A method for visualizing spatial mapping of outdoor intrusion risk

Abstract

The application discloses a kind of outdoor invasion risk spatial atlas visualization generation method, it is related to spatial data processing technical field, comprising: the spatial data of physical channel is obtained and is constructed as propagation network graph model, wherein node is associated with local risk attribute, edge is associated with propagation flux attribute and control traffic coefficient;On propagation network graph model, preset diffusion equation iteration solution is executed, and the cascade propagation risk value of each node is obtained;Control partition is demarcated based on cascade propagation risk value, and control traffic coefficient is updated;The updated control traffic coefficient is substituted back, and iteration solution and control partition demarcation are re-executed, until topological stability condition is reached, and risk spatial atlas is rendered and output;Its beneficial effect is: it can be realized in the invasion risk of blocking while avoiding excessive control to cultural travel activities.

Description

Technical Field

[0001] This invention relates to the field of spatial data processing technology, and in particular to a method for generating spatial maps of outdoor intrusion risks. Background Technology

[0002] In areas with complex terrain and natural geographical barriers, the spatial transfer of specific risks such as invasive alien species and the spread of plant diseases along agricultural routes is often difficult to spread over a large scale through natural media. Instead, it is highly dependent on the routes of human activities (such as specific outdoor routes and transportation connections), thus forming an artificial transmission network with strong path dependence.

[0003] For the above scenarios, existing spatial risk assessment technologies typically employ a grid model based on continuous space combined with a Geographic Information System (GIS). By calculating the environmental variable characteristics of each independent spatial grid, a global static risk distribution snapshot is generated. However, this assessment paradigm has the following limitations when facing artificially propagated networks with strong path dependence: Due to the unidirectional and static nature of the assessment process, when managers formulate structural control measures such as physical blocking or flow restriction for specific road sections, changes in the connectivity of physical paths inevitably lead to a redistribution of risks in the channel network. The unidirectional assessment process can only output initial risk hotspots without intervention and cannot spontaneously generate cascading responses caused by path blocking or changes in the original assessment process. As a result, the final risk distribution snapshot cannot reflect the true state of risk re-equilibrium in space after the implementation of control interventions, thus restricting the scientific formulation and optimization of control plans. Summary of the Invention

[0004] In view of the above-mentioned prior art, this application is hereby proposed. Embodiments of this application provide a method for generating a spatial map visualization of outdoor intrusion risks, which can overcome the shortcomings of traditional spatial risk assessment techniques in expressing path-dependent propagation, thereby achieving the goal of blocking intrusion risks while avoiding excessive control over cultural, sports, and tourism activities.

[0005] According to one aspect of this application, a method for generating a spatial map visualization of outdoor intrusion risk is provided, comprising: acquiring spatial data of physical channels within a target area, the physical channels including routes for outdoor activities and associated transportation connections; discretizing the spatial data and constructing a propagation network graph model; wherein each node in the propagation network graph model is associated with a local risk attribute characterizing the vulnerability of the environment to intrusion, each edge is associated with a propagation flux attribute characterizing the transfer capability of the physical channel, and a control passage coefficient characterizing the connectivity state of the physical channel, the initial value of the control passage coefficient characterizing that the corresponding physical channel is in a free passage state without intervention; using the local risk attribute of each node as the initial calculation state, using the propagation flux attribute associated with each edge and the current control passage coefficient as topology transfer parameters, in the... The propagation network graph model is used to iteratively solve a preset diffusion equation to obtain the cascade propagation risk value of each node. Based on the cascade propagation risk value, the target area is divided into control zones with different control levels. The control access coefficient associated with the edges contained in the control zone is updated according to the control level to characterize the actual connectivity changes after implementing flow restriction or blocking intervention on the corresponding physical channel. The updated control access coefficient is used as a dynamic boundary condition and substituted back into the preset diffusion equation to re-execute the iterative solution and control zone delineation until the boundary of the control zone generated in adjacent rounds reaches the preset topological stability condition. A risk space map is rendered and output, which includes the cascade propagation risk value of each node and the boundary of the control zone that has reached the topological stability condition.

