Intelligent path planning method and system for forest machine operation in hilly and mountainous regions based on satellite remote sensing

Abstract

The invention relates to the technical field of intelligent path planning, in particular to an intelligent path planning method and system for hilly and mountain forest machine operation based on satellite remote sensing, and is suitable for intelligent path planning of forestry mechanized operation under complex terrain conditions. Building a multi-dimensional energy consumption prediction model in combination with forest machine parameter data; an initial path is generated by adopting layered progressive path planning, and dynamic adjustment is performed by utilizing real-time environment data so as to adapt to a complex mountain environment; under the scene of multi-machine collaborative operation, the efficiency and safety of the final operation path are ensured through optimization of the adjusted path, the method accurately recognizes the dangerous area, optimizes the passing path, remarkably improves the safety and efficiency of forest machine operation, and is suitable for the complex operation environment of hills and mountains.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent path planning, and in particular to an intelligent path planning method and system for forestry machinery operations in hilly and mountainous areas based on satellite remote sensing, which is suitable for intelligent planning of forestry mechanized operation paths under complex terrain conditions. Background Art

[0002] Forestry mechanization is an important direction for the development of modern forestry. However, forestry machine operations in hilly and mountainous areas with complex terrain conditions place extremely high demands on path planning. Traditional forestry machine operation path planning mainly relies on manual experience, which has the following problems:

[0003] First, traditional path planning methods are difficult to adapt to the complex and changeable mountain terrain conditions. Hilly and mountainous areas have complex terrain, large slope changes, and changeable soil conditions. Traditional experience-based path planning cannot fully consider these factors, resulting in frequent safety accidents such as vehicle getting stuck and overturning during operations.

[0004] Secondly, traditional path planning methods lack systematic consideration of energy consumption. In mountainous operating environments, the energy consumption of forestry machinery is closely related to factors such as terrain, slope, and soil. Reasonable path planning can significantly reduce energy consumption. However, traditional methods often only consider the shortest distance or the shortest time, ignoring the issue of energy consumption optimization.

[0005] Thirdly, traditional path planning lacks environmental perception and dynamic adjustment capabilities. In mountainous environments, factors such as weather changes and terrain changes frequently affect operating conditions, and once the traditional method is planned, it is difficult to make real-time adjustments based on environmental changes.

[0006] In addition, when multiple forestry machines work together, traditional methods are difficult to achieve efficient coordination. The lack of a unified scheduling and conflict avoidance mechanism often leads to low operating efficiency.

[0007] Existing technologies such as CN115113616 B disclose a path planning method in an indoor environment, which divides the area into rectangular areas and covers them with a bow-shaped route. However, this method is only applicable to flat indoor floors and cannot cope with complex terrain changes in hilly and mountainous areas. It also lacks energy consumption optimization and environmental perception capabilities and is not suitable for forestry machinery operations.

[0008] Therefore, there is an urgent need for an intelligent path planning method and system that can make full use of modern remote sensing technology and comprehensively consider terrain characteristics, energy consumption optimization and environmental changes, so as to improve the operating efficiency and safety of forest machinery in hilly and mountainous areas. Summary of the invention

[0009] The purpose of the present invention is to overcome the shortcomings of the prior art and provide an intelligent path planning method and system for forestry machinery operations in hilly and mountainous areas based on satellite remote sensing. The method and system make full use of multi-source remote sensing data, combined with terrain accessibility assessment, energy consumption optimization, environmental perception and multi-machine collaboration technology, to achieve efficient, safe and energy-saving path planning for forestry machinery operations in hilly and mountainous environments.

[0010] The present invention proposes an intelligent path planning method for forestry machine operation in hilly and mountainous areas based on satellite remote sensing, comprising:

[0011] Obtain multi-source remote sensing data and forestry machinery parameter data;

[0012] Based on the multi-source remote sensing data, generating a terrain accessibility assessment model;

[0013] Constructing a multi-dimensional energy consumption prediction model based on the terrain accessibility assessment model and the forestry machinery parameter data;

[0014] Based on the multi-dimensional energy consumption prediction model, an initial path is generated through hierarchical progressive path planning;

[0015] Acquire real-time environmental data, and dynamically adjust the initial path according to the real-time environmental data to obtain an adjusted path;

[0016] When multiple forestry machines are operating simultaneously, multi-machine collaborative optimization is performed according to the adjusted path to obtain a final operating path.

[0017] Preferably, the obtaining of multi-source remote sensing data and forestry machinery parameter data comprises:

[0018] Acquire multi-source remote sensing data including satellite optical images, radar data, and aerial LiDAR point cloud data;

[0019] Obtain forest machine parameter data including weight, power, turning radius and traction;

[0020] The multi-source remote sensing data are preprocessed to obtain a high-precision digital elevation model, a soil property matrix and a tree distribution density matrix.

[0021] Preferably, generating a terrain accessibility assessment model based on the multi-source remote sensing data includes:

[0022] Construct a three-layer raster data structure including a high-precision digital elevation model, soil property matrix, and tree distribution density matrix;

[0023] Define the terrain accessibility scoring function T(x,y),

[0024] T(x,y)=w1·Slope(x,y)+w2·Roughness(x,y)+w3·SoilStrength(x,y)+w4·Vegetation(x,y)

[0025] Among them, w1, w2, w3 and w4 are adaptive weight coefficients dynamically adjusted according to factors such as season and rainfall, Slope(x, y) is the slope function, Roughness(x, y) is the terrain roughness function, SoilStrength(x, y) is the soil strength function, and Vegetation(x, y) is the vegetation distribution function;

[0026] Based on the types of forestry machinery, a traffic level matrix C is established to divide the terrain into accessible areas, restricted access areas, and prohibited access areas.

[0027] Generate a terrain accessibility grid map Tmap(x,y).

[0028] Preferably, a multi-dimensional energy consumption prediction model is constructed based on the terrain accessibility assessment model and the forestry machinery parameter data, including:

[0029] Construct a slope-sensitive energy consumption prediction function E(p1,p2) to calculate the energy consumption of the forest machine from point p1 to the adjacent point p2, where E(p1,p2) = E base ·d(p1,p2)·[1+k1·sin(θ)+k2·ΔH+k3·R(p2)], E base is the basic energy consumption coefficient, d(p1, p2) is the distance between the two points, θ is the angle between the travel direction and the contour line, ΔH is the elevation change, R(p2) is the terrain roughness of the target point, and k1, k2 and k3 are energy consumption influencing factors;

[0030] Set energy consumption limit E max and slope safety threshold θ max , ensure safe operation;

[0031] Generate the energy cost matrix Emap(x,y) as the core evaluation indicator of path search.

