference-plus-noise ratio and task timeliness , For physical layer drones The corresponding digital twin, and These are sets of drones at the physical layer and sets of digital twins at the virtual layer, respectively.
[0014] As a preferred embodiment of the UAV swarm self-organizing network method for high-altitude power grid inspection described in this invention, the construction of the three-dimensional terrain matrix includes, in the physical layer, the three-dimensional terrain matrix constructed using lidar point clouds as follows: ;
[0015] The established terrain influence function is as follows:
[0016] ,
[0017] in, It is the number of obstacles. It is the radius of curvature of the terrain. It is the maximum height difference. , and They are the first The three-dimensional coordinates of the obstacle;
[0018] The virtual layer calculates terrain gradients in real time using a digital twin. And feed it back to the physical layer.
[0019] As a preferred embodiment of the UAV swarm self-organizing network method for high-altitude power grid inspection described in this invention, the terrain influence function includes the terrain curvature radius in the terrain influence function. The adjustment is adaptive based on the point cloud density of the lidar, and satisfies the following:
[0020] ,
[0021] in, This represents the terrain complexity coefficient.
[0022] The optimal spectrum strategy includes,
[0023] Constructing a reinforcement learning state space ,in, Given the current energy state, For packet loss rate, As a task priority, This is a terrain influence function;
[0024] Based on the reinforcement learning state space, Q-learning is used to optimize the objective function to obtain the optimal spectrum of each UAV;
[0025] Based on the optimal spectrum of each UAV, the utility function of each UAV node is defined, and the optimal spectrum strategy of the UAV cluster is obtained by iteratively solving each utility function through gradient projection method.
[0026] As a preferred embodiment of the UAV swarm self-organizing network method for high-altitude power grid inspection described in this invention, the Q-learning optimization objective function includes:
[0027] ,
[0028] in, For the reward function, For learning rate, As a discount factor, This serves as the reinforcement learning state space for the next state. It is the spectrum strategy of drones. Spectrum strategy for the drone in the next state;
[0029] The reward function The calculation formula is:
[0030] ,
[0031] in, , and These are weighting coefficients. and These represent the achieved data rate and the required data rate, respectively. and These represent the energy consumed and the energy threshold, respectively.
[0032] The utility function of each UAV node is:
[0033] ,
[0034] in, Indicates the first A drone node on its selected spectrum resources The utility value below, Indicates the first The transmit power of each drone node, Indicates the first The channel gain of each drone node to itself. For background noise power density, Indicates originating from all remaining drone nodes. ( ) for the The sum of interference at each node To adjust the parameters, It is the sum of the reciprocals of the frequency intervals, the quantization of the first... The proximity of spectrum resources between a node and other nodes, where i and j are variable indices and are positive numbers;
[0035] The process of iteratively solving the utility function using the gradient projection method is as follows:
[0036] ,
[0037] in, In the During the nth iteration, the 1st The spectrum resource selection after the update for each drone node. In the During the nth iteration, the 1st The spectrum resource selection after the update for each drone node. It is the step size parameter, and it is the utility function. For spectrum resources The partial derivatives of .
[0038] As a preferred embodiment of the UAV swarm self-organizing network method for high-altitude power grid inspection described in this invention, the energy sensing topology optimization algorithm includes:
[0039] Predict the remaining energy of each drone;
[0040] Adjust the communication radius between drones based on the predicted remaining energy;
[0041] Dynamically optimize the network structure of the entire drone.
[0042] As a preferred embodiment of the UAV swarm self-organizing network method for high-altitude power grid inspection described in this invention, the biomimetic path planning algorithm includes:
[0043] Set an initial flight path for each drone and calculate the fitness value for each path;
[0044] Based on the fitness value of each path, the optimal path for the UAV is determined using the vulture search optimization algorithm;
[0045] When obstacles exist on the optimal path, a repulsive force algorithm based on the potential field method is used to recalculate the path to avoid the obstacles.
[0046] Another objective of this invention is to provide a drone swarm self-organizing network system for high-altitude power grid inspection, enabling efficient self-organizing networking and collaborative operation of drone swarms in high-altitude power grid inspection, thereby improving inspection efficiency and safety.
[0047] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a drone swarm self-organizing network system for high-altitude power grid inspection, comprising a construction module, a spectrum allocation module, a dynamic adjustment module, and a route planning module;
[0048] The building module is used to construct a bidirectional mapping unit for the physical-virtual layer;
[0049] The spectrum allocation module uses a dynamic spectrum allocation algorithm to obtain the optimal spectrum strategy for all UAV nodes.