[0006] According to another aspect of this application, a spatial map visualization generation system for outdoor intrusion risk is provided, comprising: a data acquisition module for acquiring spatial data of physical channels within a target area, wherein the physical channels include routes for outdoor activities and associated transportation connection channels; a network modeling module for discretizing the spatial data and constructing a propagation network graph model; wherein each node in the propagation network graph model is associated with a local risk attribute characterizing the vulnerability of the environment to intrusion, each edge is associated with a propagation flux attribute characterizing the transfer capability of the physical channel, and a control passage coefficient characterizing the connectivity state of the physical channel, wherein the initial value of the control passage coefficient characterizes the corresponding physical channel as being in a free passage state without intervention; and a risk solving module for using the local risk attribute of each node as the initial calculation state, and using the propagation flux attribute associated with each edge and the current control passage coefficient as topology transfer parameters, in the... The propagation network graph model is used to iteratively solve a preset diffusion equation to obtain the cascade propagation risk value of each node. A partitioning module is used to partition the target area into control partitions with different control levels based on the cascade propagation risk values, and to update the control access coefficients associated with the edges within the control partitions according to the control levels, to characterize the actual connectivity changes after implementing flow restriction or blocking interventions on the corresponding physical channels. An iterative optimization module is used to use the updated control access coefficients as dynamic boundary conditions, substitute them back into the preset diffusion equation, and re-execute the iterative solution and control partition delineation until the boundaries of the control partitions generated in adjacent rounds reach a preset topological stability condition. A graph output module is used to render and output a risk space graph, which includes the cascade propagation risk values ​​of each node and the boundaries of the control partitions that have reached the topological stability condition.

[0007] According to another aspect of this application, an electronic device is provided, including a memory and a processor, the memory being used to store computer-executable instructions, and the processor being used to execute the computer-executable instructions, which, when executed by the processor, implement the steps of the method described above.

[0008] According to another aspect of this application, a computer-readable storage medium is provided that stores computer-executable instructions thereon, which, when executed by a processor, implement the steps of the method described above.

[0009] Compared with the prior art, the spatial map visualization generation method for outdoor intrusion risk according to the embodiments of this application quantifies the cascading propagation effect across geographical barriers along physical channels by constructing a propagation network graph model, and achieves adaptive shrinkage of control zone boundaries by using dynamic back substitution of control passage coefficients, thereby blocking intrusion risks while avoiding excessive control over cultural, sports and tourism activities. Attached Figure Description

[0010] The above and other objects, features, and advantages of this application will become more apparent from the more detailed description of the embodiments of this application in conjunction with the accompanying drawings. The drawings are provided to further illustrate the embodiments of this application and form part of the specification. They are used together with the embodiments of this application to explain this application and do not constitute a limitation thereof. In the drawings, the same reference numerals generally represent the same components or steps.

[0011] Figure 1 This is a flowchart of a method for generating a spatial map visualization of outdoor intrusion risk according to the present invention.

[0012] Figure 2 This is a block diagram of a spatial map visualization generation system for outdoor intrusion risk according to the present invention.

[0013] Figure 3 This is a schematic diagram of the structure of an electronic device according to the present invention. Detailed Implementation

[0014] Hereinafter, exemplary embodiments according to this application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments of this application. It should be understood that this application is not limited to the exemplary embodiments described herein.

[0015] Exemplary method:

[0016] Figure 1 The illustration shows a method for generating a spatial map visualization of outdoor intrusion risk according to an embodiment of this application, including steps S1 to S6.

