Large transport vehicle traffic capacity dynamic evaluation method based on traffic obstacle information

Abstract

The application provides a large transport vehicle traffic capacity dynamic evaluation method based on traffic obstacle information, relates to the cross technical field of satellite remote sensing and intelligent traffic decision, and the technical field is as follows: candidate road network corridors are taken as geographic fences to call satellite images to construct width correction road topologies; turning radii are dynamically calculated based on the topological road vector surfaces, and minimum turning radius data are output; topologies are taken as references to identify height limit obstacles and slope data in combination with satellite images; road center line networks are extracted, three types of data are projected to generate traffic obstacle distribution data; traffic correlation envelopes are constructed by using vehicle traffic capacity constraints, multi-parameter linkage evaluation is performed by traversing obstacle data, and traffic risk heat maps are output. The technical problems of the prior art, such as road obstacle data update lag, turning space matching error and high height limit misjudgment rate, are solved. The technical effect of providing real-time and accurate traffic capacity prediction and path decision support for large transport vehicles and effectively avoiding traffic risks is achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of satellite remote sensing and intelligent traffic decision-making, in particular to a large transport vehicle traffic capacity dynamic evaluation method based on traffic obstacle information. BACKGROUND

[0002] The traditional method relies on manual field investigation or low-frequency inspection to obtain road obstacle data, which is inefficient and difficult to achieve dynamic update on a large scale, resulting in changes such as newly built overpasses and road construction that cannot be fed back in time.

[0003] In terms of key parameter calculation, existing turning radius calculation mostly uses simplified mathematical models or vehicle GPS trajectory fitting, without considering the spatial matching relationship between road topology structure and large vehicle turning trajectory envelope, making it difficult to accurately evaluate the body swing space risk.

[0004] At the same time, the identification of height-limited obstacles mainly relies on single laser radar elevation scanning or static database, which is easy to misjudge due to temporary obstacles or data noise.

[0005] In summary, the existing technology relies on manual investigation or low-frequency inspection for traffic height-limited obstacle state update, and cannot dynamically track changes such as newly built overpasses and road construction, resulting in technical problems such as road obstacle data update lag, turning space matching error, and high height-limited misjudgment rate. SUMMARY

[0006] In view of the above defects or improvement needs of the prior art, the present application provides a large transport vehicle traffic capacity dynamic evaluation method based on traffic obstacle information, which is used to solve the technical problems that the traffic height-limited obstacle state update in the prior art relies on manual investigation or low-frequency inspection, and cannot dynamically track changes such as newly built overpasses and road construction, resulting in technical problems such as road obstacle data update lag, turning space matching error, and high height-limited misjudgment rate.

[0007] To achieve the above purpose, the present application provides a large transport vehicle traffic capacity dynamic evaluation method based on traffic obstacle information, which comprises:

[0008] The transportation task parameters of a large transport vehicle are obtained in interaction, wherein the transportation task parameters include transportation start and end points, transportation distance deviation threshold and vehicle passing capacity constraint; the transportation distance deviation threshold is taken as a detour constraint, and feasible path planning is performed on an open source road network topology according to the transportation start and end points to generate a candidate road network corridor; the candidate road network corridor is taken as a geographic fence constraint, and a three-dimensional satellite image is called to construct a width-modified road topology; road vector surface topology relationships of the width-modified road topology are used for dynamic calculation of turning radii, and minimum turning radius distribution data are output; the width-modified road topology is taken as a road geometric reference, and obstacle parameter coupling identification is performed based on the three-dimensional satellite image, and height-limited obstacle distribution data and slope distribution data are output; a road center line network is extracted from the width-modified road topology, and the minimum turning radius distribution data, the height-limited obstacle distribution data and the slope distribution data are projected onto the road center line network to obtain passing obstacle distribution data; a vehicle passing correlation envelope is constructed using the vehicle passing capacity constraint, and multi-parameter linked road network level passing capacity evaluation is performed by dynamically traversing the passing obstacle distribution data, and a passing risk heat map is output.

[0009] In an embodiment, the candidate road network corridor is taken as a geographic fence constraint, a three-dimensional satellite image is called to construct a width-modified road topology, and the following processing is further performed:

[0010] The candidate road network corridor is taken as a geographic fence constraint, local satellite data is called and geographic boundary is cropped to obtain a three-dimensional satellite image and a local road network vector; a ground road network elevation model and a road network orthographic image map are generated based on the three-dimensional satellite image; initial road surface data are extracted from the road network orthographic image map, and topology correction is performed by superimposing the local road network vector to output vectorized road surface data; passing width dynamic calculation is performed by traversing the vectorized road surface data to obtain a width-modified road topology.

[0011] In an embodiment, the width-modified road topology is taken as a road geometric reference, and obstacle parameter coupling identification is performed based on the three-dimensional satellite image to output height-limited obstacle distribution data and slope distribution data, and the following processing is further performed:

[0012] Height-limited obstacle identification verification is performed according to the ground road network elevation model and the road network orthographic image map to locate the height-limited obstacle distribution data; road slope change feature identification is performed according to the ground road network elevation model and the width-modified road topology to output the slope distribution data.

[0013] In an embodiment, a vehicle passing correlation envelope is constructed using the vehicle passing capacity constraint, and multi-parameter linked road network level passing capacity evaluation is performed by dynamically traversing the passing obstacle distribution data to output a passing risk heat map, and the following processing is further performed:

[0014] extracting a vehicle body size, a transportation load state and a turning radius requirement from the vehicle traffic capacity constraint; constructing a vehicle envelope space based on the vehicle body size; superimposing the turning radius requirement to the vehicle envelope space to construct a turning envelope space, wherein the turning envelope space and the vehicle envelope space constitute the vehicle traffic correlation envelope; loading the transportation load state as a dynamic load parameter to a slope evaluation model; and performing a multi-parameter linked road network level traffic capacity evaluation by dynamically traversing the traffic obstacle distribution data with the vehicle envelope space, the turning envelope space and the slope evaluation model, and outputting the traffic risk heat map.

