A shared unmanned aerial vehicle intelligent scheduling method and system based on an internet of things

Abstract

The application provides a shared unmanned aerial vehicle intelligent scheduling method and system based on the Internet of Things, which comprises the following steps: acquiring sensing data of each site in actual operation to calculate comprehensive sensing values of the sites and encapsulating the comprehensive sensing values into site state vectors; constructing a multi-factor game model according to the site state vectors to obtain utility values of shared unmanned aerial vehicles of the current site; calculating a cooperation index according to the number of current tasks to be processed, the number of unmanned aerial vehicles for receiving orders and the comprehensive sensing values; generating a path scheme set without conflict in multi-machine cooperation according to the cooperative formation, the task area responsible for by each site and the residual power of shared unmanned aerial vehicles of the formation members as input; and dynamically updating the path scheme set through a neighborhood broadcast mechanism. The application effectively overcomes the problem of insufficient resource coordination and scheduling efficiency of the existing shared unmanned aerial vehicle scheduling scheme in a large-scale and dynamic environment, and improves the service capacity, energy efficiency level and business continuity of the system.

Description

Technical Field

[0001] This invention belongs to the field of intelligent scheduling of shared drones, and particularly relates to an intelligent scheduling method and system for shared drones based on the Internet of Things. Background Technology

[0002] With the acceleration of urbanization and the advancement of smart city construction, shared drones, as a new type of aerial service platform, are gradually being widely used in various scenarios such as logistics delivery, environmental monitoring, security patrol, and emergency rescue. The development of IoT technology provides fundamental support for the intelligence of shared drones. By deploying IoT sites (such as charging piles, data collection nodes, and dispatch centers) in various areas, shared drone systems can achieve real-time management and collaborative scheduling of distributed drone resources. Existing drone scheduling systems mainly rely on centralized cloud platforms or semi-distributed edge computing nodes to uniformly collect information such as user task requests, drone location and battery status, and site resource conditions, thereby completing task allocation and path planning. However, with the widespread application of shared drones in large-scale and complex environments such as cities and industrial parks, existing technologies are gradually revealing numerous problems in meeting the requirements of system scalability, real-time scheduling, and resource coordination. On the one hand, centralized scheduling faces severe communication and computing bottlenecks. The centralized uploading of massive amounts of task data and drone status information to the cloud leads to network congestion and increased decision-making delays, making it difficult to meet the real-time scheduling needs of high-density shared drone deployments. On the other hand, while existing distributed or local scheduling mechanisms alleviate the pressure of centralization to some extent, they are mostly limited to task allocation within a single site or local area, lacking a global perspective. This results in uneven resource allocation and insufficient coordination among sites or drones, often leading to high drone idle rates and resource redundancy or shortages in certain areas. Furthermore, existing solutions are weak in responding to dynamic environments and emergencies. When faced with drone malfunctions, sudden increases in task demands, or severe weather, the scheduling system struggles to respond quickly and flexibly, easily leading to task backlog, decreased system robustness, and even delays or failures of critical tasks.

[0003] To address the aforementioned issues, existing technologies urgently need a new intelligent scheduling mechanism that can simultaneously possess capabilities such as distributed perception, global collaboration, efficient local execution, and dynamic adaptation, in order to improve the overall operational efficiency and stability of shared drones in the Internet of Things environment. Summary of the Invention

[0004] The purpose of this invention is to propose an intelligent scheduling method and system for shared drones based on the Internet of Things. It addresses key issues such as resource scheduling imbalance during peak mission periods, mismatch between regional mission and resource supply and demand, and poor system response to dynamic environments and emergencies by proposing an innovative distributed perception and collaborative decision-making framework.

[0005] To achieve the above objectives, a first aspect of the present invention provides a method for intelligent scheduling of shared drones based on the Internet of Things, the method comprising the following steps:

[0006] S1. Obtain the perception data of each station in actual operation to calculate the comprehensive perception value of each station, encapsulate it into a station state vector, and synchronize it to neighboring stations and all shared drones within the station in real time through a neighborhood broadcast mechanism; wherein, the perception data includes the number of tasks to be processed, the number of available drones, the average remaining power of drones within the station, and the real-time load factor of the station.

[0007] S2. Construct a multi-factor game model based on the site state vector to obtain the utility value of the shared drone at the current site; after all the shared drones perform parallel calculations based on the multi-factor game model on all the sites they receive, select the site with the highest utility value of the shared drone as the order target, and define it as the target site.

[0008] S3. For multiple shared drones that have accepted orders at the same target site, calculate the coordination index based on the number of tasks to be processed, the number of drones that have accepted orders, and the comprehensive perception value; when the coordination index exceeds a preset threshold, select the shared drones with sufficient power and closest to the core area of ​​the task to form a coordinated formation.