[0050] The dynamic adjustment module uses an energy-sensing topology optimization algorithm to dynamically adjust the network structure according to the energy status of each UAV.
[0051] The route planning module uses a biomimetic path planning algorithm to plan the optimal flight route for each UAV.
[0052] A computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the steps of the aforementioned method for self-organizing a drone swarm for high-altitude power grid inspection.
[0053] A computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the aforementioned method for self-organizing a drone swarm for high-altitude power grid inspection.
[0054] The beneficial effects of the present invention are as follows: The present invention discloses a method for self-organizing a drone swarm for high-altitude power grid inspection. By constructing a two-way mapping unit of physical-virtual layer, real-time monitoring and feedback of drone status are realized.
[0055] A three-dimensional terrain matrix is constructed based on lidar point cloud data from UAV nodes, and a terrain influence function is established to reflect the impact of terrain on communication quality, thereby maintaining a stable communication connection even in complex mountainous environments.
[0056] In addition, the dynamic spectrum allocation algorithm can precisely control the spectrum parameters used by each UAV node, maximize the overall network transmission efficiency and service quality, and at the same time rationally allocate the spectrum resources required for energy-intensive operations, avoid unnecessary energy waste, and indirectly extend the mission execution time and service life of UAVs.
[0057] By utilizing the energy-sensing topology optimization algorithm, the network structure can be dynamically adjusted according to the energy status of each UAV, thereby extending the lifespan of the entire UAV network and ensuring the effective execution of the mission.
[0058] A biomimetic path planning algorithm is applied to plan the optimal flight route for each UAV. By combining information from the global best position, the group average position, and the random individual position, the ratio of exploration to development is adjusted through adaptive weighting to help the UAV find the optimal path. The potential field method is used to calculate the repulsive force of obstacles on the UAV and the influence of terrain is taken into account to ensure that the UAV can safely bypass obstacles and maintain a stable flight path.
[0059] This method improves the autonomy and efficiency of UAV swarms, while also enhancing the robustness and adaptability of the system. Attached Figure Description
[0060] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0061] Figure 1 This is a flowchart illustrating the overall process of a drone swarm self-organizing network method for high-altitude power grid inspection, as provided in one embodiment of the present invention.
[0062] Figure 2This is a schematic diagram of the system structure of a drone swarm self-organizing network system for high-altitude power grid inspection, provided as an embodiment of the present invention.
[0063] Figure 3 This is a schematic diagram of an electronic device structure for a drone swarm self-organizing network method for high-altitude power grid inspection, provided as an embodiment of the present invention. Detailed Implementation
[0064] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.
[0065] Example 1, referring to Figure 1 This is the first embodiment of the present invention, which provides a method for self-organizing a drone swarm for high-altitude power grid inspection, comprising:
[0066] S1: Construct a bidirectional mapping unit for the physical-virtual layer. The physical layer of the bidirectional mapping unit consists of N UAV nodes, and the virtual layer of the bidirectional mapping unit creates a digital twin for each UAV node. The bidirectional mapping unit constructs a three-dimensional terrain matrix based on the lidar point cloud data of the UAV nodes and establishes a terrain influence function to reflect the impact of terrain on communication quality.
[0067] It should be noted that the collection of drones in the physical layer is as follows: Each drone is equipped with a lidar and a millimeter-wave communication module, where N is the number of drones and i is the variable index. , and These are the three-dimensional coordinates of the i-th drone. For the i-th drone;
[0068] The set of digital twins of the virtual layer for: And synchronize the energy state of physical layer nodes in real time. Signal-to-interference-plus-noise ratio and task timeliness ; For physical layer drones The corresponding digital twin, and These are sets of drones at the physical layer and sets of digital twins at the virtual layer, respectively.
[0069] Furthermore, the three-dimensional terrain matrix constructed by the physical layer using lidar point clouds is as follows: H, W, and C represent the height, width, and number of channels of the three-dimensional terrain matrix, respectively. Let be the set of real numbers, indicating that each element in the matrix is a real number.
[0070] The established terrain influence function is as follows:
[0071] ,
[0072] in, It is the number of obstacles. It is the radius of curvature of the terrain, and the current altitude of the drone. Height of the obstacle The absolute difference It is the maximum height difference, which can be seen from Extract from, , and They are the first The three-dimensional coordinates of the obstacle, where k is the variable index and is a positive constant;
[0073] This terrain influence function is essentially applied to each obstacle. Calculate the product of its horizontal attenuation and vertical penalty, and then sum it over all obstacles;
[0074] When the drone approaches the obstacle at a horizontal position and at a similar altitude... A significant increase in the value indicates a collision risk or communication obstruction effect.
