A method for ground logistics vehicle and unmanned aerial vehicle cooperative distribution

Abstract

The application provides a ground logistics vehicle and unmanned aerial vehicle cooperative distribution method, and relates to the technical field of low-altitude logistics cooperative scheduling, which comprises the following steps: step 1, dynamically constructing a reachability cone for each unmanned aerial vehicle stop according to the to-be-distributed order; judging the spatial inclusion relationship between the destination coordinates of the to-be-distributed order and the reachability cone, determining the unmanned aerial vehicle reachable candidate order, and establishing a stop-candidate order set mapping relationship; step 2, based on the vehicle distribution set and the stop-candidate order set mapping relationship, taking the cooperative benefit cost ratio as the target, planning the driving path of the ground vehicle and iteratively generating the vehicle access sequence. The application dynamically constructs a cone, optimizes the path and task allocation, realizes the space-time coupling of the vehicle and the machine, efficiently responds to the abnormality and realizes the closed-loop incremental optimization, effectively improves the distribution efficiency and safety, reduces the operation cost, and ensures the efficient and orderly completion of the distribution task.

Description

Technical Field

[0001] This invention relates to the field of low-altitude logistics collaborative scheduling technology, and in particular to a method for collaborative delivery between ground logistics vehicles and drones. Background Technology

[0002] In urban last-mile logistics delivery scenarios, collaborative delivery between ground logistics vehicles and drones is a delivery model explored in recent years. In this model, vehicles can serve as mobile take-off and landing platforms for drones. After the vehicle travels to a preset stop, it releases the drone to complete the delivery of nearby orders. After completing its mission, the drone returns to the vehicle's next stop to rejoin it. This approach attempts to combine the carrying capacity of ground transportation with the path flexibility of air flight to improve delivery efficiency. However, in practical applications, existing collaborative delivery methods still have room for optimization when dealing with dynamic changes.

[0003] Taking a community delivery mission as an example, a delivery vehicle carrying two drones departs from a station, planning to release the drones sequentially at three stops along the way to deliver six orders in the surrounding area. During the mission, when the vehicle reaches the second stop, the system receives a new urgent delivery order approximately 1.2 kilometers away. Since task allocation is usually completed all at once before departure, it is difficult to incorporate the new order that appears during the journey into the drone's current task sequence in real time. This order is ultimately delivered by the vehicle via a detour, causing a delay of approximately 25 minutes. At the same time, one of the drones encounters crosswinds while performing its scheduled task, resulting in higher-than-expected power consumption. The remaining power may not be enough to return to the third stop as originally planned. Since the system lacks a real-time monitoring mechanism for drone power and a dynamic task adjustment mechanism, the drone chooses to land at the nearest stop, and the unfinished order needs to be delivered by another vehicle later.

[0004] The above situation reflects that existing collaborative methods mainly rely on offline static planning in the task allocation stage, making it difficult to respond in real time and dynamically adjust to factors such as order changes and drone status fluctuations that occur during the journey. This, to some extent, affects the overall operating efficiency of the vehicle-machine collaborative mode in complex environments. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide a method for collaborative delivery of ground logistics vehicles and drones, which can improve last-mile delivery efficiency and reduce reliance on manpower.

[0006] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows:

[0007] In a first aspect, a method for collaborative delivery using ground logistics vehicles and drones, the method comprising:

[0008] Step 1: Based on the orders to be assigned, dynamically construct an accessibility cone for each drone docking point; determine the spatial inclusion relationship between the destination coordinates of the orders to be assigned and the accessibility cone, identify the drone-accessible candidate orders, and establish a docking point-candidate order set mapping relationship;

[0009] Step 2: Based on the mapping relationship between the vehicle delivery set and the stop point-candidate order set, and with the goal of the collaborative benefit-cost ratio, plan the driving route for the ground vehicles and iteratively generate the vehicle access sequence;

[0010] Step 3: When a vehicle arrives at a stop according to the vehicle access sequence, the accessibility cone of the stop is reconstructed, the candidate orders covered by the accessibility cone are spatially clustered, a sub-region is assigned to each drone, and the drone flight path and return route are planned.

[0011] Step 4: Based on the planned flight path, calculate the time when the UAV completes its mission and arrives at the next rendezvous point and the time when the vehicle arrives at the same rendezvous point. If the two times are not synchronized, make dynamic adjustments to achieve vehicle-machine spatiotemporal coupling.

[0012] Step 5: When an anomaly is detected, reconstruct the micro safety cone for emergency situations, centered on the current location of the affected drone or the next planned critical point, and calculate the response strategy within the micro safety cone.

[0013] Step 6: Continuously execute steps 1 to 5 in a loop. When a new order is placed, the status is updated, or an anomaly is triggered, incremental optimization is performed on the affected links based on the results of the previous round to form a closed-loop decision until all delivery tasks are completed.

[0014] In a second aspect, a computing device includes:

[0015] One or more processors;

[0016] A storage device for storing one or more programs that, when executed by one or more processors, cause the one or more processors to implement the method.

[0017] Thirdly, a computer-readable storage medium storing a program that, when executed by a processor, implements the method.

[0018] The above-described solution of the present invention has at least the following beneficial effects:

[0019] Using dynamic accessibility cones as the core tool, this system replaces simple straight-line distance judgments. It combines real-time airspace, weather, and drone performance data to construct cones and spatially match orders with docking points. It also supports real-time inclusion of new orders and dynamic cone updates during delivery, solving the problems of static task allocation and inability to respond to changes in road conditions, equipment, and demand during the journey in collaborative delivery models. This makes drone delivery task allocation more aligned with real-world scenarios, improving the accuracy and efficiency of order allocation, effectively reducing vehicle detours, and shortening overall delivery time. Furthermore, it integrates drone accessibility constraints into the entire vehicle route planning process, iteratively generating vehicle access sequences with collaborative benefit-cost ratio as the core objective, allowing vehicle route planning to coordinate with the drone's potential service capabilities in advance. Simultaneously, based on the spatiotemporal data of drone flight path calculations, it monitors the arrival time of vehicles and drones at the rendezvous point. Dynamic verification and adjustment are performed to achieve precise coupling between vehicle driving and drone flight in time and space, effectively avoiding problems such as waiting and connection conflicts at parking points, eliminating resource idleness, and improving the overall efficiency of vehicle-drone collaborative operation. A cone-guided abnormal closed-loop response mechanism is constructed. In case of abnormal situations such as sudden drone failure, temporary airspace closure, or sudden weather changes, a miniature safety cone is quickly constructed centered on the real-time location of the affected drone. Within the safe airspace, emergency landing points, alternative flight paths, or temporary detour paths for vehicles are precisely planned. The vehicle emergency task is inserted into the access sequence with the highest priority for local replanning, replacing the simple method of landing nearby and re-delivering orders later. This effectively ensures the continuity of high-time-efficiency delivery services and improves the anti-interference ability and service completion rate of the collaborative delivery system in complex urban environments.

[0020] A real-time iterative closed-loop decision-making system was established, enabling the system to perform incremental optimizations based on the previous round of decisions. This reduces the computational load on the system, making collaborative decision-making more efficient and better suited to the dynamic changes in urban last-mile logistics, thus improving the practical implementation and operational efficiency of the method. Furthermore, by establishing drone docking points at existing urban road nodes, ground logistics vehicles serve as mobile take-off and landing platforms for drones, eliminating the need for additional dedicated infrastructure such as fixed drone landing sites. Simultaneously, the efficient collaboration between vehicles and drones enhances the efficiency of unmanned aerial vehicle (UAV) operations. The large-scale operation capability of drones effectively alleviates the objective constraints of short flight time and limited payload of drones. While reducing the infrastructure deployment costs of logistics companies, it also reduces the additional operating costs caused by equipment resource waste and order replenishment. Through the whole-process optimization of dynamic task allocation, precise spatiotemporal coordination, and rapid anomaly response, the overall delivery time of last-mile delivery is shortened. This effectively solves the delivery efficiency problem in high-time-sensitivity logistics scenarios such as instant retail, urgent pharmaceutical delivery, and fresh food delivery. At the same time, it reduces reliance on manpower, enabling the collaborative delivery mode of ground logistics vehicles and drones to achieve stable, high-frequency, large-scale operation in urban and suburban high-density areas, and comprehensively improves the service capabilities and level of urban last-mile logistics. Attached Figure Description

[0021] Figure 1 This is a flowchart illustrating a method for collaborative delivery between ground logistics vehicles and drones, provided by an embodiment of the present invention. Detailed Implementation

[0022] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0023] like Figure 1 As shown, an embodiment of the present invention proposes a method for collaborative delivery between ground logistics vehicles and drones, the method comprising the following steps:

[0024] Step 1: Based on the orders to be assigned, dynamically construct an accessibility cone for each drone docking point; determine the spatial inclusion relationship between the destination coordinates of the orders to be assigned and the accessibility cone, identify the drone-accessible candidate orders, and establish a docking point-candidate order set mapping relationship;

[0025] Step 2: Based on the mapping relationship between the vehicle delivery set and the stop point-candidate order set, and with the goal of the collaborative benefit-cost ratio, plan the driving route for the ground vehicles and iteratively generate the vehicle access sequence;

[0026] Step 3: When a vehicle arrives at a stop according to the vehicle access sequence, the accessibility cone of the stop is reconstructed, the candidate orders covered by the accessibility cone are spatially clustered, a sub-region is assigned to each drone, and the drone flight path and return route are planned.

