Scheduling method and system for low-airspace unmanned aerial vehicle group with single unmanned aerial vehicle battery changing mother ship for accompanying flight based on rabbet search algorithm

Abstract

The invention provides a scheduling method and system for a low-airspace unmanned aerial vehicle group containing single unmanned aerial vehicle battery changing mother ship accompanying flight based on an on-sun search algorithm, and relates to the technical field of unmanned aerial vehicle group scheduling and energy management. According to the method, breakthrough improvement is realized in the dimensions of cruising ability, environmental adaptability, cooperative efficiency and the like through space-time coupling constraint decoupling and a dynamic battery replacement decision-making mechanism. The method comprises the steps of unmanned aerial vehicle cluster task input and information acquisition; determining a scheduling decision variable of the unmanned aerial vehicle group accompanied by the single unmanned aerial vehicle battery-changing mother ship; determining a scheduling objective function and related constraint conditions of an unmanned aerial vehicle group accompanied by the single unmanned aerial vehicle power conversion mother ship; and according to the determined target function and the constraint condition of the target function, establishing a low-airspace unmanned aerial vehicle group scheduling model containing the single power conversion mother ship accompanying flight, and solving the scheduling model of the unmanned aerial vehicle group containing the single unmanned aerial vehicle power conversion mother ship accompanying flight by adopting an improved rabbet search algorithm based on machine learning. The method is suitable for scheduling and energy management of the unmanned aerial vehicle group.

Description

Technical Field

[0001] The present invention relates to the technical field of scheduling and energy management of unmanned aerial vehicle (UAV) swarms, and particularly to a scheduling method for a UAV swarm with a single UAV battery swapping mother ship flying in the low-altitude airspace based on the transit search algorithm. Background Art

[0002] Under the strategic background of the rapid rise of the low-altitude economy, as the main operators in the low-altitude airspace, the application scenarios of UAVs have expanded from traditional geographical mapping to urban air transportation, logistics distribution, disaster emergency rescue and other fields. However, the existing UAV swarm scheduling technologies face some bottlenecks: firstly, in long-time and long-distance tasks, UAVs are limited by their endurance. Existing static task allocation algorithms do not embed a dynamic charging and swapping scheduling mechanism, resulting in a relatively high interruption rate of long-endurance tasks; secondly, the adaptability to complex dynamic environments is insufficient. Under multi-dimensional interferences such as sudden airspace control and meteorological disturbances, the real-time path planning of traditional heuristic algorithms is reduced and it is difficult to meet the high-precision requirements; thirdly, the current technology fails to make full use of the battery swapping resources of the accompanying mother ship, resulting in resource waste and poor scheduling efficiency; moreover, the cooperative operation mechanism between UAVs and the accompanying mother ship is also lacking, affecting the overall task execution efficiency. It is worth noting that no complete scheduling solution for using the accompanying mother ship to swap batteries for UAV swarms has been seen in the existing technical documents. In particular, no scheduling framework for the integrated cooperative path optimization and dynamic matching of the battery swapping mother ship - UAV swarm has been proposed. These problems need to be solved by improving algorithms and optimizing resource management to improve the scheduling efficiency and task execution ability of UAV swarms. Such scheduling optimization can not only improve the operation efficiency of UAVs, make up for the technical gap in the field of dynamic charging and swapping resource scheduling, but also has significant theoretical value for realizing the safe and green operation of the low-altitude economy. Summary of the Invention

[0003] The object of the present invention is to propose a scheduling method for a UAV swarm with a single UAV battery swapping mother ship flying in the low-altitude airspace based on the transit search algorithm, which significantly improves the scheduling efficiency and endurance of the UAV swarm and effectively solves the problem that the endurance of existing UAVs is limited.

[0004] To achieve the above object, the present invention provides the following technical solutions:

[0005] The present invention proposes a scheduling method for a UAV swarm with a single UAV battery swapping mother ship flying in the low-altitude airspace based on the transit search algorithm, and the method includes the following steps:

[0006] Step S1: Obtain UAV swarm information, task information and battery swapping mother ship information;

[0007] Step S2: Determine the scheduling decision variables of the UAV swarm with a single UAV battery swapping mother ship flying according to the operation tasks of the UAV swarm;

[0008] Step S3: Determine the scheduling objective function of the UAV swarm accompanied by a single UAV power replacement mother ship;

[0009] Step S4: Determine the constraint conditions for the scheduling of the UAV swarm accompanied by a single UAV power replacement mother ship, so as to optimize the scheduling objective function of the UAV swarm;

[0010] Step S5: Establish a scheduling model for the UAV swarm accompanied by a single power replacement mother ship in the low altitude airspace according to the scheduling objective function of the UAV swarm and the constraint conditions of the objective function;

[0011] Step S6: Use the improved transit search algorithm based on machine learning to solve the scheduling model of the UAV swarm accompanied by a single UAV power replacement mother ship according to the scheduling decision variables of the UAV swarm, and obtain the scheduling schemes of the UAVs and the accompanying mother ships.

[0012] Furthermore, the above UAV cluster information includes the UAV set The UAV battery capacity C, the unit time energy consumption e u (v) when the UAV speed is v, the maximum speed of the UAV and the minimum speed The UAV power replacement time Δt;

[0013] The mission information includes the node set Including the starting point O and the ending point D, the set of edges And the edge The distance is d ij The discrete time window set T = (t0, t 0+r , t 0+2r …, t 0+mr ,…T max ), and r is the time window length, the latest mission completion time T max ;

[0014] The power replacement mother ship information includes the unit time energy consumption e(v) when the power replacement mother ship speed is v, the maximum speed V of the power replacement mother ship max .

[0015] Furthermore, the above UAV swarm scheduling decision variables include whether the UAV Passes through the edge Defined as x k,i,j ∈{0, 1}, the time t When the UAV Arrives at the node k,i Whether the UAV Replaces power y In the time window m of the node k,i,m ∈{0, 1}, the flight speed v of the UAV On the edge ​k,i,j Whether the battery swapping mother ship is at the node during the time window m and z m,i ∈ {0, 1}, the battery swapping mother ship is at the node during the time window m Whether it is a drone Battery swapping s k,i,m ∈ {0, 1}.