[0075] The virtual layer calculates terrain gradients in real time using a digital twin. And feed it back to the physical layer.
[0076] The terrain curvature radius in the terrain influence function, as needed. The adjustment is adaptive based on the point cloud density of the lidar, and satisfies the following:
[0077] ,
[0078] in, This represents the terrain complexity coefficient.
[0079] S2: Based on the terrain influence function, a dynamic spectrum allocation algorithm is used to obtain the optimal spectrum strategy for all UAV nodes.
[0080] Furthermore, the specific steps to obtain the optimal spectrum strategy for all drone nodes are as follows:
[0081] Constructing a reinforcement learning state space ;
[0082] in, Given the current energy state, For packet loss rate, Task priority can be determined by task timeliness. get, This is the terrain influence function; the reinforcement learning state space is used to allow the algorithm to know the current state of each drone.
[0083] Based on the reinforcement learning state space, Q-learning is used to optimize the objective function to obtain the optimal spectrum for each UAV. This step is mainly to find the optimal spectrum, that is, the best combination of radio frequency, bandwidth and transmission power, to ensure that UAVs can communicate smoothly without interfering with each other.
[0084] Based on the optimal spectrum of each drone, the utility function of each drone node is defined. The optimal spectrum strategy of the drone swarm is obtained by iteratively solving each utility function through gradient projection method. This step is the optimization step. It adopts a method similar to "game theory" so that each drone can take into account the choices of other drones and adjust its own strategy accordingly, thereby achieving a balance and reducing interference between them.
[0085] It should be noted that for the strategy of obtaining the optimal spectrum for all UAV nodes using a dynamic spectrum allocation algorithm, the algorithm does not stop working once the initial setup is complete. On the contrary, it continuously monitors the changes in the UAV's status and adjusts the radio parameters in real time according to the new situation. This is similar to a traffic control center adjusting the timing of traffic lights according to the actual road conditions, ensuring that the entire system is always in an optimal operating state.
[0086] Specifically, the Q-learning optimization objective function for:
[0087] ,
[0088] in, For the reward function, For learning rate, As a discount factor, This serves as the reinforcement learning state space for the next state. It is the spectrum strategy of drones. Let s be the spectrum policy of the drone in the next state, and s be the reinforcement learning state space;
[0089] The reward function The calculation formula is:
[0090] ,
[0091] in, , and These are weighting coefficients. and These represent the achieved data rate and the required data rate, respectively. and These represent the energy consumed and the energy threshold, respectively.
[0092] It should be noted that, in complex scenarios such as high-altitude power grid inspection, the simultaneous operation of multiple drones may lead to mutual interference between radio waves. To minimize this interference, the spectrum conflict game model calculates the utility function of each drone node; the utility function of each drone node is as follows:
[0093] ,
[0094] in, Indicates the first Each drone node in its chosen spectrum strategy The utility value below, Indicates the first The transmit power of each drone node, Indicates the first The channel gain of each drone node to itself. For background noise power density, Indicates originating from all remaining drone nodes. ( ) for the The sum of interference at each node, To adjust the parameters, It is the sum of the reciprocals of the frequency intervals, the quantization of the first... The proximity of spectrum resources between a node and other nodes, where i and j are variable indices and are positive numbers;
[0095] Furthermore, the drone adjusts its spectrum strategy according to the called function to achieve the optimal solution under Nash equalization. This method ensures effective reduction of interference even under intensive spectrum usage. Therefore, the process of iteratively solving each utility function using the gradient projection method is as follows:
[0096] ,
[0097] in, In the During the nth iteration, the 1st The spectrum resource selection after the update for each drone node. In the During the nth iteration, the 1st The spectrum resource selection after the update for each drone node. It is the step size parameter, and it is the utility function. For spectrum resources The partial derivatives; This function indicates that the update result will be projected back into the feasible region. The operation ensures that the new spectrum allocation still meets the constraints, where k is the variable index and is a positive number.
[0098] Furthermore, the step size parameter of the gradient projection method Dynamically adjust based on the current channel interference intensity, satisfying:
[0099] ,
[0100] in, Used as the baseline step size.
[0101] S3: Employs an energy-aware topology optimization algorithm to dynamically adjust the network structure based on the energy state of the UAV.