[0027] Step 4: Based on the planned flight path, calculate the time when the UAV completes its mission and arrives at the next rendezvous point and the time when the vehicle arrives at the same rendezvous point. If the two times are not synchronized, make dynamic adjustments to achieve vehicle-machine spatiotemporal coupling.

[0028] Step 5: When an anomaly is detected, reconstruct the micro safety cone for emergency situations, centered on the current location of the affected drone or the next planned critical point, and calculate the response strategy within the micro safety cone.

[0029] Step 6: Continuously execute steps 1 to 5 in a loop. When a new order is placed, the status is updated, or an anomaly is triggered, incremental optimization is performed on the affected links based on the results of the previous round to form a closed-loop decision until all delivery tasks are completed.

[0030] In this embodiment of the invention, a dynamic accessibility cone is used as the core tool to replace the simple straight-line distance judgment method. It combines real-time airspace, weather, and drone performance data to construct the cone and complete the spatial matching of orders and docking points. Simultaneously, it supports the real-time inclusion of new orders and dynamic updates of the cone during delivery, solving the problems of static task allocation and inability to respond to changes in road conditions, equipment, and demand during the journey in collaborative delivery models. This makes drone delivery task allocation more aligned with actual scenarios, improving the accuracy and efficiency of order allocation, effectively reducing vehicle detours, and shortening the overall delivery time. Furthermore, drone accessibility constraints are integrated into the entire vehicle route planning process, iteratively generating vehicle access sequences with the collaborative benefit-cost ratio as the core objective, allowing vehicle route planning and the potential service capabilities of drones to coordinate in advance. Simultaneously, based on the spatiotemporal data of drone flight path calculation, it tracks vehicle-drone arrival convergence. The timing of the docking points is dynamically verified and adjusted to achieve precise coupling between vehicle driving and drone flight in time and space. This effectively avoids problems such as waiting and connection conflicts at docking points, eliminates resource idleness, and improves the overall efficiency of vehicle-drone collaborative operations. A cone-guided abnormal closed-loop response mechanism has been constructed. In the event of abnormal situations such as sudden drone failure, temporary airspace closure, or sudden weather changes, a miniature safety cone is quickly constructed centered on the real-time location of the affected drone. Within the safe airspace, emergency landing points, alternative flight paths, or temporary detour paths for vehicles are precisely planned. The vehicle emergency task is inserted into the access sequence with the highest priority for local replanning, replacing the simple method of landing nearby and re-delivering orders later. This effectively ensures the continuity of high-efficiency delivery services and improves the anti-interference capability and service completion rate of the collaborative delivery system in complex urban environments.

[0031] A real-time iterative closed-loop decision-making system was established, enabling the system to perform incremental optimizations based on the previous round of decisions. This reduces the computational load on the system, making collaborative decision-making more efficient and better suited to the dynamic changes in urban last-mile logistics, thus improving the practical implementation and operational efficiency of the method. Furthermore, by establishing drone docking points at existing urban road nodes, ground logistics vehicles serve as mobile take-off and landing platforms for drones, eliminating the need for additional dedicated infrastructure such as fixed drone landing sites. Simultaneously, the efficient collaboration between vehicles and drones enhances the efficiency of unmanned aerial vehicle (UAV) operations. The large-scale operation capability of drones effectively alleviates the objective constraints of short flight time and limited payload of drones. While reducing the infrastructure deployment costs of logistics companies, it also reduces the additional operating costs caused by equipment resource waste and order replenishment. Through the whole-process optimization of dynamic task allocation, precise spatiotemporal coordination, and rapid anomaly response, the overall delivery time of last-mile delivery is shortened. This effectively solves the delivery efficiency problem in high-time-sensitivity logistics scenarios such as instant retail, urgent pharmaceutical delivery, and fresh food delivery. At the same time, it reduces reliance on manpower, enabling the collaborative delivery mode of ground logistics vehicles and drones to achieve stable, high-frequency, large-scale operation in urban and suburban high-density areas, and comprehensively improves the service capabilities and level of urban last-mile logistics.

[0032] In a preferred embodiment of the present invention, before step 1, in step 001, real-time data streams are accessed, including a set of orders to be delivered, real-time vehicle location and status, number of available drones and their respective real-time battery power and maximum range, real-time urban traffic conditions, a preset drone docking point network, and dynamic no-fly zones and weather information; based on order attributes, an initial screening is performed, and orders requiring personal signature are assigned to the vehicle delivery set, while the remaining orders are designated as orders to be allocated. Specifically, the system accesses all dimensions of delivery-related data in real time through preset IoT interfaces, urban traffic data platforms, airspace management systems, and logistics scheduling platforms. The data categories and access / processing methods are as follows: all order data currently awaiting delivery is obtained from the logistics order management system, including the unique identifier of each order, destination geographical coordinates (latitude and longitude), order category (such as fresh produce, medicine, and ordinary parcels), order urgency (such as priority marking for urgent orders), signature requirements (whether personal signature is required), order placement time, delivery time limit, and other core information. The system numbers and archives all orders to form a set of orders to be delivered; ground logistics data is obtained through vehicle positioning terminals (GPS / BeiDou). The system obtains real-time geographic coordinates of the vehicle through vehicle status sensors, acquiring information such as remaining driving range, current speed, payload capacity, and whether there is a malfunction. Simultaneously, it integrates with city electronic maps to match the vehicle's real-time location to the road network, providing a foundation for subsequent route planning. The system obtains the number of currently available drones and the real-time status data of each drone from the drone scheduling and management module, including real-time battery power (read from the drone battery management system as the remaining percentage, e.g., 80%), calculated by combining the drone's rated battery capacity (e.g., 5000mAh) with the formula: Remaining Battery Power (mAh) = Rated Battery Capacity × Remaining Battery Percentage); Maximum Range (calculated by combining the drone's current remaining battery power, power consumption per unit distance (e.g., 100mAh / km), and weather correction factors (e.g., 1.2 for headwinds and 0.8 for tailwinds) to calculate the maximum achievable range under current conditions, calculated as: Maximum Range (km) = Remaining Battery Power (mAh) ÷ Power Consumption per Unit Distance (mAh / km) ÷ Weather Correction Factor). The system also records basic performance parameters of each drone, such as payload capacity and flight speed.The system obtains real-time traffic speeds, congestion indices, and construction information for various road sections within the delivery area from the city's traffic management platform. It then categorizes and labels the road condition data by road segment; for example, congested sections are labeled with a speed limit of ≤20 km / h, while uncongested sections are labeled with a speed limit of ≥50 km / h, providing a basis for predicting vehicle route travel time. It also retrieves pre-set drone docking point data from the logistics dispatch platform, including the geographical coordinates of all docking points, available docking area, availability of drone take-off and landing conditions, and the presence of obstructions. All docking points are organized into a network list by region number. Finally, it obtains real-time updated data from airspace management departments. No-fly zone data includes the geographic boundary coordinates of the no-fly zone, the type of no-fly zone (such as high-voltage line areas, airport airspace, and residential rooftops), and the effective period of the no-fly zone. The system converts the no-fly zone data into a spatial polygon coordinate set. Real-time meteorological data such as wind speed, wind direction, precipitation probability, and visibility within the delivery area are obtained from the meteorological service platform and updated every 10 minutes to provide environmental constraints for UAV range calculation and flight path planning. All accessed data is converted to a unified data format (such as unifying various coordinates to the WGS84 coordinate system) and stored in the real-time database to ensure data dimension uniformity and cross-referencing capability.

[0033] The system performs order-by-order attribute identification and classification on the collected sets of orders awaiting delivery. The core filtering rules and execution process are as follows: For each order awaiting delivery, the system extracts its "signature requirements" attribute field and identifies whether this field contains key information such as "personal signature required," "personal confirmation required," or "in-person delivery required." Simultaneously, it assists in identifying order category attributes, such as large or heavy items, or fragile items requiring manual handover, as supplementary filtering criteria. If the order's signature requirements attribute explicitly states that personal signature is required, or if the order category belongs to the type requiring in-person manual handover, the system assigns the order to the vehicle delivery set. Orders within the designated delivery area will be delivered solely by ground logistics vehicles and will not be included in the drone delivery range. If the order requires delivery by proxy and does not require the recipient's signature, and the order category is a common category that drones can deliver, such as small parcels or non-fragile items, the system will classify the order as a pending order. Orders within this set will be further matched for delivery based on the drone's accessibility cone. The system will establish independent data lists for vehicle delivery sets and pending orders, recording core information such as the order number, destination coordinates, and delivery time limit for each order, and will synchronize this information to the subsequent route planning module in real time.

[0034] This embodiment provides accurate and dynamic foundational data support for subsequent accessibility cone construction, vehicle route planning, and drone task allocation through multi-dimensional and real-time multi-source data access, solving the planning deviation problems caused by data lag and single-dimensionality. Based on the initial screening of core attributes such as signature requirements, the delivery boundaries of vehicles and drones are clearly defined in advance, avoiding delivery failures caused by drone delivery requiring personal signature, improving the first delivery completion rate of orders, and reducing secondary delivery costs. The delivery task scope of vehicles and drones is distinguished at the initial stage of the delivery process, avoiding invalid drone accessibility calculations in subsequent stages, significantly reducing system computing power consumption, and improving overall scheduling efficiency. Real-time road conditions, weather, no-fly zones, and other dynamic constraints are accessed and integrated in advance, reducing the failure of planning schemes due to environmental changes and improving the stability of collaborative delivery.

