An edge computing service endurance method and system based on a flying charging unmanned aerial vehicle

Abstract

This invention proposes a method and system for extending the range of edge computing services based on in-flight charging drones, comprising a predictive energy gap detection process, a charging drone matching and path planning process, and a flight formation wireless charging process. The predictive energy gap detection process utilizes a task queue-based energy consumption estimation method to predict and proactively schedule charging needs in advance. The charging drone matching process integrates multi-dimensional constraints such as spatial distance, wind field influence, and remaining battery power, and solves the optimal matching scheme by weighting a multi-objective cost function. The wireless charging process establishes a master-slave flight formation to achieve stable control of charging power and employs a dual-mode adaptive wireless charging mechanism, automatically switching between magnetic resonance coupling charging and laser wireless charging to ensure charging efficiency. While maintaining the original flight trajectory of the drones, it achieves non-contact energy replenishment, ensuring uninterrupted operation of the edge computing service.

Description

Technical Field

[0001] This invention relates to the field of edge computing and wireless power transmission for unmanned aerial vehicles (UAVs), specifically to a method and system for extending the flight time of an UAV based on edge computing services for in-flight charging. Background Technology

[0002] With the rapid development of 5G / 6G communication and IoT technologies, the demand for low-latency, high-reliability computing services from ground user terminals (such as smartphones, IoT sensors, and autonomous vehicles) is increasing. Drones, with their advantages of flexible deployment, rapid response, and line-of-sight communication links, are widely used as aerial edge mobile servers to provide ground users with services such as compute offloading, data caching, and real-time inference.

[0003] However, the core constraint of drones lies in their extremely limited onboard battery capacity. Currently, mainstream consumer-grade multi-rotor drones typically have a flight time of only 20 to 40 minutes, and even industrial-grade drones, when equipped with edge computing loads (such as GPUs), usually have a flight time of no more than 1 to 2 hours. Once the battery is depleted, the drone must interrupt service and return to the ground to recharge or replace the battery, leading to the following problems: service interruptions degrade user experience, which is particularly detrimental to latency-sensitive applications (such as autonomous driving collaborative perception and AR / VR rendering offloading); wasted flight energy consumption to and from charging stations can account for 15% to 30% of total energy consumption; the departure of a single drone creates temporary service gaps in its original coverage area; and frequent deep charge-discharge cycles accelerate battery aging.

[0004] The main technological approaches to solving the drone battery life problem and their limitations are as follows:

[0005] (1) Ground-based fixed charging station solution: When the UAV's battery is low, it returns to the ground charging station to charge, but it must leave the service airspace and interrupt the service, resulting in a large amount of ineffective flight energy consumption during the round trip.

[0006] (2) Passive energy capture scheme in the air: The UAV replenishes its power by absorbing solar energy or energy waves emitted by ground base stations. However, the energy source is uncontrollable (solar energy is subject to weather conditions, and the energy waves of the base station decrease sharply with distance). The charging power is low and unstable, and the charging strategy is to "stop all computing services when the power is below a fixed threshold of 20%". During the charging period, all tasks are abandoned and migrated, and the service is completely interrupted.

[0007] (3) In-flight mechanical battery swapping solution: Use an in-flight power supply platform to swap batteries for UAVs, but this requires a complex mechanical docking mechanism, which poses risks of mechanical jamming and collision. The task is also interrupted during the battery swapping process, and only one UAV can be served at a time.

[0008] (4) Wired cable power supply scheme: Power is transmitted from the ground power source step by step through wired cables, but the cables severely restrict the degree of freedom of flight and pose a risk of entanglement and breakage in complex airspace.

[0009] (5) Wireless power extraction from infrastructure: UAVs can collect energy from high-voltage power transmission lines through ultra-low frequency inductive coupling, but this method is heavily dependent on the distribution of specific infrastructure and cannot be applied to any airspace.

[0010] (6) Contact-based in-flight charging solution: UAVs charge in flight through physical contact, but physical contact poses a risk of collision during flight, and mechanical contacts are easily affected by airflow disturbances, making it difficult to guarantee reliability and safety.

[0011] In summary, existing technologies all have fundamental flaws in solving the problem of drone edge computing service endurance: they either interrupt service (return to the ground to charge or replace batteries), have unreliable energy sources (passively capture environmental energy), or severely restrict flight freedom (wired power supply, reliance on infrastructure), and none of them take into account the continuity of the charging process in conjunction with the edge computing task. Summary of the Invention

[0012] To address the aforementioned shortcomings of existing technologies, this invention proposes a method and system for extending the flight range of edge computing services based on a flight-charging UAV. The aim is to enable proactive wireless charging of the service UAV (Support UAV) performing edge computing tasks during flight using a dedicated charging UAV (C-UAV). By combining predictive energy gap detection, joint charging-computing scheduling, and progressive task migration, this invention fundamentally resolves the core contradiction of service interruption during charging in UAV edge computing. The technical solution provided by this invention is as follows:

[0013] In a first aspect, an edge computing service-based battery life system for flight-charging drones includes:

[0014] Multiple service drones, each equipped with an edge computing module, a wireless charging receiver, a communication module, and a GPS positioning module;

[0015] Multiple rechargeable drones, each equipped with a large-capacity energy storage battery, a wireless charging transmitter, a GPS positioning module, and a formation flight control module;

[0016] The ground control center includes an environmental monitoring module, a charging request response module, a charging drone cluster management module, and a charging path planning module.