[0015] In an embodiment, the height-limited obstacle identification verification is performed according to the ground road network elevation model and the road network orthographic image, the height-limited obstacle distribution data is located, and the following processing is further performed:

[0016] cutting the road region elevation data from the ground road network elevation model by taking the width-corrected road topology as a mask; performing gradient calculation on the road region elevation data to obtain elevation gradient distribution data; traversing the elevation gradient distribution data by using a preset sudden elevation threshold to locate distributed sudden elevation points; locating and extracting distributed optical features of the distributed sudden elevation points in the road network orthographic image; performing permanent structure verification of the distributed sudden elevation points according to the distributed optical features to obtain distributed screened sudden points; and performing headroom height calculation on the distributed screened sudden points to output the height-limited obstacle distribution data.

[0017] In an embodiment, the headroom height calculation is performed on the distributed screened sudden points to output the height-limited obstacle distribution data, and the following processing is further performed:

[0018] locating a first road surface coordinate of a first screened sudden point based on the width-corrected road topology, and taking an elevation value of the first road surface coordinate as a first road surface elevation value; extracting a first obstacle bottom elevation of the first screened sudden point from the ground road network elevation model; calculating a difference value between the first obstacle bottom elevation and the first road surface elevation value to obtain a first single-point headroom height; performing obstacle clustering on the first screened sudden point based on spatial distribution density and optical feature consistency to obtain a first headroom height set; retrieving a first minimum headroom height as a first headroom height feature according to a descending order arrangement result of the first headroom height set; and associating the first headroom height feature and the first road surface coordinate as a first height-limited obstacle feature; and performing headroom height calculation on the distributed screened sudden points to output the height-limited obstacle distribution data.

[0019] In an embodiment, the width-corrected road topology is used to dynamically calculate the turning radius, and the minimum turning radius distribution data is outputted.

[0020] The road segment angles and effective passing widths of the intersection nodes of the width-corrected road topology are extracted. At the intersection nodes of the road vector surfaces, the road segment connection directions are used as the vehicle turning constraints, the road segment angles and effective passing widths are used to dynamically simulate the envelope space of the vehicle turning trajectory, and the swept area boundaries are located. The intersection nodes of the road vector surfaces are used as the centers, and the swept area boundaries are used to perform inversion calculation to obtain the minimum turning radius values. The minimum turning radius values, the intersection nodes of the road vector surfaces and the road segment connection directions are bound, and the minimum turning radius distribution data is outputted.

[0021] In an embodiment, the vehicle envelope space, the turning envelope space and the slope evaluation model are used to dynamically traverse the passing obstacle distribution data to perform multi-parameter linkage road network level passing capacity evaluation, and the passing risk heat map is outputted. The following processes are further performed:

[0022] The road center line network is rasterized into an evaluation unit array at a preset distance interval, and the passing obstacle distribution data is decomposed into a passing obstacle unit array according to the evaluation unit array. After a first passing obstacle unit is loaded, the vehicle envelope space is dynamically driven to verify the height-limited passing capacity, the turning envelope space is driven to verify the turning passing capacity, and the slope evaluation model is driven to verify the climbing load capacity. The first height-limited passing feasibility, the first turning passing feasibility and the first slope load matching feasibility are outputted. The first height-limited passing feasibility, the first turning passing feasibility and the first slope load matching feasibility are fused to generate a first passing risk value. The vehicle envelope space, the turning envelope space and the slope evaluation model are used to dynamically traverse the passing obstacle unit array to perform multi-parameter linkage road network level passing capacity evaluation, and a passing risk heat array is generated. The passing risk heat array is periodically dynamically rendered and restored, and the passing risk heat map is outputted.

[0023] In an embodiment, the following processes are further performed:

[0024] The passing obstacle distribution characteristics of the multiple candidate paths are extracted by projecting the candidate road network corridors onto the passing risk heat map. The candidate path passing score matrix is constructed based on the passing obstacle distribution characteristics. The target reliable passing road is located in the multiple candidate paths according to the candidate path passing score matrix.

[0025] The one or more technical solutions provided in the application have at least the following technical effects or advantages:

[0026] The embodiment of the application realizes real-time tracking of road obstacle states and dynamic refreshing of passing obstacle parameters based on a periodic updating mechanism of satellite data; through accurate simulation of the matching relationship between the turning trajectory envelope of a vehicle and road space, the problem of inaccurate risk assessment of turning space for large vehicles is completely solved; combined with cross verification of multiple remote sensing sources, the false judgment interference of height limit obstacles is significantly inhibited. Finally, real-time and accurate passing capacity prediction and path decision support are provided for large transport vehicles, effectively avoiding the technical effect of passing risks caused by newly-built height limit structures, road construction or sudden changes in slope. BRIEF DESCRIPTION OF DRAWINGS

[0027] In order to more clearly illustrate the technical solutions in the embodiments of the application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description only constitute some embodiments of the application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of these drawings.

[0028] Figure 1 A flowchart of a large transport vehicle passing capacity dynamic evaluation method based on passing obstacle information provided by the application is shown;

[0029] Figure 2 A flowchart of constructing a width correction road topology in a large transport vehicle passing capacity dynamic evaluation method based on passing obstacle information provided by the application is shown. DETAILED DESCRIPTION

[0030] The application provides a large transport vehicle passing capacity dynamic evaluation method based on passing obstacle information, which is used to solve the technical problems in the prior art that the passing height limit obstacle state update depends on manual survey or low-frequency inspection, and cannot dynamically track changes such as newly-built overpasses and road construction, resulting in road obstacle data update lag, turning space matching error and high height limit false judgment rate.

[0031] In order to make the purpose, technical solutions and advantages of the application more clear, the following will further describe the application in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the application, and are not used to limit the application. In addition, the technical features involved in each embodiment of the application described below can be combined with each other as long as they do not conflict with each other.

[0032] In the description of the present application, it should be understood that the terms "center", "longitudinal", "transverse", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise", and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, which is only for the convenience of describing the present application and simplifying the description, and does not indicate or imply that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation on the present application.

[0033] Unless otherwise explicitly stated, throughout the specification and claims, the term "comprise" or variations such as "comprises" or "comprising", etc. will be understood to encompass the stated element or components, without excluding other elements or components.

[0034] Embodiment, the flowchart of the large transport vehicle traffic capacity dynamic evaluation method based on traffic obstacle information provided by the embodiment of the present application, see Figure 1 , the method comprises:

[0035] P100: interactively obtaining the transport task parameters of the large transport vehicle, wherein the transport task parameters include the transport start and end points, the transport distance deviation threshold, and the vehicle traffic capacity constraint.

