Shortest path dynamic programming method for complex forest scene and related device

Abstract

The application discloses a shortest path dynamic programming method for a complex forest scene and related equipment, defines a geographical environment representation mode for personnel travel planning in the complex forest scene, simultaneously considers the influence of personnel elements on travel speed from two factors of load and physiological indexes, records space-time trajectory data of the personnel by using a trajectory measuring instrument, constructs a travel speed model under the coupling action of the forest environment and the personnel elements, and combines a traversal search algorithm to form a shortest path optimization algorithm in the forest scene, so that path planning auxiliary decision-making in multiple fields such as forest fires and emergency search and rescue is realized, the deficiencies of traditional methods in aspects of forest scene feature expression and travel speed function construction are solved, and the path optimization decision-making capability of emergency institutions in the complex forest scene is improved.

Description

Technical Field

[0001] This invention belongs to the field of path planning technology, and in particular relates to a shortest path dynamic planning method and related equipment for complex forest scenarios. Background Technology

[0002] Accelerated global urbanization has led to a more closely intertwined spatial distribution of forests and cities. Interactions between urban residents and natural forests have also become more frequent. The mechanisms and patterns of these interactions have gradually become a research hotspot, with wide applications in numerous fields such as global climate, ecological security, and community health. The movement patterns of people within forest spatial scenarios are a typical example of interaction research, possessing significant application value in areas such as forest fires and emergency search and rescue. For instance, the rapid arrival of firefighters in fire-fighting operations and the timely evacuation of high-risk fire areas, the planning of hiking routes for outdoor enthusiasts, and the search and rescue of casualties in the wild under extreme weather conditions all require quantitative human movement models to provide effective forest route planning services, thereby enabling better decision-making.

[0003] Path optimization research is widely applied in urban scenarios. Classical urban path planning methods mostly represent road nodes V and road edge attributes E in graph form G = (V, E). Then, traversal algorithms (such as Dijkstra's algorithm, A* algorithm, etc.) are used to generate the set of nodes with the minimum total cost C, including the starting and ending points. The cost function C of urban roads can be defined by the travel speed v of people or vehicles (for example, the cost c can be estimated by the time taken to traverse a road of length L, c = t = L / v). However, forest and urban scenarios differ significantly, and there is a lack of in-depth research on the movement patterns of people in forest scenarios, making it difficult to provide optimal path planning for people in forest areas under the coupled effects of complex geographical environments and human factors. Summary of the Invention

[0004] To overcome the shortcomings of existing technologies, this invention proposes a shortest path dynamic planning method and related equipment for complex forest scenarios. It conducts in-depth research on the geographical environment representation patterns in complex forest scenarios, analyzes the quantitative impact of human factors on travel speed, constructs a travel speed function under the coupling effect of forest environment and human factors, and provides shortest travel route planning services in forest scenarios, enabling path planning auxiliary decision-making in multiple fields such as forest fires and emergency search and rescue.

[0005] A shortest path dynamic programming method for complex forest scenarios includes the following steps:

[0006] The steps for numerical representation of forest scenes are as follows: acquire forest scene data, define a complex scene numerical representation model for personnel movement planning, and generate a set of key feature structures for forest scenes.

[0007] The steps for constructing a personnel movement sample database are as follows: considering the impact of personnel factors on movement speed from two major aspects, namely load and physiological indicators, and using a trajectory measuring instrument to record the spatiotemporal trajectory data of personnel to construct a standardized personnel movement sample database.

[0008] The steps of travel rate modeling and path optimization matching involve constructing a quantitative mapping function between geographical environment, personnel elements and travel rate, and combining it with a traversal search algorithm to search for the shortest path to a specified origin and destination.

[0009] In one embodiment,

[0010] The numerical representation of the forest scene also includes the following steps:

[0011] Forest scene data acquisition and processing: Collect multi-source heterogeneous data from multiple sensor platforms in the target forest area, use geographic data processing tools to extract digital elevation model (DEM) and vegetation type raster data, and perform operations on the above geographic data including but not limited to projection coordinate transformation, resampling, and cropping.

