Multi-scene-oriented unmanned aerial vehicle operation resource dynamic configuration and scheduling method and system

Abstract

The invention relates to the technical field of data processing, and discloses a multi-scene-oriented unmanned aerial vehicle operation resource dynamic configuration and scheduling method and a multi-scene-oriented unmanned aerial vehicle operation resource dynamic configuration and scheduling system. The method comprises the steps of generating multi-scene task constraint data through semantic analysis; constructing a dynamic situation field matrix based on a Gaussian function; establishing a continuous motion velocity field containing pressure gradient and viscosity diffusion; the vortex velocity components are fused to form a self-adaptive control velocity field; and constructing a task allocation vector field to generate an operation area configuration scheme and converting the operation area configuration scheme into a flight control instruction. According to the method, the core problems that multi-scene adaptability is insufficient and real-time performance and optimality are difficult to balance in the prior art are solved, and the intelligent scheduling capability and the collaborative operation efficiency of the unmanned aerial vehicle group in a complex and changeable environment are improved through an innovative semantic understanding and physical modeling combined method.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and in particular to a method and system for dynamic configuration and scheduling of unmanned aerial vehicle (UAV) operation resources for multiple scenarios. Background Technology

[0002] Existing drone operation resource allocation and scheduling technologies mainly employ rule-based heuristic algorithms and centralized optimization methods. These methods utilize predefined task templates and fixed parameter configurations to achieve task allocation and path planning for drone swarms. Traditional methods are typically designed specifically for single application scenarios, such as agricultural plant protection, logistics delivery, or environmental monitoring. Each scenario has its own independent algorithm architecture and parameter settings, relying on manually preset task priorities and spatial partitioning strategies to guide drone operation behavior.

[0003] Existing technologies have significant shortcomings, primarily manifested in poor scenario adaptability, difficulty in balancing real-time performance and optimality, and limited ability to handle environmental uncertainties. When faced with new application scenarios or complex tasks involving a mix of scenarios, existing methods require algorithm redesign and parameter adjustment, lacking cross-scenario versatility and transferability. Furthermore, traditional discretized path planning and centralized scheduling mechanisms experience a sharp increase in computational complexity when dealing with large-scale UAV swarms and dynamic environmental changes, making it difficult to obtain the optimal solution within a limited timeframe, and lacking robustness to unforeseen events such as communication interruptions and equipment failures. Summary of the Invention

[0004] This application provides a method and system for dynamic configuration and scheduling of UAV operation resources for multiple scenarios, which solves the core problems of insufficient adaptability to multiple scenarios and difficulty in balancing real-time performance and optimality in the existing technology. Through an innovative method that combines semantic understanding and physical modeling, it improves the intelligent scheduling capability and collaborative operation efficiency of UAV swarms in complex and ever-changing environments.

[0005] Firstly, this application provides a method for dynamic configuration and scheduling of UAV operation resources for multiple scenarios, the method comprising:

[0006] Step S1: Perform semantic parsing and constraint formalization processing on the natural language instructions input by the user to generate multi-scenario task constraint data, which includes spatial coordinates, task weights, and scenario types;

[0007] Step S2: Based on the spatial coordinates and task weights in the multi-scenario task constraint data, construct a dynamic potential field matrix that reflects the priority distribution of multi-scenario tasks by combining Gaussian functions;

[0008] Step S3: Using the dynamic potential field matrix as the driving source, establish a continuous motion velocity field for the UAV swarm that includes pressure gradient terms and viscosity diffusion terms;

[0009] Step S4: Calculate the vortex influence coefficient based on the UAV capability assessment value and the potential field strength at the current position, generate local vortex velocity components and fuse them with the continuous motion velocity field to obtain a multi-scenario adaptive control velocity field;

[0010] Step S5: Construct a task allocation vector field based on the multi-scenario adaptive control speed field and scenario type features, generate a UAV operation area configuration scheme through dynamic probability calculation, and convert it into real-time flight control commands to complete closed-loop resource scheduling.

[0011] Secondly, this application provides a dynamic configuration and scheduling system for UAV operation resources oriented towards multiple scenarios, the dynamic configuration and scheduling system for UAV operation resources oriented towards multiple scenarios includes:

[0012] The generation module is used to perform semantic parsing and constraint formalization processing on the natural language instructions input by the user, and generate multi-scenario task constraint data, which includes spatial coordinates, task weights and scenario types.

[0013] The construction module is used to construct a dynamic potential field matrix that reflects the priority distribution of multi-scenario tasks based on the spatial coordinates and task weights in the multi-scenario task constraint data through a combination of Gaussian functions.

[0014] The analysis module is used to establish a continuous velocity field of the UAV swarm, including pressure gradient terms and viscosity diffusion terms, by using the dynamic potential field matrix as the driving source.

[0015] The fusion module is used to calculate the vortex influence coefficient based on the UAV capability assessment value and the potential field strength at the current position, generate local vortex velocity components and fuse them with the continuous motion velocity field to obtain a multi-scenario adaptive control velocity field.

[0016] The conversion module is used to construct a task allocation vector field based on the multi-scenario adaptive control speed field and scenario type features, generate a UAV operation area configuration scheme through dynamic probability calculation, and convert it into real-time flight control commands to complete closed-loop resource scheduling.

[0017] Thirdly, a device for dynamic configuration and scheduling of UAV operation resources for multiple scenarios is provided, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor invokes the instructions in the memory to cause the device for dynamic configuration and scheduling of UAV operation resources for multiple scenarios to execute the aforementioned method for dynamic configuration and scheduling of UAV operation resources for multiple scenarios.

[0018] Fourthly, a computer-readable storage medium is provided, wherein instructions are stored in the computer-readable storage medium, which, when executed on a computer, cause the computer to execute the above-described method for dynamic configuration and scheduling of unmanned aerial vehicle (UAV) operation resources for multiple scenarios.

[0019] The technical solution provided in this application generates multi-scenario task constraint data by semantically parsing and formalizing constraints on natural language instructions. This overcomes the limitations of traditional UAV scheduling methods that rely on preset templates and fixed parameters, enabling a single technical architecture to simultaneously handle the operational needs of various heterogeneous scenarios such as agricultural plant protection, logistics distribution, and emergency rescue. This fundamentally solves the problem of insufficient scenario adaptability in existing technologies. Furthermore, the standardized constraint data format ensures seamless conversion and unified processing of task information across different scenarios. Based on the technical feature of constructing a dynamic potential field matrix through a combination of Gaussian functions using spatial coordinates and task weights, this innovatively introduces continuous mathematical functions into the discrete task allocation problem. This allows complex multi-task priority relationships to be quantitatively described and spatially mapped using a unified mathematical expression. Compared to traditional grid partitioning and weight allocation methods, the dynamic potential field matrix can naturally handle the time-varying characteristics and spatial continuity of task priorities, providing a smooth and differentiable driving source for subsequent motion planning. By using the dynamic potential field matrix as the driving source, a continuous velocity field containing pressure gradient and viscosity diffusion terms is established. Drawing on the Navier-Stokes equations in fluid mechanics, the discrete motion problem of the UAV swarm is transformed into a continuous field theory problem. The pressure gradient term ensures the uniform distribution of the UAV swarm in space to avoid excessive aggregation, while the viscosity diffusion term ensures the continuity and coordination of motion in adjacent areas. This continuous processing method fundamentally changes the traditional point-to-point path planning approach, making the entire UAV swarm exhibit fluid-like coordinated motion characteristics.