[0032] Preferably, based on the multi-dimensional energy consumption prediction model, an initial path is generated by hierarchical progressive path planning, including:

[0033] A quadtree structure is used to represent the operation area, and complex terrain areas are adaptively subdivided;

[0034] The improved A* algorithm is used for macroscopic path planning, and the heuristic function f(n)=g(n)+h(n)+λ·e(n) is constructed, where g(n) is the cost from the starting point to the current point, h(n) is the estimated cost from the current point to the target point, e(n) is the energy consumption factor, and λ is the energy consumption weight parameter.

[0035] Based on the macroscopic path, the dynamic programming method is applied to refine the local path segments, taking into account the steering radius constraint of the forest machine to generate a smooth path.

[0036] The terrain obstacle detection function B(path) is introduced to identify dangerous areas on the path and perform local replanning;

[0037] Generate a hierarchical path data structure PathTree including a main path and multi-level refined paths.

[0038] Preferably, the acquiring of real-time environmental data comprises:

[0039] Through the sensor system carried on the forestry machine, the airborne sensor data including visual data, radar detection data and inclination data are obtained;

[0040] Receive real-time update data from satellite remote sensing through the communication system;

[0041] The airborne sensor data and the satellite remote sensing real-time update data are integrated to obtain real-time environmental data.

[0042] Preferably, dynamically adjusting the initial path according to the real-time environmental data comprises:

[0043] Define the environmental change quantification function ΔEnv(t)=||E real (t)-E pred (t)||, where E real (t) is the real-time environmental data, E pred (t) Forecast environmental data;

[0044] When ΔEnv(t) is greater than the preset threshold τ, path adjustment is triggered;

[0045] Establish a three-level adjustment mechanism, including:

[0046] When the environment changes slightly, keep the main path unchanged and only adjust the local trajectory;

[0047] When the environmental changes are moderate, keep the starting and ending points unchanged and replan some path segments;

[0048] When the environmental changes exceed the safety threshold, the path is completely replanned;

[0049] Generate a dynamically adjusted new path PathTree' and calculate the deviation index from the original path.

[0050] Preferably, when multiple forestry machines are operating simultaneously, multi-machine collaborative optimization is performed according to the adjusted path, including:

[0051] Based on the Voronoi diagram and the principle of energy balance, the operation area is adaptively divided into Area(i)=f(P i ,E i ,T i ), where P i is the performance parameter of the forestry machine, E i is the remaining energy, T i Provide task completion time estimates;

[0052] Constructing the spatiotemporal conflict detection function Conflict(p i ,p j ), when the distance between two forest machines is less than the safety distance d in the same period of time safe Identify as a conflict;

[0053] When a conflict is detected, the path is reallocated based on the principle of minimizing energy consumption;

[0054] Design the task coordination function Collab(M,T) to realize the coordination modes such as capability complementation, regional relay and abnormal support;

[0055] Generate the multi-machine collaborative path set FinalPathSet and the job scheduling schedule Schedule.

[0056] Preferably, the method further comprises:

[0057] Real-time monitoring of forestry machinery operation status and progress;

[0058] When an abnormal situation occurs during the operation, the emergency plan will be automatically activated;

[0059] Record operation tracks and energy consumption data, and generate operation reports;

[0060] Based on historical operation data, the parameters of the terrain accessibility assessment model and the multi-dimensional energy consumption prediction model are optimized.

[0061] The intelligent path planning system for hilly and mountainous forestry machinery operations based on satellite remote sensing includes:

[0062] Data acquisition module, used to acquire multi-source remote sensing data and forest machine parameter data;

[0063] A terrain assessment module, used to generate a terrain accessibility assessment model including a terrain accessibility scoring function based on the multi-source remote sensing data;

[0064] An energy consumption prediction module, used to construct a multidimensional energy consumption prediction model including a slope-sensitive energy consumption prediction function according to the terrain accessibility assessment model and the forestry machinery parameter data;

[0065] A path planning module, used to generate an initial path including a main path and multi-level refined paths through hierarchical progressive path planning based on the multi-dimensional energy consumption prediction model;

[0066] Environmental perception module, used to obtain real-time environmental data including airborne sensor data and satellite remote sensing real-time update data;

[0067] A path adjustment module, used to dynamically adjust the initial path based on the degree of environmental change according to the real-time environmental data to obtain an adjusted path;

[0068] A multi-machine coordination module is used to perform multi-machine coordination optimization including conflict detection and task coordination according to the adjusted path when multiple forestry machines are operating simultaneously, so as to obtain a final operation path;

[0069] Communication module, used to transmit data between modules and exchange data with the forest machine control system;

[0070] Storage module, used to store remote sensing data, model parameters, planned paths and historical operation data;

[0071] Processor, used to perform calculation and control functions of each module.

[0072] The intelligent path planning method and system for forestry machine operation in hilly and mountainous areas based on satellite remote sensing provided by the present invention achieves the following beneficial effects through technologies such as multi-source remote sensing data fusion, terrain accessibility assessment, multi-dimensional energy consumption prediction, hierarchical progressive path planning, environmental perception and dynamic adjustment, and multi-machine collaborative optimization:

[0073] 1) Accurately identify dangerous areas and optimal paths in complex mountainous environments through multi-source remote sensing data fusion and terrain accessibility assessment, significantly improving operational safety;

[0074] 2) Through the slope-sensitive energy consumption prediction model and hierarchical progressive path planning, the energy-optimized path generation is achieved, and fuel consumption is reduced by 30% to 45%;

[0075] 3) Through real-time environmental perception and dynamic path adjustment mechanism, the system can respond to emergencies such as weather changes and terrain changes, improving system adaptability and robustness;

[0076] 4) Through multi-machine collaborative optimization, the conflict problem when multiple forestry machines are working at the same time is solved, and the operating efficiency is improved by more than 50%;

