Automatic replenishment and transportation communication method for unmanned aerial vehicle cluster

Abstract

The invention discloses an unmanned aerial vehicle cluster automatic replenishment and transportation communication method, which realizes efficient task allocation and conflict resolution through state information interaction between unmanned aerial vehicle nodes and supply stations. The method has the advantages of efficient task allocation and execution, strong conflict resolution capability, real-time dynamic adjustment, optimized communication protocol, accurate state perception and management, flexible expansibility, high reliability and safety and the like.

Description

Technical Field

[0001] This invention relates to an automatic replenishment and transportation communication method for unmanned aerial vehicle (UAV) swarms. Background Technology

[0002] With the rapid development of automation and intelligent technologies, autonomous mobile drone swarms are playing an increasingly significant role in logistics, rescue, and industrial applications. Particularly in the field of material transportation, drone swarms are valued for their flexibility, adaptability to complex terrain, and potential for distributed collaboration. Typical application scenarios include: in remote areas or urban environments, large swarms of drones collect supplies from multiple resupply stations, deliver them to designated service points (such as warehouses, hospitals, or mobile recipients), and dynamically adjust transportation strategies based on real-time demand.

[0003] Traditional drone airlift systems typically rely on centralized scheduling, with a central server coordinating the paths and task allocation of all nodes. However, this model faces significant challenges in dynamic, high-load, or communication-constrained environments. Centralized systems are prone to efficiency degradation due to single points of failure or network latency, and struggle to quickly adapt to sudden changes in service point demand or real-time fluctuations in supply station inventory. Furthermore, existing communication protocols (such as ZigBee, MQTT, or ROS-based protocols) are primarily designed for low-bandwidth or static network scenarios, lacking support for distributed autonomous decision-making, real-time material matching, and high-priority tasks.

[0004] To address these challenges, this invention proposes a communication method specifically designed for automated replenishment and transportation of unmanned aerial vehicle (UAV) swarms. This method achieves efficient and reliable material transportation through distributed state management, dynamic resource matching, and a decentralized conflict resolution mechanism, making it particularly suitable for applications in dynamic environments and those with high real-time requirements. Summary of the Invention

[0005] This invention provides an automatic replenishment and transportation communication method for drone swarms to address the problems existing in the prior art.

[0006] The technical solutions adopted in this invention are as follows:

[0007] An automated replenishment and transportation communication method for drone swarms includes the following steps:

[0008] (1) The UAV node sends a resource request message containing UAV node status information to the supply station;

[0009] (2) The supply station receives and processes resource request messages and sorts the requests according to the timestamp in the UAV node status information using the first-in-first-out principle;

[0010] (3) The UAV node receives feedback information sent by the supply station. If the feedback result is that its resource request can be met, it performs material loading operation, updates its own status information and sends status update information to the control center and the receiving point. The supply station updates its inventory record synchronously. If the feedback result is that the resource request cannot be met, it selects another supply station to send the resource request or waits for the supply station to update its inventory before making the request.

[0011] (4) The control center generates and sends task instructions to the drone nodes based on the material demand information submitted by the receiving point and the status information of the drone nodes and supply stations. After receiving the task instructions, the drone nodes execute the corresponding tasks according to the instructions and provide real-time feedback of their status information to the control center during the task execution process.

[0012] (5) During the transportation process, the UAV node makes autonomous decisions based on its own status information and task requirements, combined with the surrounding environment information and the interaction information with other UAV nodes, and dynamically adjusts the flight path and speed according to the comprehensive scoring algorithm. When there is a queue conflict at the supply station or a conflict in the allocation of materials, the UAV node uses the local status table to assess the possibility of the queue conflict, prioritizes the supply station with a waiting time within the preset threshold, and compares its own request time to resolve the conflict problem.

[0013] Furthermore, in step (2), if the request times are the same, the drone identifiers are sorted from smallest to largest according to their lexicographical order to determine whether the drone node's resource request can be satisfied, and the processing result is notified to the drone node and the control center through status broadcast.

[0014] Furthermore, the status information of the drone node includes: the three-dimensional coordinate position of the drone node, battery level, load capacity, current load, movement speed, and timestamp;

[0015] The status information of the supply station includes: the three-dimensional coordinates of the supply station, inventory, queue length, service capacity, and availability flag.

[0016] Furthermore, in step (4), the material demand information includes the material type, quantity, and delivery deadline.

[0017] Furthermore, the communication protocol between the UAV nodes, supply stations, control centers, and receiving points uses Protocol Data Units (PDUs) for information exchange. The PDUs include the following fields:

[0018] Message type, sender identifier, receiver identifier, PDU timestamp, interaction information, and cyclic redundancy check (CRC) for verifying data integrity.