[0009] In the coordinated formation, the shared drone closest to the mission core area serves as the leader, and the other shared drones serve as formation members.

[0010] S4. Based on the cooperative formation, the respective task areas, and the remaining battery power of the shared drones of the formation members as inputs, with the optimization objectives of minimizing path length, energy consumption, and the overlap of the flight paths of the formation members, and combined with battery constraints, task time constraints, and cooperative obstacle avoidance constraints, a set of conflict-free multi-drone cooperative path schemes is generated.

[0011] S5. For the cooperative formation, execute the path of the path scheme set, collect feedback data from all shared UAVs after execution, and dynamically update the path scheme set through the neighborhood broadcast mechanism in combination with the updated site state vector.

[0012] Furthermore, the comprehensive perception value is obtained by weighted summation of the current number of tasks to be processed, the number of available drones, the average remaining battery power of drones in the site, and the real-time load factor of the site; wherein the real-time load factor is determined based on the current number of tasks to be processed and the number of available drones.

[0013] The site status vector includes the number of currently pending tasks, the number of available drones, the average remaining battery power of drones within the site, the site's real-time load factor, and the comprehensive perception value.

[0014] Furthermore, the multi-factor game model is expressed as:

[0015]

[0016] Among them, R i For site S i The average revenue per task in the pending assignments reflects the task's business priority or economic value. j d j Current remaining battery power, C i This is a comprehensive perceived value used to reflect S. i Task pressure and resource load, D ij For d j With site S i physical distance, e c U represents the energy consumption per unit distance, where ∈ is a small positive number to avoid the denominator approaching zero; γ, α, β, and λ are the adjustment weights for revenue, energy consumption, load, and safety factor set by this invention, respectively; U ij It is the utility value; The safety margin constraint for endurance is used to simulate the remaining battery power of a shared drone after completing a mission, preventing the risk of drones failing to return due to remote distance or insufficient battery power.

[0017] Furthermore, after all the currently shared drones perform parallel calculations on all the received sites based on a multi-factor game model, they select the site with the highest utility value of the shared drones as the order target. If the utility value of the site is less than the utility threshold, the shared drones at the current site choose to remain on standby. This is to ensure that when resources are unevenly distributed in terms of task pressure, some drones can be automatically put on standby to reduce the risk of local overload.

[0018] Furthermore, when the coordination index exceeds a preset threshold, shared drones with sufficient power and close to the core area of ​​the mission are selected to form a coordinated formation. Specific steps include:

[0019] The design criteria are used to select collaborative formations that actually participate in the collaboration; these criteria include:

[0020] Battery priority: Select the top m non-shared HMIs with the highest remaining battery power;

[0021] Location priority: Based on the distance between the drone and the mission center point, the shared drone closest to the core mission area is selected first.

[0022] Furthermore, the objective function for the optimization objective is:

[0023]

[0024] Among them, P j For drones d j The path to be optimized, L j This is the total length of the path. For d j Expected completion Required energy consumption, Φ j Indicates the path P to be optimized. j With coordinated formation The overlap index of other member paths is expressed as a spatial-temporal coincidence measure of track intersection; κ1, κ2, and κ3 are adjustment factors that can be flexibly set according to the station task density or airspace complexity.

[0025] Furthermore, the path to be optimized is subject to the following constraints:

[0026] Power constraint: Ensure that the energy consumption in the path to be optimized does not exceed a preset threshold of the remaining power of the shared drones of the formation members;

[0027] Task deadline constraints: All paths to be optimized must meet the given task deadline within their respective task areas;

[0028] Collaborative obstacle avoidance constraint: Avoid tracks where the overlap index between the path to be optimized and other paths in the collaborative formation exceeds a preset threshold.

[0029] Furthermore, the feedback data includes task completion rate, path energy consumption, and path execution fluctuation.

[0030] Furthermore, S5 specifically includes:

[0031] Feedback data from shared UAVs executing the path scheme set in the collaborative formation is collected and summarized into a local feedback matrix for the target site;

[0032] The target station adjusts its station state vector according to the local feedback matrix to generate an updated comprehensive perception value, expressed as:

[0033]

[0034] in, This is the original site's comprehensive perception value. The updated perception value is represented by δ, which is the historical state fusion factor. For coordinated formation The average completion rate of medium tasks The average value of the path execution deviation. ω1, ω2, and ω3 are the average remaining battery power of each drone, and ω1, ω2, and ω3 are the weights of the feedback factor.

[0035] Finally, the target site broadcasts the updated site state vector to neighboring sites and the system, forming an iterative update of the global graph.

