Construction machine transportation method, device, equipment and medium based on terrain data

Abstract

Embodiments of the present disclosure disclose a construction machinery transportation method and device based on terrain data, equipment and medium. A specific embodiment of the method comprises: constructing a three-dimensional scene model of a construction-related road section according to acquired terrain data, mapping data and overhead line data of the construction-related road section; generating an initial arrangement state of a target construction machinery; simulating movement of the target construction machinery along a preset route in the three-dimensional scene model to generate an initial transportation scheme; for the initial transportation scheme, performing a first step: performing automatic collision checking on the movement process of the target construction machinery; in response to the collision detection result meeting a preset condition, performing a second step: adjusting the transportation path and transportation posture; generating a corresponding transportation instruction sequence and executing transportation of the target construction machinery. This embodiment can realize automatic obstacle avoidance for construction transportation, thereby reducing collision accidents and manual intervention.

Description

Technical Field

[0001] The embodiments of this disclosure relate to the field of computer technology, and more specifically to construction machinery transportation methods, apparatus, equipment, and media based on terrain data. Background Technology

[0002] Currently, in the planning of transportation for construction machinery in power grid construction, mountain wind power, and complex terrain areas, the transportation routes and machinery postures are usually determined by combining two-dimensional topographic maps or digital elevation models (DEMs) with human experience. The transportation plan is mainly based on static analysis of design drawings, with little consideration given to the dynamic spatial relationships between construction machinery and terrain, features, and overhead lines.

[0003] However, when using the above methods to transport construction machinery, the following technical problems often arise: The reliance on human experience in transportation planning leads to high subjectivity and low accuracy. It cannot automatically identify spatial conflicts in complex terrain, resulting in frequent collisions of construction machinery during the entry process. This necessitates repeated stops for adjustments or even rework, seriously affecting project progress and transportation safety.

[0004] The information disclosed in this background section is only intended to enhance the understanding of the background of the inventive concept, and therefore may contain information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] The summary portion of this disclosure is intended to provide a brief overview of the concepts, which will be described in detail in the detailed description portion. This summary portion is not intended to identify key or essential features of the claimed technical solutions, nor is it intended to limit the scope of the claimed technical solutions.

[0006] Some embodiments of this disclosure provide methods, apparatus, electronic devices, and computer-readable media for transporting construction machinery based on terrain data to address one or more of the technical problems mentioned in the background section above.

[0007] In a first aspect, some embodiments of this disclosure provide a method for transporting construction machinery based on terrain data, including: constructing a three-dimensional scene model of the construction-related road section based on acquired terrain data, plotting data, and overhead line data; generating an initial arrangement state of the target construction machinery in the three-dimensional scene model by embedding a pre-loaded three-dimensional model of the target construction machinery into the three-dimensional scene model; simulating the movement of the target construction machinery along a preset route in the three-dimensional scene model based on the initial arrangement state to generate an initial transportation plan; and performing the following first step on the initial transportation plan: performing automatic collision verification on the movement process of the target construction machinery based on the initial transportation plan to generate a collision detection result; and performing the following second step in response to the collision detection result meeting preset conditions: adjusting the transportation path and transportation posture in the initial transportation plan to generate an optimized transportation plan; generating a transportation instruction sequence corresponding to the optimized transportation plan; and executing the transportation of the target construction machinery according to the transportation instruction sequence.

[0008] Secondly, some embodiments of this disclosure provide a construction machinery transportation device based on terrain data, comprising: a construction unit configured to construct a three-dimensional scene model of the construction-related road segment based on acquired terrain data, plotting data, and overhead line data of the construction-related road segment; a generation unit configured to generate an initial arrangement state of the target construction machinery in the three-dimensional scene model by embedding a pre-loaded three-dimensional model of the target construction machinery into the three-dimensional scene model; a simulation unit configured to simulate the movement of the target construction machinery along a preset route in the three-dimensional scene model based on the initial arrangement state, to generate an initial transportation plan; and an execution unit configured to perform the following first step for the initial transportation plan: performing automatic collision verification on the movement process of the target construction machinery based on the initial transportation plan to generate a collision detection result; and, in response to the collision detection result satisfying a preset condition, performing the following second step: adjusting the transportation path and transportation posture in the initial transportation plan to generate an optimized transportation plan; generating a transportation instruction sequence corresponding to the optimized transportation plan; and executing the transportation of the target construction machinery according to the transportation instruction sequence.

[0009] Thirdly, some embodiments of this disclosure provide an electronic device, including: one or more processors; and a storage device having one or more programs stored thereon, such that when the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any implementation of the first aspect.

[0010] Fourthly, some embodiments of this disclosure provide a computer-readable medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the method as described in any implementation of the first aspect.

[0011] The above-described embodiments of this disclosure have the following beneficial effects: The terrain-based construction machinery transportation method of some embodiments of this disclosure can achieve automated obstacle avoidance in construction transportation, thereby reducing collision accidents and human intervention. Specifically, the frequent collisions in traditional construction transportation processes are due to the lack of deep fusion and three-dimensional visualization of multi-source data such as terrain, features, and overhead lines, making it impossible to achieve automatic collision verification and closed-loop optimization of transportation plans. Based on this, the terrain-based construction machinery transportation method of some embodiments of this disclosure first constructs a three-dimensional scene model of the construction-related road sections based on the acquired terrain data, plotting data, and overhead line data. This achieves high-precision fusion of multi-source heterogeneous data, constructing a full-element three-dimensional visualization environment including terrain, features, and overhead lines, providing a realistic and reliable digital foundation for subsequent transportation simulation. Then, by embedding the pre-loaded three-dimensional model of the target construction machinery into the three-dimensional scene model, the initial arrangement state of the target construction machinery in the three-dimensional scene model is generated. This achieves precise anchoring and attitude initialization of the construction machinery and its three-dimensional scene model in three-dimensional space, ensuring that the starting point of the motion simulation conforms to physical reality, and providing accurate initial boundary conditions for subsequent path planning. Secondly, based on the initial layout, the target construction machinery is simulated to move along a preset route in the 3D scene model to generate an initial transportation plan. This achieves collaborative planning of the machinery's movement path and terrain posture, generating a standardized initial plan containing path point sets and posture sequences, providing iterative input data for collision verification. Finally, for the initial transportation plan, the following first step is executed: Based on the initial transportation plan, automatic collision verification is performed on the movement process of the target construction machinery to generate collision detection results. This achieves refined collision detection for continuous movement processes, accurately locating collision positions and types through discretization sampling and envelope mapping, providing a quantitative basis for optimization. In response to the collision detection results meeting preset conditions, the following second step is executed: The transportation path and transportation posture in the initial transportation plan are adjusted to generate an optimized transportation plan. This achieves closed-loop optimization based on collision detection results, automatically generating optimized plans that eliminate collision risks and satisfy kinematic constraints through classification, sorting, strategy matching, and plan fusion. A transportation instruction sequence corresponding to the optimized transportation plan is generated, and the transportation of the target construction machinery is executed according to the transportation instruction sequence. It has achieved the automated transformation from optimization scheme to mechanical execution, and completed safe and autonomous transportation in complex terrain through instruction generation, spatial synchronous verification, and dynamic feedback adjustment. Attached Figure Description

[0012] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and elements are not necessarily drawn to scale.

[0013] Figure 1 This is a flowchart of some embodiments of a construction machinery transportation method based on terrain data according to the present disclosure; Figure 2 These are schematic diagrams of some embodiments of a construction machinery transportation device based on terrain data according to this disclosure; Figure 3 This is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure. Detailed Implementation

[0014] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0015] It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other.

[0016] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0017] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0018] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0019] This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0020] refer to Figure 1The diagram illustrates a flow 100 of some embodiments of a terrain-based construction machinery transportation method according to the present disclosure. This terrain-based construction machinery transportation method includes the following steps: Step 101: Based on the obtained terrain data, plotting data, and overhead line data of the construction-related road sections, construct a three-dimensional scene model of the construction-related road sections.

[0021] In some embodiments, the executor of the above-described terrain data-based construction machinery transportation method (e.g., an electronic device) can be hardware or software. When the computing device is hardware, it can be implemented as a distributed cluster of multiple servers or terminal devices, or as a single server or a single terminal device. When the computing device is software, it can be installed in the hardware devices listed above. It can be implemented as multiple software programs or software modules to provide distributed services, or as a single software program or software module. No specific limitations are made here.

[0022] In other embodiments, the aforementioned executing entity can construct a three-dimensional scene model of the construction-related road sections based on the acquired terrain data, plotting data, and overhead line data. The construction-related road sections can be road segments that construction machinery needs to pass through; for example, a 5-kilometer-long temporary mountain road connecting the construction site and material storage area. The terrain data can be spatial three-dimensional point cloud information reflecting the surface morphology; for example, the terrain data can be mountainous surface elevation point clouds obtained through UAV LiDAR scanning. The plotting data can be manually marked location and attribute information of features within the construction area; for example, the plotting data can be bridge locations, river boundaries, and outlines of existing buildings marked on construction drawings. The overhead line data can be three-dimensional spatial location information of power transmission lines above the construction area; for example, the overhead line data can include tower coordinates, conductor sag curves, and line voltage levels. The three-dimensional scene model can be a digital three-dimensional environment scene of the construction-related road sections; for example, the three-dimensional scene model can be a virtual simulation scene including mountainous terrain, roads, bridges, power towers, and cables.

