A method and system for low-altitude economy of unmanned aerial vehicle landing and taking off

Abstract

The application discloses a kind of unmanned plane remote take-off and landing method and system for low-altitude economy, it is related to a kind of non-electric variable control or regulation system technical field, comprising: constructing the distributed network by the mobile airport with edge computing capability, each node real-time acquisition and pre-process multi-source disaster data;Based on pre-processed data, through distributed bidding and dynamic queuing mechanism, the optimal task execution node is determined by inter-node autonomous cooperation;Unmanned plane adopts flight segment type rolling path planning, and realizes the accurate scheduling of airspace and apron resource in combination with landing reservation mechanism;After landing, synchronous completion of material loading and unloading and battery replacement, and trigger resource state network synchronization and new round of task bidding cycle.This application overcomes the limitations of traditional centralized system in the decision paralysis when disaster area communication is interrupted, with the help of task bidding and rolling planning mechanism, significantly improves the accuracy of resource matching and the safety of flight path in complex and variable environment.

Description

Technical Field

[0001] This invention relates to the field of control or regulation systems for non-electrical variables, and more particularly to a method and system for remote take-off and landing of unmanned aerial vehicles (UAVs) for low-altitude economic purposes. Background Technology

[0002] Currently, in emergency relief efforts following major natural disasters (such as earthquakes and floods), rapidly establishing reliable logistics and transportation channels is crucial for saving lives and minimizing losses. Traditional ground transportation is often disrupted, making low-altitude drone logistics a core technology for maintaining lifelines. To extend the operational radius of drones and achieve precise delivery of supplies and intelligence gathering in widely dispersed disaster areas, drone take-off and landing technology has been introduced into disaster relief scenarios. This technology allows drones to land, resupply, and perform new missions at different locations within the disaster area, rather than having to return to their initial take-off point. Existing technology achieves this function by deploying mobile airports. These mobile airports are modular platforms that can be transported to the disaster area by vehicle or helicopter, integrating automatic take-off and landing guidance, fast charging / battery swapping, and simple communication relay functions. The typical operating mode is as follows: the rescue command center, based on the initial disaster situation, pre-deploys multiple mobile airports and plans flight missions connecting these nodes for drones. Drones sequentially visit each mobile airport to complete the loading and unloading of supplies and refueling, thereby achieving cyclical support to multiple disaster-stricken areas and initially solving the problem of limited operational radius from a single fixed base.

[0003] Existing methods for remote take-off and landing of drones based on mobile airports, applied in disaster relief scenarios, are fundamentally based on static path pre-planning according to preliminary disaster information. That is, in the early stages of rescue, fixed flight sequences and resupply nodes are set for drones based on limited information. This method has a fundamental flaw: the system cannot effectively adapt to the dynamic evolution of the disaster situation and the unpredictability of temporary distress requests. The direct impact is rigid dispatching of rescue resources and delayed response. In real disaster environments, the location of trapped personnel, the type and priority of material needs change rapidly. Existing technologies struggle to incorporate this sudden information in real time during mission execution and dynamically adjust the drone missions across the entire network.

[0004] Therefore, it is necessary to improve an existing method and system for remote take-off and landing of unmanned aerial vehicles (UAVs) for low-altitude economy in order to solve the above problems. Summary of the Invention

[0005] This invention overcomes the shortcomings of the prior art and provides a method and system for remote take-off and landing of unmanned aerial vehicles (UAVs) for low-altitude economy, aiming to solve the problem of rigid dynamic response to disasters caused by static planning and resource silos in the prior art.

[0006] To achieve the above objectives, the technical solution adopted by this invention is: a method for remote take-off and landing of unmanned aerial vehicles (UAVs) for low-altitude economic purposes, comprising:

[0007] S1. Construct a distributed mobile airport network, deploy several mobile airports with edge computing capabilities in the disaster area to form a dynamic take-off and landing node group, collect multi-source disaster data in real time and perform edge preprocessing.

[0008] S2. Each mobile airport acts as an edge node, generating a task bidding package based on preprocessed multi-source disaster data, broadcasting it to all edge nodes and receiving feedback bidding information to determine the optimal task execution node, including: take-off node and landing node.

[0009] S3. The drones controlled by the takeoff node adopt segment-based rolling path planning and initiate landing reservations to the landing node. The landing node confirms the airspace and apron scheduling based on resource status and task queue.

[0010] S4. The drone lands according to the reservation permission and simultaneously completes the loading and unloading of materials, battery replacement or mission payload adjustment on the take-off and landing platform.

[0011] S5. After the take-off and landing operation is completed, the mobile airport of the landing node updates its local resource inventory and synchronizes it to the entire network, triggering a new round of task bidding cycle based on the resource status change.

[0012] In a preferred embodiment of the present invention, step S2 further includes a dynamic queue-jumping mechanism: when a temporary high-priority task occurs, the system triggers the dynamic queue-jumping mechanism to interrupt the flight of the current non-critical task UAV and guide it to land at a designated mobile airport to complete a rapid task switchover; the dynamic queue-jumping mechanism includes: generating and broadcasting a high-priority task bidding package, identifying and interrupting non-critical task UAVs, selecting temporary landing points, and simultaneously preparing rapid battery swapping equipment and materials required for the new task; the priority score of the temporary high-priority task exceeds a preset threshold, and the non-critical task is a task with a priority score lower than the preset threshold.

[0013] In a preferred embodiment of the present invention, the edge preprocessing in step S1 includes:

[0014] The collected multi-source disaster data is cleaned and outliers are removed. Among them, environmental data is smoothed and corrected using the sliding window mean method, demand data is filtered for invalid requests based on preset geographical boundaries and data verification rules, and equipment data is removed by threshold judgment method to eliminate erroneous readings.

[0015] Valid data is classified and archived, and assigned dynamic priorities; demand data is prioritized based on the core criteria of life rescue > medical supplies > living supplies > general reconnaissance, and priority scores are calculated in combination with the urgency of mission reporting time and the number of people trapped.

[0016] Key information is extracted and compressed by using feature extraction algorithms to extract key information from the fused dataset and then using data desensitization and format optimization techniques for lightweight compression.

[0017] A cross-validation mechanism is activated to compare the data collected by environmental sensors with the data transmitted back by the UAV, and to verify the spatial overlap between the location reported by the ground terminal and the image data.

[0018] Add a timestamp to each type of key information and set an update cycle. When the data has not been updated within the update cycle, automatically send a data request to the associated data collection terminal.

[0019] In a preferred embodiment of the present invention, in step S2, the task bidding package includes the task objective, requirement priority score, required material type and quantity, target location coordinates, latest response time limit, and flight safety boundary; the optimal task execution node is selected through a multi-dimensional comprehensive adaptability model, which includes response speed adaptability score, task completion probability adaptability score, resource consumption and redundancy adaptability score, coordination and supply connection adaptability score, historical task reliability adaptability score, task conflict and load pressure adaptability score, and resource recovery capability adaptability score, and is dynamically adjusted in combination with environmental urgency correction factor and time sensitivity correction factor.

[0020] In a preferred embodiment of the present invention, the segment-based rolling path planning in step S3 includes:

[0021] The complete flight path of the drone from the take-off point to the landing point is divided into multiple consecutive short-distance segments, and the end point of each segment is set as a temporary check point.

[0022] Based on real-time updated environmental, demand, and equipment data, the route is replanned and verified segment by segment.

[0023] When planning each flight segment, high-risk areas marked in the environmental data are avoided, and dynamic adjustments are made during the mission execution process.

[0024] When the drone flies to the temporary checkpoint, it replans the path for the next flight segment based on the latest received updated data. The updated data includes real-time changes in environmental data, adjustments to demand data, and real-time consumption of the drone's remaining battery power in the equipment data.

