Multi-aircraft cooperative unmanned aerial vehicle scheduling method and system

Abstract

The invention discloses a multi-unmanned aerial vehicle scheduling method and system, and relates to the technical field of unmanned aerial vehicle scheduling, and the method comprises the following steps: obtaining the path response data of an unmanned aerial vehicle task switching process, and recognizing an execution interference region through a clustering algorithm; carrying out feature separation on the interference area and carrying out scheduling evaluation based on a result; forming a scheduling window and extending to obtain a plurality of first scheduling windows according to the evaluated and reconstructed buffer time section; in the windows, deducing an avoidance area for the path intersection section of the unmanned aerial vehicle, and generating a candidate scheduling path set; and inserting tasks through a multi-target matching algorithm to obtain a multi-aircraft task time sequence coordination matrix for unmanned aerial vehicle scheduling. According to the method, the problems of execution interference and conflict caused by path intersection and time sequence overlapping due to task switching in a multi-unmanned aerial vehicle cooperative task are effectively solved, and the scheduling safety and efficiency are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of drone scheduling, and more specifically, to a multi-machine collaborative drone scheduling method and system. Background Art

[0002] With the rapid development of drone technology and the continuous expansion of its application areas, the capabilities of individual drones in mission execution are becoming increasingly apparent. However, faced with large-scale, complex, and diverse mission requirements, single-drone operations can no longer meet the dual requirements of efficiency and safety. Therefore, multi-drone collaborative scheduling has become an important direction for the development of drone systems.

[0003] Multi-drone collaborative scheduling involves key technologies such as path planning, task allocation, time coordination, and conflict avoidance for multiple drones. The goal is to enable efficient, coordinated, and safe joint operations within a limited airspace. This scheduling process not only considers the flight performance and mission requirements of individual drones, but also dynamically responds to factors such as task switching, changing flight environments, communication limitations, and unexpected interference.

[0004] However, when multiple drones collaborate to perform tasks, due to the path intersection and timing overlap during the task switching process, execution interference and conflicts occur between drones, making it difficult to effectively identify and quantify the interference area, which in turn affects the continuity and flexibility of task scheduling, and makes it impossible to achieve safe scheduling and path optimization adjustment for inserting new tasks. Summary of the Invention

[0005] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a multi-machine collaborative drone scheduling method and system, which solves the problem of execution interference and conflict between drones due to path intersection and timing overlap during the task switching process when multiple drones collaborate to perform tasks by identifying the interference area in the task switching and dynamically reconstructing the scheduling window.

[0006] To achieve the above object, the present invention provides the following technical solutions: A multi-machine collaborative UAV scheduling method comprises the following steps: obtaining path response data of a UAV task switching process, and using a clustering algorithm to identify execution interference areas of the UAV task switching process; performing feature separation on the execution interference areas, and performing scheduling evaluation on the UAV tasks based on the feature separation results; reconstructing the buffer time segments in the UAV task scheduling based on the evaluation results to obtain a scheduling window, and extending the time range before and after the scheduling window to obtain a plurality of first scheduling windows; performing avoidance area deduction on a plurality of UAV path intersection segments in the plurality of first scheduling windows to obtain a candidate scheduling path set; and inserting task matching into the candidate scheduling path set based on a multi-target matching algorithm to obtain a multi-machine task timing coordination matrix for use in UAV scheduling.

[0007] In a preferred embodiment, the path response data of the drone task switching process is obtained, and a clustering algorithm is used to identify the execution interference area of ​​the drone task switching process. Specifically, the path response data of the drone task switching process is collected in real time and a spatiotemporal response matrix is ​​constructed, wherein the path response data includes three-dimensional trajectory coordinates, velocity vectors, and heading angle deviations; acceleration mutation points and heading oscillation intervals in the spatiotemporal response matrix are extracted to identify potential interference feature points; a density peak clustering algorithm is used to perform spatial density scanning on the potential interference feature points to obtain a cluster core area; a preset radius is expanded outward with the cluster core area as the center, and the execution interference area is obtained in combination with preset geographic fence constraints.