[0023] In some optional implementations of certain embodiments, the aforementioned execution entity may construct a three-dimensional scene model of the construction-related road section based on the acquired terrain data, plotting data, and overhead line data, which may include the following steps: The first step is to perform format conversion and coordinate normalization on the aforementioned terrain data to generate standardized terrain data. This standardized terrain data can be terrain point cloud data that has undergone coordinate unification and format standardization. For example, it can be a standard format point set converted from the original LiDAR point cloud into a unified coordinate system. In practice, firstly, the terrain data (original terrain point cloud file) is read, and its original coordinate system and data format are identified. Then, the point cloud coordinates are converted to a project-wide unified spatial reference system, and the data format is standardized to a standardized LAS or XYZ format to generate standardized terrain data.

[0024] The second step involves classifying ground points and removing outliers from the standardized terrain data to generate a filtered terrain point cloud model. Outliers can be isolated noise points with significant spatial distance or density differences from surrounding points in the point cloud. The filtered terrain point cloud model can be a point cloud dataset (point cloud model) that primarily retains surface features after removing non-ground points and outliers. For example, the filtered terrain point cloud model can be point cloud data that purely represents ground elevation undulations after removing vegetation, vehicle, and other points. In practice, firstly, a ground point classification algorithm is used to identify and label ground points and non-ground points in the standardized terrain data. Then, statistical filtering or radius filtering algorithms are used to remove isolated outliers. Finally, the ground point classification results are retained, and the filtered terrain point cloud model is output. For example, a progressive triangulation algorithm is used to separate ground points and tree points from a LiDAR point cloud in a mountainous area, and then outliers with abnormal flight altitudes are removed to obtain a pure surface point cloud model.

[0025] The third step involves registering the filtered terrain point cloud model with the plotted data to generate a fused terrain and feature model. This registration process precisely aligns the filtered terrain point cloud model with the plotted data in space. The fused terrain and feature model can be a composite model formed by overlaying the terrain point cloud and plotted features; for example, it could be a composite model formed by overlaying road centerlines and river boundary markers onto the terrain point cloud. In practice, firstly, the filtered terrain point cloud model and the plotted data are imported into the same spatial reference frame. Then, an iterative nearest-point algorithm is used to find the optimal spatial transformation matrix, achieving precise alignment between the point cloud and vector data (e.g., by iteratively calculating the rigid body transformation (rotation and translation) between two point sets, the average distance error between the source point cloud (filtered terrain point cloud model) and the target point cloud (reference point set extracted from the plotted data) is minimized, thus finding the optimal rotation matrix and translation vector). Finally, the registered plotted attributes are attached to the corresponding terrain locations, outputting the fused terrain and feature model. For example, by registering the road centerline vector data with the terrain point cloud, the road markings can be accurately projected onto the corresponding terrain surface.

[0026] The fourth step involves reconstructing the fused terrain and feature model to generate a 3D geometric mesh model. This 3D geometric mesh model can be a continuous triangular mesh surface model reconstructed from point cloud data. For example, it could be a mountain surface mesh containing millions of triangular faces, generated from terrain point clouds. In practice, firstly, the point cloud in the fused terrain and feature model is triangulated to establish topological connections between points. Then, a surface reconstruction algorithm is used to generate a continuous triangular mesh surface, filling holes and optimizing boundaries. Finally, a 3D geometric mesh model is output for visualization and analysis. For example, a Poisson surface reconstruction algorithm is used to generate a continuous triangular mesh model from the mountain point cloud, repairing holes formed in areas with missing data.

[0027] The fifth step involves integrating the aforementioned 3D geometric mesh model with the overhead line data to generate an interactive 3D scene model. In practice, firstly, a 3D conductor model is generated based on the tower coordinates and conductor parameters from the overhead line data. Then, the conductor model is spatially aligned and overlaid with the 3D geometric mesh model. Finally, interactive features of the scene (e.g., scaling, rotation, distance measurement) are configured, and an interactive 3D scene model is output.

[0028] Step 102: By embedding the preloaded 3D model of the target construction machinery into the 3D scene model, the initial layout state of the target construction machinery in the 3D scene model is generated.

[0029] In some embodiments, the execution entity can generate the initial arrangement state of the target construction machinery in the 3D scene model by embedding a pre-loaded 3D model of the target construction machinery into the 3D scene model. The 3D model of the target construction machinery can be a 3D digital model reflecting the geometry and dimensions of the construction machinery. For example, the 3D model of the target construction machinery can be a precise 3D model of a QUY260 crawler crane, including components such as the chassis, tracks, turntable, and boom. The initial arrangement state can be the initial position, orientation, and attitude information of the construction machinery in the 3D scene. For example, the initial arrangement state can be the state parameters of a crane located at the starting point of a road segment, with its front facing due north and its chassis horizontally placed.

[0030] In some optional implementations of certain embodiments, the execution entity can generate the initial arrangement state of the target construction machinery in the three-dimensional scene model by embedding the preloaded three-dimensional model of the target construction machinery into the three-dimensional scene model, which may include the following steps: The first step is to generate a digital parameter model of the machinery based on the aforementioned 3D model of the target construction machinery. This digital parameter model can be a structured data model containing the machinery's geometric dimensions, kinematic parameters, and collision boundaries. For example, the digital parameter model could include the length, width, and height dimensions, wheelbase, track width, minimum turning radius, and bounding box data of each component of a crawler crane. In practice, firstly, the 3D model file of the target construction machinery is parsed to extract its geometric mesh and the dimensional parameters of each component. Then, a structured parameter model of the machinery is constructed, defining the relative positional relationships between moving parts and collision boundaries. Finally, a digital parameter model suitable for spatial calculations is output. For example, parsing the crane's FBX model file, extracting the body dimensions as 12m × 3.5m × 3.2m, defining the track height as 0.5m above the ground, and outputting a JSON format model file containing these parameters as the digital parameter model.

[0031] The second step involves determining the initial position and orientation of the digital parameter model in the 3D scene model's spatial coordinate system based on the aforementioned transportation starting point information for the construction-related road sections and the 3D scene model itself, thereby generating a positioning matrix. The transportation starting point information can be the coordinates of the starting position and initial orientation of the construction machinery. For example, it could be the location at the road section entrance with coordinates (104.32°E, 35.67°N) and the requirement that the machinery should face due east. The positioning matrix can be a 4×4 spatial transformation matrix describing the position and orientation of the machinery model in the 3D scene. For instance, it could be a transformation matrix that translates the machinery model to the starting point coordinates, rotates it to a specified orientation, and scales it to its actual size. In practice, first, the 3D coordinates of the transportation starting point and the initial orientation angle of the machinery are read. Then, translation and rotation transformation matrices are constructed based on the starting point position and orientation, and multiplied together to obtain the complete positioning matrix. Finally, the spatial transformation matrix applicable to the machinery model is output. For example, with the starting coordinates (500, 200, 150) and the orientation due east (rotated 90 degrees around the Z-axis), construct a 4×4 transformation matrix that includes translation and rotation.

[0032] The third step involves importing the aforementioned digital parameter model into the aforementioned 3D scene model using the aforementioned positioning matrix to generate a preliminary embedded mechanical model. This preliminary embedded mechanical model can be a mechanical model placed in the 3D scene according to the positioning matrix but not yet validated. For example, it could be a crane model placed at the starting position but without checking for penetration of the ground. In practice, firstly, the coordinates of each vertex in the digital parameter model are multiplied by the positioning matrix to achieve spatial transformation. Then, the transformed mechanical geometry is added to the scene graph of the 3D scene model. Finally, a mechanical model with preliminary spatial positioning is generated. For example, each of the 2000 vertices of the crane model is multiplied by the positioning matrix to ensure the crane precisely appears at the starting position and faces due east, resulting in a crane model embedded in the scene.

[0033] The fourth step involves performing spatial interference and bottom clearance checks on the initially embedded mechanical model based on the terrain data in the aforementioned 3D scene model to generate the initial layout state. In practice, the aforementioned spatial interference can refer to the mechanical model penetrating or overlapping with the terrain or feature models in 3D space. For example, spatial interference could be the crane tracks sinking below the ground, the boom passing through overhead lines, or the vehicle body overlapping with a mountain. The aforementioned bottom clearance check is a process of verifying whether a safe ground clearance is maintained between the mechanical chassis and the ground surface. For example, the bottom clearance check can be used to check whether there is a safety gap of at least 200mm between the bottom of the crane tracks and the ground. First, the vertical distance between each point on the bottom of the mechanical model and the terrain surface below is calculated to detect any penetration or insufficient clearance. Then, the spatial overlap between each component of the mechanical model and the terrain, features, and overhead lines in the 3D scene is checked. Finally, if both the interference and clearance meet the safety requirements, the initial layout state is confirmed and output. For example, the inspection revealed that the left track of the crane penetrated 5cm into a protruding rock. The position of the machine was adjusted until the chassis was completely in contact with the ground and the gap was uniform, confirming that the arrangement was feasible.

[0034] Step 103: Based on the initial layout, simulate the movement of the target construction machinery along a preset route in the three-dimensional scene model to generate an initial transportation plan.

[0035] In some embodiments, the execution entity can simulate the movement of the target construction machinery along a preset route in the three-dimensional scene model based on the initial arrangement state to generate an initial transportation plan. The preset route can be a pre-planned or designed centerline for construction machinery transportation; for example, it can be the centerline of a road from point A to point B on a design drawing or a temporarily planned transportation corridor. The initial transportation plan can be a preliminary feasible transportation plan generated based on the preset route, including path, attitude, and speed. For example, the initial transportation plan can be a text file containing a list of path point coordinates, vehicle pitch / tilt angles at each point, and suggested driving speeds.

