Warehouse logistics AGV and unmanned aerial vehicle cooperative scheduling system

Abstract

The invention discloses a warehouse logistics AGV and unmanned aerial vehicle cooperative scheduling system, and relates to the technical field of intelligent logistics scheduling, the system comprises a plurality of AGVs, unmanned aerial vehicles and a warehouse internal communication network, an intelligent docking platform is integrated at the tops of the AGVs, the AGV comprises an LED array beacon, a UWB accurate positioning module, a mechanical self-locking or electromagnetic adsorption mechanism and a wireless communication module, and the AGV and the unmanned aerial vehicle are connected through a distributed negotiation decision protocol. The AGV and the unmanned aerial vehicle can perform autonomous cooperative scheduling, after a demand side broadcasts a cooperative request data packet, an idle device calculates cooperative cost based on the state of the idle device and replies an intention response, the demand side selects a cooperative partner according to a preset rule, the cooperative partner and the cooperative partner determine a temporary convergence point through a dynamic convergence point algorithm, and the AGV drives to the convergence point and lights an LED array beacon. The unmanned aerial vehicle flies to a convergence point, visual identification and UWB positioning are fused to achieve accurate approaching, and finally butt joint and fixing are completed through a mechanical self-locking or electromagnetic adsorption mechanism, and cargo transfer is executed.

Description

Technical Field

[0001] This invention relates to the field of intelligent logistics scheduling technology, and more specifically, to a collaborative scheduling system for warehousing and logistics AGVs and drones. Background Technology

[0002] In smart warehousing scenarios involving high-frequency sorting of multiple product categories, the collaborative model of AGV ground transfer and drone three-dimensional delivery has become a key path to overcome the dual bottlenecks of planar congestion and vertical inefficiency. However, the existing technology system has multiple hidden constraints. Although the centralized scheduling architecture can achieve global path planning, it relies on cloud server centralized decision-making, which can easily lead to latency of hundreds of milliseconds when the number of concurrent devices exceeds fifty. Moreover, the failure of a single scheduling node can cause regional operation paralysis, making it difficult to adapt to the dynamic fluctuations in orders in flexible production. More importantly, the use of closed communication protocols by different manufacturers' equipment and the lack of standardized collaborative interfaces have led to frequent instruction parsing deviations and resource contention when multiple brands are mixed in operation.

[0003] The technical shortcomings in the positioning and docking process are even more significant: the multipath propagation effect formed by the dense metal shelves in the warehouse degrades the accuracy of traditional Bluetooth positioning to the meter level, which cannot meet the high precision requirements for docking; the success rate of single visual positioning solution drops sharply in the shadow of the shelves or in the environment of LED lighting flicker; at the same time, the existing solution adopts a fixed coordinate convergence mode, which does not take into account the ground turning constraints of AGV and the difference in flight ceiling of UAV, resulting in overlapping and redundant paths between the two parties. In the scenario of high-frequency distribution of multiple SKUs, the waiting time for a single collaboration exceeds two minutes.

[0004] The lack of dynamic interaction mechanisms further exacerbates efficiency losses: UAV flight control systems generally lack relative motion compensation capabilities, making them prone to landing attitude deviations when faced with minor bumps or steering deviations during AGV operation; the mechanical docking mechanism adopts a rigid interlocking design, with an attitude fault tolerance rate of less than five degrees, and the traditional communication module has a packet loss rate of over 10% in dense metal environments, leading to interruptions in docking command transmission. These problems combined have resulted in a long-term equipment collaboration success rate of less than 85%, hindering the advancement of warehousing from single-point automation to group collaborative intelligence.

[0005] Therefore, existing technologies suffer from problems such as poor central scheduling resistance to interference, single and easily biased positioning, and lack of dynamic negotiation and adaptation connection mechanisms. Summary of the Invention

[0006] In order to overcome the problems of poor central scheduling resistance, single positioning prone to deviation, and lack of dynamic negotiation and adaptation docking mechanism in the existing technology, the present invention discloses a collaborative scheduling system for warehousing and logistics AGVs and drones, which can effectively solve the above-mentioned technical problems.

[0007] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows:

[0008] A warehouse logistics AGV and drone collaborative scheduling system includes multiple AGVs, multiple drones, and an in-warehouse communication network. Each AGV integrates an intelligent docking platform on its top, which includes dynamically illuminated or extinguishable LED array beacons, a UWB precise positioning module, a mechanical self-locking or electromagnetic adsorption mechanism, and a wireless communication module. The system achieves autonomous collaborative scheduling of AGVs and drones through a distributed negotiation decision-making protocol. The scheduling process includes:

[0009] The demand side broadcasts collaborative request data packets through the warehouse's internal communication network;

[0010] When an idle AGV or drone receives a request, it calculates the collaboration cost based on its own status and responds with an indication of intent.

[0011] The demand side acts as a temporary auctioneer, selecting a partner from the responses based on preset rules;

[0012] The two parties determine a temporary rendezvous point through a dynamic rendezvous point algorithm. The AGV drives to the rendezvous point and lights up the LED array beacon. The drone flies to the rendezvous point and achieves precise approach by integrating visual recognition and UWB positioning.

[0013] After docking and fixing is completed through mechanical self-locking or electromagnetic adsorption mechanisms, the goods are transferred.

[0014] Preferably, the LED array beacon is used to generate an optical identifier, which encodes the AGV's device identifier, task status, and docking request information. The optical identifier is a QR code or an AUCO code.