[0008] In a preferred embodiment, the feature separation of the execution interference area is specifically performed as follows: An execution interference region topology map is constructed, and the Laplace matrix eigenvectors of the execution interference region topology map are calculated using the spectral clustering algorithm. The execution interference region is divided based on the Laplace matrix eigenvectors to obtain several interference core regions. The convex hull boundaries of several interference core regions are extracted to obtain a dynamic separation interface.

[0009] In a preferred embodiment, the scheduling evaluation of UAV tasks based on the feature separation results is specifically as follows: constructing a three-dimensional grid unit within the dynamic separation interface, and using the Poisson distribution algorithm to calculate the probability of UAV spacing conflict between grid units; using the Markov chain algorithm to predict the heading angle conflict state transition path, and constructing a scheduling evaluation matrix based on the UAV spacing conflict probability.

[0010] In a preferred embodiment, the buffer time segment in the UAV task scheduling is reconstructed based on the evaluation results to obtain a scheduling window, and the time range before and after the scheduling window is extended to obtain a plurality of first scheduling windows. Specifically, the maximum tolerable delay in the scheduling evaluation matrix is ​​analyzed to obtain the initial buffer time segment; with the interference core area as the center, the time sequence of the task preparation phase is traced back forward, and the time sequence of the task exit phase is extended backward to obtain the extended time axis; On the extended time axis, the initial buffer time segment is expanded according to a preset ratio to obtain a scheduling window; the overlapping intervals of the scheduling windows of adjacent drones are detected, and the overlapping areas are nonlinearly stretched on the time axis to obtain several first scheduling windows.

[0011] In a preferred embodiment, in several first scheduling windows, avoidance area deduction is performed on the intersection sections of multiple drone paths to obtain a set of candidate scheduling paths. Specifically, for each first scheduling window, the predicted path centerlines of all drones in the window are extracted; the intersection angles of the path centerlines and the spatiotemporal coordinates of the intersection points are calculated to construct a three-dimensional conflict heat map; the gradient mutation area is identified in the conflict heat map, and a dynamic avoidance sphere is generated with the mutation point as the center of the sphere; the dynamic avoidance sphere is deformed along the path centerline according to a preset deformation degree to obtain an ellipsoid; the ellipsoid is expanded into an avoidance channel along the tangent direction to obtain a set of candidate scheduling paths.

[0012] In a preferred embodiment, the candidate scheduling path set is inserted with task matching based on the multi-objective matching algorithm to obtain a multi-machine task timing coordination matrix, specifically: a multi-objective function is constructed, and each path solution of the candidate scheduling path set is converted into a chromosome gene sequence, wherein the chromosome gene sequence includes waypoint sequences and timestamps of multiple drones; the drone waypoint sequences in the chromosome gene sequence are randomly exchanged to obtain several populations; based on the several populations, the improved NSGA-II algorithm is used to solve the multi-objective function to obtain a Pareto optimal solution set; and the Pareto optimal solution set is encoded as a multi-machine task timing coordination matrix.

[0013] In a preferred embodiment, the encoding of the Pareto optimal solution set into a multi-machine task timing coordination matrix is ​​specifically as follows: sorting the Pareto optimal solutions in descending order and selecting several non-dominated solutions; performing time deconvolution on the path nodes of each non-dominated solution to obtain the action sequence of each drone; discretizing the action sequence according to preset time slices, and marking the drone state in each time slice to obtain several marked non-dominated solutions; merging several marked non-dominated solutions to obtain a three-dimensional tensor; and compressing the three-dimensional tensor to obtain the multi-machine task timing coordination matrix.

[0014] The technical effects and advantages of the multi-machine coordinated UAV dispatching method and system of the present invention are as follows: 1. The present invention collects the three-dimensional trajectory, speed and heading deviation information of the UAV in real time during the task switching phase, and uses the density peak clustering algorithm to accurately identify the execution interference area, thereby enhancing the ability to perceive potential conflict points during the task switching process. Secondly, the spectral clustering algorithm is combined to perform feature separation on the interference area, and a dynamic separation interface is constructed to achieve fine division of complex intersection areas and precise positioning of the interference core area, thereby improving the accuracy of conflict assessment. Based on this, the Poisson distribution and Markov chain model are used to predict the probability of UAV spacing conflict and the heading angle state transition, and a scheduling evaluation matrix is ​​constructed to provide data support for the scientific reconstruction of the buffer time segment and the dynamic extension of the scheduling window.