[0036] In some optional implementations of certain embodiments, the execution entity may, based on the initial arrangement state, simulate the movement of the target construction machinery along a preset route in the three-dimensional scene model to generate an initial transportation plan, which may include the following steps: The first step involves spatial sampling based on the aforementioned preset route and 3D scene model, using curvature variation density to generate a discrete sequence of path sampling points. This curvature variation density sampling can be achieved by dynamically adjusting the sampling point spacing according to the path's curvature. For example, the curvature variation density could be one point every 5 meters on straight sections, and a denser sampling density of one point every 1 meter on curves with greater curvature. In practice, first, the geometric data of the preset route (e.g., a polyline or polysegment line) and the 3D scene model are acquired. Then, the curvature of each point on the path is calculated, and the sampling step size is dynamically determined based on the curvature value. Smaller step sizes are used where curvature is greater (e.g., curves), and larger step sizes are used where curvature is less (e.g., straight sections). Finally, sampling is performed along the path according to the calculated step size, generating a discrete sequence of path sampling points containing attributes such as position, tangent direction, and curvature. For example, for an S-shaped curve, one point is sampled every 10 meters on straight sections, and the density is increased to one point every 2 meters on the curve, ultimately generating a sequence of 300 ordered points as the path sampling point sequence.

[0037] The second step involves mapping the spatial attitude and contour of the machinery at different path positions point by point, based on the aforementioned path sampling point sequence and the mechanical pose in the initial layout state, generating a path-attitude mapping table. This path-attitude mapping table can be a table recording the mapping relationship between the construction machinery's corresponding pose (position and orientation) at each path point. For example, it can be a data structure where the key is the path point ID, and the values ​​are the vehicle's coordinates, heading angle, roll angle, and pitch angle at that point. In practice, firstly, the reference pose (position, orientation, roll, pitch) of the construction machinery at the starting point is obtained from the initial layout state. Then, the path sampling point sequence is traversed. For each path point, based on the mechanical pose of the previous point, combined with the tangent direction, normal direction (from the scene terrain), and mechanical kinematic constraints of the current path point, the mechanical pose at that point, including the position of its bounding box vertex, is calculated. Finally, each path point is associated with its corresponding mechanical pose and bounding box vertex coordinates to generate the path-attitude mapping table.

[0038] The third step involves performing spatial interference detection based on the path and attitude mapping table and the terrain data in the 3D scene model, generating path safety verification results. This spatial interference detection can involve placing the mechanical model from the path and attitude mapping table point-by-point into the 3D scene and detecting whether the mechanical envelope overlaps with the terrain, features, or overhead lines at each location. The path safety verification results can be the conclusions of collision pre-detection of the mechanical attitude and scene at each sampling point on the path. For example, the path safety verification results can be a Boolean array marking which path points have vehicle bounding boxes interfering with the terrain or features. In practice, first, the mechanical model from the path and attitude mapping table is placed point-by-point into the 3D scene. Then, the overlap between the mechanical envelope and the terrain, features, or overhead lines at each location is detected. Finally, all interference detection results are summarized to generate path safety verification results. For example, the detection found that at point 80, the crane boom is only 1.2 meters away from the 10kV line above (the safety distance should be 2 meters), and at point 150, the right track is sunk 3 centimeters into the ground.

[0039] The fourth step involves adjusting the path point sequence and corresponding attitude based on the path safety verification results to generate an initial transportation plan. First, the path points causing interference and their types are identified based on the path safety verification results. Then, interference is eliminated by locally offsetting the path point coordinates or adjusting the mechanical attitude parameters. Finally, the adjusted path point sequence and attitude parameters are integrated to output the initial transportation plan. For example, adjusting the boom angle at point 80 from 30 degrees to 25 degrees to increase clearance, and offsetting the path at point 150 outward by 0.5 meters to avoid protruding rocks, generates the initial transportation plan.

[0040] Step 104, for the initial transportation plan, perform the following first step: Step 1041: Based on the initial transportation plan, perform automatic collision verification on the movement process of the target construction machinery to generate collision detection results.

[0041] In some embodiments, the aforementioned executing entity can perform automatic collision verification on the movement process of the target construction machinery based on the initial transportation plan to generate collision detection results. The aforementioned movement process can be a continuous travel process of the construction machinery from a starting point to an end point. The aforementioned automatic collision verification can be a process in which a computer program automatically detects whether spatial interference occurs between the machinery and the scene geometry during transportation. For example, the aforementioned automatic collision verification can be software detecting frame-by-frame whether the vehicle model penetrates models such as road surfaces, bridges, and cables. The aforementioned collision detection results can be detection conclusions containing information on the collision location, collision type, and severity. For example, the aforementioned collision detection results can be a detection report indicating a collision between the chassis and the ground at the 3rd second or insufficient clearance between the boom and the guide wire at the 15th second.

[0042] In addressing the aforementioned technical problems using technical solutions, the application scenarios—mountainous power grid projects, wind power construction, and the transportation of large construction machinery in complex terrain areas—often present the following challenges: multiple interference risk points exist along the transportation path, and these risk points are interconnected. Simple global replanning leads to a surge in computational load, while path adjustment and attitude adjustment are mutually restrictive, making it difficult to efficiently generate feasible optimization solutions while ensuring safety, resulting in increased time and computational costs. Therefore, the following characteristics are required for this application scenario: prioritizing high-risk areas, coordinating path and attitude optimization, and avoiding global recalculation through local adjustments. In some optional implementations of certain embodiments, the execution entity may perform automatic collision verification on the movement process of the target construction machinery based on the initial transportation plan to generate collision detection results, which may include the following steps: The first step involves performing time-uniform motion interpolation between path sampling points based on the transportation path information, transportation posture information, and speed information from the initial transportation plan, thereby generating a keyframe sequence. The transportation path information can be a sequence of geometric coordinates of the machinery's travel route; for example, it could be crane trajectory data composed of 200 three-dimensional coordinate points. The transportation posture information can be data on the changes in the machinery's posture parameters during travel; for example, it could include time-series data of the chassis pitch angle, roll angle, and steering angle at each path point. The speed information can be speed planning data for the machinery traveling on different road sections; for example, it could be segmented speed data at 0.5 m / s on flat roads, 0.2 m / s on curves, and 0.3 m / s on slopes. The path sampling points can be discrete locations selected at certain intervals along the transportation path; for example, a point with coordinate and posture information selected every 1 meter along the path. The aforementioned time-uniform motion interpolation can be a process of calculating the motion time between adjacent sampling points based on velocity information, and interpolating the intermediate mechanical position and attitude parameters between path sampling points at fixed time intervals (e.g., every 0.1 seconds). The aforementioned keyframe sequence can be a sequence of mechanical state snapshots sampled time-uniformly during motion; for example, the keyframe sequence can be a state dataset recording mechanical position and attitude every 0.1 seconds. In practice, firstly, path points, attitude, and velocity data are extracted from the initial transportation plan. Then, the motion time between adjacent sampling points is calculated based on velocity information, and intermediate states are interpolated at fixed time intervals. Finally, a time-uniformly distributed keyframe sequence is output.

[0043] The second step is to determine the mechanical bounding box based on the keyframe sequence. This bounding box can be a simplified geometry that encloses the outer contour of the mechanical model; for example, it could be an axis-aligned cuboid bounding box with a length of 12 meters, a width of 3.5 meters, and a height of 4 meters. In practice, first, the position and orientation data of the mechanical model in the keyframes are parsed. Then, the coordinates of the eight vertices of the minimum bounding box in the current pose are calculated based on the mechanical's geometric dimensions. Finally, the mechanical bounding box corresponding to each keyframe is output. For example, when the crane is located at (500, 200, 150) and horizontal, the calculated bounding box vertex coordinates are within the range of ±6 meters in the X direction, ±1.75 meters in the Y direction, and 0 to 4 meters in the Z direction.

[0044] The third step involves determining the overlapping areas between the mechanical bounding box and the feature models based on the terrain data and feature models of the aforementioned 3D scene model, in order to generate a potential interference region set. The terrain data can be 3D point cloud or mesh data reflecting the undulating terrain, for example, 5cm resolution mountain elevation mesh data generated by LiDAR scanning. The feature models can be 3D models of artificial and natural objects within the construction area, such as power tower models, tree models, building models, and bridge models. The potential interference region set can be a collection of areas where the mechanical bounding box and feature models may overlap or spatially conflict, for example, three overlapping areas near a bend and close to a mountain, and two overlapping areas near a power tower. In practice, firstly, a rough spatial overlap detection is performed between the mechanical bounding box and the terrain data and feature models in the 3D scene. Then, the overlapping areas and their corresponding keyframe numbers are recorded. Finally, the potential interference region set is output. For example, the detection found that the bounding box in frames 50-80 overlapped with the mountain and in frames 120-150 overlapped with the power tower, generating an information set containing two interference regions.