[0025] In a preferred embodiment of the present invention, the landing reservation and airspace and apron scheduling confirmation in step S3 include: the UAV sends reservation information to the landing node through short-range communication technology of the distributed mobile airport network. The reservation information includes the UAV's current location, the expected time window for arrival at the landing node, the required apron type, and the resupply requirements.

[0026] The airspace and landing site scheduling confirmation includes: the landing node reviews the landing site based on the local real-time resource status, including the landing site vacancy status in the equipment data, battery inventory, other UAV landing plans in the current task queue, and the local airspace occupancy status in the environmental data; if there is no resource conflict, the specific landing site number and the expected landing guidance signal frequency are specified through the scheduling confirmation command; if there is a resource conflict, the UAV is coordinated to adjust the landing time or recommended to an adjacent backup landing node.

[0027] In a preferred embodiment of the present invention, step S4, which involves simultaneously completing material loading and unloading, battery replacement, or mission payload adjustment, includes:

[0028] After the drone lands, the landing site linkage equipment automatically locks the drone's fuselage.

[0029] The robotic arm of the fast battery swapping system automatically removes the old battery and installs the new battery based on the battery status in the equipment data.

[0030] Based on the type and quantity of materials in the demand data, the material loading and unloading robotic arm can simultaneously complete the unloading and loading of materials.

[0031] For mission load adjustments, the dedicated replacement equipment simultaneously removes the original load module, installs the new mission load, and completes equipment power-on testing and parameter debugging.

[0032] During the operation, the edge computing module collects the operating data of each operating device in real time and compares it with the preset operating standards. If any abnormality occurs, the fault tolerance mechanism is triggered.

[0033] In a preferred embodiment of the present invention, step S5, updating the local resource inventory includes: based on the operation results of S4, correcting the material inventory, battery inventory, and floor status parameters, and recording equipment failures or material losses.

[0034] The synchronization to the entire network includes: broadcasting the updated resource data to all other edge nodes in a standardized format, containing only core resource change information;

[0035] The triggering of a new round of task bidding cycle includes: after each edge node updates the network resource status database, it automatically re-pulls the emergency needs in the task pool to be responded to, and generates a differentiated bidding strategy based on the updated resource distribution data.

[0036] In a preferred embodiment of the present invention, step S1, the construction of the distributed mobile airport network includes:

[0037] Node deployment sites are selected based on real-time disaster maps of the disaster area, prioritizing coverage of areas with high disaster density and urgent rescue needs, avoiding aftershock risk areas and flood-prone areas, ensuring that each node can cover a rescue range of five to ten kilometers, and controlling the distance between nodes within the coverage range of short-range communication.

[0038] Each mobile airport is equipped with an embedded edge computing module, a multi-source data acquisition module, a resource storage module, and a takeoff and landing control module; the multi-source data acquisition module includes environmental sensors, an image receiving unit, and a wireless signal receiver.

[0039] Inter-node communication is achieved through short-range communication technologies, including microwave, LoRa, or temporary networking methods.

[0040] This invention provides a remote take-off and landing system for unmanned aerial vehicles (UAVs) for low-altitude economic purposes, comprising:

[0041] The distributed node deployment module is used to build a mobile airport network in disaster areas and complete node deployment.

[0042] The multi-source data acquisition module is used to collect disaster data in real time, including environmental, demand, and equipment data.

[0043] The edge data preprocessing module is used to clean, weight, verify, and manage the timeliness of the collected data.

[0044] The task bidding and scheduling module is used to generate task bidding packages, select the optimal node, and trigger dynamic queue insertion.

[0045] The flight segment path planning module is used to realize dynamic path planning for UAVs in a rolling manner, segment by segment.

[0046] The take-off and landing resource reservation module is used to process drone landing reservations and confirm airspace and apron scheduling.

[0047] The ground synchronous operation module is used to complete the loading and unloading of materials, battery swapping, and payload adjustment of the UAV;

[0048] The resource synchronization update module is used to update the local resource inventory and synchronize it to the entire network to trigger the bidding cycle.

[0049] This invention addresses the shortcomings of the prior art and has the following beneficial effects:

[0050] (1) This invention constructs a distributed mobile airport network and endows each node with edge computing capabilities, enabling each mobile airport to independently collect and process multi-source disaster data without relying on the rear command center. In scenarios where communication is interrupted or information is frequently updated at the disaster site, this distributed architecture ensures autonomous collaboration and decision-making among nodes, directly avoiding the risk of the traditional centralized system being paralyzed due to the loss of connection at the center. Compared with the existing technology that relies excessively on stable communication links for global static planning, this method significantly improves the robustness and adaptability of the system. Furthermore, the real-time data processing of edge nodes provides a precise and dynamic decision-making basis for subsequent task scheduling, thereby enabling the continuous and efficient operation of rescue response in a highly uncertain environment in the disaster area.

[0051] (2) This invention introduces a task bidding mechanism and dynamic queue-jumping rules. Each mobile airport generates a bidding package based on its local resource status and broadcasts it. The optimal take-off and landing nodes are selected through a multi-dimensional comprehensive adaptability model. In the dynamic evolution of the disaster situation, this mechanism can quickly match tasks and resources and prioritize responses to high-urgent rescue needs. It directly solves the problem of traditional predefined task chains becoming invalid due to information updates. Compared with the limitations of rigid task allocation and inability to adjust midway in existing technologies, this method realizes flexible allocation of rescue resources and real-time optimization of task execution. Furthermore, by interrupting non-critical tasks and quickly switching to high-priority tasks, it ensures the timely fulfillment of core needs such as life rescue and improves the overall rescue efficiency and success rate.

[0052] (3) This invention adopts segment-based rolling path planning and landing reservation scheduling, which divides the UAV flight path into short segments and replans them in real time, and dynamically adjusts the route in combination with environmental changes; at the same time, the landing node confirms the airspace and apron position based on the resource status, avoiding take-off and landing conflicts; in the scenario of sudden environmental changes and frequent obstacles in the disaster area, this dynamic planning method directly ensures the flight safety and mission adaptability of the UAV. Compared with the defects of the existing technology, which is prone to failure due to environmental interference, this method significantly reduces the risk of mission interruption; furthermore, through precise take-off and landing resource reservation and scheduling, the waiting time of the UAV is reduced, and the seamless connection between mission execution and resource replenishment is realized, supporting the continuity and stability of rescue operations.

[0053] (4) This invention triggers a new round of task bidding cycle by updating the resource inventory after takeoff and landing operations and synchronizing it with the entire network, so that each mobile airport can dynamically adjust task allocation based on the latest resource status. In disaster scenarios where rescue resources are limited and demands are varied, this closed-loop mechanism directly ensures the real-time consistency of resource information and the continuous optimization of task scheduling. Compared with the resource mismatch problem caused by isolated resource management and delayed updates in the prior art, this method realizes the efficient utilization and rapid redistribution of rescue resources. Furthermore, through automated resource synchronization and bidding triggering, the system can continuously adapt to the evolution of the disaster situation and maintain the dynamic balance and long-term sustainability of rescue scheduling. Attached Figure Description

[0054] To more clearly illustrate the technical solutions in the embodiments of the present invention or 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 only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0055] Figure 1 This is a flowchart of a preferred embodiment of the present invention;

[0056] Figure 2 This is a task bidding and dynamic scheduling diagram according to a preferred embodiment of the present invention;

[0057] Figure 3 This is a preferred embodiment of the path planning and reservation scheduling diagram of the present invention. Detailed Implementation

[0058] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0059] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein. Therefore, the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0060] Application Overview:

[0061] This application addresses a dynamic task scheduling and resource coordination system for drones based on mobile airports in disaster emergency logistics and rescue scenarios. While existing technologies achieve forward-looking take-off and landing points through mobile airports, their core limitations lie in the unpredictable nature of task planning and the localized static nature of resource management. Existing systems rely on predefined task chains established in the early stages of rescue based on incomplete information; once disaster information is updated, the entire task chain becomes suboptimal or even ineffective. The system cannot, during task execution, suspend, merge, or replan the routes of multiple drones already in flight based on real-time, multi-source distress information. Furthermore, the batteries and supplies stored at each mobile airport are isolated, unable to be proactively and intelligently allocated and replenished according to dynamic changes in the needs of the disaster area.