[0035] In a preferred embodiment of the present invention, step 1 includes:

[0036] Step 100: Based on the dynamic no-fly zone and meteorological information in the accessed real-time data stream, and the remaining flight range of the UAV, for each candidate UAV landing point, construct an accessibility cone centered on the corresponding landing point. The accessibility cone is a continuous and safe airspace volume formed by taking the landing point as the vertex, the maximum reachable radius of the UAV as the generatrix, and removing the portion intersecting with the dynamic no-fly zone. Specifically, the construction of the accessibility cone is based on a single UAV landing point, combined with real-time environmental constraints and UAV performance parameters, to complete the spatial modeling of the safe airspace. The specific process is as follows: retrieve the core data of the landing point to be calculated from the real-time database, including the geographical coordinates (latitude, longitude, altitude) of the landing point, the spatial coordinate set of the dynamic no-fly zone accessed in step 001, real-time meteorological information (wind speed, wind direction, visibility), and the remaining flight range of available UAVs. Flight range (calculated in step 001: remaining battery power ÷ power consumption per unit distance ÷ weather correction factor); First, environmental calibration is performed on the maximum reachable radius of the drone. If the real-time weather is headwind (wind speed ≥ 5m / s), the headwind will increase the drone's flight resistance and increase power consumption per unit distance. Therefore, the original maximum reachable radius is multiplied by a correction factor of 0.9 to adapt to the actual endurance. If the weather is tailwind (wind speed ≥ 5m / s), the tailwind can reduce flight resistance and reduce power consumption per unit distance. Therefore, it is multiplied by a correction factor of 1.05 to make full use of the environmental advantages. If there is no obvious wind direction (wind speed < 5m / s), the impact of the environment on the drone's endurance can be ignored, and the correction factor is 1.0. Finally, the formula for calculating the maximum reachable radius of the drone after calibration is: Maximum reachable radius after calibration (km) = maximum flight range calculated in step 001 × weather correction factor.

[0037] Using the current candidate drone landing point as the spatial vertex and the calibrated maximum reachable radius of the drone as the generatrix, construct an initial conical airspace volume along the direction perpendicular to the ground (considering drone flight altitude limitations, generally with an upper limit of 120 meters). The spatial range of this initial cone covers all theoretically reachable airspace around the landing point. Its base is a circular area with the landing point's projection on the ground as the center and the calibrated maximum reachable radius as the radius, and its height is the drone's maximum flight altitude (e.g., 120 meters). Then, calculate the spatial intersection between the real-time accessed dynamic no-fly zone spatial polygon coordinate set (e.g., high-voltage line areas, airport airspace, residential rooftop no-fly zones) and the initial cone. First, the two-dimensional geographic coordinates of the no-fly zone are converted into three-dimensional spatial coordinates including altitude (e.g., the altitude of the high-voltage line area is 10 to 20 meters, and the altitude of the ground no-fly zone is 0 to 120 meters). Then, the spatial overlap between each no-fly zone and the initial cone is judged one by one. If the spatial range of a certain no-fly zone intersects with the initial cone, the intersecting area is removed from the initial cone. After removing all the intersecting parts of the no-fly zones, the remaining continuous airspace volume without no-fly zone overlap is the final dynamic accessibility cone of the docking point. The system stores the accessibility cone of each docking point in the form of a three-dimensional spatial coordinate set, specifying the key parameters such as the boundary range, effective height, and ground projection area of ​​the cone.

[0038] Step 101: Determine the spatial inclusion relationship between the destination coordinates of the orders to be assigned and the accessibility cone of each drone docking point. If the destination of the order to be assigned is located within the accessibility cone of any drone docking point, then mark the corresponding order to be assigned as a drone reachable candidate order. Specifically, this includes: extracting the destination geographic coordinates (latitude and longitude) of each order from the list of orders to be assigned generated in Step 001, and supplementing the altitude of the location (the default is ground altitude, such as 50 meters in a typical urban area), and converting it into a three-dimensional spatial coordinate system (WGS84 coordinate system) consistent with the accessibility cone; for each order to be assigned, determine its spatial inclusion relationship with the accessibility cones of all candidate drone docking points in turn. First, calculate the straight-line distance between the projection point of the order destination on the ground and the projection point of the docking point on the ground; second, determine the straight-line distance between the projection point of the order destination on the ground and the projection point of the docking point on the ground. The system first checks if the distance is less than or equal to the calibrated maximum reachable radius of the accessibility cone at the docking point, and if the altitude of the order's destination is within the effective flight altitude range of the cone (e.g., 0 to 120 meters). Then, it further determines if the three-dimensional spatial coordinates of the order's destination fall within the effective airspace of the cone after removing no-fly zones (no no-fly zone overlap). If all three conditions are met, the destination of the order to be assigned is determined to be within the accessibility cone of the docking point, and the order is marked as a drone-accessible candidate order. If none of the accessibility cones at any docking point contain the order, it is reassigned to the vehicle delivery set in step 001. The system adds a unique tag to each order marked as a drone-accessible candidate order, records its docking point number, cone matching criteria (e.g., ground distance, altitude, no no-fly zone overlap), and compiles these into a drone-accessible candidate order list.

[0039] Step 102: Establish the correspondence between each drone docking point and the drone reachable candidate orders it can serve, denoted as the docking point-candidate order set mapping relationship. Specifically, this includes: the system iterates through all drone reachable candidate orders, classifying and statistically analyzing them according to the docking point number matched with the order. If an order matches only one docking point's reachability cone, it is directly assigned to that docking point's candidate order set. If an order matches multiple docking point reachability cones simultaneously, such as when the order is located in the overlapping area of ​​two docking point cones, the straight-line distance from the order to each matching docking point is calculated, and the order is assigned to the nearest docking point's candidate order set (distance calculation formula is the same as in step 101). Using the docking point number as the primary key, construct a key-value pair docking point-candidate order set mapping relationship, where the primary key is the unique docking point number and the value is the docking point pair. The system generates a list of all available candidate order numbers for all drones and records the core information of each order (destination coordinates, delivery time limit, order category). For example, the candidate order set corresponding to docking point 001 is [order 005, order 008, order 012], and the candidate order set corresponding to docking point 002 is [order 003, order 010]. The system synchronizes this mapping relationship to the vehicle route planning module in real time to ensure that subsequent steps can directly call it. The system is equipped with a real-time monitoring mechanism. If the real-time data stream in step 001 changes, such as the addition of new orders to be assigned, dynamic adjustment of no-fly zones, or changes in the remaining flight range of drones, the cone reconstruction in step 100 is automatically triggered, the spatial judgment in step 101 is re-executed, and the docking point-candidate order set mapping relationship is updated to ensure that the mapping relationship is always consistent with the real-time environment and order status.

[0040] This embodiment constructs a dynamic accessibility cone, combining the theoretical range of drones with real-time no-fly zones and weather conditions to replace simple straight-line distance judgment. This defines the effective delivery airspace for drones at each docking point, preventing drones from entering no-fly zones or experiencing insufficient endurance due to environmental factors, thus improving drone flight safety. By judging the inclusion relationship in three-dimensional space, it filters out truly drone-deliverable candidate orders, avoiding including orders outside the drone's reach or located in no-fly zones in the drone delivery plan, reducing ineffective planning and improving order matching accuracy. Simultaneously, it establishes a structured mapping relationship to improve overall scheduling efficiency. Both the accessibility cone and the mapping relationship support real-time dynamic updates, enabling rapid response to dynamic factors such as new orders, no-fly zone adjustments, and weather changes, solving the problem of static planning being unable to adapt to environmental changes, ensuring that the vehicle-to-drone collaborative delivery solution always fits the actual operating scenario. By establishing a mapping relationship between docking points and candidate orders, the drone service capabilities of each docking point are clearly defined in advance, improving the overall efficiency of vehicle-to-drone collaboration and reducing unnecessary vehicle detours.

[0041] In a preferred embodiment of the present invention, step 2 includes:

[0042] Step 200: Merge the vehicle's starting position, all demand locations in the vehicle delivery set, and all drone docking points to form a set of points to be visited; initialize the vehicle access sequence, where the initial element of the vehicle access sequence is the vehicle's starting position. Specifically, the system first extracts three types of core location data from the real-time database and merges them. The first type is the vehicle's starting position, which is the initial departure position of the ground logistics vehicle, such as a logistics station, and this position is recorded in full in the form of latitude and longitude coordinates. The second type is the demand locations in the vehicle delivery set, which are the destination coordinates of all orders assigned to the vehicle delivery set in step 001. The delivery location of each order is considered an independent demand location. The third type is all drone docking points, which are all candidate drones participating in the accessibility cone construction in step 100. The system deduplicates the three types of location data mentioned above. If a vehicle delivery order location overlaps with a drone docking point location, only one point is retained. Then, all points are uniformly converted into geographic coordinates in the WGS84 coordinate system to form a structured set of points to be accessed. At the same time, each point is assigned a unique identifier, such as starting point-001, vehicle order point-005, and docking point-002. The system also records the type attributes of each point, which are divided into three categories: starting location, vehicle delivery point, and drone docking point. The system creates a data structure called vehicle access sequence, which is an ordered list of points. In the initial state, only the vehicle starting location is recorded as a unique element. For example, the initial value of the vehicle access sequence is [logistics station (longitude: 116.40, latitude: 39.90)].