[0045] The fourth step involves performing geometric intersection processing on the 3D model of the target construction machinery and the 3D scene model within the aforementioned potential interference region set to generate preliminary collision pairs and collision depth data. This geometric intersection processing can involve loading a machinery triangular mesh model and a scene mesh model within the potential interference region, performing precise triangle-to-triangle intersection calculations, detecting collisions, and calculating the penetration depth. The initial collision pairs can be pairs of intersecting objects detected through the set intersection processing. For example, the collision depth data can be the penetration distance or volume measurement of the intersection point. For example, the collision depth data could be a 3cm penetration depth between the track and the ground. In practice, firstly, a high-precision machinery triangular mesh model and a scene mesh model are loaded within the potential interference region. Then, precise triangle-to-triangle intersection calculations are performed to detect collisions and calculate the penetration depth. Finally, the preliminary collision pairs and collision depth data are output. For example, at frame 60, an intersection of a track triangle and a terrain triangle is detected with a penetration depth of 3cm, recorded as a collision pair of (track, ground, 3cm).

[0046] The fifth step involves filtering the initial collision pairs and collision depth data to remove duplicate and false collision data, generating a cleaned collision pair list. Duplicate data can be redundant data resulting from multiple recordings of the same collision event across different keyframes or sampling points. For example, the same track colliding with the ground might be recorded 10 times in 10 consecutive frames. False collision data can be spurious collision detection results caused by algorithm tolerance, numerical errors, or geometric approximations. For example, two model surfaces that are only tangent (penetration depth of 0) might be misclassified as a collision, or there might be noise collisions with a penetration depth less than 0.1 mm. The cleaned collision pair list can be the set of valid collision events retained after filtering to remove duplicate records and false detection results. For example, merging the same track-ground collision recorded repeatedly in 10 consecutive frames into one record and removing noise collisions with a penetration depth less than 0.1 mm results in a list containing 3 true collision pairs. In practice, first, duplicate collision records caused by consecutive frames are identified and marked. Then, false collision data caused by algorithm tolerance is detected and removed. Finally, the cleaned-up collision list is output after removing redundancy and spurious data. For example, collisions where the same track is detected with the ground in 10 consecutive frames are merged into one collision event; noisy collisions with a penetration depth of less than 0.1 mm are removed.

[0047] Step 6: For each collision pair in the above-mentioned cleaned collision pair list, classify the events according to the collision geometry, collision object type, and collision depth to generate preliminary interference event classification results. The collision geometry can be a description of the spatial geometric attributes at the time of the collision, such as point contact, line contact, surface contact, contact area, penetration depth, etc. The collision object type can be the category of environmental objects involved at the time of the collision, such as terrain, overhead lines, trees, buildings, or temporary facilities. The collision depth can be the spatial distance between the mechanical model and the obstacle, such as the penetration distance of a crane track sinking 3 cm below the ground. The preliminary interference event classification results can be a set of interference events grouped by collision type, such as 3 terrain interference events, 2 clearance interference events, and 1 lateral scraping event. In practice, first, the geometric features of each collision pair are analyzed to determine whether it is point contact or surface contact. Then, classification and labeling are performed according to the type of collision object (terrain / wire / tree). Finally, the system outputs the classification results of interference events grouped by collision type. For example, track-ground collisions are classified as terrain interference, boom-guided collisions as clearance interference, and vehicle-body-tree collisions as side scrapes.

[0048] Step 7: Based on a preset safety threshold rule base, risk assessment values ​​are assigned to the preliminary interference event classification results to generate a list of interference events with risk assessment information. The safety threshold rule base can be a set of rules storing various collision safety judgment standards. For example, it could include rules such as chassis ground clearance ≥10cm, guide clearance ≥2m, and lateral safety distance ≥0.5m. The interference event list can be a structured collision event list containing risk assessment information. For example, it could include records such as (Event Type: Terrain Interference, Location: Point 120, Risk Level: High). In practice, first, the preset safety threshold rule base is called to obtain the safety standards for various collisions. Then, the collision depth is compared with the safety thresholds to calculate the risk level. Finally, risk assessment information is assigned to each interference event, and an interference event list is generated. For example, a terrain interference collision depth of 3cm exceeds the 10cm safety threshold and is assessed as high risk; a guide clearance of less than 0.5m is below the 2m threshold and is also assessed as high risk.

[0049] Step 8: Based on the aforementioned list of interference events, generate collision detection results. In practice, first, summarize the risk assessment information of all interference events. Then, sort and organize the interference events according to risk level and location. Finally, output collision detection results containing complete collision information. For example, generate a comprehensive collision detection report containing 3 high-risk terrain collisions, 2 high-risk clearance collisions, and 1 medium-risk lateral scrape.

[0050] The above-described steps, as an inventive point of this disclosure, solve the technical problem mentioned in the background art: "There are multiple different types of interference risk points on the transportation path, and these risk points are coupled with each other. Simple global replanning will lead to a surge in computational load. At the same time, path adjustment and attitude adjustment are mutually restrictive, making it difficult to efficiently generate feasible optimization schemes while ensuring safety, resulting in increased time and computational costs." The reasons for the above technical problems are as follows: Traditional methods lack deep fusion and 3D visualization of multi-source data, making it impossible to achieve automatic collision verification and closed-loop optimization of transportation schemes. Furthermore, path and attitude planning are separated, making it difficult to coordinate adjustments in complex terrain. The present invention first extracts collision events and groups them by spatial clustering, then sorts them by risk priority and performs coordinated local optimization of path and attitude in each region. Finally, it ensures the optimization quality through local resimulation and verification, and integrates each local scheme into a global scheme. This achieves priority processing of high-risk areas, coordinated optimization of path and attitude, and avoidance of global recalculation by local adjustments, saving the computational resources and time costs required for global replanning.

[0051] Step 1042: In response to the collision detection result meeting the preset conditions, the following second step is executed: Step 10421: Adjust the transportation path and transportation posture in the initial transportation plan to generate an optimized transportation plan.

[0052] In some embodiments, the aforementioned executing entity may adjust the transportation path and transportation attitude in the initial transportation plan to generate an optimized transportation plan. The optimized transportation plan may be the final transportation plan after eliminating the risk of collision. For example, the optimized transportation plan may be a safe transportation plan that shifts the curve that would collide with the mountainside in the original plan outward by 2 meters and adjusts the chassis attitude by 5 degrees.

[0053] In addressing the technical challenges mentioned above, the application scenarios—mountainous power grid projects, wind power construction, and the transportation of large construction machinery in complex terrain areas—often present the following challenges: multiple interference risk points exist along the transportation path, and these risk points are interconnected. Simple global replanning leads to a surge in computational load, while path adjustment and attitude adjustment are mutually restrictive, making it difficult to efficiently generate feasible optimization solutions while ensuring safety, thus increasing time and computational costs. Considering the following requirements for this application scenario: prioritizing high-risk areas, coordinating path and attitude optimization, and avoiding global recalculation through local adjustments, we have decided to adopt the following solution: In some optional implementations of certain embodiments, the executing entity may adjust the transportation path and transportation posture in the initial transportation plan to generate an optimized transportation plan, which may include the following steps: The first step involves extracting corresponding collision information from the hard and soft collision events in the collision detection results to generate a list of collision events to be optimized. Hard collision events can be severe collisions where machinery penetrates an obstacle, such as a crane track sinking 3 cm into the ground or a boom overlapping a power line by 2 cm. Soft collision events can be potential risk events where the clearance between machinery and an obstacle is less than a safety threshold, such as a crane boom being 1.5 meters from a 10kV line (the safety distance should be 2 meters). The corresponding collision information can include detailed data such as the location, type, depth, and involved components of the collision. For example, the collision information could include the collision point coordinates (512.3, 245.6, 178.2), the collision type being terrain interference, the penetration depth being 3 cm, and the involved component being the left track. The list of collision events to be optimized can be a list of collision events requiring optimization. For example, the list could include 5 terrain interference events, 3 clearance interference events, and 1 lateral scrape event. In practice, firstly, information such as the location, type, and depth of all collision events is extracted from the collision detection results. Then, hard and soft collision events are distinguished and their priority weights are assigned. Finally, a list of collision events to be optimized is generated.

[0054] The second step involves generating multiple clusters of interference regions to be processed based on the aforementioned list of collision events to be optimized. These clusters can be sets of regions formed by clustering spatially adjacent collision events. For example, they could include a cluster formed by three terrain collisions in a curve area and a cluster formed by two clearance collisions in a straight section. In practice, first, the spatial distance between each event in the list of collision events to be optimized is calculated. Then, events with a spatial distance less than a threshold are clustered into the same interference region cluster. Finally, multiple clusters of interference regions to be processed are output. For example, three terrain collisions at a curve with a distance of less than 5 meters are clustered into region cluster A, and two clearance collisions at a straight section are clustered into region cluster B.

[0055] The third step involves prioritizing the multiple interference region clusters to generate a sorted interference region cluster set. This sorted cluster set can be a collection of interference regions ordered by priority; for example, it could be a sequence where high-risk curve clusters are ranked first, medium-risk straight clusters second, and low-risk ramp clusters third. In practice, first, the maximum risk level and number of collision events in each interference region cluster are assessed. Then, the clusters are sorted from highest to lowest risk level. Finally, the sorted interference region cluster set is output. For example, cluster A, containing 3 high-risk collisions, is ranked first, and cluster B, containing 2 medium-risk collisions, is ranked second.