[0019] Furthermore, the comprehensive scoring algorithm selects the most suitable individual or group to perform the task based on the comprehensive score of the drone nodes. The comprehensive score is calculated by weighting the risk score and energy consumption score, and the formula for calculating the comprehensive score is as follows:

[0020] S=w R ∙R +w E ∙E ,

[0021] Where: R: Standardized risk score, calculated using the following formula:

[0022] ,

[0023] Where, d: distance from the drone to the mission target; dmax: maximum possible distance in the mission scenario, used for standardization; Bmax: maximum battery capacity; Cmax: maximum load capacity; P1, P2, P3: weighting coefficients of various risk factors; C MDID : Represents the load capacity of a drone node, i.e., the maximum weight of supplies a drone can carry; B MDID : Represents the battery level of the drone node;

[0024] E: Standardized energy expenditure score, calculated using the following formula:

[0025] ,

[0026] Where w: weight of materials required for the task, a, b, c: coefficients of various energy consumption factors, w R Risk weight, w E : The weight of energy consumption, and w R +w E =1;

[0027] The standardized formula for calculating risk and energy consumption scores is as follows:

[0028] ,

[0029] ,

[0030] Among them, R min ,R max The minimum and maximum possible values ​​of the risk score;

[0031] Emin, Emax: The minimum and maximum possible values ​​of the energy expenditure score.

[0032] Furthermore, in step (5), a conflict resolution mechanism is used to handle queuing conflicts at supply stations and conflicts in the allocation of supplies. The conflict resolution mechanism includes the following rules:

[0033] (1) The supply station maintains a dynamic request queue to record the resource requests of the UAV and their arrival times;

[0034] (2) According to the requested arrival time T PDU The drones are sorted according to the first-in, first-out principle. If multiple drones request the same time, they are sorted in ascending order according to the lexicographical order of the drone identifier MDID.

[0035] (3) The supply station has a service capacity of S SSID The upper limit is set, and requests are processed in queue order. After processing, the current queue length Q of the supply station is updated. SSID And notify the drone cluster via status broadcast;

[0036] (4) If the waiting time of the drone in the queue exceeds the threshold Tt imeout The supply station sent a rejection response, and the drone needs to select another supply station.

[0037] (5) The UAV uses its local state table to assess the possibility of queuing conflicts and dynamically adjusts its request behavior; it calculates the expected waiting time T at the supply station. wait If T wait Exceeding the preset threshold T threshold The drone chose another resupply station;

[0038] (6) In the event of a conflict in the allocation of supplies, if the supply station's inventory I SSID If there is a deficiency, a conflict notification is broadcast; after receiving the conflict notification, the drone compares its own request time T. PDU The drone that requests resources earliest will be allocated resources first. If the request times are the same, the drones will be sorted lexicographically by their MDID identifier.

[0039] Furthermore, when the drone uses its local state table to assess the likelihood of queuing conflicts, it calculates the expected waiting time T at the supply station according to the following formula. wait ,

[0040] ,

[0041] If T wait ≤T threshold The drone sent a request to the supply station;

[0042] If T wait >T threshold The drone selects another resupply station from the status table.

[0043] The distance d between the drone and the supply station can be calculated using the following formula:

[0044] ,

[0045] Among them: Q SSID The current queue length at the supply station;

[0046] S SSID The service capacity of a supply station, i.e., the number of drones it can serve per minute;

[0047] Lat MDID Lon MDID Alt MDID These are the latitude, longitude, and altitude of the drone, respectively.

[0048] Lat SSID Lon SSID Alt SSID These are the latitude, longitude, and altitude of the supply station, respectively.

[0049] The present invention has the following beneficial effects:

[0050] 1. Through autonomous decision-making algorithms, UAVs can independently establish the most suitable grouping strategy for the task based on comprehensive scores and execute group maneuvers, improving the efficiency and accuracy of task allocation. The adoption of a decentralized conflict resolution mechanism effectively handles supply station queuing conflicts and resource allocation conflicts, ensuring the efficiency and fairness of resource acquisition and task execution.

