Task-driven unmanned aerial vehicle reconnaissance path planning modeling method

Abstract

The invention discloses a task-driven unmanned aerial vehicle reconnaissance path planning modeling method, which is characterized in that a simulation modeling method research of unmanned aerial vehicle path planning is carried out around an unmanned aerial vehicle reconnaissance task, a staged path simulation modeling framework is provided, and a complete reconnaissance task process is divided into four stages of takeoff, entering, reconnaissance and recovery. Simulation input parameters and calculation models of all stages are clearly defined. In a simulation environment, combining real-time attitude information of the unmanned aerial vehicle to realize dynamic prediction of a takeoff stage end point; constructing an automatic generation algorithm of a region and point reconnaissance path by adopting a configurable path template defined by XML (Extensible Markup Language) and an affine transformation method; through a line segment offset and corner smoothing processing algorithm, automatic generation of a complex linear path is realized. Finally, the effectiveness and practicability of the simulation modeling method are verified through multiple groups of experimental cases, and a complete solution is provided for systematic simulation and engineering application of unmanned aerial vehicle path planning.

Description

Technical Field

[0001] This invention relates to a simulation modeling method, and more particularly to a task-driven UAV reconnaissance path planning modeling method. Background Technology

[0002] System-on-systems simulation is a core tool in modern military fields for tactical verification, equipment evaluation, and command decision-making. It simulates the dynamic game between opposing forces in multi-domain space by constructing highly realistic virtual battlefield environments. In this context, unmanned aerial vehicles (UAVs), as an important reconnaissance tool, have been widely applied in all aspects of military operations. The quality of their path planning directly determines the efficiency and completeness of battlefield intelligence gathering, thus affecting the credibility of the entire simulation results. Therefore, developing a UAV path planning method that can quickly respond to operational scenarios, flexibly adapt to complex constraints, and seamlessly integrate into the simulation engine has become an urgent need to improve the effectiveness of system-on-systems simulation.

[0003] Currently, research on UAV path planning for system-on-system confrontation simulation faces challenges from both algorithmic and engineering perspectives. At the algorithmic level, scholars both domestically and internationally have conducted extensive research, proposing a series of optimization methods such as particle swarm optimization, artificial bee colony optimization, genetic algorithms, dung beetle algorithms, directed search tree algorithms, A* algorithms, and improved potential field methods, aiming to solve optimal or suboptimal path finding problems under specific constraints. However, most of these methods focus on single-stage or single-type path generation, generally lacking an integrated modeling framework for the entire "takeoff-entry-reconnaissance-recovery" process of reconnaissance missions. This makes it difficult to directly embed the generated paths into simulation scenarios, requiring significant manual intervention for segmentation, splicing, and correction, severely limiting scenario editing efficiency. At the engineering application level, scholars both domestically and internationally currently focus on system-level simulation technology research, tending to promote the overall development of system-on-system confrontation simulation technology from a top-level design perspective. However, there is still room for further deepening in translating theoretical frameworks into concrete practices, easily neglecting compatibility considerations for user-input UAV reconnaissance tactical parameters, resulting in certain limitations in the flexibility of the planned paths, making it difficult to fully adapt to the high requirements for UAV autonomy in system-on-system confrontation simulations. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention discloses a task-driven UAV reconnaissance path planning method, the technical solution of which is as follows:

[0005] A mission-driven UAV reconnaissance path planning method, characterized by comprising the following steps:

[0006] Step S1: Construct a phased path planning modeling framework, dividing the UAV reconnaissance mission into the take-off phase, entry phase, reconnaissance phase and recovery phase, and defining the input parameters and calculation models for each phase, wherein the input parameters include reconnaissance type, geographic information and motion constraints;

[0007] Step S2: Based on the reconnaissance type, generate the reconnaissance phase path using the corresponding path generation algorithm:

[0008] When the reconnaissance type is area reconnaissance or point reconnaissance, a path generation algorithm based on XML templates and affine transformations is used.

[0009] When the reconnaissance type is line reconnaissance, a path generation algorithm based on geometry calculation is used;

[0010] Step S3: Based on the connection relationship of each stage path, integrate the paths of the takeoff stage, entry stage, reconnaissance stage and recovery stage to form a complete mission path.