[0036] In a second aspect, the present invention provides an intelligent scheduling system for shared unmanned aerial vehicles based on the Internet of Things, the system comprising:

[0037] Distributed sensing module: Used to acquire sensing data from each station in actual operation, calculate the comprehensive sensing value of each station, encapsulate it into a station state vector, and synchronize it in real time to neighboring stations and all shared drones within the station through a neighborhood broadcast mechanism; The sensing data includes the number of tasks to be processed, the number of available drones, the average remaining battery power of drones within the station, and the real-time load factor of the station.

[0038] Game allocation module: used to construct a multi-factor game model based on the site state vector to obtain the utility value of the shared drone at the current site; after all the shared drones perform parallel calculations based on the multi-factor game model on all the sites they receive, the site with the highest utility value of the shared drone is selected as the order target and defined as the target site.

[0039] Elastic Formation Module: Used to calculate a coordination index for multiple shared drones that have accepted orders at the same target site, based on the number of tasks to be processed, the number of drones accepting orders, and the comprehensive perception value; when the coordination index exceeds a preset threshold, the shared drones with sufficient power and closest to the core area of ​​the task are selected to form a coordinated formation.

[0040] In the coordinated formation, the shared drone closest to the mission core area serves as the leader, and the other shared drones serve as formation members.

[0041] The path optimization module is used to generate a set of conflict-free path schemes for multi-drone cooperation based on the cooperative formation, the task area each drone is responsible for, and the remaining battery power of the shared drones of the formation members as inputs. The optimization objectives are to minimize the path length, energy consumption, and the overlap of the flight paths of the formation members. It combines battery constraints, task time constraints, and cooperative obstacle avoidance constraints.

[0042] Feedback Update Module: This module executes the path scheme set for the cooperative formation, collects feedback data from all shared UAVs after execution, and dynamically updates the path scheme set through a neighborhood broadcast mechanism in conjunction with the updated site state vector.

[0043] The beneficial technical effects of the present invention are at least as follows:

[0044] This invention constructs a multi-layered perception-decision-execution closed-loop mechanism, enabling IoT sites to achieve real-time perception of regional task load and UAV resource status, forming an information sharing and collaborative perception network among multiple sites.

[0045] Building upon this foundation, the system empowers drones with autonomous collaboration and dynamic optimization capabilities. Drones can make intelligent decisions based on the mission environment, site load, and local resource status, flexibly deciding whether to accept tasks, execute collaboratively, or coordinate in formation. This achieves a fusion of global collaboration across sites and efficient local execution. Furthermore, the system possesses an adaptive optimization mechanism for unforeseen scenarios, supporting drones in dynamically adjusting task allocation and execution strategies under dynamic conditions such as resource shortages, sudden increases in tasks, and sudden environmental changes. This ensures the stability and scheduling robustness of the shared drone system in highly dynamic and complex environments.

[0046] Through the above innovative design, this invention effectively overcomes the problem of insufficient resource coordination and scheduling efficiency of existing shared drone scheduling schemes in large-scale, dynamic environments, and improves the system's service capabilities, energy efficiency, and business continuity. Attached Figure Description

[0047] The present invention will be further described with reference to the accompanying drawings, but the embodiments in the drawings do not constitute any limitation on the present invention. For those skilled in the art, other drawings can be obtained based on the following drawings without creative effort.

[0048] Figure 1 This is a flowchart of an intelligent scheduling method for shared drones based on the Internet of Things according to the present invention. Detailed Implementation

[0049] Embodiments of the present invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0050] like Figure 1 As shown in the figure, an embodiment of the present invention provides a method for intelligent scheduling of shared drones based on the Internet of Things, the method comprising:

[0051] S1. Obtain the perception data of each station in actual operation to calculate the comprehensive perception value of each station, encapsulate it into a station state vector, and synchronize it to neighboring stations and all shared drones within the station in real time through a neighborhood broadcast mechanism; wherein, the perception data includes the number of tasks to be processed, the number of available drones, the average remaining power of drones within the station, and the real-time load factor of the station.

[0052] Specifically, in this step, the system uses various stations S on the Internet of Things. i As the starting point for perception and scheduling, a multi-dimensional dynamically weighted perception mechanism with dynamic adaptability was designed. This mechanism is used to integrate the key tasks and resource status of each site in real time, providing a reliable basis for subsequent UAV task allocation decisions. Site S i In actual operation, it is necessary to sense the number T of tasks to be processed in real time. i Number of available drones U i The average remaining battery power E of the drones within the site i and the site's real-time load factor L i Considering the frequent task tidal effects and dynamic changes in resource distribution in shared drone applications within the Internet of Things (IoT) environment, traditional simple averaging or fixed weighting mechanisms are insufficient to comprehensively reflect the site pressure status. Therefore, this invention proposes an adaptive weighting model based on task dynamics. By dynamically adjusting the weight ratios of task density, drone resources, and energy consumption factors, a comprehensive perception value C for the site is generated. i As a key input for multi-site resource collaboration.