[0062] The reason why existing technologies struggle to achieve dynamic optimization in disaster relief scenarios lies in the fundamental mismatch between their decision-making models and the highly uncertain environment of disaster areas. Their core algorithms typically perform one-time global optimizations based on deterministic conditions, while the core characteristics of disaster relief are incomplete, asymmetric, and continuously updated information. Such static models cannot absorb real-time changes for rolling decision-making. In terms of system architecture, there is an over-reliance on centralized computation through stable communication links with the rear command center. Given the frequent damage to communication infrastructure in disaster areas, once the central command loses contact with the mobile command center, the entire system's decision-making process collapses.

[0063] This application breaks with convention by abandoning the centralized architecture that relies on stable backend communication and global static planning, and innovatively proposes a distributed rescue scheduling method based on edge autonomy and swarm intelligence.

[0064] Exemplary method:

[0065] like Figure 1 As shown, a method for remote take-off and landing of unmanned aerial vehicles (UAVs) for low-altitude economy includes:

[0066] S1. Construct a distributed mobile airport network, deploy several mobile airports with edge computing capabilities in the disaster area to form a dynamic take-off and landing node group, collect multi-source disaster data in real time and perform edge preprocessing.

[0067] S2. Each mobile airport acts as an edge node, generating a task bidding package based on preprocessed multi-source disaster data, broadcasting it to all edge nodes and receiving feedback bidding information to determine the optimal task execution node, including: take-off node and landing node.

[0068] When a temporary high-priority task occurs, the system triggers a dynamic queue-jumping mechanism to interrupt the flight of the current non-critical task drone and guide it to land at a designated mobile airport to complete a rapid task switch.

[0069] S3. The drones controlled by the takeoff node adopt segment-based rolling path planning and initiate landing reservations to the landing node. The landing node confirms the airspace and apron scheduling based on resource status and task queue.

[0070] S4. The drone lands according to the reservation permission and simultaneously completes the loading and unloading of materials, battery replacement or mission payload adjustment on the take-off and landing platform.

[0071] S5. After the take-off and landing operation is completed, the mobile airport of the landing node updates its local resource inventory and synchronizes it to the entire network, triggering a new round of task bidding cycle based on the resource status change.

[0072] This invention proposes a method for remote take-off and landing of unmanned aerial vehicles (UAVs) based on a distributed mobile airport network. Step S1 focuses on building an adaptive and intelligent group of take-off and landing nodes.

[0073] In step S1, the deployment planning of mobile airports is carried out, and a distributed mobile airport network is constructed.

[0074] A distributed mobile airport network refers to a network system composed of several geographically dispersed mobile airports with data interaction and collaborative decision-making capabilities. The mobile airports do not operate in isolation, but achieve information exchange through short-range communication technologies, including microwave, LoRa, or temporary networking methods, ultimately forming a dynamic cluster of take-off and landing nodes covering the disaster area. Compared with existing mobile airport networks, which are mostly based on a central-node architecture, all data needs to be uploaded to the rear command center for processing. The nodes only execute instructions and do not have autonomous decision-making capabilities. Once the central communication is interrupted, the nodes become paralyzed.

[0075] When constructing a distributed mobile airport network, node deployment site selection should be based on real-time disaster maps of the disaster area, prioritizing coverage of areas with high disaster density and urgent rescue needs, such as the vicinity of temporary shelters and medical points. At the same time, dangerous areas such as aftershock risk zones and flood-prone areas should be avoided to ensure that each node can cover a rescue range of five to ten kilometers, and the distance between nodes should be controlled within the coverage range of short-range communication to avoid network blind spots.

[0076] After site selection, each mobile airport is configured with hardware to enable edge computing capabilities. The mobile airport configuration includes: an embedded edge computing module, a multi-source data acquisition module, a resource storage module, and a takeoff and landing control module. The embedded edge computing module is equipped with lightweight data processing algorithms and supports real-time data processing. The multi-source data acquisition module includes environmental sensors, an image receiving unit, and a wireless signal receiver. The resource storage module is used to store spare batteries and rescue supplies.

[0077] The hardware configuration of this application enables the mobile airport to become an intelligent node, capable of independently completing data collection and preliminary processing without waiting for central instructions.

[0078] Dynamic take-off and landing node clusters emphasize that these nodes can adjust their roles and resource status in real time according to the evolution of the disaster situation, rather than being fixed physical points.

[0079] Initiate the multi-source disaster data collection process;

[0080] Multi-source disaster data refers to data from multiple channels that reflects different dimensions of the disaster area, specifically including three categories:

[0081] The first category is environmental data, which is collected in real time by the environmental sensors on the mobile airport to collect real-time weather conditions and terrain obstacle information of the area.

[0082] The second category is demand data, which is rescue information reported by ground rescuers through portable terminals, including: location of people trapped, number of people, type of supplies needed, priority of demand, and image data of the disaster area transmitted back.

[0083] The third category is equipment data, including the remaining battery power, payload capacity, and current location of drones in operation, as well as the battery inventory and apron vacancy status of each mobile airport.

[0084] This application utilizes multi-source data collection to continuously capture the dynamics of the disaster area and ensure data timeliness.

[0085] The edge computing modules of each mobile airport will perform edge preprocessing on the collected multi-source disaster data, including:

[0086] The mobile airport's local algorithm first cleans the collected data and removes outliers;

[0087] In environmental data, the jump values ​​of sensors caused by vibration and severe weather are smoothed and corrected by the sliding window averaging method;

[0088] In the demand data, duplicate information mistakenly reported by ground terminals and invalid requests with coordinates outside the disaster area are directly filtered out based on preset geographical boundaries and data verification rules.

[0089] In the equipment data, real-time parameters such as the drone's remaining battery power and airport apron status are evaluated using a threshold judgment method to eliminate erroneous readings from faulty sensors, ensuring the accuracy of the basic data.

[0090] Classify and archive valid data, and assign dynamic priorities;

[0091] The demand data is prioritized based on the core criteria of life rescue > medical supplies > living supplies > general reconnaissance. At the same time, it takes into account the urgency of the mission reporting time and the number of people trapped, and calculates specific priority scores through quantitative calculation to ensure that high-urgency and high-impact missions are responded to first.

[0092] Environmental data is defined as constraint data, providing insurmountable hard constraints for UAV path planning, that is, defining flight safety boundaries based on environmental conditions to prevent UAVs from entering dangerous areas;

[0093] Equipment data is classified as resource data. Information such as the inventory of materials at mobile airports and the status parameters of drones must be recorded in real time and synchronized to the local resource ledger. This will provide accurate resource support for subsequent task bidding among mobile airports and ensure that task allocation matches resource capabilities.

[0094] For example, the priority score calculation formula is as follows: ,in, As the basic weight for task type, The urgency coefficient of the task is calculated based on the superposition of hazard factors. The time decay coefficient for the rescue window is determined using an exponential decay model. The coordination complexity coefficient is calculated based on the number of required coordination nodes and the technical difficulty. To determine the fit between currently available resources and task requirements, The weights for each dimension are dynamically adjusted based on the disaster situation to ensure that the model always aligns with real-world needs. This is an information update intensity correction term, calculated based on data update frequency and importance.