[0015] 2. This invention generates dynamically deformable avoidance channels by deducing the avoidance areas of path intersections within the scheduling window, ensuring safe flight distances between drones. Finally, combined with an improved multi-objective optimization algorithm, it implements task insertion and matching for candidate paths, generates a multi-machine task timing coordination matrix, effectively coordinates drone timing, avoids task conflicts, and achieves real-time dynamic scheduling and conflict avoidance during multi-drone task switching. This not only improves the safety and reliability of task execution, but also optimizes path utilization and scheduling efficiency, significantly reduces flight delays and conflict risks, and meets the complex requirements of large-scale, multi-task drone collaborative operations. It has broad application prospects and practical value. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 The figure is a flow chart of a multi-machine collaborative UAV scheduling method of the present invention.

[0017] Figure 2 This is a structural schematic diagram of a multi-machine collaborative UAV scheduling system of the present invention. DETAILED DESCRIPTION

[0018] The following will provide a clear and complete description of the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0019] Example 1, Figure 1 The present invention provides a multi-machine coordinated UAV scheduling method, which includes the following steps: S1, obtain the path response data of the UAV task switching process and use the clustering algorithm to identify the execution interference area of ​​the UAV task switching process; In this example, the path response data of the UAV task switching process is obtained, and a clustering algorithm is used to identify the execution interference area of ​​the UAV task switching process. Specifically: Real-time collection of path response data during the UAV task switching process and construction of a spatiotemporal response matrix. The path response data includes three-dimensional trajectory coordinates, velocity vectors, and heading angle deviations. Extract acceleration mutation points and heading oscillation intervals in the spatiotemporal response matrix to identify potential interference feature points; The density peak clustering algorithm is used to scan the spatial density of potential interference feature points to obtain the cluster core area; The execution interference area is obtained by expanding the preset radius outward from the cluster core area and combining it with the preset geo-fence constraints.

[0020] It should be noted that by calculating the velocity change rate at consecutive time points, the acceleration greater than the set threshold (such as 2 m / s 2 ) as a mutation point and analyze the continuous change of the heading angle. If there is high-frequency back-and-forth oscillation (for example, the heading angle fluctuation frequency is greater than 2Hz within 5 seconds), it is marked as a heading oscillation interval, and potential interference feature points are obtained. The interference feature points are input into the density peak clustering algorithm for spatial cluster analysis. This method automatically identifies the core points of clusters by calculating the local density of each point and the distance to the point with higher density. The steps are as follows: Calculate the local density ρ of each feature point; For each point, calculate its minimum distance δ to all points with higher density; Points with high ρ and high δ values ​​are selected as cluster centers to automatically form multiple cluster core areas; With the core area of ​​each cluster as the center, a spherical area with a radius of R is expanded outward to preliminarily generate a candidate set of execution interference areas.

[0021] To ensure the spatial rationality and safety of the identified areas, the system further filters them based on pre-set geofencing constraints. For example, if a candidate area overlaps with a no-fly zone, a building, or an area with abnormal ground elevation, the boundaries of that area are removed or adjusted. Ultimately, a set of high-confidence execution interference regions is obtained. Each interference region is described by attributes such as center coordinates, boundary radius, and time window, which are used for subsequent feature separation and scheduling reconstruction.

[0022] S2, feature separation of the execution interference area, and scheduling evaluation of the UAV task based on the feature separation results; In this example, feature separation is performed on the execution interference area, specifically: Constructing an execution interference area topology map and using a spectral clustering algorithm to calculate the Laplacian matrix eigenvectors of the execution interference area topology map; The execution interference region is divided based on the Laplace matrix eigenvector to obtain several interference core regions; The convex hull boundaries of several interference core areas are extracted to obtain the dynamic separation interface.

[0023] It should be noted that first, all interference feature points are extracted from the execution interference area. These feature points usually include locations where the heading angle fluctuates frequently, locations where acceleration changes suddenly, etc., and their spatial positions are marked with three-dimensional coordinates.