[0020] The technical feature of calculating the vortex influence coefficient and generating local vortex velocity components based on UAV capability assessment values ​​and current position potential field strength cleverly introduces a vortex dynamics mechanism to solve the conflict avoidance problem between UAVs. The calculation of the vortex influence coefficient simultaneously considers the individual capability differences of UAVs and their importance in the mission space, enabling UAVs with strong capabilities and located in key positions to gain greater local influence. The local vortex velocity components achieve graceful avoidance by generating a rotational effect. Compared with traditional hard constraints and replanning mechanisms, the vortex mechanism provides a more natural and efficient conflict solution. The technical feature of constructing a task allocation vector field based on multi-scenario adaptive control velocity field and scene type features and generating a work area configuration scheme through dynamic probability calculation transforms the abstract task allocation problem into a specific probability calculation problem. The task allocation vector field comprehensively considers global motion trends and local scene features, and dynamic probability calculation ensures the quantitative comparability and decision transparency of task allocation results. Finally, it is converted into real-time flight control commands to complete closed-loop resource scheduling, forming a complete technical link from high-level semantic understanding to low-level hardware control. In specific multi-scenario UAV operation applications, the introduction of semantic parsing algorithms enables non-professional users to directly describe complex operational requirements using natural language. Gaussian function combination algorithms ensure the reasonable distribution and smooth transition of tasks with different priorities in space. Continuous motion velocity field algorithms, drawing on fluid dynamics theory, realize coordinated control of large-scale UAV swarms. Vortex avoidance algorithms solve the conflict problem in multi-UAV systems through a physics-inspired approach. Probabilistic allocation algorithms guarantee the scientific nature and interpretability of task allocation. The organic combination of these algorithmic features provides a theoretical basis and implementation path for the dynamic configuration and scheduling of UAV operation resources in complex multi-scenario environments. Attached Figure Description

[0021] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a schematic diagram of an embodiment of the dynamic configuration and scheduling method for UAV operation resources for multiple scenarios in this application.

[0023] Figure 2 This is a schematic diagram of an embodiment of a dynamic configuration and scheduling system for unmanned aerial vehicle (UAV) operation resources for multiple scenarios, as described in this application.

[0024] Figure 3 This is a schematic block diagram of the structure of a dynamic configuration and scheduling device for drone operation resources for multiple scenarios in an embodiment of the present invention. Detailed Implementation

[0025] This application provides a method and system for dynamic configuration and scheduling of UAV operational resources for multiple scenarios. The terms first, second, third, fourth, etc. (if present) in the specification, claims, and accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms include or have, and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0026] For ease of understanding, the specific process of the embodiments of this application is described below. Please refer to [link / reference]. Figure 1 One embodiment of the dynamic configuration and scheduling method for UAV operation resources for multiple scenarios in this application includes:

[0027] Step S1: Perform semantic parsing and constraint formalization processing on the natural language instructions input by the user, and obtain multi-scenario task constraint data, which includes spatial coordinates, task weights and scenario types;

[0028] Step S2: Based on the spatial coordinates and task weights in the multi-scenario task constraint data, construct a dynamic potential field matrix that reflects the priority distribution of multi-scenario tasks by combining Gaussian functions;

[0029] Step S3: Using the dynamic potential field matrix as the driving source, establish a continuous velocity field for the UAV swarm that includes pressure gradient terms and viscosity diffusion terms;

[0030] Step S4: Calculate the vortex influence coefficient based on the UAV capability assessment value and the potential field strength at the current position, generate local vortex velocity components and fuse them with the continuous motion velocity field to generate a multi-scenario adaptive control velocity field;

[0031] Step S5: Construct a task allocation vector field based on the multi-scenario adaptive control speed field and scenario type features, generate a UAV operation area configuration scheme through dynamic probability calculation, and convert it into real-time flight control commands to complete closed-loop resource scheduling.

[0032] It is understood that the executing entity of this application can be a dynamic configuration and scheduling system for UAV operation resources for multiple scenarios, or it can be a terminal or a server; the specific implementation is not limited here. This application's embodiment uses a server as an example for illustration.

[0033] Specifically, semantic parsing and constraint formalization first decompose natural language instructions into lexical units through lexical analysis, then perform syntactic parsing to identify grammatical structures and semantic relationships, thereby extracting scene identifiers, task type descriptors, and constraint condition descriptors to obtain a set of semantic elements. The pre-trained domain knowledge base contains standardized parameter templates for various scenarios such as agricultural plant protection, logistics distribution, and emergency rescue. Semantic elements are mapped to specific numerical parameters through matching mapping. Spatial coordinate extraction obtains latitude and longitude values ​​through geographic information parsing. Task weights are calculated as numerical importance coefficients using a priority evaluation algorithm. Scene types are determined by a classification algorithm for scene identifiers to identify specific operational modes. Multi-scenario task constraint data, as a structured output, contains the basic data required for all subsequent computational steps.

[0034] The dynamic potential field matrix is ​​constructed based on the spatial distribution characteristics of the Gaussian function. The single-point potential field distribution is obtained by calculating the Euclidean distance between the task point and each point in space, combined with the exponential decay characteristic of the Gaussian function, to obtain the spatial influence range. The task priority potential field multiplies the single-point potential field distribution by the task weight as the amplitude coefficient, resulting in a weighted spatial intensity distribution. The multi-scenario task priority distribution is obtained by numerically superimposing all task priority potential fields in the same spatial coordinate system, reflecting the competitive and cooperative relationships between different tasks in space. The influence radius parameter is automatically set according to the scenario type: a larger radius is set for agricultural plant protection scenarios to cover continuous operation areas, while a smaller radius is set for logistics and distribution scenarios to accurately locate delivery points. Matrix processing discretizes the continuous spatial distribution into a computable numerical matrix format.

[0035] The continuous velocity field of the UAV swarm is established based on fluid dynamics principles. The task-driven force vector is obtained by performing gradient calculations on the dynamic potential field matrix. The gradient calculation calculates the potential field difference between adjacent grid points and divides it by the spatial distance to obtain the force vector pointing towards the high-priority region. The pressure gradient term is calculated by statistically analyzing the number density of UAVs in a local area. When the density exceeds a threshold, an outward repulsive force is generated to prevent excessive aggregation of UAVs in the same area. The viscosity diffusion term performs Laplace operator operations on the velocity field to eliminate abrupt changes and discontinuities, ensuring smooth velocity transitions between adjacent regions. Solving the dynamic equations involves vector synthesis of the three force components according to the laws of physics to obtain the velocity vector at each spatial point, resulting in a continuous velocity field describing the coordinated movement trend of the UAV swarm throughout the entire operational space.

[0036] Multi-scenario adaptive control velocity field generation is achieved through a fusion algorithm of local vortex mechanism and global velocity field. The UAV capability assessment value comprehensively considers multiple dimensions such as remaining battery power, payload capacity, and sensor accuracy, and obtains the comprehensive capability value through weighted summation. The potential field intensity at the current position is directly obtained from the dynamic potential field matrix according to the UAV's real-time coordinates. The vortex influence coefficient is obtained by multiplying the capability assessment value by the potential field intensity and then normalizing it by dividing by the sum of the capabilities of all UAVs. In the construction of the local vortex field, the distance weighting factor is normalized by dividing the radial distance by the maximum effective distance, and the spatial attenuation coefficient adopts an exponential function to ensure rapid attenuation of long-distance influence. The vortex intensity value is distributed around the UAV to obtain a rotationally symmetric force field distribution. The tangential velocity component is calculated by multiplying the vortex intensity value by a unit tangential vector, and then vector-superimposed with the continuous motion velocity field to obtain a comprehensive control field with both global guidance and local avoidance functions.