[0017] And an integrated air-ground communication network connecting service drones, charging drones, and ground control centers, with communication between service drones, charging drones, and ground control centers via cellular networks or proprietary data links.

[0018] Secondly, a method for extending the battery life of a drone based on edge computing services for drones that charge while in flight includes the following steps:

[0019] Register all service drones and charging drones, establish a global drone status database, and set response time intervals; connect to meteorological services and local sensor networks to collect environmental parameters of the operational airspace in real time;

[0020] Based on the existing service drone deployment locations and the estimated charging demand heat map, the ground control center deployed several charging drones to various standby airspaces for low-speed patrol and standby.

[0021] When a service drone detects that its expected power demand is positive, it submits a charging request to the ground control center. The control center adds the service drone that receives the request to the global charging request queue, updates the basic information of all charging drones, and uniformly parses and updates the global status database.

[0022] The ground control center sorts all service drones in the global charging request queue into a task processing priority queue according to their hierarchical order.

[0023] Determine if the number of idle charging drones is greater than or equal to the number of service drones in the task processing priority queue. If it is, move all service drones in the task processing priority queue to the matching queue; otherwise, move service drones in the task processing priority queue that are equal to the number of available charging drones to the matching queue according to task priority.

[0024] Check the task priority of the service drone in each ongoing charging task in turn. If the task priority of the service drone is lower than the highest priority of the service drone in the task processing priority queue, interrupt the charging task, set the charging drone of the charging task to an idle state, add the service drone of the charging task to the task processing priority queue, and move the service drone with the highest priority in the task processing priority queue into the matching queue.

[0025] The service cost of each pair of serving drones and idle charging drones in the matching queue is calculated using a multi-objective cost function; then, the optimal drone pairing is obtained by using an allocation algorithm.

[0026] By combining global state information, environmental parameters, and optimal UAV pairing, a path planning algorithm is used to obtain a globally optimal set of flight paths.

[0027] Each charging drone flies to the target area according to the path in the flight path set, establishes communication with the service drone and forms a master-slave formation, and uses flight control algorithms to always maintain the optimal charging position;

[0028] When the relative positional deviation between the charging drone and the service drone is less than a threshold, precise positioning magnetic resonance coupling charging is used; otherwise, laser wireless charging is used.

[0029] After the charging task is completed, the charging drone detaches from the service drone. If the charging drone's battery level reaches the warning threshold, it returns to the ground; otherwise, it cruises at low speed and remains on standby.

[0030] Preferably, the formula for calculating the expected electricity demand is:

[0031]

[0032] in, This refers to the drone's remaining battery power. It is a safety margin of power. This is the estimated total energy consumption of the task queue, calculated when... When the value is positive, it indicates that the drone is no longer able to complete all tasks and needs to request charging services.

[0033] The preferred hierarchical sorting rule is as follows:

[0034] First-level sorting: by task deadline Increasing sort;

[0035] Secondary sorting: When the time interval between task deadlines is less than or equal to the response time interval, sorting is based on the estimated power demand. Ascending sort; Task deadline The calculation formula is:

[0036]

[0037] in It is the instantaneous total power consumption.

[0038] Preferably, the calculation formula for the multi-objective cost function is as follows:

[0039]

[0040] in, Let represent the Euclidean spatial distance between charging drone i and service drone j. This indicates the remaining battery power of the charging drone. This indicates the full charge capacity of the charging drone i. This represents the correction factor for flight time due to wind field conditions, based on the maximum flight speed of the charging drone. Current wind speed vector Angle with flight path The calculation formula is as follows:

[0041]

[0042] Weight , and Adjustments will be made dynamically based on work performance.

[0043] Preferably, the allocation algorithm is any one of the following: Hungarian algorithm, auction algorithm, minimum cost maximum flow algorithm, greedy algorithm, or Jonker-Volgenant algorithm.

[0044] Preferably, the path planning algorithm is any one of the following: A* algorithm, Dijkstra's algorithm, RRT algorithm, PRM algorithm, ant colony optimization algorithm, or genetic algorithm.