[0024] The spatial connections between these feature points are then constructed into a topological graph. Based on the spatial proximity of the feature points, the system assigns a correlation strength to each pair of points to describe their mutual interference relationship. The closer the distance and the more similar the interference variations, the higher the correlation strength.

[0025] Based on the constructed topology map, the interference area is divided using a spectral clustering algorithm. Spectral clustering, a graph-based clustering method, analyzes the spatial distribution of interference points and divides the interference area into several sub-areas. Each sub-area exhibits highly consistent interference characteristics. These sub-areas are referred to as interference core areas. Each core area has a high density of feature points, consistent fluctuation patterns, and strong spatial coupling. In actual deployments, the system dynamically adjusts the number of areas based on the distribution density of interference points, typically setting it to three to five core areas.

[0026] For each interference core region, the system further extracts its boundary contours. This method calculates the minimum bounding box of all feature points in the region, forming a three-dimensional convex hull model. This convex hull boundary effectively describes the core region's external shape and distribution profile. The boundaries of multiple interference core regions collectively form a dynamic separation interface, which serves as the basis for demarcating the internal structure of the interference region.

[0027] In this example, the UAV mission is scheduled and evaluated based on the feature separation results, specifically: A three-dimensional grid cell is constructed within the dynamic separation interface, and the Poisson distribution algorithm is used to calculate the probability of UAV spacing conflict between grid cells. The Markov chain algorithm is used to predict the heading angle conflict state transition path, and a scheduling evaluation matrix is ​​constructed based on the UAV spacing conflict probability.

[0028] It should be noted that within each dynamic separation interface, a three-dimensional evaluation grid with spatial resolution is established. This grid is divided into multiple three-dimensional cells at a set interval, covering the entire interference core area. Each cell represents a local evaluation area, which is used to count and evaluate the probability of UAV interference within the area. Each grid cell records information such as the number of passing drones, flight speed, heading changes, and flight time. Using the relative distance between drones, the probability of co-occurrence, and historical mission data, the system generates statistics to determine the degree of spatial congestion within each grid cell over different time periods. Based on these statistical results, the system probabilistically estimates the probability of a distance conflict between any two drones within the same grid cell. The higher the probability, the more likely the area will become a mission conflict hotspot. Based on known drone flight paths and historical heading data, the system models the heading trends of drones within the core interference area. By analyzing the sequence of heading angle changes, the system predicts the heading state paths of each drone over different time periods. Combining flight speed and trajectory curvature, the system can predict time periods where intersections, convergences, or collisions are likely to occur, thereby inferring the likelihood of heading conflicts within a specific grid cell.

[0029] After obtaining spatial conflict probabilities and heading state transition predictions for each grid cell, the system aggregates these assessment results to construct a task scheduling evaluation matrix. This matrix, based on drone number, records the conflict risk level for each drone within each interference core area. The scheduling evaluation matrix serves as a key reference for subsequent scheduling window construction, buffer time adjustments, and path deduction. It quantifies the interference level and scheduling risk for each drone at a specific time and spatial location.

[0030] S3, based on the evaluation results, reconstructing the buffer time segment in the UAV task scheduling to obtain a scheduling window, and extending the time range before and after the scheduling window to obtain a number of first scheduling windows; In this example, based on the evaluation results, the buffer time segment in the UAV task scheduling is reconstructed to obtain the scheduling window, and the time range before and after the scheduling window is extended to obtain several first scheduling windows, specifically: Analyze the maximum tolerable delay in the scheduling evaluation matrix to obtain the initial buffer time segment; Taking the interference core area as the center, trace back the time sequence of the task preparation phase and extend the time sequence of the task exit phase backward to obtain the extended time axis; On the extended time axis, the initial buffer time segment is expanded according to a preset ratio to obtain a scheduling window; The overlapping intervals of the scheduling windows of adjacent UAVs are detected, and the overlapping areas are nonlinearly stretched on the time axis to obtain several first scheduling windows.