[0012] In one embodiment,

[0013] The numerical representation of the forest scene further includes the following steps: The forest area is divided into closely adjacent regular grids, each grid being a cell, where the spatial resolution of the cell is L. L is the grid resolution, which determines the fineness of spatial discretization and is an important parameter for calculating slope and constructing a database of pedestrian movement. The cell size has a significant impact on model accuracy.

[0014] In one embodiment,

[0015] The numerical representation of the forest scene also includes the following steps:

[0016] Facing a regular grid that divides the forest area into closely adjacent spaces, each grid is a cell;

[0017] For the geographical attribute A of the cell geo To characterize;

[0018] Geographical attribute A of a cell geo The attributes used to characterize the spatial structure of forests include, but are not limited to, the following:

[0019] A geo ={ID, elevation, roughness, VD, VT, weather};

[0020] Here, ID is the unique identifier of the cell, elevation represents terrain height in meters, roughness is the surface roughness, which can be represented by the mean u(N) and variance σ(N) of the point cloud elevation values ​​within the surface area, reflecting the ease or difficulty of traversing the surface, and VD represents vegetation density, which can be represented by the point cloud density in the area from the surface to the height of the person. The higher the density, the more difficult the traversal. VT represents vegetation type, such as shrubs, trees, grasslands, wastelands, and water bodies. Dense shrubs will make it almost impossible for people to walk, while wasteland areas are easier to traverse quickly. Weather is a set of meteorological parameters, including temperature T, humidity H, wind speed WS, and wind direction WD. When the temperature is too high, it will significantly affect the speed of movement.

[0021] In one embodiment,

[0022] The steps for constructing the personnel movement sample database also include the following steps:

[0023] Personnel indicator data collection and processing steps:

[0024] Define a personnel status attribute A that conforms to the forest scenario. per Used to quantitatively describe the characteristics of personnel during movement, including but not limited to the following attributes:

[0025] A per ={ID,PID,X,Y,load,physiology,time};

[0026] In this system, ID is the unique identifier for each person, PID is the unique identifier for each trajectory point, and (X, Y) are the spatial geographic coordinates of the person, representing the location of the trajectory point, such as latitude and longitude coordinates (113.12310, 23.86581). Load represents the weight carried by the person. For example, a firefighter carries a 15kg hose, while the weight carried by an ordinary person is generally negligible (0kg). Physiology is a set of physiological indicators, including heart rate (HR), blood oxygen saturation (SaO2), and other physiological parameters. During high-intensity exercise, the heart rate will remain at a high level, physical strength will gradually deplete, blood oxygen concentration will decrease, and walking speed will gradually decrease. Time is the moment the trajectory point was generated, such as 2023-02-16 12:30:30. Simple time difference calculations can be performed between different times to calculate the time interval between the generated trajectory points. The spatial coordinates of the person can be obtained using trajectory measuring equipment, the weight can be accurately measured using a weighing scale, and the physiological indicators can be obtained using professional physiological measuring instruments. The walking characteristics of people in a forest scene differ depending on their status attributes.

[0027] In one embodiment,

[0028] The steps for constructing the personnel movement sample database also include the following steps:

[0029] Steps for constructing trajectory stamp samples based on spatiotemporal alignment:

[0030] Under a unified spatial coordinate system, the collected personnel status attribute data is spatiotemporally aligned with the cellular geospatial data. Personnel trajectory points will fall on the cells of the corresponding coordinates, and trajectory points that are temporally adjacent will form a trajectory stamp sample.