[0016] Furthermore, through the completion time and total energy consumption of the drone swarm flight mission, the scheduling objective function of the drone swarm - battery swapping mother ship is determined, specifically:

[0017]

[0018] Among them, represents the time when all drones arrive at the end point at the latest, α is the time weight value, β is the energy consumption weight value, represents the total energy consumption of the drone swarm and the battery swapping mother ship, where represents the flight energy consumption of the drone swarm, represents the energy consumption of the battery swapping mother ship moving between time windows, d prev(m,i) represents the moving distance required to move to node i during the time window m.

[0019] Furthermore, the above - mentioned constraints for drone swarm scheduling include path continuity constraints, time continuity constraints including battery swapping time, power constraints, time window matching constraints between the battery swapping mother ship and the drone swarm, service capacity limit constraints of the battery swapping mother ship, residence time constraints of the battery swapping mother ship, moving constraints of the battery swapping mother ship, and task time feasibility constraints.

[0020] Furthermore, the above - mentioned path continuity constraints are used to ensure that the drones complete the task from the starting point to the end point, specifically:

[0021]

[0022] It means that each drone starts from the starting point O and only selects one subsequent node;

[0023]

[0024] It means that each drone finally arrives at the end point D and only selects one previous node to enter;

[0025]

[0026] It means that for intermediate nodes, the number of incoming and outgoing edges is equal to ensure path continuity;

[0027] The time continuity constraints including battery swapping time are used to ensure the time sequence and the impact of battery swapping time, specifically:

[0028]

[0029] Among them, t k,j is the time to reach node j, which needs to satisfy the flight time from i to j and the possible battery swapping time Δt·∑ m y k,i,m ;

[0030]

[0031] M is a maximum constant. When x k,i,j = 0 is a relaxation constraint, and when x k,i,j = 1, it ensures the continuity of time;

[0032] By setting the auxiliary variable Y k,i to determine the battery capacity constraint, the auxiliary variable is:

[0033]

[0034] is the conditional judgment for calculating the battery capacity. Y k,i = 1 means that the UAV k swaps batteries at node i, then the battery capacity is reset to full and then decreased; Y k,i = 0 means no battery swapping, then the battery capacity is decreased;

[0035] During the flight of the UAV swarm, the battery capacity needs to meet the flight mission. The specific constraints are as follows:

[0036]

[0037] If Y k,i = 0, then the battery capacity is decreased. If Y k,i = 1, then the battery capacity is reset to full and then decreased. M ensures that the constraint only takes effect under the corresponding conditions;

[0038]

[0039] is the non - negativity constraint of the UAV battery capacity to ensure the feasibility of the mission;

[0040]

[0041] is the limit on the number of battery swaps for the UAV. Each UAV can swap batteries at most once at the same node i to prevent repeated battery swapping;

[0042] The time window matching constraint between the battery swapping mother ship and the UAV swarm is specifically:

[0043]

[0044] If yk,i,m = 1, the battery swapping time t k,i must be completed within the time window [t0 + m·r, t0 + (m + 1)·r]; M ensures that the constraint only takes effect during battery swapping;

[0045] The battery swapping mother ship service capacity limit constraint is used to limit the service capacity of the mother ship, specifically:

[0046]

[0047] Indicates that within the time window m of node i, the battery swapping mother ship can serve at most times of battery swapping;

[0048] The battery swapping mother ship residence time constraint is specifically:

[0049]

[0050] Indicates that if the battery swapping mother ship is at node i within the time window m, it must stay for at least Δt time to ensure that the battery swapping mother ship has enough time to serve the UAV;

[0051] The battery swapping mother ship movement constraint is specifically:

[0052]

[0053] Among them, Indicates the actual moving distance of the battery swapping mother ship from the previous node i prev to node i, V max ·(r - Δt) represents the maximum movable distance of the battery swapping mother ship between time windows; is the slack term, Indicates the logical judgment value of the continuous residence state of the battery swapping mother ship to ensure that the movement of the battery swapping mother ship conforms to the constraints of speed and time;

[0054] The task time feasibility constraint is specifically:

[0055]

[0056] Ensure that the time t k,D for each UAV k to reach the end point D does not exceed the maximum deadline T max of the task, t k,D refers to the specific time for UAV k to reach the end point, T max is the latest completion time allowed for the task;

[0057]

[0058] Indicates the flight speed v k,i,j of each UAV k on the edge (i, j), and requires the minimum speed And the maximum speed Between

[0059] Furthermore, the above-mentioned step S6 is specifically as follows:

[0060] Step S61: Initial solution generation and fitness evaluation: Randomly generate diverse initial solutions. Each solution contains decision variables, such as path decision variables, time window allocation decisions, UAV speed decision variables, and decision variables for UAV matching with the power exchange mother ship. And each decision variable needs to meet the corresponding constraint conditions, and calculate the fitness of each solution;

[0061] Step S62: Fusion optimization strategy: According to the nature of the decision variables, different algorithms are respectively used to fuse and optimize the decision variables in the solution vector;

[0062] Step S63: Search stage: Analogy the optimization process of the solution to the dynamic adjustment mechanism of planetary orbital motion and photometric change. Use the transit stage to detect the improvement of the solution fitness. If the transit phenomenon occurs, use the planetary stage for global update; otherwise, use the neighbor stage for local search update;

[0063] Step S64: Exploration stage: Further broaden the exploration scope of the solution space for the output solution population in the search stage, and further break the local convergence through multimodal perturbation;

[0064] Step S65: Iteration and decoding: Calculate the fitness of the new solution population to obtain the optimal solution of the current solution population. Repeat the above operation steps until the maximum number of iterations is reached to obtain the solution vector with the lowest fitness value; Decode the optimal solution vector to obtain the scheduling scheme of the UAV and the accompanying mother ship.

[0065] The scheduling method of a UAV group with a single UAV accompanied by a power exchange mother ship in low altitude based on the transit search algorithm described in the present invention can be fully implemented by computer software. Therefore, correspondingly, the present invention also provides a scheduling system of a UAV group with a single UAV accompanied by a power exchange mother ship in low altitude based on the transit search algorithm. The system includes a storage device, and the storage device is used to execute the above-mentioned method and steps.