[0053] Its core calculation model is as follows:

[0054]

[0055] Among them, T i Indicates site S i The current number of tasks to be assigned, U i For S i Number of standby drones, E i For S i The average remaining battery power of available drones (values ​​have been normalized to the [0,1] range), L i The load factor of a site is calculated as follows: This is used to reflect the relationship between mission density and drone availability. t w u w e w l To achieve dynamic adaptive weighting, the system adjusts its weights in real time based on historical task fluctuations, site type, and the task tidal effect within the current time window. Specifically, the weight values ​​will be automatically adjusted for different periods, such as peak and off-peak delivery times. For example, during peak periods, the system will focus on task density and site load, assigning a weight to specific areas. t and w l Higher weighting is given to drones, while during off-peak hours, more attention is paid to drone energy consumption and battery life to improve performance. e The weight.

[0056] Furthermore, complete the comprehensive perception value C. iAfter calculation, site S i C i As a core indicator, with T i U i E i L i They are encapsulated together as a site state vector S i ={C i ,T i U i E i ,L i}, and synchronize in real time to neighboring sites and all shared drones within the site through a neighborhood broadcast mechanism. j The above design effectively ensures that the entire shared drone system can achieve a low-latency, distributed global perception network in a large-scale Internet of Things environment, providing highly reliable environmental input for the distributed game allocation of drones in subsequent steps. It solves the problems of limited site resource perception and information silos in traditional scheduling schemes, and significantly improves the intelligence and coordination of the overall system scheduling.

[0057] S2. Construct a multi-factor game model based on the site state vector to obtain the utility value of the shared drone at the current site; after all the shared drones perform parallel calculations based on the multi-factor game model on all the sites they receive, select the site with the highest utility value of the shared drone as the order target, and define it as the target site.

[0058] In this step, the station state vector S output from step one is used. i ={C i ,T i U i E i ,L i This invention proposes a multi-objective distributed game-theoretic task allocation mechanism, aiming to solve the problem of how to achieve fair, efficient, and secure task acceptance decisions for shared drones in an IoT environment, facing multiple site task requests, dynamically uneven resource distribution, and complex environmental constraints. Each standby shared drone... j Received from surrounding stations S i After receiving the status information, the system needs to make an intelligent order-accepting choice based on task benefits, drone energy consumption, site congestion, and its own battery life and safety.

[0059] Furthermore, this scheme designs a multi-factor game model that integrates revenue, energy consumption, load, and flight safety, as d j The core utility function is as follows:

[0060]

[0061] Among them, R iFor site S i The average revenue per task in the pending assignments reflects the task's business priority or economic value. j d j The current remaining battery power (normalized to [0,1]), C i The comprehensive sensing value of the site, calculated by the dynamic weighted sensing model in step one, is used to reflect S i Task pressure and resource load, D ij For d j With site S i The physical distance (in kilometers), e c λ represents the energy consumption per unit distance (e.g., 0.01 means 0.01 units of electricity are consumed per kilometer of flight), where ∈ is a small positive number to avoid the denominator approaching zero. γ, α, β, and λ are the adjustment weights for the revenue, energy consumption, load, and safety factors set by this invention, respectively. For example, β can be set larger during peak mission periods to strengthen the congestion avoidance mechanism.

[0062] Understandably, the innovation of this invention lies in U ij The function not only balances revenue and energy consumption, but also introduces a safety margin constraint on battery life for real-world use scenarios of shared drones. This simulation realistically measures the remaining battery power of a drone after completing a mission, preventing the risk of drones failing to return due to remote locations or insufficient battery power. It achieves integrated game theory decision-making that combines mission allocation and battery life safety, unlike traditional game theory models that only consider revenue and energy consumption.

[0063] drones j For all received sites S i After performing parallel computation, the utility value U is selected. ij The largest site The strategy for targeting orders is as follows:

[0064]

[0065] Where θ is a system-defined utility threshold, below which d j The option is to remain on standby until the next round of competition before participating in order-taking decisions. This mechanism has adaptive competition-sleep switching capabilities, ensuring that when resources are unevenly distributed across tasks, some drones are automatically put on standby, reducing the risk of localized overload.

[0066] Based on the above model, d j The order acceptance decision is not only related to the task revenue R i and remaining power E j Directly related, and also incorporates C from step one. iThis value enables dynamic avoidance of high-load sites, alleviates local scheduling congestion, and ensures a balanced distribution of overall system resources. Specifically, the newly added range safety item can forcibly reduce the utility value U when the battery level approaches the mission execution threshold. ij This allows drones to automatically avoid situations where the mission range is too far or they cannot return safely after completing the mission, significantly improving the system's mission completion rate and operational safety.