[0051] 2. Drone nodes and resupply stations can update and broadcast status information in real time, improving the cluster's coordination capabilities and task execution accuracy. It supports large-scale cluster operations, adapting to logistics and replenishment needs of varying scales, and exhibits excellent scalability. Detailed Implementation

[0052] This invention discloses an automatic replenishment and transportation communication method for unmanned aerial vehicle (UAV) swarms, comprising the following steps:

[0053] (1) The UAV node sends a resource request message containing UAV node status information to the supply station;

[0054] (2) The supply station receives and processes resource request messages and sorts the requests according to the timestamp in the UAV node status information using the first-in-first-out principle;

[0055] (3) The UAV node receives feedback information sent by the supply station. If the feedback result is that its resource request can be met, it performs material loading operation, updates its own status information and sends status update information to the control center and the receiving point. The supply station updates its inventory record synchronously. If the feedback result is that the resource request cannot be met, it selects another supply station to send the resource request or waits for the supply station to update its inventory before making the request.

[0056] (4) The control center generates and sends task instructions to the drone nodes based on the material demand information submitted by the receiving point and the status information of the drone nodes and supply stations. After receiving the task instructions, the drone nodes execute the corresponding tasks according to the instructions and provide real-time feedback of their status information to the control center during the task execution process.

[0057] (5) During the transportation process, the UAV node makes autonomous decisions based on its own status information and task requirements, combined with the surrounding environment information and the interaction information with other UAV nodes, and dynamically adjusts the flight path and speed according to the comprehensive scoring algorithm. When there is a queue conflict at the supply station or a conflict in the allocation of materials, the UAV node uses the local status table to assess the possibility of the queue conflict, prioritizes the supply station with a waiting time within the preset threshold, and compares its own request time to resolve the conflict problem.

[0058] The entities and roles corresponding to this invention are as follows. Based on the functional descriptions of the entities and roles, the method of this invention will be further explained.

[0059] 1: Drone Node MD.

[0060] The drone node (MD) is an autonomous mobile robot responsible for retrieving supplies from supply stations, transporting them to service points, and making distributed decisions during the transportation process. It maintains and updates its local status table, recording information such as supply station inventory and its own battery level.

[0061] The drone node MD possesses independent computing, communication, and navigation capabilities, and can optimize mission execution based on local status tables and control center instructions. It also exchanges status, requests, and feedback with supply stations, the control center, and other drone node MDs.

[0062] The unique identifier MDID is a 64-bit unsigned integer generated based on the device's hardware serial number (such as IMEI) to ensure global uniqueness. It is assigned by the manufacturer and stored in non-volatile memory during the firmware initialization of the drone node MD. It is used for communication addressing, task allocation, and conflict resolution.

[0063] The key attributes of the drone node MD are shown in Table 1.

[0064] Table 1

[0065] <![CDATA[Three-dimensional coordinate position: P MDID =(Lat MDID , Lon MDID , Alt MDID )]]> <![CDATA[Latitude Lat MDID , 4-byte floating point number, unit: degree; Longitude Lon MDID , 4-byte floating point number, unit: degree; Altitude Alt MDID , 4-byte floating point number, unit: meter; Obtain data through the built-in GPS or Beidou module of MD, with an accuracy of ±5 meters, used to calculate the distance to the supply station or service point]]> <![CDATA[Electric quantity B MDID > Percentage (0-100%, 1-byte unsigned integer); obtained in real-time by the MD Battery Management System (BMS), with an error of ±1%. Used to influence the weighting of refueling station selection. <![CDATA[Load capacity C MDID > Maximum carryable weight, in grams, 4-byte fixed value, stored in firmware. Used to constrain cargo loading. <![CDATA[Current load L MDID > Current weight of carried supplies, in kilograms, 4-byte fixed-point digital (MD) sensor, updated at mission start / end, with an error of ±0.1 kg. Remaining load capacity is calculated using this information. <![CDATA[Moving speed V MDID > Average moving speed, in meters per second, 1 byte of fixed-point data <![CDATA[T status ]]> Status timestamp, 8 bytes of UTC time in milliseconds, indicating the data generation time, NTP synchronization. <![CDATA[T PDU ]]> Resource request message timestamp, 8 bytes UTC time, in milliseconds. Used for conflict optimization mechanisms.

[0066] Interactive information:

[0067] Drone Status: S MDID =(MDID,P MDID B MDID C MDID ,L MDID V MDID ,Tstatus );

[0068] Current status of supplies carried by the drone: M MDID = (MDID,Cargo list , T cargo ) ;

[0069] Cargo list ={(Ti,Qi,L) MDID )|i=1,2,…,N};

[0070] Data interaction rules:

[0071] 1) The drone broadcasts S to the control center every 5 seconds. MDID Event-driven transmissions can be triggered when the battery level drops below 30% or the location changes by more than 100 meters. Resource requests are sent to supply stations, and status broadcasts, including inventory and queuing information, are received.