[0011] Preferably, the generation of the takeoff phase path includes: if the user specifies an entry path, the takeoff phase endpoint is the starting point of the entry path; if the user does not specify an entry path but provides a takeoff extension distance, the takeoff phase endpoint is calculated based on the UAV's real-time attitude information, wherein the real-time attitude information includes yaw angle and pitch angle.

[0012] Preferably, the path generation algorithm based on XML templates and affine transformations includes: predefining a normalized path template and storing it in XML format, wherein the template includes circular, figure-eight, or W-shaped paths; calculating the minimum bounding box (MBB) of the target region and obtaining its center point, width, height, and rotation angle; performing an affine transformation on the normalized path template using the MBB, including proportional scaling, rotation, and translation; and optimizing the sequence of the transformed path to connect to the end point of the stage.

[0013] Preferably, the line reconnaissance path generation algorithm based on geometric calculation includes: shifting the user-input linear path by forward and reverse line segments to generate shifted line segments; smoothing the corners of the shifted line segments and generating a set of inflection points by solving for the intersection points of the line segments or the intersection points of the extended lines; merging the forward shifted path, the reverse shifted path, and the set of inflection points to remove duplicates; and performing three-dimensional coordinate transformation and sequence optimization on the deduplicated point set to form a reconnaissance path.

[0014] This invention also discloses a mission-driven UAV reconnaissance path planning device, characterized in that it comprises:

[0015] The phased path modeling module is used to build a phased path planning modeling framework, which divides the UAV reconnaissance mission into the take-off phase, entry phase, reconnaissance phase and recovery phase, and defines the input parameters and calculation models for each phase.

[0016] The reconnaissance path generation module is used to generate reconnaissance phase paths based on the reconnaissance type, including:

[0017] The region / point reconnaissance path generation submodule is used to execute a path generation algorithm based on XML templates and affine transformations;

[0018] The line reconnaissance path generation submodule is used to execute a path generation algorithm based on geometry calculations;

[0019] The path integration module is used to integrate the paths of each stage to form a complete task path.

[0020] Preferably, the phased path modeling module includes a takeoff phase processing unit, used to: if the user specifies an entry path, set the takeoff phase endpoint as the starting point of the entry path; if the user does not specify an entry path but provides a takeoff extension distance, calculate the takeoff phase endpoint based on the UAV's real-time attitude information.

[0021] Preferably, the region / point reconnaissance path generation submodule includes: a template storage unit for storing predefined normalized XML path templates; a minimum bounding box (MBB) calculation unit for calculating the minimum bounding box (MBB) of the target region; an affine transformation unit for performing affine transformations on the template using the MBB; and a path optimization unit for optimizing the sequence of the transformed path.

[0022] Preferably, the affine transformation unit is configured to: scale and rotate the point coordinates in the path template proportionally, based on the width, height and rotation angle of the MBB; translate the scaled and rotated points to the center of the MBB to obtain two-dimensional coordinates; and add height information to obtain three-dimensional coordinates.

[0023] Preferably, the line reconnaissance path generation submodule includes: a line segment offset unit for performing forward and reverse line segment offset on the user-input linear path; a corner processing unit for smoothing the corners of the offset line segments and generating a set of inflection points; a path merging unit for merging the forward offset path, the reverse offset path, and the set of inflection points, and removing duplicates; and a coordinate transformation unit for performing three-dimensional coordinate transformation and sequence optimization.

[0024] The present invention also discloses a non-volatile storage medium, characterized in that the non-volatile storage medium includes a stored program, wherein the program, when running, controls the device where the non-volatile storage medium is located to execute the above-described method.

[0025] The present invention also discloses a terminal device, characterized in that the terminal device includes: a processor, a memory, a communication interface, and a bus; the processor, the memory, and the communication interface are connected through the bus and communicate with each other; the memory stores executable program code; the processor reads the executable program code stored in the memory to run a program corresponding to the executable program code, so as to execute the method described above.