[0066] The present invention also provides a computer-readable storage medium. A computer program is stored on the computer-readable storage medium, and when the computer program is run by a processor, it executes the scheduling method of a UAV group with a single UAV accompanied by a power exchange mother ship in low altitude based on the transit search algorithm described in any one of the above.

[0067] The present invention also provides a computer device, which includes a memory and a processor. A computer program is stored in the memory. When the processor runs the computer program stored in the memory, the processor executes the scheduling method of a drone swarm with a single drone power exchange mother ship flying in the low-altitude area based on the transit search algorithm described in any one of the above.

[0068] The beneficial effects of the present invention are as follows:

[0069] The present invention proposes a cooperative scheduling method for a flying escort power exchange mother ship - drone swarm based on an improved transit search algorithm. Through the decoupling of spatio-temporal coupling constraints and a dynamic power exchange decision-making mechanism, breakthrough improvements are achieved in dimensions such as endurance, environmental adaptability, and cooperation efficiency. Compared with traditional methods, it effectively solves the problem of the endurance time limit of drones in long-term tasks and provides the possibility for the power supply of drones. At the same time, this scheduling method shows excellent adaptability and accuracy in dynamic and complex environments, optimizes resource utilization, improves the overall task execution efficiency, enhances the cooperative operation mechanism between drones and the mother ship, realizes the Pareto optimality of energy consumption and efficiency, and provides a complete solution with both theoretical innovation and engineering practical value for large-scale drone operations in the low-altitude economic scenario.

[0070] Furthermore, the present invention balances the resource requirements of the power exchange mother ship and drones by setting time window matching, service capacity limitations, and movement constraints, reducing resource waste; also forms a closed-loop guarantee mechanism by setting path, time, and power constraints to ensure task integrity and security; also realizes global cooperation through the constraint linkage between drones and the mother ship (such as time window alignment, power exchange decision-making), shortening the total task time and reducing energy consumption.

[0071] The present invention is applicable to the scheduling and energy management of drone swarms. Description of the Drawings

[0072] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for use in the description of the specific embodiments or the prior art. Obviously, the following drawings are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0073] Figure 1 It is a flowchart of a scheduling method for a drone swarm with a single drone power exchange mother ship flying in the low-altitude area based on the transit search algorithm proposed by the present invention;

[0074] Figure 2It is a flowchart of the scheduling model of the drone swarm with a single drone swapping power mother ship in low altitude solved by the transit search algorithm improved based on machine learning in the present invention, and obtaining the scheduling schemes of drones and the accompanying mother ship;

[0075] Figure 3 It is the power swapping decision scheduling diagram of 10 drones in the present invention;

[0076] Figure 4 It is the power swapping decision scheduling diagram of 20 drones in the present invention;

[0077] Figure 5 It is the power swapping decision scheduling diagram of 40 drones in the present invention;

[0078] Figure 6 It is the performance comparison diagram between the improved transit search algorithm proposed in the present invention and the existing algorithm. Detailed implementation manners

[0079] In the following description, specific details such as specific system structures and technologies are presented for the purpose of illustration rather than limitation, so as to thoroughly understand the embodiments of the present application. However, those skilled in the art should clearly understand that the present application can also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid unnecessary details from interfering with the description of the present application.

[0080] The following further details the specific implementation manners of the present invention in conjunction with the accompanying drawings. The following implementation manners will help those skilled in the art to further understand the present invention, but do not limit the present invention in any form. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several changes and improvements can still be made, and these all belong to the protection scope of the present invention.

[0081] Embodiment 1. Refer to Figure 1 To illustrate this embodiment, this embodiment proposes a scheduling method for a drone swarm with a single drone swapping power mother ship in low altitude based on the transit search algorithm, which significantly improves the scheduling efficiency and endurance of the drone swarm and effectively solves the problem that the endurance of existing drones is limited.

[0082] The scheduling method includes the following steps, as Figure 1 shown:

[0083] Step S1: Obtain drone cluster information, task information, and power swapping mother ship information;

[0084] Step S2: Determine the scheduling decision variables of the drone swarm with a single drone swapping power mother ship according to the operation tasks of the drone cluster;

[0085] Step S3: Determine the scheduling objective function of the UAV swarm accompanied by a single UAV power replacement mother ship;

[0086] Step S4: Determine the constraint conditions for the scheduling of the UAV swarm accompanied by a single UAV power replacement mother ship, so as to optimize the scheduling objective function of the UAV swarm;

[0087] Step S5: According to the scheduling objective function of the UAV swarm and the constraint conditions of the objective function, establish a scheduling model for the UAV swarm accompanied by a single power replacement mother ship in the low-altitude airspace;

[0088] Step S6: Use the improved transit search algorithm based on machine learning to solve the scheduling model of the UAV swarm accompanied by a single UAV power replacement mother ship according to the scheduling decision variables of the UAV swarm, and obtain the scheduling schemes of the UAVs and the accompanying mother ship.

[0089] This embodiment proposes a cooperative scheduling method for the accompanying power replacement mother ship-UAV swarm based on the improved transit search algorithm. Through the decoupling of spatio-temporal coupling constraints and the dynamic power replacement decision-making mechanism, breakthrough improvements are achieved in dimensions such as endurance, environmental adaptability, and cooperation efficiency. Compared with traditional methods, it effectively solves the problem of the UAV endurance time limit in long-term tasks and provides the possibility for the power supply of UAVs. At the same time, this scheduling method shows excellent adaptability and accuracy in dynamic and complex environments, optimizes resource utilization, improves the overall task execution efficiency, enhances the cooperative operation mechanism between UAVs and mother ships, realizes the Pareto optimality of energy consumption and efficiency, and provides a complete solution with both theoretical innovation and engineering practical value for large-scale UAV operations in the low-altitude economic scenario.