[0036] Specifically, the embodiment obtains a set of key parameters required by the large transport vehicle to perform the transport task through a human-computer interaction mode, obtains the transport task parameters, and specifically includes the transport start and end points defining the geographic coordinates of the transport task, the maximum distance tolerance value representing the maximum distance allowed for path planning to deviate from the optimal path, the transport distance deviation threshold for constraining the range of detours, and the constraint condition reflecting the physical characteristics of the vehicle itself, covering the vehicle body outline size, minimum turning radius, load state, and other direct influence on road traffic safety properties The vehicle traffic capacity constraint.

[0037] P200: performing feasible path planning on an open source road network topology based on the transport distance deviation threshold as the detour constraint, and generating a candidate road network corridor.

[0038] It should be understood that the open source road network topology is a digital model of the road network constructed based on public geographic information data such as OpenStreetMap, which includes topological connection relationships and basic attributes such as road intersections as nodes and road segments as edges. The basic attributes here include but are not limited to road grade and direction.

[0039] The candidate road network corridor is a set of multiple feasible paths that satisfy the detour constraint, forming a strip-shaped spatial region.

[0040] The embodiment takes the transportation distance deviation threshold as a rigid constraint condition of the path bypassing range, takes the transportation start and end points as path endpoints in the open source road network topology, searches all feasible paths with a total length not exceeding "the shortest path length + the transportation distance deviation threshold" through a shortest path algorithm such as Dijkstra or A* algorithm, and finally aggregates to generate a spatial strip region covering these paths, that is, the candidate road network corridor.

[0041] P300: taking the candidate road network corridor as a geographic fence constraint, calling stereoscopic satellite images to construct a width-modified road topology.

[0042] In an implementation manner, referring to Figure 2 taking the candidate road network corridor as a geographic fence constraint, calling stereoscopic satellite images to construct a width-modified road topology, the method provided by the embodiment comprises the following steps:

[0043] P310: taking the candidate road network corridor as a geographic fence constraint, performing local satellite data calling and geographic boundary clipping to obtain stereoscopic satellite images and local road network vectors.

[0044] P320: fitting to generate a road network elevation model and a road network orthographic image based on the stereoscopic satellite images.

[0045] P330: extracting initial road surface data from the road network orthographic image, and performing topology correction by superimposing the local road network vectors to output vectorized road surface data.

[0046] P340: performing dynamic calculation of the passing width by traversing the vectorized road surface data to obtain a width-modified road topology.

[0047] It should be understood that the geographic fence constraint is to take a geographic coordinate range as a spatial screening condition to limit the target region of data operation, and the stereoscopic satellite images are three-dimensional image data of the ground surface obtained through satellite stereoscopic imaging technology, and contain elevation information.

[0048] In the embodiment, the spatial range of the candidate road network corridor is taken as a geographic fence constraint to call stereoscopic satellite images covering the region, and a road network model containing accurate passing width information, that is, the width-modified road topology, is constructed through road surface extraction and topology correction processing.

[0049] Specifically, the embodiment takes the candidate road network corridor as a spatial boundary to screen and clip satellite raw data covering the region to obtain local stereoscopic satellite images, and extracts road vector line data of the same region from an open source map to obtain the local road network vectors.

[0050] Based on the cropped stereo satellite images, two core geographic information products are generated by photogrammetry technology, one is the road network elevation model, and the other is the road network orthophoto map.

[0051] The road network elevation model is a continuous grid surface model representing the ground elevation, including the height of roads, buildings and other ground objects, and the road network orthophoto map is a geometrically corrected non-distorted optical satellite image, which eliminates the influence of terrain undulations and sensor tilt.

[0052] The road network elevation model and the road network orthophoto map obtained in this embodiment use conventional mature technology and will not be described here.

[0053] The road area vector data containing geometric errors is extracted from the road network orthophoto map to obtain the initial road surface data, and then the spatial position correction and topological relationship optimization are performed by superimposing the local road network vector to output the vectorized road surface data with improved geometric accuracy.

[0054] Each vectorized road surface unit in the vectorized road surface data is traversed, and the actual passing width is dynamically calculated according to the geometric shape of each unit such as road edge curvature and intersection expansion area, and finally the width corrected road topology with width attribute is generated, providing reliable data support for subsequent spatial analysis such as turning radius.

[0055] P400: Based on the road vector surface topological relationship of the width corrected road topology, the turning radius is dynamically calculated, and the minimum turning radius distribution data is output.

[0056] In one implementation, based on the road vector surface topological relationship of the width corrected road topology, the turning radius is dynamically calculated, and the minimum turning radius distribution data is output, and the method steps P400 provided by the present application include:

[0057] P410: Extracting a plurality of road segment angles and a plurality of effective passing widths of a plurality of road vector surface intersection nodes in the width corrected road topology.

[0058] P420: At the plurality of road vector surface intersection nodes, the connection direction of the plurality of road segments is used as the vehicle turning constraint, the vehicle turning trajectory envelope space is dynamically simulated according to the plurality of road segment angles and the plurality of effective passing widths, and the plurality of swept area boundaries are located.

[0059] P430: Taking the plurality of road vector surface intersection nodes as the center, the plurality of swept area boundaries are inversely calculated to obtain a plurality of minimum turning radius values.

[0060] P440: Binding the plurality of minimum turning radius values, the plurality of road vector surface intersection nodes and the plurality of road segment connection directions, and outputting the minimum turning radius distribution data.

[0061] Specifically, the road vector surface intersection node refers to a spatial point where a plurality of road surface vector polygons intersect, representing the geometric center position of a road intersection or a branch point. The road segment included angle is the plane angle difference between the directions of the center line vectors of adjacent road segments, such as 90 degrees at a crossroad. The effective passing width refers to the actual passable road surface width after deducting the kerbstone and temporary obstacles, which is dynamically calculated based on the road surface boundary geometry.

[0062] In the present embodiment, the spatial coordinates of all road intersection nodes are extracted from the width-corrected road topology model, the direction included angle between the connected road segments at each node is calculated, such as by using the dot product of the adjacent road center line vectors and the inverse trigonometric function, and the effective passing width values of the road segments near the node are obtained, such as by removing the fixed obstacle occupied area based on road surface buffer analysis, to form a structured parameter set containing the node position, road segment included angle and passing width.