[0037] The task allocation vector field is constructed based on a feature vector extraction algorithm that incorporates multi-scenario adaptive control velocity field and scene type features. Scene type features are encoded using predefined feature templates: agricultural plant protection scenarios prioritize area coverage, while logistics and distribution scenarios prioritize point-to-point accuracy. The task type feature vector is multiplied point-by-point with the velocity field values ​​to obtain a spatial distribution reflecting the suitability of task execution. The allocation probability value is calculated by inputting the UAV's current position into the task allocation vector field for numerical lookup, combining it with capability assessment values ​​for weighted processing, and then normalizing the probability using a softmax function. The target operating area for the UAV is determined based on the maximum probability principle. Control command generation converts the area coordinates into a waypoint sequence, and combines this with the scene type to determine flight altitude, velocity parameters, and operational parameters, resulting in a standardized flight control command format.

[0038] In one specific embodiment, step S1 includes:

[0039] Lexical analysis and syntactic parsing are performed on natural language instructions to obtain a set of semantic elements containing scene identifiers, task type descriptors, and constraint condition descriptors;

[0040] The set of semantic elements is input into a pre-trained domain knowledge base for matching and mapping to obtain a standardized combination of task parameters.

[0041] Based on the standardized task parameter combination, spatial coordinates and task weights are extracted to obtain coordinate values ​​and weight values;

[0042] Scenes are classified based on scene identifiers to obtain scene types, which are then combined with spatial coordinates and task weights to obtain multi-scene task constraint data.

[0043] Specifically, the data processing mechanisms for lexical analysis and syntactic parsing are based on the hierarchical architecture of natural language processing. Lexical analysis first segments the continuous input character stream according to lexical boundaries, identifies independent lexical units through maximum matching algorithms and dictionary lookups, and assigns part-of-speech tags such as nouns, verbs, and adjectives to each word. Syntactic parsing constructs a dependency tree based on the lexical analysis results, and determines the grammatical function and semantic role of each component in the sentence through subject-verb-object relationship identification and modification relationship analysis. Scene identifier extraction uses keyword matching algorithms to maintain a dictionary containing feature words of various work scenarios. When identifier words in fields such as agriculture, logistics, and rescue are detected, their position and context information in the sentence are recorded. Task type descriptor extraction focuses on identifying verb phrases expressing specific work behaviors, such as action words like monitoring, spraying, and delivery, and determines their attributes as core actions of the task through verb semantic role labeling. Constraint descriptor extraction obtains restrictive descriptions such as immediate, full coverage, and high precision through the identification of temporal adverbs, spatial prepositions, and quantitative adjectives. The semantic element set organizes and stores these three types of information in a structured format.

[0044] The matching and mapping process of the pre-trained domain knowledge base adopts a strategy combining semantic similarity calculation and template matching. Semantic similarity calculation converts each element in the semantic element set into a word vector representation, and then calculates the cosine distance with the pre-stored standard term vectors in the knowledge base; the smaller the distance value, the higher the semantic similarity. Template matching uses a conditional rule engine to search for predefined parameter configuration templates based on the combination pattern of scene identifiers and task type descriptors. Each template contains standard parameter settings and constraints for that scene. When multiple matching results are found, a confidence-weighted fusion algorithm is used to weight and average the parameters of each matching item according to its matching score to obtain the final standardized task parameter combination. The standardized task parameter combination includes structured data such as spatial range parameters, priority coefficients, job mode identifiers, and resource requirements. Each parameter is labeled with its data type, value range, and unit information.

[0045] Spatial coordinate extraction and task weight extraction are based on parameter parsing and quantification evaluation algorithms. Spatial coordinate extraction identifies geographically relevant parameters from standardized task parameter combinations. When parameters contain specific coordinate values, they are extracted directly. When parameters contain geographic entity names, they are converted to latitude and longitude coordinates using geocoding services. When parameters contain relative location descriptions, absolute coordinates are derived using spatial inference algorithms combined with known location information. Task weight extraction employs a multi-dimensional evaluation algorithm, comprehensively analyzing three dimensions: task time urgency, spatial importance, and resource consumption. Time urgency is quantified by the difference between the task deadline and the current time. Spatial importance is assessed based on the economic and ecological value of the operational area. Resource consumption is calculated using estimated drone working time and energy consumption. The scores of the three dimensions are weighted and summed to obtain the final weight value, which is then normalized to ensure it falls within the range of zero to one.

[0046] Scene classification employs a feature-matching-based decision tree classification algorithm. The decision tree is constructed based on feature combinations of scene identifiers and predefined classification rules. Agricultural scene classification determines the scene type by detecting the presence and combination patterns of agricultural feature words such as farmland, crops, and plant protection. Logistics scene classification identifies the scene type by detecting logistics feature words such as delivery, transportation, and warehousing. Emergency scene classification confirms the scene type by detecting emergency feature words such as rescue, disaster, and emergency. When the input contains multiple scene features, a feature weight voting mechanism determines the dominant scene type. Each feature word is assigned a different voting weight based on its importance within the scene, and the scene type with the highest final vote score is selected as the primary scene. Scene types are represented as enumerated constants, including predefined types such as agricultural operations, logistics delivery, emergency rescue, and environmental monitoring. Each type is associated with specific operational parameters and constraints. Multi-scene task constraint data is obtained through data encapsulation operations, combining spatial coordinates, task weights, and scene types according to a predefined data structure to obtain a composite data object containing complete constraint information.

[0047] When a natural language instruction to monitor pests in a tea garden is received, lexical analysis identifies four lexical units: tea garden (noun), pests (noun), monitoring (verb), and operation (noun). Syntactic parsing determines that tea garden is the operation object, pest monitoring is a composite task type, and operation is the execution instruction. Semantic element extraction yields the scene identifier "tea garden agriculture," the task type descriptor "pest pest monitoring," and the constraint descriptor "periodic inspection." During the pre-trained domain knowledge base matching process, tea garden agriculture shows the highest similarity to the agricultural operation template. Pest pest monitoring matches the parameter configuration of the monitoring task, and periodic inspection matches the time-periodic constraint. Standardized task parameter combinations generate detailed configurations including the boundary coordinates of the tea garden plot, monitoring frequency requirements, and flight path parameters. Spatial coordinate extraction converts the textual description of the tea garden into a specific set of geographic boundary coordinates. Task weights are determined by assessing the importance of pest monitoring's impact on crop yield, resulting in a medium-priority weight value. Scene classification categorizes tea garden agriculture as an agricultural scene type, associating it with parameters such as the standard flight altitude and monitoring sensor configuration for agricultural operations. The final multi-scenario task constraint data integrates the precise geographical range of the tea garden, the priority weight of pest monitoring, and the operation mode identifier of the agricultural scenario, resulting in a standardized input data format for subsequent potential field construction and drone scheduling algorithms.

[0048] In one specific embodiment, step S2 includes:

[0049] Based on the spatial coordinates in the multi-scenario task constraint data, the Gaussian potential field contribution function of each task point is constructed to obtain the potential field distribution of a single point.