[0043] Step 201: Repeat the following operations until all points in the set of points to be visited are added to the vehicle access sequence. Specifically, for each candidate point in the set not yet added to the vehicle access sequence, calculate the increase in the vehicle travel path when inserting the corresponding candidate point between any two adjacent points in the current vehicle access sequence. Determine the collaborative benefit-cost ratio of the corresponding candidate point based on the docking point-candidate order set mapping relationship. The numerator of the collaborative benefit-cost ratio is determined based on the number of drone-accessible candidate orders covered within the accessibility cone of the corresponding candidate point and the urgency of those orders. The denominator is the additional time cost for the vehicle to detour to the corresponding candidate point, specifically including... The system first checks whether all points in the set of points to be visited have been added to the vehicle access sequence. If there are still points in the set that have not been added to the sequence, the calculation process for candidate points in this round is started. If all points in the set of points to be visited have been added to the sequence, the iteration process is terminated, and the system directly proceeds to the final sequence confirmation stage in step 202. For each candidate point in the set of points to be visited that has not yet been added to the sequence, the system will simulate all possible scenarios of inserting the candidate point between any two adjacent points in the current access sequence, and calculate the increase in the vehicle travel path corresponding to each insertion method. Specifically, the calculation is first based on the real-time urban traffic data accessed in step 001. The original travel path length LAB between two adjacent points A and B in the current sequence is calculated in kilometers. This length is adjusted based on real-time traffic conditions; for example, the effective length is adjusted according to the actual traffic speed in congested areas. Next, the travel path length LAC from point A to candidate point C and the travel path length LCB from candidate point C to point B are calculated. Finally, the path increase under this insertion method is calculated using the formula: Path increase (km) = LAC + LCB - LAB. Simultaneously, the system incorporates the average vehicle speed under real-time traffic conditions; for example, the average vehicle speed is 30 km / h in unobstructed areas and 10 km / h in congested areas, converting the path increase into an additional... The external time cost is calculated as follows: External time cost (minutes) = Path increase ÷ Average vehicle speed × 60 (1 hour equals 60 minutes). The synergy benefit cost ratio is the core indicator for measuring the insertion value of candidate points. Its numerator is the synergy benefit value of the candidate point, and the denominator is the external time cost. The specific calculation process is as follows: The numerator is the calculation of the synergy benefit value. If the candidate point is a drone docking point, the docking point-candidate order set mapping relationship established in step 102 is retrieved to obtain the number N of drone reachable candidate orders within the reachability cone of the docking point. At the same time, each order is assigned a weight W for urgency. For example, the weight W for urgent orders is 2, and the weight W for ordinary orders is 1.The collaborative benefit value equals the sum of the urgency weights of all candidate orders. The formula is: Collaborative Benefit Value = Number of Urgent Orders × 2 + Number of Regular Orders × 1. If a candidate point is a vehicle delivery point, and such points do not have drone service capabilities, then the collaborative benefit value is the urgency weight of the order itself. For example, the weight W for urgent orders requiring personal signature is 2, and the weight W for regular orders is 1. The denominator is the additional time cost, in minutes. The additional time cost value calculated using the above steps for this insertion method is used to finally calculate the collaborative benefit cost ratio. The formula is: Collaborative Benefit Cost Ratio = Collaborative Benefit Value ÷ Additional Time Cost. The system will calculate the corresponding collaborative benefit cost ratio for each insertion method for each candidate point and record the ratio and the corresponding insertion location information.

[0044] Step 202: From all candidate points and their insertion positions, select the candidate point and its insertion position that maximizes the collaborative benefit-cost ratio. Insert the corresponding candidate point into the insertion position of the vehicle access sequence. After all points have been inserted, the final vehicle access sequence is obtained. The stopping points in the vehicle access sequence are collaborative operation nodes with the capability to provide drone-accessible candidate order services. Specifically, the system summarizes the collaborative benefit-cost ratios corresponding to all candidate points and all insertion positions, and selects the set of solutions with the largest ratio. Specifically, inserting a candidate point C between two adjacent points A and B in the current sequence has the highest collaborative benefit-cost ratio among all available solutions. The system then inserts the candidate point C corresponding to this optimal solution into the vehicle access sequence. The system updates the vehicle access sequence between points A and B in the vehicle access sequence; at the same time, it removes candidate point C from the set of points to be accessed, thus completing this round of iteration. The system repeats the iteration process from step 201 to step 202 until all points in the set of points to be accessed are inserted into the vehicle access sequence. The complete sequence generated at this time is the final vehicle access sequence. All points marked as drone docking points in the sequence are collaborative operation nodes with the ability to provide drone-accessible candidate order services. For example, the final sequence is presented as: [Logistics Station, Dock Point-001, Vehicle Order Point-005, Dock Point-002, Vehicle Order Point-010, Dock Point-003, Vehicle Order Point-015].

[0045] This embodiment incorporates the candidate order service capabilities of drone docking points into the core indicators of vehicle route planning. It quantifies the benefits of drone services and the costs of vehicle detours through a collaborative benefit-cost ratio, avoiding meaningless detours and ensuring that route planning meets vehicle delivery needs while maximizing the value of drone collaborative delivery. The calculation of collaborative benefit values ​​incorporates the urgency of orders, prioritizing the delivery efficiency of urgent orders and addressing the issue of prioritizing distance over timeliness in route planning, making it suitable for high-time-sensitivity scenarios such as instant retail and urgent pharmaceutical delivery. Through an iterative approach of selecting the optimal insertion scheme in each round, compared to a one-time global planning, it better reflects the dynamic changes in real-time urban traffic conditions. Each iteration is calculated based on the latest data, resulting in a more practically valuable vehicle access sequence. The final sequence retains only docking points with drone service capabilities as collaborative operation nodes, preventing vehicles from going to docking points without available orders, reducing unnecessary travel, lowering vehicle fuel consumption and time costs, and improving overall delivery efficiency.

[0046] In a preferred embodiment of the present invention, step 3 includes:

[0047] Step 300: When a vehicle travels according to the vehicle access sequence and arrives at a docking point, based on the current dynamic no-fly zone and weather information, as well as the remaining battery power of available drones, the system reconstructs the current accessibility cone for the corresponding docking point and obtains the candidate orders for drones to be delivered currently covered by the accessibility cone, as the candidate order set for the current batch. Specifically, after the ground logistics vehicle travels to a drone docking point according to the vehicle access sequence generated in step 202 and completes docking, the system immediately synchronizes the full real-time data at that moment, and updates the dynamic no-fly zone and weather information. The system retrieves the following data: the spatial coordinates of the flight area (e.g., temporarily controlled airspace, newly added no-fly zones for construction), real-time weather information (including wind speed, wind direction, visibility, etc., consistent with the data source for the calibration parameters in step 100), the remaining battery power of all available drones (read in real-time via the drone battery management system, in mAh), and their current payload status. Simultaneously, it retrieves the geographic coordinates (latitude, longitude, and altitude) of the docking point. Referring to the cone construction logic in step 100, it reconstructs the accessibility cone of the docking point at the current moment based on the aforementioned real-time data. First, it calculates the maximum range of the drone in the current state. Maximum range (km) = Remaining UAV battery power (mAh) ÷ Power consumption per unit distance (mAh / km) ÷ Weather correction factor (0.9 for headwind ≥ 5m / s, 1.05 for tailwind ≥ 5m / s, 1.0 for calm < 5m / s); Using the docking point as the apex and the calculated maximum range as the generatrix, and combining it with current flight altitude restrictions (such as the 120-meter upper limit commonly used for urban low-altitude logistics), an initial conical airspace is constructed; areas intersecting with the current dynamic no-fly zone are eliminated, ultimately forming a continuous and safe accessibility cone for the docking point at the current moment, ensuring the cone's range is completely... Adapt to the real-time environment and the remaining battery life of the drone; the system retrieves all drone reachable candidate orders corresponding to the docking point in the docking point-candidate order set mapping relationship established in step 102, and verifies the real-time status of each order one by one, that is, removing orders that have been delivered, canceled, or exceeded the delivery time limit, and retaining orders that are still in the pending delivery state and whose destination coordinates fall within the reachability cone after this reconstruction. These orders are summarized to form the candidate order set of the current batch, and the core information of each order, such as the destination three-dimensional coordinates (latitude, longitude, altitude), delivery time limit, and order category, is recorded.

[0048] Step 301: Cluster the candidate order set of the current batch according to spatial location. Combining the remaining battery power and maximum range of each available drone, dynamically allocate a sub-region within the reachability cone for each available drone. This sub-region ensures that multiple order points assigned to the same drone are located within the corresponding sub-region, and that the total flight path of the corresponding drone completing all assigned orders and returning to the next rendezvous point is within its range. Specifically, this includes: the system uses a spatial density clustering algorithm to classify the candidate order set of the current batch, and calculates the straight-line distance between any two order locations based on the ground projection coordinates of the order destination (calculation formula: straight-line distance (km) = ...). Orders within a distance of ≤1 km are grouped into the same cluster, ultimately dividing the current batch of candidate orders into several spatial clusters. Orders within each cluster are concentrated in location, facilitating centralized drone delivery. For each available drone, the maximum feasible flight path length for completing the delivery task and returning is calculated. First, the drone's remaining range is calculated: Remaining range (km) = Remaining battery power (mAh) ÷ Power consumption per unit distance (mAh / km) ÷ Weather correction factor. Considering the drone's need to return to the vehicle's next predetermined docking point (next hub)... (For each rendezvous point), 10% of the remaining battery power needs to be reserved as a safety redundancy. Therefore, the actual usable range (km) = remaining range × 0.9 (0.9 is the percentage coefficient of the remaining range after deducting the 10% safety redundancy). At the same time, the estimated flight path length of the drone from the current docking point to any cluster group, to complete the delivery of all orders in that group, and then back to the next rendezvous point is calculated. The estimated flight path length (km) = distance from the current docking point to the center of the cluster group + total path length between order points within the cluster group + distance from the last order point in the cluster group to the next rendezvous point.