[0090] Embodiment 2. Refer to Figures 2 to 5 To illustrate this embodiment, this embodiment specifically describes a scheduling method for a UAV swarm accompanied by a single UAV power replacement mother ship in the low-altitude airspace proposed in the above Embodiment 1;

[0091] Step S1: Obtain UAV cluster information, task information, and power replacement mother ship information;

[0092] Specifically:

[0093] The UAV cluster information includes the UAV set The UAV battery capacity C, the unit time energy consumption e u (v) when the UAV speed is v, the maximum UAV speed and the minimum speed The UAV power replacement time Δt;

[0094] The task information includes the node set Including the starting point O and the ending point D, the set of edges And the edge The distance is d ij , the set of discrete time windows T = (t0, t 0+r , t 0+2r …, t 0+mr , … T max ), and r is the time window length, the latest completion time of the task T max ;

[0095] The information of the battery swapping mother ship includes the energy consumption per unit time e(v) when the speed of the battery swapping mother ship is v, and the maximum speed V of the battery swapping mother ship max .

[0096] Step S2: According to the operation tasks of the UAV cluster, determine the scheduling decision variables of the UAV group with a single UAV accompanied by a battery swapping mother ship;

[0097] Specifically:

[0098] The scheduling decision variables of the UAV group include whether the UAV passes through the edge defined as x k,i,j ∈ {0, 1}, the time t when the UAV arrives at the node , whether the UAV k,i swaps batteries y in the time window m at the node ∈ {0, 1}, the flight speed v k,i,m of the UAV on the edge , whether the battery swapping mother ship is located at the node k,i,j in the time window m and z ∈ {0, 1}, whether the battery swapping mother ship swaps batteries for the UAV m,i at the node in the time window m s k,i,m ∈ {0, 1}.

[0099] Step S3: Determine the scheduling objective function of the UAV group with a single UAV accompanied by a battery swapping mother ship;

[0100] Specifically:

[0101] Considering the completion time T max and the total energy consumption E totla of the UAV group flight task, set the comprehensive objective function:

[0102]

[0103] Among them, max k∈K (t k,D - t0) represents the time (task completion time) when all UAVs arrive at the end point at the latest, and α is the time weight value. It represents the total energy consumption of the UAV swarm, which consists of the flight energy consumption of the UAV swarm and the moving energy consumption of the power exchange mother ship. β is the energy consumption weight value. Specifically, It represents the flight energy consumption of the UAV swarm, that is, the sum of the energy consumption of each UAV on each path (the energy consumption per unit time at speed v × flight time). It represents the energy consumption of the power exchange mother ship moving during the time window, d prev(m,i) It represents the moving distance required to move to node l during the time window m.

[0104] Step S4: Determine the constraint conditions for the scheduling of the UAV swarm with a single-UAV power exchange mother ship in accompaniment, so as to optimize the scheduling objective function of the UAV swarm;

[0105] Specifically:

[0106] To optimize the comprehensive objective function, the scheduling of the UAV swarm with a single-UAV power exchange mother ship in accompaniment satisfies the following constraint conditions;

[0107] Step S4.1 Path continuity constraint: It is used to ensure that the UAVs complete the task from the starting point to the ending point.

[0108]

[0109] It means that each UAV starts from the starting point O and selects only one subsequent node.

[0110]

[0111] It means that each UAV finally reaches the ending point D and selects only one previous node to enter.

[0112]

[0113] It means that for the intermediate nodes, the number of incoming and outgoing edges is equal to ensure path continuity.

[0114] In this embodiment, the path continuity constraint is set. Its function is to ensure the coherence of the UAV path from the starting point to the ending point and avoid path interruption or repeated access to nodes. The effect is to reduce invalid paths: avoid the UAVs getting into an infinite loop or missing key nodes during the task, directly reducing the total flight distance and energy consumption. Task integrity: Ensure that all UAVs complete the task from the starting point to the ending point.

[0115] Step S4.2 Time continuity constraint including power exchange time: Ensure the time sequence and the influence of power exchange time.

[0116]

[0117] Among them, t k,j is the time to reach node j, and it needs to satisfy the flight time from i to j and the possible battery swapping time Δt·∑ m y k,i,m 。

[0118]

[0119] M is a maximum constant. When x k,i,j = 0 is a relaxation constraint, and when x k,i,j = 1, it ensures time continuity.

[0120] This embodiment sets the time continuity constraint. Its function is to ensure the logical consistency of time variables by introducing the influence of time sequence and battery swapping time consumption. The effect is to avoid time conflicts: through the time window constraint, it ensures the reasonable time sequence of the UAV during battery swapping and flight, reducing task delays. Precise scheduling: Combining the battery swapping time consumption, it optimizes the allocation of time resources and supports the optimization objectives related to the total task time or timeliness in the objective function.

[0121] Step S4.3 Battery Constraint:

[0122] Define the auxiliary variable Y k,i :

[0123]

[0124] For the conditional judgment of calculating the battery power, Y k,i = 1 means that UAV k swaps batteries at node i, then the battery power is reset to full charge and then decreases; Y k,i = 0 means no battery swapping, then the battery power decreases.

[0125] The battery power of the UAV swarm needs to meet the flight mission during flight. The specific constraints are as follows:

[0126]

[0127] If Y k,i = 0, then the battery power decreases. If Y k,i = 1, then the battery power is reset to full charge and then decreases. M ensures that the constraint takes effect only under the corresponding conditions.

[0128]

[0129] This is the non - negativity constraint of the UAV battery power to ensure task feasibility.

[0130]

[0131] This is the limit on the number of battery swaps for the UAV. Each UAV can swap batteries at most once at the same node i to prevent repeated battery swapping.

[0132] This embodiment sets a power constraint, which is used to manage the power consumption and battery replacement strategy of the UAV, ensuring that the power always meets the flight requirements. The effect is to prevent power depletion: through battery replacement and consumption calculations, it avoids mission failures caused by insufficient power of the UAV, improving the reliability of the system. Dynamic battery replacement decision: Through the auxiliary variable Y k,i Distinguish between battery replacement and non-battery replacement scenarios, optimize the battery replacement location selection strategy, ensure the orderly battery replacement of the UAV swarm, and improve the system coordination.