[0045] Preferably, the flight control algorithm is as follows:

[0046] During formation maintenance, the service drone acts as the navigator, and its state vector is defined as follows:

[0047]

[0048] in To serve the three-dimensional spatial coordinates of the drone, For flight speed, The yaw angle is given; the charging drone, acting as a follower, has its state vector defined as:

[0049]

[0050] in The three-dimensional spatial coordinates of the charging drone, For flight speed, Yaw angle; desired formation offset This indicates the desired relative position that the charging drone should maintain with respect to the service drone, determined by the optimal horizontal coupling distance of the wireless charging module. Determined jointly with the current flight direction of the service drone:

[0051]

[0052] To compensate for the vertical installation misalignment between the receiving coil at the bottom of the service drone and the transmitting coil at the top of the charging drone, the control law drives the formation error to converge to zero:

[0053]

[0054]

[0055] The proportional gain matrix determines the convergence speed; The differential gain matrix determines the damping characteristics; and It is a positive definite diagonal matrix, whose diagonal elements adaptively adjust according to the intensity of environmental disturbances: they increase as the disturbance increases. To improve tracking rigidity, the tracking stiffness should be appropriately reduced when the disturbance decreases in order to reduce energy consumption.

[0056] Compared with the prior art, the beneficial effects achieved by the present invention are:

[0057] This invention enables wireless charging during flight, eliminating the need for the S-UAV to return to a ground charging station, while edge computing services remain continuously provided throughout the charging process. Instead of simply halting computing services during charging, a charging-aware, progressive task migration mechanism selectively migrates only tasks exceeding the effective computing capacity after charging degradation, maximizing the preservation of local service quality and resolving the fundamental contradiction of inevitable service interruptions during charging in existing solutions.

[0058] Compared to passive environmental energy harvesting solutions (solar energy, base station energy waves), the C-UAV of this invention carries a large-capacity energy storage battery and actively dispatches it to the S-UAV on demand. The energy source is controllable and the charging power is stable, which can meet the real-time energy requirements of high-power edge computing loads (such as GPU inference and video transcoding). The dual-mode adaptive wireless charging mechanism enables the system to automatically switch between magnetic resonance coupling and laser transmission charging methods according to the flight environment (airflow stability, formation distance), balancing efficiency and reliability.

[0059] This invention employs a predictive energy gap detection mechanism. The S-UAV proactively calculates the estimated stored energy and task deadline based on the current task queue's estimated energy consumption, rather than waiting for the energy level to drop to a fixed threshold before triggering an alarm. This mechanism incorporates the energy required to complete the task into the remaining time calculation, enabling advance prediction and proactive scheduling of charging needs. This avoids the passive predicament of running out of energy and the C-UAV not being able to arrive in time, as is common in fixed-threshold schemes, fundamentally ensuring the safety of task execution.

[0060] The dual-mode adaptive charging mechanism of this invention enables the system to cope with complex airflow environments; the charging preemption and priority scheduling mechanism ensures global energy security constraints in multi-S-UAV scenarios. The system has good scalability and is suitable for wide-area, long-endurance operational scenarios such as urban air traffic, marine monitoring, border patrol, and disaster relief. Attached Figure Description

[0061] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0062] Figure 1This is a schematic diagram of the process of an edge computing service for extending the battery life of a drone based on flight charging, according to the present invention.

[0063] Figure 2 This is a two-dimensional schematic diagram of a basic charging scenario for a single C-UAV in an embodiment of the present invention;

[0064] Figure 3 This is a two-dimensional schematic diagram of a complex request scheduling scenario involving multiple C-UAVs in an embodiment of the present invention. Detailed Implementation

[0065] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0066] To make the above-mentioned objectives, features and effects of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0067] Example 1: A method for extending the battery life of a drone based on edge computing services for in-flight charging, such as... Figure 1 As shown, it includes the following steps:

[0068] S101. Activate the drone edge charging network. All service drones (S-UAVs) and charging drones (C-UAVs) complete registration, report their own device parameters and location coordinates, establish a global drone status database, and set global parameters such as response time interval. .

[0069] S102, the ground control center, based on the existing S-UAV deployment locations and the estimated charging demand heat map, deployed several C-UAVs to various standby airspaces for low-speed patrol and standby.

[0070] S103 connects to meteorological services and local sensor networks to collect environmental parameters of the operational airspace in real time, including wind speed and direction, air temperature and pressure, and electromagnetic interference intensity, and constructs a dynamic environmental model to provide input for subsequent flight time prediction, path planning and energy consumption calculation.

[0071] S104. The control center monitors all S-UAV charging request messages in real time. When any S-UAV detects its estimated power demand... When the value is positive (insufficient to complete all tasks), a charging request is proactively submitted to the control center. The control center adds the received request to the global charging request queue for further processing. Estimated power demand. The calculation formula is:

[0072]

[0073] in, This is the current remaining battery power of the drone. This is a safety margin of power (reserved for safe return to the base station). This is the estimated total energy consumption of the current task queue, calculated when... When the value is positive, it indicates that the drone is currently unable to complete all tasks and needs to request charging services.

[0074] S105. When the control center receives a charging request from an S-UAV, it will proactively update the basic information of all C-UAVs, including location, remaining battery power, and mission execution progress. The control center will then parse this information and update the global status database accordingly.