[0056] Fourth, for each interference region cluster in the sorted interference region cluster set, perform the following steps sequentially according to the sorting information: Sub-step one involves calling a preset path adjustment strategy library and a preset attitude adjustment strategy library to generate a local optimization scheme for the aforementioned cluster of interference regions. The sorting information can be priority data determining the processing order of the interference region clusters; for example, the sorting information could be based on a collision depth of 3cm and a risk assessment that prioritizes high-risk curve clusters. The preset path adjustment strategy library can be a collection of strategies storing various path adjustment methods; for example, it could include strategies such as lateral path offset, vertical lift, curve smoothing, and obstacle avoidance. The preset attitude adjustment strategy library can be a collection of strategies storing various attitude adjustment methods; for example, it could include strategies such as chassis pitch adjustment, roll adjustment, boom angle adjustment, and vehicle steering adjustment. The local optimization scheme can be a path and attitude adjustment scheme for a single interference region cluster; for example, the local optimization scheme could be to offset the path outward by 1 meter in the curve region while simultaneously adjusting the chassis roll angle from 5 degrees to 3 degrees. In practice, the first step is to identify the collision type combination of the interference region cluster. Then, suitable path adjustment strategies are matched from the path adjustment strategy library, and suitable attitude adjustment strategies are matched from the attitude adjustment strategy library. Finally, a local optimization scheme for the cluster is output. For example, for a curve terrain collision cluster, a strategy of 2 meters lateral path offset and a strategy of reducing chassis roll angle by 2 degrees are invoked.

[0057] Sub-step two involves determining the matching degree between the ground normal vector and the mechanical attitude at each sampling point on the optimized path, based on the aforementioned local optimization scheme and the terrain data in the 3D scene model, and generating an attitude fitness evaluation result. The matching degree between the ground normal vector and the mechanical attitude can be a measure of the consistency between the mechanical chassis normal vector and the ground normal vector. For example, the matching degree can be the cosine of the angle between the chassis normal vector (0.2, 0, 0.98) and the ground normal vector (0.25, 0, 0.97), which is 0.998. The attitude fitness evaluation result can be a quantitative evaluation of the degree of matching between the mechanical attitude and the terrain. For example, a matching degree of 0.998 can be rated as excellent, 0.950 as good, and 0.850 as poor. In practice, firstly, the ground normal vector at each sampling point on the optimized path is extracted. Then, the cosine of the angle between the mechanical chassis normal vector and the ground normal vector is calculated as the matching degree. Finally, the attitude fitness evaluation result is output.

[0058] Step three involves fine-tuning the transport attitude parameters in the aforementioned local optimization scheme based on the attitude adaptability evaluation results, generating a terrain-adaptive attitude fine-tuning scheme. This terrain-adaptive attitude fine-tuning scheme can be a scheme that precisely adjusts the attitude according to terrain features. For example, it could be a scheme that adjusts the local slope from 15 degrees to 14.5 degrees to ensure complete track contact. In practice, firstly, sampling points with attitude adaptability evaluation results of poor or good are selected. Then, the mechanical attitude parameters are fine-tuned point-by-point based on the ground normal vector. Finally, the terrain-adaptive attitude fine-tuning scheme is output. For example, if the ground normal vector changes by 5 degrees at point 150, the chassis pitch angle is fine-tuned from 15 degrees to 15.8 degrees.

[0059] Sub-step four involves fusing the aforementioned attitude fine-tuning scheme with the path adjustment portion of the local optimization scheme to generate local candidate schemes for terrain matching. These local candidate schemes for terrain matching can be those where both the path and attitude match the terrain well. For example, a local candidate scheme for terrain matching could be a combined scheme with a path offset of 1.2 meters, a chassis pitch angle of 14.5 degrees, and a roll angle of 2.3 degrees. In practice, firstly, the geometric data after path adjustment is spatially aligned with the attitude data after attitude fine-tuning. Then, a one-to-one correspondence between path points and attitude parameters is established. Finally, the local candidate schemes for terrain matching are output. For example, 200 offset path points are fused with the corresponding 200 fine-tuned attitude parameters to form a candidate scheme.

[0060] Sub-step five involves substituting the aforementioned candidate local optimization schemes into the initial transportation scheme and performing a local re-simulation in the 3D scene model to generate a simulated transportation segment. This simulated transportation segment can be transportation process segment data generated by the local re-simulation; for example, it can be simulated motion data from 70 points, from the 80th path point to the 150th path point. In practice, firstly, the path and attitude of the corresponding region in the initial transportation scheme are replaced with the local candidate scheme. Then, the motion process of this segment is simulated in the 3D scene model. Finally, the simulated transportation segment is output. For example, simulating the motion of a curved area from point 80 to point 150 generates simulated data of 70 keyframes.

[0061] Sub-step six involves re-performing collision checks on the simulated transport segment within the aforementioned interference region cluster and its preceding and following buffer zones to generate local verification results. The preceding and following buffer zones can be additional spatial ranges extending before the start point and after the end point of the interference region cluster. These buffer zones are used to verify the smoothness of the transition between the locally optimized scheme and adjacent unoptimized regions, ensuring that no new collision risks arise due to abrupt changes in path or attitude when entering or exiting the optimized region. For example, if the interference region cluster covers path points 100 to 150, points 80 to 100 can be set as the preceding buffer zone, and points 150 to 170 as the following buffer zone, to detect the safety of the transition section. The local verification results can be the results of re-collision checks on the locally optimized scheme. For example, the local verification results could be a conclusion that no collisions were detected in the optimized curve area, and the clearance distance was greater than 2 meters. In practice, firstly, precise collision detection is performed within the interference region cluster. Then, collision detection is also performed in the preceding and following buffer zones to verify the smoothness of the transition. Finally, the local verification results are output. For example, if no collisions are detected within the area and no new collisions are found within the buffer zone, the verification passes.

[0062] Sub-step seven: In response to the local verification results meeting the preset verification conditions, a corresponding local verification scheme is generated. The preset verification conditions can be criteria for determining whether an optimized scheme passes verification. For example, the preset verification conditions could be zero hard collision events and all soft collision event clearance greater than 20% of the safety threshold. The local verification scheme can be a segment of the optimized scheme that passes verification in a local area. For example, the local verification scheme could be an optimized segment that passes verification after a 1.2-meter path offset in a curve area and attitude fine-tuning. In practice, first, it is determined whether the local verification results meet the preset requirements of zero hard collision and clearance. Then, the optimized segments that pass verification and their parameters are recorded. Finally, the corresponding local verification scheme is output.

[0063] The fifth step involves integrating the various local verification schemes to generate an optimized transportation plan. In practice, firstly, the local verification schemes are sequentially assembled into the global transportation plan according to the order of the interference region clusters. Then, the smoothness of the transition at the boundary between adjacent region clusters is addressed. Finally, the complete optimized transportation plan is output. For example, the curve optimization plan is assembled with the straight-line optimization plan, and cubic spline interpolation is used to smooth the transition at the boundary to generate a complete optimized transportation plan.

[0064] The above-described operational steps, combined with step 10422, constitute an inventive point of this disclosure, solving the technical problem mentioned in the background: "Multiple different types of interference risk points exist on the transportation path, and these risk points are coupled with each other. Simple global replanning leads to a surge in computational load. At the same time, path adjustment and attitude adjustment are mutually restrictive, making it difficult to efficiently generate feasible optimization schemes while ensuring safety, resulting in increased time and computational costs." The reasons for the above technical problems are as follows: Traditional methods lack deep fusion and 3D visualization of multi-source data, making it impossible to achieve automatic collision verification and closed-loop optimization of transportation schemes. Furthermore, path and attitude planning are separated, making it difficult to coordinate adjustments in complex terrain. The present invention first extracts collision events and groups them by spatial clustering, then sorts them by risk priority and performs coordinated local optimization of path and attitude in each region. Finally, it ensures the optimization quality through local resimulation and verification, and inherits each local scheme as a global scheme. This achieves intelligent closed-loop optimization that prioritizes high-risk areas, coordinates path and attitude optimization, avoids global recalculation through local adjustments, and adopts the optimized results after verification, saving the computational resources and time costs required for global replanning.

[0065] Step 10422: Generate the transportation instruction sequence corresponding to the optimized transportation plan, and execute the transportation of the target construction equipment according to the transportation instruction sequence.

[0066] In some embodiments, the executing entity can generate a transportation instruction sequence corresponding to the optimized transportation scheme, and execute the transportation of the target construction equipment according to the transportation instruction sequence. The transportation instruction sequence can be a time-ordered set of instructions executable by the mechanical control system. For example, the transportation instruction sequence can be a sequence containing 200 CAN bus instructions, each instruction including a timestamp, instruction type, and parameter value.