[0072] 2) Nearby drones acquire S via point-to-point communication MDID It is used for collaborative task allocation, and the control center receives it to monitor the cluster status.

[0073] 3) After loading or unloading supplies, the drone sends an M to the control center or service point. MDID The supply station updates its inventory records simultaneously, and the service point receives the records to confirm delivery.

[0074] 2: Supply Station SS.

[0075] Supply Stations (SS) are fixed or mobile supply points for materials and resources, providing loading and charging services for drone nodes (MD) and periodically broadcasting status information. Supply Stations (SS) possess inventory management and communication capabilities, maintaining local queues and material records.

[0076] The unique identifier SSID is a 48-bit unsigned integer generated based on the last 6 bytes of the MAC address. It is assigned during the initialization of the SS network interface at the data source supply station and stored in a local database.

[0077] The key attributes of the Supply Station SS are shown in Table 2.

[0078] Table 2

[0079] <![CDATA[Three-dimensional coordinate position P SSID =(Lat SSID ,Lon SSID , Alt SSID )]]> <![CDATA[Latitude Lat SSID , 4-byte floating point number, unit: degree; Longitude Lon SSID , 4-byte floating point number, unit: degree; Altitude Alt SSID , 4-byte floating point number, unit: meter; Obtain data through the built-in GPS or Beidou module of the SSID, the SS built-in GPS module (fixed SS) or the dynamic positioning system (mobile SS), update frequency 1Hz]]> <![CDATA[Inventory: I SSID ={SSID, (Ti, Qi, Ui) | i = 1, 2, …, N}]]> Ti: Material Type ID (2-byte unsigned integer, e.g., 0x0001 represents medicine). Qi: Material Quantity (4-byte unsigned integer, unit: piece); Ui: Last Update Time, 8-byte UTC timestamp, unit: milliseconds; Data source: SS Inventory Management System, updated in real time. Used for MD material matching and supply station selection. <![CDATA[Queue length Q SSID > The current number of MDs in the queue (2-byte unsigned integer). Data source: SS request counter, updated in real time. ETA calculation and supply station selection weights used to influence MDs. <![CDATA[Service capacity: S SSID > Number of MDs that can be served per minute (2-byte unsigned integer). Data source: SS design specification, fixed value. Used to calculate queue waiting time. <![CDATA[Availability flag A SSID > 1 byte, 0x00 is unavailable, 0x01 is available <![CDATA[T status ]]> Status timestamp, 8-byte UTC timestamp in milliseconds, indicating the data generation time, NTP synchronization.

[0080] Interactive information:

[0081] S SSID =(SSID,P SSID Q SSID ,S SSID A SSID ,I SSID Tstation ).

[0082] Data interaction rules:

[0083] The supply station periodically (every 10 seconds) broadcasts S SSID To the nearest drone and control center, or to trigger a broadcast state S when the queue changes. SSID The drone and control center receive S SSID Update the local state table to handle supply requests.

[0084] 3: Control Center (CC).

[0085] The Control Center (CC) is a high-level management entity of the system, responsible for task allocation, cluster monitoring, and dynamic coordination, but does not directly interfere with the MD's path planning or supply station selection.

[0086] The unique identifier CCID is a 32-bit unsigned integer generated based on the server instance ID. It is assigned by the system administrator during control center CC initialization. It is used for task broadcasting and status feedback addressing.

[0087] The key attributes of the Control Center (CC) are shown in Table 3.

[0088] Table 3

[0089] <![CDATA[Task ID ]]> The task's unique hash, 16 bytes, is generated using SHA-1. <![CDATA[SP ID ]]> Target service point ID, a 48-bit unsigned integer. <![CDATA[D j :]]> <![CDATA[Material requirements, in the same format as I SSID > <![CDATA[P j :]]> Priority: 1-byte unsigned integer, 1 is the highest, 5 is the lowest. <![CDATA[Task list T CC > <![CDATA[{(Task ID ,SP ID ,D j ,P j )∣j=1,2,…,M}]]> <![CDATA[Monitoring data M CC > {(MDID,Status),(SSID,Status)}

[0090] Interactive information:

[0091] C CCID =(CCID,Type,Content,Tcommand)

[0092] Type: Instruction type (1 byte, 0x01 task allocation, 0x02 forced instruction)

[0093] Triggering drone task execution Type=0x01: ((Task) ID SP ID D j ,P j ))

[0094] Adjust cluster behavior Type=0x02:(Action,Params), such as rerouting or pausing.