[0063] The vehicle turning constraint refers to the legal vehicle turning rule defined in the topology connection direction, such as the prohibition of U-turn at a T-shaped intersection. The vehicle turning trajectory envelope space refers to the three-dimensional space region swept by the vehicle body during turning, the boundary of which is determined by the vehicle body size, turning angle, wheel angle and road geometry. It should be noted that the vehicle turning trajectory envelope space covers the universal envelope of all types of passing vehicles, and the turning sweeping path of different vehicle models is simulated through geometric kinematics simulation, which universally represents the minimum physical space boundary required for safe vehicle passing. The swept area boundary is the outer contour polygon of the envelope space, representing the minimum physical space range required for vehicle turning.

[0064] For each road vector surface intersection node (i.e. road intersection node), the drivable path of the vehicle is defined in the topology connection direction, and the road segment included angle and effective passing width of the node are combined to dynamically simulate the vehicle turning process through kinematic geometry model. Specifically, the vehicle body sweeping path under the maximum steering angle of the front wheel is calculated with the rear axle center as the reference, and the swept area boundary is generated as the geometric basis for reversing the turning radius.

[0065] By analogy, at the plurality of road vector surface intersection nodes, the vehicle turning constraint is defined in the connection direction of the plurality of road segments, and the plurality of road segment included angles and the plurality of effective passing widths are used to dynamically simulate the vehicle turning trajectory envelope space, and the plurality of swept area boundaries are located.

[0066] Taking a road vector plane intersection node as a center, all legal turning directions corresponding to a swept area boundary polygon of the node are traversed, a minimum circumscribed circle radius capable of completely containing the boundary polygon under each turning path is calculated, a radius value obtained by iteratively solving a circle equation fitted through polygon vertices is taken as a minimum passable turning radius of the turning direction, it is ensured that a vehicle body envelope space is always located within a road passing area when the vehicle turns, and finally a plurality of minimum turning radius values at the plurality of road vector plane intersection nodes are obtained.

[0067] The plurality of minimum turning radius values, the plurality of road vector plane intersection nodes and the plurality of road segment connection directions are bound for data structural recording, and minimum turning radius distribution data is output, which can support a minimum turning radius distribution database for spatial query and path analysis.

[0068] P500: Taking a width corrected road topology as a road geometry reference, obstacle parameter coupling identification is performed based on the stereoscopic satellite image, and height-limited obstacle distribution data and slope distribution data are output.

[0069] In an implementation manner, the width corrected road topology is taken as the road geometry reference, the obstacle parameter coupling identification is performed based on the stereoscopic satellite image, and the height-limited obstacle distribution data and the slope distribution data are output, and the method step P500 provided by the present application comprises the following steps:

[0070] P510: Height-limited obstacle identification verification is performed according to the ground road network elevation model and the road network orthographic image map, and the height-limited obstacle distribution data is located.

[0071] P520: Road slope change feature identification is performed according to the ground road network elevation model and the width corrected road topology, and the slope distribution data is output.

[0072] In an implementation manner, the height-limited obstacle identification verification is performed according to the ground road network elevation model and the road network orthographic image map, and the height-limited obstacle distribution data is located, and the method step P510 provided by the present application comprises the following steps:

[0073] P511: Taking the width corrected road topology as a mask, road region elevation data is obtained by cutting the ground road network elevation model.

[0074] P512: Gradient calculation is performed on the road region elevation data to obtain elevation gradient distribution data.

[0075] P513: A preset sudden elevation threshold is used to traverse the elevation gradient distribution data to locate distributed elevation sudden change points.

[0076] P514: Distributed optical features of the distributed elevation sudden change points are located and extracted in the road network orthographic image map.

[0077] P515: performing permanent structure verification of the distributed height break points according to the distributed optical features to obtain distributed screening break points.

[0078] P516: performing clearance height calculation on the distributed screening break points to output the height-limited obstacle distribution data.

[0079] In an implementation, the clearance height calculation is performed on the distributed screening break points to output the height-limited obstacle distribution data, and the method step P516 provided by the application comprises:

[0080] P516-1: positioning a first road surface coordinate of a first screening break point based on the width-corrected road topology, and taking an elevation value of the first road surface coordinate as a first road surface elevation value.

[0081] P516-2: extracting a first obstacle bottom elevation of the first screening break point from the ground surface road network elevation model.

[0082] P516-3: calculating a difference between the first obstacle bottom elevation and the first road surface elevation value to obtain a first single-point clearance height.

[0083] P516-4: performing obstacle clustering based on spatial distribution density and optical feature consistency at the first screening break point by analogy to obtain a first clearance height set.

[0084] P516-5: retrieving a first minimum clearance height as a first clearance height feature according to a descending order arrangement result of the first clearance height set.

[0085] P516-6: associating the first clearance height feature and the first road surface coordinate as a first height-limited obstacle feature.

[0086] P516-7: performing clearance height calculation on the distributed screening break points by analogy to output the height-limited obstacle distribution data.

[0087] Specifically, the width-corrected road topology of the road surface vector polygon obtained in the foregoing is taken as a spatial mask, and road region elevation data in a road region within a ground surface road network elevation model (DSM) is cropped, the road region elevation data only retains grid elevation values within a road surface range, eliminates surrounding ground object interference, and provides a pure road elevation reference for subsequent gradient calculation.

[0088] Gradient calculation is performed on the road area elevation data, preferably by a spatial convolution algorithm such as the Sobel operator, to calculate the elevation change rate pixel by pixel, and generate elevation gradient distribution raster data representing the steepness of the road surface, wherein high gradient values correspond to steep slopes or abrupt points.

[0089] A preset abrupt elevation threshold such as a height difference of ≥2 meters is used to traverse the elevation gradient distribution data, compare the gradient values pixel by pixel, and screen grid points exceeding the threshold to locate distributed elevation abrupt points, forming an initial coordinate set of candidate height-restricted obstacles. It should be understood that the abrupt points here may be overpass pillars, road bumps, etc.

[0090] After locating the coordinates of the distributed elevation abrupt points in the road network DOM (DOM), the distributed optical features of the coordinates are extracted. The specific extraction method is to obtain the texture features of the abrupt point area, such as the roughness of concrete, the color features, such as the gray system, and the shape features, such as the columnar or arched shape, to obtain the distributed optical features corresponding to the distributed elevation abrupt points, providing a basis for subsequent permanent structure verification.