[0067] In addressing the technical problems mentioned above by adopting technical solutions, and considering the application scenario—autonomous or high-precision remote-controlled operation of high-value, large-sized, and irregularly shaped construction machinery (such as tunnel boring machines, large modular building components, wind turbine blade transport vehicles, and special engineering vehicles) in complex, unstructured terrain (such as temporary roads at construction sites, mountainous areas, and mining areas)—the following technical problems often arise: During transportation, deviations exist between the actual terrain environment and the pre-generated optimized plan (such as road subsidence, temporary obstacles, and GPS signal fluctuations), causing the actual driving trajectory and posture of the machinery to deviate from the expected plan. If not detected and corrected in time, this may lead to machinery overturning, collisions, or jamming, resulting in major safety accidents and economic losses. Given the following requirements for this application scenario: During transportation, deviations exist between the actual terrain environment and the pre-generated optimized plan (such as road subsidence, temporary obstacles, and GPS signal fluctuations), causing the actual driving trajectory and posture of the machinery to deviate from the expected plan. If not detected and corrected in time, this may lead to machinery overturning, collisions, or jamming, resulting in major safety accidents and economic losses, we have decided to adopt the following solution: In some optional implementations of certain embodiments, the execution entity can generate a transportation instruction sequence corresponding to the optimized transportation scheme, and execute the transportation of the target construction equipment according to the transportation instruction sequence, which may include the following steps: The first step, based on the optimized transportation scheme described above, is to generate structured scheme data. This structured scheme data includes: discrete pathpoint sequences, continuous attitude sequences, speed information, obstacle avoidance information, and a set of operational instructions. The structured scheme data can be standardized transportation scheme data; for example, it can be JSON format data containing pathpoint arrays, attitude arrays, and speed arrays, organized from the optimized scheme. The discrete pathpoint sequences can be sets of path coordinate points sampled at certain intervals; for example, it can be a set of 500 three-dimensional coordinate points (x, y, z) sampled every 0.5 meters. The continuous attitude sequences can be sequences of continuously changing attitude parameters corresponding to the pathpoints; for example, it can be time-series data of chassis pitch angle, roll angle, and steering angle corresponding to 500 pathpoints. The speed information mentioned above can be the planned speed of the machinery on different road sections. For example, the speed information could be segmented data showing a speed of 0.3 m / s from points 0-100, 0.5 m / s from points 100-300, and 0.2 m / s from points 300-500. The obstacle avoidance information mentioned above can be the location and avoidance action data of obstacles requiring deceleration or detour. For example, the obstacle avoidance information could be an instruction to decelerate to 0.1 m / s and sound the horn at the location of a power tower at point 150. The set of operating instructions mentioned above can be a set of operating instructions other than driving instructions. For example, the set of operating instructions could include instructions for adjusting the boom angle, sounding the horn, and turning on the lights. In practice, the first step is to load the optimized transportation plan file. Then, the file content is parsed, extracting the path information into an ordered list of coordinate points, the attitude information into a list of Euler angles or quaternions corresponding to the path points or arranged in a time series, the velocity planning into a velocity-time curve or velocity-path mapping, and the operational requirements for specific road segments into markers or instruction objects. Finally, this data is organized into an internal structure or class instance to generate structured scheme data. For example, parsing a JSON scheme generates a data object containing 1000 path points, 1000 corresponding attitudes, a velocity curve, and "horn" instructions at 5 location points.

[0068] The second step involves converting the discrete path point sequences, continuous attitude sequences, speed information, and operation instruction sets in the structured data into control instruction primitives recognizable by the underlying controller, based on the interface protocol of the preset target construction machinery's control system, to generate a list of control instruction primitives. The interface protocol of the preset target construction machinery's control system can be a communication protocol standard supported by the mechanical controller. For example, the interface protocol could be a communication standard supporting the CANopen protocol, defining the instruction frame format as an 11-bit ID plus 8 bytes of data. The underlying controller can be a hardware device that directly controls the mechanical actuators. For example, the underlying controller could be a PLC controller or an ECU electronic control unit responsible for driving the motor and hydraulic valves. The control instruction primitives can be the smallest instruction unit recognizable by the underlying controller. For example, the control instruction primitives could be atomic instructions such as "forward 0.5 meters," "turn left 15 degrees," or "adjust pitch angle 2 degrees." The aforementioned list of control command primitives can be a set of command primitives arranged sequentially. For example, the list could contain 50 primitive commands such as "forward 0.5 meters," "pitch angle +2 degrees," "forward 0.3 meters," and "turn -10 degrees." In practice, firstly, the target machine's control system interface protocol document is read to determine the command frame format. Then, each action in the structured scheme is converted into a command primitive conforming to the protocol format. Finally, the list of control command primitives is output. The third step involves sequentially and causally sorting the instruction primitives in the control instruction primitive list according to the preset temporal logic and spatial order of the transportation process, generating a preliminary instruction sequence. The preset temporal logic and spatial order of the transportation process can be logical rules that determine the order in which instructions are executed. For example, the temporal logic and spatial order could be a rule that executes forward instructions before turning instructions, or arrives at the position before adjusting the attitude. The preliminary instruction sequence can be an initial instruction sequence sorted by time and space. For example, the preliminary instruction sequence could be a sequence where the first 100 instructions are forward and the last 50 are turning instructions, obtained by sorting the instruction primitives by timestamp. In practice, firstly, the instruction primitives are sorted according to the timestamp information in the transportation plan. Then, the causal relationships between instructions are checked (e.g., turning before forward). Finally, the sorted preliminary instruction sequence is output.

[0069] The fourth step involves inserting terrain synchronization command primitives into the preliminary command sequence based on the terrain data in the aforementioned 3D scene model, to generate a terrain-adaptive synchronization command sequence. These terrain synchronization command primitives can be atomic control commands inserted based on terrain undulation characteristics to adjust the machine's attitude to adapt to ground changes. For example, a terrain synchronization command primitive could be a command primitive for inserting "chassis pitch angle + 5 degrees" when an uphill start is detected. The terrain-adaptive synchronization command sequence can be a complete command sequence after inserting the terrain synchronization command primitives, enabling the machine to dynamically adjust its attitude according to terrain changes during operation. In practice, firstly, key locations in the 3D scene with drastic terrain changes (e.g., uphill start, downhill end, pothole edges) are identified. Then, the corresponding terrain synchronization command primitives are inserted at the preliminary command sequence indices corresponding to these locations. Finally, a complete sequence containing all terrain synchronization command primitives is output. For example, if an uphill is detected at 10 meters along the path, a "pitch angle + 5 degrees" command is inserted; if a roll is detected at 30 meters along the path, a "roll angle + 2 degrees" command is inserted.

[0070] The fifth step involves driving the pre-defined digital twin model of the construction machinery in a 3D simulation environment to execute the aforementioned terrain-adaptive synchronization instruction sequence, thereby generating simulated execution trajectory and state data. The 3D simulation environment can be a virtual 3D space used to simulate mechanical motion; for example, it can be a high-fidelity simulation platform built on the Unity3D engine, incorporating terrain and mechanical models. The pre-defined digital twin model of the construction machinery can be a virtual simulation model corresponding one-to-one with the real machinery; for example, it can be a high-precision simulation model of a tracked crane containing complete kinematic and dynamic parameters. The simulated execution trajectory and state data can be mechanical motion data generated during simulation; for example, it can be the position coordinates and attitude angle data at 500 time points output by the simulation. In practice, first, the digital twin model of the construction machinery is loaded into the 3D simulation environment. Then, the model is driven to execute the synchronization instruction sequence one by one. Finally, the position and attitude data at each moment during the simulation are recorded. For example, in Unity3D, driving the crane model to execute 200 instructions outputs 500 frames of position and attitude data.

[0071] The sixth step involves comparing the proposed trajectory and status data with the optimized transportation plan to generate a command execution verification report. This report can be a verification report generated by comparing simulation results with the expected plan; for example, it might include a maximum position deviation of 0.05 meters, a maximum attitude deviation of 1.2 degrees, and a conclusion indicating whether the verification passed. First, the trajectory and status data output from the simulation are compared point-by-point with the original optimized plan. Then, the position and attitude deviations at each comparison point are calculated. Finally, the command execution verification report is output. For example, if the calculation finds a maximum position deviation of 0.05 meters (allowable by 0.10 meters) and a maximum attitude deviation of 0.8 degrees (allowable by 1.0 degree), the report conclusion would be "passed."

[0072] Step 7: Based on the deviation data in the command execution verification report, adjust the control parameters in the terrain-adaptive synchronous command sequence to generate a transportation command sequence. The deviation data can be the difference between the simulated execution trajectory and the expected plan; for example, it could be a specific difference of 0.03 meters in the X-direction and 0.8 degrees in the pitch angle at point 200. The control parameters can be parameters used to adjust the command execution accuracy; for example, they could be the proportional coefficient, integral coefficient, derivative coefficient, and speed compensation value of the PID controller. In practice, first, analyze the deviation data exceeding the threshold in the verification report to locate the command where the deviation occurred. Then, adjust the control parameters of the corresponding command (e.g., increase speed compensation, modify PID parameters). Finally, output the calibrated transportation command sequence. For example, if a lag deviation of 0.03 meters is found at point 80, adjust the speed compensation coefficient of the forward command from 1.0 to 1.05.

[0073] The eighth step involves sending the transport instruction sequence via the vehicle control bus to the onboard control system corresponding to the target construction machinery, thereby driving the target vehicle to transport the machinery. The vehicle control bus can be a communication system between controllers within the machinery; for example, it can be a CAN bus conforming to the SAE J1939 standard, used for transmitting control instructions and status feedback. The onboard control system can be an electronic control system installed on the construction machinery; for example, it can be a complete machine control system integrating GPS, IMU, and motion controllers, responsible for parsing and executing instructions. In practice, firstly, the transport instruction sequence is converted into a bus frame format via a CAN bus adapter. Then, it is sent frame by frame to the onboard control system according to the timing sequence. Finally, the onboard control system parses the instructions and drives the actuators to complete the transport task. For example, sending 200 J1939 protocol frames at 10ms intervals to the crane ECU via the CAN bus allows the crane to automatically complete a 500-meter journey on a mountain road.