[0075] S106. For all tasks in the request queue, sort them hierarchically to obtain a task processing priority queue (Priority()). The priority sorting rule is as follows:

[0076] First-level sorting: by task deadline Increasing sort;

[0077] Secondary sorting: Task deadline intervals are less than or equal to According to the required power consumption Ascending sort. Task deadline. The calculation formula is:

[0078]

[0079] in It is the instantaneous total power consumption (including flight power consumption, computing power consumption and communication power consumption).

[0080] S107. Determine if the number of available C-UAVs is greater than the total number of tasks in Priority(); if it is greater than or equal to, proceed to step S112; if it is less than, proceed to step S108.

[0081] S108. Move the S-UAVs in Priority() that are equal to the number of C-UAVs into the Match queue Match() according to their priority.

[0082] S109. Does a C-UAV exist in the current charging network? If not, proceed to step S112; if it does, proceed to step S110.

[0083] S110. Check if the task priorities of all current charging S-UAVs are greater than or equal to the highest priority of the task in Priority(). If they are all greater than or equal to the highest priority of the task in Priority(), proceed to step S112. If there is a task with a priority lower than the highest priority of the task in Priority(), proceed to step S111.

[0084] S111: Interrupt the lowest priority service in the charging S-UAV, set the service's C-UAV to idle, and re-add the S-UAV to Priority(); move the highest priority S-UAV in Priority() into Match(), and proceed to step S109.

[0085] S112. Calculate the service cost for each pair (C-UAC, S-UAV) using the multi-objective cost function Cost(), and then use a matching algorithm to calculate the optimal match. The multi-objective weighted cost function Cost() is:

[0086]

[0087] in, Indicates C-UAV i With S-UAV j Euclidean distance between them Indicates C-UAV i Its remaining battery power Indicates C-UAV i Full charge capacity, This represents the correction factor for flight time caused by the wind field. Since the wind field alters the ground-to-ground speed of the C-UAV, affecting arrival time and flight energy consumption, this invention employs a ground speed ratio model to quantify the wind speed vector into a flight time correction factor. And based on the maximum flight speed of C-UAV Current wind speed vector Angle with flight path The calculation, as a one-dimensional cost function, participates in the matching decision. The calculation formula is:

[0088]

[0089] Weight Adjust dynamically according to work situation: in case of emergency Increase the size of the vehicle, prioritizing the nearest C-UAV to reduce arrival time; when the working environment is harsh. Increase capacity, prioritizing drone safety, especially in emergencies. The reduction allows for responses from more distant C-UAVs.

[0090] S113. Combining global state information, environmental parameters, and the optimal pairing of (C-UAV, S-UAV), a path planning algorithm is used to obtain the globally optimal set of flight paths, Paths.

[0091] S114. Each C-UAV flies to the target area according to the path set Paths, establishes communication with the target UAV, and forms a master-slave formation. A flight control algorithm is used to maintain the optimal charging position at all times. The optimal charging mode is selected based on environmental conditions. The specific flight control algorithm is as follows:

[0092] During formation holding, the S-UAV acts as the navigator, and its state vector is defined as:

[0093]

[0094] in The three-dimensional spatial coordinates of the S-UAV For flight speed, This refers to the yaw angle. The S-UAV maintains its original service flight path and broadcasts its status at fixed intervals.

[0095] As a follower, the state vector of a C-UAV is defined as follows:

[0096]

[0097] in The three-dimensional spatial coordinates of the C-UAV For flight speed, This is the yaw angle.

[0098] Expected formation offset This indicates the desired relative position that the C-UAV should maintain with respect to the S-UAV, determined by the optimal horizontal coupling distance of the wireless charging module. Determined together with the current flight direction of the S-UAV:

[0099]

[0100] This refers to the vertical mounting deviation between the bottom receiving coil of the S-UAV and the top transmitting coil of the C-UAV, which is determined by the model structure. The negative sign indicates that the C-UAV is located behind and below the S-UAV to reduce wake effects while maintaining charging coil alignment.

[0101] The control law is designed as a pilot-follower PD (proportional-derivative) tracking law, driving the formation error to converge to zero:

[0102]

[0103]

[0104] The first term is the positional deviation ratio: it drives the C-UAV to converge relative to the S-UAV offset. The target location The proportional gain matrix determines the convergence speed;

[0105] The second term is the velocity deviation differential term: it suppresses overshoot and oscillation, synchronizing the C-UAV velocity with the S-UAV velocity. The differential gain matrix determines the damping characteristics;

[0106] and It is a positive definite diagonal matrix, whose diagonal elements adaptively adjust according to the intensity of environmental disturbance (gust level): increasing as the disturbance increases. To improve tracking rigidity, the tracking stiffness should be appropriately reduced when the disturbance decreases in order to reduce energy consumption.

[0107] S115, When the relative positional deviation between C-UAV and S-UAV Less than the threshold At that time, the drones can stably maintain the target distance. At this point, precise magnetic resonance coupling charging can be used to achieve higher charging power and efficiency.

[0108] S116, When the relative positional deviation between C-UAV and S-UAV Greater than the threshold At this time, the distance between drones will fluctuate greatly, making it difficult to maintain target spacing. At this point, laser wireless charging with greater positioning tolerance is adopted to maintain a stable charging power while ensuring the safety of the drone.