[0077] 5) It reduces equipment loss as a whole, prolongs equipment service life by about 25%, and improves the economic and ecological benefits of forestry mechanized operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] Figure 1 A flow chart of an intelligent path planning method for hilly and mountainous forestry machine operations based on satellite remote sensing provided by an embodiment of the present invention;

[0079] Figure 2 A schematic diagram of a multi-source remote sensing data fusion and preprocessing process provided by an embodiment of the present invention;

[0080] Figure 3 A schematic diagram of a terrain accessibility assessment model provided by an embodiment of the present invention;

[0081] Figure 4 A schematic diagram of a multi-dimensional energy consumption prediction model provided by an embodiment of the present invention;

[0082] Figure 5 A schematic diagram of hierarchical progressive path planning provided by an embodiment of the present invention;

[0083] Figure 6 A schematic diagram of environmental perception and dynamic path adjustment provided by an embodiment of the present invention;

[0084] Figure 7 A schematic diagram of multi-machine collaborative optimization provided by an embodiment of the present invention;

[0085] Figure 8 A structural block diagram of an intelligent path planning system for hilly and mountainous forestry machine operations based on satellite remote sensing provided by an embodiment of the present invention;

[0086] Fig. 9 A schematic diagram of a practical application scenario provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0087] Please refer to the attached Figure 1-9 , the present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0088] Example 1: Overall process of intelligent path planning method for hilly and mountainous forestry machine operation based on satellite remote sensing

[0089] See also Figure 1 This embodiment provides an intelligent path planning method for hilly and mountainous forestry machine operations based on satellite remote sensing. The overall process of the method includes:

[0090] Step S1: Acquire multi-source remote sensing data and forestry machinery parameter data.

[0091] Step S2: Generate a terrain accessibility assessment model based on the multi-source remote sensing data.

[0092] Step S3: constructing a multi-dimensional energy consumption prediction model based on the terrain accessibility assessment model and the forestry machinery parameter data.

[0093] Step S4: Based on the multi-dimensional energy consumption prediction model, an initial path is generated through hierarchical progressive path planning.

[0094] Step S5: acquiring real-time environment data, and dynamically adjusting the initial path according to the real-time environment data to obtain an adjusted path.

[0095] Step S6: When multiple forestry machines are operating simultaneously, multi-machine collaborative optimization is performed according to the adjusted path to obtain a final operation path.

[0096] like Figure 1 As shown in the figure, this method first establishes a basic data set of the operating area through multi-source remote sensing data and forest machine parameters, and then builds a terrain accessibility assessment model based on these data as a safety guarantee basis for path planning. Next, a multidimensional energy consumption prediction model is constructed in combination with the characteristics of forest machines, so that path planning not only considers safety but also optimizes energy efficiency. On this basis, a hierarchical progressive path planning algorithm is used to generate the initial path to ensure the global optimality and local feasibility of the path. Subsequently, the system continuously obtains real-time environmental data, dynamically adjusts the initial path according to environmental changes, and improves the system's adaptability to environmental changes. Finally, when multiple forest machines are operating at the same time, the system will also perform multi-machine collaborative optimization to avoid conflicts and improve overall operating efficiency.

[0097] Through this progressive and mutually coordinated method, the present invention achieves safe, efficient and energy-saving path planning for forestry machinery operations in hilly and mountainous environments, effectively overcoming the limitations of traditional methods in complex terrain.

[0098] Embodiment 2: Method for acquiring multi-source remote sensing data and forest machine parameter data

[0099] like Figure 2 As shown, this embodiment describes in detail a method for acquiring multi-source remote sensing data and forest machine parameter data, specifically including:

[0100] Acquire multi-source remote sensing data including satellite optical images, radar data and aerial LiDAR point cloud data. Specifically, satellite optical images mainly come from high-resolution remote sensing satellites (such as GF-2, WorldView-3, etc.), with a resolution of preferably 0.5-2 meters; radar data mainly comes from synthetic aperture radar (SAR) satellites (such as Sentinel-1, TerraSAR-X, etc.), which are used to obtain soil moisture and surface cover; aerial LiDAR point cloud data is obtained through the laser radar system carried by drones or manned aircraft, and the point cloud density is preferably 8-20 points per square meter.

[0101] Obtain the forestry machine parameter data including the weight, power, turning radius and traction. The forestry machine parameter data is mainly obtained from the forestry machine's technical manual and actual tests, including but not limited to: the forestry machine's overall weight (tons), engine power (kW), minimum turning radius (meters), maximum traction (kN), track or tire width (meters), ground contact pressure (kPa), maximum climbing grade (degrees), etc. These parameters are important basic data for subsequent energy consumption prediction and path planning.

[0102] The multi-source remote sensing data is preprocessed to obtain a high-precision digital elevation model, a soil property matrix, and a tree distribution density matrix. The preprocessing process includes: geometric correction, atmospheric correction, and radiation correction for satellite images; scattering correction and terrain correction for radar data; noise removal, ground point / non-ground point classification, and coordinate conversion for LiDAR point clouds. Through these preprocessing, valuable terrain feature information is extracted from the original data to form three key data layers: a high-precision digital elevation model (DEM), a soil property matrix (including parameters such as moisture, hardness, and bearing capacity), and a tree distribution density matrix.

[0103] Preferably, the three data layers use a unified geographic coordinate system and the same spatial resolution (usually 5 meters grid) to facilitate subsequent processing and analysis. These pre-processed data will serve as the basic input for terrain accessibility assessment.

[0104] Example 3: Method for generating terrain accessibility assessment model

[0105] like Figure 3 As shown, this embodiment describes in detail a method for generating a terrain accessibility assessment model, which specifically includes:

[0106] A three-layer raster data structure is constructed, including a high-precision digital elevation model, a soil property matrix, and a tree distribution density matrix. This data structure uses a unified spatial reference system, usually the local UTM projection coordinate system, to ensure the accuracy of distance and area calculations. The three layers of data are completely spatially corresponding, and each raster cell contains elevation values, soil property values, and tree density values.