[0095] Subsequently, key information is extracted and compressed. Due to the limited bandwidth of short-range communication links in the disaster area, synchronizing the complete fused dataset to the entire network could easily cause channel congestion. Therefore, the edge computing module uses feature extraction algorithms to extract key information that plays a crucial role in task decision-making from the fused dataset.

[0096] The extracted key information is compressed in a lightweight manner, and data anonymization and format optimization techniques are used to remove redundant fields and convert image features into coordinate parameters; the text information is also converted into structured encoding, which greatly reduces the data volume.

[0097] To prevent erroneous data caused by terminal failure or signal interference from affecting decision-making, the edge computing module will activate a cross-validation mechanism.

[0098] The wind speed data collected by the environmental sensors is compared with the air wind speed data transmitted back by the UAV. If the difference exceeds 3 m / s, the average of the two data is used for correction based on terrain factors. The spatial overlap between the stranded location reported by the ground terminal and the suspected stranded area identified in the image data is checked. If the overlap is less than 60%, a secondary confirmation command is triggered, and a location verification request is sent to the ground terminal via short-range communication.

[0099] Meanwhile, to ensure data timeliness, timestamps are added to each type of key information, and update cycles are set: flight support data is updated every 3 minutes, rescue demand data is updated in real time, and resource scheduling data is updated every 1 minute. When a certain type of data has not been updated beyond the update cycle, the edge node will automatically send a data request to the associated data acquisition terminal. If there is no response after 3 consecutive requests, the terminal will be marked as offline, and its data will be removed from the fused dataset to avoid making decisions based on outdated information.

[0100] Step S1 involves scientifically deploying mobile airports with edge computing capabilities to build a distributed network, collecting multi-source disaster data in real time, including environmental, demand, and equipment data. Through edge preprocessing processes such as cleaning and denoising, classification and weighting, extraction and compression of key information, cross-validation, and timeliness management, the messy raw data is transformed into accurate and efficient decision support data. This not only solves the shortcomings of traditional centralized architectures that rely on back-end communication and have lagging data processing, but also lays a real-time dynamic data foundation for subsequent task scheduling. Under this premise, Step S2 will rely on this high-quality data to realize task bidding and dynamic queue scheduling among mobile airports, promoting the flexible and efficient allocation of rescue resources.

[0101] like Figure 2 As shown, a distributed task bidding and dynamic scheduling mechanism is used to achieve autonomous matching and priority adaptation of rescue tasks, breaking the rigid limitations of the centralized allocation of existing technologies.

[0102] The mission bidding package is a standardized information package formed by edge nodes based on multi-source data preprocessed by S1, integrating the core requirements and execution constraints of the rescue mission. It contains key information such as mission objectives, requirement priority scores, required material types and quantities, target location coordinates, latest response time limit, and flight safety boundaries, and is used for mission collaboration and competition evaluation among nodes.

[0103] Edge nodes are mobile airports with edge computing and collaborative decision-making capabilities. They are both data processing units and the main entities for task bidding and execution.

[0104] The optimal task execution node is the mobile airport most suitable for completing the task, selected through multi-dimensional evaluation. It is divided into takeoff nodes responsible for scheduling drone takeoffs and landing nodes for receiving drone resupply upon landing. The dynamic queue-jumping mechanism is a scheduling rule triggered by temporary high-priority tasks. It is used to interrupt non-critical tasks and prioritize the rapid execution of core tasks.

[0105] Each mobile airport acts as an edge node, calling upon the demand data, equipment data, and environmental data output by S1 to extract core mission elements. From the demand data, it obtains the coordinates of the stranded location, the type of required supplies, the demand priority score, and the number of stranded personnel. From the equipment data, it matches the payload capacity and remaining battery power of locally available drones. From the environmental data, it extracts the flight safety boundary of the target area. Then, it integrates this information according to a preset standardized format to generate a mission bidding package. Compared to existing technologies, where mission allocation is directly assigned to execution nodes by the rear command center based on a pre-planned mission chain, and the nodes only passively receive instructions without participating in mission feasibility assessment, the command center cannot monitor the resource dynamics of each node in real time, which can easily lead to mismatches such as having missions but no resources or resources being idle but no missions.

[0106] This application enables edge nodes to autonomously generate bidding packages. Since each node directly controls local real-time resources and environmental data, it can accurately reflect the feasibility and cost of task execution, avoid information lag in remote allocation, and ensure that all nodes can quickly parse the bidding content, thus improving collaboration efficiency.

[0107] After generating the task bidding package, each edge node broadcasts the task bidding package to all other edge nodes through the short-range communication technology of the distributed mobile airport network built by S1.

[0108] Receiving feedback on bidding information means that after receiving a bidding package, other edge nodes, based on their own device data, environmental data, and local task queue, evaluate their comprehensive ability to execute the task, generate bidding response information including execution cost, estimated arrival time, estimated resource consumption, probability of task completion, and current task busyness, and then feed it back to the edge node that initiated the bidding package.

[0109] For example, after receiving a bid package, an edge node retrieves the remaining battery power, payload capacity, and current location of its local drone from the device data, as well as the battery inventory and apron availability of the mobile airport. Combined with the wind speed and terrain obstacle information from the local area to the target area in the environmental data, it calculates the estimated flight time and power consumption to complete the task. Then, based on the number of tasks it has already undertaken, it determines whether it has the capacity to undertake new tasks, and finally forms a complete bid response. Compared with the centralized allocation in the existing technology, nodes do not need to report their own status, and the command center can only allocate tasks based on initial information, which cannot cope with dynamic changes in node resources.

[0110] The broadcast and feedback mechanism of this application allows the initiating node to comprehensively collect real-time capability information of all nodes, providing a complete basis for optimal allocation. Furthermore, short-range communication does not rely on the rear command center, and even if communication is interrupted, nodes can still autonomously complete information exchange to ensure uninterrupted task allocation.

[0111] After receiving all bidding responses, the edge node that initiated the bidding package starts the optimal task execution node selection process. The optimal task execution node refers to the edge node with the strongest overall execution capability, lowest resource consumption, and fastest response speed. It is divided into takeoff nodes and landing nodes. The former is responsible for dispatching drones to perform the mission, and the latter is responsible for landing and resupplying the drones after they complete the mission.

[0112] The selection process employs a multi-dimensional comprehensive evaluation model. The overall suitability score for each bidding node is calculated using the following representative formula, and the node with the highest score is selected as the optimal execution node:

[0113] The overall formula for overall fit is: ;in The response speed score is determined by both the estimated arrival time and the urgency of the task; the faster the response and the higher the task priority, the higher the score. The task completion probability is assigned a score, which is calculated based on the current equipment status and environmental risk coefficient. The higher the completion probability, the higher the score. The score is based on resource consumption and redundancy adaptation, taking into account the energy consumption, load occupancy and spare resources required for the task. The lower the consumption and the more redundancy, the higher the score. To assess the coordination and resupply integration score, the efficiency of coordination and the matching degree of resupply resources between candidate landing nodes and candidate takeoff nodes of UAVs are evaluated. The smoother the connection and the more sufficient the resupply, the higher the score. The historical task reliability adaptation score is calculated by weighting the completion rate, timeliness and frequency of anomalies of the past tasks at that node. The more stable the historical performance, the higher the score. The task conflict and load pressure adaptation score is determined by assessing the temporal and spatial overlap and resource competition between the current task load of the node and new tasks. The lighter the load and the fewer the conflicts, the higher the score. The resource recovery capability score measures a node's ability to quickly restore energy, repair equipment, and redeploy after completing a task. The shorter the recovery cycle and the higher the stability, the higher the score. These are the basic weighting coefficients for each dimension, which are dynamically adjusted based on the task type. These are correction factors for environmental urgency and time sensitivity; For the rate of environmental degradation, This is a task delay risk index. As a weighting factor for resource recovery, This represents the network connectivity coefficient.