[0031] It should be noted that the maximum allowable time delay for each drone within the core interference area is first extracted from the scheduling evaluation matrix. This delay value is calculated by combining historical mission performance, flight speed, path complexity, and mission type, representing the maximum acceptable delay for the mission without affecting the overall process. Based on this maximum tolerable delay, an initial buffer time segment is set for each mission segment. This serves as a time protection zone during mission execution to absorb time variations caused by local path adjustments.

[0032] Furthermore, with the core interference zone as the center, the timeline traces back to the mission preparation phase to obtain the adjustable time before the mission enters the zone. Simultaneously, the exit phase after mission completion is extended backward to form a complete mission interference impact zone. By analyzing the UAV's mission state transition moments (such as takeoff, steering, cruising, and landing), these traced and extended time periods are integrated into the mission timeline to form an extended time zone, providing greater flexibility for scheduling window generation.

[0033] In the extended timeline, the initial buffer time segment is magnified by a preset ratio. For example, in high-risk areas, a larger magnification factor can be set to increase scheduling redundancy; in lower-risk areas, the original buffer ratio can be retained. The magnified time period is defined as the scheduling window. This window covers the high-risk periods before and after the mission's critical nodes, providing a time tolerance for operations such as path avoidance and intersection prediction.

[0034] Finally, an overlap analysis is performed on all UAV scheduling windows to identify areas where multiple scheduling windows overlap within the same time period. These overlapping areas are high-risk for conflict and require specific management. To avoid mission conflicts, the time axis of these overlapping areas is nonlinearly stretched. This involves differentially adjusting the time spacing of certain time periods to distribute tasks more sparsely within the same physical time period, thereby reducing the probability of overlap. After this stretching process, several first scheduling windows are generated. Each first scheduling window has clear start and end times, information about the UAV mission segments it covers, and redundant time, providing a temporal structure for subsequent path deduction and conflict avoidance.

[0035] S4, in a plurality of first scheduling windows, performing avoidance area deduction for the intersection segments of multiple UAV paths to obtain a set of candidate scheduling paths; In this example, in several first scheduling windows, avoidance area deduction is performed on multiple UAV path intersection segments to obtain a set of candidate scheduling paths, specifically: For each first scheduling window, extract the predicted path center lines of all UAVs within the window; Calculate the intersection angle of the path centerlines and the spatial and temporal coordinates of the intersection points to construct a three-dimensional conflict heat map; Identify the gradient mutation area in the conflict heat map and generate a dynamic avoidance sphere with the mutation point as the center; The dynamic avoidance sphere is deformed according to the preset deformation degree along the center line of the path to obtain an ellipsoid The ellipsoid is expanded into an avoidance channel along the tangent direction to obtain a set of candidate scheduling paths.

[0036] It should be noted that for each first scheduling window, the predicted flight paths of all participating drones within that time period are extracted. Each path is composed of multiple consecutive flight points. Through fitting analysis, the spatial center trajectory is extracted as the path centerline. The path centerline represents the drone's primary flight direction and form within a specific time period and serves as the basis for subsequent intersection analysis and conflict assessment.

[0037] Furthermore, spatial analysis is performed on each pair of path centerlines to identify their intersection points and intersection angles. If there is spatial overlap between paths and the intersection angle is less than a set threshold, it is identified as a high-risk intersection segment. For each high-risk intersection, its corresponding time coordinates are extracted to create a time- and space-based intersection dataset for subsequent dynamic avoidance zone construction.

[0038] A three-dimensional conflict heat map is created within each scheduling window by combining the intersection dataset. This map assesses the conflict risk level at each location in the space based on factors such as path density, intersection angle, and time overlap, and is annotated with color depth or numerical value. Heat maps are used to identify locations and areas where risk changes significantly in the local space, providing intuitive avoidance reference information. Areas with significant changes in risk values, i.e., risk gradient mutation points, are detected in the heat map. A three-dimensional spherical avoidance zone is constructed with these mutation points as the center. The size of the sphere is determined by the drone's speed, intersection density, and scheduling window width. The avoidance sphere can move dynamically over time, representing the path conflict areas that require priority avoidance.