[0049] Based on clustering results and drone endurance, the system assigns sub-regions within the reachability cone to each available drone. The allocation is based on spatial clusters, prioritizing the designation of complete clusters as dedicated sub-regions for a single drone, ensuring all order points assigned to the same drone are within that sub-region. The system then verifies the total flight path length for the drone to complete all order deliveries within the sub-region and return to the next rendezvous point. If the total flight path length is less than or equal to the actual available endurance, the sub-region allocation is confirmed. If the total flight path length exceeds the actual available endurance, the cluster is split into 2 to 3 smaller independent sub-clusters based on the spatial distance between orders. These sub-clusters are then reassigned to drones with more remaining battery power (assigning them new sub-regions), or some orders within the cluster are adjusted to the sub-regions already assigned to the remaining drones. After all sub-regions are allocated, the system defines the delivery boundaries (the three-dimensional spatial coordinate range of the sub-region) for each drone, ensuring no order omissions, no area overlap, and that each drone's workload remains within its endurance range.

[0050] Step 302 involves planning a flight path for each drone within its assigned sub-region to each order point, and planning its return route to the next predetermined docking point of the vehicle. Specifically, for each drone's assigned sub-region, the system plans the flight path to each order point using the shortest path priority principle. Using the order points within the sub-region as nodes, the system calculates the flight distance between any two order points (considering no-fly zone avoidance, non-linear distance), constructing a path distance matrix between order points. Based on the distance matrix, the system plans the optimal flight sequence, starting from the docking point and sequentially visiting all order points within the sub-region according to the principle of shortest total flight distance. The calculation formula is: Total flight distance (km) = Distance from docking point to the first order point + (Distance between adjacent order points) ensures that the drone's flight path within the sub-region is unique and does not involve detours, maximizing delivery efficiency; after completing the flight path planning for order points within the sub-region, the system connects to the vehicle access sequence, retrieves the geographical coordinates of the vehicle's next scheduled stop (next rendezvous point), and plans the drone's return route from the last order point in the sub-region back to the rendezvous point. First, the straight-line distance from the last order point to the next rendezvous point is calculated, and then, combined with real-time no-fly zones and weather information, a safe route is planned to avoid obstacles, ensuring the entire route remains within legal airspace; the length of the return route is verified, and the return route length (km) + sub-region If the total flight distance within the domain is less than or equal to the actual usable range of the drone, and exceeds this range, the order allocation or flight order within the sub-region will be readjusted until the range constraint is met. At the same time, the estimated time for the vehicle to travel from the current stop point to the next rendezvous point (the travel time calculated in step 201) will be combined to plan the flight speed of the return route, ensuring that the drone's flight rhythm matches the vehicle's travel rhythm. The system will integrate the flight path and return route of each drone within its sub-region into a complete flight plan, record parameters such as the distance, estimated flight time, and altitude limit of each segment, and synchronize them to the drone flight control system to provide accurate guidance for the drone to perform delivery tasks.

[0051] In this embodiment, after the vehicle arrives at the docking point, the accessibility cone is reconstructed to adapt to real-time changes in no-fly zones, weather, and the drone's remaining battery power. This avoids order allocation failures caused by using static planning before departure, ensuring that all orders in the current batch are within the drone's actual reach. Spatial clustering assigns orders with concentrated locations to the same drone, reducing drone detours. The actual available range is calculated based on the remaining battery power, and the total flight path is verified, preventing drones from failing to return due to insufficient power and improving drone operation safety. Each drone is assigned a dedicated sub-region to avoid overlapping delivery paths and task conflicts. Flight routes are planned according to the shortest path to reduce invalid flight time and improve the order delivery efficiency of a single drone. When planning the return route, the next rendezvous point of the vehicle is anchored, and the matching of route length and range is verified to prevent situations where the drone cannot rendezvous with the vehicle after completing delivery. Sub-region allocation and route planning both support real-time adjustments. If the drone's battery power fluctuates suddenly or the order status changes, cluster groups can be quickly split / merged, and routes can be replanned, solving the problem that static task allocation cannot cope with unexpected situations during the journey.

[0052] In a preferred embodiment of the present invention, step 4 includes:

[0053] Step 400: Based on the flight path planned for each drone, calculate the flight time for each drone to complete all assigned orders in the current batch, and calculate the return time for the corresponding drone from the last order point to the next predetermined stop point (i.e., the next rendezvous point) in the vehicle access sequence. Add the two to obtain the total time for the drone to reach the next rendezvous point. Specifically, this includes: the system retrieves the precise flight path planned for each drone within a sub-region, extracts the total flight distance of the path, and combines it with the drone's actual flight speed in the urban low-altitude environment (the average flight speed after comprehensively considering no-fly zone avoidance and weather conditions, not the maximum flight speed) to calculate the flight time to complete all assigned orders in the current batch. The calculation formula is: Delivery order flight time (minutes) = Total distance of precise flight path within the sub-region (kilometers) ÷ Actual average flight speed of the drone (kilometers). The time is calculated as follows: (miles / hour) × 60, where multiplying by 60 converts the hourly calculation result to minutes to ensure consistent time measurement. The system retrieves the total return flight distance of the drone from the last order point to the next rendezvous point. Using the same actual average flight speed as the delivery flight, the system calculates the individual return time using the formula: Return time (minutes) = Total return flight distance (km) ÷ Actual average flight speed of the drone (km / hour) × 60. The delivery order flight time and the return time are summed to obtain the total time for each drone to complete the current batch of delivery tasks and reach the next rendezvous point. The formula is: Total drone time (minutes) = Delivery order flight time + Return time. The system calculates and records the total time for each available drone individually to ensure that the data corresponds one-to-one without confusion.

[0054] Step 401: Based on real-time traffic data, predict the vehicle travel time from the current stop to the next rendezvous point. Compare the total time of each drone with the vehicle travel time. If the total time of any drone is greater than the vehicle travel time, reduce the number of tasks assigned to that drone or transfer some orders to the remaining available drones, and replan the flight path of the corresponding drone until the total drone time is less than or equal to the vehicle travel time. If the total time of any drone is less than the vehicle travel time and the difference exceeds a preset threshold, adjust the drone's task execution order or suggest adjusting the next rendezvous point. The system determines the location of the rendezvous point and ensures that the time difference between the two is within a preset range, achieving spatiotemporal coupling between the vehicle and the drone. Specifically, this involves: Based on the real-time urban traffic data accessed in step 001, the system extracts road traffic information for ground logistics vehicles traveling from their current stop to the next rendezvous point. This information includes the total length of the travel route, the real-time congestion index for each road segment, and the traffic speed. Combined with the vehicle's actual average travel speed under different road conditions (e.g., 30 km / h on smooth roads and 10 km / h on congested roads), the system calculates the travel time in segments and then sums them to obtain the total travel time to the next rendezvous point. The calculation formula is: Vehicle travel time (minutes) = (Length of each road segment ÷ Actual average speed of vehicles on the corresponding road segment) × 60; The system compares the total time for each drone to reach the next rendezvous point with the total time for the vehicle to reach the rendezvous point. Based on the comparison results, it judges two cases separately. The preset time difference threshold is set at 5 minutes. This threshold is a reasonable time difference for vehicle-drone connection in urban last-mile logistics delivery, which avoids drones waiting for a long time and also prevents vehicles from excessively accelerating to match the drones. In the first case, if the total time of a certain drone is greater than the vehicle's travel time, it is determined that the drone cannot arrive at the next rendezvous point synchronously with the vehicle. There is a possibility that the drone will have to wait for the vehicle after returning to base, or even cause the vehicle to be delayed due to waiting. For the issue of delayed missions, a mission load adjustment strategy needs to be implemented. In scenario two, if the total time of a drone is less than the vehicle's travel time, and the time difference exceeds a preset threshold of 5 minutes, it is determined that the drone will arrive at the rendezvous point significantly earlier than the vehicle, indicating idle drone resources and wasted flight time. In this case, a mission execution or path adjustment strategy needs to be implemented. For the adjustment in scenario one, for drones exceeding the total time limit, the number of missions assigned to them should be reduced or some orders transferred. Specifically, one or two orders that are farther away should be removed from the drone's current order allocation and transferred to other drones with sufficient remaining battery power and a total time less than the vehicle's travel time. Using a drone, the system then replans the flight path and return route for the drone after removing orders, and calculates its total time again according to the formula in step 400. This process is repeated until the drone's total time is less than or equal to the vehicle's travel time. For adjustments in scenario two, where the total time is significantly less than the vehicle's travel time, the system first attempts to adjust the task execution order. This involves replanning the order of order access within the sub-region and appropriately increasing the continuity of the flight path to slightly extend the flight time by 3 to 5 minutes. If, after adjusting the execution order, the time difference still exceeds a preset threshold, the system sends a suggestion to the vehicle route planning module to fine-tune the location of the next rendezvous point, prioritizing... The new rendezvous point is a road node between the current stop and the original rendezvous point that has the conditions for drone take-off and landing. While shortening the vehicle's travel distance, the return distance of the drone is appropriately increased. After recalculating the vehicle-drone time, the time difference between the two is controlled within a preset range of 5 minutes. When the difference between the total arrival time of all drones and the vehicle's travel time is controlled within the range of total drone time ≤ vehicle travel time and time difference not exceeding 5 minutes, it is determined that the vehicle-drone spatiotemporal synchronization adjustment is completed, realizing the spatiotemporal coupling of the vehicle and the drone. The system locks the current drone task allocation scheme, flight path planning scheme and vehicle travel path scheme, and issues execution instructions.