[0095] Data interaction rules:

[0096] Control Center Broadcast C CCID Send to all drones, or send to a specific MDID.

[0097] UAV receives C CCID To execute tasks or instructions.

[0098] 4: Receiver point SP.

[0099] The receiving point (SP) is the point of receipt for material delivery. It can be static (such as a fixed warehouse) or dynamic (such as mobile personnel). It submits requests to the CC and receives materials delivered by the MD.

[0100] The unique identifier SPID is a 48-bit unsigned integer generated based on the device MAC address or registration ID. It is allocated during SP initialization and stored locally. It is used for target addressing in task instructions.

[0101] The key attributes of the receiving point SP are shown in Table 4.

[0102] Table 4

[0103] <![CDATA[Three-dimensional coordinate position P SPID = (Lat SPID , Lon SPID , Alt SPID )]]> <![CDATA[Same format as P MDID， The GPS module of the SP (dynamic SP) or fixed configuration (static SP). Calculate the transport distance from MD to SP]]> <![CDATA[Requirement D SPID > <![CDATA[{(SPID,T i Q i Deadline i A list of material requirements for each i=1,2,…,K, including type Ti, quantity Qi, and delivery deadline. i 8-byte UTC timestamp. Data comes from SP's demand management system and is updated in real time. Used for task allocation and MD resource matching.

[0104] Data interaction rules:

[0105] Submit requirement D to CC SPID Receive task status.

[0106] Note: Entity IDs 1-4 do not use sequential encoding, but use differential encoding.

[0107] 5: Protocol architecture.

[0108] Protocol Data Unit (PDU) Definition: PDU APP =(Type,Source ID Dest ID ,T PDU Payload, CRC);

[0109] Type: Message type (1-byte unsigned integer, 0x01 status broadcast, 0x02 task instruction, 0x03 resource request, 0x04 confirmation feedback, 0x05 request submission).

[0110] SourceID: Sender identifier, a 64-bit unsigned integer, compatible with MDID (64-bit), SSID, and SPID (48-bit). Empty spaces are filled with specific bytes.

[0111] DestID: Receiver identifier (64-bit CCID, 32-bit CCID, 0xFF for broadcast).

[0112] T PDU : PDU timestamp, 8-byte UTC timestamp, in milliseconds, sender clock, NTP synchronization.

[0113] Payload: Interactive information (variable length, serialized S) MDID M MDID ,SSSID C CCID D SPID ).

[0114] CRC: Checksum, 4 bytes, CRC32.

[0115] Interaction scenarios:

[0116] 0x01: The drone broadcasts the SMDID, and the supply station broadcasts the SSSID, DestID=0xFF.

[0117] 0x02: Control center sends C CCID ,DestID=0xFF or MDID.

[0118] 0x03: Drone sends M MDID ,DestID=SSID.

[0119] 0x04: Supply station / receiving point sends feedback, DestID=MDID or CCID

[0120] 0x05: Receiver sends D SPID ,DestID=CCID.

[0121] Data encoding rules:

[0122] Encodes payload, supports variable-length data, Payload=TLV List .

[0123] TLV List ={(Tag i ,Length i Value i )|i=1,…,N}:

[0124] Tag i Data type (2 bytes, 0x0001S) MDID 0x0002M MDID 0x0003S SSID 0x0004C CCID 0x0005D SPID ).

[0125] Length i : Length 2 bytes.

[0126] Value i Serialize interactive information (such as the SMDID field).

[0127] Network layer:

[0128] It uses IPv6 addresses, with a route TTL of 1 byte and a maximum value of 255. If a connection to the peer is lost after 255 attempts or 255 hops, the peer is considered disconnected, marked as disabled, and removed from the cluster.

[0129] 6. Independent decision-making.

[0130] Autonomous decision-making is the core capability of the automated replenishment and transportation communication protocol for drone swarms, aiming to enable drone MDs to independently execute tasks and achieve dynamic coordination in a distributed environment. Within a drone swarm, drones autonomously select the most suitable individuals or groups to perform the task, demonstrating the autonomous decision-making nature of the drone swarm.

[0131] Single drone mission: Select the drone with the smallest overall score S from the cluster as the mission executor.

[0132] Multi-drone mission: Select the top k drones with the smallest overall score S from the cluster to form a mission execution team.

[0133] Overall score S=w R ∙R +w E ∙E

[0134] R: Standardized risk score;

[0135] E: Standardized energy expenditure score;

[0136] w R Risk weighting;

[0137] w E : The weight of energy consumption, and w R +w E =1;

[0138] ,

[0139] d: Distance from the drone to the mission target, expressed as P MDID and target location (e.g., P) SPID It was calculated.