[0026] Beneficial effects

[0027] (1) By combining real-time attitude data of the UAV, dynamic prediction of the endpoint of the takeoff phase was achieved;

[0028] (2) By using XML templates and affine transformation methods, flexible and automatic generation of reconnaissance paths for regions and points was achieved;

[0029] (3) Through line segment offset and corner processing algorithms, the smooth construction and automatic generation of line reconnaissance paths were realized;

[0030] (4) It can effectively generate UAV paths that meet the actual task requirements, and has good adaptability and engineering practical value;

[0031] (5) It not only improves the automation level of path planning, but also lays the foundation for further research on multi-UAV collaboration and dynamic environment adaptation. Attached Figure Description

[0032] Figure 1 Flowchart of path planning algorithms for regional reconnaissance and point reconnaissance;

[0033] Figure 2 This is a schematic diagram of the path template;

[0034] Figure 3. Flowchart of the line reconnaissance path planning algorithm;

[0035] Figure 4. Schematic diagram of line segment offset model;

[0036] Figure 5. Schematic diagram of the intersection of the extension lines of the line segments;

[0037] Figure 6. Schematic diagram of two line segments directly intersecting;

[0038] Figure 7 A schematic diagram illustrating waypoint prediction during the takeoff phase of an unmanned aerial vehicle (UAV).

[0039] Figure 8 Diagram of point reconnaissance;

[0040] Figure 9 is a schematic diagram of a W-shaped reconnaissance path (example area 1);

[0041] Figure 10 is a schematic diagram of a W-shaped reconnaissance path (example area 2);

[0042] Figure 11 is a schematic diagram of the figure-eight reconnaissance path (example area 1);

[0043] Figure 12 is a schematic diagram of the figure-eight reconnaissance path (example area 2);

[0044] Figure 13 is a schematic diagram of the reconnaissance route (Example Route 1);

[0045] Figure 14 is a schematic diagram of the reconnaissance route (Example Route 2). Detailed Implementation

[0046] To address the aforementioned challenges, this invention focuses on the practical needs of UAV reconnaissance missions in system-on-systems confrontation simulations, and conducts research on mission-driven UAV reconnaissance path planning algorithms. First, a phased path planning modeling framework is proposed, dividing the complete UAV reconnaissance mission process into four stages: takeoff, entry, reconnaissance, and recovery. The simulation input parameters, connection relationships, and computational models for each stage are clearly defined, laying a solid foundation for the automated generation of the entire process path. Second, a region / point reconnaissance path planning algorithm based on XML templates and affine transformations is designed. By predefined configurable path templates and combining them with the minimum bounding box of the target region for affine transformations, efficient and flexible generation of various standard reconnaissance modes is achieved. Third, a line reconnaissance path planning algorithm based on geometric calculations is developed. Through forward / reverse line segment offsets and corner processing, the challenges of path smoothness and continuity when reconnaissance along complex linear targets are solved, significantly improving the engineering usability of the path.

[0047] I. Path Planning Algorithm Modeling

[0048] All calculations in this invention are based on the ENU (East(x)-North(y)-Sky(z)) coordinate system.

[0049] (I) Task-driven phased path planning modeling framework

[0050] In UAV mission-level simulation scenarios, reconnaissance operations are a complete process involving multiple interconnected stages, encompassing takeoff from the ground control station / pilot's location, flight to the target area to perform reconnaissance (point reconnaissance, line reconnaissance, area reconnaissance) missions, and return to the takeoff and landing point for recovery upon mission completion. To meet the requirements of system-on-systems confrontation simulation for efficient, coherent, and engineering-practical path planning, this paper proposes a phased path modeling framework. This framework divides the UAV reconnaissance mission system into four logical stages: Take-off, Ingress, Reconnaissance, and Outgress. The paths of each stage are seamlessly connected through set operations to form a complete mission path model.

[0051]

[0052] In the formula, This represents the set of all path points for the entire UAV mission. , , , Define the path point sets for the UAV takeoff, entry, reconnaissance, and recovery phases respectively.

[0053] 1. User input parameter definition

[0054] The flexibility of path generation depends on the tactical parameters input by the user. Table 1 defines the key input parameters and their functions, covering reconnaissance type, geographic information, motion constraints, and dynamic adjustment options. The parameter design follows the principle of engineering practicality, supporting both mandatory and optional configurations to accommodate complex simulation scenarios.

[0055] Table 1 User Input Parameter Table

[0056]

[0057] 2. Path of each stage

[0058] (1) Takeoff phase

[0059]

[0060] In the formula, the definition is... This marks the starting point of the drone's takeoff phase, i.e., the position of the ground control station / pilot; [Definition] This marks the end of the takeoff phase for the unmanned aerial vehicle (UAV).