[0114] Finally, after weighted fusion and dynamic correction, each adaptation score generates a comprehensive evaluation value, which is used to rank candidate nodes and prioritize the one with the highest adaptation degree to execute tasks, ensuring that the system can achieve efficient, stable and sustainable collaborative operation in complex environments.

[0115] The dynamic queue-jumping mechanism is an emergency scheduling rule designed by the system for temporary high-priority tasks. It means that when a task with a priority score exceeding a preset threshold appears in the demand data collected by S1, the current non-critical task is automatically interrupted to prioritize the execution of the high-urgent task.

[0116] The system monitors the priority of all tasks being executed across the network in real time through a distributed network. When the demand data for a temporary high-priority task is transmitted to any edge node, the node immediately generates a high-priority task bidding package and broadcasts it to the entire network. At the same time, the system automatically identifies drones that are currently executing non-critical tasks and whose demand priority score is lower than a preset threshold, and sends them a task interruption command.

[0117] After receiving instructions, the drone accesses environmental and equipment data from the S1 system to select the nearest mobile airfield with available landing sites and rapid battery swapping capabilities as a temporary landing point. The system simultaneously notifies the temporary landing point to reserve landing sites, prepare rapid battery swapping equipment, and prepare supplies needed for the new mission. After adjusting its flight path and landing, the drone simultaneously unloads supplies, replaces batteries, or adjusts its payload for the current mission. It then quickly receives new high-priority mission instructions and transitions from the temporary landing point to a new takeoff point to execute its mission. In contrast, existing technologies lack a dynamic queueing mechanism, where all missions are executed in a pre-planned order. Even life-threatening high-priority missions require waiting for preceding low-priority missions to complete, potentially leading to missed rescue windows.

[0118] Step S2 utilizes distributed task bidding, allowing each edge node to autonomously assess its capabilities and provide bidding information. This fully integrates real-time resource and environmental data, ensuring precise matching between task allocation and node capabilities, thus avoiding resource waste. Bidding information feedback and comprehensive suitability calculations, through multi-dimensional quantitative evaluation, select the optimal node, significantly improving task completion success rate and efficiency. The dynamic queue-jumping mechanism directly addresses the pain point of slow response times for temporary high-priority tasks by interrupting non-critical tasks to prioritize core tasks, aligning with the paramount need for life in disaster relief scenarios.

[0119] Step S2 relies on the high-quality data preprocessed in S1. Each mobile airport edge node autonomously generates a task bidding package and broadcasts it. After receiving feedback, it uses a multi-dimensional comprehensive adaptability model to select the optimal take-off and landing nodes. At the same time, it uses a dynamic queue-jumping mechanism to interrupt non-critical tasks and prioritize high-priority rescue needs. This completely breaks the rigidity of traditional centralized task allocation and achieves precise dynamic matching of rescue resources and task requirements. Based on this, step S3 will focus on the segment-based rolling path planning of UAVs and the airspace and apron reservation scheduling of landing nodes to provide flight and site support for efficient mission landing.

[0120] like Figure 3As shown, by using dynamic path planning and precise take-off and landing reservation scheduling, the problem of traditional static paths being unable to adapt to changes in the disaster area environment and conflicts in take-off and landing resources is solved, providing flight and site support for UAVs to perform missions efficiently and safely.

[0121] Segment-based rolling path planning refers to a dynamic planning method that breaks down the complete flight path of a UAV from takeoff to landing into multiple consecutive short-distance segments, and replans and verifies the path segment by segment by combining real-time updated environmental data and mission status.

[0122] Unlike existing technologies that determine a fixed path for the entire process at once, this approach achieves real-time adaptation to dynamic environments through segmented planning and rolling updates.

[0123] After completing each flight segment, the drone replans the next segment based on the latest perception data, effectively avoiding sudden obstacles, weather changes, and no-fly zones, ensuring that the path is always optimal.

[0124] Landing reservation is a request for takeoff and landing resources initiated by a drone to the target landing node after the planning of each flight segment is completed. It includes information such as the expected landing time, the type of landing site required, and resupply needs, so that the landing node can coordinate resources in advance.

[0125] Airspace and apron scheduling confirmation is the process by which landing nodes review drone landing reservations based on their real-time resource status and task queues, clarify whether landing is permitted, designate specific apron locations, and coordinate airspace avoidance.

[0126] When a drone undergoing takeoff node control initiates segment-based rolling path planning, path splitting rules are determined based on the preprocessed environmental data, demand data, and equipment data from S1.

[0127] The route splitting is based on: the distribution of terrain obstacles and wind speed variation areas in the environmental data; the location of the mission target and the latest response time in the requirements data; and the remaining battery power and payload status of the drone in the equipment data. The entire route is divided into several segments of 5-10 kilometers each, with the end of each segment set as a temporary checkpoint to ensure that the drone still has sufficient battery power to handle route adjustments at the end of each segment. After splitting, the drone plans the optimal route for the first segment based on the initial environmental data provided at the takeoff node, avoiding high-risk areas marked in the environmental data during the planning process.

[0128] Once the first flight segment is planned, the UAV will immediately carry out the flight mission, while simultaneously receiving real-time updated data from the takeoff node during the flight.

[0129] The updated data includes real-time changes in environmental data in S1, such as sudden gusts of wind, new terrain obstacle information, adjustments to required data such as minor adjustments to the mission target location, and real-time consumption of the device's remaining battery power. When the drone flies to the temporary checkpoint of the first flight segment, a rolling update process is initiated to replan the path for the second flight segment based on the latest received data.

[0130] During replanning, if environmental conditions along the original planned route deteriorate, such as strong winds appearing in previously safe areas or new temporary high-priority missions requiring airspace, the route will be automatically adjusted to avoid risky areas or give way to high-priority mission routes. If the drone's remaining battery power is lower than expected, the route will be optimized to shorten the flight distance, prioritizing reaching the nearest mobile airport for refueling. This process is repeated until the drone completes the final flight segment and arrives near the landing node.

[0131] After completing the planning of the first flight segment, the drone made a landing reservation to the landing node.

[0132] The reservation information is transmitted via short-range communication technology through a distributed mobile airport network. The information includes the drone's current location, the estimated time window for arrival at the landing node, the type of landing site required (e.g., a dedicated landing site for large drones), and whether resupply services such as battery replacement or cargo loading and unloading are needed.

[0133] Once the landing node receives the reservation, it immediately retrieves the local real-time resource status, including the vacancy status of the landing site in the equipment data, battery inventory, other drone landing plans in the current task queue, and the local airspace occupancy status in the environmental data.

[0134] The landing node processes this information rapidly via an edge computing module. If the expected landing time does not conflict with the existing task queue and there are matching available landing sites and the necessary resupply resources, a scheduling confirmation command is sent to clearly inform the drone of the specific landing site number, the expected landing guidance signal frequency, and precautions. If there is a resource conflict, the drone is coordinated to adjust its landing time or recommended to an adjacent backup landing node, and the drone's flight segment planning system is updated simultaneously to guide it in adjusting its subsequent flight path. Upon receiving the scheduling confirmation command, the drone immediately updates its flight path and landing parameters, entering the stable approach phase. As it approaches the landing node, its onboard high-precision positioning system completes signal docking with the ground guidance equipment, achieving centimeter-level precision landing.

[0135] In existing technologies, drone path planning mostly adopts one-time global static planning, with the entire path fixed before takeoff. This cannot cope with sudden environmental changes in disaster areas, such as new terrain obstacles caused by aftershocks or sudden gusts of wind, nor can it adapt to dynamic adjustments during mission execution, such as temporary changes in mission priority. Therefore, drones often fail to complete missions due to environmental deterioration along the path, or are forced to hover and wait after arriving at the landing point because they have not reserved takeoff and landing resources in advance, which seriously affects rescue efficiency.