[0050] Based on the task weights, the single-point potential field distribution of each task point is weighted according to priority to obtain the task priority potential field.

[0051] Multiple task priority potential fields are superimposed and combined according to their spatial location to obtain a multi-scenario task priority distribution;

[0052] The dynamic potential field matrix is ​​obtained by setting the influence radius parameter based on the scene type and performing matrix processing on the priority distribution of tasks in multiple scenes.

[0053] Specifically, based on the spatial coordinate information in the multi-scenario task constraint data, the coordinates of each task point are used as the center position of the Gaussian function. By calculating the Euclidean distance from any point in space to the task center point, a spatial influence distribution centered on that task point is constructed. The Gaussian potential field contribution function adopts a two-dimensional Gaussian distribution model, where the task point coordinates are used as the mean parameter, and the standard deviation parameter controls the diffusion degree of the influence range. Spatial locations closer to the task center receive higher potential field strength values, while the potential field strength decreases exponentially with increasing distance. The single-point potential field distribution is formed by calculating Gaussian function values ​​point by point on a preset spatial grid. Each grid point corresponds to a potential field strength value, forming a continuous distribution surface with the task point as the peak. The calculation process of the single-point potential field distribution divides the space into a regular grid structure. The grid accuracy determines the spatial resolution of the potential field calculation. Higher grid accuracy can more accurately describe the spatial variation characteristics of the potential field, but at the same time, it increases the computational complexity.

[0054] Priority-weighted processing applies task weights as multiplication coefficients to each numerical point in the single-point potential field distribution. The task weights are directly derived from normalized weight values ​​in multi-scenario task constraint data; a larger weight indicates a higher task priority and a stronger potential field. Weighting is achieved through element-wise matrix multiplication, multiplying the task weights by each element in the single-point potential field distribution matrix to obtain the adjusted potential field intensity distribution. The task priority potential field maintains the spatial shape characteristics of the original single-point potential field distribution, but the overall intensity level is proportionally adjusted according to task importance. Higher-priority tasks have higher potential field peaks and correspondingly stronger potential field intensity within their influence range. Priority-weighted processing solves the problem of existing technologies being unable to distinguish task importance, achieving quantitative expression and spatial mapping of task priority through a numerical weighting mechanism.

[0055] The superposition and combination of multiple task priority potential fields is performed by point-by-point numerical accumulation. When multiple tasks have overlapping influence ranges in space, the potential field strength of the overlapping area is equal to the sum of the values ​​of the potential fields of each task priority at that location. The superposition process traverses all grid points, summing the potential field contributions of all tasks at each location to form a comprehensive multi-scenario task priority distribution. The multi-scenario task priority distribution reflects the comprehensive importance level of different areas in the entire operational space. Areas with dense tasks and high priorities exhibit high potential field strength, while areas with sparse tasks or low priorities exhibit low potential field strength. The spatial competition relationship between tasks is also considered during the superposition process. When multiple high-priority tasks are spatially adjacent, their potential field superposition will form a stronger attraction area, guiding more UAV resources to concentrate in that area.

[0056] The influence radius parameter is set based on the different characteristics of different scenario types. In agricultural scenarios, due to the need for large-scale continuous operations, the influence radius is set to a larger value to cover a wide planting area. In logistics scenarios, which emphasize precise point-to-point delivery, the influence radius is set to a smaller value to achieve accurate positioning. In emergency scenarios, involving emergency rescue, the influence radius is dynamically adjusted according to the disaster area and rescue needs. The influence radius parameter directly affects the diffusion degree of the Gaussian potential field contribution function; a larger influence radius makes the potential field distribution more gradual and widespread, while a smaller influence radius makes the potential field distribution steeper and more concentrated. Matrix processing converts the continuous multi-scenario task priority distribution into a discrete numerical matrix format. The rows and columns of the matrix correspond to the coordinate indices of the spatial grid, and the matrix element values ​​correspond to the potential field intensity at the corresponding grid location. The dynamic potential field matrix, as the final output, contains the potential field distribution information of the entire workspace. The matrix dimension is determined by the precision of the spatial grid division, and the range of element values ​​reflects the distribution characteristics of task priorities.

[0057] When handling multi-scenario operations involving both orchard pest and disease monitoring and logistics delivery, the multi-scenario task constraint data provides the central coordinates and high-priority weights for the orchard monitoring task, and the target coordinates and medium-priority weights for the delivery task. During the construction of the Gaussian potential field contribution function, the orchard monitoring task constructs a Gaussian distribution with its central coordinates as the peak value. The influence radius is set to a relatively large value based on the agricultural scenario type to cover the entire orchard area, forming a broad and gentle single-point potential field distribution. The delivery task constructs another Gaussian distribution with its target coordinates as the center, and the influence radius is set to a smaller value based on the logistics scenario type to achieve precise positioning, forming a concentrated and steep single-point potential field distribution. Priority weighting multiplies the high-weight value of the orchard monitoring task with its single-point potential field distribution to obtain a high-intensity task priority potential field, while the medium-weight value of the delivery task is multiplied with its single-point potential field distribution to obtain a medium-intensity task priority potential field. During the superposition and combination process, the potential fields of the two tasks are numerically accumulated according to their spatial locations. In the orchard area, a high-intensity distribution dominated by the monitoring task is formed, while a medium-intensity distribution dominated by the delivery task is formed at the delivery target point. The potential field intensity in other areas of the space exhibits gradient changes according to the distance decay law. The final generated dynamic potential field matrix integrates the spatial distribution and priority information of the two different scenario tasks.

[0058] In one specific embodiment, step S3 includes:

[0059] The gradient components of the potential field are calculated based on the dynamic potential field matrix to obtain the task driving force vector pointing to the high-priority task region.

[0060] Gradient calculation of the pressure field is performed based on the spatial distribution density of the drone swarm to obtain the pressure gradient term that prevents excessive concentration of drones.

[0061] Based on the requirement of continuous motion of the UAV swarm, the velocity field is subjected to diffusion processing to obtain a viscosity diffusion term that maintains a smooth transition of the velocity field.

[0062] By solving the dynamic equations for the task-driving force vector, pressure gradient term, and viscosity diffusion term, we obtain the continuous velocity field of the UAV swarm, which describes the coordinated motion of the UAV swarm in a multi-scenario operation environment.

[0063] Specifically, spatial derivative operations are performed on the dynamic potential field matrix. Gradient calculation is obtained by calculating the difference in potential field strength between each grid point and its adjacent grid points, and then dividing by the grid spacing to obtain the gradient vector at that location. The task driving force vector is obtained by taking the negative of the potential field gradient, because the gradient points in the direction of the fastest potential field growth, and the driving force needs to point towards the high potential field region, i.e., the high-priority task region. During gradient calculation, for a two-dimensional potential field matrix, the x-direction gradient of each internal grid point is obtained by subtracting the potential field value of the left grid point from the potential field value of the right grid point and then dividing by twice the grid spacing; the y-direction gradient is obtained by subtracting the potential field value of the lower grid point from the potential field value of the upper grid point and then dividing by twice the grid spacing. The gradient calculation of boundary grid points uses forward differencing or backward differencing methods to ensure that all grid points obtain effective gradient values. The magnitude of the task driving force vector reflects the attraction strength of that location towards the high-priority task region, and its direction points towards the nearest high potential field region, forming a force field distribution that guides the UAV to move towards the task center.