[0067] S3. For multiple shared drones that have accepted orders at the same target site, calculate the coordination index based on the number of tasks to be processed, the number of drones that have accepted orders, and the comprehensive perception value; when the coordination index exceeds a preset threshold, select the shared drones with sufficient power and closest to the core area of ​​the task to form a coordinated formation.

[0068] In the coordinated formation, the shared drone closest to the mission core area acts as the leader, and the other shared drones act as formation members.

[0069] Specifically, this step follows immediately after the order-taking negotiation process in step two, and targets the same website. For multiple drones receiving orders, a flexible collaborative formation mechanism was designed. This mechanism aims to dynamically generate collaborative formations based on task density, historical load, and drone status, thereby solving the problems of insufficient single-drone operation capabilities and inefficient multi-drone collaboration in large-scale tasks or high-density task areas.

[0070] In the Internet of Things (IoT) environment, the drone that successfully received the order... j Gathered into the order collection set To achieve adaptability in the collaboration mechanism, the site First, based on the current workload Combined with the number of drones receiving orders, k, and the comprehensive site perception value from step one. Calculate the synergy index

[0071]

[0072] Here, k represents the number of pending tasks at the current site, and k represents the number of drones accepting orders. The overall site load factor is the one sensed in real time during step one, and μ is the historical load adjustment coefficient, used to improve the system's willingness to collaborate under sustained high load. For example, in a park logistics scenario, if... (15 tasks to be executed at distributed delivery points), k = 4, And μ = 0.5, then

[0073] Among them, when When the system-defined collaboration threshold ω is exceeded (e.g., ω = 4), the site... If the current resources are determined to require collaborative operation, the elastic formation mechanism will be triggered; otherwise, each drone will execute subsequent tasks independently in single-machine mode.

[0074] After entering the collaboration process, the site right Chinese drones j The subset of participants who actually engage in collaboration is selected based on the following two indicators.

[0075] Battery priority: Select E j The top m drones by (remaining battery power);

[0076] Position priority: combined with d c,j The drones are sorted by distance from the mission center point, with priority given to drones closer to the core mission area.

[0077] For example, in the case where k=4 above, if the remaining battery power of drones d1, d2, d3, and d4 are {0.8, 0.7, 0.4, 0.9} respectively, and d4 and d1 are closer to the mission core area, the system will include drones d4, d1, and d2. A temporary formation was formed, and the remaining d3 were assigned as standby aircraft to be deployed in case of changes in mission requirements or to be added to the mission.

[0078] This flexible collaborative mechanism, compared to traditional full-formation schemes, reduces ineffective collaboration and lowers resource redundancy. After completing the collaborative formation generation, the system... Temporary roles assigned to members:

[0079] The drone closest to the mission's core area is automatically designated as the team leader;

[0080] Other member machines are marked as squad members.

[0081] The team leader will be responsible for local sub-region division, collaborative path coordination, and local collaboration strategy design in the subsequent steps (step four).

[0082] This solution perfectly addresses the triple challenges of task requirements, resource availability, and dynamic adaptability in the IoT environment for shared drone systems. It forms a complete closed loop of order acceptance and collaboration structure, providing an efficient, flexible, and scenario-aware collaborative foundation for subsequent collaborative execution and path optimization. It has good feasibility and innovation for implementation.

[0083] S4. Based on the cooperative formation, the respective task areas, and the remaining battery power of the shared UAVs of the formation members as inputs, and with the optimization objectives of minimizing path length, energy consumption, and the overlap of the flight paths of the formation members, a set of conflict-free multi-aircraft cooperative path schemes is generated by combining battery constraints, task time constraints, and cooperative obstacle avoidance constraints.

[0084] Furthermore, in this step, the system receives the cooperative formation output from step three. Designed to coordinate multiple drones j This invention proposes a collaborative path optimization model for tasks performed in the same area, providing structured, low-conflict, and energy-balanced collaborative path solutions for subsequent execution phases. Specifically addressing the characteristics of shared unmanned aerial vehicle (UAV) systems in IoT scenarios, this invention proposes a multi-UAV collaborative flight path-energy consumption joint optimization model to solve common problems in multi-UAV collaborative operations, such as path overlap, flight path congestion, and unbalanced energy consumption distribution. This model is based on collaborative formation. Each person is responsible for their own task area and team members' machine d j Remaining power E j For the input, design the following optimization objective function:

[0085]

[0086] Among them, P j For drones d j The path to be optimized, L j This is the total length of the path. For d j Expected completion Required energy consumption, Φ j P represents j and The overlap index of other member paths (defined as a spatial-temporal coincidence measure of track intersection). κ1, κ2, and κ3 are adjustment factors that can be flexibly set according to station task density or airspace complexity.