[0136] This application adopts segment-based rolling path planning, which incorporates dynamic factors such as environmental changes and mission adjustments into the planning process in real time through segment-by-segment updates, greatly improving the safety and adaptability of the path. The landing reservation and scheduling confirmation mechanism allows landing nodes to coordinate resources in advance, avoiding resource conflicts after the drone arrives and ensuring efficient connection of take-off and landing operations.

[0137] Through segment-based rolling planning, UAVs can avoid unexpected risks in real time, ensuring flight safety and solving the problem that static paths in existing technologies cannot cope with the dynamic environment of disaster areas. Landing reservations and scheduling confirmations enable advance coordination of takeoff and landing resources, avoiding resource waste and conflicts, and improving the overall system efficiency. All planning and reservation processes are completed within the distributed mobile airport network. Even if a partial communication link is interrupted, the UAV can still complete the planning and execution of the current segment based on the latest acquired data, ensuring uninterrupted mission execution.

[0138] Step S3, relying on the real-time data support of S1 and the optimal take-off and landing nodes determined in S2, dynamically adjusts the UAV's flight path through segment-based rolling path planning. This avoids sudden environmental risks and mission changes in disaster areas in real time. Simultaneously, by leveraging landing reservation and airspace / apron scheduling confirmation mechanisms, it coordinates take-off and landing resources in advance, resolving the problems of poor adaptability and resource conflicts inherent in traditional static planning. Building on this foundation, Step S4 focuses on practical aspects such as material loading and unloading, battery replacement, and mission payload adjustment after the UAV's precise landing, achieving an efficient closed-loop for mission coordination.

[0139] Step S4 uses an automated synchronous operation mode to achieve rapid resupply and mission handover after the drone lands.

[0140] The authorization without prior reservation is the result of the S3 airspace and apron scheduling confirmation issued by the landing node to the UAV. It contains specific landing instructions, apron information, and operational requirements. It is the core basis for the UAV to perform landing and subsequent operations.

[0141] Synchronous operation mode refers to an efficient operation method in which, after the drone lands, it can complete the loading and unloading of materials, battery replacement, and mission payload adjustment in parallel within the same time period. It is different from the traditional serial operation process and improves the efficiency of ground operation through process optimization and equipment linkage.

[0142] The rapid battery swapping system is an automated battery replacement device deployed at mobile airports. It can quickly remove and install new batteries for drones without manual intervention. Mission payload adjustment involves replacing or adjusting equipment carried by the drone, such as relief supply boxes, reconnaissance cameras, and communication relay modules, according to new mission requirements.

[0143] Based on the reservation permission issued at the landing node, the drone smoothly lands on the designated apron by connecting with the ground guidance equipment through a high-precision positioning system. After landing, the corresponding linkage equipment at the apron immediately activates, automatically locking the drone's fuselage to ensure stability during the operation, significantly reducing the operation time compared to traditional manual guided landings.

[0144] Next, the system initiates the synchronous operation process; the robotic arm of the rapid battery swapping system automatically identifies the battery interface type based on the battery status of the drone in the S1 device data, accurately completes the removal of the old battery and the installation of the new battery, and the entire battery swapping process is controlled within 1-2 minutes.

[0145] At the same time, the material loading and unloading robotic arm can simultaneously complete operations such as opening the hatch, unloading materials (e.g., transferring relief supplies carried by the drone to the mobile airport storage area), and loading materials (e.g., loading medical supplies or living supplies required for a new mission into the drone's cargo hold) based on the type and quantity of materials in the demand data.

[0146] For scenarios requiring payload adjustments, such as switching from a material delivery mission to a disaster reconnaissance mission, the dedicated replacement equipment will simultaneously remove the original payload module, install the new mission payload, and complete equipment power-on testing and parameter debugging to ensure that the payload functions normally.

[0147] During synchronous operations, the mobile airport's edge computing module collects real-time operational data from each piece of equipment and compares it with preset operational standards. If any anomalies occur, such as improper battery installation or deviations in the quantity of materials loaded, a local alarm is immediately triggered, and a fault-tolerance mechanism is activated. The battery swapping operation is then re-executed, and the position of the loading / unloading robotic arm is adjusted to ensure operational quality. After the operation is completed, the system automatically unlocks the aircraft, sends an operation completion command to the drone, and simultaneously synchronizes the operation results, such as battery replacement status, quantity of materials loaded / unloaded, and load adjustment status, to the equipment data in S1, updating the local resource ledger.

[0148] In step S4, after the drone lands precisely according to the reservation permission, it completes the loading and unloading of materials, battery replacement and mission payload adjustment in parallel through a multi-device collaborative synchronous operation mode. The fully automated operation replaces the traditional manual serial process, which greatly shortens the ground dwell time and improves the efficiency and accuracy of resupply. This lays the foundation for the drone to be quickly put into the next round of missions. After that, step S5 will trigger a new round of mission bidding cycle through resource inventory updates and full network synchronization, realizing the dynamic closed-loop operation of rescue dispatch.

[0149] Step S5 involves real-time updates of resource status and synchronization across the entire network to trigger dynamic iterations of the task bidding cycle.

[0150] Local resource inventory updates refer to the process by which the mobile airport at the landing node, after completing the loading and unloading of supplies, battery replacement, and mission payload adjustment in S4, corrects and records its own resource status in real time. Network-wide synchronization involves synchronizing the updated local resource information to all edge nodes via short-range communication technology within the distributed mobile airport network. Mission bidding loops, based on the updated network-wide resource status, restart the S2 mission generation, bidding, and node selection processes to achieve dynamic reallocation of rescue resources.

[0151] After the drone completes its ground operations in S4, the mobile airport at the landing node immediately initiates the local resource inventory update process.

[0152] The edge computing module calls the device data template of S1 and corrects various resource parameters according to the actual operation results: In terms of material inventory, if the material unloading is completed, the inventory quantity of the corresponding material is increased; if the material loading is completed, the inventory quantity of the corresponding material is decreased; in terms of battery inventory, the old batteries replaced by the drone are marked as waiting to be charged, and the inventory quantity of newly replaced batteries is decreased; in terms of landing status, the landing space occupied by the drone is updated from occupied to idle, and the operation completion time of the landing space is recorded simultaneously.

[0153] In addition, if equipment failures occur during the operation, such as abnormalities in the battery swapping system or material losses, they will also be recorded in the local resource ledger to provide a basis for subsequent equipment maintenance and resource allocation.

[0154] After the local resource inventory is updated, the landing node broadcasts the updated resource data to all other edge nodes in a standardized format through the distributed mobile airport network built by S1. To avoid data transmission congestion caused by limited communication bandwidth in the disaster area, the synchronized data only includes core resource change information, such as increases or decreases in material inventory, available battery quantity, and changes in apron status, rather than a complete resource ledger.

[0155] Upon receiving the synchronization information, each receiving node immediately updates its local database of network-wide resource status to ensure that all edge nodes maintain consistent resource information, providing unified and accurate basic data for subsequent task bidding.

[0156] Once the network-wide resource status is synchronized, the system automatically triggers a new round of task bidding. It re-retrieves emergency requests from the pending task pool and, combined with updated resource distribution data, generates differentiated bidding strategies for each mobile airport.

[0157] This application completely solves the problems of asynchronous technical resource information and untimely task adjustments in the existing technology through automated local resource updates, efficient full network synchronization, and dynamic bidding cycle triggering mechanism.