[0109] S117. Determine whether the C-UAV is interrupted before the task is completed; if the C-UAV is interrupted, proceed to step S111; if it is not interrupted, continue the charging task, and finally proceed to step S119.

[0110] S118. Charging task completed. C-UAV and S-UAV detach. S-UAV returns safely after completing all tasks.

[0111] S119. Has the current C-UAV's battery level reached the warning threshold? If the condition is met, proceed to step S120; otherwise, proceed to step S102.

[0112] S120: The current C-UAV sends a shutdown notification to the control center and then returns safely.

[0113] Figure 2This demonstrates a basic scenario where an S-UAV runs out of power while performing an edge computing task, and a C-UAV provides wireless charging during flight. Specific parameters and procedures are as follows:

[0114] System initialization phase (S101~S103): In a working area of ​​1km×1km×100m (area radius L=500m), the ground control center (GCC) is located at the exact center of the area (500m, 500m), and monitors meteorological data (wind speed 3m / s, southeast wind, temperature 25°C, air pressure 1013hPa) and UAV status information in real time, and sets global parameters: mission response time interval. The GCC is equipped with a C-UAV, a dedicated hexacopter platform for charging, carrying a 400Wh energy storage battery, and a low battery warning value. .

[0115] Initially, an S-UAV equipped with an edge computing module enters the area to perform edge computing tasks at coordinates (150m, 250m) and reports its basic parameters to GCC. Based on the global UAV heat map, GCC estimates the charging needs and dispatches a C-UAV to conduct low-altitude patrols in the designated area and listen for charging requests in real time.

[0116] Calculation and charging request (S104): At a certain moment, the S-UAV located at (300m, 300m, 50m) receives a new calculation task and updates its status information: remaining power. Safety margin The estimated total energy consumption of the current task queue Instantaneous total power consumption (Including flight power consumption of 50W, computing power consumption of 35W, and communication power consumption of 15W), and calculating your own estimated parameters:

[0117]

[0118]

[0119] Due to calculation For positive, This means that a 10Wh energy deficit will occur in about 0.35 hours. In order to ensure mission continuity, the S-UAV immediately sends a charging request to the GCC.

[0120] Status Update and Priority Sorting (S105~S106): Upon receiving the request, the control center immediately establishes a charging task: After sorting the task into two levels of priority based on "deadline 0.35h, required power 10Wh", it adds the task to the Priority() queue. It finds that there is 1 idle C-UAV with coordinates (280m, 580m, 60m) and 300Wh of remaining power, which meets the current task quantity and charging requirements. The task is then moved to the matching queue Match()={S-UAV}.

[0121] Cost Calculation and Matching (S107~S112): Calculate the Cost() cost of the current C-UAV (280m, 580m, 60m) and S-UAV (300m, 300m, 50m).

[0122] Euclidean spatial distance:

[0123]

[0124] Take weight , , ;; , The wind field correction factor is calculated using the ground speed ratio model, and the maximum flight speed of the C-UAV is... Wind speed Angle between wind direction and flight path ,but

[0125]

[0126]

[0127] Since only one pair (C-UAC, S-UAV) is directly confirmed as a match.

[0128] Path planning and formation flying (S113~S114): The path planning algorithm, based on the environmental model and pairing conditions, plans the C-UAV to fly from (280m, 580m, 60m) along the optimal path to the estimated S-UAV position. After the C-UAV reaches (450m, 350m, 50m), it establishes a data link with the S-UAV to form a master-slave formation. The S-UAV maintains its original flight path as the navigator, and the C-UAV acts as the follower, using a PD control law to maintain the formation. The S-UAV's state vector is:

[0129]

[0130] Expected formation offset Based on horizontal offset distance Vertical offset distance Calculations show that the optimal distance between C-UAV and S-UAV is approximately 1 meter.

[0131]

[0132] Dual-mode charging selection (S115-S116): The current wind speed is 3m / s, the airflow is stable, the optimal distance of the formation is 1m, and the distance between S-UAV and C-UAV is maintained at about 1m (within the magnetic resonance coupling charging range of 0.8~1.2m). Therefore, the magnetic resonance coupling charging mode (mode A) is activated, the charging power is maintained at about 200W, and the charging efficiency is about 80%.

[0133] During a certain period of charging, the system encountered strong local gusts of wind, causing significant fluctuations in the formation distance between the S-UAV and C-UAV (<0.8m, >1.2m). The drone charging connection was almost interrupted or a collision occurred, and the charging efficiency plummeted to 40%. At this point, the system actively increased the distance, maintaining the formation distance at around 1.5m (>1m, <2m), and used wireless laser charging, maintaining the charging power at around 120W, with a charging efficiency of approximately 80%.

[0134] Charging complete and return (S117~S120): After 5 minutes of charging, the formation reaches (700m, 250m, 50m), and the S-UAV has obtained 12Wh (120% of the 10Wh power shortage). The negative value was restored, and the C-UAV and S-UAV safely disengaged from formation.