[0015] The UWB precision positioning module communicates with the UAV, providing centimeter-level accuracy in relative distance and angle measurement data.

[0016] Preferably, the collaborative request data packet includes the requester's location information, task type, cargo specifications, and task urgency.

[0017] The self-state includes the current location, remaining battery power, and task queue status, and the coordination cost is calculated based on the travel distance, time consumption, and energy consumption.

[0018] Preferably, the preset rules of the distributed negotiation decision-making protocol include selecting the responder with the optimal coordination cost or the responder with the shortest expected arrival time. The decision-making process is executed independently and in a distributed manner by the demand side without the need for intervention from a central server.

[0019] Preferably, the dynamic meeting point is a dynamic coordinate point predicted based on the current motion state of the AGV and the drone. The dynamic meeting point algorithm comprehensively considers the predetermined path of the AGV and the flight capability of the drone, and determines the optimization objective by minimizing the total travel cost of both parties.

[0020] Preferably, the UAV switches to an autonomous precision control mode within a preset range near the rendezvous point, achieves coarse positioning by identifying LED array beacons through an onboard visual sensor, and achieves fine positioning by combining the relative pose data obtained by the UWB precision positioning module.

[0021] Preferably, the flight control system of the UAV integrates an active compliant control algorithm, which dynamically compensates for the relative motion generated by the movement of the AGV based on relative pose data, so as to realize the dynamic landing or hovering loading and unloading of the UAV on the moving AGV.

[0022] Preferably, the mechanical self-locking or electromagnetic adsorption mechanism is used to secure the UAV landing gear during the docking process;

[0023] The intelligent docking platform is also equipped with a cargo transfer mechanism, and the cargo transfer is completed through the cargo transfer mechanism or the release mechanism of the drone itself, so as to achieve rapid transfer.

[0024] Preferably, the wireless communication module is a communication module that supports low-latency data transmission, used to realize real-time data interaction between AGV and drone, and between AGV and drone and warehousing infrastructure.

[0025] Preferably, when the drone needs to distribute goods to multiple AGVs, it establishes cooperative relationships with each AGV in turn through multiple rounds of distributed negotiation and decision-making, calculates the dynamic rendezvous point of each round of cooperation and completes the transfer of goods until the distribution task is completed;

[0026] After the mission is completed, the drone autonomously returns to the docking point, and the AGV drives to the target location to perform subsequent operations.

[0027] Compared with existing technologies, the advantages of this invention are as follows: This technical solution uses a distributed negotiation decision-making protocol to replace the traditional central scheduling architecture. The demand party broadcasts a collaboration request containing location, task type, cargo specifications, and urgency level through a low-latency communication network within the warehouse. Idle AGVs or drones, based on their own location, remaining power, and task queue status, calculate the comprehensive collaboration cost of travel distance, time consumption, and energy consumption, and then respond with an intention response. The demand party, acting as a temporary auctioneer, independently selects a partner according to the rule of optimal collaboration cost or shortest expected arrival time. Because the decision-making process does not require intervention from a central server, it avoids the risk of systemic paralysis caused by central node failure and eliminates the problem of data transmission lag in the cloud, thus improving the timeliness of response and anti-interference capability in multi-device concurrent scenarios. To address the shortcomings of single positioning being prone to deviation, an innovative fusion scheme of LED array beacons and UWB precise positioning is adopted. The LED array dynamically generates QR codes or ArUco codes containing device identification, task status, and docking request information for the drone's onboard visual sensor to identify. Instead of coarse positioning, the UWB module synchronously transmits high-precision relative pose data to achieve fine positioning. The two complement each other, effectively resisting the impact of multipath interference from warehouse metal shelves and changes in light on positioning, ensuring the stability and accuracy of docking positioning. To solve the problem of the lack of a dynamic negotiation and matching docking mechanism, a dynamic rendezvous point algorithm is designed. Combining the AGV's predetermined path and the drone's flight capability, a temporary rendezvous point is determined with the goal of minimizing the total travel cost for both parties. When the drone approaches, it switches to an autonomous and precise control mode. Its flight control system's active compliant control algorithm dynamically compensates for the relative motion generated by the AGV's movement based on real-time pose data. Then, the landing gear is stabilized and fixed through mechanical self-locking or electromagnetic adsorption mechanisms. With the help of the intelligent docking platform's cargo transfer mechanism or the drone's own release mechanism, a rapid transfer is completed. This avoids path redundancy and waiting losses at fixed rendezvous points, and improves the success rate of mobile docking through dynamic attitude compensation and reliable docking structure. The multi-round negotiation mechanism is more suitable for multi-AGV cargo distribution scenarios, achieving dual optimization of efficiency and reliability in the entire process. Attached Figure Description

[0028] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary. For those skilled in the art, other embodiments can be derived from the provided drawings without creative effort.

[0029] Figure 1 This is a system architecture diagram of the present invention;

[0030] Figure 2 This is a flowchart illustrating the collaborative scheduling process of the present invention. Detailed Implementation

[0031] The accompanying drawings are for illustrative purposes only and should not be construed as limiting the scope of this patent.

[0032] To better illustrate this embodiment, some parts in the accompanying drawings may be omitted, enlarged, or reduced, and do not represent the actual product dimensions;

[0033] It will be understood by those skilled in the art that certain well-known structures and their descriptions may be omitted in the accompanying drawings.