[0107] Define the terrain accessibility scoring function T(x,y), and its calculation formula is:

[0108] T(x,y)=w1·Slope(x,y)+w2·Roughness(x,y)+w3·SoilStrength(x,y)+w4·Vegetation(x,y)

[0109] Among them, w1, w2, w3 and w4 are adaptive weight coefficients that are dynamically adjusted according to factors such as season and rainfall, usually satisfying w1+w2+w3+w4=1, and all weight values ​​are positive numbers; Slope(x,y) is a slope function, which represents the terrain slope value at the position (x,y), usually calculated by DEM data; Roughness(x,y) is a terrain roughness function, which represents the flatness of the surface, usually calculated by the local variance of DEM; SoilStrength(x,y) is a soil strength function, which represents the bearing capacity of the soil and is extracted through the soil attribute matrix; Vegetation(x,y) is a vegetation distribution function, which represents the impact of tree density on traffic and is calculated through the tree distribution density matrix.

[0110] Preferably, in the dry season, the weight configuration can be set to w1 = 0.4, w2 = 0.3, w3 = 0.1, w4 = 0.2; and in the rainy season, due to significant changes in soil conditions, the weight configuration can be adjusted to w1 = 0.3, w2 = 0.2, w3 = 0.4, w4 = 0.1. The system will automatically adjust these weights according to real-time meteorological data to adapt to different environmental conditions.

[0111] Based on the type of forestry machine, the access level matrix C is established to divide the terrain into passable area, restricted access area and prohibited access area. Specifically, when the T(x,y) value is less than the threshold T1, the area is marked as a passable area (C1); when the T(x,y) value is between the threshold T1 and t2, the area is marked as a restricted access area (C2); when the T(x,y) value is greater than the threshold T2, the area is marked as a prohibited access area (C3). The setting of thresholds T1 and T2 is related to the specific type of forestry machine, and the threshold settings for heavy forestry machines and light forestry machines are different.

[0112] Generate a terrain accessibility grid map Tmap(x,y). This map divides the entire work area into areas of different access levels, represented by different colors: green for accessible areas, yellow for restricted access areas, and red for prohibited access areas. This visual representation intuitively shows the accessibility of the work area, which is convenient for reference in subsequent path planning.

[0113] The output result Tmap(x,y) of the terrain accessibility assessment model will be used as a safety constraint for subsequent path planning to ensure that the forest machine always travels within a safe area.

[0114] Example 4: Method for constructing a multidimensional energy consumption prediction model

[0115] like Figure 4 As shown, this embodiment describes in detail a method for constructing a multidimensional energy consumption prediction model, which specifically includes:

[0116] The slope-sensitive energy consumption prediction function E(1, p2) is constructed to calculate the energy consumption of the forest machine from point p1 to the adjacent point p2. The calculation formula is:

[0117] E(1,p2)=E base ·d(p1,p2)·[1+k1·sin(θ)+k2·ΔH+k3·R(p2)],

[0118] Among them, E base is the basic energy consumption coefficient, with the unit of kilojoules / meter, which indicates the basic energy consumption of the forest machine for every meter it moves on flat terrain. This value is related to the weight and power of the forest machine. d(p1, p2) is the Euclidean distance between two points, with the unit of meter. θ is the angle between the moving direction and the contour line, with a value range of [-π / 2, π / 2]. When the forest machine moves along the contour line, θ = 0, when the forest machine moves upward perpendicular to the contour line, θ = π / 2, and when the forest machine moves downward perpendicular to the contour line, θ = -π / 2. ΔH is the elevation change, with the unit of meter, indicating the height difference from p1 to p2. R(p2) is the terrain roughness at the target point p2, which is dimensionless and usually ranges from [0, 1]. k1, k2 and k3 are energy consumption influencing factors, which respectively indicate the degree of influence of slope aspect, elevation change and terrain roughness on energy consumption.

[0119] Preferably, for typical forestry machines, the value range of k1 is [0.5, 1.5], the value range of k2 is [0.1, 0.3], and the value range of k3 is [0.2, 0.6]. These parameters can be calibrated through actual test data to adapt to different types of forestry machines.

[0120] Set energy consumption limit E max and slope safety threshold θ max , to ensure safe operation. Energy consumption limit E max It is usually set to 3-5 times the basic energy consumption to identify the path sections with abnormally high energy consumption; the slope safety threshold θ max It is determined based on the maximum climbing ability of the forest machine and is usually set to 25-35 degrees. max or the slope exceeds θ max When the path segment is marked as infeasible, the machine will be prevented from getting stuck in high energy consumption or dangerous areas.

[0121] Generate an energy cost matrix Emap(x,y) as the core evaluation index for path search. For each grid cell (x,y) in the operation area, calculate the energy consumption value from the cell to its eight adjacent cells to form an energy cost matrix. This matrix will be used in subsequent path planning to evaluate the energy cost of different paths and achieve the path selection with the best energy consumption.

[0122] The multidimensional energy consumption prediction model not only takes into account the distance factor, but also comprehensively considers the impact of slope aspect, elevation change and terrain roughness on energy consumption, so that path planning can minimize energy consumption and improve the economic benefits of forest machinery operations while ensuring safety.

[0123] Example 5: Hierarchical progressive path planning method

[0124] like Figure 5 As shown, this embodiment describes in detail the hierarchical progressive path planning method, which specifically includes:

[0125] The quadtree structure is used to represent the operating area, and the complex terrain area is adaptively subdivided. The quadtree is a hierarchical spatial data structure that recursively divides the operating area into four sub-areas until the preset accuracy requirements are met. In areas with complex terrain and drastically changing traffic conditions, the system automatically performs a more detailed division; while in areas with flat terrain and stable traffic conditions, a coarser division is used to achieve a balance between accuracy and computational efficiency.

[0126] The improved A* algorithm is used for macroscopic path planning to construct the heuristic function f(n), whose calculation formula is:

[0127] f(n)=g(n)+h(n)+λ·e(n),

[0128] Among them, g(n) is the actual cost from the starting point to the current node n, usually expressed by the length of the path that has been traveled; h(n) is the estimated cost from the current node n to the target point, usually expressed by the Euclidean distance; e(n) is the energy consumption factor, which represents the energy consumption evaluation value of the current node n, obtained through the energy consumption cost matrix Emap; λ is the energy consumption weight parameter, which is used to adjust the importance of energy consumption in path planning, and the value range is usually [0.5,2]. It is worth noting that the traditional A* algorithm only considers the two factors of g(n) and h(n), while the present invention introduces the energy consumption factor e(n) to enable path planning to give priority to paths with lower energy consumption.