[0017] To facilitate understanding, we will use the Tiger Leaping Gorge-Haba Snow Mountain area in the Greater Shangri-La region as an example to illustrate the invasive risk management scenario. This area is a typical region where high biodiversity and high density of outdoor cultural and sports tourism activities overlap. The terrain rises sharply from the Jinsha River valley (altitude about 1800 meters) to the Haba Snow Mountain alpine meadow (altitude about 4200 meters), with a vertical difference of over 2000 meters. It traverses multiple vegetation zones, including subtropical dry-hot valleys, montane broad-leaved forests, mixed coniferous and broad-leaved forests, alpine shrubs, and alpine meadows. The main invasive species in the area is *Ageratina adenophora*, a perennial herbaceous plant of the Asteraceae family native to Central America. It has now spread widely and formed a dominant community in areas below 2500 meters above sea level in the Jinsha River valley, posing a serious threat to native vegetation. The seeds of *Ageratum adenophorum* are small, numerous, and have pappus. Under natural conditions, they are mainly dispersed by wind and water. However, the complex terrain of the region severely restricts their spread—deep canyons, steep cliffs, and high mountain passes form natural geographical barriers, making it difficult for them to spread widely to high-altitude areas through natural means. Nevertheless, the region's intensive outdoor cultural and recreational activities provide artificial dispersal channels for *Ageratum adenophorum* to cross these terrain barriers: seeds carried by hikers' shoes, seed-bearing soil attached to off-road vehicle tires, and propagule fragments carried by camping equipment can transport *Ageratum adenophorum* seeds from low-altitude valleys to alpine meadows above 3000 meters—elevations that would have been inaccessible to natural dispersal for decades—within hours. These alpine meadows, precisely because of their high soil exposure, weak competition from native vegetation, and proximity to water sources, pose a high risk of invasive colonization.

[0018] Assume the following outdoor cultural and tourism infrastructure exists in the area: Tiger Leaping Gorge High-Altitude Hiking Route (hereinafter referred to as Route A), starting from Qiaotou Town, passing through Naxi Yage and the intermediate inn to Haba Village, with a total length of approximately 22 kilometers, which is the most popular multi-day hiking route in the area; Haba Snow Mountain Climbing Route (hereinafter referred to as Route B), starting from Haba Village to Haba Snow Mountain Base Camp, with a total length of approximately 12 kilometers; Jinsha River Valley Cycling Route (hereinafter referred to as Route C), starting from Qiaotou Town along the Jinsha River Valley to Walnut Garden and then to Daju Township, with a total length of approximately 18 kilometers. The Qiaotou Town Visitor Center Parking Lot (hereinafter referred to as Shuttle Station T1) is the common starting point for both Route A and Route C, and Haba Village (hereinafter referred to as Shuttle Station T2) is the connecting point between Route A and Route B. The intermediate inn has a camping area (hereinafter referred to as Campsite C1), and the Haba Snow Mountain Base Camp has a camping area (hereinafter referred to as Campsite C2).

[0019] It should be noted that the above-mentioned Tiger Leaping Gorge-Haba Snow Mountain area scenario is merely a specific example selected for ease of understanding the technical solution of this application. The method of this application is applicable to any area possessing the following combination of characteristics: complex terrain, the presence of natural geographical barriers that prevent specific risks from spreading over a large scale through natural media, and human outdoor activity routes forming physical channels across these geographical barriers. In addition to the spread of invasive species, this method is also applicable to scenarios with similar path-dependent transmission characteristics, such as the spread of plant diseases along agricultural routes and the spread of forest fire risks along firebreaks. The specific species names, altitude values, pedestrian traffic data, and parameter values ​​mentioned below are all exemplary and do not constitute a limitation on the scope of protection of this application.

[0020] The following steps describe how to generate an intrusion risk spatial map based on the above scenario.

[0021] like Figure 1 As shown, in step S1, spatial data of physical channels within the target area are obtained. The physical channels include routes for outdoor activities and associated transportation connections.

[0022] In the aforementioned example, the physical corridor spatial data to be acquired includes: GPS trajectory data for routes A, B, and C (recording the complete direction and mileage information of the routes in geospatial space), coordinate data for shuttle stations T1 and T2, and coordinate data for campsites C1 and C2. In addition, geospatial base map data is also required, including a DEM elevation model covering the area (used to extract elevation and slope at each location), water system distribution vector data (used to calculate the distance to water sources at each location), land cover classification raster data (used to determine the land cover type at each location), and suitable habitat probability raster data for the target invasive species *Ageratum adenophorum* (generated based on suitable habitat prediction models such as MaxEnt, recording the suitable habitat probability of *Ageratum adenophorum* in each raster cell within the area in different seasons).