[0034] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0035] Example

[0036] This embodiment is applied to the regional core warehousing center of a large e-commerce company. The center has a three-story structure with a total warehousing area of ​​tens of thousands of square meters. It is divided into five functional areas: inbound temporary storage area, automated storage and retrieval system (AS / RS) storage area, split picking area, verification and packaging area, and outbound collection area. During the Double 11 promotion, the warehousing center processes more than 100,000 orders per day. Among them, small and medium-sized items, such as clothing, 3C accessories, household goods, and beauty samples, account for more than 70% and need to be quickly transferred from the AS / RS storage area to the split picking area for sorting.

[0037] Traditional operating models revealed significant bottlenecks during peak sales periods: the densely packed shelves in automated storage and retrieval systems made AGV ground movement susceptible to aisle congestion, resulting in low transfer efficiency; goods on high-rise shelves required manual retrieval using climbing equipment, which was not only time-consuming but also prone to errors; the central dispatch system was prone to response delays when equipment concurrency surged, leading to coordination failures. To address these pain points, the warehouse center deployed an AGV and drone collaborative dispatch system based on this technical solution. It adopted a collaborative mode of drone high-rise retrieval + AGV ground connection, combined with distributed negotiation decision-making technology, to achieve autonomous equipment matching and efficient coordination without the need for centralized server management. The system deployed a total of twenty AGVs, ten drones, and a low-latency communication network covering the entire area, specifically adapted to the high-concurrency transfer demands during peak sales periods.

[0038] Please see Figure 1-2 A warehouse logistics AGV and drone collaborative scheduling system includes multiple AGVs, multiple drones, and an in-warehouse communication network. Each AGV integrates an intelligent docking platform on its top, which includes dynamically illuminated or extinguishable LED array beacons, a UWB precise positioning module, a mechanical self-locking or electromagnetic adsorption mechanism, and a wireless communication module. The system achieves autonomous collaborative scheduling of AGVs and drones through a distributed negotiation decision-making protocol. The scheduling process includes:

[0039] The demand side broadcasts collaborative request data packets through the warehouse's internal communication network;

[0040] When an idle AGV or drone receives a request, it calculates the collaboration cost based on its own status and responds with an indication of intent.

[0041] The demand side acts as a temporary auctioneer, selecting a partner from the responses based on preset rules;

[0042] The two parties determine a temporary rendezvous point through a dynamic rendezvous point algorithm. The AGV drives to the rendezvous point and lights up the LED array beacon. The drone flies to the rendezvous point and achieves precise approach by integrating visual recognition and UWB positioning.

[0043] After docking and fixing is completed through mechanical self-locking or electromagnetic adsorption mechanisms, the goods are transferred.

[0044] The LED array beacon is used to generate optical identifiers, which encode the AGV's device identifier, task status, and docking request information. The optical identifiers are either QR codes or AUCO codes.

[0045] The UWB precision positioning module communicates with the UAV, providing centimeter-level accuracy in relative distance and angle measurement data.

[0046] The collaborative request data packet includes the requester's location information, task type, cargo specifications, and task urgency.

[0047] The self-state includes the current location, remaining battery power, and task queue status, and the coordination cost is calculated based on the travel distance, time consumption, and energy consumption.

[0048] The preset rules of the distributed negotiation decision-making protocol include selecting the responder with the optimal coordination cost or the responder with the shortest expected arrival time. The decision-making process is executed independently and in a distributed manner by the demand side without the need for intervention from a central server.

[0049] The dynamic meeting point is a dynamic coordinate point predicted based on the current motion state of the AGV and the drone. The dynamic meeting point algorithm comprehensively considers the predetermined path of the AGV and the flight capability of the drone, and determines the minimum total travel cost of both parties as the optimization objective.

[0050] The UAV switches to autonomous precision control mode within a preset range near the rendezvous point, and achieves coarse positioning by recognizing LED array beacons through onboard visual sensors, and achieves fine positioning by combining the relative pose data obtained by the UWB precision positioning module.

[0051] The UAV's flight control system integrates an active compliant control algorithm. This algorithm dynamically compensates for the relative motion generated by the movement of the AGV based on relative pose data, enabling the UAV to dynamically land on the moving AGV or hover in the air for loading and unloading.

[0052] The mechanical self-locking or electromagnetic adsorption mechanism is used to secure the UAV landing gear during the docking process;

[0053] The intelligent docking platform is also equipped with a cargo transfer mechanism, and the cargo transfer is completed through the cargo transfer mechanism or the release mechanism of the drone itself, so as to achieve rapid transfer.

[0054] The wireless communication module is a communication module that supports low-latency data transmission and is used to realize real-time data interaction between AGV and drone, and between AGV and drone and warehousing infrastructure.

[0055] When a drone needs to distribute goods to multiple AGVs, it establishes cooperative relationships with each AGV in turn through multiple rounds of distributed negotiation and decision-making, calculates the dynamic rendezvous point of each round of cooperation and completes the transfer of goods until the distribution task is completed.

[0056] After the mission is completed, the drone autonomously returns to the docking point, and the AGV drives to the target location to perform subsequent operations.

[0057] AGV and intelligent docking platform configuration: The AGV is a medium-sized wheeled AGV specially selected for the promotion. It has bidirectional driving and autonomous obstacle avoidance capabilities. It is equipped with LiDAR, binocular vision sensors and ultrasonic sensors, which can realize 360-degree environmental perception. It can accurately identify shelves, personnel, other equipment and ground debris, and dynamically adjust the driving path and speed. The AGV is powered by a high-capacity lithium battery and supports fast charging. A single charge can meet the continuous operation needs during the promotion. The side of the vehicle is equipped with an emergency stop button and status indicator lights to facilitate manual emergency intervention and status monitoring.