[0061] If the user specifies the access path ,but for The starting point, that is:

[0062]

[0063] In the formula, use express The starting point.

[0064] If the user does not specify an entry route, but provides the drone takeoff extension distance... The UAV attitude data and extended range provided by the simulation platform are then implemented. calculate Location:

[0065]

[0066] In the formula, Define the yaw angle of the drone; Define the drone's pitch angle.

[0067] (2) Entering the stage

[0068] The entry phase connects the end of the takeoff phase, the entry route, and the start of the reconnaissance phase, as follows:

[0069]

[0070] In the formula, the definition is... This marks the beginning of the reconnaissance phase.

[0071] (3) Reconnaissance phase

[0072] The reconnaissance phase connects the entry point to the reconnaissance route and the reconnaissance path, namely:

[0073]

[0074] In the formula, the definition is... This marks the end of the phase. The path is dynamically generated based on the reconnaissance type (area / point / line). The specific algorithm will be discussed in subsequent chapters.

[0075] (4) Recycling stage

[0076] The recovery phase connects the end point of the reconnaissance phase, the exit route, and the drone's starting point, namely:

[0077]

[0078] In the formula, the definition is... End of the drone reconnaissance phase

[0079] (II) Region and point reconnaissance path planning based on XML templates and affine transformations

[0080] To address the path generation requirements for both area and point reconnaissance, this section proposes a reconnaissance path planning algorithm based on XML templates and affine transformations. This method utilizes predefined standardized path templates and geometric transformation techniques to achieve efficient generation and flexible adaptation of reconnaissance paths, meeting the requirements for path configurability and real-time performance in system-on-systems simulations. The algorithm flow is as follows: Figure 1 As shown, it consists of 4 steps.

[0081] Step 1: Predefine various normalized path templates, such as circular (point reconnaissance), figure-eight (area reconnaissance), W-shaped (area reconnaissance), etc., and store them in XML format;

[0082] Step 2: Calculate the MBB of the target region;

[0083] Step 3: Perform an affine transformation on the path template using MBB, and calculate the transformed reconnaissance path;

[0084] Step 4: Optimize the sequence of the converted reconnaissance paths and calculate the final reconnaissance path.

[0085] (1) XML reconnaissance path template

[0086] We use XML to define normalized path templates to ensure cross-platform compatibility and scalability.

[0087] (1) <route>Tags. Use <route>The `<path>` tag defines the root element of the reconnaissance path. The `name` attribute defines the name of the reconnaissance path, such as "Circle" for a circle. The `closed` attribute defines whether the path is closed, with `true` for closed and `false` for open.

[0088] (2) <point>Tags. Use <point>for <route>The sub-tag defines a point in the reconnaissance path template. <point>The x attribute in the tag defines the horizontal coordinate (east). The y attribute defines the vertical coordinate (north). .

[0089] Normalized path templates can be defined in circular, figure-eight, and W shapes. There are a total of o points, of which As shown in the table below:

[0090] Table 2 Predefined template XML

[0091]

[0092] Normalized reconnaissance path such as Figure 2 As shown.

[0093] (2) Calculate MBB

[0094] Assume the irregular reconnaissance area consists of n path points ( , Represents the coordinates of the east direction. It consists of coordinates representing the north direction, i.e. The calculation can be performed using the classic rotating caliper method (see existing techniques: Freeman H, Shapira R. Determining the minimum-area encasing rectangle for an arbitrary closed curve[J]. Communications of the ACM, 1975, 18(7):409-413. Shamos, Michael Ian. Computational geometry. Yale University, 1978.). MBB. Assume that MBB is The rectangle formed by the *th edge of the convex hull can be used to calculate the width and height of the center point of the MBB:

[0095]

[0096]

[0097]

[0098] In the formula, the center point of MBB is defined as... , Represents the coordinates of the east direction. Represents the coordinates of the north direction; definition Define the width of the MBB; Let the height of the MBB be ; define the four vertices of the MBB as . , Represents the coordinates of the east direction. Represents the north direction coordinates; define the counterclockwise rotation angle of MBB relative to the original coordinate system's X-axis as the rotation angle is... .

[0099] (3) Affine transformation

[0100] Affine transformation is used to map the normalized path template to the actual geographic space.

[0101] Step 1: Scale coordinates proportionally. Using the length and width of the MBB, scale and rotate the x and y coordinates of all path points in the reconnaissance path template proportionally.