[0055] This embodiment calculates the time taken by the vehicle and the drone based on the flight path, return route, and real-time road conditions, replacing the straight-line distance estimation method. This makes the time calculation results more consistent with the actual operation scenario, avoiding the time deviation problem of the vehicle and drone arriving at the rendezvous point from the source. A differentiated adjustment strategy is implemented based on the vehicle-drone time comparison results, solving the waiting problem caused by the drone's total time exceeding the limit and avoiding resource idleness caused by the drone arriving too early. This ensures a high degree of matching between the vehicle's driving rhythm and the drone's flight rhythm, improving the efficiency of vehicle-drone coordination. Clearly defined time difference preset thresholds and order adjustment quantities are set, providing specific quantitative basis for vehicle-drone time adjustments and avoiding task chaos or reduced delivery efficiency caused by irregular adjustments, ensuring the scientific and operable nature of the adjustment process. Dynamic adjustments eliminate time conflicts in vehicle-drone connection, ensuring that the drone can rendezvous with the vehicle in a timely manner after completing delivery, avoiding delays in subsequent delivery tasks due to waiting or poor connection, and ensuring the continuity of the overall delivery process. This avoids the waste of power caused by the drone waiting for a long time, and also prevents the vehicle from excessively accelerating to accommodate the drone, optimizing the resource utilization efficiency of both the drone and the vehicle, indirectly reducing the operating costs of logistics and delivery.

[0056] In a preferred embodiment of the present invention, step 5 includes:

[0057] Step 500: When a sudden drone malfunction, temporary airspace closure, or abnormal weather change is detected, the affected drone is identified. Using the current location of the affected drone or its planned next critical point as the center, and combining the current dynamic no-fly zone and weather information with the drone's remaining battery power, a miniature safety cone is constructed for emergency situations. This miniature safety cone is used to find a safe emergency landing point or alternative path. Specifically, the system constructs a multi-dimensional anomaly detection system through the drone flight control system, airspace management system, and real-time weather monitoring module to capture various abnormal signals in real time, including equipment fault codes from the drone flight control system (such as power system failure, positioning system anomaly), temporary no-fly zone activation notifications from the airspace management system, and real-time data from the weather monitoring module. In the event of sudden severe weather data, such as a sudden increase in wind speed to above 8 m / s or sudden heavy rainfall, the system immediately determines that an anomaly has occurred in the collaborative delivery process and initiates an anomaly response procedure. Based on the source and coverage of the anomaly signal, the system accurately locates the affected drones. If the problem is a local equipment failure of the drone, the system directly identifies the faulty drone through its unique device identifier. If the problem is a temporary closure of airspace, all drones performing delivery tasks within the coverage area of ​​the temporary no-fly zone are identified as affected. If the problem is a sudden weather change, all drones within the area affected by the severe weather are designated as affected drones. At the same time, the system records the core status of each affected drone, including its current real-time three-dimensional coordinates, remaining battery power, number of completed delivery orders, and location of uncompleted delivery orders.

[0058] For each affected drone, the system immediately extracts the core real-time data for cone construction. If the drone can transmit location information normally, its current location is used as the core reference point. If the drone's positioning fails, its next planned key point, such as the next order point or navigation point on the planned route, is used as the core reference point. Simultaneously, the system retrieves the currently updated dynamic no-fly zone spatial coordinate set, real-time weather information (wind speed, wind direction, visibility), the drone's remaining battery power (mAh), and power consumption per unit distance (mAh / km), among other performance parameters. Referring to the accessibility cone construction logic and considering the specificities of emergency scenarios, the system constructs a miniature safety cone for emergency situations, using the core reference point of the affected drone as the vertices. The specific calculation and construction process is as follows: First, the maximum emergency range of the drone with its current remaining battery power is calculated. The maximum emergency range (km) = UAV remaining battery power (mAh) ÷ power consumption per unit distance (mAh / km) × 0.8, where multiplying by 0.8 is to reserve 20% of emergency battery power to cope with sudden energy consumption during emergency flight and ensure flight safety. Using the core reference point as the apex and the calculated maximum emergency range as the generatrix, combined with the altitude restrictions for low-altitude emergency flights in urban areas (0 meters to 80 meters, reducing flight altitude to improve the safety of emergency operations), an initial conical emergency airspace is constructed. The spatial intersection of this initial airspace with the current dynamic no-fly zone and the area affected by sudden severe weather is eliminated, and finally a continuous airspace volume without safety hazards is formed, namely a miniature safety cone. This cone defines a unique safe airspace range for subsequent search for emergency landing points and planning alternative routes, ensuring that all emergency operations are carried out within the cone range.

[0059] Step 501: Calculate the response strategy within the miniature safety cone. If the corresponding drone can continue flying, plan a safe path to avoid the obstacle within the miniature safety cone and guide the drone to the new target point. If the corresponding drone cannot continue flying and vehicle intervention is required, plan a temporary detour path for the vehicle to the apex of the miniature safety cone, and insert the corresponding temporary task as the highest priority into the vehicle access sequence for local replanning. Specifically, this includes: the system comprehensively verifies the flight capability of each affected drone, based on criteria including whether the drone's remaining battery power meets the short-distance flight requirements within the cone, whether the power / navigation system is functioning normally, and whether current weather conditions support the drone's continued flight. For flight operations, drones are categorized into two groups based on verification results: those that can continue flying and those that cannot, with corresponding response strategies developed for each. If an affected drone is determined to be able to continue flying, the system plans alternative routes within its corresponding miniature safety cone. First, the new target point for the drone is identified, prioritizing alternative drone docking points within the cone, unobstructed public airspace nodes, or the closest uncompleted order point within the cone to the current location. Starting from the drone's current location and ending at the new target point, a safe flight path is planned within the miniature safety cone, avoiding all obstacles (no-fly zones, buildings, areas affected by severe weather). The path planning follows the principles of shortest distance and highest safety, while simultaneously verifying the overall trajectory of the path. The flight distance is less than or equal to the drone's maximum emergency range, ensuring the drone can successfully reach the new target point. The planned safe path is pushed to the drone's flight control system in real time, guiding the drone to fly to the new target point and continue its executable delivery task. Simultaneously, the drone's mission planning information is updated. If the affected drone is determined to be unable to continue flying, the system immediately initiates a coordinated intervention process with ground vehicles to ensure the proper handling of the drone and unfinished orders. First, the apex of the miniature safety cone (i.e., the drone's core reference point, current location, or planned next critical point) is determined as the vehicle's emergency intervention point; this point is the drone's awaiting recovery / disposal location. Based on real-time urban traffic data, ground logistics vehicles... The system plans a temporary detour route from the vehicle's current location to the emergency intervention point, prioritizing routes with high traffic efficiency and short distances. It also verifies the total length and travel time of the temporary detour route to ensure the vehicle can arrive quickly. This temporary detour task is marked as the highest priority and inserted into the vehicle's current access sequence for local replanning. The insertion position is before the point the vehicle is about to reach, ensuring the vehicle prioritizes the emergency intervention task, while other existing tasks are postponed accordingly. The system simultaneously updates the vehicle's route plan and pushes information such as the coordinates of the emergency intervention point and the drone's status to the vehicle terminal, guiding the vehicle to retrieve the drone. Any unfinished delivery orders from the drone are then completed by the vehicle.Once all response strategies are planned, the system monitors the execution status of drones and vehicles in real time. As the drones fly along the planned path, they continuously report location and battery level information. When vehicles travel according to the replanned access sequence, their progress is synchronized in real time, ensuring the effective implementation of anomaly response strategies. Simultaneously, anomalies, response measures, and execution progress are synchronized to the logistics dispatch platform, achieving full-process status visibility.

[0060] This embodiment constructs a multi-dimensional anomaly detection system to achieve real-time capture and rapid judgment of anomalies such as drone malfunctions, temporary airspace closures, and sudden weather changes. This addresses the issue of delayed anomaly response in collaborative modes and enhances the system's anti-interference capabilities in complex urban environments. A miniature safety cone is constructed around the drone's current location or planned key points. Emergency flight range is calculated based on remaining battery power, and emergency power is reserved. Dangerous airspace is eliminated, defining a unique safe zone for all emergency operations. This prevents drones from flying into no-fly zones or areas with severe weather under abnormal conditions, ensuring the safety of drone equipment and operations. Based on the drone's flight capability assessment, a differentiated response plan is developed. Drones capable of continuing flight are guided to safe paths to continue their missions. When a drone is unable to fly, an emergency vehicle intervention is initiated, avoiding resource waste or mission stagnation caused by a single response method and improving the flexibility of anomaly handling. The temporary detour task of drone recovery is set as the highest priority and inserted into the vehicle access sequence for local replanning without global route adjustment. This enables rapid vehicle intervention while minimizing the impact on the original delivery tasks. At the same time, unfinished orders from drones are taken over by vehicles to avoid order delays or omissions and ensure the continuity of high-efficiency delivery services. Anomalies, response strategies, and the execution progress of vehicles and drones are synchronized to the logistics dispatching platform in real time, enabling full-process visualization of anomaly handling. This facilitates real-time monitoring and manual intervention by dispatchers, improving the overall scheduling and control capabilities of collaborative delivery.