[0058] The AGV features a square intelligent docking platform integrated at the top center. The platform's dimensions are precisely matched to the drone's landing gear, and its edges are equipped with anti-slip protrusions and cushioning strips to prevent cargo from slipping or equipment from colliding during docking. The core component configuration is as follows:

[0059] LED Array Beacon: Composed of dozens of high-brightness SMD LEDs in a regular array, driven by an embedded controller, it can dynamically switch between generating QR codes or ArUco codes for optical identification. The identification code information includes the AGV's unique device number, current task status (idle / awaiting docking / transferring / fault), docking request type (receiving goods / sending goods), and remaining battery power. The brightness of the LEDs can be adaptively adjusted after detecting the warehouse light intensity by a photoresistor, ensuring that the drone can be clearly identified in brightly lit outbound areas or in dimly lit deep shelves. The array is encapsulated with a transparent acrylic protective cover, which has dustproof and waterproof functions and can adapt to the dusty and clean environment in the warehouse.

[0060] UWB Precision Positioning Module: Employs an industrial-grade high-precision positioning chip, with a built-in UWB positioning tag and dedicated communication unit. The positioning tag collects AGV position data in real time via ultra-wideband pulse signals, while the communication unit establishes a dedicated communication link with the drone using dual-frequency bands, transmitting data on the relative distance, angle, and height difference between the AGV and the drone. Positioning accuracy reaches the centimeter level. The module incorporates an anti-interference algorithm, effectively resisting electromagnetic interference generated by metal shelves and warehousing equipment, ensuring stable and reliable positioning data in complex environments.

[0061] Mechanical self-locking mechanism: It adopts a four-claw buckle self-locking structure, consisting of a fixed base, movable buckles, and micro stepper motors. The fixed base is installed on the docking platform surface, with positioning grooves corresponding to the drone landing gear support points. Rubber buffer pads are placed in the grooves. The movable buckles are driven by four motors respectively, enabling radial extension and retraction. When the drone landing gear falls into the positioning groove, the infrared sensor built into the base detects the landing gear signal, and the motor drives the buckles to extend synchronously and lock tightly into the pre-set slots of the landing gear. The locking force meets the fixing requirements of the drone under full load. After the cargo transfer is completed, the motor drives the buckles to retract synchronously, releasing the drone. The entire action is responsive and the locking is stable.

[0062] Wireless communication module: Adopting an industrial-grade low-latency communication module, supporting time division multiple access technology, it can simultaneously establish communication links with drones, warehouse management systems, and other AGVs. Data transmission latency is controlled at an extremely low level. The module has an automatic frequency hopping function, which can quickly switch to a backup frequency band when a certain frequency band is interfered with, ensuring communication continuity. At the same time, it has a built-in data encryption algorithm to encrypt transmitted location data, control commands, and other information to prevent data leakage or tampering.

[0063] Cargo Transfer Mechanism: Features a built-in small belt conveyor with a non-slip rubber belt width adapted to common small and medium-sized goods. Driven by a stepper motor, the conveying speed is adaptively adjusted according to the weight and size of the goods. An infrared sensor at the conveyor entrance detects whether the goods are in place. The exit seamlessly connects to the storage compartment inside the AGV, which is equipped with partitions to prevent multiple items from being mixed. When the drone places goods on the conveyor belt, the sensor triggers a signal, the conveyor starts, and transfers the goods to the storage compartment. If the AGV needs to deliver goods to the drone, the conveyor reverses its direction, transporting the goods from the storage compartment to the designated location on the docking platform.

[0064] Drone Configuration: A hexacopter drone specifically designed for major promotional events is selected. The fuselage is made of carbon fiber composite material, resulting in a lightweight yet high-strength structure. Its maximum payload meets the weight requirements of small to medium-sized goods, and its flight time is adapted to the high-frequency pickup-docking-return process during promotional events. The drone is equipped with a high-precision flight control system, multi-dimensional sensing modules, and an efficient cargo-grabbing mechanism. The fuselage features status indicator lights and an emergency landing device. The core configuration is as follows:

[0065] Airborne vision sensor: Composed of a 10-megapixel high-definition industrial camera and a dedicated image processing unit, the camera is mounted on the bottom gimbal of the drone and features 360-degree rotation, autofocus, and adaptive light adjustment. It can clearly capture images of the AGV docking platform under various lighting conditions. The image processing unit has a built-in fast recognition algorithm that can analyze the identification information generated by the LED array beacon in a very short time, extract the AGV equipment number and position data, and achieve coarse positioning. The gimbal has electronic image stabilization to counteract the shaking during drone flight, ensuring clear and stable images.

[0066] UWB Precision Positioning Module: Fully compatible with the module model mounted on the AGV, it has a built-in UWB positioning base station and data filtering unit. The positioning base station receives the pulse signal transmitted by the AGV positioning tag and calculates the relative distance between the two through the time-of-flight algorithm. The data filtering unit performs Kalman filtering noise reduction on the raw positioning data to eliminate errors caused by environmental interference, and transmits the processed precise relative distance and angle data to the UAV flight control system, providing core data for precise positioning.