[0039] The dynamic avoidance spheres are deformed according to the path centerline, elongating them appropriately along the flight path to form an ellipsoidal structure, enhancing the path's maneuverability. The system then extends multiple ellipsoids tangentially along the path to create a more connected and circumventable avoidance channel. This channel is flexible and adaptable to varying mission speeds and headings.

[0040] Ultimately, the system constructs multiple avoidance paths based on the original path and the spatial structure of the avoidance corridor. Each path avoids key intersection risk points and is feasible within the scheduling window. All path plans are aggregated to form a candidate scheduling path set, providing path support for subsequent task insertion and multi-machine coordination.

[0041] S5, based on the multi-objective matching algorithm, insert task matching into the candidate scheduling path set to obtain the multi-machine task timing coordination matrix and use it for UAV scheduling.

[0042] In this example, the candidate scheduling path set is matched with inserted tasks based on the multi-objective matching algorithm to obtain the multi-machine task timing coordination matrix, which is specifically: A multi-objective function is constructed, and each path solution of the candidate scheduling path set is converted into a chromosome gene sequence, wherein the chromosome gene sequence includes waypoint sequences and timestamps of multiple UAVs; Randomly exchange the UAV waypoint sequences in the chromosome gene sequence to obtain several populations; Based on several populations, the improved NSGA-II algorithm is used to solve the multi-objective function and obtain the Pareto optimal solution set; The Pareto optimal solution set is encoded as a multi-machine task timing coordination matrix.

[0043] It should be noted that the multi-objective function includes minimizing total mission duration, minimizing path intersection conflicts, minimizing inter-task waiting time, maximizing available scheduling window utilization, and minimizing flight energy consumption. Furthermore, for each path in the candidate scheduling path set, the system serializes it into a representation containing the following: a sequence of path points for each drone, the time of each path point, and information about the mission segments included in each path.

[0044] Furthermore, by shuffling and cross-rearranging the tasks in multiple scheduling paths, multiple initial matching solutions are formed, called populations. Each population represents a possible UAV task sequencing solution with different path combinations and execution time structures.

[0045] In this example, based on several populations, the improved NSGA-II algorithm is used to solve the multi-objective function and obtain the Pareto optimal solution set, which is: Evaluate the fitness of each population solution under multiple objective functions; Compare the non-dominance of each solution and select the optimal solution; Perform crossover and mutation processing on the optimal solution to generate a new solution; Repeat the evaluation and screening process and gradually converge to the optimal matching set.

[0046] In this example, the Pareto optimal solution set is encoded as a multi-machine task timing coordination matrix, specifically: Sort the Pareto optimal solutions in descending order and select several non-dominated solutions; Perform temporal deconvolution on the path nodes of each non-dominated solution to obtain the action sequence of each drone; Discretize the action sequence into preset time slices, mark the drone state in each time slice, and obtain several marked non-dominated solutions; Merge several annotated non-dominated solutions to obtain a three-dimensional tensor; The three-dimensional tensor is compressed to obtain the multi-machine task timing coordination matrix.

[0047] It should be noted that the set of non-dominated solutions is first ranked by comprehensive scheduling objective. Ranking metrics include total task duration, path conflict intensity, scheduling window utilization, and energy consumption evaluation. Based on the ranking results, several optimal non-dominated solutions are selected as representative scheduling schemes to ensure coverage of various scheduling preferences, such as speed priority, safety priority, or energy conservation priority.

[0048] In addition, for each selected non-dominated solution, the system extracts its corresponding multi-UAV path node sequence, including the time information, spatial position and mission status of the flight points.

[0049] The system uses a time-reversal algorithm to restore each drone's complete mission action sequence, point by point, from the final mission objective forward. This sequence includes: startup phase; takeoff process; mid-stage navigation; path avoidance; mission execution; return path; landing and exit. The action sequence is divided into equally spaced time slices at a uniform time resolution, for example, each slice is five time units. In each time slice, the drone's state is annotated, such as "waiting," "moving," "mission execution," "avoidance," or "idle." The annotated time slices of all drones form multiple separate two-dimensional matrices, each corresponding to a set of non-dominated solutions.