[0135] Since the C-UAV currently has approximately 260Wh of remaining power, it has not reached the warning threshold. Therefore, it was dispatched by GCC again and proceeded to the next dispatch point (800m, 500m, 60m) to continue low-speed cruising.

[0136] The S-UAV no longer accepts new tasks. When it flies to (800m, 200m, 50m), it completes all remaining edge computing tasks and then safely returns to the base station.

[0137] Figure 3 The demonstration showed four C-UAVs distributed in four directions, with two in service and two idle and on standby. Three S-UAVs requested charging almost simultaneously, and one high-priority S-UAV obtained service through a preemption mechanism.

[0138] System initialization (S101~S103): Four C-UAVs are deployed in a 1000m×1000m×100m working area (L=500m), positioned at the top, left, right, and bottom. Current environment: wind speed 4m / s, southerly wind, temperature 23°C, moderate electromagnetic interference level. Both C-UAVs and S-UAVs have completed registration. Global parameters are set: task response time interval. The GCC's C-UAV is a dedicated hexacopter platform for charging, carrying a 400Wh energy storage battery and a low-power warning value. .

[0139] Calculation and charging request (S104): At this time, the one located on the right... Below Left side Almost simultaneously detected The status is positive, and each drone has proactively sent a charging request to the control center. The current status of each S-UAV is as follows:

[0140] Located at (720m, 500m, 50m), remaining power Safety margin Estimated total energy consumption of the task queue Instantaneous total power consumption :

[0141]

[0142] If the value is positive, it indicates that all tasks cannot be completed within the deadline, triggering a charging request that requires approximately 18Wh of additional power.

[0143] Located in the lower area (550m, 270m, 45m), remaining power Safety margin Estimated total energy consumption of the task queue Instantaneous total power consumption =60W:

[0144]

[0145] If the value is positive, a charging request will be sent, requiring approximately 13Wh of additional power.

[0146] Located in the left region (100m, 420m, 50m), it has just received an urgent, computationally intensive task. Remaining battery power... Safety margin Estimated total energy consumption of the task queue , :

[0147]

[0148] It is a positive value, and In just 10 minutes, a charging request is sent immediately, requiring approximately 15Wh of additional power.

[0149] Status Update (S105): After receiving three charging requests, the control center updates the global status database, including weather information and information on each UAV, and prepares to schedule tasks. The UAV information is updated as follows:

[0150] Located in the upper region (500m, 800m, 55m), with a remaining power of 200Wh, currently... Provide charging services ( =8Wh =3min);

[0151] Located in the left area (250m, 500m, 50m), with a remaining battery capacity of 180Wh, currently charging... Provide charging services ( =15Wh, =30min);

[0152] Located in the right area (750m, 400m, 50m), with 250Wh of battery remaining, idle and ready for low-speed cruising;

[0153] Located in the lower region (400m, 400m, 60m), with 150Wh of battery remaining, idle and ready for low-speed cruising.

[0154] The number of available C-UAVs is 2 ( , ).

[0155] Priority sorting (S106): The three charging requests are sorted first-level according to their task deadlines: Then perform a secondary sort: because and The difference of 1 minute is less than the response time interval. Therefore, the local arrangement is performed according to the increasing required power: Therefore, S4 is ranked before S3. Final priority queue .

[0156] Resource determination and preemption (S107~S111): With 2 available C-UAVs, move the first 2 S-UAVs from Priority() into the matching queue, i.e., the highest priority ones. and secondary priority . Priority() remaining Then, proceed to the preemptive judgment process.

[0157] Detected (Serve )and (Serve All are in the service, sorted according to first-level priority:

[0158]

[0159] Right now The level of urgency is higher than If the conditions are met, it is necessary to seize the opportunity. Charging services: Set the status to idle. Re-add to the end of Priority(), then add the highest priority from Priority(). Move into Match().

[0160] Double-check if it's possible to preempt: Enjoying The charging services, after comparison and evaluation, No priority The system is at a high level, so it stops preempting the process and prepares for task matching.

[0161] At this point, Match() contains 3 S-UAVs: Three C-UAVs are available: , quantity matching. Waiting for an idle drone to appear.

[0162] Multi-objective cost calculation and optimal matching (S112): C-UAV maximum flight speed Wind speed Angle between the southerly wind and the flight path The wind field correction factor is:

[0163]

[0164] The following calculates Cost(), to... Calculate the cost for example:

[0165] Coordinates (250, 500, 50) Coordinates (100, 420, 50) L = 500m is the radius of the working area. Weighting is applied. , , Wind field correction factor .

[0166]

[0167]

[0168] by Calculate the cost for example:

[0169] Coordinates (750, 400, 50) Coordinates (720, 500, 50):

[0170]

[0171]

[0172] Similarly, the remaining combinations are calculated to obtain the complete cost matrix:

[0173]

[0174] In this embodiment, the Hungarian algorithm is used as the matching algorithm to obtain the optimal match: , , The total cost is 1.164, which is the global optimum.