[0067] Flight Control System: Employs a high-performance flight control motherboard integrating a GPS / BeiDou dual-mode navigation module, an inertial measurement unit (IMU), and an active compliant control algorithm. The navigation module can collect the UAV's coordinate data within the warehouse in real time, and combine this with the shelf location and aisle information provided by the warehouse management system to plan the optimal flight path. The IMU collects the UAV's pitch angle, roll angle, yaw angle, and flight speed data hundreds of times per second to ensure stable flight attitude. The active compliant control algorithm receives relative attitude data transmitted from the UWB module and dynamically adjusts the UAV's flight attitude and speed through a PID control algorithm, accurately compensating for the relative motion generated by the AGV's movement and achieving precise docking during movement.

[0068] Cargo gripping mechanism: It adopts an electric gripper structure, consisting of two symmetrical grippers and a micro servo motor. The inner side of the grippers is wrapped with silicone anti-slip pads to avoid damaging the product packaging. The gripper spacing can be freely adjusted within a certain range to adapt to small and medium-sized products of different sizes. The gripping mechanism has a built-in pressure sensor. When the gripper contacts the product, the sensor detects the pressure value and feeds it back to the flight control system to ensure that the gripping force is moderate, preventing the product from falling off and avoiding injury. The gripping mechanism has a fast response capability and can complete the gripping and releasing action in a very short time, which is suitable for the high-frequency operation needs during promotional periods.

[0069] Status monitoring and emergency module: Integrates battery management system, motor speed sensor, attitude monitoring unit and emergency landing device. The battery management system monitors the remaining battery power, voltage and temperature in real time. When the power is lower than the preset threshold, it sends a low power warning to the flight control system and uploads the information to the warehouse management system. The motor speed sensor monitors the speed of each rotor motor. When an abnormal speed occurs, it immediately alarms and initiates fault-tolerant control. The attitude monitoring unit judges the flight attitude of the UAV in real time. When there is severe shaking or deviation from the predetermined path, it initiates an emergency stabilization program. The emergency landing device can automatically deploy the parachute in extreme cases to ensure the safety of equipment and cargo.

[0070] Warehouse communication network deployment: A two-layer communication architecture with full coverage of fiber optic backbone network and wireless APs is adopted. The fiber optic backbone network connects the core switches of each functional area, and the transmission rate meets the stable transmission requirements of large data volumes during peak sales periods. A total of fifty industrial-grade wireless APs are deployed on the top of the shelf aisles, at the junction of functional areas and on the warehouse pillars. They adopt the latest wireless communication protocol, support dual-band simultaneous operation, and achieve signal coverage without dead zones in the warehouse area. The signal coverage radius can meet the spatial requirements of large warehouses.

[0071] The network employs a service quality priority scheduling mechanism, prioritizing the transmission of positioning data and control commands between AGVs and drones to ensure the transmission of critical data. It also utilizes load balancing technology to automatically switch some devices to nearby APs when too many devices are connected to a particular wireless AP, preventing network congestion. Furthermore, the network is equipped with firewalls, intrusion detection systems, and data encryption gateways to comprehensively protect network security, prevent external attacks and data leaks, and ensure stable system operation during peak sales periods.

[0072] This embodiment takes a typical high-frequency task during the Double 11 promotion, in which a drone retrieves goods from a high-level automated storage area and works with an AGV to transfer the goods to the picking area. The AGV then transfers the goods to the picking area. The embodiment details the collaborative scheduling process, which is divided into six stages: task triggering, distributed negotiation and decision-making, dynamic rendezvous point determination, precise docking, goods transfer, and task completion. The entire process requires no manual intervention and is suitable for the high concurrency requirements during the promotion.

[0073] The warehouse management system generates picking tasks in real time based on order data. When the inventory of a certain type of product in the picking area is lower than the preset threshold, it immediately sends a picking instruction to the drone corresponding to the automated storage and retrieval system. The instruction includes core information such as the shelf number, shelf location, product name and quantity of the target product, and marks the task priority (high priority by default during major promotions).

[0074] After receiving instructions, the drone uses GPS / BeiDou navigation modules combined with warehouse positioning data to plan the optimal flight path to the target shelf. The path avoids lighting equipment, fire pipes, and other drones in flight on the shelf. After starting its rotors and flying towards the target shelf, the drone uses onboard visual sensors to identify the shelf number and shelf position, adjusts its flight attitude to be directly in front of the goods, controls the goods gripping mechanism to extend and align with the goods, and after the gripper contacts the goods, the pressure sensor detects the gripping signal, the gripper tightens, and the drone adjusts its attitude to slowly fly away from the shelf, while uploading the successful retrieval information to the warehouse management system.

[0075] After the drone flies away from the shelf, it broadcasts a collaborative request data packet to the warehouse communication network via its wireless communication module. This data packet contains the following core information: drone equipment number, current location coordinates (precisely collected by the navigation module), task type (goods transfer, from the automated storage and retrieval system to the picking area), cargo information (goods name, weight, dimensions, packaging type and quantity), task priority (high priority), and the drone's current status (remaining battery power, flight attitude, cargo grabbing status).

[0076] The warehouse communication network pushes collaboration request data packets to idle AGVs between the automated storage and picking areas (the system maintains the list of idle AGVs through real-time heartbeat detection; busy or faulty AGVs do not receive requests). In this embodiment, five idle AGVs receive the request. After receiving the request, each AGV immediately starts the collaboration cost calculation program and calculates the comprehensive cost of collaborating with the drone based on its own status parameters.