[0023] like Figure 1 As shown, in step S2, the spatial data is discretized and constructed into a propagation network graph model; wherein, each node in the propagation network graph model is associated with a local risk attribute that characterizes the vulnerability of the environment to intrusion, each edge is associated with a propagation flux attribute that characterizes the physical channel transfer capability, and a control passage coefficient that characterizes the physical channel connectivity state. The initial value of the control passage coefficient indicates that the corresponding physical channel is in a free passage state without intervention.

[0024] In this step, the physical channel spatial data obtained in step S1 is abstracted into a directed weighted graph structure—a propagation network graph model. The reason for using a graph structure instead of the commonly used continuous spatial grid model is that in areas with extremely fragmented terrain, such as the Greater Shangri-La region, the human-induced spread of invasive species is strictly limited to accessible physical pathways—hikers will not deviate from the route to cross cliffs, vehicles can only travel along roads, and cyclists can only proceed along predetermined cycling routes. The carrying and release of invasive reproductive organisms occurs along these physical pathways, rather than spreading uniformly in continuous space. The grid model assumes that risk is uniformly transmitted between adjacent grids, failing to express crucial geographical facts such as "two hillsides only 2 kilometers apart are completely inaccessible in a propagation sense because they are separated by a canyon." The graph structure, on the other hand, uses physical pathways as its framework, and the connections between nodes and edges faithfully reflect the topology of actual accessible paths, ensuring that risk propagation calculations are strictly performed along accessible pathways.

[0025] The process of constructing the propagation network graph model is as follows:

[0026] Node set Generation: Routes A, B, and C are discretized and sampled at 500-meter intervals, generating a series of path nodes on each route. Shuttle stations T1 and T2, and campsites C1 and C2 are directly added to the node set as functional nodes. In the aforementioned example, route A (22 km) generates approximately 44 path nodes, route B (12 km) generates approximately 24 path nodes, and route C (18 km) generates approximately 36 path nodes. Adding the 4 functional nodes, a total of approximately 108 nodes are generated. For ease of explanation in the numerical calculations below, seven representative nodes are selected for illustration—nodes For Qiaotou Town parking lot (shuttle station T1, altitude 1800 meters), node The node is Naxi Yage (altitude 2300 meters) on Route A. This is a midway inn (campsite C1, altitude 2600 meters), a node. The endpoint of Route A is Haba Village (connecting station T2, altitude 2700 meters), node. The endpoint of Route B is Haba Base Camp (campsite C2, altitude 4100 meters). The node is the walnut orchard (elevation 1900 meters) on route C. The endpoint of Route C is Daju Township (elevation 1800 meters).

[0027] Edge set Generation: Directed edges are established between adjacent discrete nodes on the same route, with the direction being the designed travel direction (bidirectional edges are established for bidirectional routes). Cross-route connecting edges are established at the junctions where routes intersect—nodes. (Connecting Station T1) simultaneously connects to the first segment of Route A ( ) and the first edge of route C ( ),node (Connecting Station T2) simultaneously connects the last edge of Route A and the first edge of Route B. Establish a bidirectional edge between the campsite node and the nearest path node on its route—campsite C1 ( Connect with the adjacent node on route A, campsite C2 ( It connects to the end point of route B.

[0028] The local risk attributes that each node needs to be associated with and the propagation flux attributes that each edge needs to be associated with are determined in the following ways.

[0029] In a preferred embodiment, the local risk attribute is determined as follows:

[0030] Acquire environmental variable data of the node's location and the suitability probability data of the target species. The environmental variable data includes land cover type, terrain slope, and distance to water source.

[0031] Based on land cover type, topographic slope, and distance to water source, the corresponding environmental factor sub-values ​​were calculated; among them, the environmental factor sub-values ​​represent the degree to which environmental variables promote the colonization and spread of the target species.

[0032] The environmental vulnerability score is synthesized by weighted summation of the component values ​​of each environmental factor. The weight coefficients of each component value in the weighted summation are determined by parameter search with known risk distribution record points as positive samples and random background points as negative samples, with the goal of maximizing classification accuracy.

[0033] The local risk attribute is obtained by multiplying the environmental vulnerability score by the sampled value of the suitability probability data at that node.