[0129] Preferably, the value of λ can be adjusted according to the different requirements of the task: when the task prioritizes energy saving, λ can be set to a larger value (such as 1.5-2); when the task prioritizes timeliness, λ can be set to a smaller value (such as 0.5-1).

[0130] Based on the macro path, the dynamic programming method is applied to refine the local path segments, taking into account the turning radius constraint of the forestry machine to generate a smooth path. The path obtained by macro planning may have a turning radius that is too small and does not conform to the motion characteristics of the forestry machine. It is necessary to make the path smoother through local optimization. The dynamic programming method divides the path into segments, and each segment of the path finds a smooth path with optimal energy consumption while satisfying the turning radius constraint of the forestry machine.

[0131] The terrain obstacle detection function B(path) is introduced to identify dangerous areas on the path and perform local replanning. The terrain obstacle detection function B(path) checks along the planned path whether there are obstacles or dangerous areas that are not identified by the terrain accessibility assessment model. When such areas are detected, the system automatically searches for alternative paths around the area to ensure that the forest machine can pass safely.

[0132] Generate a hierarchical path data structure PathTree containing the main path and multi-level refined paths. Path Tree is a tree data structure whose root node is the main path and each child node represents a refinement of a section of the main path. This structure enables the system to access and adjust the path at different accuracy levels according to actual needs, improving the flexibility and adaptability of path planning.

[0133] The hierarchical progressive path planning method achieves the unity of global optimality and local feasibility through progressive processing from macro to micro. The generated path not only meets the global goal of optimal energy consumption, but also can adapt to the movement characteristics and terrain complexity of the forest machine.

[0134] Embodiment 6: Method for obtaining real-time environmental data

[0135] like Figure 6 As shown, this embodiment describes in detail a method for obtaining real-time environmental data, which specifically includes:

[0136] The sensor system on the forestry machine can be used to obtain onboard sensor data including visual data, radar detection data and inclination data. The forestry machine sensor system usually includes: high-definition cameras (forward, rear and surround) to obtain visual information of the surrounding environment; millimeter-wave radar to detect the distance and speed of obstacles in front; laser radar (optional) to generate a three-dimensional point cloud of the surrounding environment; IMU (inertial measurement unit) to measure the attitude, inclination and acceleration of the forestry machine; GPS / RTK positioning system to obtain the precise position of the forestry machine. These sensors collect data in real time at a frequency of 10-30Hz, providing basic information for environmental perception.

[0137] Receive satellite remote sensing real-time update data through the communication system. The communication system adopts a multi-level communication architecture, including short-distance communication (such as WiFi, Bluetooth), medium-distance communication (such as 5G / 4G mobile network) and long-distance communication (such as satellite communication). In areas with good signal coverage, the system gives priority to using 5G / 4G networks to receive meteorological data, remote sensing data updates and other environmental information; in areas with weak or no signal, it switches to satellite communication mode to ensure the continuity of data transmission. Satellite remote sensing real-time update data mainly includes the latest meteorological information, terrain change information and emergency warning information.

[0138] The airborne sensor data and the satellite remote sensing real-time update data are fused to obtain real-time environmental data. Data fusion uses a multi-sensor information fusion algorithm to integrate data from different sources and types into a unified environmental perception result. During the fusion process, the data will be synchronized in time, spatially aligned, and evaluated for reliability to remove noise and outliers, thereby improving the accuracy and reliability of environmental perception.

[0139] Preferably, data fusion uses an algorithm based on Kalman filtering, which can effectively handle sensor delays, errors, and failures. The fused real-time environmental data will be used for subsequent dynamic path adjustments, enabling the system to respond to environmental changes in a timely manner.

[0140] Embodiment 7: Path dynamic adjustment method based on real-time environmental data

[0141] like Figure 6 As shown, this embodiment describes in detail a method for dynamically adjusting a path based on real-time environmental data, specifically including:

[0142] Define the environmental change quantification function ΔEnv(t), and its calculation formula is:

[0143] ΔEnv(t)=||E real (t)-E pred (t)||,

[0144] Among them, E real (t) is the real-time environmental data at time t, including actual observation values ​​of terrain, soil, vegetation, etc.; E pred (t) is the predicted environmental data at time t, that is, the predicted value of the environmental model used in path planning; ||·|| represents the vector norm, which is used to calculate the degree of difference between two environmental data sets. The environmental change quantification function converts environmental changes into quantifiable values, which makes it easier for the system to automatically determine whether the path needs to be adjusted.

[0145] When ΔEnv(t) is greater than the preset threshold τ, path adjustment is triggered. The setting of the threshold τ requires a balance between the adjustment frequency and the system response sensitivity, and is usually set according to the environmental stability of the operating area and the task requirements. In areas with large environmental changes (such as changeable meteorological conditions in mountainous areas), τ can be set smaller to increase the system's sensitivity to environmental changes; in areas with relatively stable environments, τ can be set larger to reduce unnecessary path adjustments.

[0146] A three-level adjustment mechanism is established, including: when the environmental changes are small, keep the main path unchanged and only adjust the local trajectory; when the environmental changes are moderate, keep the starting and ending points unchanged and re-plan some path segments; when the environmental changes exceed the safety threshold, completely re-plan the path. The three-level adjustment mechanism adopts different levels of adjustment strategies according to the degree of environmental changes, minimizing the computational cost and path changes caused by the adjustment while ensuring safety.

[0147] Specifically, when ΔEnv(t) is between τ and 1.5τ, the system performs a first-level adjustment and only fine-tunes the local trajectory, such as avoiding obstacles in a small range or optimizing trajectory smoothness; when ΔEnv(t) is between 1.5τ and 2.5τ, the system performs a second-level adjustment and re-plans some path segments while keeping the starting and ending points unchanged; when ΔEnv(t) is greater than 2.5τ, the system performs a third-level adjustment and completely re-plans the entire path to cope with significant environmental changes.

[0148] Generate a dynamically adjusted new path PathTree' and calculate the deviation index from the original path. The adjusted path still maintains the data structure of PathTree, but the content has been updated according to environmental changes. The system also calculates the deviation index between the new and old paths, including the path length change rate, energy consumption estimate change rate, and key point offset degree. These indicators are used to evaluate the necessity and effect of adjustments on the one hand, and provide reference information for forest machine operators on the other.