[0140] dmax: The maximum possible distance in the task scenario, used for standardization.

[0141] Bmax: Maximum battery level, set to 100% (full charge percentage).

[0142] Cmax: The maximum load capacity, determined based on the maximum value of all CMDIDs in the drone cluster.

[0143] p1, p2, p3: Weighting coefficients of each risk factor, which can be adjusted according to task requirements.

[0144] ,

[0145] w: Weight of materials required for the task, derived from Qi in the task requirement DSPID, unit is C. MDID Consistent.

[0146] a, b, c: Coefficients of various energy consumption factors, which can be adjusted according to the performance of the drone.

[0147] ,

[0148] ,

[0149] Rmin, Rmax: Minimum and maximum possible values ​​of the risk score; Emin, Emax: Minimum and maximum possible values ​​of the energy consumption score.

[0150] 7: Conflict resolution mechanism.

[0151] In automated replenishment and transportation systems for drone swarms, conflict is an unavoidable challenge in distributed collaboration, especially in resource-competitive scenarios. A decentralized resolution mechanism is employed to ensure the efficiency and fairness of resource acquisition and task execution.

[0152] Common types of collisions in drone swarms include:

[0153] Supply station queuing conflict: Multiple drones (MDs) simultaneously send resource requests to the same supply station (SS), causing the queue length Q at the supply station to increase. SSID The service capacity is insufficient to meet all requests simultaneously.

[0154] Resource allocation conflict: Multiple drones request limited resources from the same supply station, resulting in insufficient resources to meet all requests, which may lead to uneven distribution or mission failure.

[0155] The supply station maintains a dynamic request queue that records the resource requests from drones and their arrival times.

[0156] 8: Queue management is based on the following rules.

[0157] Sort by: Request arrival time (T) PDU The drones are sorted according to the First-In-First-Out (FIFO) principle using 8-byte UTC timestamps. If multiple drones request the same time (with a difference of less than 10 milliseconds), they are sorted in ascending order according to the drone identifier MDID (64-bit unsigned integer).

[0158] Service allocation: Supply stations are allocated based on service capacity S SSIDThe maximum number of drones that can be served per minute (a 2-byte unsigned integer) is used to process requests in queue order. The queue length Q is updated after processing is complete. SSID (2-byte unsigned integer), and notify the cluster via status broadcast (Type=0x01, every 10 seconds or when the queue changes).

[0159] Timeout mechanism: If the drone waits in the queue for more than the threshold T timeout (Protocol preset, e.g., 60 seconds), the supply station sends a rejection feedback (Type=0x04, Payload contains the rejection reason, Dest). ID =MDID), the drone needs to choose another resupply station.

[0160] The implementation process is as follows:

[0161] The drone sends a resource request (MMDID, Type=0x03, DestID=SSID, SourceID=MDID).

[0162] The supply station receives the request, updates the queue, and records T. PDU Supply stations are sorted by FIFO and MDID, and resources are allocated accordingly.

[0163] The supply station broadcast has been updated to S SSID , including Q SSID The system notifies the drone queue status. If a rejection is received, the drone reassesses other resupply stations.

[0164] Optimization of drone state awareness:

[0165] The drone utilizes a local status table (which stores the S of the supply station) SSID Assess the likelihood of queuing conflicts and dynamically adjust request behavior:

[0166] 1) Waiting Time Assessment: The expected waiting time for the drone to reach the resupply station is as follows:

[0167] ,

[0168] 2) Supply station selection:

[0169] If T wait ≤Tthreshold (protocol preset, e.g., 30 seconds), the drone sends a request to the resupply station.

[0170] If T wait >Tthreshold, the drone selects another resupply station from the status table, based on distance d and inventory I. SSID .

[0171] ,

[0172] Comparison of drone mission requirements D SPID T in i、 Q i With supply station inventory I SSID T in i、 Q i Prioritize supply stations that meet your needs.

[0173] 9: Resolution of conflicts in the allocation of resources.

[0174] 1) The drone sends a resource request M MDID Including demand T i、 Q i Record the request time T PDU .

[0175] 2) The supply station receives the request and checks the inventory. SSID If Qi is insufficient, broadcast a conflict notification (Type=0x04, Payload contains T). i、 Q i Shortage information, Dest ID =0xFF).

[0176] 3) The drone receives the conflict notification and compares its own request time T. PDU The drone that requests supplies earliest will be given priority in allocation.

[0177] 4) If the request times are the same (the difference is less than 10 milliseconds), sort them in ascending order by MDID lexicographical order.