[0102]

[0103] In the formula: definition The relative offset in the X direction after scaling and rotation; definition The relative offset in the Y direction after scaling and rotation.

[0104] Step 2: 2D / 3D coordinate transformation, using the center of the MBB to convert to actual 2D coordinates:

[0105]

[0106]

[0107] In the formula: Define a set of three-dimensional coordinate points; The first one in the definition template The two-dimensional coordinates of each path point after transformation; definition The height represented by the point; Define a second-order identity matrix.

[0108] (4) Path optimization

[0109] Adjust the order of waypoints to connect the endpoints of the entry phase. For a closed path, iterate through all points in the path and select the point with the specified distance. The nearest point is chosen as the entry point; for open paths, the first and last points in the path are compared, and the point with the shortest distance is selected. The nearest point is used as the entry point. The remaining points are arranged in order, i.e.:

[0110]

[0111] In the formula, Defined as The A point, which is related to Closest distance.

[0112] (III) Line reconnaissance path planning based on geometric calculation

[0113] To address the need for generating reconnaissance paths for linear targets such as roads and boundaries, this section proposes a path planning algorithm based on geometric computation. This method generates a reconnaissance path parallel to the target's linear path while maintaining the offset distance through geometric offset and smoothing processes. This ensures path continuity and flyability, meeting the requirements of adaptability to complex terrain and engineering practicality in system-on-systems simulations. The algorithm flow is as follows: Figure 3 As shown, it consists of 4 steps.

[0114] Step 1: Using the line segment offset model, perform forward and reverse line segment offsets on the user-input reconnaissance path.

[0115] Step 2: Smooth the inflection points of the offset line segments generated in Step 1 to form an inflection point set.

[0116] Step 3: Merge and deduplicate the forward path offset points, reverse path offset points, and inflection point sets generated in Step 1 and Step 2.

[0117] Step 4: Perform 3D coordinate transformation and sequence optimization on the deduplicated reconnaissance point set to form the final reconnaissance path.

[0118] (1) Line segment offset model

[0119] Assume a set of n forward-oriented two-dimensional points that constitute a continuous reconnaissance path. Reverse two-dimensional point set ,in .

[0120] Construct a line segment offset model and calculate respectively and The offset line segment. This patent will now explain using a forward path as an example. For each line segment Its offset The subsequent offset point is:

[0121]

[0122] In the formula, Define line segment The offset point; definition The direction of progress; Definition from Starting vertical direction ( Rotate 90° counterclockwise.

[0123] like Figure 4 As shown, This represents the offset point of the reverse path.

[0124] (2) Corner processing model

[0125] Taking the forward path as an example, for each line segment The corner is calculated as follows.

[0126] Step 1: Calculate separately , Offset line segment , :

[0127]

[0128] In the formula, definition The starting point of the line segment; definition The endpoint of the line segment; definition The starting point of the line segment; definition The endpoint of the line segment.

[0129] Step 2: Intersection , intersection of line segments

[0130] when , When the segments are not collinear, the intersection points of the line segments or the intersection points of their extensions can be found. The following equation can then be solved:

[0131]

[0132] In the formula, s defines the intersection point along... Scale parameters (from) Set off), ;t is defined as the intersection point along Scale parameters (from) Set off), ;

[0133] Then the solution for the intersection point:

[0134]

[0135] In the formula, Define the intersection point of two offset lines or the intersection point of their extensions.

[0136] when , When collinear, and Solution with concurrent points and intersections:

[0137]

[0138] Step 3: Solve for the set of corner points:

[0139] like , The extensions of the lines intersect, that is Corner point set:

[0140]

[0141] In the formula Define the set of corner points.

[0142] like , The intersection point lies on any one of the line segments, that is Then the set of corner points:

[0143]

[0144] like Figure 5-6 As shown, Points such as "etc." represent the offset points of the reverse path, distinguishing them from the points of the forward path.

[0145] (3) Path generation

[0146] Step 1: Based on the above line segment offset and corner processing model, calculate respectively and offset path and :

[0147]

[0148]

[0149]

[0150] In the formula, Define the forward reconnaissance path, including the starting point offset, corner point, ending point offset, and ending point correction point; Define the reverse reconnaissance path, including the starting point offset, corner point, starting point offset, and reverse endpoint correction point; Define the endpoint correction point of the forward path; Define the endpoint correction point for the reverse path; Definition of the first One positive path offset point; Definition of the first One reverse path offset point.