[0175] Path planning and formation (S113~S114): Using path planning algorithms for... , , Plan the optimal flight path separately. Fly from left to lower left direction; Fly from the right nearest ; Flying from below The path planning takes into account the influence of a 4 m / s southerly wind. The flight path follows the headwind and flies at low altitude to reduce wind resistance. The three pairs of formations form master and slave formations respectively to carry out recharging missions.

[0176] Dual-mode charging (S115~S116): With a current wind speed of 4 m / s, the drone formation operates smoothly, consistently maintaining a formation distance deviation below the threshold. All three formations utilize magnetic resonance coupling charging mode (Mode A), with charging power set to 180W respectively. ), 150W ), 160W ).

[0177] Completion and Return (S117~S120): (Most urgent task) After charging for 6 minutes, the battery capacity will be replenished to 18Wh. The value returned to negative, and then the aircraft flew towards [location] again according to GCC scheduling. Replenish the remaining required power. After charging for 10 minutes, the battery capacity is replenished to 21.6Wh. After 7 minutes of charging, the battery capacity was replenished to 15.6Wh. Once charging was complete, each C-UAV detached from the formation. Since the remaining battery capacity of each C-UAV was higher than the warning threshold... Once again dispatched by GCC, the S-UAVs proceeded to the target area and cruised at low speed. After replenishing their power, each S-UAV completed its remaining tasks and returned safely. The entire mission was completed without interruption, achieving its objective.

[0178] Example 2: An edge computing service-based endurance system for drones that charge in flight, comprising:

[0179] Multiple service drones (S-UAVs), each equipped with an edge computing module, a wireless charging receiver (including a magnetic resonance receiving coil and a laser photovoltaic cell array), a communication module, and a GPS positioning module;

[0180] Multiple rechargeable drones (C-UAVs) are equipped with a large-capacity energy storage battery, a wireless charging transmitter (including a magnetic resonance transmitting coil and a laser diode array), a GPS positioning module, and a formation flight control module.

[0181] The ground control center (GCC) includes an environmental monitoring module, a charging request response module, a C-UAV cluster management module (maintaining the C-UAV status table and performing preemption scheduling), and a charging path planning module.

[0182] It also includes an integrated air-to-ground communication network that connects S-UAVs, C-UAVs, and ground control centers, with communication between S-UAVs, C-UAVs, and ground control centers via cellular networks or proprietary data links.

[0183] Example 3: The computer-readable storage medium of this example stores a computer program that, when executed by a processor, implements the steps in the edge computing service endurance method based on a flight-charging drone in Example 1.

[0184] The computer-readable storage medium in this embodiment can be an internal storage unit of the terminal, such as the terminal's hard disk or memory; the computer-readable storage medium in this embodiment can also be an external storage device of the terminal, such as a plug-in hard disk, smart memory card, secure digital card, flash memory card, etc. equipped on the terminal; furthermore, the computer-readable storage medium can include both the terminal's internal storage unit and external storage devices.

[0185] The computer-readable storage medium of this embodiment is used to store computer programs and other programs and data required by the terminal. The computer-readable storage medium can also be used to temporarily store data that has been output or will be output.

[0186] Example 4: The computer device of this example includes a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the steps in the edge computing service endurance method based on a flying charging drone in Example 1.

[0187] In this embodiment, the processor can be a central processing unit, or other general-purpose processors, digital signal processors, application-specific integrated circuits, off-the-shelf programmable gate arrays or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor, etc. The memory can include read-only memory and random access memory, and provides instructions and data to the processor. A portion of the memory can also include non-volatile random access memory. For example, the memory can also store device type information.

[0188] Those skilled in the art will clearly understand that each implementation can be achieved using software plus the necessary general-purpose hardware platform, or of course, hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0189] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A flight charging drone-based edge computing service-based extended battery life system, characterized in that, include: Multiple service drones, each equipped with an edge computing module, a wireless charging receiver, a communication module, and a GPS positioning module; Multiple rechargeable drones, each equipped with a high-capacity energy storage battery, a wireless charging transmitter, a GPS positioning module, and a formation flight control module; The ground control center includes an environmental monitoring module, a charging request response module, a charging drone cluster management module, and a charging path planning module. And an integrated air-ground communication network connecting service drones, charging drones, and ground control centers, with communication between service drones, charging drones, and ground control centers via cellular networks or proprietary data links.