[0031] In one embodiment,

[0032] The trajectory stamp sample is a vector sample, with its direction consistent with the time flow, i.e., it is generated from the previously generated trajectory point N. before The trajectory point N generated after pointing after The attribute space of a trajectory stamp is denoted as A. trj Includes, but is not limited to, the following attributes:

[0033] A trj ={ID, NID before NID after , distance, slope, time, speed; A per A geo};

[0034] Where / D is the unique identifier for the track stamp. NID before It is the number of the trajectory point that was generated first in time. This number can be used to query all attribute information of the corresponding person's trajectory point. per NID after This refers to the number of the track point generated later. Distance is the geographical distance between two track points in the track stamp, i.e., the track stamp length. Slope is the slope of the track stamp vector, which can be calculated by the track stamp length distance and the elevation difference ΔH between the track points. after -elevation before The representation is slope = ΔH / distance. Time represents the time taken for a person to traverse the length of the track stamp, calculated from the time difference between two track points, i.e., time = timea. fter -time before Speed ​​represents the movement rate of a person along the track marker, i.e., speed = distance / time. A per and A geo These correspond to the status attributes and geographical attributes of individuals at different times and locations. A sample database of human movement can be constructed from these movement trajectory stamps, containing key geographical environment and human characteristics information during the movement process, providing crucial data support for the analysis of patterns in human movement characteristics.

[0035] In one embodiment,

[0036] The travel rate modeling and path optimization steps also include the following steps:

[0037] Modeling of travel speed under the coupling of geographical environment and human factors:

[0038] A fitting equation f is used to construct a functional mapping relationship between movement speed and geographical environment and personnel elements. Through training and iteration with sample data, the parameter set θ of the fitting equation f is determined, so as to realize the quantitative calculation of personnel movement speed in forest scene.

[0039] The mathematical expression of the fitting equation is diverse, including but not limited to linear polynomials and nonlinear equations such as Gaussian distribution functions and deep neural networks. Through iterative training with sample data, the parameter set θ of the fitting equation f can be determined, thereby enabling the quantitative calculation of the movement speed of people in a forest scene. Taking a convolutional neural network as an example, the weight parameters w and bias b of network f are the model parameters to be determined. When continuously trained iteratively using trajectory stamp sample data (X, Y), the initial model parameters w0 and b0 can converge to stable values ​​w′ and b′ with small loss errors. When new trajectory stamp sample data X′ is input, network f can accurately estimate the movement speed value Y′.

[0040] In one embodiment,

[0041] The travel rate modeling and path optimization steps also include the following steps:

[0042] Construction of trajectory stamp samples based on spatiotemporal alignment: The forest space is discretized into a set of cells C = {c0, c1, ..., c2}. n}, where n is the total number of cells within the region; corresponding to the cell set, construct the shortest time set T = {t0, t1, ..., t}. n The optimal path pointing set R = {r0, r1, ..., r} n};

[0043] The shortest time set records the time taken for the shortest path from the starting cell to all cells in the forest space, with an initial default value of ∞;

[0044] The optimal path pointing set records the cell pointing in the neighborhood of the shortest travel path. The pointing can be replaced by {0, 1, ..., 7} instead of the eight directions in the Moore neighborhood.

[0045] The initial default value of the pointed-to value is -1;

[0046] When the start and end coordinates are set, the corresponding spatial location is the start cell c. iand endpoint cell c j Selected;

[0047] cell c i The corresponding shortest time t i Updated to 0, the value being pointed to retains the default value r. i =-1;

[0048] Generate cell c i Trajectory stamp A to each neighboring cell in the Moore neighborhood trj Then, use the travel rate function f to calculate the temporary set of travel time. Furthermore, it can be deduced from cell c i The set of times of arrival at neighboring cells

[0049] When T i If the value is less than the corresponding value in the shortest time set T, update the time value and change the cell c. i The pointer relationship value of the cell is updated in the pointer set R;

[0050] After completing one iteration update, retrieve the minimum value in T that has not been iterated over, such as cell c. i If a cell has already been iterated over, it will no longer be selected. Repeat the above update steps until all cells have been traversed.