[0133] Step S4.4 Battery replacement mother ship and UAV swarm time window matching constraint:

[0134]

[0135] If y k,o,m = 1, the battery replacement time t k,i must be completed within the time window [t0 + m·r, t0 + (m + 1)·r]; M ensures that the constraint only takes effect during battery replacement.

[0136] This embodiment sets a battery replacement mother ship time window matching constraint, which is used to align the UAV battery replacement time with the time window of the mother ship. The effect is efficient resource utilization: ensuring that the UAV only replaces the battery within the available time window of the mother ship, achieving the time matching between the UAV and the battery replacement mother ship during the dynamic flight process. Coordinated scheduling: Supporting the orderly battery replacement of multiple UAVs within the time window, shortening the total waiting time, and directly optimizing the time-related indicators in the objective function.

[0137] Step S4.5 Battery replacement mother ship service capacity limit constraint: Restrict the service capacity of the mother ship.

[0138]

[0139] Indicates that within the time window m of node i, the battery replacement mother ship can serve at most times of battery replacement.

[0140] This embodiment sets a mother ship service capacity limit constraint, which is used to limit the maximum number of battery replacements of the mother ship within a single time window. The effect is load balancing: preventing the mother ship from overloading services within a certain period of time and reducing the risk of service delay.

[0141] Step S4.6 Battery replacement mother ship residence time constraint:

[0142]

[0143] Indicates that if the battery replacement mother ship is at node i within the time window m, it must stay for at least Δt time to ensure that the battery replacement mother ship has enough time to serve the UAV.

[0144] This embodiment sets a constraint on the residence time of the mother ship. Its function is to require the mother ship to stay at the node for enough time to complete the battery swapping service. The effect is service reliability: ensuring the integrity of the battery swapping operation and avoiding battery swapping failure caused by the premature departure of the mother ship.

[0145] Step S4.7 Constraint on the movement of the battery swapping mother ship:

[0146]

[0147] Among them, represents the actual moving distance of the battery swapping mother ship from the previous node i prev to node i, and V max ·(r - Δt) represents the maximum movable distance of the battery swapping mother ship within the time window; is a relaxation term, represents the logical judgment value of the continuous residence state of the battery swapping mother ship, ensuring that the movement of the battery swapping mother ship conforms to the constraints of speed and time.

[0148] This embodiment sets a constraint on the movement of the mother ship. Its function is to limit the moving distance of the mother ship between different time windows. The effect is safety guarantee: restricting the navigation speed of the mother ship and thus the moving distance to avoid accident risks caused by the mother ship's speed exceeding the threshold. Time window connection: ensuring the reachability of the mother ship's position in adjacent time windows and avoiding service interruption caused by excessive movement.

[0149] Step S4.8 Constraint on the feasibility of task time:

[0150]

[0151] Ensure that the time t k,D for each drone k to reach the end point D does not exceed the maximum deadline T max of the task, where t k,D refers to the specific time for drone k to reach the end point, and T max is the latest completion time allowed for the task.

[0152]

[0153] represents the flight speed v k,i,j of each drone k on the edge (i, j), and it is required to be between the minimum speed and the maximum speed .

[0154] Step S5: Establish a scheduling model for the UAV swarm accompanied by a single battery swapping mother ship in the low altitude airspace according to the scheduling objective function of the UAV swarm and the constraint conditions of the objective function;

[0155] Step S6: Solve the scheduling model of the UAV swarm with a single UAV power replacement mother ship in flight using a transit search algorithm improved based on machine learning to obtain the scheduling plans for the UAVs and the accompanying mother ships.

[0156] Specifically:

[0157] As Figure 2 shown:

[0158] Step S61: Initial solution generation and fitness evaluation: Randomly generate diverse initial solutions. Each solution contains decision variables, such as path decision variables, time window allocation decisions, UAV speed decision variables, and decision variables for UAV matching with the power replacement mother ship. And each decision variable needs to meet the corresponding constraint conditions, and calculate the fitness of each solution;

[0159] Step S62: Fusion optimization strategy: According to the nature of the decision variables, different algorithms are used to fuse and optimize the decision variables in the solution vector respectively;

[0160] Step S63: Search stage: Analogy the optimization process of the solution to the dynamic adjustment mechanism of planetary orbital motion and photometric change. Use the transit stage to detect the improvement of the solution fitness. If the transit phenomenon occurs, use the planetary stage for global update; otherwise, use the neighbor stage for local search update;

[0161] Step S64: Exploration stage: Further broaden the exploration scope of the solution space for the output solution population in the search stage, and further break the local convergence through multi-modal perturbation;

[0162] Step S65: Iteration and decoding: Calculate the fitness of the new solution population to obtain the optimal solution of the current solution population. Repeat the above operation steps until the maximum number of iterations is reached to obtain the solution vector with the lowest fitness value; Decode the optimal solution vector to obtain the scheduling plans for the UAVs and the accompanying mother ships.

[0163] More specifically:

[0164] A1 galaxy stage: Initial solution generation and fitness evaluation.

[0165] A1.1 Initial solution generation, generate diverse initial solutions that meet the constraint conditions to lay the foundation for global search.

[0166] Randomly generate ns initial solutions. The set of solutions is defined as S = {S1, S2, …, S ns}, where each solution S p contains decision variables p is the index of the solution.

[0167] The generation of path decision variables meets the constraint conditions:

[0168]

[0169] Among them, It represents that the probability of the drone k choosing to navigate along the edge (i, j) is inversely proportional to the distance d ij ; in addition, the sum of probabilities from node i to all possible subsequent nodes j' is 1 to ensure the path continuity of the drone.

[0170] Generation of time window allocation decision variables :

[0171]

[0172] Among them, It represents the time when the drone k arrives at node i, which follows a uniform distribution defined on the interval (t0 + m·r, t0 + (m + 1)·r - Δt), and unbiased exploration of the search space is ensured through random sampling.

[0173] Initialization of speed decision variables satisfies the condition:

[0174]

[0175] Among them, It represents the navigation speed of the drone k on the edge (i, j), which follows a uniform distribution defined on the interval .