[0149] The dynamic path adjustment method enables the system to respond to environmental changes in real time, avoid risks in a timely manner, maintain the optimality and safety of the path, and greatly improves the system's adaptability and reliability in complex and changeable mountain environments.

[0150] Example 8: Multi-machine collaborative optimization method

[0151] like Figure 7 As shown, this embodiment describes in detail the multi-machine collaborative optimization method, which specifically includes:

[0152] Based on the Voronoi diagram and the principle of energy consumption balance, the operation area is adaptively divided, and the calculation formula is:

[0153] Area(i)=f(P i ,E i ,T i ),

[0154] Where Area(i) represents the operating area assigned to the i-th forest machine; P i E is the performance parameter of the forestry machinery, including operation efficiency, operation accuracy, etc. i T is the remaining energy, indicating the remaining fuel or electricity of the forest machine; iThe task completion time estimate represents the time required to complete the assigned task, and the time required to complete the assigned task. The division of the work area will take these three factors into consideration to ensure that the task load of each forest machine matches its capacity and avoid the situation where some forest machines are overloaded while others are idle.

[0155] Preferably, the operation area division adopts an improved Voronoi diagram algorithm, which introduces the capacity weight and energy consumption balance factor on the basis of the traditional Voronoi diagram, making the area division more reasonable. For forestry machines with strong capacity and more residual energy, a larger operation area is allocated; for forestry machines with weak capacity and less residual energy, a smaller operation area is allocated.

[0156] Constructing the spatiotemporal conflict detection function Conflict(p i ,p j ), when the distance between two forest machines is less than the safety distance d in the same period of time safe The spatial-temporal conflict detection not only considers the spatial position, but also the time factor, that is, predicting the positions of different forestry machines at different time points and judging whether they will appear in adjacent positions at the same time. safe It is usually set to 2-3 times the size of the forest machine to ensure sufficient safety margin.

[0157] When a conflict is detected, the path is reallocated based on the principle of minimizing energy consumption. Conflict resolution strategies include: time staggering, that is, adjusting the speed of the forestry machines so that they pass through potential conflict points at different times; path offset, that is, appropriately adjusting the path of one or two forestry machines to avoid spatial conflicts; task reallocation, that is, redividing the operating area in extreme cases to fundamentally eliminate the possibility of conflict. The system will choose the strategy with the least increase in energy consumption to resolve the conflict.

[0158] The task coordination function Collab(M,T) is designed to realize the coordination modes such as capability complementation, regional relay and abnormal support. The task coordination function designs the optimal coordination strategy according to the type of forestry machine, operation content and environmental conditions. The capability complementation mode is suitable for the joint operation of different types of forestry machines, such as the cooperation between harvesters and transport vehicles; the regional relay mode is suitable for large-scale continuous operation, and the next forestry machine can seamlessly connect after the previous forestry machine completes the area; the abnormal support mode dispatches nearby forestry machines to provide assistance when a forestry machine encounters difficulties or failures.

[0159] Generate the multi-machine collaborative path set FinalPathSet and the operation scheduling schedule Schedule. The multi-machine collaborative path set contains the final operation paths of all forestry machines. These paths have been conflict-detected and collaboratively optimized to ensure safe and efficient collaboration among multiple machines. The operation scheduling schedule specifies in detail the operation time, speed, and key point passing time of each forestry machine, providing precise operation guidance for forestry machine operators.

[0160] The multi-machine collaborative optimization method solves the problems of task allocation, conflict avoidance and collaborative operation when multiple forestry machines operate simultaneously, significantly improving the overall operation efficiency and resource utilization.

[0161] Embodiment 9: Method for realizing additional functions

[0162] This embodiment describes in detail the additional function implementation method of the present invention, which specifically includes:

[0163] Real-time monitoring of the operation status and progress of the forestry machine. The system collects and transmits information such as the location, speed, fuel consumption, and operation completion of the forestry machine in real time through sensors and communication modules on the forestry machine. This information is centrally displayed and analyzed in the control center, allowing managers to grasp the operation status in real time and discover and solve problems in a timely manner.

[0164] When an abnormal situation occurs during the operation, the emergency plan is automatically activated. The system defines a variety of abnormal situations and their handling plans, such as forest machine failure, extreme weather, terrain collapse, etc. When an abnormal situation is detected, the system automatically activates the corresponding emergency plan, such as sending an alarm, adjusting the path, arranging rescue, etc., to minimize losses and risks.

[0165] Record operation trajectory and energy consumption data and generate operation reports. The system will record the actual movement trajectory, energy consumption, operation time and other data of the forest machine in detail, and compare and analyze them with the planned data to generate a detailed operation report. The report content includes operation efficiency evaluation, energy consumption analysis, problem summary and improvement suggestions, etc., providing reference for subsequent operations.

[0166] Based on historical operation data, the parameters of the terrain accessibility assessment model and the multi-dimensional energy consumption prediction model are optimized. The system collects and analyzes historical operation data, and automatically adjusts the model parameters through machine learning algorithms to make the model prediction results more consistent with the actual situation. This continuous optimization mechanism enables the system to continuously learn and adapt to different operating environments, improving the accuracy and efficiency of planning.

[0167] Preferably, the parameter optimization adopts a combination of gradient descent and genetic algorithm, which can converge quickly and avoid falling into local optimum. The continuously optimized model can more accurately predict terrain conditions and energy consumption, providing more reliable basic data for path planning.

[0168] Example 10: Intelligent path planning system for hilly and mountainous forestry machine operations based on satellite remote sensing

[0169] like Figure 8 As shown, this embodiment provides an intelligent path planning system for hilly and mountainous forestry machine operations based on satellite remote sensing, the system comprising:

[0170] Data acquisition module 1 is used to acquire multi-source remote sensing data and forestry machine parameter data. This module is responsible for collecting and preprocessing data from various data sources, including satellite remote sensing data receiving unit, aerial data acquisition unit, forestry machine parameter library access unit and data preprocessing unit. The data acquisition module converts the raw data into a standardized data format, providing a basis for subsequent module processing.

[0171] The terrain assessment module 2 is used to generate a terrain accessibility assessment model including a terrain accessibility scoring function based on the multi-source remote sensing data. The module includes a terrain feature extraction unit, a soil condition analysis unit, a vegetation distribution processing unit and a accessibility scoring calculation unit. The terrain assessment module integrates and analyzes the multi-source data to generate a terrain accessibility grid map representing the accessibility difficulty of different areas.