[0050] Subsequently, multiple labeling matrices are coded and integrated, and they are organized into a three-dimensional structure according to the mission number, UAV number, and time slice number, which is expressed as follows: the first dimension is the UAV number, the second dimension is the time slice number, The third dimension is the action state number.

[0051] Finally, to reduce scheduling control costs and improve real-time computing efficiency, the system compresses the data of the three-dimensional structure. The compression method includes: identifying state redundancy within adjacent time slices; merging action segments; eliminating invalid segments; and encoding and compressing state segments.

[0052] Example 2, Figure 2 The present invention provides a multi-machine collaborative UAV scheduling system, which includes an interference identification module, a scheduling evaluation module, a scheduling reconstruction module, a path deduction module, and a task coordination module: The interference identification module is used to obtain the path response data of the UAV task switching process and use the clustering algorithm to identify the execution interference area of ​​the UAV task switching process; The scheduling evaluation module is used to perform feature separation on the execution interference area and perform scheduling evaluation on the UAV mission based on the feature separation results; A scheduling reconstruction module is used to reconstruct the buffer time segment in the UAV task scheduling according to the evaluation results to obtain a scheduling window, and extend the time range before and after the scheduling window to obtain a plurality of first scheduling windows; A path deduction module is used to perform avoidance area deduction on multiple UAV path intersection segments in a plurality of first scheduling windows to obtain a candidate scheduling path set; The task coordination module is used to insert task matching into the candidate scheduling path set based on the multi-objective matching algorithm, obtain the multi-machine task timing coordination matrix and use it for UAV scheduling.

[0053] The above formulas are all dimensionless and numerical calculations. The formulas are obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters in the formulas are set by technicians in this field according to actual conditions.

[0054] The above embodiments may be implemented in whole or in part through software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments may be implemented in whole or in part in the form of a computer program product.

[0055] Those skilled in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0056] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0057] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0058] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A multi-machine coordinated UAV scheduling method, characterized in that: The following steps are involved: Obtain the path response data of the UAV task switching process and use clustering algorithm to identify the execution interference area of ​​the UAV task switching process; Perform feature separation on the execution interference area and conduct scheduling evaluation on the UAV mission based on the feature separation results; According to the evaluation results, the buffer time segment in the UAV task scheduling is reconstructed to obtain the scheduling window, and the time range before and after the scheduling window is extended to obtain several first scheduling windows; In several first scheduling windows, avoidance area deduction is performed on multiple UAV path intersection segments to obtain a candidate scheduling path set; Based on the multi-objective matching algorithm, task matching is performed on the candidate scheduling path set to obtain the multi-machine task timing coordination matrix and use it for UAV scheduling.

2. The multi-machine coordinated UAV scheduling method according to claim 1 is characterized in that: The path response data of the UAV task switching process is obtained, and a clustering algorithm is used to identify the execution interference area of ​​the UAV task switching process, specifically: Real-time collection of path response data during the UAV task switching process and construction of a spatiotemporal response matrix. The path response data includes three-dimensional trajectory coordinates, velocity vectors, and heading angle deviations. Extract acceleration mutation points and heading oscillation intervals in the spatiotemporal response matrix to identify potential interference feature points; The density peak clustering algorithm is used to scan the spatial density of potential interference feature points to obtain the cluster core area; The execution interference area is obtained by expanding the preset radius outward from the cluster core area and combining it with the preset geo-fence constraints.

3. The multi-machine coordinated UAV scheduling method according to claim 2 is characterized in that: The feature separation of the execution interference area is specifically as follows: Constructing an execution interference area topology map and using a spectral clustering algorithm to calculate the Laplacian matrix eigenvectors of the execution interference area topology map; The execution interference region is divided based on the Laplace matrix eigenvector to obtain several interference core regions; The convex hull boundaries of several interference core areas are extracted to obtain the dynamic separation interface.

4. The multi-machine coordinated UAV scheduling method according to claim 3 is characterized in that: The scheduling evaluation of the UAV mission is performed based on the feature separation results, specifically: A three-dimensional grid cell is constructed within the dynamic separation interface, and the Poisson distribution algorithm is used to calculate the probability of UAV spacing conflict between grid cells. The Markov chain algorithm is used to predict the heading angle conflict state transition path, and a scheduling evaluation matrix is ​​constructed based on the UAV spacing conflict probability.