[0091] The specific technical implementation of performing permanent structure verification according to the distributed optical features is as follows:

[0092] The gray level co-occurrence matrix is used to calculate the roughness and contrast to quantitatively analyze the texture structure complexity of the optical features corresponding to the abrupt points; the separation threshold of vegetation and non-structure in the RGB / HSV color space is extracted to analyze the optical features corresponding to the abrupt points to obtain the spectral reflection characteristics; and whether the abrupt points correspond to the optical features of the bridge arch, the gantry rectangle, etc. is judged based on the contour polygon fitting to analyze the abrupt points corresponding to the optical features to obtain the geometric regularity.

[0093] A decision tree of optical features is constructed to traverse the texture structure complexity, spectral reflection characteristics, and geometric regularity of each abrupt point. First, low-complexity textures such as vegetation shadows, high-green components such as vegetation, or non-structured contours such as temporary roadblocks are excluded. Then, multi-scale sliding window verification is performed to exclude local noise, and finally, only high-texture-complexity, low-vegetation-reflectivity, and contour-geometric-regularity permanent structures with high confidence are retained. The distributed screening abrupt points that have passed the multi-dimensional optical verification are output. The distributed screening abrupt points refer to the spatial coordinate set of the permanent height-restricted obstacle candidate points such as overpass pillars and gantry pedestals.

[0094] The clear height of the distributed screening abrupt points is calculated, and the height-restricted obstacle distribution data is output. The height-restricted obstacle distribution data is a spatial vector data set, each record of which contains obstacle coordinates, type, and calculated minimum clear height value, used to represent the road network height-restricted constraint.

[0095] The embodiment calculates the clearance height of the distributed screening mutation point, and outputs the specific technical implementation of the height-limited obstacle distribution data as follows.

[0096] Since the method of calculating the clearance height of each mutation point is the same, the embodiment takes the first height-limited obstacle feature of the first screening mutation point as an example to elaborate the technical solution in detail.

[0097] Based on the width correction road topology positioning of the road surface projection coordinates corresponding to the first screening mutation point, the first road surface coordinates are obtained, and then the spatial projection algorithm is used to vertically map the mutation point to the road surface polygon to obtain the road surface coordinate point directly below as the elevation reference position. Subsequently, the elevation value of the projection coordinate point is extracted from the ground road network elevation model as the road surface elevation reference value associated with the current obstacle to obtain the first road surface elevation value. The elevation value is essentially the road surface elevation.

[0098] The first obstacle bottom elevation of the first screening mutation point is extracted from the ground road network elevation model. The elevation value can be the lowest point elevation of the overpass.

[0099] The difference between the first obstacle bottom elevation and the first road surface elevation value is calculated to obtain the first single-point clearance height representing the local minimum passing height of the obstacle corresponding to the mutation point.

[0100] Using the same technical means to obtain the single-point clearance height, based on the spatial distribution density of other elevation mutation points within a radius of 5 meters and the optical feature consistency such as texture structure complexity and color histogram similarity consistency extracted from the road network orthographic image, obstacle clustering analysis is performed. Specifically, the DBSCAN density clustering algorithm is used to merge spatially adjacent and optically feature-matched mutation points to form a point cluster belonging to the same physical obstacle, and aggregate all single-point clearance height values in the cluster to form a first clearance height set representing the height measurement value sequence of the same obstacle at different positions, providing a data basis for subsequent extraction of the most unfavorable passing height feature.

[0101] According to the descending order arrangement result of the first clearance height set, the first minimum clearance height is retrieved as the first clearance height feature. The first clearance height feature represents the most unfavorable position height of the obstacle for vehicle passing, ensuring that the safety evaluation covers the most stringent scenario and avoiding the risk of misjudgment due to excessive local height allowance.

[0102] The determined first clearance height feature is associated with the corresponding first road surface coordinates in terms of spatial attributes to construct a structured record containing the coordinate point longitude and latitude, minimum clearance height value, and obstacle type code, forming a first height-limited obstacle feature representing the spatial position and passing constraint of a single height-limited obstacle.

[0103] Traverse all distributed screening mutation points, loop execute P516-1 to P516-6 standardization process, calculate the clearance height of the distributed screening mutation points, output the height limit obstacle distribution data.

[0104] According to the road elevation model and the width-corrected road topology, the road slope change characteristics are identified, and the specific technology of outputting the slope distribution data is as follows:

[0105] First, each face unit in the width-corrected road topology is taken as a processing unit, and the elevation grid points in the coverage range of the road face are extracted from the road elevation model; then the spatial gradient algorithm such as Sobel operator convolution kernel is used to calculate the two-dimensional directional derivative of the elevation point array, and the slope value of each grid point in the road face is generated; then the slope values of all grid points in the section are aggregated according to the geometric segmentation of the road topology (such as every 10 meters), and the arithmetic mean value is calculated as the representative slope of the section; finally, the slope values of all road sections in the road network are bound with their spatial positions and topological directions, and the slope distribution data is output, each record in the slope distribution data contains the coordinates of the start and end points of the road section, the length and the average slope value, which completely represents the slope change characteristics of the road network.

[0106] P600: Extract the road centerline network from the width-corrected road topology, and obtain the traffic obstacle distribution data by projecting the minimum turning radius distribution data, height limit obstacle distribution data and slope distribution data onto the road centerline network.

[0107] This embodiment first uses the centerline extraction algorithm to process the width-corrected road topology, specifically, traverses each road face vector polygon, generates a geometric centerline through Voronoi graph skeletonization or boundary line equidistant offset convergence calculation, and then constructs a topologically coherent road centerline network based on node connection rules.

[0108] Taking the road intersection node as the anchor point, associate several minimum turning radius values in the minimum turning radius distribution data to the corresponding node of the centerline network, and inherit the turning topology relationship. Through spatial nearest neighbor analysis, vertically map several height limit points in the height limit obstacle distribution data to the nearest road centerline segment, record the linear reference distance of the projection point from the start point of the road segment, and bind the clearance height value. Perform spatial overlay analysis on the vector line segment of the slope distribution data and the road centerline, cut the slope line segment according to the centerline segmentation, calculate the weighted average slope value in each centerline segment, and finally integrate to generate the traffic obstacle distribution data.

[0109] The traffic obstacle distribution data takes a road center line network as a spatial carrier, and each center line segment stores three types of attributes: a set of turning radii at nodes, a sequence of height limit points along a line, and a road segment slope value, forming a structured linear obstacle database supporting dynamic traffic evaluation.