[0172] The energy consumption prediction module 3 is used to construct a multidimensional energy consumption prediction model including a slope-sensitive energy consumption prediction function based on the terrain accessibility evaluation model and the forest machine parameter data. The module includes an energy consumption basic parameter calculation unit, a slope impact analysis unit, a terrain impact evaluation unit, and an energy consumption cost matrix generation unit. The energy consumption prediction module can accurately estimate the energy consumption of the forest machine under different terrain conditions, and provide a basis for energy consumption optimization path planning.

[0173] The path planning module 4 is used to generate an initial path including a main path and a multi-level refined path through hierarchical progressive path planning based on the multi-dimensional energy consumption prediction model. The module includes a regional quadtree partition unit, a macro path search unit, a local path optimization unit and an obstacle detection unit. The path planning module adopts a hierarchical progressive strategy to ensure the feasibility and smoothness of the local path while ensuring the global optimum.

[0174] The environment perception module 5 is used to obtain real-time environmental data including airborne sensor data and satellite remote sensing real-time update data. This module includes an airborne sensor data acquisition unit, a satellite data receiving unit, a data fusion processing unit and an environmental change detection unit. The environment perception module can monitor the changes in the operating environment in real time and provide a basis for dynamic path adjustment.

[0175] The path adjustment module 6 is used to dynamically adjust the initial path based on the degree of environmental change according to the real-time environmental data to obtain an adjusted path. This module includes an environmental change assessment unit, an adjustment strategy selection unit, a local path replanning unit, and a global path replanning unit. The path adjustment module can adopt an adjustment strategy of a corresponding level according to the degree of environmental change to maintain the optimality and safety of the path.

[0176] The multi-machine coordination module 7 is used to perform multi-machine coordination optimization including conflict detection and task coordination according to the adjusted path when multiple forestry machines are operating at the same time, so as to obtain the final operation path. The module includes an operation area division unit, a conflict detection and resolution unit, a task coordination strategy unit and a scheduling plan generation unit. The multi-machine coordination module can coordinate the operations of multiple forestry machines, avoid conflicts and improve overall efficiency.

[0177] The communication module 8 is used to transmit data between modules and exchange data with the forest machine control system. The module supports multiple communication modes, including wired network, wireless LAN, mobile communication network and satellite communication, to ensure the reliability and real-time performance of data transmission.

[0178] Storage module 9 is used to store remote sensing data, model parameters, planned paths and historical operation data. This module adopts a hierarchical storage architecture to store data of different types and frequencies in different types of storage devices to optimize storage efficiency and access speed.

[0179] The processor 10 is used to execute the calculation and control functions of each module. The processor can be a general-purpose CPU, or a dedicated DSP or FPGA. The appropriate processing hardware is selected according to actual needs to ensure the real-time performance and reliability of the system.

[0180] The modules of this system work closely together, and data flows smoothly, forming a complete intelligent path planning solution. The system realizes the intelligent processing of the entire process from data acquisition, terrain assessment, energy consumption prediction to path planning, environmental perception, path adjustment and multi-machine collaboration, providing efficient, safe and energy-saving path planning services for forestry machinery operations in hilly and mountainous areas.

[0181] Example 11: Application Case

[0182] like Fig. 9 As shown, this embodiment shows a practical application case, verifying the feasibility and effectiveness of the present invention. This case selected a hilly forest area in Sanming City, Fujian Province as the test site, with an area of ​​about 350 hectares, an altitude of 200-800 meters, an average slope of 25-35 degrees, and the main tree species are Chinese fir and Masson pine.

[0183] First, the system acquired multi-source remote sensing data including high-resolution satellite images, SAR radar data, and aerial LiDAR point clouds, as well as detailed parameters of the test forest machine. Through data preprocessing, a 5-meter resolution digital elevation model, soil property matrix, and tree distribution density matrix were generated.

[0184] Based on these data, the system built a terrain accessibility assessment model, dividing the test area into accessible areas (68%), restricted access areas (22%), and prohibited access areas (10%). At the same time, a multi-dimensional energy consumption prediction model was built to accurately predict the energy consumption of different paths.

[0185] The system uses a hierarchical progressive path planning method to generate an initial operation path. During the actual operation, the system obtains real-time environmental data through the sensor system and satellite communication system on the forest machine. When it detects that local rainfall has caused changes in soil conditions, the system automatically triggers path adjustment and successfully avoids potential risk areas for the vehicle to get stuck.

[0186] In the multi-machine coordination test, the system coordinated three different types of forestry machines to operate simultaneously, and achieved efficient coordination through reasonable area division and conflict avoidance mechanism. Compared with traditional manual planning, the overall operation efficiency was improved by 47.3%, fuel consumption was reduced by 31.5%, and equipment wear was reduced by 23.8%.

[0187] The test results show that the method and system provided by the present invention can effectively solve the path planning problem in forestry machinery operations in hilly and mountainous areas, improve the safety, efficiency and economy of operations, and have good practical value and promotion prospects.

[0188] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. An intelligent path planning method for forestry machine operation in hilly and mountainous areas based on satellite remote sensing, characterized in that: include: Obtain multi-source remote sensing data and forestry machinery parameter data; Based on the multi-source remote sensing data, generating a terrain accessibility assessment model; Constructing a multi-dimensional energy consumption prediction model based on the terrain accessibility assessment model and the forestry machinery parameter data; Based on the multi-dimensional energy consumption prediction model, an initial path is generated through hierarchical progressive path planning; Acquire real-time environmental data, and dynamically adjust the initial path according to the real-time environmental data to obtain an adjusted path; When multiple forestry machines are operating simultaneously, multi-machine collaborative optimization is performed according to the adjusted path to obtain a final operating path.

2. The method according to claim 1, characterized in that The obtaining of multi-source remote sensing data and forest machine parameter data includes: Acquire multi-source remote sensing data including satellite optical images, radar data, and aerial LiDAR point cloud data; Obtain forest machine parameter data including weight, power, turning radius and traction; The multi-source remote sensing data are preprocessed to obtain a high-precision digital elevation model, a soil property matrix and a tree distribution density matrix.