[0102] Furthermore, the energy-aware topology optimization algorithm aims to extend the lifespan of the entire UAV network by predicting node energy consumption and dynamically adjusting the network structure, while ensuring effective mission execution. Its specific steps are as follows:
[0103] S301: Predict the remaining energy of each drone;
[0104] Specifically, an LSTM-based node lifetime prediction model is used for prediction, and the calculation formula is as follows:
[0105] ,
[0106] in, To predict energy consumption using LSTM, For the current time step No. The remaining energy of each node For nodes In time The velocity vector; Represents the time from the initial moment to the current moment. ,node The cumulative velocity square integral is used to quantify the impact of nodal motion on energy consumption;
[0107] The input to the LSTM is a normalized energy consumption index that takes into account the historical energy state of the node and its motion, which is reflected by the velocity square integral.
[0108] S302: Adjust the communication radius between drones based on the predicted remaining energy;
[0109] The specific calculation formula is as follows:
[0110] ,
[0111] in, For communication radius, For the maximum communication radius, This is a critical energy level, i.e., a set threshold energy level. When the energy level of a node falls below this value, special attention needs to be paid to its energy status. The first prediction obtained based on the LSTM model Energy consumption value of each node;
[0112] When the node energy is high, the communication radius is close to its maximum value; when the node energy is close to the critical value, the communication radius decreases significantly, reducing energy consumption.
[0113] S303: Dynamically optimizes the network structure of the entire UAV;
[0114] The optimization process employs a topology optimization objective function, the specific calculation formula of which is as follows:
[0115] ,
[0116] in, Indicates the first The position of each node With the mission location The Euclidean distance between them These are weighting coefficients used to balance the importance of the two optimization objectives. The trace of the network's Laplace matrix reflects the complexity of the network's topology. A larger trace indicates a more dispersed network, while a smaller trace indicates a more compact network. Let i be the total number of nodes, and i be the index of the variable.
[0117] The constraints are: ,and, The total number of nodes. Let be the number of connected components, and let the objective function of this topology optimization represent the rank of the Laplace matrix being equal to the number of nodes minus the number of connected components; this objective function effectively guarantees the connectivity and robustness of the UAV network, that is, if the goal is to maintain the entire UAV swarm as a single connected network ( If this condition ensures that the rank of the Laplace matrix is , then this condition guarantees that the rank of the Laplace matrix is . This means that all eigenvalues except for one zero eigenvalue are positive, thus ensuring good network connectivity.
[0118] The goal of this optimization problem is to find a node selection strategy and network topology that can both efficiently complete the task and extend the network's lifespan. Specifically, it achieves this goal through two steps: prioritizing nodes with sufficient energy and close to the task location, and adjusting the network topology to ensure network connectivity and prevent network splitting due to node energy depletion.
[0119] S4: Employs a biomimetic path planning algorithm to plan the optimal flight route for each drone.
[0120] Specifically, the steps of the biomimetic path planning algorithm are as follows:
[0121] S401: Set an initial flight path for each UAV and calculate the fitness value for each path:
[0122] ,
[0123] in, For the fitness function, For path length, The average signal-to-noise ratio. , and There are 3 weighting coefficients. The sum of the second partial derivatives of the terrain influence function is used to measure the unevenness of the terrain surface. If the sum of the second partial derivatives of the terrain influence function is large, the terrain is more complex or steep, which may have an adverse effect on the flight and communication of drones.
[0124] It is important to note that the fitness function is designed to comprehensively consider multiple key factors in path planning and assign appropriate weights to each factor in order to find the optimal path. Specifically, minimizing path length reduces flight time and energy consumption; the shorter the path, the higher the fitness. Maximizing the average signal-to-noise ratio (SINR) ensures the stability and reliability of data transmission; a higher SINR helps improve communication quality, thus increasing fitness. Considering changes in terrain surface helps avoid selecting paths with complex terrain that may affect flight safety and communication quality; large terrain changes can lead to a decrease in fitness.
[0125] S402: Based on the fitness value of each path, the optimal path for the UAV is determined using the vulture search optimization algorithm.
[0126] ,
[0127] in, Indicates the first A drone in time step new location, Current global best position and For adaptive weights:
[0128] ,
[0129] ,
[0130] and This represents the current iteration number. It represents the total number of iterations. and It is a random number, usually taking values in the range of 1000-1000. Between these, randomness is introduced to increase the diversity of the search. Let be the average position of all individuals in the current population, representing the position of the th individual. A drone in time step Current location The position of an individual randomly selected from the current population;
[0131] The vulture search optimization algorithm consists of three parts: first, guiding the drone to approach the known optimal location; second, encouraging individuals to move closer to the group center to promote group convergence; and third, introducing randomness to maintain the diversity of the search space and avoid getting trapped in local optima.