[0110] P700: A vehicle traffic envelope is constructed using vehicle traffic capacity constraints, and a multi-parameter linked road network level traffic capacity evaluation is performed by dynamically traversing the traffic obstacle distribution data, and a traffic risk heat map is output.

[0111] In an implementation, a vehicle traffic envelope is constructed using vehicle traffic capacity constraints, and a multi-parameter linked road network level traffic capacity evaluation is performed by dynamically traversing the traffic obstacle distribution data, and a traffic risk heat map is output, and the method step P700 provided by the present application comprises:

[0112] P710: The vehicle body size, transportation load state, and turning radius requirement are extracted from the vehicle traffic capacity constraints.

[0113] P720: A vehicle envelope space is constructed based on the vehicle body size.

[0114] P730: The turning radius requirement is superimposed on the vehicle envelope space to construct a turning envelope space, wherein the turning envelope space and the vehicle envelope space constitute the vehicle traffic correlation envelope.

[0115] P740: The transportation load state is loaded as a dynamic load parameter into a slope evaluation model.

[0116] P750: The vehicle envelope space, turning envelope space, and slope evaluation model are used to dynamically traverse the traffic obstacle distribution data to perform a multi-parameter linked road network level traffic capacity evaluation, and the traffic risk heat map is output.

[0117] In an implementation, the vehicle envelope space, turning envelope space, and slope evaluation model are used to dynamically traverse the traffic obstacle distribution data to perform a multi-parameter linked road network level traffic capacity evaluation, and the traffic risk heat map is output, and the method step P750 provided by the present application comprises:

[0118] P751: The road center line network is rasterized into an evaluation unit array at a preset distance interval, and the traffic obstacle distribution data is decomposed into a traffic obstacle unit array according to the evaluation unit array.

[0119] P752: After loading the first passage obstacle unit, dynamically drive the vehicle envelope space to verify the height-limited passage capability, drive the turning envelope space to verify the turning passage capability, drive the slope assessment model to verify the climbing load capacity, and output the first height-limited passage feasibility, the first turning passage feasibility and the first slope load matching feasibility.

[0120] P753: Fuse the first height-limited passage feasibility, the first turning passage feasibility and the first slope load matching feasibility to generate a first passage risk value.

[0121] P754: Analogous to the vehicle envelope space, the turning envelope space and the slope assessment model, dynamically traverse the passage obstacle unit array to perform multi-parameter linkage road network level passage capability evaluation, and generate a passage risk heat array.

[0122] P755: Periodically dynamically render the passage risk heat array to output the passage risk heat map.

[0123] Specifically, based on the foregoing, the vehicle passage capability constraint includes vehicle body size, transportation load state and turning radius requirement. The vehicle body size specifically includes the maximum length, width and height values of the vehicle contour. The transportation load state is the current loading mass and distribution of the vehicle, such as 40 tons full load and 1.8m centroid height. The turning radius requirement refers to the minimum safe turning circle radius allowed by the vehicle steering system, such as 12m for articulated vehicles. The above three data provide input reference for constructing the passage envelope.

[0124] A local coordinate system is established along the vehicle body direction with the center of the rear axle of the vehicle as the coordinate origin, generating a cubic OBB model with dimensions equal to the length x width x height of the vehicle, as the vehicle envelope space. The vehicle envelope space represents the maximum spatial occupation range of the vehicle in a static state, and is used for height-limited obstacle collision detection.

[0125] Based on the steering kinematics model, a circular trajectory corresponding to the turning radius requirement is generated in the horizontal plane with the center of the rear axle of the vehicle as the rotation reference point. The vehicle envelope cube is stretched along the trajectory to form a swept volume space, which is the turning envelope space.

[0126] The turning envelope space and the vehicle envelope space jointly constitute the vehicle passage associated envelope, which is used for spatial intrusion simulation in the turning scenario.

[0127] The transportation load state is loaded into the slope assessment model, and the correlation rule between load and climbing ability is established through vehicle dynamics principle. Specifically, the function of the slope assessment model is to simulate the actual climbing performance of the vehicle on different slope roads based on the total mass distribution of the vehicle under the current load state, combined with the engine output torque characteristics, transmission system efficiency and tire rolling resistance coefficient.

[0128] The slope assessment model determines the maximum slope threshold at which the vehicle can safely pass under this load condition by calculating the dynamic balance between the available traction force of the vehicle's drive wheels and the total resistance that needs to be overcome to climb the slope. This provides a dynamically changing load adaptability criterion for subsequent slope obstacle verification.

[0129] This embodiment employs three computational engines: a vehicle envelope space for height restriction verification, a turning envelope space for turning verification, and a slope assessment model for climbing verification. It loads traffic obstacle distribution data segment by segment along the road centerline network and performs three types of verifications simultaneously: comparing the vehicle envelope height with the height restriction value, detecting the topological inclusiveness of the turning envelope space and the road boundary, and calculating the matching between the current slope and the load-bearing climbing capacity. Finally, the fused results generate a heat map of traffic risk for the entire road network.

[0130] Specifically, the road centerline network is linearly referenced and cut at preset intervals (e.g., 10 meters) to generate a discretized array of evaluation units, where each unit is an independent evaluation object.

[0131] Based on the evaluation unit array, the traffic obstacle distribution data is simultaneously decomposed according to the same linear reference rule. Specifically, the turning radius parameter is bound to the centerline node, the height limit parameter is associated with the nearest unit according to the projection distance, and the slope value is distributed to each unit according to the length weight, forming a traffic obstacle unit array that is spatially aligned with the evaluation unit array.

[0132] The first obstacle unit is the first 10-meter road segment. After loading the first obstacle unit, the three-engine verification is triggered. Specifically, the vehicle envelope space is driven to verify the height restriction passage capability by extracting the net height of all height restriction points within the unit, taking the minimum value and comparing it with the vehicle envelope height. If the vehicle envelope space meets the minimum net height, feasibility is output; otherwise, it is 1. The turning envelope space is driven to verify the turning passage capability by superimposing the turning envelope space onto the road network orthophoto at road intersection nodes and detecting whether the envelope exceeds the road surface boundary through grid occupancy analysis. If it does not exceed the boundary, feasibility is output; otherwise, it is 1. The slope evaluation model is driven to verify the climbing load capacity by inputting the unit slope value into the slope evaluation model. If the unit slope value is less than the current load climbing threshold, feasibility is output; otherwise, it is 1.