[0077] The AGV's own status parameters include: current position coordinates (collected by the AGV's built-in positioning module), remaining power (monitored in real time by the battery management system), task queue status (no tasks to be executed for any of the five AGVs), and storage warehouse idle status (all idle). The collaborative cost calculation adopts a weighted scoring method, with the following weights: travel distance (40%), time consumption (30%), and energy consumption (30%). The travel distance is the straight-line distance from the AGV's current position to the expected meeting point (preliminary estimate based on historical data); the time consumption is calculated based on the AGV's travel speed and the channel congestion (real-time push from the warehouse management system); and the energy consumption is calculated based on the travel distance, cargo weight, and the AGV energy consumption model (trained from historical operating data).

[0078] After each of the five AGVs completes its collaborative cost calculation, it replies to the drone via wireless communication. The response information includes the AGV's device number, current location, remaining battery power, collaborative cost score, and storage capacity. The drone, as the requester, receives the response information as a temporary auctioneer. Because the task is of high priority, the preset rule is the shortest estimated arrival time. The drone's built-in decision-making program sorts the estimated arrival times of the five AGVs and determines that AGV-12 has the shortest estimated arrival time. It then sends a cooperation confirmation instruction to AGV-12, which includes the drone's device number, current location, and docking preparation requirements. At the same time, it sends a rejection instruction to the other four AGVs, which update their status to idle and re-add them to the idle list to wait for the next request.

[0079] The entire negotiation and decision-making process is executed independently by drones without the need for central server intervention. The time from request broadcast to cooperation confirmation is extremely short, ensuring efficient response during major promotional periods.

[0080] After receiving the cooperation confirmation instruction, AGV-12 establishes a dedicated communication link with the drone through the wireless communication module. The two parties exchange motion status data in real time and start the dynamic rendezvous point algorithm to calculate the temporary rendezvous point. The motion data transmitted by the drone to the AGV includes: current flight speed, flight direction, remaining flight time, maximum flight altitude limit and cargo grasping stability. The motion data transmitted by the AGV to the drone includes: current travel speed, travel direction, remaining battery power, maximum travel speed and preset travel path (the conventional path from the current position to the picking area).

[0081] The dynamic rendezvous point algorithm aims to minimize the total travel cost for both parties and ensure synchronized arrival times. It comprehensively considers the following constraints: AGVs must avoid busy passageways at the exit of the automated storage area (during peak sales periods, personnel and equipment flow is dense); drones must avoid obstacles on top of the shelves and other flying drones; the rendezvous point must be located near the AGV's preset path and within the drone's flight radius; and the rendezvous point must have sufficient open space for docking. The algorithm generates multiple candidate rendezvous points through multiple rounds of iterative calculations, then calculates the travel cost and arrival time difference for each candidate point, and finally selects the candidate point with the lowest total travel cost and the smallest arrival time difference as the dynamic rendezvous point.

[0082] In this embodiment, the dynamic rendezvous point is determined to be the open area of ​​the passageway on the west side of the exit of the automated storage rack area. This location facilitates the AGV to arrive quickly after exiting the passageway, and also provides unobstructed flight and docking space for the UAV. Furthermore, the time difference between the arrival of both parties at this point is controlled within a very short period of time to avoid long waiting times. The coordinate data of the rendezvous point is synchronously transmitted to the navigation modules of the UAV and AGV as the target for both parties to travel.

[0083] After receiving the rendezvous point coordinates, the AGV-12 initiates its driving program. The navigation module, combined with real-time channel congestion data pushed by the warehouse management system, plans the optimal path from its current location to the rendezvous point. During the journey, the LiDAR and binocular vision sensors scan the surrounding environment in real time. When personnel or temporary obstacles are detected ahead, the AGV automatically slows down and fine-tunes its path. When channel congestion is detected, it immediately switches to an alternative path to ensure rapid arrival. When the AGV is approximately a certain distance from the rendezvous point, the LED array beacon of the intelligent docking platform automatically lights up, generating an ArUco code containing its own device number, docking status, and request to receive cargo. At the same time, the UWB precise positioning module is activated, establishing a communication link with the UWB module of the drone and beginning to transmit relative pose data.

[0084] After receiving the rendezvous point coordinates, the drone's navigation module plans a flight path from its current location to the rendezvous point, sets a safe flight altitude (avoiding personnel and equipment), and starts its rotors to fly towards the rendezvous point. During flight, the flight control system corrects the flight attitude in real time. When encountering airflow or other interference that causes a slight deviation, it immediately adjusts the rotor speed to return to the predetermined path. When the drone is about a certain distance from the rendezvous point, the flight control system automatically switches to autonomous precision control mode, the onboard vision sensor is activated, and it takes pictures of the area below. The image processing unit quickly analyzes the ArUco code in the image, extracts the AGV position information, achieves coarse positioning, and guides the drone to the area above the AGV.

[0085] When the drone is approximately a certain distance from the AGV, the UWB precision positioning module takes the lead in positioning, transmitting precise relative distance, angle, and altitude difference data to the flight control system. The flight control system's active compliant control algorithm combines this data with attitude data collected by the inertial measurement unit to dynamically adjust the drone's flight attitude and speed. At this time, the AGV is still moving slowly. The algorithm calculates the AGV's speed and direction in real time, controlling the drone to move synchronously and accurately compensating for relative motion to ensure that the drone is always directly above the AGV docking platform. When the drone descends to approximately a certain height from the docking platform, the laser rangefinder sensor is activated to monitor the altitude difference with the platform in real time, controlling the drone to descend slowly. The descent speed is dynamically adjusted according to the altitude difference to ensure a smooth approach.