5. The multi-machine coordinated UAV scheduling method according to claim 4 is characterized in that: According to the evaluation results, the buffer time segment in the UAV task scheduling is reconstructed to obtain a scheduling window, and the time range before and after the scheduling window is extended to obtain several first scheduling windows, specifically: Analyze the maximum tolerable delay in the scheduling evaluation matrix to obtain the initial buffer time segment; Taking the interference core area as the center, trace back the time sequence of the task preparation phase and extend the time sequence of the task exit phase backward to obtain the extended time axis; On the extended time axis, the initial buffer time segment is expanded according to a preset ratio to obtain a scheduling window; The overlapping intervals of the scheduling windows of adjacent UAVs are detected, and the overlapping areas are nonlinearly stretched on the time axis to obtain several first scheduling windows.

6. The multi-machine coordinated UAV scheduling method according to claim 5, characterized in that: In the first scheduling windows, avoidance area deduction is performed on the intersection sections of multiple UAV paths to obtain a candidate scheduling path set, specifically: For each first scheduling window, extract the predicted path center lines of all UAVs within the window; Calculate the intersection angle of the path centerlines and the spatial and temporal coordinates of the intersection points to construct a three-dimensional conflict heat map; Identify the gradient mutation area in the conflict heat map and generate a dynamic avoidance sphere with the mutation point as the center; The dynamic avoidance sphere is deformed according to a preset deformation degree along the center line of the path to obtain an ellipsoid; The ellipsoid is expanded into an avoidance channel along the tangent direction to obtain a set of candidate scheduling paths.

7. The multi-machine coordinated UAV dispatching method according to claim 6, characterized in that: The candidate scheduling path set is inserted into task matching based on the multi-objective matching algorithm to obtain a multi-machine task timing coordination matrix, which is specifically: A multi-objective function is constructed, and each path solution of the candidate scheduling path set is converted into a chromosome gene sequence, wherein the chromosome gene sequence includes waypoint sequences and timestamps of multiple UAVs; Randomly exchange the UAV waypoint sequences in the chromosome gene sequence to obtain several populations; Based on several populations, the improved NSGA-II algorithm is used to solve the multi-objective function and obtain the Pareto optimal solution set; The Pareto optimal solution set is encoded as a multi-machine task timing coordination matrix.

8. The multi-machine coordinated UAV dispatching method according to claim 7, characterized in that: The Pareto optimal solution set is encoded into a multi-machine task timing coordination matrix, specifically: Sort the Pareto optimal solutions in descending order and select several non-dominated solutions; Perform temporal deconvolution on the path nodes of each non-dominated solution to obtain the action sequence of each drone; Discretize the action sequence into preset time slices, mark the drone state in each time slice, and obtain several marked non-dominated solutions; Merge several annotated non-dominated solutions to obtain a three-dimensional tensor; The three-dimensional tensor is compressed to obtain the multi-machine task timing coordination matrix.

9. A multi-machine coordinated UAV dispatching system, applied to a multi-machine coordinated UAV dispatching method according to any one of claims 1 to 8, characterized in that: It includes interference identification module, scheduling evaluation module, scheduling reconstruction module, path deduction module and task coordination module: The interference identification module is used to obtain the path response data of the UAV task switching process and use the clustering algorithm to identify the execution interference area of ​​the UAV task switching process; The scheduling evaluation module is used to perform feature separation on the execution interference area and perform scheduling evaluation on the UAV mission based on the feature separation results; A scheduling reconstruction module is used to reconstruct the buffer time segment in the UAV task scheduling according to the evaluation results to obtain a scheduling window, and extend the time range before and after the scheduling window to obtain a plurality of first scheduling windows; A path deduction module is used to perform avoidance area deduction on multiple UAV path intersection segments in a plurality of first scheduling windows to obtain a candidate scheduling path set; The task coordination module is used to insert task matching into the candidate scheduling path set based on the multi-objective matching algorithm, obtain the multi-machine task timing coordination matrix and use it for UAV scheduling.

Images