[0087] Understandably, the innovation of this invention lies in addressing the frequent flight path interference and high-density airspace conflict problems faced by shared unmanned aerial vehicle (UAV) systems during collaborative operations. The model introduces Φ for the first time... j As a cooperative sparsity regularization term, it uses the intensity of space-time conflict in the track as a constraint to actively guide the avoidance strategy in the cooperative path generation process.

[0088] Furthermore, during the modeling process, path P j Subject to the following constraints:

[0089] Power constraints: Ensure that the estimated energy consumption after path design does not exceed E j The ρ threshold (e.g., ρ = 0.8).

[0090] Task time constraints: All P j Must The task must satisfy the given deadline t. deadline,j .

[0091] Collaborative obstacle avoidance constraints: Path design needs to avoid obstacles and... Φ in other paths j Tracks with conflict intensity exceeding δ.

[0092] For example, if d1, d2 Both are close to 0.8E j And Φ j If the overlap exceeds 30%, the model will automatically adjust the optimization strategy to balance energy consumption and path sparsity under the drive of the objective function.

[0093] The output of this step is the optimized path scheme set. This means determining the optimal mission execution path for each UAV within the coordinated formation, providing execution-level input for mission execution and feedback in step five. Through this path optimization model, the system achieves a closed-loop modeling of route optimization, energy consumption balance, and path sparsity in multi-UAV collaborative tasks, providing an efficient collaborative path strategy for shared UAV scheduling systems and ensuring that subsequent collaborative execution stages have stronger safety, energy efficiency, and mission timeliness.

[0094] S5. For the cooperative formation, execute the path of the path scheme set, collect feedback data from all shared UAVs after execution, and dynamically update the path scheme set through the neighborhood broadcast mechanism in combination with the updated site state vector.

[0095] Specifically, this step constructs a feedback mechanism for the execution results of UAV missions and a dynamic update model for site state vectors and global graphs, ensuring that the shared UAV scheduling system can form a complete closed loop of perception, allocation, coordination, optimization, and feedback.

[0096] After the task is completed, coordinate formation Each drone in the d j Based on the optimized path generated in step four The system returns its execution status to the site in the form of structured feedback data. To achieve a more reasonable map update mechanism, this invention proposes a multi-index feedback vector model, specifying that each d j The following core feedback data needs to be reported: Task completion rate r j (Completion rate per unit task point), path energy consumption (along Actual electricity consumption), path execution fluctuation ξ j (actual track deviation) (Degree of deviation in planning). The above data is summarized as follows: site Local feedback matrix

[0097] Furthermore, the site take over Then, the feedback fusion update model is used to adjust the local state vector. Specifically, based on history The value and the feedback from the new round of drone execution are updated using the following perception update formula:

[0098]

[0099] in, This is the original site's comprehensive perception value. The updated perception value is represented by δ, which is the historical state fusion factor. for The average completion rate of medium tasks The average value of the path execution deviation. This represents the average remaining battery power of each drone. ω1, ω2, and ω3 are the weights of the feedback factors, for example, ω1 = 0.5, ω2 = 0.3, and ω3 = 0.2, reflecting that the system places greater emphasis on the impact of task completion rate on the perceived value.

[0100] Furthermore, the innovation lies in the fact that this perception update model is not simply a task completion count, but for the first time in a shared UAV system, it integrates task completion performance, collaborative execution accuracy, and remaining resource status to form a composite feedback-perception coupled dynamic graph modeling scheme. This scheme can dynamically adapt to factors such as environmental interference and fluctuations in UAV task execution performance within the system, and automatically adjust the site accordingly.

[0101] Ultimately, the site Update the state vector The data is broadcast to neighboring sites and the system graph layer, forming an iterative update of the global graph G′, thereby achieving real-time status updates of the system.

[0102] This step improves the closed-loop design of the shared UAV intelligent scheduling system through a modeling mechanism of multi-indicator feedback, perception fusion, and map update. It enhances the system's real-time adaptability to task execution performance and dynamic environments, provides a more accurate perception basis for a new round of distributed game and collaborative scheduling, and ensures that the system has long-term, efficient, and intelligent operation capabilities in multi-task and dynamic environments.

[0103] This invention also provides an intelligent scheduling system for shared unmanned aerial vehicles (UAVs) based on the Internet of Things (IoT), the system comprising:

[0104] Distributed sensing module: Used to acquire sensing data from each station in actual operation, calculate the comprehensive sensing value of each station, encapsulate it into a station state vector, and synchronize it in real time to neighboring stations and all shared drones within the station through a neighborhood broadcast mechanism; The sensing data includes the number of tasks to be processed, the number of available drones, the average remaining battery power of drones within the station, and the real-time load factor of the station.