[0051] Based on the iteratively updated sets T and R, the optimal travel route for any specified starting point and ending point can be generated.

[0052] In one embodiment,

[0053] A terminal includes a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that...

[0054] The processor implements the above-described method when executing the computer program.

[0055] In one embodiment,

[0056] The aforementioned terminal also includes a data acquisition unit and a display, characterized in that,

[0057] The data acquisition device includes at least a thermometer, hygrometer, anemometer, trajectory measuring instrument, weighing instrument, and physiological measuring instrument;

[0058] The display shows the path planning results.

[0059] In one embodiment,

[0060] A computer-readable storage medium storing a computer program, characterized in that the computer program, when executed by a processor, is the method described above.

[0061] The beneficial effects of this invention are:

[0062] Addressing the path planning needs in forest scenarios such as forest fires and wilderness search and rescue, this study analyzes the characteristics of movement patterns under different forest geographical environments and individual physiological factors. A novel movement rate model coupling geographical environment and human factors is constructed, and combined with a traversal search algorithm, a shortest path optimization algorithm in forest scenarios is formed. This solves the shortcomings of traditional methods in forest scenario feature representation and movement rate function construction, and improves the path optimization decision-making capabilities of emergency agencies in complex forest scenarios. Attached Figure Description

[0063] Figure 1 This is a flowchart of a shortest path dynamic programming method for complex forest scenarios according to an embodiment of the present invention.

[0064] Figure 2 This is a schematic diagram of cell partitioning and Moore neighborhood definition in the forest scene of this invention.

[0065] Figure 3 This is a schematic diagram illustrating the spatiotemporal alignment of the sample stamp with the geographical environment in this invention.

[0066] Figure 4 This is a flowchart of the path optimization steps based on a traversal search algorithm according to an embodiment of the present invention.

[0067] Figure 5 This is a system composition diagram of a shortest path dynamic programming device for complex forest scenarios according to an embodiment of the present invention.

[0068] Figure 6 This is a schematic diagram of the structure of a terminal provided in an embodiment of the present invention. Detailed Implementation

[0069] The preferred embodiments of the invention will be described in further detail below.

[0070] like Figure 1 As shown, an embodiment of the shortest path dynamic programming method for complex forest scenarios includes the following steps:

[0071] S1. Divide the forest area into closely adjacent regular grids, with each grid being a single cell, such as... Figure 2 As shown, the spatial resolution of the cell is L, which can be set to 30 meters or 5 meters, or other values.

[0072] S2, Obtain the geographic attributes A of all cells. geo Includes, but is not limited to, the following attributes:

[0073] A geo={ID, elevation, roughness, VD, VT, weather};

[0074] S3, Personnel Status Attribute A per Includes, but is not limited to, the following attributes:

[0075] A per ={ID,PID,X,Y,load,physiology,time};

[0076] S4. Perform spatiotemporal alignment processing between personnel status attribute data and cellular geospatial data.

[0077] S5. Generate trajectory stamp samples, whose attribute space is denoted as A. trj Construct a sample database of personnel movement, A trj Includes, but is not limited to, the following attributes:

[0078] A trj ={ID, NID before NID after , distance, slope, time, speed; A per A geo};

[0079] S6. Modeling of travel speed under the coupling of geographical environment and human factors:

[0080] speed=f(A trj A per A geo );

[0081] S7. Path optimization based on traversal search algorithm:

[0082] Route = Search(f; A trj A per A geo ).