[0176] Generation of decision variables for the matching of the drone and the power exchange mother ship satisfies the condition:

[0177]

[0178] Among them, It represents whether the drone k exchanges power within the time window m at node i. If it exchanges power, the value is 1, and at the same time satisfies otherwise it is with a value of 0, and at the same time is randomly generated by the random probability p swap and the arrival time whether it is within the time window. Among them, p swap is the power exchange probability, represents the time when the drone k arrives at node i, rand() is a random number generator used to generate a uniform random number within [0, 1], t0 + m·r is the start time of the time window m, and t0 + (m + 1)·r - Δt is the end time of the time window m minus the power exchange time consumption.

[0179] A1.2 Fitness evaluation of solutions: Evaluate the quality of each solution to guide subsequent optimization.

[0180] Calculate the fitness F(S of each solution in set S p )

[0181]

[0182] where F(S p ) is the weighted sum of the time cost and energy cost to complete the task, and α and β are trade-off coefficients. Further, a penalty term is introduced to eliminate infeasible solutions:

[0183]

[0184] where μ is the penalty coefficient. When the power or time , add a penalty term to ensure the rationality of the solution.

[0185] A2 Fusion optimization strategy.

[0186] Input population S and output population S′.

[0187] Perform fusion optimization on the existing set of solutions S to improve the quality of the solutions and reduce unnecessary spatial exploration. According to the nature of the decision variables, different algorithms are used to perform fusion optimization on the decision variables in the solution vector as follows:

[0188] A2.1 Improve the path and speed decision variables through crossover and fusion, respectively:

[0189] Crossover optimization of path decision variables.

[0190] Based on the optimal solution S in the solution set S obtained from A1 * , introduce the crossover probability and update the path decision variable of each solution S in the solution set S p :

[0191]

[0192] where represents the new path decision variable of drone k, rand() is a random function used to generate a uniform random number in [0,1], p

[0193] c is the crossover probability, is the path decision variable of drone k in the optimal solution S * , and p is the path decision of drone k in the current solution S.

[0194] Fusion optimization of speed decision variables.

[0195] Based on A1, obtain the optimal solution S in the solution set S * , introduce a fusion factor to update each solution S in the solution set S p 's speed decision variable:

[0196]

[0197] Among them, represents the new flight speed of the drone k on the route (i,j), the optimal solution S * the flight speed of the drone k in is the current solution S p the flight speed of the drone k in, and δ is the fusion factor.

[0198] A2.2 Local optimization of the battery swapping time window based on LSTM prediction improvement.

[0199] Considering the discreteness of the battery swapping time window, make full use of the space search efficiency and accuracy of the machine learning algorithm, and further introduce the LSTM algorithm to optimize and improve the battery swapping time window:

[0200]

[0201] Among them, m new represents the best battery swapping time window of the drone k at node i in the new solution. LSTM() is a long short-term memory network, which can map the input features (time, battery level, distance) to the time window index output prediction value, is the time when the drone k arrives at node l, is the current battery level of the drone, represents the total remaining path distance. LSTM predicts the best battery swapping time window m based on the current time the current battery level and the total remaining path distance. new .

[0202] A3. Search stage.

[0203] Analogize the solution optimization process to the dynamic regulation mechanism of planetary orbital motion and photometric change, and use the transit stage (A3.1) to detect the fitness improvement of the solution (that is, refine the population S' from A2). If the transit phenomenon occurs, use the planetary stage (A3.2) for global update; otherwise, use the neighbor stage (A3.3) for local search update.

[0204] A3.1 Transit stage.

[0205] For the candidate solution population S'={S'1,S'2,…,S' ns}Perform fitness evaluation, and quantitatively analyze the relative quality of the solution through photometric calculation, such as L b :

[0206]

[0207] where d b represents the distance between the solution S′ p and the current optimal solution. For each solution vector S′ in the solution population S′ p calculate the initial photometric value After that, perform a small random perturbation on S′ p to obtain a new corresponding solution vector S″ p , and recalculate the photometric value If it indicates that a transit phenomenon occurs and enters the planetary stage; otherwise, enter the neighbor stage.

[0208] A3.2 Planetary stage.

[0209] Perform global refinement on the solution vector S″ detected with transit, and simulate the orbital adjustment of the planet affected by the gravitational force of the star. Combine the current optimal solution S p with the perturbed solution S″ * by weighted fusion to generate S p as follows: z as follows:

[0210]

[0211] In the formula, c1 is the gravitational balance coefficient, which controls the contribution weight of the historical optimal solution; R L is the random perturbation force, which simulates the uncertainty of celestial body motion. Based on the galaxy center S r (the optimal solution S of the initial population 0 or the current optimal solution S * ), further adjust the specific solution vector as shown in the following formula:

[0212]

[0213] where g ∼ {1, 2, 3}) represents randomly selecting the perturbation mode; c2 and c3 represent the perturbation amplitude coefficients, which determine the search range.

[0214] A3.3 Neighbor stage.

[0215] Perform refined adjustment on the perturbed solution S″ not detected with transit to simulate the local orbital fine-tuning driven by the stellar radiation pressure, and combine the perturbed solution S″ p with the galaxy center S p to generate an intermediate solution, as shown in the following formula: r as follows:

[0216]

[0217] In the formula, c4 is the centroid offset coefficient, which controls the conservatism of local search. Further, based on the intermediate solution and the galaxy center, the corresponding solution vector is further adjusted as shown in the following formula:

[0218]

[0219] where c5 and c6 are the local perturbation step sizes. After performing the transit judgment and perturbation update on the initial solution population S′ = {S′1, S′2, …, S′ ns}, a new solution population S″′ = {S″′1, S″′2, …, S″′ ns} is output

[0220] A4. Exploration stage.