2. A method for extending the battery life of an edge computing service based on a flying-charging drone, characterized in that, Includes the following steps: Register all service drones and charging drones, establish a global drone status database, and set response time intervals; It connects to meteorological services and local sensor networks to collect environmental parameters of the operational airspace in real time. Based on the existing service drone deployment locations and the estimated charging demand heat map, the ground control center deployed several charging drones to various standby airspaces for low-speed patrol and standby. When a service drone detects that its expected power demand is positive, it submits a charging request to the ground control center. The control center adds the service drone that receives the request to the global charging request queue, updates the basic information of all charging drones, and uniformly parses and updates the global status database. The ground control center sorts all service drones in the global charging request queue into a task processing priority queue according to their hierarchical order. Determine if the number of idle charging drones is greater than or equal to the number of service drones in the task processing priority queue. If it is greater than or equal to the number of service drones in the task processing priority queue, move all service drones in the task processing priority queue into the matching queue. Otherwise, service drones in the task processing priority queue that are equal to the number of available charging drones will be moved into the matching queue according to task priority. Check the task priority of the service drone in each ongoing charging task in turn. If the task priority of the service drone is lower than the highest priority of the service drone in the task processing priority queue, interrupt the charging task, set the charging drone of the charging task to an idle state, add the service drone of the charging task to the task processing priority queue, and move the service drone with the highest priority in the task processing priority queue into the matching queue. The service cost of each pair of serving drones and idle charging drones in the matching queue is calculated using a multi-objective cost function; then, the optimal drone pairing is obtained by using an allocation algorithm. By combining global state information, environmental parameters, and optimal UAV pairing, a path planning algorithm is used to obtain a globally optimal set of flight paths. Each charging drone flies to the target area according to the path in the flight path set, establishes communication with the service drone and forms a master-slave formation, and uses flight control algorithms to always maintain the optimal charging position; When the relative positional deviation between the charging drone and the service drone is less than a threshold, precise positioning magnetic resonance coupling charging is used; otherwise, laser wireless charging is used. After the charging task is completed, the charging drone detaches from the service drone. If the charging drone's battery level reaches the warning threshold, it returns to the ground; otherwise, it cruises at low speed and remains on standby.

3. The method for extending the battery life of a drone based on edge computing services for in-flight charging, as described in claim 2, is characterized in that... The formula for calculating the estimated electricity demand is as follows: ； in, This refers to the drone's remaining battery power. It is a safety margin of power. This is the estimated total energy consumption of the task queue, calculated when... When the value is positive, it indicates that the drone is no longer able to complete all tasks and needs to request charging services.

4. The method for extending the battery life of a drone based on edge computing services for in-flight charging, as described in claim 3, is characterized in that... The hierarchical sorting rules are as follows: First-level sorting: by task deadline Increasing sort; Secondary sorting: When the time interval between task deadlines is less than or equal to the response time interval, sorting is based on the estimated power demand. Ascending sort; Task deadline The calculation formula is: ； in It is the instantaneous total power consumption.

5. The method for extending the battery life of a drone based on edge computing services for in-flight charging, as described in claim 2, is characterized in that... The formula for calculating the multi-objective cost function is as follows: ； in, Let represent the Euclidean spatial distance between charging drone i and service drone j. This indicates the remaining battery power of the charging drone. This indicates the full charge capacity of the charging drone i. This represents the correction factor for flight time due to wind field conditions, based on the maximum flight speed of the charging drone. Current wind speed vector Angle with flight path The calculation formula is as follows: ； Weight , and Adjustments will be made dynamically based on work performance.

6. The method for extending the battery life of an edge computing service based on a flight-charging drone according to claim 5, characterized in that, The allocation algorithm is any one of the following: Hungarian algorithm, auction algorithm, minimum cost maximum flow algorithm, greedy algorithm, or Jonker-Volgenant algorithm.

7. The method for extending the battery life of a drone based on edge computing services for in-flight charging, as described in claim 5, is characterized in that... The path planning algorithm is any one of the following: A* algorithm, Dijkstra's algorithm, RRT algorithm, PRM algorithm, ant colony optimization algorithm, or genetic algorithm.

8. A method for extending the battery life of an edge computing service based on a flight-charging drone according to claim 2, characterized in that, The flight control algorithm is specifically as follows: During formation maintenance, the service drone acts as the navigator, and its state vector is defined as follows: ； in To serve the three-dimensional spatial coordinates of the drone, For flight speed, The yaw angle is given; the charging drone, acting as a follower, has its state vector defined as: ； in The three-dimensional spatial coordinates of the charging drone, For flight speed, Yaw angle; desired formation offset This indicates the desired relative position that the charging drone should maintain with respect to the service drone, determined by the optimal horizontal coupling distance of the wireless charging module. Determined jointly with the current flight direction of the service drone: ； To compensate for the vertical installation misalignment between the receiving coil at the bottom of the service drone and the transmitting coil at the top of the charging drone, the control law drives the formation error to converge to zero: ； ； The proportional gain matrix determines the convergence speed; The differential gain matrix determines the damping characteristics; and It is a positive definite diagonal matrix, whose diagonal elements adaptively adjust according to the intensity of environmental disturbances: they increase as the disturbance increases. To improve tracking rigidity, the tracking stiffness should be appropriately reduced when the disturbance decreases in order to reduce energy consumption.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps in the edge computing service endurance method for a flight-charging drone as described in any one of claims 2-8.

10. A computer device comprising a processor, a memory, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps in the edge computing service endurance method for a flight-charging drone as described in any one of claims 2-8.

Images