[0083] The personnel trajectory stamp sample set is the data foundation for constructing the personnel movement rate function in the forest scene. Under a unified spatial coordinate system, the collected personnel status attribute data is spatiotemporally aligned with the cellular geospatial data, such as... Figure 3 As shown. Personnel trajectory points will fall on the corresponding coordinate cell grid. Trajectory points that are temporally adjacent will form a trajectory stamp sample. The trajectory stamp is a vector sample, with its direction consistent with the time flow, i.e., it is generated by the first generated trajectory point N. before The trajectory point N generated after pointing after For example, N before The time value is 2023-02-16 12:30:30, N afterThe time value is 2023-02-16 12:31:00, and there are no other trajectory points within the time period formed by the two times, i.e., N. before With N after These are adjacent trajectory points. The attribute space of a trajectory stamp is denoted as A. trj Includes, but is not limited to, the following attributes:

[0084] A trj ={ID, NID before NID after , distance, slope, time, speed; A per A geo};

[0085] Here, ID is a unique identifier for the track stamp. NID before It is the number of the trajectory point that was generated first in time. This number can be used to query all attribute information of the corresponding person's trajectory point. per NID after This refers to the number of the track point generated later. Distance is the geographical distance between two track points in the track stamp, i.e., the track stamp length. Slope is the slope of the track stamp vector, which can be calculated by the track stamp length distance and the elevation difference ΔH between the track points. after -elevation before The representation is slope = ΔH / distance. Time represents the time taken for a person to traverse the length of the track stamp, calculated from the time difference between two track points, i.e., time = time... after -time before Speed ​​represents the movement rate of a person along the track marker, i.e., speed = distance / time. A per and A geo These correspond to the status attributes and geographical attributes of individuals at different times and locations. A sample database of human movement can be constructed from these movement trajectory stamps, containing key geographical environment and human characteristics information during the movement process, providing crucial data support for the analysis of patterns in human movement characteristics.

[0086] In a discretized forest space, to achieve path planning with a specified start and end point, a traversal search algorithm is needed to calculate the optimal route.

[0087] Route = Search(f; A trj A per A geo );

[0088] Search algorithms come in various forms, such as breadth-first search and depth-first search algorithms, including but not limited to Dijkstra's algorithm, Floyd's algorithm, and A* algorithm. Taking Dijkstra's breadth-first search algorithm as an example... Figure 4 As shown, in one embodiment, the general calculation process is as follows:

[0089] Step S301: Assume the forest space is discretized into a set of cells C = {c0, c1, ..., c2}. n}, where n is the total number of cells within the region; corresponding to the cell set, construct the shortest time set T = {t0, t1, ..., t}. n The optimal path pointing set R = {r0, r1, ..., r} n The shortest time set T records the time taken for the shortest path from the starting cell to all cells in the forest space; the initial default value of T is ∞, indicating that it is unreachable in the time dimension; the optimal path pointing set R records the cell pointing within the neighborhood of the shortest path, and the pointing can use {0, 1, ..., 7} to replace the eight directions in the Moore neighborhood, such as... Figure 2 As shown, these correspond to due north, northeast, east, southeast, south, southwest, west, and northwest, respectively. The initial default value for the pointing value is -1, indicating that it does not point in any direction. When the start and end coordinates are set, the corresponding spatial location start cell c i and endpoint cell c j Selected, cell c i The corresponding shortest time t i Updated to 0, the value being pointed to retains the default value r. i =-1;

[0090] Step S302: Select the minimum value t in set T. i The corresponding cell is the cell c to be iterated. i Each cell can only be selected once;

[0091] Step S303: Generate cell c i Trajectory stamp A to each neighboring cell in the Moore neighborhood trj ;

[0092] Step S304: Calculate the temporary set of travel time using the travel rate function f.

[0093] Step S305: Calculate from cell c i The set of times of arrival at neighboring cells

[0094] Step S306: Compare when T i The value is less than the corresponding value in the shortest time set T;

[0095] Step S307: When T i If the value is less than the corresponding value in the shortest time set T, update the time value and change the cell c. i The pointer relationship value of the cell is updated in the pointer set R;

[0096] After completing one iteration update, retrieve the minimum value in T that has not been iterated over, such as cell c. i If a cell has already been iterated over, it will no longer be selected. Repeat steps S303 to S307 until all cells have been traversed.