[0064] The spatial distribution density calculation of the UAV swarm is based on the real-time location information of all UAVs at the current moment. A kernel density estimation algorithm is used to calculate the UAV aggregation degree at each location on a spatial grid. Kernel density estimation employs a Gaussian kernel function, constructing a Gaussian distribution centered on each UAV location. The density distribution of all UAVs is superimposed to obtain the density distribution of the entire space. The pressure field gradient calculation treats the density distribution as a pressure distribution. The pressure gradient term is obtained by performing gradient operations on the density matrix. The pressure gradient points in the direction of decreasing density, generating a repulsive force away from high-density areas. The calculation process of the pressure gradient term is similar to that of the potential field gradient, but with the opposite sign, because the pressure gradient needs to generate an outward diffusion effect. When the UAV density in a certain area is too high, the pressure gradient vector in that area points outwards, pushing UAVs to diffuse towards lower-density areas, preventing excessive aggregation of UAVs at the same location that could lead to collision risks or resource waste. The strength of the pressure gradient term is proportional to the local UAV density; higher-density areas generate stronger repulsive forces.

[0065] The velocity field diffusion process employs the Laplace operator to smooth the current velocity field. The Laplace operator calculates the velocity difference between each grid point and its neighboring grid points, eliminating abrupt changes and discontinuities in the velocity field through weighted averaging. The viscosity diffusion term is calculated by performing a Laplace operation on each component of the velocity field. The Laplace value of the x-direction velocity component equals the x-direction velocity at that point minus the average of the x-direction velocities of four adjacent points. The y-direction velocity component is processed in the same way. This diffusion process ensures that UAVs in adjacent spatial locations have similar motion trends, avoiding the unreasonable situation of adjacent UAVs moving in completely opposite directions. The strength of the viscosity diffusion term is controlled by the viscosity coefficient. A higher viscosity coefficient results in a stronger diffusion effect and a smoother velocity field, but it also reduces the sensitivity of the velocity field to local changes in the task. The viscosity diffusion term simulates the viscous drag effect in a fluid, giving the UAV swarm motion characteristics similar to the continuity and coordination of a fluid.

[0066] The dynamic equations are solved by vector synthesis of the task-driving force vector, pressure gradient term, and viscosity diffusion term, using a simplified form based on the Navier-Stokes equations for numerical solution. The solution process first involves vector addition of the three force components at each grid point to obtain the resultant force vector at that location. Then, according to Newton's second law, the resultant force is divided by the equivalent mass to obtain the acceleration vector. Finally, the velocity vector is obtained through numerical integration. Numerical integration employs the Euler method or the Runge-Kutta method, calculating the velocity value at the next moment based on the current velocity and acceleration. Boundary conditions need to be set during the solution process; typically, a zero-velocity boundary condition is set at the boundary of the operational area to prevent the UAVs from flying out of the operational range. The continuous velocity field of the UAV swarm, as the solution result, contains velocity vector information for each location in the entire operational space. The direction of the velocity vector indicates the expected direction of movement for the UAV at that location, and the magnitude of the velocity vector indicates the speed of movement.

[0067] When handling multi-scenario operations involving both orchard monitoring and emergency delivery, the dynamic potential field matrix shows that the orchard area has a high potential field strength, while the delivery target point has a moderate potential field strength. Potential field gradient calculation generates a task-driven force vector pointing towards the orchard center in the area surrounding the orchard, and a driving force vector pointing towards the delivery target point around the delivery target point. When multiple drones gather at the orchard entrance, the spatial distribution density of the drone swarm reaches its peak at this location. Density gradient calculation generates a pressure gradient term that diffuses into and around the orchard, propelling some drones to disperse to different areas of the orchard to perform monitoring tasks. Velocity field diffusion processing ensures that adjacent monitoring drones maintain a coordinated cruising speed and flight direction, avoiding mutual interference between rows of fruit trees. Delivery drones, due to the high urgency and clear objective of their tasks, experience a larger task-driven force vector strength, and are less affected by the pressure gradient term, primarily moving along a straight path towards the delivery target. During the solution of the dynamic equations, the vector synthesis of the three force components enables the monitoring drones to form an orderly cruising pattern within the orchard, while the delivery drones form a fast, direct trajectory. The velocity field of the entire drone swarm reflects the coordinated execution and resource optimization characteristics of multi-scenario tasks.

[0068] In one specific embodiment, step S4 includes:

[0069] The vortex influence coefficient of each UAV is obtained by performing a weighted product calculation based on the UAV capability assessment value and the potential field strength at the current location.

[0070] A local vortex field is constructed based on the vortex influence coefficient and vortex influence radius to obtain the rotational speed distribution of each UAV.

[0071] Tangential velocity components are calculated from the rotational velocity distribution to obtain local vortex velocity components that prevent UAV collisions and overlaps.

[0072] By vector superposition and fusion of the local vortex velocity components and the continuous motion velocity field of the UAV swarm, a multi-scenario adaptive control velocity field with both global coordination and local avoidance functions is obtained.

[0073] Specifically, based on a comprehensive calculation of multi-dimensional capability indicators, the capability assessment value is obtained by weighted summation of key parameters such as remaining battery power, payload capacity, sensor accuracy, and flight stability. The remaining battery power value is directly obtained from the UAV's battery management sensor; the payload capacity is calculated based on the ratio of the current onboard equipment weight to the maximum payload; sensor accuracy is determined by equipment specifications; and flight stability is assessed using GPS positioning accuracy and attitude sensor data. The potential field strength at the current location is directly obtained by indexing the dynamic potential field matrix according to the UAV's real-time coordinates. The coordinate index is achieved by converting geographic coordinates into matrix row and column indices. The vortex influence coefficient is normalized by multiplying the UAV's capability assessment value by the current location's potential field strength and then dividing by the sum of the products of all UAV capability assessment values ​​and their corresponding potential field strengths, ensuring that the sum of the vortex influence coefficients of all UAVs equals one. UAVs with high capabilities and located in high potential field strength positions receive larger vortex influence coefficients, indicating stronger influence and priority in the local area. The vortex influence coefficient ranges from zero to one, and its distribution reflects the capability differences and positional advantages within the UAV swarm.

[0074] The local vortex field is constructed based on the vortex influence coefficient and a preset vortex influence radius parameter. The vortex influence radius is determined according to the physical dimensions of the UAV, safety distance requirements, and operational accuracy needs. The local vortex field employs a vortex function model, constructing a rotationally symmetric velocity distribution within the vortex influence radius, centered on the UAV's current position. The vortex intensity is zero at the center, increases with radial distance, reaches a peak at a specific radius, and then decays with further distance. The rotational velocity distribution is obtained by calculating the angular velocity and radial distance of each spatial point within the vortex field. The angular velocity is determined by the vortex influence coefficient and the vortex intensity function, while the radial distance is calculated using the Euclidean distance from that point to the UAV's center. The rotational velocity distribution exhibits a ring-like structure, with lower rotational velocities closer to the UAV's center, higher velocities at medium distances, and gradually decreasing to zero at greater distances. The direction of the rotational velocity distribution follows the right-hand rule, forming clockwise or counterclockwise rotational patterns.