[0105] Game allocation module: used to construct a multi-factor game model based on the site state vector to obtain the utility value of the shared drone at the current site; after all the shared drones perform parallel calculations based on the multi-factor game model on all the sites they receive, the site with the highest utility value of the shared drone is selected as the order target and defined as the target site.

[0106] Elastic Formation Module: Used to calculate a coordination index for multiple shared drones that have accepted orders at the same target site, based on the number of tasks to be processed, the number of drones accepting orders, and the comprehensive perception value; when the coordination index exceeds a preset threshold, the shared drones with sufficient power and closest to the core area of ​​the task are selected to form a coordinated formation.

[0107] In the coordinated formation, the shared drone closest to the core mission area serves as the leader, and the other shared drones serve as formation members.

[0108] The path optimization module is used to generate a set of conflict-free path schemes for multi-drone cooperation based on the cooperative formation, the task area each drone is responsible for, and the remaining battery power of the shared drones of the formation members as inputs. The optimization objectives are to minimize the path length, energy consumption, and the overlap of the flight paths of the formation members. It combines battery constraints, task time constraints, and cooperative obstacle avoidance constraints.

[0109] Feedback Update Module: This module executes the path scheme set for the cooperative formation, collects feedback data from all shared UAVs after execution, and dynamically updates the path scheme set through a neighborhood broadcast mechanism in conjunction with the updated site state vector.

[0110] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0111] In the embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection of apparatuses or units may be electrical, mechanical, or other forms.

[0112] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0113] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims

1. A method for intelligent scheduling of shared unmanned aerial vehicles (UAVs) based on the Internet of Things (IoT), characterized in that, The method includes the following steps: S1. Obtain the perception data of each station in actual operation to calculate the comprehensive perception value of each station, encapsulate it into a station state vector, and synchronize it to neighboring stations and all shared drones within the station in real time through a neighborhood broadcast mechanism; wherein, the perception data includes the number of tasks to be processed, the number of available drones, the average remaining power of drones within the station, and the real-time load factor of the station. S2. Construct a multi-factor game model based on the site state vector to obtain the utility value of the shared drone at the current site; after all the shared drones perform parallel calculations based on the multi-factor game model on all the sites they receive, select the site with the highest utility value of the shared drone as the order target, and define it as the target site. S3. For multiple shared drones that have accepted orders at the same target site, calculate the coordination index based on the number of tasks to be processed, the number of drones that have accepted orders, and the comprehensive perception value; when the coordination index exceeds a preset threshold, select the shared drones with sufficient power and closest to the core area of ​​the task to form a coordinated formation. In the coordinated formation, the shared drone closest to the mission core area acts as the leader, and the other shared drones act as formation members; the coordination index is calculated as follows: The station generates the following data based on the number of pending tasks and the number of drones accepting orders at the current station, combined with the comprehensive perception values ​​of each station: ； in, As a synergy index, This represents the number of pending tasks at the current site. The number of drones accepted for orders. This refers to the comprehensive sensing values ​​of the stations in real time. The historical load adjustment coefficient is used to enhance the system's willingness to cooperate under continuous high load; S4, based on the cooperative formation, the task area each is responsible for, and the remaining battery power of the shared UAVs of the formation members as input, with the optimization objective of minimizing path length, energy consumption, and the overlap of the flight paths of the formation members, combined with battery constraints, task time constraints, and cooperative obstacle avoidance constraints, a set of conflict-free multi-aircraft cooperative path schemes is generated. S5. For the cooperative formation, execute the path of the path scheme set, collect feedback data from all shared UAVs after execution, and dynamically update the path scheme set through the neighborhood broadcast mechanism in combination with the updated site state vector. The comprehensive sensing value is obtained by weighted summation of the current number of tasks to be processed, the number of available drones, the average remaining battery power of drones in the site, and the real-time load factor of the site; wherein the real-time load factor is determined based on the current number of tasks to be processed and the number of available drones. The site status vector includes the number of currently pending tasks, the number of available drones, the average remaining battery power of drones within the site, the site's real-time load factor, and the comprehensive perception value. The multi-factor game model is expressed as follows: ； in, For the site The average revenue per task in the pending assignments reflects the task's business priority or economic value. express Current remaining battery power, The comprehensive perceived value is used to reflect... The level of task pressure and resource load, for With the site physical distance, This represents the energy consumption per unit distance. The number should be a small positive number to avoid the denominator approaching zero; , , and These are the adjustment weights for revenue, energy consumption, load, and safety factors set for this invention, respectively. It is the utility value; To establish a safety margin constraint for endurance, this is used to simulate the remaining battery power of a shared drone after completing a mission, preventing the risk of drones failing to return due to remote locations or insufficient battery power.