[0061] In a preferred embodiment of the present invention, step 6 includes:

[0062] Step 600: Continuously monitor the real-time data stream for new order inputs, vehicle location updates, drone status updates, traffic condition changes, dynamic no-fly zones and weather information changes, and abnormal event triggers. When any status change or new event is detected, determine the smallest affected link. If a new order is input, add the new order to the pending orders and update the reachability cone and stop point-candidate order set mapping relationship. If the vehicle location or status is updated, start from the corresponding position in step 2 or step 3 to perform local path adjustments. If the drone status is updated or an abnormal event is triggered, start from the corresponding link in step 4 or step 5 to perform task reassignment or abnormal response. Specifically, this includes: system setup 7×24. The 24 / 7 real-time data stream monitoring module connects to all data sources across the entire collaborative delivery process, continuously collecting and verifying changes in six core data categories: new order input data from the logistics order management system, vehicle location and status update data from vehicle positioning and status sensors, drone status update data from the drone battery management and flight control system, traffic condition change data from the urban traffic management platform, dynamic no-fly zone change data from the airspace management system, and weather information change data from the meteorological service platform. It also connects to an anomaly detection system to capture trigger signals for various abnormal events in real time. The monitoring module refreshes data every 10 seconds to ensure that all dynamic changes are captured immediately. There is no data lag or omission; the monitoring module compares and analyzes each data refresh result, comparing the current data with the baseline data from the previous closed-loop calculation to determine if there are any data changes or new event triggers. If the value, attribute, or status of a certain category of data differs from the baseline data, it is determined that a dynamic change has occurred. If the anomaly detection system detects signals such as faults, no-fly zones, or sudden weather changes, it is determined that an abnormal event has been triggered. Once a dynamic change or new event is detected, the monitoring module immediately issues an early warning and initiates the location process for the affected links. Based on the business logic and data relationships of each link in the delivery process, the system accurately identifies the type of dynamic change or event trigger based on the type of dynamic change or event trigger. Identify the smallest affected business link and clarify the corresponding optimization entry point to avoid wasting computing power and causing process chaos due to global adjustments. The specific judgment and entry rules are as follows: If a new order is detected, the system first completes the attribute identification of the new order according to the rules in step 001, adds the new orders that require personal signature to the vehicle delivery set, and adds the remaining new orders to the pending orders; then, only for the drone docking points in the area to which the new order belongs, the system re-executes the accessibility cone reconstruction in step 100, the spatial inclusion relationship judgment in step 101, and the mapping relationship update in step 102, that is, only the docking point-candidate order set mapping relationship affected by the new order is updated, and the docking point mapping relationship remains unchanged.If a vehicle location or status update is detected, such as a vehicle speed adjustment due to road conditions or a temporary vehicle stop, the system determines that this change affects the vehicle path planning and UAV task assignment stages. It then directly intervenes at the corresponding position in the vehicle access sequence planning stage (steps 200-202) or the UAV sub-region allocation and path planning stage (steps 300-302). If the vehicle only updates its location and has not yet reached the next stop point, it intervenes at the candidate point path increment calculation stage (step 201) to perform local path adjustments. If the vehicle has reached the stop point and its status has changed, it intervenes at the reachability cone reconstruction stage (step 300) to re-adapt UAV task assignment. If a UAV status update is detected (such as fluctuations in remaining battery power or adjustments in flight speed), the system will re-adapt the UAV task assignment. If a change is triggered by an abnormal event, the system determines that the change affects the vehicle-machine spatiotemporal coupling and abnormal response process. It then directly intervenes at the corresponding position in the vehicle-machine time comparison and dynamic adjustment step (steps 400-401) or the abnormal response strategy planning step (steps 500-501). If it is only a routine drone status update, it recalculates and adjusts the vehicle-machine time from the drone total time calculation step (step 400). If it is triggered by an abnormal event, it directly initiates the abnormal response process (step 500). The system uniformly archives information such as the detected dynamic change / event trigger type, the smallest affected link, and the corresponding optimization intervention position, and synchronizes this information in real time to the delivery decision optimization module, providing a clear execution basis for the incremental optimization in step 601.

[0063] Step 601: Based on the vehicle access sequence, stop-candidate order set mapping relationship, and flight path planning results generated in the previous cycle, incremental optimization is performed only on the affected local areas, rather than global recalculation. This process continues to form a real-time iterative closed-loop decision system until all delivery tasks are completed. Specifically, after receiving synchronization information from the affected links, the delivery decision optimization module immediately retrieves all baseline decision results generated in the previous cycle from the real-time database, including the vehicle access sequence, stop-candidate order set mapping relationship, drone sub-region allocation scheme, drone flight path and return route planning results, and vehicle-drone spatiotemporal data. The coupling adjustment scheme, etc., keeps the baseline results unaffected by dynamic changes / event triggers unchanged as the basis for this optimization. The system only performs targeted incremental optimization calculations for the least affected links determined in step 600. All optimization processes are based on the original baseline results and do not involve recalculation of unaffected links. Specifically, if the optimization entry point is the update of the docking point-candidate order set mapping relationship, only the docking points affected by new orders are recalculated according to the calculation rules of steps 100 to 102 to determine the reachability cone and order matching relationship. The mapping relationships of other docking points directly use the baseline results and do not need to be recalculated. If the optimization entry point is a local adjustment of the vehicle access sequence, only the candidate points affected by changes in vehicle status / position are recalculated according to the calculation rules in step 201 for the increase in path and the cost-benefit ratio. The order of the remaining determined points remains unchanged. Only the points that need adjustment are locally inserted or the order between adjacent points is fine-tuned. There is no need to rebuild the set of points to be visited and iterate globally. If the optimization entry point is UAV task assignment and path planning, only the UAVs whose status is updated are recalculated according to the calculation rules in step 301 for the remaining range and sub-region allocation. Their flight paths are replanned according to the rules in step 302. The tasks and paths of the remaining UAVs are recalculated. The solution directly uses the baseline results without re-clustering and reassignment. If the optimization entry point is vehicle-machine spatiotemporal coupling adjustment, only the total time to reach the rendezvous point for the drones whose status has been updated is recalculated according to the calculation rules in step 400, and compared with the vehicle travel time obtained according to the calculation rules in step 401. Local adjustments are then made according to the differential adjustment rules in step 401, while the vehicle-machine time matching schemes for the remaining drones remain unchanged. If the optimization entry point is abnormal response, only the abnormal response process is executed for the affected drones according to the rules in steps 500 to 501. The delivery schemes for the remaining drones and vehicles operating normally use the baseline results and are unaffected.All the formulas and rules for incremental optimization calculations are consistent with the original calculation rules of the corresponding links to ensure the consistency of the optimization results. For example, when recalculating the drone's endurance mileage, the formula is still used: remaining endurance mileage (km) = remaining battery power of the drone (mAh) ÷ power consumption per unit mileage (mAh / km) ÷ meteorological correction factor. When recalculating the vehicle path increment, the formula is still used: path increment (km) = LAC + LCB - LAB. The system integrates the results obtained from local incremental optimization calculations with the benchmark decision results of the previous round, only replacing the original results of the affected links, and keeping the unaffected results unchanged, forming a new decision result for this cycle and updating it to the real-time database as the benchmark data for the next cycle; The system repeats the real-time monitoring and determination of affected links in step 600, and the local incremental optimization and result update in step 601, forming a data-driven and real-time iterative closed-loop decision-making system. This system always operates continuously with the logic of capturing changes, locating links, locally optimizing, and updating results until the logistics scheduling center platform feedbacks that all distribution orders have been completed, there are no pending orders to be assigned, and no in-transit tasks, then the system terminates the closed-loop iteration process.;

[0064] In this embodiment, a full-dimensional monitoring system covering orders, vehicle status, road conditions, airspace, and meteorology is built, refreshing data at high frequency and conducting comparative analysis to ensure that all dynamic changes and abnormal events in the distribution process are captured in the first time, solving the core problem that static planning cannot adapt to the dynamic urban environment; Based on business logic and data correlation relationships, only the smallest business links affected by dynamic changes are determined, avoiding indiscriminate global adjustments, reducing the system's computational volume and data processing load, and improving the efficiency of decision-making optimization, enabling the system to adapt to the real-time decision-making needs of high-density distribution scenarios; Only targeted optimization is carried out on the affected links, and the unaffected decision results are directly used without global re-planning, avoiding problems such as interruption of vehicle distribution tasks and path confusion caused by overall adjustments, and ensuring the continuity of the distribution process to the greatest extent, adapting to the needs of high-timeliness distribution scenarios such as urgent medical delivery and instant retail; Through the continuous cycle of monitoring, positioning, optimization, and updating, the decision-making plan for collaborative distribution is always synchronized with real-time order requirements, vehicle status, and environmental changes, realizing the dynamic iterative optimization of the decision-making plan, and improving the adaptability and flexibility of ground vehicle and drone collaborative distribution in complex urban environments; Local incremental optimization always uses the same calculation formulas and planning rules as the original links, avoiding conflicts in decision-making results caused by inconsistent optimization rules, ensuring the scientificity, consistency, and executability of the full-process distribution decision-making, and improving the completion rate and accuracy of order distribution; The closed-loop decision-making system realizes the full-process automated operation from dynamic change capture to local optimization execution, without the need for manual participation in global planning and adjustment, only requiring manual assistance in special abnormal situations, reducing the dependence on manual scheduling, and improving the automation and intelligence level of urban last-mile logistics distribution scheduling.