[0178] 5) Confirmation of the allocation of the winning drone to the supply station (Type=0x04, Dest) ID =MDID. Other drones can select other resupply stations from the status table or wait for inventory updates.

[0179] To reduce conflicts in resource allocation, supply stations update inventory levels in real time. SSID And broadcast it.

[0180] 10: Performance optimization.

[0181] In automated replenishment and transportation systems for drone swarms, performance optimization is key to ensuring efficient, reliable, and scalable operation. For example:

[0182] Delay management: Minimize communication latency in task allocation, resource requests, and conflict resolution to ensure real-time response.

[0183] Bandwidth optimization: Reduces bandwidth usage for broadcast and peer-to-peer communication, improving network efficiency.

[0184] To address the above issues, the following strategies are adopted:

[0185] 1) Adaptive broadcast frequency

[0186] In the current design, the resupply station broadcasts every 10 seconds, and the drone broadcasts every 5 seconds. To reduce latency and adapt to dynamic changes, a state-based adaptive adjustment is introduced:

[0187] Supply station: If the queue length Q SSID If the Qi value increases by more than 2 or decreases by more than 10% in the inventory ISSID, a broadcast is immediately triggered to update the Tstation. If there is no significant change, the interval is extended to 15 seconds.

[0188] Drone: If battery B MDID If BMDID is less than 30%, the broadcast interval is increased from 5 seconds to 10 seconds. If BMDID is greater than or equal to 30%, the interval remains at 5 seconds.

[0189] 2) Fast request and response

[0190] If Q SSID SSID (Queue not full), the supply station immediately sends a confirmation feedback Type=0x04, Dest ID =MDID.

[0191] Supply Station Pre-allocation Broadcast S SSID At that time, if QSSID=0 and inventory I SSID Meeting demand Q i For Qi with ≥DSPID, add a pre-allocation flag (1 byte, 0x01 indicates immediate allocation) to the Payload.

[0192] 3) Bandwidth optimization

[0193] The PDU's payload uses TLV (Tag-Length-Value) encoding. To reduce bandwidth usage, TLV encoding is optimized for dynamic encoding: only changing interactive information is transmitted. For example, a supply station broadcasts S... SSID When, if I SSID Some T i Q i No changes, only changes, reduces TLV List length.

[0194] 11: Security mechanisms.

[0195] Communication security mechanisms are crucial for ensuring system reliability and data trustworthiness. Existing technologies have many network communication security algorithms that can be used directly and are not within the scope of this invention.

[0196] ​The above description is only a preferred embodiment of the present invention. It should be noted that those skilled in the art can make several improvements without departing from the principle of the present invention, and these improvements should also be considered within the scope of protection of the present invention.

Claims

1. An unmanned aerial vehicle cluster automatic replenishment and transportation communication method, characterized in that: The method comprises the following steps: (1) The unmanned aerial vehicle node sends a resource request message including the state information of the unmanned aerial vehicle node to the supply station; (2) The supply station receives and processes the resource request message, and sorts the requests according to the time stamps in the state information of the unmanned aerial vehicle node according to the first-in first-out principle; (3) The unmanned aerial vehicle node receives the feedback information sent by the supply station, and if the feedback result can meet the resource request, the unmanned aerial vehicle node performs a material loading operation, updates the state information, and sends the state update information to the control center and the receiving point, and the supply station synchronously updates the inventory record; If the feedback result cannot meet the resource request, the unmanned aerial vehicle node reselects other supply stations to send resource requests or waits for the supply station inventory to be updated before making a request; (4) The control center generates a task instruction to the unmanned aerial vehicle node according to the material demand information submitted by the receiving point and in combination with the state information of the unmanned aerial vehicle node and the supply station, and the unmanned aerial vehicle node receives the task instruction and executes the corresponding task according to the instruction content, and feeds back the state information to the control center in real time during the task execution; (5) The unmanned aerial vehicle node makes autonomous decision according to the comprehensive scoring algorithm during the transportation process according to the state information and task demand of the unmanned aerial vehicle node, the surrounding environment information and the interaction information with other unmanned aerial vehicle nodes, dynamically adjusts the flight path and speed, and when a queue conflict or a material allocation conflict occurs, the unmanned aerial vehicle node evaluates the possibility of the queue conflict by using the local state table, and preferentially selects a supply station with a waiting time within a preset threshold, and compares the request time to solve the conflict problem.