3. The method according to claim 1, characterized in that Based on the multi-source remote sensing data, a terrain accessibility assessment model is generated, including: Construct a three-layer raster data structure including a high-precision digital elevation model, soil property matrix, and tree distribution density matrix; Define the terrain accessibility scoring function T(x,y), where T(x,y)=w1·Slope(x,y)+w2·Roughness(x,y)+w3·SoilStrength(x,y)+w4·Vegetation(x,y) Among them, w1, w2, w3 and w4 are adaptive weight coefficients dynamically adjusted according to factors such as season and rainfall, Slope(x,y) is the slope function, Roughness(x,y) is the terrain roughness function, SoilStrength(x,y) is the soil strength function, and Vegetation(x,y) is the vegetation distribution function; Based on the forest machine type, a traffic level matrix C is established to divide the terrain into the passable area, restricted pass area and prohibited pass area; Generate a terrain accessibility grid map Tmap(x,y).

4. The method according to claim 1, characterized in that: According to the terrain accessibility assessment model and the forestry machinery parameter data, a multi-dimensional energy consumption prediction model is constructed, including: Construct a slope-sensitive energy consumption prediction function E(p1,p2) to calculate the energy consumption of the forest machine from point p1 to the adjacent point p2, where E(p1,p2) = E base ·d(p1,p2)·[1+k1·sin(θ)+k2·ΔH+k3·R(p2)], E base is the basic energy consumption coefficient, d(p1, p2) is the distance between the two points, θ is the angle between the travel direction and the contour line, ΔH is the elevation change, R(p2) is the terrain roughness of the target point, and k1, k2 and k3 are energy consumption influencing factors; Set energy consumption limit E max and slope safety threshold θ max , ensure safe operation; Generate the energy cost matrix Emap(x,y) as the core evaluation indicator of path search.

5. The method according to claim 1, characterized in that Based on the multi-dimensional energy consumption prediction model, an initial path is generated through hierarchical progressive path planning, including: A quadtree structure is used to represent the operation area, and complex terrain areas are adaptively subdivided; The improved A* algorithm is used for macroscopic path planning, and the heuristic function f(n)=g(n)+h(n)+λ·e(n) is constructed, where g(n) is the cost from the starting point to the current point, h(n) is the estimated cost from the current point to the target point, e(n) is the energy consumption factor, and λ is the energy consumption weight parameter. Based on the macroscopic path, the dynamic programming method is applied to refine the local path segments, taking into account the steering radius constraint of the forest machine to generate a smooth path. The terrain obstacle detection function B(path) is introduced to identify dangerous areas on the path and perform local replanning; Generate a hierarchical path data structure PathTree including a main path and multi-level refined paths.

6. The method according to claim 1, characterized in that The obtaining of real-time environmental data includes: Through the sensor system carried on the forestry machine, the airborne sensor data including visual data, radar detection data and inclination data are obtained; Receive real-time update data from satellite remote sensing through the communication system; The airborne sensor data and the satellite remote sensing real-time update data are integrated to obtain real-time environmental data.

7. The method according to claim 1, characterized in that Dynamically adjusting the initial path according to the real-time environmental data includes: Define the environmental change quantification function ΔEnv(t)=||E real (t)-E pred (t)||, where E real (t) is the real-time environmental data, E pred (t) Forecast environmental data; When ΔEnv(t) is greater than the preset threshold τ, path adjustment is triggered; Establish a three-level adjustment mechanism, including: When the environment changes slightly, keep the main path unchanged and only adjust the local trajectory; When the environmental changes are moderate, keep the starting and ending points unchanged and replan some path segments; When the environmental changes exceed the safety threshold, the path is completely replanned; Generate a dynamically adjusted new path PathTree' and calculate the deviation index from the original path.

8. The method according to claim 1, characterized in that When multiple forestry machines are operating simultaneously, multi-machine collaborative optimization is performed according to the adjusted path, including: Based on the Voronoi diagram and the principle of energy balance, the operation area is adaptively divided into Area(i)=f(P i ,E i ,T i ), where P i is the performance parameter of the forestry machine, E i is the remaining energy, T i Provide task completion time estimates; Constructing the spatiotemporal conflict detection function Conflict(p i ,p j ), when the distance between two forest machines is less than the safety distance d in the same period of time safe Identify as a conflict; When a conflict is detected, the path is reallocated based on the principle of minimizing energy consumption; Design the task coordination function Collab(M,T) to realize the coordination modes such as capability complementation, regional relay and abnormal support; Generate the multi-machine collaborative path set FinalPathSet and the job scheduling schedule Schedule.

9. The method according to claim 1, characterized in that: The method further comprises: Real-time monitoring of forestry machinery operation status and progress; When an abnormal situation occurs during the operation, the emergency plan will be automatically activated; Record operation tracks and energy consumption data, and generate operation reports; Based on historical operation data, the parameters of the terrain accessibility assessment model and the multi-dimensional energy consumption prediction model are optimized.

10. An intelligent path planning system for hilly and mountainous forestry machine operations based on satellite remote sensing that implements any of the above methods of claims 1-9, characterized in that: include: Data acquisition module, used to acquire multi-source remote sensing data and forest machine parameter data; A terrain assessment module, used to generate a terrain accessibility assessment model including a terrain accessibility scoring function based on the multi-source remote sensing data; An energy consumption prediction module, used to construct a multidimensional energy consumption prediction model including a slope-sensitive energy consumption prediction function according to the terrain accessibility assessment model and the forestry machinery parameter data; A path planning module, used to generate an initial path including a main path and multi-level refined paths through hierarchical progressive path planning based on the multi-dimensional energy consumption prediction model; Environmental perception module, used to obtain real-time environmental data including airborne sensor data and satellite remote sensing real-time update data; A path adjustment module, used to dynamically adjust the initial path based on the degree of environmental change according to the real-time environmental data to obtain an adjusted path; A multi-machine coordination module is used to perform multi-machine coordination optimization including conflict detection and task coordination according to the adjusted path when multiple forestry machines are operating simultaneously, so as to obtain a final operation path; Communication module, used to transmit data between modules and exchange data with the forest machine control system; Storage module, used to store remote sensing data, model parameters, planned paths and historical operation data; Processor, used to perform calculation and control functions of each module.

Images