[0158] Exemplary system:

[0159] A remote take-off and landing system for unmanned aerial vehicles (UAVs) for low-altitude economy includes:

[0160] The distributed node deployment module is used to build a mobile airport network in disaster areas and complete node deployment.

[0161] The multi-source data acquisition module is used to collect disaster data in real time, including environmental, demand, and equipment data.

[0162] The edge data preprocessing module is used to clean, weight, verify, and manage the timeliness of the collected data.

[0163] The task bidding and scheduling module is used to generate task bidding packages, select the optimal node, and trigger dynamic queue insertion.

[0164] The flight segment path planning module is used to realize dynamic path planning for UAVs in a rolling manner, segment by segment.

[0165] The take-off and landing resource reservation module is used to process drone landing reservations and confirm airspace and apron scheduling.

[0166] The ground synchronous operation module is used to complete the loading and unloading of materials, battery swapping, and payload adjustment of the UAV;

[0167] The resource synchronization update module is used to update the local resource inventory and synchronize it to the entire network to trigger the bidding cycle;

[0168] Specifically, the distributed node deployment module is used to build a mobile airport network covering the disaster area and to complete the scientific deployment of nodes, which is the physical foundation of the entire system.

[0169] The module uses the LT-MP-200 modular mobile take-off and landing platform as its core carrier, combined with a Trimble R12i GNSS receiver to acquire real-time geographic location information. Node site selection is completed based on the real-time disaster map of the disaster area. Priority deployment is given to areas with high rescue demand, such as temporary shelters and medical points, avoiding dangerous areas such as aftershock risk zones. This ensures that each node covers a range of 5-10 kilometers, and the node spacing is controlled within the coverage area of ​​the Haige Communications HH-600 microwave communication equipment, avoiding network blind spots. After deployment, the module uses a Semtech SX1302 LoRa module to achieve initial networking between nodes, providing communication support for subsequent multi-source data acquisition and task bidding modules. The output node location, communication coverage, and other basic data will serve as the basis for the acquisition boundaries of the multi-source data acquisition module.

[0170] The multi-source data acquisition module is used to collect three core types of data in the disaster area in real time: environment, needs, and equipment, providing raw data input for system decision-making. This module is equipped with an environmental sensor group including a Sensirion SHT35 temperature and humidity sensor and a Gill WindMaster Pro 3D anemometer, collecting real-time environmental data such as weather and terrain obstacles; it receives data on trapped locations and material needs reported by ground rescue personnel via the Beidou Hailiao HL200 portable terminal; and it uses a DJI DS-M100 monitoring module to collect data on the remaining battery power and payload capacity of connected drones, as well as equipment data such as mobile airport battery inventory and apron status recorded by the Delixi DLX-LX-01 infrared sensor. The collected data is transmitted in real time to the edge data preprocessing module via a Huawei ME909s-8215G industrial module. Environmental data serves as a constraint for path planning, demand data provides a basis for task priority determination, and equipment data supports resource scheduling decisions.

[0171] The edge data preprocessing module cleans, weights, verifies, and manages the timeliness of raw data transmitted from the multi-source data acquisition module, outputting high-quality decision-making data. This module relies on the NVIDIA Jetson AGX Orin edge computing unit to run lightweight algorithms: correcting abrupt changes in environmental sensor values ​​using a sliding window mean method; filtering invalid requests from the required data based on preset geographical boundaries; eliminating erroneous readings from device data using a threshold judgment method; calculating scores using a priority formula; and initiating a cross-validation mechanism and adding timestamps. The processed, standardized data is synchronized to the task bidding scheduling module, providing accurate data support for the generation of task bidding packages.

[0172] The task bidding and scheduling module generates task bidding packages based on pre-processed edge data, filters optimal execution nodes, and triggers a dynamic queue-jumping mechanism to achieve precise matching of tasks and resources. This module integrates task objectives and priority scores from demand data, resource status from equipment data, and safety boundaries from environmental data using the UNO-2484G industrial controller. It generates standardized bidding packages containing elements such as target location and latest response time, and broadcasts them across the entire network. After receiving bidding information such as execution costs and estimated arrival times from each node, it runs a multi-dimensional adaptation algorithm to filter optimal takeoff and landing nodes. When an emergency task is inserted, the module interrupts non-critical mission drones via the DJI DB-M300 command transmitter, guiding them to land at a designated node. The filtered optimal node information is transmitted to the flight path planning module as the core basis for determining the path's start and end points.

[0173] The flight segment path planning module is used to achieve dynamic, rolling path planning for UAVs based on takeoff and landing nodes determined by the task bidding and scheduling module and real-time environmental data from the edge preprocessing module. This module uses the Atlas 200I DKA2 edge AI computing box. Based on the terrain obstacle distribution, wind speed variation areas, and remaining battery power of the UAV in the environmental data, it divides the entire path into several segments of 5-10 kilometers, with the endpoint of each segment designated as a temporary checkpoint. When the UAV reaches a checkpoint, the module replans the path for the next flight segment based on the latest environmental data (such as sudden gusts) and demand data (such as minor adjustments to the mission position), avoiding risky areas or shortening the distance to ensure battery power. The planned flight segment path and estimated arrival time are synchronized to the takeoff and landing resource reservation module as core parameters for landing reservations.

[0174] The takeoff and landing resource reservation module receives path information from the flight segment path planning module, processes UAV landing reservations and confirms airspace and apron scheduling, ensuring efficient utilization of takeoff and landing resources. If there are no resource conflicts, it provides the specific apron number and guidance signal frequency; if conflicts exist, it coordinates adjustments to landing times or recommends alternative nodes, and sends the scheduling results back to the flight segment path planning module to adjust subsequent paths. Simultaneously, the module synchronizes successfully reserved takeoff and landing information to the ground synchronization operation module, triggering the resource preparation process in advance.

[0175] The ground synchronization operation module, based on the scheduling confirmation information from the takeoff and landing resource reservation module, synchronously completes material loading and unloading, battery swapping, and payload adjustment after the UAV lands, enabling rapid task transition. This module is equipped with an EHang Intelligent EH-SW-002 automatic battery swapping robotic arm, and a KUKA KR C4 compact collaborative robotic arm, which simultaneously performs door opening and material loading / unloading operations according to the material type and quantity in the demand data. During the operation, the module transmits real-time progress data to the resource synchronization update module as the basis for updating resource status.

[0176] The resource synchronization update module receives the operation results from the ground synchronization module, updates the local resource inventory, and synchronizes it to the entire network, triggering a new round of task bidding cycle and forming a system closed loop. After the UAV completes battery swapping and material adjustment, the system automatically verifies the payload status and takeoff conditions. Once confirmed, it releases takeoff performance permission, and the relevant data is synchronized to the task bidding and scheduling module to participate in the next round of task response. The entire process achieves low-latency closed-loop control through edge computing and cloud collaboration, significantly improving the efficiency of multi-UAV cluster operations and resource turnover rate, providing a replicable technical paradigm for the large-scale operation of urban low-altitude logistics networks.

[0177] Based on the preferred embodiments of the present invention described above, those skilled in the art can make various changes and modifications without departing from the inventive concept. The technical scope of this invention is not limited to the contents of the specification, but must be determined according to the scope of the claims.