[0097] Step S308: Update the final sets T and R to generate the optimal travel route for any specified start and end point.

[0098] like Figure 5 As shown, in one embodiment,

[0099] A shortest path dynamic programming device 10 for complex forest scenarios includes the following modules:

[0100] The forest scene numerical representation module 101 includes a forest scene data acquisition and processing unit 1011 and a forest feature structure representation unit 1012. The forest scene data acquisition and processing unit 1011 is used to acquire forest scene data, and the forest feature structure representation unit 1012 is used to generate a set of key feature structures of the forest scene by using a complex scene numerical representation mode oriented towards personnel movement planning.

[0101] The personnel movement sample database construction module 102 includes a personnel data indicator acquisition and processing unit 1021 and a trajectory stamp sample construction unit 1022. The personnel data indicator acquisition and processing unit 1021 is used to collect personnel characteristics such as load and physiological indicators and to perform quantitative description. The trajectory stamp sample construction unit 1022 is used to construct trajectory stamp samples based on spatiotemporal alignment to form a standardized personnel movement sample database.

[0102] The travel speed modeling and path optimization module 103 includes a travel speed modeling unit 1031 and a path optimization unit 1032. The travel speed modeling unit 1031 uses statistical analysis and deep learning methods to construct a quantitative mapping function between the geographical environment, personnel elements and travel speed. The path optimization unit 1032 uses a traversal search algorithm to calculate the optimal path for a specified start and end point, providing the shortest path planning for forest fire rescue teams in complex forest scenarios.

[0103] The module / unit referred to in this invention refers to a type of module / unit that can be... Figure 6 The processor 13 in the middle acquires a series of computer-readable instruction segments that are capable of performing a fixed function, and these segments are stored in Figure 6 It is stored in memory 12.

[0104] like Figure 6 The diagram shown is a structural schematic of a terminal provided in an embodiment of the present invention. Figure 6 As shown, the terminal 1 may include a data acquisition unit 11, a memory 12, a processor 13, and a display 14.

[0105] The data acquisition unit 11 includes a thermometer, hygrometer, an anemometer, a trajectory measuring instrument, a weighing instrument, and a physiological measuring instrument.

[0106] Memory 12 may include one or more random access memory (RAM) and one or more non-volatile memory (NVM). The RAM can be directly read and written by the processor 13, and can be used to store executable programs (e.g., machine instructions) of other running programs, as well as user and application data. The RAM may include static random-access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), etc.

[0107] Non-volatile memory can also store executable programs and user and application data, and can be pre-loaded into random access memory for direct reading and writing by the processor 13. Non-volatile memory can include disk storage devices and flash memory.

[0108] Memory 12 is used to store one or more computer programs. The one or more computer programs are configured to be executed by processor 13. The one or more computer programs include multiple instructions that, when executed by processor 13, can implement a shortest path dynamic programming method for complex forest scenarios, which is executed on terminal 1.

[0109] In other embodiments, such as Figure 6 The terminal 1 shown also includes an external memory interface for connecting to an external memory to expand the storage capacity of the terminal 1.

[0110] Processor 13 may include one or more processing units, such as an application processor (AP), a modem processor, a graphics processing unit (GPU), an image signal processor (ISP), a controller, a video codec, a digital signal processor (DSP), and / or a neural network processing unit (NPU). These different processing units may be independent devices or integrated into one or more processors.

[0111] Processor 13 provides computational and control capabilities, for example, processor 13 is used to execute computer programs stored in memory 12 to implement the shortest path dynamic programming method for complex forest scenarios described above.

[0112] Display 14 is used to display the path planning results.

[0113] This invention also provides a computer-readable storage medium storing a computer program, which includes program instructions. When the program instructions are executed, the method implemented can be referred to the methods in the above embodiments of this invention.