[0086] Once the drone's landing gear falls into the positioning groove of the AGV docking platform, the infrared sensor built into the base immediately detects the landing gear signal and sends a docking signal to the AGV controller. The AGV controller then drives the four movable latches of the mechanical self-locking mechanism to extend synchronously, locking the drone's landing gear into the slots and securing the drone. At this point, the drone and the AGV confirm the docking is complete through a two-way communication module.

[0087] After docking is completed, the drone sends a cargo transfer preparation signal to the AGV. The AGV starts the belt conveyor of the cargo transfer mechanism and adjusts the conveying speed to a suitable level according to the weight of the cargo. After receiving the conveyor ready signal, the drone's flight control system controls the grippers of the cargo gripping mechanism to slowly release and place the goods smoothly on the conveyor belt. After the infrared sensor at the conveyor entrance detects the cargo signal, the conveyor starts to operate and transports the goods to the storage compartment inside the AGV. The partitions in the storage compartment ensure that the goods are stored in a fixed location.

[0088] Once the goods have fully entered the storage compartment, the position sensor inside the compartment sends a goods receiving completion signal to the AGV controller. The AGV controller forwards this signal to the drone via the communication module. The drone's flight control system controls the gripping mechanism to reset and simultaneously sends a release request command to the AGV. The AGV controller drives the mechanical self-locking mechanism to retract synchronously, releasing the drone's landing gear. The LED array beacon turns off, the UWB module stops working, and the drone starts its rotors to slowly ascend. Once it reaches a safe altitude, it adjusts its flight direction and returns to the preset docking point, while simultaneously uploading the goods transfer completion information to the warehouse management system.

[0089] After receiving the goods, AGV-12 updates its task status to "transferring". The navigation module replans the optimal path from the rendezvous point to the picking area, prioritizing the dedicated green channel during the promotional period. During the journey, AGV continuously interacts with the warehouse management system to obtain real-time channel congestion data and dynamically adjusts its speed. Upon arriving at the designated location in the picking area, AGV starts the conveyor of the goods transfer mechanism to reverse, transporting the goods to the sorting table. The barcode scanner next to the sorting table scans the product barcode and sends a goods delivery completion signal to AGV after confirming receipt.

[0090] After receiving the signal, the AGV updates its status to idle and determines whether it needs to be charged based on its remaining power: if the power is sufficient, it drives to the vicinity of the automated storage area to wait for the next collaboration request; if the power is lower than the preset threshold, it automatically drives to the AGV charging area and completes docking and charging through the automatic charging device. At the same time, it uploads the execution data of this task (driving trajectory, docking time, cargo information, etc.) to the warehouse management system.

[0091] After the drone arrives at the docking point, it completes docking and charging through the automatic charging device. At the same time, it uploads information such as flight data, pickup location, docking status, and remaining battery power for this mission to the warehouse management system. The warehouse management system integrates the mission data of the AGV and the drone, generates a goods transfer record, and updates the goods inventory in the picking area in a synchronous manner, thus completing the closed loop of this collaborative mission.

[0092] During major sales events, there is a frequent scenario where drones retrieve goods from the inbound temporary storage area and distribute the same batch of goods to AGVs in multiple individual picking areas. This system achieves efficient adaptation through multi-round distributed negotiation decision-making. Taking the example of a drone picking up eight identical items and needing to distribute them to AGVs in eight different individual picking areas, the specific process is as follows:

[0093] After the drone retrieves eight items from the temporary storage area, it broadcasts a collaborative request data packet to the warehouse network for the first time. It then selects an available AGV near the first target picking area and, through distributed negotiation, chooses a cooperating AGV, determines a dynamic rendezvous point, and completes the docking and transfer of the first item. After the first item is transferred, the drone immediately updates the remaining quantity of goods, initiates the second round of negotiation, and broadcasts a request to an available AGV near the second target picking area. It repeats the above process to complete the transfer of the second item. After completing eight rounds of negotiation and transfer, all items are distributed to the corresponding AGVs.

[0094] Each round of negotiation is executed independently. The dynamic rendezvous point is calculated separately based on the position and movement status of each AGV to ensure that the total travel cost of each round of collaboration is minimized. At the same time, the drone flight control system monitors the remaining battery power and the product grabbing status in real time. When the remaining battery power is lower than the preset threshold, the transfer of products of the negotiated AGVs is completed first, and then the drone returns to charge. If no idle AGV responds in a round of negotiation, the drone pauses the negotiation, hovers at the temporary docking point, and sends a scheduling request to the warehouse management system. The system then coordinates AGVs from other areas to provide support, ensuring that the mission is not interrupted during the promotion period.

[0095] If communication between the AGV and the drone is interrupted during collaboration, both parties immediately activate the backup communication link (switch to the backup frequency band); if the backup link still cannot be restored, the AGV stops moving and illuminates the LED array beacon, which flashes continuously to maintain its stable state; the drone hovers at its current position and sends a communication request at very short intervals, while uploading abnormal information to the warehouse management system; if communication still cannot be restored within the preset time, the drone initiates the emergency procedure, carrying the goods back to the original pickup point to avoid loss of goods.

[0096] If the drone fails to recognize the AGV marker due to sudden changes in light or dust obstruction, it immediately switches to pure UWB positioning mode and adjusts its position using the precise data transmitted by the UWB module. If the UWB positioning is deviated due to interference from the metal shelves, the drone increases its flight altitude to expand the visual recognition range and re-captures the AGV marker until accurate positioning is restored. If both positioning methods are inaccurate, the drone sends a positioning anomaly alarm to the warehouse management system and hovers while waiting. The system then dispatches nearby AGVs to assist in positioning using light signals to ensure successful docking.