[0132] S403: When there are obstacles on the optimal path, the repulsive force algorithm based on the potential field method is used to recalculate the path to avoid the obstacles.
[0133] The formula for calculating the repulsive force based on the potential field method is:
[0134] ,
[0135] in, This represents the total repulsive force exerted on the drone by all obstacles. This represents the obstacle threat level; a higher value indicates a more dangerous obstacle or the need for more aggressive avoidance maneuvers. The unit direction vector points towards the drone. Current location Let K be the coordinates of the kth obstacle. For drones The current position coordinates, The gradient of the terrain influence function;
[0136] In addition, for each obstacle, based on its threat level... and A repulsive force is calculated, with the force increasing as the distance increases and decreasing as the distance increases. Considering the impact of terrain on UAV flight, the direction and magnitude of the repulsive force are adjusted using the gradient of the terrain influence function. The total repulsive force is obtained by summing the repulsive forces generated by all obstacles, and is used to adjust the UAV's flight direction in real time, guiding it away from obstacles and avoiding complex terrain. Furthermore, information from the global optimal position, the group average position, and random individual positions is combined, and the exploration-exploration ratio is adjusted through adaptive weighting to help the UAV find the optimal path. The repulsive force of obstacles on the UAV is calculated using the potential field method, taking into account the terrain influence, to ensure the UAV can safely bypass obstacles and maintain a stable flight path.
[0137] Example 2, refer to Figure 3 The second embodiment of the present invention differs from the first embodiment in that:
[0138] As attached Figure 3 An electronic device shown includes:
[0139] Processor, memory, communication interface;
[0140] The memory is used to store the executable instructions of the processor;
[0141] The processor is configured to execute the aforementioned method for self-organizing a drone swarm for high-altitude power grid inspection by executing the executable instructions.
[0142] A readable storage medium storing a computer program, characterized in that, when the computer program is executed by a processor, it implements the above-mentioned method for self-organizing a drone swarm for high-altitude power grid inspection.
[0143] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.
[0144] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-including system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.
[0145] More specific examples of computer-readable media (a non-exhaustive list) include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.
[0146] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.
[0147] Example 3, referring to Figure 2 This is the third embodiment of the present invention, which provides a drone swarm self-organizing network system for high-altitude power grid inspection, including a construction module, a spectrum allocation module, a dynamic adjustment module, and a route planning module;
[0148] Modules are used to construct bidirectional mapping units for the physical-virtual layer;
[0149] The spectrum allocation module uses a dynamic spectrum allocation algorithm to obtain the optimal spectrum strategy for all UAV nodes;
[0150] The dynamic adjustment module uses an energy-sensing topology optimization algorithm to dynamically adjust the network structure based on the energy status of each UAV.
[0151] The route planning module uses a biomimetic path planning algorithm to plan the optimal flight route for each drone.
[0152] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.
Claims
1. A method for self-organizing unmanned aerial vehicle (UAV) swarms for high-altitude power grid inspection, characterized in that: include, A two-way mapping unit for the physical-virtual layer is constructed. The physical layer of the two-way mapping unit consists of N UAV nodes, and the virtual layer of the two-way mapping unit creates a digital twin for each UAV node. The two-way mapping unit constructs a three-dimensional terrain matrix based on the lidar point cloud data of the UAV nodes and establishes a terrain influence function to reflect the impact of terrain on communication quality. Based on the terrain influence function, a dynamic spectrum allocation algorithm is used to obtain the optimal spectrum strategy for all UAV nodes; the optimal spectrum strategy includes, Constructing a reinforcement learning state space ,in, Given the current energy state, For packet loss rate, As a task priority, This is a terrain influence function; Based on the reinforcement learning state space, Q-learning is used to optimize the objective function to obtain the optimal spectrum of each UAV; Based on the optimal spectrum of each UAV, the utility function of each UAV node is defined, and the optimal spectrum strategy of the UAV cluster is obtained by iteratively solving each utility function through the gradient projection method. The Q-learning optimization objective function includes, in, For the reward function, For learning rate, As a discount factor, This serves as the reinforcement learning state space for the next state. It is the spectrum strategy of drones. Spectrum strategy for the drone in the next state; The reward function The calculation formula is: in, , and These are weighting coefficients. and These represent the achieved data rate and the required data rate, respectively. and These represent the energy consumed and the energy threshold, respectively. The utility function of each UAV node is: in, Indicates the first A drone node on its selected spectrum resources The utility value below, Indicates the first The transmit power of each drone node, Indicates the first The channel gain of each drone node to itself. For background noise power density, Indicates originating from all remaining drone nodes. For the The sum of interference from each node; To adjust the parameters, It is the sum of the reciprocals of the frequency intervals, the quantization of the first... The proximity of spectrum resources between a node and other nodes, where i and j are variable indices and are positive numbers; An energy-sensing topology optimization algorithm is used to dynamically adjust the network structure based on the energy state of the UAV. A biomimetic path planning algorithm is used to plan the optimal flight route for each drone.