[0075] The calculation of the tangential velocity component converts the rotational velocity distribution into velocity vector components in Cartesian coordinates, with the tangential direction determined by the perpendicular direction of the radial vector. For any point within the vortex field, the radial vector relative to the vortex center is first calculated, then the radial vector is rotated 90 degrees counterclockwise to obtain the tangential unit vector. The magnitude of the tangential velocity is equal to the rotational velocity value at that point. The local vortex velocity component is obtained by multiplying the magnitude of the tangential velocity by the tangential unit vector, forming a two-dimensional velocity vector with x and y components. The calculation of the vortex velocity components ensures that adjacent UAVs exhibit a tendency to orbit each other. When two UAVs are too close, their respective vortex fields interact to produce an avoidance effect. The magnitude of the tangential velocity component is proportional to the vortex influence coefficient; the more capable UAV generates a stronger vortex field, exerting a greater repulsive effect on surrounding UAVs. The vortex velocity component also exhibits distance attenuation characteristics; the influence gradually weakens at locations far from the vortex center, ensuring that the vortex effect only operates within a local range.

[0076] Vector superposition and fusion involves point-by-point vector addition of local vortex velocity components to the continuous velocity field of the UAV swarm. The final velocity at each spatial grid point equals the continuous velocity vector at that point plus the vortex velocity component vector generated by all UAVs at that point. The superposition process requires traversing the vortex fields of all UAVs, calculating the contribution of each vortex field at the target grid point, and then adding it to the corresponding component of the continuous velocity field. The multi-scenario adaptive control velocity field, as the fusion result, possesses both global coordination and local avoidance functions. The global coordination function originates from the response of the continuous velocity field to task priorities, while the local avoidance function originates from the handling of inter-UAV conflicts by the vortex velocity components. The fused velocity field maintains a strong attraction effect in high-priority task areas, while producing an appropriate dispersion effect in densely populated UAV areas and a bypass effect in UAV proximity areas. The numerical distribution of the velocity field reflects the balance between multi-scenario task requirements and UAV swarm coordination constraints.

[0077] In one specific embodiment, a local vortex field is constructed based on the vortex influence coefficient and the vortex influence radius to obtain the rotational speed distribution of each UAV, including:

[0078] The radial distance is normalized based on the vortex influence coefficient to obtain the distance weighting factor;

[0079] The vortex effect range is set according to the vortex influence radius, and the spatial attenuation coefficient is obtained by performing exponential decay calculation on the distance weighting factor.

[0080] The vortex intensity value at each spatial location is obtained by multiplying the vortex influence coefficient and the spatial attenuation coefficient.

[0081] Based on the vortex intensity value, radial and tangential velocity components are decomposed into the surrounding spatial points to obtain the rotational velocity distribution describing the rotational motion pattern around the UAV.

[0082] Specifically, based on the vortex influence coefficient as a standardization benchmark, the actual distances from each spatial point around the UAV to the UAV's center are divided by the vortex influence coefficient for standardization. The radial distance is obtained by calculating the Euclidean distance between the coordinates of a spatial grid point and the UAV's current position. The Euclidean distance is calculated using the distance formula between two points, by adding the squares of the differences in the x and y coordinates and taking the square root. Normalization involves dividing the actual radial distance by the vortex influence coefficient. UAVs with larger vortex influence coefficients have smaller normalized distance values ​​for points at the same actual distance, while UAVs with smaller vortex influence coefficients have larger normalized distance values ​​for points at the same actual distance. The distance weighting factor is calculated as the reciprocal of the normalized distance; points closer to each other have a larger weighting factor, and points farther apart have a smaller weighting factor. The calculation of the distance weighting factor ensures that UAVs with strong capabilities and located in high-potential fields have a stronger influence on surrounding spatial points, reflecting the moderating effect of UAV capability differences on the vortex field distribution.

[0083] The vortex effect range is set based on the vortex influence radius parameter, which is determined comprehensively based on the UAV's safe flight distance, sensor detection range, and operational accuracy requirements. Exponential decay calculation uses an exponential function to attenuate the distance weighting factor. The exponential decay function is constructed in a negative exponential form to ensure that the distance weighting factor decays rapidly with increasing radial distance. The base of the exponential decay is usually chosen as the base of the natural logarithm, and the exponent parameter is determined by the ratio of radial distance to the vortex influence radius. When the radial distance equals the vortex influence radius, the decay coefficient drops to a certain proportion of its original value; when the radial distance exceeds the vortex influence radius, the decay coefficient approaches zero. The spatial decay coefficient is obtained by multiplying the distance weighting factor by the exponential decay function. The decay coefficient maintains a high value near the center of the UAV, decreases sharply at the boundary of the vortex influence radius, and approaches zero outside the boundary. The distribution characteristics of the spatial decay coefficient ensure that the vortex effect only operates within a predetermined range, avoiding unnecessary mutual interference over long distances.

[0084] The vortex intensity calculation involves point-by-point multiplication of the vortex influence coefficient and the spatial attenuation coefficient. This multiplication is performed independently at each spatial grid point. The vortex influence coefficient serves as a global coefficient applied to all grid points within the UAV's influence range, while the spatial attenuation coefficient acts as a position-dependent local coefficient, adjusting the vortex intensity at different locations. The numerical distribution of the vortex intensity values ​​exhibits a ring-shaped structure centered on the UAV. The vortex intensity value at the center is zero. As the radial distance increases, the vortex intensity value first increases and then decreases, reaching a peak at a specific radius, before attenuating to zero with further distance. The peak position of the vortex intensity value is jointly determined by the vortex influence coefficient and the attenuation function parameter, typically located at a certain proportion of the vortex influence radius. The magnitude of the vortex intensity value directly affects the rotational speed at that location; higher intensity values ​​result in faster rotational motion, while locations with zero intensity value produce no rotational effect. The distribution of vortex intensity values ​​at various spatial locations forms a two-dimensional intensity field describing the intensity of rotational motion around the UAV.

[0085] Velocity component decomposition, based on vortex intensity values ​​and spatial geometry, converts the vortex intensity at each spatial point into radial and tangential velocity components. The radial velocity component is calculated as the projection of the vortex intensity value onto the radial direction, defined as the direction from the UAV center towards the target spatial point. This radial velocity component is typically set to zero because vortex motion primarily manifests as tangential rotation rather than radial diffusion or contraction. The tangential velocity component is calculated as the projection of the vortex intensity value onto the tangential direction, defined as a 90-degree counter-clockwise rotation from the radial direction. The magnitude of the tangential velocity is equal to the vortex intensity value at that point. The directions of the tangential velocity components follow the right-hand rule, forming a rotational velocity field around the UAV center. The rotational velocity distribution is formed by combining the velocity components of all spatial points. The velocity distribution exhibits a spiral or ring-shaped streamlined structure, with streamline density reflecting the magnitude of the rotational velocity and streamline direction reflecting the direction of rotational motion. The calculation of the rotational velocity distribution solves the problem of rigid UAV avoidance mechanisms in existing technologies, generating a smooth orbital trajectory through a continuous rotational velocity field.

[0086] In one specific embodiment, step S5 includes:

[0087] Feature vectors are extracted based on the multi-scenario adaptive control speed field and scenario type to obtain task type feature vectors;

[0088] Based on the task type feature vector, the task allocation weight is calculated for each point in the space to obtain the task allocation vector field.

[0089] Input the current position and capability assessment value of the UAV into the task allocation vector field to perform probability matrix operation, and obtain the allocation probability value of each UAV for the corresponding task area;

[0090] Based on the assigned probability value, the region affiliation is determined and control commands are generated to obtain a UAV operation area configuration scheme containing flight path and operation parameters, which is then converted into real-time flight control commands.