[0133] The final output includes a triplet result of the feasibility of passing through the first height restriction, the feasibility of passing through the first turn, and the feasibility of matching the first slope load.

[0134] Wherein, any feasibility is set to 1 to trigger a global impassable determination, ensuring the conservatism and safety of risk assessment, in a non-1 feasibility scenario, the first height limit passage feasibility is the height difference between the vehicle envelope space and the minimum net height, the first turning passage feasibility is the minimum distance between the turning envelope space and the road outer boundary, and the first slope load matching feasibility is the critical slope difference between the maximum climbing ability of the vehicle under the current load state and the unit slope value.

[0135] In a non-1 feasibility scenario, based on a preset weight distribution, the first height limit passage feasibility, the first turning passage feasibility and the first slope load matching feasibility are fused to generate a first passage risk value.

[0136] When any feasibility is set to 1, the first passage risk value is 1, indicating that it is not feasible.

[0137] By analogy, the vehicle envelope space, the turning envelope space and the slope evaluation model are used to dynamically traverse the passage obstacle unit array to perform multi-parameter linked road network level passage capacity evaluation, and finally generate a passage risk heat array consistent with the spatial distribution of the road center line.

[0138] The passage risk heat array is converted into a continuous grid surface through a spatial interpolation algorithm, a color gradient is assigned according to the risk value range, and is superimposed on the road network base map and periodically refreshed to output a dynamically visualized passage risk heat map, which reflects the distribution of the road network passage risk in real time.

[0139] The embodiment realizes dynamic quantification of road network passage capacity with changes in vehicle attributes and environment through dynamic updating of satellite data, multi-obstacle coupling analysis and spatialized risk assessment, and provides a reliable reference for large freight vehicle passage route selection.

[0140] Based on the periodic updating mechanism of satellite data, the embodiment realizes real-time tracking of road obstacle states and dynamic refreshing of passage obstacle parameters; by accurately simulating the matching relationship between the vehicle turning trajectory envelope and the road space, the problem of inaccurate risk assessment of large vehicles turning in space is completely solved; combined with multi-source remote sensing cross verification, the false judgment interference of height limit obstacles is significantly suppressed. Ultimately, real-time and accurate passage capacity prediction and path decision support are provided for large transport vehicles, effectively avoiding the technical effects of passage risks caused by newly built height limit structures, road construction or sudden changes in slope.

[0141] In one implementation manner, the method provided by the present application further includes:

[0142] P810: Extract a plurality of passage obstacle distribution characteristics of a plurality of candidate paths by projecting the candidate road network corridor to the passage risk heat map.

[0143] P820: constructing a candidate path traffic score matrix based on the plurality of traffic obstacle distribution characteristics.

[0144] P830: screening a positioning target reliable traffic road from the plurality of candidate paths according to the candidate path traffic score matrix.

[0145] Specifically, in the embodiment, the candidate road network corridor is spatially superimposed and analyzed with the traffic risk heat map, a risk value sequence of all grid cells passed by each path is extracted through a geographic information system tool, and the sequence corresponding to the path with a traffic risk value of 1 is directly eliminated, and a plurality of candidate paths are screened.

[0146] Further, the overall average risk, the highest risk point value, the number of height limit obstacles, and the cumulative proportion of steep slope sections of the plurality of candidate paths are calculated, and a plurality of obstacle distribution characteristic vectors quantitatively representing path traffic safety are formed.

[0147] Based on the obstacle distribution characteristic vectors of each candidate path, the features of different dimensions are first normalized to score values in the range of zero to one, wherein a high score represents a low risk, and then a weight coefficient is assigned according to the importance of the three types of obstacles, i.e., height limit, slope, and turning, to generate a comprehensive safety score of each path, and finally a two-dimensional matrix with paths as rows and feature scores as columns is constructed to realize the structured and quantitative comparison of candidate path traffic safety.

[0148] According to the candidate path traffic score matrix, multi-level screening is performed, specifically, high-risk paths with a comprehensive score lower than a safety threshold (such as a defined threshold of 0.6) are eliminated first, and then in the remaining paths, the non-dominated sorting algorithm is used to screen a Pareto optimal solution set with the highest safety score and the shortest path as dual objectives; finally, a detailed risk report of the top three paths is output to assist users in selecting target traffic roads with safety and efficiency.

[0149] The above only describes the preferred embodiments of the present application, but the protection scope of the present application is not limited thereto, any person skilled in the art can easily think of changes or replacements within the technical range disclosed by the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for dynamic evaluation of the traffic capacity of large transport vehicles based on traffic obstacle information, characterized in that, include: The transportation task parameters of large transport vehicles are obtained interactively, wherein the transportation task parameters include transportation start and end points, transportation distance deviation threshold and vehicle traffic capacity constraints. Using the transportation distance deviation threshold as a detour constraint, feasible path planning is performed on the open-source road network topology based on the transportation start and end points to generate candidate road network corridors; The candidate road network corridors are used as geofence constraints, and stereo satellite imagery is retrieved to construct a width-corrected road topology. Based on the road vector surface topology relationship of the width-corrected road topology, the turning radius is dynamically calculated, and the minimum turning radius distribution data is output. Using the width-corrected road topology as the road geometric reference, obstacle parameter coupling identification is performed based on the stereo satellite imagery, and height-restricted obstacle distribution data and slope distribution data are output. The road centerline network is extracted from the width-corrected road topology, and the traffic obstacle distribution data is obtained by spatially projecting the minimum turning radius distribution data, height restriction obstacle distribution data, and slope distribution data onto the road centerline network. A vehicle traffic association envelope is constructed using vehicle traffic capacity constraints. The road network-level traffic capacity assessment is performed by dynamically traversing the traffic obstacle distribution data and outputting a traffic risk heat map. Specifically, the candidate road network corridors are used as geofencing constraints, and stereo satellite imagery is retrieved to construct a width-corrected road topology, including: The candidate road network corridors are used as geofence constraints to retrieve local satellite data and crop geographical boundaries, thereby obtaining stereo satellite imagery and local road network vectors. Based on the stereo satellite imagery, a surface road network elevation model and a road network orthophoto map are generated. Initial road surface data is extracted from the road network orthophoto map, and topology correction is performed by superimposing the local road network vector to output vectorized road surface data. The vectorized road surface data is traversed to dynamically calculate the traffic width and obtain the width-corrected road topology. A vehicle traffic association envelope is constructed using vehicle capacity constraints. The traffic obstacle distribution data is dynamically traversed to perform a multi-parameter linked road network-level traffic capacity assessment, outputting a traffic risk heatmap, including: Extract vehicle body dimensions, transport load status, and turning radius requirements from the vehicle traffic capacity constraints; The vehicle envelope space is constructed based on the vehicle body dimensions; The turning radius requirement is superimposed on the vehicle envelope space to construct the turning envelope space, wherein the turning envelope space and the vehicle envelope space constitute the vehicle traffic association envelope; The transport load status is loaded as a dynamic load parameter into the slope evaluation model; The vehicle envelope space, turning envelope space, and slope assessment model are used to dynamically traverse the traffic obstacle distribution data to perform a multi-parameter linked road network-level traffic capacity assessment and output the traffic risk heat map.