[0221] Finally, for the solution population S”' output in the search stage, the exploration range of the solution space is further broadened, and the local convergence is further broken through multi-modal perturbation to enhance the traversal ability of the solution space. Based on a certain solution vector S″′ p output in A3 and the galaxy center, multiple random perturbation terms are introduced to calculate S″′ p as shown in the following formula:

[0222]

[0223] where c7 is the stellar gravitational attenuation coefficient, which controls the retention ratio of historical solutions; c8 and c9 are perturbation factors, c8 is the benchmark intensity, and c9 is the distance attenuation factor. P ∼ U(0, 1) is the random perturbation factor. h ∼ {1, 2, 3, 4} is the conditional variable. After applying multi-modal perturbation to the solution population S″′ = {S″′1, S″′2, …, S″′ ns} output in the search stage, a new solution population S”” = {S″″1, S″″2, …, S″″ ns} is generated.

[0224] A5. Calculate the fitness of the new solution population according to the above step A1.2 to obtain the optimal solution minF(S””) of the current solution population. Repeat the operation steps A2 - A5 until the maximum iteration number maxgen is reached to obtain the solution vector with the lowest fitness value; decode the optimal solution vector to obtain the scheduling scheme of the UAV and the accompanying mother ship, as Figures 3 to 5 shown.

[0225] Embodiment 3. A scheduling method for a drone swarm with a single drone power - swapping mother ship flying in the low - altitude area based on the transit search algorithm proposed in the above - mentioned embodiments can be fully implemented by computer software. Therefore, correspondingly, this embodiment proposes a scheduling system for a drone swarm with a single drone power - swapping mother ship flying in the low - altitude area based on the transit search algorithm. The system includes:

[0226] A storage device for obtaining drone cluster information, mission information, and power - swapping mother ship information;

[0227] A storage device for determining scheduling decision variables of a drone swarm with a single drone power - swapping mother ship flying according to the operation tasks of the drone cluster;

[0228] A storage device for determining the scheduling objective function of a drone swarm with a single drone power - swapping mother ship flying;

[0229] A storage device for determining the constraint conditions for the scheduling of a drone swarm with a single drone power - swapping mother ship flying, so as to optimize the scheduling objective function of the drone swarm;

[0230] A storage device for establishing a scheduling model for a drone swarm with a single power - swapping mother ship flying in the low - altitude area according to the scheduling objective function of the drone swarm and the constraint conditions of the objective function;

[0231] A storage device for solving the scheduling model of a drone swarm with a single drone power - swapping mother ship flying by using a transit search algorithm improved based on machine learning according to the scheduling decision variables of the drone swarm, and obtaining the scheduling schemes of the drones and the accompanying mother ships.

[0232] Embodiment 4. This embodiment provides a computer - readable storage medium. A computer program is stored on the computer - readable storage medium. When the computer program is run by a processor, it executes the scheduling method for a drone swarm with a single drone power - swapping mother ship flying in the low - altitude area based on the transit search algorithm described in any one of the above - mentioned embodiments.

[0233] Embodiment 5. This embodiment provides a computer device. The device includes a memory and a processor. A computer program is stored in the memory. When the processor runs the computer program stored in the memory, the processor executes the scheduling method for a drone swarm with a single drone power - swapping mother ship flying in the low - altitude area based on the transit search algorithm described in any one of the above - mentioned embodiments.

[0234] A computer device provided by this embodiment. The hardware device in this part is of a general model and is not shown in the form of a diagram. The system includes a processor and a memory. The processor and the memory can be connected through a bus or other means. The memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs, non-transitory computer-executable programs, and modules, as well as corresponding program instructions / modules. The processor executes various functional applications and data processing of the processor by running the non-transitory software programs, instructions, and modules stored in the memory, so as to implement the scheduling method and steps of the drone swarm with a single drone power-changing mother ship flying in the low-altitude area based on the transit search algorithm in the above method embodiment.

[0235] Embodiment Six. Refer to Figure 6 Describe this embodiment. This embodiment compares the transit search algorithm improved based on machine learning with the existing algorithms proposed in the above embodiments;

[0236] By setting different numbers of drone swarms, the transit search algorithm improved based on machine learning is respectively compared with the genetic algorithm and the simulated annealing algorithm. The comparison metrics are respectively the quality of the solution (average fitness), the convergence time (in seconds), and the number of iterations (average).

[0237] After calculation, the results are as Figure 6 shown. It can be seen from the figure that based on the number of 10 drones, the average value of the solution of the improved transit search algorithm is 38186296.57, the convergence time is 7.24, and the number of iterations is 38.10. Compared with the genetic algorithm and the simulated annealing algorithm, it is the best in terms of the quality of the solution, the convergence time, and the number of iterations. Further, on the basis of increasing the number of drones to 20 and 40, it can be seen that as the number of drones increases, the improved transit search algorithm is still superior to the genetic algorithm and the simulated annealing algorithm in terms of the quality of the solution, the convergence time, and the number of iterations.

[0238] Those skilled in the art can understand that the above are only the preferred embodiments of the present invention. The features described in each embodiment and / or claim of the present disclosure can be combined or combined in various ways, even if such combinations or combinations are not explicitly recorded in the present disclosure. It is not used to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, for those skilled in the art, they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

[0239] Although the preferred embodiments of the present invention have been described, additional changes and modifications can be made to these embodiments by those skilled in the art once they learn the basic creative concept. Therefore, the appended claims are intended to be construed to include the preferred embodiments as well as all changes and modifications that fall within the scope of the present invention. Obviously, those skilled in the art can make various changes and variations to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention is also intended to include these modifications and variations.

Claims

1. A scheduling method for a drone swarm with a single drone swapping power and accompanied by a mother ship in low airspace based on the transit search algorithm, characterized in that, The method is as follows: S1: Obtain the information of the UAV cluster, mission information, and the information of the power exchange mother ship; S2: Determine the scheduling decision variables of the UAV group with a single-UAV power exchange mother ship accompanying flight according to the operation tasks of the UAV cluster; S3: Determine the scheduling objective function of the UAV group with a single-UAV power exchange mother ship accompanying flight; S4: Determine the constraint conditions for the scheduling of the UAV group with a single-UAV power exchange mother ship accompanying flight, so as to optimize the scheduling objective function of the UAV group; S5: Establish a scheduling model of the UAV group with a single power exchange mother ship accompanying flight in the low-altitude airspace according to the scheduling objective function of the UAV group and the constraint conditions of the objective function; S6: Use the transit search algorithm improved based on machine learning to solve the scheduling model of the UAV group with a single-UAV power exchange mother ship accompanying flight according to the scheduling decision variables of the UAV group, and obtain the scheduling schemes of the UAVs and the accompanying mother ships.