[0151] Step Two: Merge and After deduplication, a set of two-dimensional coordinate points for the reconnaissance route is obtained, namely:

[0152] ,

[0153] In the formula, Define the merged path ,common One, of which ; The distance threshold for determining the repetition of coordinate points is set to, for example, 0.001.

[0154] (4) 3D coordinate transformation and path point optimization

[0155] The three-dimensional coordinate transformation and path point optimization rule-based regional reconnaissance methods will not be elaborated here.

[0156] II. Experiment and Results Description

[0157] Based on the actual reconnaissance mission requirements of UAVs in the research project, the following four sets of experiments were designed.

[0158] (a) End point prediction of UAV takeoff phase

[0159] Set the starting position of the drone The user inputs L = 10 meters, and the simulation platform obtains its yaw angle. Pitch angle The algorithm presented in this paper can then predict the endpoint of the drone's takeoff phase. ,See Figure 7 .

[0160] (II) Planning the reconnaissance route

[0161] For point reconnaissance, since the rotation angle of its MBB has no effect on the reconnaissance path, it can be set to 0, i.e., parallel to the X-axis. Let the UAV's current position be (3.6, 3.6, 0.5), ready to enter the reconnaissance phase. The user-input reconnaissance mission parameters are as follows: , , If the terrain around the target point is flat, then the algorithm in this paper can generate a reconnaissance path. Figure 8 .

[0162] (III) Planning the reconnaissance route in the area

[0163] Assume the UAV's current position is (0.8, 1.0, 0.5), and it is ready to enter the reconnaissance phase. The user inputs the reconnaissance mission parameters as follows: Example area 1 ( ) and example area 1 ( )(like Figure 9-12 As shown), perform respectively , If the parameters are that the ground around the target point is flat, then the reconnaissance path can be generated by the algorithm in this paper as shown in 9-12.

[0164] (iv) Planning the reconnaissance route

[0165] Assume the UAV's current position is (0.5, 2.5, 0.5), and it is ready to enter the reconnaissance phase. The user inputs the reconnaissance mission parameters as follows: Example Route 1 ( ) and Example Route 2 ( )(like Figure 13-14 (as shown) , Assuming the terrain around the target point is flat, the algorithm in this paper can generate a reconnaissance path as shown in Figures 13-14.

[0166] This invention, based on UAV reconnaissance missions and addressing the needs of UAV path planning simulation modeling, proposes a mission-driven, phased path generation framework covering four mission phases: takeoff, entry, reconnaissance, and recovery. The parameters and calculation models for each phase are defined in detail. By combining real-time UAV attitude data, dynamic prediction of the takeoff phase endpoint is achieved; using XML templates and affine transformation methods, flexible generation of area and point reconnaissance paths is realized; and through line segment offset and corner processing algorithms, smooth construction of line reconnaissance paths is achieved.

[0167] This invention constructs a complete set of experimental cases for verification. Through multiple sets of experiments, the path generation process under different reconnaissance missions is simulated, verifying the practicality and flexibility of the proposed modeling framework and algorithm, and providing strong support for engineering practice applications.

[0168] Experimental results show that this invention can effectively generate UAV paths that meet the requirements of actual tasks, demonstrating good adaptability and engineering practical value. This method not only improves the automation level of path planning but also lays the foundation for further research on multi-UAV collaboration and dynamic environment adaptation.

[0169] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention. The scope of protection claimed by the appended claims and their equivalents is defined.< / point> < / route> < / point> < / point> < / route> < / route>

Claims

1. A task-driven UAV reconnaissance path planning method, characterized in that, Includes the following steps: Step S1: Construct a phased path planning modeling framework, dividing the UAV reconnaissance mission into the take-off phase, entry phase, reconnaissance phase and recovery phase, and defining the input parameters and calculation models for each phase, wherein the input parameters include reconnaissance type, geographic information and motion constraints; Step S2: Based on the reconnaissance type, generate the reconnaissance phase path using the corresponding path generation algorithm: When the reconnaissance type is area reconnaissance or point reconnaissance, a path generation algorithm based on XML templates and affine transformations is used. When the reconnaissance type is line reconnaissance, a path generation algorithm based on geometry calculation is used; Step S3: Based on the connection relationship of each stage path, integrate the paths of the takeoff stage, entry stage, reconnaissance stage and recovery stage to form a complete mission path.