2. The UAV cluster automatic restocking and transportation communication method of claim 1, wherein: In step (2), if the request time is the same, the unmanned aerial vehicle identifiers are sorted in ascending order according to the dictionary order to determine whether the resource request of the unmanned aerial vehicle node can be met, and the processing result is notified to the unmanned aerial vehicle node and the control center by state broadcasting.

3. The UAV cluster automatic restocking and transportation communication method of claim 1, wherein: The state information of the unmanned aerial vehicle node includes the three-dimensional coordinate position, the power, the load capacity, the current load, the moving speed and the time stamp of the unmanned aerial vehicle node; The state information of the supply station includes the three-dimensional coordinate position, the inventory, the queue length, the service capacity and the availability flag of the supply station.

4. The UAV swarm automatic restocking and transportation communication method of claim 1, wherein: In step (4), the material demand information includes the type, the quantity and the delivery deadline of the material.

5. The UAV swarm automatic restocking and transportation communication method of claim 1, wherein: The communication protocol between the unmanned aerial vehicle node, the supply station, the control center and the receiving point adopts a protocol data unit (PDU) for information interaction, and the PDU includes the following fields: message type, sender identifier, receiver identifier, PDU time stamp, interaction information and cyclic redundancy check (CRC) for checking data integrity.

6. The UAV swarm automatic restocking and transportation communication method of claim 1, wherein: The comprehensive scoring algorithm selects the individual or group most suitable for executing the task based on the comprehensive score of the unmanned aerial vehicle node, and the comprehensive score is calculated by weighting the risk score and the energy consumption score, and the comprehensive score calculation formula is: S = w R R w E E , wherein: R: the normalized risk score, and the calculation formula is: ， Wherein, d: the distance from the UAV to the task target, dmax: the maximum possible distance in the task scene, for standardization; Bmax: the maximum value of the power; Cmax: the maximum value of the load capacity; P1, P2, P3: the weight coefficients of the risk factors; C MDID : represents the load capacity of the UAV node, that is, the maximum weight of the goods that the UAV can carry; B MDID : represents the power of the UAV node; E: the normalized energy consumption score, and the calculation formula is: ， where w: weight of materials required for the task, a, b, c: coefficients of energy consumption factors, w R : weight of risk, w E : weight of energy consumption, and w R +w E =1; The calculation formula of the normalized risk and energy consumption scores is: ， ， wherein R min ,R max : minimum and maximum possible values for the risk score; Emin, Emax: the minimum and maximum possible values of the energy consumption score.

7. The UAV swarm automatic restocking and transportation communication method of claim 1, wherein: In step (5), the queuing conflict and the material allocation conflict of the supply station are handled by a conflict resolution mechanism, which includes the following rules: (1) The supply station maintains a dynamic request queue to record the resource requests of the UAV and their arrival times; (2) According to the request arrival time T PDU , the first-in first-out principle is sorted, and if the request time of multiple UAVs is the same, the UAV identifier MDID is sorted in ascending order according to the dictionary order; (3) The supply station with service capacity S SSID The request is processed in the order of the queue, and the current queue length Q of the supply station is updated after the processing is completed SSID And the UAV cluster is informed through state broadcast (4) If the waiting time of the UAV in the queue exceeds the threshold Tt imeout , the supply station sends a rejection feedback, and the UAV needs to select other supply stations again; (5) The UAV evaluates the possibility of queuing conflict by using the local state table, dynamically adjusts the request behavior; calculates the expected waiting time T of the supply station wait If T wait exceeds the preset threshold T threshold , the UAV selects other supply stations; (6) In the conflict of material distribution, if the supply station inventory I SSID is insufficient, broadcast conflict notification; after receiving the conflict notification, the UAV compares the request time T PDU , and the UAV with the earliest request time is given priority in the distribution of materials. If the request times are the same, the UAVs are sorted in lexicographic order according to the UAV identifier MDID.

8. The UAV cluster automatic restocking and transportation communication method of claim 7, wherein: When the drone evaluates the possibility of a queue conflict using the local state table, the expected waiting time T for the replenishment station is calculated according to the following formula wait , ， If T wait ≤ T threshold , the UAV sends a request to the supply station; If T wait > T threshold The drone selects other supply stations from the state table, The distance d between the UAV and the supply station is calculated according to the following formula: ， where: Q SSID is the current queue length of the replenishment station; S SSID the service capacity of the replenishment station, i.e. the number of drones that can be served per minute; Lat MDID ,Lon MDID ,Alt MDID are latitude, longitude and altitude of the UAV, respectively; Lat SSID ,Lon SSID ,Alt SSID are the latitude, longitude and altitude of the replenishment station, respectively.

Images