[0074] Step 9: During the execution of the aforementioned transportation instruction sequence, based on the real-time collected vehicle pose data and the expected pose data in the optimized transportation plan, execution deviation tracking and dynamic tolerance determination are performed to generate a determination result. The aforementioned vehicle pose data can refer to the real-time collected vehicle position and attitude information. For example, the vehicle pose data can be the current position coordinates (x, y, z) and attitude angles (pitch, roll, yaw) data collected by RTK-GPS and IMU sensors. The aforementioned expected pose data can refer to the theoretical position and attitude of the vehicle at the corresponding moment planned in the optimized transportation plan. For example, the aforementioned expected pose data can be the planned data that the vehicle should be located at coordinates (100.5, 50.2, 0.3) with a pitch angle of 2.5° at 100 seconds. The aforementioned deviation tracking can refer to the process of continuously monitoring the difference between the actual pose and the expected pose. For example, the aforementioned deviation tracking can be the real-time calculation of the Euclidean distance and attitude angle difference between the current position and the planned position at a frequency of 10Hz. The aforementioned dynamic tolerance determination refers to the process of dynamically adjusting the deviation tolerance threshold based on the current terrain and operating conditions. For example, the dynamic tolerance determination could be an adaptive determination that sets a position tolerance of 0.1m on flat road sections and relaxes it to 0.2m on rugged road sections. The aforementioned determination result can refer to the conclusion that the deviation exceeds the dynamic tolerance range. For example, the aforementioned "determination result" could be a Boolean value of "out of limit" or "normal". In practice, firstly, the vehicle's current position coordinates and attitude angle data are collected in real time at a frequency of 10Hz using onboard RTK-GPS and IMU sensors to generate vehicle pose data. Then, the expected pose data corresponding to the current timestamp is read from the optimized transportation plan, including the expected position coordinates and expected attitude angles. Finally, the actual pose and the expected pose are compared frame by frame to calculate the position deviation (Euclidean distance) and attitude deviation (angle difference between each axis), and the tolerance threshold is dynamically adjusted according to the current terrain conditions to generate the determination result.

[0075] Step 10: In response to the above determination result not meeting the preset conditions, a correction operation is performed on the currently pending or executing instruction in the above transportation instruction sequence, and the corrected instruction is determined as the transportation instruction sequence. The instruction correction operation includes at least: a deceleration instruction, a stop instruction, and an attitude correction instruction. The "preset conditions" can refer to deviation threshold conditions for determining whether a correction operation is needed. For example, the "preset conditions" can include a combination of conditions where the position deviation is ≤0.1m and the attitude deviation is ≤1.0°. The execution instruction correction operation can refer to an operation that modifies the current or pending instruction. For example, the execution instruction correction operation can replace the current forward instruction with a deceleration instruction and add attitude adjustment parameters. The deceleration instruction can refer to a control instruction that reduces the vehicle's speed. For example, the deceleration instruction can be a CAN bus instruction that reduces the target speed from 0.5m / s to 0.1m / s. The stop instruction can refer to a control instruction that immediately stops the vehicle. For example, the stop instruction can be an emergency stop instruction that sets the target speed to 0 and activates the parking brake. The aforementioned attitude correction command can refer to a control command that adjusts the vehicle's pitch angle, roll angle, or steering angle. For example, the attitude correction command could be a command to increase the chassis pitch angle by 3° to adapt to the current uphill gradient. In practice, firstly, the judgment result generated in step nine is received, and it is determined whether the preset conditions are met. For example, the judgment result is "exceeding limits," and the exceeding limit type is excessive position deviation. Then, the corresponding command correction operation is selected according to the deviation type and severity. For example, if the position deviation is 0.35m, exceeding the tolerance of 0.3m, a deceleration command is selected to reduce the target speed from 0.5m / s to 0.1m / s; if the attitude deviation exceeds the tolerance, an attitude correction command is selected to adjust the pitch angle or roll angle; if the deviation continues to increase and may endanger safety, a stop command is selected. Finally, the correction command is sent to the vehicle control system via the vehicle control bus, replacing or inserting it into the currently pending command sequence.

[0076] The eleventh step involves completing the transport instruction sequence as described above, thus completing the transport of the target construction equipment. In practice, firstly, the vehicle control system confirms that the last instruction in the transport instruction sequence has been successfully issued and executed. For example, confirming that the ACK receipt for the last "stop and turn off the engine" instruction has been received. Then, the final vehicle pose data is collected using RTK-GPS and IMU sensors and compared with the target pose in the optimized transport plan to confirm that the vehicle has accurately arrived at its destination and is safely stopped. Finally, a transport completion report is generated, recording the actual transport trajectory, total deviation statistics, all corrective operation logs, and final arrival accuracy, thus completing the transport of the target construction equipment.

[0077] The above-described operating steps, as an inventive point of this disclosure, solve the technical problem mentioned in the background art: "During transportation execution, there is a deviation between the actual terrain environment and the pre-generated optimized plan (such as road subsidence, temporary obstacles, GPS signal fluctuations), causing the actual driving trajectory and attitude of the machinery to deviate from the expected plan. If not detected and corrected in time, it may lead to machinery overturning, collision, or jamming, causing major safety accidents and economic losses." The reasons for the above-mentioned technical problem are as follows: Existing transportation control methods rely solely on pre-planned paths and command sequences for open-loop control, without introducing real-time pose feedback and deviation judgment mechanisms. They cannot perceive trajectory deviations and attitude anomalies during actual driving. Once unexpected deviations occur, they can only rely on manual judgment and handling by the operator, resulting in slow response speed and low reliability. This invention achieves closed-loop real-time control of the transportation process by collecting vehicle pose data in real time during the execution of a transportation command sequence and comparing it frame-by-frame with the expected pose data in the optimized transportation plan. This allows for deviation tracking and dynamic tolerance determination. When the determination result does not meet the preset conditions, it automatically executes command correction operations such as deceleration, stopping, or attitude correction. This enables the machinery to travel safely and accurately according to the optimized plan in complex terrain, saving the manpower costs of manual monitoring and emergency handling, and avoiding safety accidents and equipment maintenance costs caused by the accumulation of deviations.

[0078] In some optional implementations of certain embodiments, the aforementioned execution entity may further perform the following steps: The first step, in response to the collision detection results not meeting the preset conditions, is to determine the optimized transportation scheme as the initial transportation scheme and continue executing the first step. In practice, firstly, it is determined whether the collision detection results meet the preset conditions (e.g., zero hard collisions and all clearance margins are greater than 20% of the safety threshold). If met, the current scheme is output and the iteration ends; if not met (i.e., there are hard collisions or insufficient clearance), the optimized transportation scheme generated in this optimization is used as the new initial transportation scheme. Then, the first step is skipped, and automatic collision verification is re-executed. Finally, the above iterative loop is repeated until the collision detection results meet the preset conditions. For example, after three rounds of iteration, the number of collision events decreases from 15 to 0, meeting the conditions and outputting the final scheme.

[0079] The above-described embodiments of this disclosure have the following beneficial effects: The terrain-based construction machinery transportation method of some embodiments of this disclosure can achieve automated obstacle avoidance in construction transportation, thereby reducing collision accidents and human intervention. Specifically, the frequent collisions in traditional construction transportation processes are due to the lack of deep fusion and three-dimensional visualization of multi-source data such as terrain, features, and overhead lines, making it impossible to achieve automatic collision verification and closed-loop optimization of transportation plans. Based on this, the terrain-based construction machinery transportation method of some embodiments of this disclosure first constructs a three-dimensional scene model of the construction-related road sections based on the acquired terrain data, plotting data, and overhead line data. This achieves high-precision fusion of multi-source heterogeneous data, constructing a full-element three-dimensional visualization environment including terrain, features, and overhead lines, providing a realistic and reliable digital foundation for subsequent transportation simulation. Then, by embedding the pre-loaded three-dimensional model of the target construction machinery into the three-dimensional scene model, the initial arrangement state of the target construction machinery in the three-dimensional scene model is generated. This achieves precise anchoring and attitude initialization of the construction machinery and its three-dimensional scene model in three-dimensional space, ensuring that the starting point of the motion simulation conforms to physical reality, and providing accurate initial boundary conditions for subsequent path planning. Secondly, based on the initial layout, the target construction machinery is simulated to move along a preset route in the 3D scene model to generate an initial transportation plan. This achieves collaborative planning of the machinery's movement path and terrain posture, generating a standardized initial plan containing path point sets and posture sequences, providing iterative input data for collision verification. Finally, for the initial transportation plan, the following first step is executed: Based on the initial transportation plan, automatic collision verification is performed on the movement process of the target construction machinery to generate collision detection results. This achieves refined collision detection for continuous movement processes, accurately locating collision positions and types through discretization sampling and envelope mapping, providing a quantitative basis for optimization. In response to the collision detection results meeting preset conditions, the following second step is executed: The transportation path and transportation posture in the initial transportation plan are adjusted to generate an optimized transportation plan. This achieves closed-loop optimization based on collision detection results, automatically generating optimized plans that eliminate collision risks and satisfy kinematic constraints through classification, sorting, strategy matching, and plan fusion. A transportation instruction sequence corresponding to the optimized transportation plan is generated, and the transportation of the target construction machinery is executed according to the transportation instruction sequence. It has achieved the automated transformation from optimization scheme to mechanical execution, and completed safe and autonomous transportation in complex terrain through instruction generation, spatial synchronous verification, and dynamic feedback adjustment.