2. The method according to claim 1, characterized in that, The generation of the takeoff phase path includes: if the user specifies an entry path, the takeoff phase endpoint is the starting point of the entry path; if the user does not specify an entry path but provides a takeoff extension distance, the takeoff phase endpoint is calculated based on the UAV's real-time attitude information, wherein the real-time attitude information includes yaw angle and pitch angle.

3. The method according to claim 1, characterized in that, The path generation algorithm based on XML templates and affine transformations includes: predefining a normalized path template and storing it in XML format, wherein the template includes circular, figure-eight, or W-shaped paths; calculating the minimum bounding box (MBB) of the target region and obtaining its center point, width, height, and rotation angle; performing an affine transformation on the normalized path template using the minimum bounding box, including proportional scaling, rotation, and translation; and optimizing the sequence of the transformed path to connect to the end point of the stage.

4. The method according to claim 1, characterized in that, The line reconnaissance path generation algorithm based on geometric calculation includes: shifting the user-input linear path by forward and reverse line segments to generate shifted line segments; smoothing the corners of the shifted line segments and generating a set of inflection points by solving the intersection points of the line segments or the intersection points of the extended lines; merging the forward shifted path, the reverse shifted path, and the set of inflection points to remove duplicates; and performing three-dimensional coordinate transformation and sequence optimization on the deduplicated point set to form a reconnaissance path.

5. A mission-driven UAV reconnaissance path planning device, characterized in that, include: The phased path modeling module is used to build a phased path planning modeling framework, which divides the UAV reconnaissance mission into the take-off phase, entry phase, reconnaissance phase and recovery phase, and defines the input parameters and calculation models for each phase. The reconnaissance path generation module is used to generate reconnaissance phase paths based on the reconnaissance type, including: The region / point reconnaissance path generation submodule is used to execute a path generation algorithm based on XML templates and affine transformations; The line reconnaissance path generation submodule is used to execute a path generation algorithm based on geometry calculations; The path integration module is used to integrate the paths of each stage to form a complete task path.

6. The apparatus according to claim 5, characterized in that, The phased path modeling module includes a takeoff phase processing unit, used for: If the user specifies an entry path, then the end point of the takeoff phase is set as the starting point of the entry path; If the user does not specify an entry path but provides a takeoff extension distance, the takeoff phase endpoint will be calculated based on the UAV's real-time attitude information.

7. The apparatus according to claim 5, characterized in that, The region / point reconnaissance path generation submodule includes: Template storage unit, used to store predefined normalized XML path templates; Minimum bounding box (MBB) calculation unit, used to calculate the minimum bounding box (MBB) of the target region; Affine transformation unit, used to perform affine transformation on the template using the minimum bounding box (MBB); The path optimization unit is used to optimize the sequence of the transformed path.

8. The apparatus according to claim 5, characterized in that, The affine transformation unit is configured as follows: The point coordinates in the path template are scaled and rotated proportionally based on the width, height and rotation angle of the minimum bounding box (MBB). The scaled and rotated points are translated to the center of the minimum bounding box (MBB) to obtain two-dimensional coordinates. Add height information to obtain three-dimensional coordinates.

9. The apparatus according to claim 5, characterized in that, The line reconnaissance path generation submodule includes: The line segment offset unit is used to offset line segments in the forward and reverse directions of the linear path input by the user. The corner processing unit is used to smooth the corners of the offset line segments and generate a set of inflection points. The path merging unit is used to merge forward offset paths, reverse offset paths, and inflection point sets, and to remove duplicates. The coordinate transformation unit is used for 3D coordinate transformation and sequence optimization.

10. A non-volatile storage medium, characterized in that, The non-volatile storage medium includes a stored program, wherein the program, when executed, controls the device where the non-volatile storage medium is located to perform the method described in any one of claims 1 to 4.

11. A terminal device, characterized in that, The terminal device includes: a processor, a memory, a communication interface, and a bus; the processor, the memory, and the communication interface are connected through the bus and communicate with each other; the memory stores executable program code; the processor reads the executable program code stored in the memory to run a program corresponding to the executable program code, so as to execute the method as described in any one of claims 1-4 above.

Images