[0114] The computer-readable storage medium can be the internal memory of the terminal described in the above embodiments, such as the terminal's hard disk or memory. Alternatively, it can be an external storage device for the terminal, such as a plug-in hard disk, smart media card (SMC), secure digital card (SD), flash card, etc.

[0115] In some embodiments, a computer-readable storage medium may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application program required for at least one function, etc.; and the data storage area may store data created based on the use of the terminal, etc.

[0116] The above description, in conjunction with specific preferred embodiments, provides a further detailed explanation of the present invention. It should not be construed that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art, various simple deductions or substitutions can be made without departing from the inventive concept, and all such modifications and substitutions should be considered to fall within the patent protection scope defined by the submitted claims.

Claims

1. A shortest path dynamic programming method for complex forest scenarios, characterized by: Includes the following steps: The steps for numerical representation of forest scenes are as follows: acquire forest scene data, define a complex scene numerical representation mode for personnel movement planning, and generate a set of key feature structures for forest scenes, including: a regular grid that divides the forest area into closely adjacent grids, with each grid being a cell. The steps for constructing a personnel movement sample database are as follows: considering the impact of personnel elements on movement speed from two major factors, namely load and physiological indicators, and using a trajectory measuring instrument to record the spatiotemporal trajectory data of personnel to construct a standardized personnel movement sample database. The steps include: constructing trajectory stamp samples based on spatiotemporal alignment: under a unified spatial coordinate system, the collected personnel status attribute data and cellular geographic space are spatiotemporally aligned. Personnel trajectory points will fall on the cells of the corresponding coordinates, and trajectory points that are temporally adjacent will constitute a trajectory stamp sample. The steps of modeling travel speed and optimizing and matching paths involve constructing a quantitative mapping function between geographical environment, personnel elements and travel speed based on a trajectory stamp sample set, and combining it with a traversal search algorithm to search for the shortest path to a specified origin and destination. The travel rate modeling and path optimization steps also include the following steps: Modeling of travel speed under the coupling of geographical environment and human factors: Adopting the fitting equation A functional mapping relationship between travel speed and geographical environment and personnel elements is constructed, and the fitting equation is determined through training and iteration with sample data. parameter set To achieve quantitative calculation of the movement speed of people in forest scenes; Construction of trajectory stamp samples based on spatiotemporal alignment: The forest space is discretized into a set of cells. ,in, Given the total number of cells within the region; for the corresponding cell set, construct the shortest time set. and the set of optimal paths ; The shortest time set records the time taken for the shortest path from the starting cell to all cells in the forest space, with an initial default value of ∞; The optimal path pointer set records the cell pointers within the neighborhood of the shortest path. The pointers are used... To replace the eight directions in the Moore neighborhood; The initial default value of the pointed-to value is -1; When setting the start and end coordinates, the corresponding spatial location is the start cell. and terminal cell Selected; cell The corresponding shortest time Updated to 0, the value being pointed to retains its default value. ; Generating cells Trajectory stamps to each neighboring cell in the Moore neighborhood Then use the travel rate function Temporary set for calculating travel time Furthermore, it can be deduced from the cell The set of times of arrival at neighboring cells ; when The value is less than the shortest time set. When the corresponding value is found, update the time consumption value and change the cell. The pointer relation value of the cell is updated to the pointer set. middle; After completing one iteration update, retrieve it again. Find the minimum value that has not been iterated over, and repeat the above update steps until all cells have been traversed; Update the set iteratively to the final set. and It generates the optimal travel route for any specified start and end point.

2. The shortest path dynamic programming method for complex forest scenarios as described in claim 1, characterized in that, The numerical representation of the forest scene also includes the following steps: Forest scene data acquisition and processing steps: Collect multi-source heterogeneous data from multiple sensing platforms in the target forest area, use geographic data processing tools to extract digital elevation model (DEM) and vegetation type raster data, and perform at least one of the following operations on the above geographic data: projection coordinate transformation, resampling, and cropping.