2. The scheduling method of a drone swarm with a single drone swapping power on a low-altitude aircraft carrier in companion flight based on the transit search algorithm according to claim 1, characterized in that, The information of the UAV cluster includes the set of UAVs The battery capacity C of the UAV, and the energy consumption per unit time e when the UAV speed is v u (v), the maximum speed of the UAV and the minimum speed The power change time Δt of the UAV; The task information includes a set of nodes including a starting point O and an ending point D, and a set of edges and the edge has a distance of d ij , and a set of discrete time windows T = (t0, t 0+r , t 0+2 r…, t 0+mr , … T max ), where r is the time window length and T is the latest completion time of the task max ; The information of the battery swapping mother ship includes the energy consumption per unit time \(e(v)\) when the speed of the battery swapping mother ship is \(v\), and the maximum speed \(V\) of the battery swapping mother ship max .

3. The scheduling method of a drone swarm with a single drone power exchange mother ship flying in the low-altitude area based on the transit search algorithm according to claim 2, characterized in that, The scheduling decision variables of the UAV swarm include UAVs Whether it passes through the edge Defined as x k,i,j ∈{0, 1}, the time t when the UAV arrives at the node , the time window m of the UAV k,i at the node Whether to change the battery y ∈{0, 1}, the flight speed v of the UAV k,i,m at the edge , whether the battery-changing mother ship is at the nodev k,i,j at the time window m and z m,i ∈{0, 1}, the battery-changing mother ship at the time window m at the node Whether it is for the UAV To change the battery s k,i,m ∈{0, 1}.

4. A scheduling method for a drone swarm with a single drone swapping power carrier flying in low altitude airspace based on the transit search algorithm according to claim 1, characterized in that, Determine the scheduling objective function of the UAV group - power exchange mother ship through the completion time and total energy consumption of the UAV group flight tasks, specifically: Among them, represents the time when all UAVs reach the end point at the latest, α is the time weight value, and β is the energy consumption weight value. represents the flight energy consumption of the UAV swarm. represents the energy consumption of the power exchange carrier moving within the time window, and d prev(m,i) represents the moving distance required to move to node i at time window m.

5. The scheduling method of a drone swarm with a single drone power replacement mother ship flying in the low-altitude area based on the transit search algorithm according to claim 1, characterized in that The constraint conditions for the scheduling of the UAV group include path continuity constraint, time continuity constraint including power exchange time consumption, power constraint, time window matching constraint between the power exchange mother ship and the UAV group, service capacity limit constraint of the power exchange mother ship, residence time constraint of the power exchange mother ship, movement constraint of the power exchange mother ship, and task time feasibility constraint.

6. The scheduling method of a UAV group with a single-UAV power exchange mother ship accompanying flight in the low-altitude airspace based on the transit search algorithm according to claim 5, characterized in that Path continuity constraint: Time continuity constraint including power exchange time consumption: where t k,j is the time to reach node j, is the flight time from i to j, Δt·∑ m y k,i,m is the battery swapping time, and M is a large constant; By setting the auxiliary variable Y k,i Determine the power quantity constraint, auxiliary variables: The power constraint of the UAV group when performing flight tasks is specifically: Time window matching constraint between the power exchange mother ship and the UAV group: Service capacity limit constraint of the power exchange mother ship: Residence time constraint of the power exchange mother ship: Movement constraint of the power exchange mother ship: Among them, represents the actual moving distance of the power exchange mother ship from the previous node i prev to node i, and V max ·(r - Δt) represents the maximum movable distance of the power exchange mother ship during the time window; is a relaxation term, represents the logical judgment value of the continuous residence state of the power exchange mother ship to ensure that the movement of the power exchange mother ship conforms to the constraints of speed and time; Task time feasibility constraint:

7. The scheduling method of an unmanned aircraft group with a single unmanned aircraft power replacement mother ship flying in a low-altitude area based on the transit search algorithm according to claim 6, characterized in that S6 is specifically: S61: Initial solution generation and fitness evaluation: Randomly generate diverse initial solutions, each solution contains decision variables, such as path decision variables, time window allocation decisions, UAV speed decision variables, decision variables for the matching of UAVs and the power exchange mother ship, and each decision variable needs to meet the corresponding constraint conditions, and calculate the fitness of each solution; S62: Fusion optimization strategy: According to the nature of the decision variables, different algorithms are respectively used to fuse and optimize the decision variables in the solution vector; S63: Search stage: Analogy the optimization process of the solution to the dynamic adjustment mechanism of planetary orbital motion and photometric change, use the transit stage to detect the improvement of the fitness of the solution, if the transit phenomenon occurs, use the planetary stage for global update; otherwise, use the neighbor stage for local search update; S64: Exploration stage: Further broaden the exploration scope of the solution space for the output solution population in the search stage, and further break the local convergence through multi-modal perturbation; S65: Iteration and decoding: Calculate the fitness of the new solution population, obtain the optimal solution of the current solution population, repeat the above operation steps until the maximum number of iterations is reached, and obtain the solution vector with the lowest fitness value; decode the optimal solution vector to obtain the scheduling schemes of the UAVs and the accompanying mother ships.

8. A scheduling system for a drone swarm with a single drone for in-flight battery swapping accompanying an aircraft carrier in low-altitude airspace based on the transit search algorithm, characterized in that, The system includes a storage device for performing the method and steps described in claim 1.

9. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium, and when the computer program is run by a processor, it executes a scheduling method for a drone swarm with a single drone power replacement carrier flying in the low-altitude area based on the transit search algorithm described in any one of claims 1-7.

10. A computer device, characterized in that, The device includes a memory and a processor. A computer program is stored in the memory, and when the processor runs the computer program stored in the memory, the processor executes a scheduling method for a drone swarm with a single drone power replacement carrier flying in the low-altitude area based on the transit search algorithm described in any one of claims 1-7.

Images