[0097] If the AGV experiences a motor failure or sensor malfunction during operation, it immediately sends a fault signal to the drone and simultaneously alerts the warehouse management system. The drone then cancels docking and restarts the negotiation process to select another AGV. If the drone experiences a single rotor failure during flight, the flight control system immediately activates the fault-tolerant control program, controlling the drone to land smoothly at the nearest emergency docking point. At the same time, it alerts the system and arranges for maintenance personnel to handle the situation, ensuring the safety of the equipment and goods.

[0098] To adapt to the high concurrency demands during peak sales periods, the system adopts three core optimization measures: First, pre-monitoring of equipment status: the warehouse management system monitors the battery level and fault status of all AGVs and drones in real time, schedules charging of low-battery equipment in advance, and handles minor faults promptly to ensure equipment availability; second, dynamic path optimization: based on order distribution data during peak sales periods, high-frequency transfer routes are planned in advance, and dedicated green channels are set up to reduce equipment congestion; third, task priority scheduling: tasks are prioritized according to the urgency of orders, and collaborative tasks corresponding to urgent orders are executed first to ensure timely order fulfillment.

[0099] The same or similar labels correspond to the same or similar parts;

[0100] The terms used to describe positional relationships in the accompanying drawings are for illustrative purposes only and should not be construed as limiting this patent.

[0101] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the implementation of the present invention. For those skilled in the art, other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all implementation methods here. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the claims of the present invention.

Claims

1. A collaborative scheduling system for warehousing and logistics AGVs and drones, characterized in that, The system includes multiple AGVs, multiple drones, and an in-warehouse communication network. Each AGV has an integrated intelligent docking platform on its top. This platform includes dynamically illuminated or extinguishable LED array beacons, a UWB precision positioning module, a mechanical self-locking or electromagnetic adsorption mechanism, and a wireless communication module. The system achieves autonomous collaborative scheduling of the AGVs and drones through a distributed negotiation decision-making protocol. The scheduling process includes: The demand side broadcasts collaborative request data packets through the warehouse's internal communication network; When an idle AGV or drone receives a request, it calculates the collaboration cost based on its own status and responds with an indication of intent. The demand side acts as a temporary auctioneer, selecting a partner from the responses based on preset rules; The two parties determine a temporary rendezvous point through a dynamic rendezvous point algorithm. The AGV drives to the rendezvous point and lights up the LED array beacon. The drone flies to the rendezvous point and achieves precise approach by integrating visual recognition and UWB positioning. After docking and fixing is completed through mechanical self-locking or electromagnetic adsorption mechanisms, the goods are transferred.

2. The system according to claim 1, characterized in that, The LED array beacon is used to generate optical identifiers, which encode the AGV's device identifier, task status, and docking request information. The optical identifiers are either QR codes or AUCO codes. The UWB precision positioning module communicates with the UAV, providing centimeter-level accuracy in relative distance and angle measurement data.

3. The system according to claim 1, characterized in that, The collaborative request data packet includes the requester's location information, task type, cargo specifications, and task urgency. The self-state includes the current location, remaining battery power, and task queue status, and the coordination cost is calculated based on the travel distance, time consumption, and energy consumption.

4. The system according to claim 1, characterized in that, The preset rules of the distributed negotiation decision-making protocol include selecting the responder with the optimal coordination cost or the responder with the shortest expected arrival time. The decision-making process is executed independently and in a distributed manner by the demand side without the need for intervention from a central server.

5. The system according to claim 1, characterized in that, The dynamic meeting point is a dynamic coordinate point predicted based on the current motion state of the AGV and the drone. The dynamic meeting point algorithm comprehensively considers the predetermined path of the AGV and the flight capability of the drone, and determines the minimum total travel cost of both parties as the optimization objective.

6. The system according to claim 1, characterized in that, The UAV switches to autonomous precision control mode within a preset range near the rendezvous point, and achieves coarse positioning by recognizing LED array beacons through onboard visual sensors, and achieves fine positioning by combining the relative pose data obtained by the UWB precision positioning module.

7. The system according to claim 6, characterized in that, The UAV's flight control system integrates an active compliant control algorithm. This algorithm dynamically compensates for the relative motion generated by the movement of the AGV based on relative pose data, enabling the UAV to dynamically land on the moving AGV or hover in the air for loading and unloading.

8. The system according to claim 1, characterized in that, The mechanical self-locking or electromagnetic adsorption mechanism is used to secure the UAV landing gear during the docking process; The intelligent docking platform is also equipped with a cargo transfer mechanism, and the cargo transfer is completed through the cargo transfer mechanism or the release mechanism of the drone itself, so as to achieve rapid transfer.

9. The system according to claim 1, characterized in that, The wireless communication module is a communication module that supports low-latency data transmission and is used to realize real-time data interaction between AGV and drone, and between AGV and drone and warehousing infrastructure.

10. The system according to claim 1, characterized in that, When a drone needs to distribute goods to multiple AGVs, it establishes cooperative relationships with each AGV in turn through multiple rounds of distributed negotiation and decision-making, calculates the dynamic rendezvous point of each round of cooperation and completes the transfer of goods until the distribution task is completed. After the mission is completed, the drone autonomously returns to the docking point, and the AGV drives to the target location to perform subsequent operations.

Images