3. The shortest path dynamic programming method for complex forest scenarios as described in claim 1, characterized in that, The numerical representation of the forest scene also includes the following steps: Forest feature structure representation steps for personnel movement planning: Geographical attributes of the cells To characterize; Geographical attributes of cells Used to characterize the spatial structure of forests, it includes the following attributes: ； in, It is the unique identifier of a cell; Indicates terrain elevation; Indicates surface roughness; Indicates vegetation density; Indicates vegetation type; It is a set of meteorological parameters, which at least includes temperature. ,humidity Wind speed ,wind direction parameter.

4. The shortest path dynamic programming method for complex forest scenarios as described in claim 1, characterized in that, The steps for constructing the personnel movement sample database also include the following steps: Personnel indicator data collection and processing steps: Define personnel status attributes that conform to the firefighting scenario. It is used to quantitatively describe the characteristics of personnel during a march and includes the following attributes: ； in, It is a unique identification number for the personnel; It is a unique identifier for the trajectory point; These are the spatial geographic coordinates of a person, representing the location of a trajectory point; It refers to the load-bearing capacity, indicating the weight of the fire-fighting equipment carried by personnel. It is a set of physiological indicators of a person, including at least heart rate. Blood oxygen saturation Physiological parameters; It is the moment when the trajectory point is generated; Collect and process the aforementioned personnel status attribute data.

5. The shortest path dynamic programming method for complex forest scenarios as described in claim 1, characterized in that, The trajectory stamp sample is a vector sample, with its direction aligned with the time flow, i.e., it consists of the trajectory points generated earlier. The trajectory point generated after pointing The attribute space of a trajectory stamp is denoted as It includes the following attribute items: ； in, It is a unique identifier for the trajectory stamp; It is the number of the trajectory point that occurred first in time; It is the number of the trajectory point generated later; It is the geographical distance between two trajectory points marked by the trajectory stamp; It is the slope of the trajectory stamp vector; This indicates the time taken for a person to complete the entire length of the trajectory stamp. This indicates the speed at which personnel move along the track marker; and These correspond to the status attributes and geographical attributes of personnel in a given time and space.

6. A shortest path dynamic programming device for complex forest scenarios, the device being used to implement the method as described in any one of claims 1 to 5, the device operating on a terminal, characterized in that, Includes the following modules: The forest scene numerical representation module 101 includes a forest scene data acquisition and processing unit 1011 and a forest feature structure representation unit 1012. The forest scene data acquisition and processing unit 1011 is used to acquire forest scene data, and the forest feature structure representation unit 1012 is used to generate a set of key feature structures of the forest scene by using a complex scene numerical representation mode oriented towards personnel movement planning. The personnel movement sample database construction module 102 includes a personnel data indicator acquisition and processing unit 1021 and a trajectory stamp sample construction unit 1022. The personnel data indicator acquisition and processing unit 1021 is used to collect personnel characteristics and perform quantitative descriptions. The trajectory stamp sample construction unit 1022 is used to construct trajectory stamp samples based on spatiotemporal alignment to form a standardized personnel movement sample database. The personnel characteristics include load and physiological indicators. The travel speed modeling and path optimization module 103 includes a travel speed modeling unit 1031 and a path optimization unit 1032. The travel speed modeling unit 1031 uses various methods to construct a quantitative mapping function between the geographical environment, personnel elements and travel speed. The path optimization unit 1032 uses a traversal search algorithm to calculate the optimal path to a specified origin and destination, providing the shortest path planning for forest fire rescue teams in complex forest scenarios. The various methods for constructing the geographical environment include statistical analysis methods and deep learning methods.

7. A terminal comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method as described in any one of claims 1 to 5.

8. The terminal as described in claim 7 further includes a data collector and a display, characterized in that, The data acquisition device includes at least a thermometer, hygrometer, anemometer, trajectory measuring instrument, weighing instrument, and physiological measuring instrument; The display shows the path planning results.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 5.

Images