[0065] Embodiments of the present invention also provide a computing device, including: a processor and a memory storing a computer program, wherein the computer program, when executed by the processor, performs the method described above. All implementations in the above method embodiments are applicable to this embodiment and can achieve the same technical effects.

[0066] Embodiments of the present invention also provide a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the method described above. All implementations in the above method embodiments are applicable to this embodiment and can achieve the same technical effects.

[0067] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for collaborative delivery using ground logistics vehicles and drones, characterized in that, The method includes: Step 1: Based on the orders to be assigned, dynamically construct an accessibility cone for each drone docking point; determine the spatial inclusion relationship between the destination coordinates of the orders to be assigned and the accessibility cone, identify the drone-accessible candidate orders, and establish a docking point-candidate order set mapping relationship; Step 2: Based on the mapping relationship between the vehicle delivery set and the stop point-candidate order set, and with the goal of the collaborative benefit-cost ratio, plan the driving route for the ground vehicles and iteratively generate the vehicle access sequence; Step 3: When a vehicle arrives at a stop according to the vehicle access sequence, the accessibility cone of the stop is reconstructed, the candidate orders covered by the accessibility cone are spatially clustered, a sub-region is assigned to each drone, and the drone flight path and return route are planned. Step 4: Based on the planned flight path, calculate the time when the UAV completes its mission and arrives at the next rendezvous point and the time when the vehicle arrives at the same rendezvous point. If the two times are not synchronized, make dynamic adjustments to achieve vehicle-machine spatiotemporal coupling. Step 5: When an anomaly is detected, reconstruct the micro safety cone for emergency situations, centered on the current location of the affected drone or the next planned critical point, and calculate the response strategy within the micro safety cone. Step 6: Continuously execute steps 1 to 5 in a loop. When a new order is placed, the status is updated, or an anomaly is triggered, incremental optimization is performed on the affected links based on the results of the previous round to form a closed-loop decision until all delivery tasks are completed.

2. The method for collaborative delivery of ground logistics vehicles and drones according to claim 1, characterized in that, Before step 1, access real-time data streams, including the set of orders to be delivered, real-time location and status of vehicles, number of available drones and their respective real-time battery level and maximum range, real-time city traffic conditions, preset drone docking point network, and dynamic no-fly zones and weather information; Based on order attributes, an initial screening is performed, and orders requiring personal signature are assigned to the vehicle delivery collection, while the remaining orders are designated as pending allocation orders.

3. The method for collaborative delivery of ground logistics vehicles and drones according to claim 2, characterized in that, Step 1 includes: Based on the dynamic no-fly zone and weather information in the accessed real-time data stream, as well as the remaining flight range of the UAV, for each candidate UAV landing point, an accessibility cone is constructed with the corresponding landing point as the center. The accessibility cone is a continuous and safe airspace volume formed with the landing point as the vertex, the maximum reachable radius of the UAV as the generatrix, and the part intersecting with the dynamic no-fly zone space is removed. The spatial inclusion relationship between the destination coordinates of the order to be assigned and the accessibility cone of each drone docking point is determined. If the destination of the order to be assigned is located within the accessibility cone of any drone docking point, the corresponding order to be assigned is marked as a drone reachable candidate order. Establish a correspondence between each drone docking point and the candidate drone orders that it can serve, denoted as the docking point-candidate order set mapping relationship.

4. The method for collaborative delivery of ground logistics vehicles and drones according to claim 3, characterized in that, Step 2 includes: The vehicle's starting location, all demand locations in the vehicle delivery set, and all drone docking points are merged to form a set of points to be visited; the vehicle access sequence is initialized, with the initial element of the vehicle access sequence being the vehicle's starting location. Repeat the following operations until all points in the set of points to be visited are added to the vehicle access sequence. That is, for each candidate point in the set of points to be visited that has not yet been added to the vehicle access sequence, calculate the increase in the vehicle travel path when inserting the corresponding candidate point between any two adjacent points in the current vehicle access sequence. Based on the mapping relationship between the stop point and the candidate order set corresponding to the candidate point, determine the collaborative benefit cost ratio of the corresponding candidate point. The numerator of the collaborative benefit cost ratio is determined based on the number of drone-accessible candidate orders covered in the accessibility cone of the corresponding candidate point and the urgency of the orders. The denominator is the additional time cost for the vehicle to detour to the corresponding candidate point. From all candidate points and their insertion positions, select the candidate point and its insertion position that maximizes the collaborative benefit-cost ratio, and insert the corresponding candidate point into the insertion position of the vehicle access sequence. After all points have been inserted, the final vehicle access sequence is obtained. The stopping points in the vehicle access sequence are collaborative operation nodes with the ability to provide drone-accessible candidate order services.

5. The method for collaborative delivery of ground logistics vehicles and drones according to claim 4, characterized in that, Step 3 includes: When a vehicle travels according to the vehicle access sequence and arrives at a stop, based on the current dynamic no-fly zone and weather information and the remaining battery power of the available drones, the accessibility cone of the corresponding stop at the current moment is reconstructed, and the candidate orders of drones to be delivered covered by the accessibility cone are obtained as the candidate order set for the current batch. The candidate order set of the current batch is clustered according to spatial location. Combining the remaining power and maximum range of each available drone, a sub-region within the reachability cone is dynamically allocated to each available drone. The sub-region ensures that multiple order points allocated to the same drone are located within the corresponding sub-region, and the total flight path of the corresponding drone to complete all allocated orders and return to the next rendezvous point is within its range. Plan a flight path for each drone within its assigned sub-area to each order point, and plan its return route to the next scheduled stop on the vehicle.

6. The method for collaborative delivery of ground logistics vehicles and drones according to claim 5, characterized in that, Step 4 includes: Based on the flight path planned for each drone, the flight time for each drone to complete all assigned orders in the current batch is calculated, and the return time for the corresponding drone to fly from the last order point to the next predetermined stop point, i.e. the next rendezvous point, in the vehicle access sequence is calculated. The two are added together to obtain the total time for the drone to arrive at the next rendezvous point. Based on real-time traffic data, the system predicts the travel time of vehicles from the current stop to the next rendezvous point. The total time of each drone is compared with the vehicle travel time. If the total time of a drone is greater than the vehicle travel time, the number of tasks assigned to the corresponding drone is reduced or some orders are transferred to the remaining available drones. The flight path of the corresponding drone is then replanned until the total drone time is less than or equal to the vehicle travel time. If the total time of a drone is less than the vehicle travel time and the difference exceeds a preset threshold, the task execution order of the drones is adjusted or the location of the next rendezvous point is suggested to be adjusted until the time difference between the two is within a preset range, thus achieving spatiotemporal coupling between vehicles and drones.

7. The method for collaborative delivery of ground logistics vehicles and drones according to claim 6, characterized in that, Step 5 includes: When a sudden drone malfunction, temporary airspace closure, or abnormal weather changes are detected, the affected drone is identified. Centered on the current location of the affected drone or its next planned key point, and combined with the current dynamic no-fly zone and weather information, as well as the remaining battery power of the corresponding drone, a miniature safety cone is constructed for emergency situations. The miniature safety cone is used to find a safe emergency landing point or alternative path. Within the miniature safety cone, a response strategy is calculated. If the corresponding drone can continue flying, a safe path is planned within the miniature safety cone to avoid obstacles and guide the drone to a new target point. If the corresponding drone cannot continue flying and vehicle intervention is required, a temporary detour path is planned for the vehicle to the apex of the miniature safety cone, and the corresponding temporary task is inserted as the highest priority into the vehicle access sequence for local replanning.

8. The method for collaborative delivery of ground logistics vehicles and drones according to claim 7, characterized in that, Step 6 includes: Continuously monitor the real-time data stream for new order inputs, vehicle location updates, drone status updates, traffic condition changes, dynamic no-fly zones and weather information changes, and abnormal event triggers. When any status change or new event is detected, determine the smallest affected link. If a new order is input, add the new order to the pending orders and update the accessibility cone and stop point-candidate order set mapping relationship. If the vehicle location or status is updated, start from the corresponding position in step 2 or step 3 to perform local path adjustment. If the drone status is updated or an abnormal event is triggered, start from the corresponding link in step 4 or step 5 to perform task reassignment or abnormal response. Based on the vehicle access sequence, stop-candidate order set mapping relationship, and flight path planning results generated in the previous cycle, incremental optimization is performed only on the affected local areas, rather than global recalculation; the above process continues to form a real-time iterative closed-loop decision system until all delivery tasks are completed.

9. A computing device, characterized in that, include: One or more processors; A storage device for storing one or more programs, which, when executed by one or more processors, cause the one or more processors to implement the method as described in any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a program that, when executed by a processor, implements the method as described in any one of claims 1 to 8.

Images