[0080] Further reference Figure 2 As an implementation of the methods shown in the above figures, this disclosure provides some embodiments of a construction machinery transportation device based on terrain data. These device embodiments are similar to... Figure 1Corresponding to the method embodiments shown, this terrain data-based construction machinery transportation device can be specifically applied to various electronic devices.

[0081] like Figure 2 As shown, a construction machinery transportation device 200 based on terrain data includes: a construction unit 201, a generation unit 202, a simulation unit 203, and an execution unit 204. The construction unit 201 is configured to: construct a three-dimensional scene model of the construction-related road sections based on acquired terrain data, plotting data, and overhead line data. The generation unit 202 is configured to: generate an initial arrangement state of the target construction machinery in the three-dimensional scene model by embedding a pre-loaded three-dimensional model of the target construction machinery into the three-dimensional scene model. The simulation unit 203 is configured to: simulate the movement of the target construction machinery along a preset route in the three-dimensional scene model based on the initial arrangement state to generate an initial transportation plan. The execution unit 204 is configured to: perform the following first step on the initial transportation plan: perform automatic collision verification on the movement process of the target construction machinery based on the initial transportation plan to generate a collision detection result. In response to the collision detection result meeting preset conditions, perform the following second step: adjust the transportation path and transportation posture in the initial transportation plan to generate an optimized transportation plan. Generate the transportation instruction sequence corresponding to the above optimized transportation scheme, and execute the transportation of the above target construction equipment according to the above transportation instruction sequence.

[0082] It is understandable that the units described in the terrain-based construction machinery transport device 200 are related to the reference... Figure 1 The steps described in the method correspond to each other. Therefore, the operations, features, and beneficial effects described above for the method are also applicable to the construction machinery transportation device 200 based on terrain data and the units contained therein, and will not be repeated here.

[0083] The following is for reference. Figure 3 It shows a schematic diagram of the structure of an electronic device (e.g., an electronic device) 300 suitable for implementing some embodiments of the present disclosure. Figure 3 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments of this disclosure.

[0084] like Figure 3As shown, the electronic device 300 may include a processing unit (e.g., a central processing unit, a graphics processing unit, etc.) 301, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 302 or a program loaded from a storage device 308 into a random access memory (RAM) 303. The RAM 303 also stores various programs and data required for the operation of the electronic device 300. The processing unit 301, ROM 302, and RAM 303 are interconnected via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0085] Typically, the following devices can be connected to I / O interface 305: input devices 306 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 307 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 308 including, for example, magnetic tapes, hard disks, etc.; and communication devices 309. Communication device 309 allows electronic device 300 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 3 An electronic device 300 with various devices is shown; however, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively. Figure 3 Each box shown can represent a device or multiple devices as needed.

[0086] In particular, according to some embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, some embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication device 309, or installed from storage device 308, or installed from ROM 302. When the computer program is executed by processing device 301, it performs the functions defined in the methods of some embodiments of this disclosure.

[0087] It should be noted that, in some embodiments of this disclosure, the computer-readable medium described above may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In some embodiments of this disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In some embodiments of this disclosure, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0088] In some implementations, clients and servers can communicate using any currently known or future-developed network protocol such as HTTP (Hypertext Transfer Protocol) and can interconnect with digital data communication (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), the Internet (e.g., the Internet of Things), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future-developed networks.

[0089] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently without being assembled into the electronic device. The aforementioned computer-readable medium carries one or more programs, which, when executed by the electronic device, cause the electronic device to: construct a three-dimensional scene model of the construction-related road section based on acquired terrain data, plotting data, and overhead line data; generate an initial arrangement state of the target construction machinery in the three-dimensional scene model by embedding a pre-loaded three-dimensional model of the target construction machinery into the three-dimensional scene model; simulate the movement of the target construction machinery along a preset route in the three-dimensional scene model based on the initial arrangement state to generate an initial transportation plan; and, for the initial transportation plan, perform the following first step: perform automatic collision verification on the movement process of the target construction machinery based on the initial transportation plan to generate a collision detection result; and, in response to the collision detection result meeting preset conditions, perform the following second step: adjust the transportation path and transportation posture in the initial transportation plan to generate an optimized transportation plan; generate a transportation instruction sequence corresponding to the optimized transportation plan, and execute the transportation of the target construction machinery according to the transportation instruction sequence.

[0090] Computer program code for performing operations of some embodiments of this disclosure can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0091] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0092] The units described in some embodiments of this disclosure can be implemented in software or hardware. The described units can also be housed in a processor; for example, a processor may be described as including a construction unit, a generation unit, a simulation unit, and an execution unit. The names of these units do not necessarily limit the unit itself; for example, a construction unit may be described as "a unit that constructs a three-dimensional scene model of the construction-related road section based on acquired terrain data, plotting data, and overhead line data of the construction-related road section."

[0093] The functions described above in this document can be performed at least in part by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), system-on-a-chip (SoCs), complex programmable logic devices (CPLDs), and so on.

[0094] The above description is merely a selection of preferred embodiments of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in the embodiments of this disclosure.

Claims

1. A method for transporting construction machinery based on terrain data, comprising: Based on the acquired terrain data, plotting data, and overhead line data of the construction-related road sections, a three-dimensional scene model of the construction-related road sections is constructed. By embedding the preloaded 3D model of the target construction machinery into the 3D scene model, the initial layout state of the target construction machinery in the 3D scene model is generated; Based on the initial layout, the target construction machinery is simulated to move along a preset route in the three-dimensional scene model to generate an initial transportation plan; For the initial transportation plan, perform the following first step: Based on the initial transportation plan, automatic collision verification is performed on the movement process of the target construction machinery to generate collision detection results; In response to the collision detection result meeting the preset conditions, the following second step is executed: The transportation path and transportation posture in the initial transportation plan are adjusted to generate an optimized transportation plan; Generate a transportation instruction sequence corresponding to the optimized transportation scheme, and execute the transportation of the target construction equipment according to the transportation instruction sequence.

2. The method according to claim 1, wherein, The method further includes: In response to the collision detection result not meeting the preset conditions, the optimized transportation plan is determined as the initial transportation plan, and the first step is continued.

3. The method according to claim 1, wherein, The step of constructing a three-dimensional scene model of the construction-related road sections based on the acquired terrain data, plotting data, and overhead line data includes: The terrain data is subjected to format conversion and coordinate normalization to generate standardized terrain data; The standardized terrain data is subjected to ground point classification and outlier removal to generate a filtered terrain point cloud model. The filtered terrain point cloud model is registered with the plotted data to generate a fused terrain and feature model. The fused terrain and feature model is reconstructed to generate a three-dimensional geometric mesh model; The three-dimensional geometric mesh model is integrated with the overhead line data to generate an interactive three-dimensional scene model.

4. The method according to claim 1, wherein, The step of generating the initial layout state of the target construction machinery in the 3D scene model by embedding the preloaded 3D model of the target construction machinery into the 3D scene model includes: Based on the three-dimensional model of the target construction machinery, a digital parameter model of the machinery is generated; Based on the transportation starting point information of the construction-related road sections and the three-dimensional scene model, the initial position and orientation of the digital parameter model in the three-dimensional scene model spatial coordinate system are determined, and a positioning matrix is ​​generated. The digital parameter model is imported into the three-dimensional scene model through the positioning matrix to generate a preliminary embedded mechanical model; Based on the terrain data in the three-dimensional scene model, spatial interference and bottom gap verification are performed on the initially embedded mechanical model to generate the initial layout state.

5. The method according to claim 1, wherein, Based on the initial layout state, simulating the movement of the target construction machinery along a preset route in the three-dimensional scene model to generate an initial transportation plan includes: Based on the preset route and the three-dimensional scene model, spatial sampling is performed according to the curvature change density to generate a discrete path sampling point sequence. Based on the path sampling point sequence and the mechanical pose in the initial arrangement state, the spatial pose and contour of the mechanical at different path positions are mapped point by point to generate a path and pose mapping table. Based on the path and attitude mapping table and the terrain data in the 3D scene model, spatial interference detection is performed to generate path safety verification results. Based on the path safety verification results, the path point sequence and corresponding attitude are adjusted to generate an initial transportation plan.

6. A construction machinery transportation device based on terrain data, comprising: The construction unit is configured to construct a three-dimensional scene model of the construction-related road section based on the acquired terrain data, plotting data, and overhead line data of the construction-related road section. The generation unit is configured to generate the initial arrangement state of the target construction machinery in the three-dimensional scene model by embedding the preloaded three-dimensional model of the target construction machinery into the three-dimensional scene model; The simulation unit is configured to simulate the movement of the target construction machinery along a preset route in the three-dimensional scene model based on the initial layout state, so as to generate an initial transportation plan; The execution unit is configured to perform the following first step for the initial transportation plan: Based on the initial transportation plan, automatic collision verification is performed on the movement process of the target construction machinery to generate collision detection results; In response to the collision detection result meeting the preset conditions, the following second step is executed: The transportation path and transportation posture in the initial transportation plan are adjusted to generate an optimized transportation plan; Generate a transportation instruction sequence corresponding to the optimized transportation scheme, and execute the transportation of the target construction equipment according to the transportation instruction sequence.

7. An electronic device, comprising: One or more processors; Storage device, on which one or more programs are stored, When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of claims 1-5.

8. A computer-readable medium having a computer program stored thereon, wherein, When the program is executed by the processor, it implements the method as described in any one of claims 1-5.

Images