2. The intelligent scheduling method for shared unmanned aerial vehicles based on the Internet of Things according to claim 1, characterized in that, After all the currently shared drones perform parallel calculations based on a multi-factor game model on all the stations they receive, they select the station with the highest utility value of the shared drones as the order target. If the utility value of the station is less than the utility threshold, the shared drones at the current station choose to remain on standby. This is to ensure that when resources are unevenly distributed in terms of task pressure, some drones can be automatically put on standby to reduce the risk of local overload.

3. The intelligent scheduling method for shared drones based on the Internet of Things according to claim 1, characterized in that, When the coordination index exceeds a preset threshold, shared drones with sufficient power and close to the core mission area are selected to form a coordinated formation. The specific steps include: The design criteria are used to select collaborative formations that actually participate in the collaboration; these criteria include: Battery Priority: Select the shared drone with the highest remaining battery level (m); Location priority: Based on the distance between the drone and the mission center point, the shared drone closest to the core mission area is selected first.

4. The intelligent scheduling method for shared drones based on the Internet of Things according to claim 1, characterized in that, The objective function for the optimization objective is: ； in, For drones The path to be optimized This is the total length of the path. for Expected completion Required energy consumption Indicates the path to be optimized. With coordinated formation The overlap index of other member paths is expressed as a spatial-temporal coincidence measure of track intersections; , , As an adjustment factor, it can be flexibly set according to the site task density or spatial complexity.

5. The intelligent scheduling method for shared drones based on the Internet of Things according to claim 4, characterized in that, The path to be optimized is subject to the following constraints: Power constraint: Ensure that the energy consumption in the path to be optimized does not exceed a preset threshold of the remaining power of the shared drones of the formation members; Task deadline constraints: All paths to be optimized must meet the given task deadline within their respective task areas; Collaborative obstacle avoidance constraint: Avoid tracks where the overlap index between the path to be optimized and other paths in the collaborative formation exceeds a preset threshold.

6. The intelligent scheduling method for shared unmanned aerial vehicles based on the Internet of Things according to claim 1, characterized in that, The feedback data includes task completion rate, path energy consumption, and path execution fluctuation.

7. The intelligent scheduling method for shared unmanned aerial vehicles based on the Internet of Things according to claim 6, characterized in that, S5 specifically includes: Feedback data from shared UAVs executing the path scheme set in the collaborative formation is collected and summarized into a local feedback matrix for the target site; The target station adjusts its station state vector according to the local feedback matrix to generate an updated comprehensive perception value, expressed as: ； in, This is the original site's comprehensive perception value. The updated perception value, As a historical state fusion factor, For coordinated formation The average completion rate of medium tasks The average value of the path execution deviation. This represents the average remaining battery power of all drones. , , The weights of the feedback factors; Finally, the target site broadcasts the updated site state vector to neighboring sites and the system, forming an iterative update of the global graph.

8. A system for implementing the IoT-based intelligent scheduling method for shared unmanned aerial vehicles as described in claim 1, characterized in that, The system includes: Distributed sensing module: Used to acquire sensing data from each station in actual operation, calculate the comprehensive sensing value of each station, encapsulate it into a station state vector, and synchronize it in real time to neighboring stations and all shared drones within the station through a neighborhood broadcast mechanism; The sensing data includes the number of tasks to be processed, the number of available drones, the average remaining battery power of drones within the station, and the real-time load factor of the station. Game allocation module: used to construct a multi-factor game model based on the site state vector to obtain the utility value of the shared drone at the current site; after all the shared drones perform parallel calculations based on the multi-factor game model on all the sites they receive, the site with the highest utility value of the shared drone is selected as the order target and defined as the target site. Elastic Formation Module: Used to calculate a coordination index for multiple shared drones that have accepted orders at the same target site, based on the number of tasks to be processed, the number of drones accepting orders, and the comprehensive perception value; when the coordination index exceeds a preset threshold, the shared drones with sufficient power and closest to the core area of ​​the task are selected to form a coordinated formation. In the coordinated formation, the shared drone closest to the mission core area serves as the leader, and the other shared drones serve as formation members. The path optimization module is used to generate a set of conflict-free multi-drone cooperative path schemes based on the cooperative formation, the task area each drone is responsible for, and the remaining battery power of the shared drones of the formation members as inputs. The optimization objectives are to minimize the path length, energy consumption, and the overlap of the flight paths of the formation members. It combines battery constraints, task time constraints, and cooperative obstacle avoidance constraints. Feedback Update Module: This module executes the path scheme set for the cooperative formation, collects feedback data from all shared UAVs after execution, and dynamically updates the path scheme set through a neighborhood broadcast mechanism in conjunction with the updated site state vector.

Images