[0091] Specifically, based on the numerical distribution characteristics of the multi-scenario adaptive control velocity field and the semantic identifiers of scene types, feature vector extraction employs a combination of principal component analysis and feature encoding. The multi-scenario adaptive control velocity field contains velocity vector information for each grid point in space. Feature extraction first performs statistical analysis on the velocity field, calculating statistics such as the mean, variance, maximum, and minimum velocity magnitudes. Then, it calculates the distribution characteristics of velocity directions, including the main flow direction, divergence degree, and rotation intensity. Scene type feature encoding prioritizes area coverage for agricultural scenes, point-to-point accuracy for logistics scenes, and time sensitivity for emergency scenes. Task type feature vectors are obtained by weighted fusion of velocity field statistical features and scene type encoding. The weighting coefficients are determined based on the importance of velocity field features in different scenarios: agricultural scenes prioritize uniformity of velocity field coverage, logistics scenes prioritize convergence, and emergency scenes prioritize rapid response. The dimensions of the feature vectors include key indicators such as spatial coverage, motion coordination, task responsiveness, and resource utilization, each expressed through specific numerical quantification.

[0092] Task allocation weights are calculated based on the task type feature vector to assess the adaptability of each point in space. The weight calculation employs an algorithm combining distance weighting and feature matching. The task allocation weight for each grid point in space is obtained by calculating the similarity between the point's velocity feature and the task type feature vector. The similarity calculation uses cosine similarity or the reciprocal of Euclidean distance as the metric. Distance weighting considers the physical distance from the spatial point to the task center; points closer to the task center receive higher base weights, and points farther away receive lower base weights. Distance weights are calculated using a Gaussian decay function. Feature matching weights are based on the degree of matching between the point's velocity feature and the task requirements. Matching velocity magnitude with task intensity requirements, velocity direction with task execution direction, and velocity change rate with task dynamism all influence the feature matching weights. The task allocation vector field is obtained by multiplying the distance weights by the feature matching weights. The value of each point in the vector field represents the suitability of that location for performing the corresponding task; higher values ​​indicate a more suitable location for assigning a task to a UAV. The distribution of the task allocation vector field reflects the differences in adaptability of different regions in space to different types of tasks.

[0093] The probability matrix operation takes the UAV's current position and capability assessment value as input parameters, and performs queries and calculations within the task allocation vector field. The UAV's current position is obtained through GPS coordinates, which are converted into a grid index in the task allocation vector field, from which the task suitability value for that position is extracted. The capability assessment value includes comprehensive indicators such as the UAV's remaining battery power, payload capacity, sensor accuracy, and flight stability, and is obtained as a single capability assessment value through weighted summation. The probability calculation uses a softmax function to exponentially transform and normalize the UAV's suitability values ​​in each task region, ensuring that the sum of the allocation probabilities for all task regions equals one. The formula for calculating the allocation probability value is: multiply the suitability value for a task region by the capability assessment value, take the exponent, and then divide by the sum of the exponents of the corresponding values ​​for all task regions. Rows in the probability matrix correspond to different UAVs, columns correspond to different task regions, and matrix elements represent the probability value of a corresponding UAV being assigned to its corresponding task region. The probability matrix operation solves the problem of the lack of quantitative basis for task allocation in existing technologies, expressing the rationality and priority of task allocation through numerical probabilities.

[0094] The assignment of a drone to a specific task area follows the maximum probability principle, selecting the task area with the highest probability value as the target area for each drone. The determination process iterates through each row of the probability matrix, finding the element with the largest value in that row; the corresponding column index represents the drone's assigned task area. Control commands are generated based on the assigned task area coordinates and scenario type parameters, producing a complete command set including waypoint sequences, flight altitude, flight speed, and operational parameters. Waypoint sequences are generated using a path planning algorithm, determining the optimal flight path from the drone's current location to the target task area, considering obstacle avoidance and energy optimization. Flight altitude is automatically set according to the scenario type: a safe altitude above crop canopy for agricultural scenarios, a transport altitude avoiding buildings for logistics scenarios, and a suitable altitude for observation and rescue in emergency scenarios. Operational parameters include sensor configuration, data acquisition frequency, flight mode, and other specific operational settings, matched to the task type and drone capabilities. The drone operational area configuration scheme, as the final output, integrates the task assignment results and execution parameters of all drones. Real-time flight control commands are sent to the drone flight controller via a standardized communication protocol, with command formats conforming to mainstream drone control interface specifications such as MAVSDK or ROS.

[0095] The above describes the method for dynamic configuration and scheduling of UAV operation resources for multiple scenarios in the embodiments of this application. The following describes the system for dynamic configuration and scheduling of UAV operation resources for multiple scenarios in the embodiments of this application. Please refer to [link / reference]. Figure 2 One embodiment of the dynamic configuration and scheduling system for UAV operation resources for multiple scenarios in this application includes:

[0096] The generation module is used to perform semantic parsing and constraint formalization processing on the natural language instructions input by the user, and generate multi-scenario task constraint data, which includes spatial coordinates, task weights and scenario types.

[0097] The construction module is used to construct a dynamic potential field matrix that reflects the priority distribution of multi-scenario tasks based on the spatial coordinates and task weights in the multi-scenario task constraint data through a combination of Gaussian functions.

[0098] The analysis module is used to establish a continuous velocity field of the UAV swarm, including pressure gradient terms and viscosity diffusion terms, by using the dynamic potential field matrix as the driving source.

[0099] The fusion module is used to calculate the vortex influence coefficient based on the UAV capability assessment value and the potential field strength at the current position, generate local vortex velocity components and fuse them with the continuous motion velocity field to obtain a multi-scenario adaptive control velocity field.

[0100] The conversion module is used to construct a task allocation vector field based on the multi-scenario adaptive control speed field and scenario type features, generate a UAV operation area configuration scheme through dynamic probability calculation, and convert it into real-time flight control commands to complete closed-loop resource scheduling.

[0101] above Figure 2 The dynamic configuration and scheduling system for UAV operation resources oriented to multiple scenarios in this embodiment of the invention is described in detail from the perspective of modular functional entities. The dynamic configuration and scheduling device for UAV operation resources oriented to multiple scenarios in this embodiment of the invention is described in detail from the perspective of hardware processing.

[0102] Reference Figure 3 This invention also provides a device for dynamic configuration and scheduling of drone operation resources for multiple scenarios. This device can be a server, and its internal structure can be as follows: Figure 3As shown, this multi-scenario UAV operation resource dynamic configuration and scheduling device includes a processor, memory, display screen, input device, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory of this multi-scenario UAV operation resource dynamic configuration and scheduling device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database of this multi-scenario UAV operation resource dynamic configuration and scheduling device stores the data corresponding to this embodiment. The network interface of this multi-scenario UAV operation resource dynamic configuration and scheduling device is used for communication with external terminals via network connection. When the computer program is executed by the processor, it implements the above-described method.

[0103] Those skilled in the art will understand that Figure 3 The structure shown is merely a block diagram of a portion of the structure related to the present invention and does not constitute a limitation on the dynamic configuration and scheduling equipment for UAV operation resources in multiple scenarios to which the present invention is applied.

[0104] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when the instructions are executed on a computer, cause the computer to perform the steps of the method for dynamic configuration and scheduling of UAV operation resources for multiple scenarios.