2. The method for self-organizing UAV swarms for high-altitude power grid inspection as described in claim 1, characterized in that: The bidirectional mapping unit includes, The physical layer consists of the following drones: Each drone is equipped with a lidar and a millimeter-wave communication module, where N is the number of drones and i is the variable index. , and They are the first The three-dimensional coordinates of the drone For the i-th drone; The set of digital twins for the virtual layer is as follows: And synchronize the energy state of physical layer nodes in real time. Signal-to-interference-plus-noise ratio and task timeliness ; For physical layer drones The corresponding digital twin, and These are sets of drones at the physical layer and sets of digital twins at the virtual layer, respectively.
3. The method for self-organizing UAV swarms for high-altitude power grid inspection as described in claim 2, characterized in that: The construction of the three-dimensional terrain matrix includes the physical layer constructing the three-dimensional terrain matrix using lidar point clouds. H, W, and C represent the height, width, and number of channels of the three-dimensional terrain matrix, respectively. Let be the set of real numbers, indicating that each element in the matrix is a real number; The established terrain influence function is as follows: in, It is the number of obstacles. It is the radius of curvature of the terrain. It is the maximum height difference. , and They are the first The three-dimensional coordinates of the obstacle; The virtual layer calculates terrain gradients in real time using a digital twin. And feed it back to the physical layer.
4. The method for self-organizing UAV swarms for high-altitude power grid inspection as described in claim 3, characterized in that: The terrain influence function includes the terrain curvature radius in the terrain influence function. The adjustment is adaptive based on the point cloud density of the lidar, and satisfies the following: in, This represents the terrain complexity coefficient.
5. The method for self-organizing UAV swarms for high-altitude power grid inspection as described in claim 4, characterized in that: The process of iteratively solving the utility function using the gradient projection method is as follows: in, In the During the nth iteration, the 1st The spectrum resource selection after the update for each drone node. In the During the nth iteration, the 1st The spectrum resource selection after the update for each drone node. It is the step size parameter, and it is the utility function. For spectrum resources The partial derivatives; To update the result and project it back into the feasible region, k is the variable index, which is a positive integer.
6. The method for self-organizing UAV swarms for high-altitude power grid inspection as described in claim 5, characterized in that: The energy-sensing topology optimization algorithm includes: Predict the remaining energy of each drone; Adjust the communication radius between drones based on the predicted remaining energy; Dynamically optimize the network structure of the entire drone.
7. The method for self-organizing UAV swarms for high-altitude power grid inspection as described in claim 6, characterized in that: The biomimetic path planning algorithm includes, Set an initial flight path for each drone and calculate the fitness value for each path; Based on the fitness value of each path, the optimal path for the UAV is determined using the vulture search optimization algorithm; When obstacles exist on the optimal path, a repulsive force algorithm based on the potential field method is used to recalculate the path to avoid the obstacles.
8. A drone swarm self-organizing network system for high-altitude power grid inspection, employing the drone swarm self-organizing network method for high-altitude power grid inspection as described in any one of claims 1 to 7, characterized in that, It includes a construction module, a spectrum allocation module, a dynamic adjustment module, and a route planning module; The building module is used to construct a bidirectional mapping unit for the physical-virtual layer; The spectrum allocation module uses a dynamic spectrum allocation algorithm to obtain the optimal spectrum strategy for all UAV nodes. The dynamic adjustment module uses an energy-sensing topology optimization algorithm to dynamically adjust the network structure according to the energy status of each UAV. The route planning module uses a biomimetic path planning algorithm to plan the optimal flight route for each UAV.
9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method for self-organizing a drone swarm for high-altitude power grid inspection as described in any one of claims 1 to 7.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the method for self-organizing a drone swarm for high-altitude power grid inspection as described in any one of claims 1 to 7.