2. The method for dynamic evaluation of the traffic capacity of large transport vehicles based on traffic obstacle information as described in claim 1, characterized in that, Using the width-corrected road topology as the road geometric reference, obstacle parameter coupling identification is performed based on the stereo satellite imagery, outputting height-restricted obstacle distribution data and slope distribution data, including: Based on the aforementioned surface road network elevation model and road network orthophoto map, height restriction obstacles are identified and verified, and the distribution data of the height restriction obstacles are located. Based on the surface road network elevation model and width-corrected road topology, road slope variation characteristics are identified, and the slope distribution data is output.

3. The method for dynamic evaluation of the traffic capacity of large transport vehicles based on traffic obstacle information as described in claim 2, characterized in that, Based on the aforementioned surface road network elevation model and road network orthophoto map, height restriction obstacles are identified and verified, and the distribution data of the height restriction obstacles are located, including: Using the width-corrected road topology as a mask, road area elevation data is obtained by cropping from the surface road network elevation model; Gradient calculation is performed on the elevation data of the road area to obtain elevation gradient distribution data; The elevation gradient distribution data is traversed using a preset abrupt elevation threshold to locate distributed elevation abrupt points. The distributed optical features of the distributed elevation change points are located and extracted from the orthophoto map of the road network. Based on the distributed optical characteristics, a permanent structural verification of the distributed elevation abrupt points is performed to obtain distributed screened abrupt points. The clearance height of the distributed screening mutation points is calculated, and the distribution data of the height-restricted obstacles is output.

4. The method for dynamic evaluation of the traffic capacity of large transport vehicles based on traffic obstacle information as described in claim 3, characterized in that, The clearance height of the distributed screening mutation points is calculated, and the distribution data of the height-restricted obstacles is output, including: Based on the width-corrected road topology, the first road surface coordinates of the first selected abrupt change point are located, and the elevation value of the first road surface coordinates is used as the first road surface elevation value. Extract the bottom elevation of the first obstacle at the first screened mutation point from the surface road network elevation model; The difference between the bottom elevation of the first obstacle and the elevation of the first road surface is calculated to obtain the first single-point clearance height; By analogy, obstacle clustering is performed at the first screening mutation point based on spatial distribution density and optical feature consistency to obtain the first set of clearance heights; The first minimum clearance height is retrieved as the first clearance height feature based on the descending sorting result of the first clearance height set. The first clearance height feature and the first road surface coordinates are used as the first height-restricted obstacle feature; Similarly, the clearance height of the distributed screening mutation points is calculated, and the distribution data of the height-restricted obstacles is output.

5. The method for dynamic evaluation of the traffic capacity of large transport vehicles based on traffic obstacle information as described in claim 1, characterized in that, Based on the road vector surface topology relationship of the width-corrected road topology, the turning radius is dynamically calculated, and the minimum turning radius distribution data is output, including: Extract multiple road segment angles and multiple effective passage widths from the intersection nodes of multiple road vector planes in the width-corrected road topology; At the intersection of the multiple road vector planes, the vehicle turning direction is constrained by the connection direction of multiple road segments. The vehicle turning trajectory envelope space is dynamically simulated based on the included angle of the multiple road segments and the multiple effective passage widths to locate the boundaries of multiple sweeping areas. Using the intersection nodes of the multiple road vector surfaces as centers, inversion calculations are performed based on the boundaries of the multiple swept areas to obtain multiple minimum turning radius values; Bind the multiple minimum turning radius values, multiple road vector surface intersection nodes, and multiple road segment connection directions to output the minimum turning radius distribution data.

6. The method for dynamic evaluation of the traffic capacity of large transport vehicles based on traffic obstacle information as described in claim 1, characterized in that, The vehicle envelope space, turning envelope space, and slope assessment model are used to dynamically traverse the traffic obstacle distribution data to perform a multi-parameter linked road network-level traffic capacity assessment, outputting the traffic risk heatmap, including: The road centerline network is rasterized into an evaluation unit array at preset distance intervals, and the traffic obstacle distribution data is decomposed into a traffic obstacle unit array based on the evaluation unit array; After loading the first obstacle unit, the vehicle envelope space is dynamically driven to verify the height restriction passage capability, the turning envelope space is driven to verify the turning passage capability, the slope evaluation model is driven to verify the climbing load capability, and the first height restriction passage feasibility, the first turning passage feasibility, and the first slope load matching feasibility are output. By integrating the feasibility of passing through the first height restriction, the feasibility of passing through the first turn, and the feasibility of matching the first slope with the load, a first passage risk value is generated; By analogy, the vehicle envelope space, turning envelope space and slope evaluation model are used to dynamically traverse the traffic obstacle unit array to perform multi-parameter linkage road network-level traffic capacity assessment and generate a traffic risk heat map array. The passage risk heat map is output by periodically and dynamically rendering and restoring the passage risk heat map.

7. The method for dynamic evaluation of the traffic capacity of large transport vehicles based on traffic obstacle information as described in claim 1, characterized in that, Also includes: Multiple obstacle distribution features of multiple candidate paths are extracted by projecting the candidate road network corridors onto the traffic risk heat map; A candidate path access score matrix is ​​constructed based on the distribution characteristics of the multiple access obstacles. Based on the candidate path accessibility score matrix, a reliable access road to the target is selected from multiple candidate paths.

Images