[0105] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0106] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part 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 multi-scenario UAV operation resource dynamic configuration and scheduling 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 the present invention. 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.

[0107] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for dynamic configuration and scheduling of unmanned aerial vehicle (UAV) operational resources for multiple scenarios, characterized in that, The method includes: Step S1: Perform semantic parsing and constraint formalization processing on the natural language instructions input by the user to generate multi-scenario task constraint data, which includes spatial coordinates, task weights, and scenario types; Step S2: Based on the spatial coordinates and task weights in the multi-scenario task constraint data, construct a dynamic potential field matrix that reflects the priority distribution of multi-scenario tasks by combining Gaussian functions; Step S3: Using the dynamic potential field matrix as the driving source, establish a continuous motion velocity field for the UAV swarm that includes pressure gradient terms and viscosity diffusion terms; Step S4: Calculate the vortex influence coefficient based on the UAV capability assessment value and the potential field strength at the current position, generate local vortex velocity components and fuse them with the continuous motion velocity field to obtain a multi-scenario adaptive control velocity field; Step S5: Construct a task allocation vector field based on the multi-scenario adaptive control speed field and scenario type features, generate a UAV operation area configuration scheme through dynamic probability calculation, and convert it into real-time flight control commands to complete closed-loop resource scheduling.

2. The method for dynamic configuration and scheduling of UAV operation resources for multiple scenarios as described in claim 1, characterized in that, Step S1 includes: Lexical analysis and syntactic parsing are performed on the natural language instructions to obtain a set of semantic elements containing scene identifiers, task type descriptors, and constraint condition descriptors; The set of semantic elements is input into a pre-trained domain knowledge base for matching and mapping to obtain a standardized combination of task parameters. Based on the standardized task parameter combination, the spatial coordinates and task weights are extracted respectively to obtain coordinate values ​​and weight values; The scene is classified according to the scene identifier to obtain the scene type, and then combined with the spatial coordinates and task weights to obtain the multi-scene task constraint data.

3. The method for dynamic configuration and scheduling of UAV operation resources for multiple scenarios as described in claim 1, characterized in that, Step S2 includes: Based on the spatial coordinates in the multi-scenario task constraint data, a Gaussian potential field contribution function for each task point is constructed to obtain the potential field distribution at a single point. Based on the task weights, the single-point potential field distribution of each task point is weighted according to priority to obtain the task priority potential field. Multiple task priority potential fields are superimposed and combined according to their spatial positions to obtain a multi-scenario task priority distribution; The influence radius parameter is set based on the scenario type, and the priority distribution of the multi-scenario tasks is matrixed to obtain the dynamic potential field matrix.

4. The method for dynamic configuration and scheduling of UAV operation resources for multiple scenarios as described in claim 1, characterized in that, Step S3 includes: The potential field gradient components are calculated based on the dynamic potential field matrix to obtain the task driving force vector pointing to the high-priority task region. The pressure field gradient is calculated based on the spatial distribution density of the drone swarm to obtain the pressure gradient term that prevents excessive concentration of drones. According to the requirement of continuous motion of the UAV swarm, the velocity field is diffused to obtain the viscosity diffusion term that maintains a smooth transition of the velocity field. The task-driving force vector, pressure gradient term, and viscosity diffusion term are solved by dynamic equations to obtain the continuous velocity field of the UAV swarm, which describes the coordinated movement of the UAV swarm in a multi-scenario operation environment.

5. The method for dynamic configuration and scheduling of UAV operation resources for multiple scenarios as described in claim 1, characterized in that, Step S4 includes: The vortex influence coefficient of each UAV is obtained by performing a weighted product calculation based on the UAV capability assessment value and the potential field strength at the current position. Based on the vortex influence coefficient and vortex influence radius, a local vortex field is constructed to obtain the rotational speed distribution of each UAV; Tangential velocity components are calculated from the rotational velocity distribution to obtain the local vortex velocity components that prevent UAV collisions and overlaps. The local vortex velocity component is vector-superimposed and fused with the continuous motion velocity field of the UAV swarm to obtain the multi-scenario adaptive control velocity field that has both global coordination and local avoidance functions.

6. The method for dynamic configuration and scheduling of UAV operation resources for multiple scenarios as described in claim 5, characterized in that, The process of constructing a local vortex field based on the vortex influence coefficient and vortex influence radius to obtain the rotational speed distribution of each UAV includes: The radial distance is normalized based on the vortex influence coefficient to obtain the distance weighting factor; The vortex effect range is set according to the vortex influence radius, and the spatial attenuation coefficient is obtained by performing an exponential decay calculation on the distance weighting factor. The vortex influence coefficient and the spatial attenuation coefficient are multiplied to obtain the vortex intensity value at each spatial location. Based on the vortex intensity value, radial and tangential velocity components are decomposed on the surrounding spatial points to obtain the rotational velocity distribution describing the rotational motion pattern around the UAV.

7. The method for dynamic configuration and scheduling of UAV operation resources for multiple scenarios as described in claim 1, characterized in that, Step S5 includes: Based on the multi-scenario adaptive control velocity field and the scenario type, feature vectors are extracted to obtain task type feature vectors; Based on the task type feature vector, the task allocation weight is calculated for each point in the space to obtain the task allocation vector field. Input the current position and capability assessment value of the UAV into the task allocation vector field to perform probability matrix operation, and obtain the allocation probability value of each UAV for the corresponding task area; Based on the assigned probability value, the region affiliation is determined and control commands are generated to obtain the UAV operation area configuration scheme containing flight path and operation parameters, and then converted into the real-time flight control commands.

8. A dynamic configuration and scheduling system for unmanned aerial vehicle (UAV) operation resources for multiple scenarios, characterized in that, The method for dynamic configuration and scheduling of UAV operation resources for multiple scenarios as described in any one of claims 1-7, wherein the dynamic configuration and scheduling system for UAV operation resources for multiple scenarios comprises: The generation module is used to perform semantic parsing and constraint formalization processing on the natural language instructions input by the user, and generate multi-scenario task constraint data, which includes spatial coordinates, task weights and scenario types. The construction module is used to construct a dynamic potential field matrix that reflects the priority distribution of multi-scenario tasks based on the spatial coordinates and task weights in the multi-scenario task constraint data through a combination of Gaussian functions. The analysis module is used to establish a continuous velocity field of the UAV swarm, including pressure gradient terms and viscosity diffusion terms, by using the dynamic potential field matrix as the driving source. The fusion module is used to calculate the vortex influence coefficient based on the UAV capability assessment value and the potential field strength at the current position, generate local vortex velocity components and fuse them with the continuous motion velocity field to obtain a multi-scenario adaptive control velocity field. The conversion module is used to construct a task allocation vector field based on the multi-scenario adaptive control speed field and scenario type features, generate a UAV operation area configuration scheme through dynamic probability calculation, and convert it into real-time flight control commands to complete closed-loop resource scheduling.

9. A device for dynamic configuration and scheduling of unmanned aerial vehicle (UAV) operational resources for multiple scenarios, characterized in that, The method includes a memory and a processor, wherein the memory stores a computer program that can run on the processor, and the processor executes the computer program to implement the method for dynamic configuration and scheduling of UAV operation resources for multiple scenarios as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is run by the processor, it causes the processor to execute the method for dynamic configuration and scheduling of unmanned aerial vehicle (UAV) operation resources for multiple scenarios as described in any one of claims 1 to 7.

Images