Claims

1. A method for remote take-off and landing of unmanned aerial vehicles (UAVs) for low-altitude economic purposes, characterized in that, Includes the following steps: S1. Construct a distributed mobile airport network, deploy several mobile airports with edge computing capabilities in the disaster area to form a dynamic take-off and landing node group, collect multi-source disaster data in real time and perform edge preprocessing. S2. Each mobile airport acts as an edge node, generating a task bidding package based on preprocessed multi-source disaster data. This package is broadcast to all edge nodes and receives feedback bidding information to determine the optimal task execution node, including takeoff and landing nodes. The task bidding package includes: task objectives, priority score, required material type and quantity, target location coordinates, latest response time limit, and flight safety boundary. The optimal task execution node is selected through a multi-dimensional comprehensive adaptability model. The model includes response speed adaptability score, task completion probability adaptability score, resource consumption and redundancy adaptability score, coordination and replenishment adaptability score, historical task reliability adaptability score, task conflict and load pressure adaptability score, and resource recovery capability adaptability score, and is dynamically adjusted in combination with environmental urgency correction factor and time sensitivity correction factor. S3. The drones controlled by the takeoff node adopt segment-based rolling path planning and initiate landing reservations to the landing node. The landing node confirms the airspace and apron scheduling based on resource status and task queue. S4. The drone lands according to the reservation permission and simultaneously completes the loading and unloading of materials, battery replacement or mission payload adjustment on the take-off and landing platform. S5. After the take-off and landing operation is completed, the mobile airport of the landing node updates its local resource inventory and synchronizes it to the entire network, triggering a new round of task bidding cycle based on the resource status change; Step S3, segment-based rolling path planning, includes: The complete flight path of the drone from the take-off point to the landing point is divided into multiple consecutive short-distance segments, and the end point of each segment is set as a temporary check point. Based on real-time updated environmental, demand, and equipment data, the route is replanned and verified segment by segment. When planning each flight segment, high-risk areas marked in the environmental data are avoided, and dynamic adjustments are made during the mission execution process. When the drone flies to the temporary checkpoint, it replans the path for the next flight segment based on the latest received updated data. The updated data includes real-time changes in environmental data, adjustments to demand data, and real-time consumption of the drone's remaining battery power in the equipment data.

2. The method for remote take-off and landing of unmanned aerial vehicles (UAVs) for low-altitude economic purposes according to claim 1, characterized in that: Step S2 also includes a dynamic queue-jumping mechanism: when a temporary high-priority task occurs, the system triggers the dynamic queue-jumping mechanism to interrupt the flight of the current non-critical task drone and guide it to land at a designated mobile airport to complete a rapid task switchover; the dynamic queue-jumping mechanism includes: generating and broadcasting a high-priority task bidding package, identifying and interrupting non-critical task drones, selecting temporary landing points, and simultaneously preparing rapid battery swapping equipment and materials required for the new task; the priority score of the temporary high-priority task exceeds a preset threshold, and the non-critical task is a task with a priority score lower than the preset threshold.

3. The method for remote take-off and landing of unmanned aerial vehicles (UAVs) for low-altitude economic purposes according to claim 1, characterized in that: The edge preprocessing in step S1 includes: The collected multi-source disaster data is cleaned and outliers are removed. Among them, environmental data is smoothed and corrected using the sliding window mean method, demand data is filtered for invalid requests based on preset geographical boundaries and data verification rules, and equipment data is removed by threshold judgment method to eliminate erroneous readings. Valid data is classified and archived, and assigned dynamic priorities; demand data is prioritized based on the core criteria of life rescue > medical supplies > living supplies > general reconnaissance, and priority scores are calculated in combination with the urgency of mission reporting time and the number of people trapped. Key information is extracted and compressed by using feature extraction algorithms to extract key information from the fused dataset and then using data desensitization and format optimization techniques for lightweight compression. A cross-validation mechanism is activated to compare the data collected by environmental sensors with the data transmitted back by the UAV, and to verify the spatial overlap between the location reported by the ground terminal and the image data. Add a timestamp to each type of key information and set an update cycle. When the data has not been updated within the update cycle, automatically send a data request to the associated data collection terminal.

4. A method for remote take-off and landing of unmanned aerial vehicles (UAVs) for low-altitude economic purposes according to claim 1, characterized in that: The landing reservation and airspace and apron scheduling confirmation in step S3 include: the UAV sends reservation information to the landing node through short-range communication technology of the distributed mobile airport network. The reservation information includes the UAV's current location, the expected time window for arrival at the landing node, the required apron type, and the resupply requirements. The airspace and landing site scheduling confirmation includes: the landing node reviews the landing site based on the local real-time resource status, including the landing site vacancy status in the equipment data, battery inventory, other UAV landing plans in the current task queue, and the local airspace occupancy status in the environmental data; if there is no resource conflict, the specific landing site number and the expected landing guidance signal frequency are specified through the scheduling confirmation command; if there is a resource conflict, the UAV is coordinated to adjust the landing time or recommended to an adjacent backup landing node.

5. A method for remote take-off and landing of unmanned aerial vehicles (UAVs) for low-altitude economic purposes according to claim 1, characterized in that: In step S4, the simultaneous completion of material loading and unloading, battery replacement, or mission load adjustment includes: After the drone lands, the landing site linkage equipment automatically locks the drone's fuselage. The robotic arm of the fast battery swapping system automatically removes the old battery and installs the new battery based on the battery status in the equipment data. Based on the type and quantity of materials in the demand data, the material loading and unloading robotic arm can simultaneously complete the unloading and loading of materials. For mission load adjustments, the dedicated replacement equipment simultaneously removes the original load module, installs the new mission load, and completes equipment power-on testing and parameter debugging. During the operation, the edge computing module collects the operating data of each operating device in real time and compares it with the preset operating standards. If any abnormality occurs, the fault tolerance mechanism is triggered.

6. A method for remote take-off and landing of unmanned aerial vehicles (UAVs) for low-altitude economic purposes according to claim 1, characterized in that: In step S5, updating the local resource inventory includes: based on the operation results of S4, correcting the material inventory, battery inventory, and floor status parameters, and recording equipment failures or material losses. The synchronization to the entire network includes: broadcasting the updated resource data to all other edge nodes in a standardized format, containing only core resource change information; The triggering of a new round of task bidding cycle includes: after each edge node updates the network resource status database, it automatically re-pulls the emergency needs in the task pool to be responded to, and generates a differentiated bidding strategy based on the updated resource distribution data.

7. A method for remote take-off and landing of unmanned aerial vehicles (UAVs) for low-altitude economic purposes according to claim 1, characterized in that: In step S1, constructing the distributed mobile airport network includes: Node deployment sites are selected based on real-time disaster maps of the disaster area, prioritizing coverage of areas with high disaster density and urgent rescue needs, avoiding aftershock risk areas and flood-prone areas, ensuring that each node can cover a rescue range of five to ten kilometers, and controlling the distance between nodes within the coverage range of short-range communication. Each mobile airport is equipped with an embedded edge computing module, a multi-source data acquisition module, a resource storage module, and a takeoff and landing control module; the multi-source data acquisition module includes environmental sensors, an image receiving unit, and a wireless signal receiver. Information exchange between nodes can be achieved through short-range communication technologies, including microwave, LoRa, or temporary networking methods.

8. A remote take-off and landing system for unmanned aerial vehicles (UAVs) for low-altitude economic purposes, based on the remote take-off and landing method for UAVs for low-altitude economic purposes as described in any one of claims 1-7, characterized in that, include: The distributed node deployment module is used to build a mobile airport network in disaster areas and complete node deployment. The multi-source data acquisition module is used to collect disaster data in real time, including environmental, demand, and equipment data. The edge data preprocessing module is used to clean, weight, verify, and manage the timeliness of the collected data. The task bidding and scheduling module is used to generate task bidding packages, select the best nodes, and trigger dynamic queue insertion. The flight segment path planning module is used to realize dynamic path planning for UAVs in a rolling manner for each flight segment. The take-off and landing resource reservation module is used to process drone landing reservations and confirm airspace and apron scheduling. The ground synchronous operation module is used to complete the loading and unloading of materials, battery swapping, and payload adjustment of UAVs; The resource synchronization update module is used to update the local resource inventory and synchronize it to the entire network to trigger the bidding cycle.

Images