[0034] It should be noted that the selection of the above three environmental variables is based on known laws of invasion ecology. Land cover type determines the competitive relationship between invasive species and native vegetation—bare land and degraded grassland have weak native vegetation competition, making them the easiest for invasive species to establish themselves; forest edges have moderate competition; and dense forests and water bodies have the lowest risk of invasion. Topographic slope affects the probability of invasive species propagules (seeds, rhizome fragments) remaining on the ground—propagules are easily retained and germinate on low-slope surfaces, while on steep slopes, propagules are easily washed away or rolled down by rainwater, making establishment difficult. Distance from water sources affects the likelihood of waterborne propagation—in areas near streams and lakes, propagules can be secondary dispersed to a wider area via water flow.

[0035] Describing the above process using mathematical expressions, environmental vulnerability scoring... The calculation formula is:

[0036] ;

[0037] in, For nodes Land cover type of the location For terrain slope, Distance to water source; , , These are mapping functions for the environmental factor component values ​​corresponding to the three environmental variables; , , These are the corresponding weighting coefficients. The local risk attribute is... ,in For the target species at this node location, quarter The sampled value of the fitness probability.

[0038] In the aforementioned example, using nodes (Qiaotou Town parking lot) as an example: the land cover type at this location is bare land and degraded grassland. Take the higher value), the slope is about 5 degrees (low slope, (Take the higher value), approximately 200 meters from the Jinsha River (near the water source). (Take the highest value). The environmental vulnerability score is obtained by weighted summation of the three component values. The probability of *Eupatorium adenophorum* thunbergii thunbergii* thriving in summer at this location (river valley at an altitude of 1800 meters). Therefore, the node The local risk attribute is This reflects that even without considering human-caused transmission, the location already has a high risk of intrusion based solely on the local environmental conditions.

[0039] Then by node (Haba Base Camp) as an example: the land cover type at this location is alpine meadow and scree bare ground. (Take a medium to high value), slope approximately 18 degrees (medium slope), (Take the median value), approximately 2000 meters from the nearest stream (distant water source). (Take the lower value). Environmental vulnerability score The probability of *Eupatorium adenophorum* thunbergii* thriving in summer at this location (4100 meters above sea level). —The current altitude far exceeds its known distribution upper limit, but the suitable habitat prediction model shows that if the reproductive organisms are carried here, there is still a certain probability of colonization. Node The local risk attribute is This value is extremely low, indicating that the location is unlikely to be invaded by natural conditions alone.

[0040] The local risk attributes of all representative nodes were calculated using the method described above, and the results are as follows: Nodes (Qiaotou Town) is 0.65, node (Naxi Yage) is 0.25, node (Midway Inn) is 0.20, node (Haba Village) has a value of 0.15, node (Haba Base Camp) is 0.03, node (Walnut Orchard) is 0.58, node (Daju Township) is 0.55. It can be observed that low-altitude river valley nodes ( , , The local risk attributes of high-altitude nodes are generally high, while those at high altitudes are relatively high. , The local risk attribute is very low, which is consistent with the actual pattern that Agastache rugosa is currently concentrated in the area below 2,500 meters above sea level.

[0041] In a preferred embodiment, the propagation flux attribute is determined as follows:

[0042] Obtain the traffic flow data of the physical channels corresponding to each side, as well as the activity carrying coefficient of each physical channel; wherein, the activity carrying coefficient is determined based on the measured carrying amount of the target species' reproductive bodies in the surface attachment samples of outdoor activity carriers passing through the physical channels.

[0043] Multiply the traffic flow data by the activity carrying coefficient to obtain the propagation flux attribute of each side association.

[0044] It should be noted that the activity carrying coefficient was determined based on measured data. The specific method was as follows: at the start and end sampling stations of each type of activity route, soil samples were randomly collected from the soles of hikers' shoes, the tires of cyclists, and the bottoms of campers' tents. The number of *Ageratum adenophorum* seeds and rhizome fragments in the samples were tested in a laboratory. The activity carrying coefficient was calculated as the ratio of the average carrying amount per person per activity type to the type with the highest carrying amount. In the aforementioned examples, the measured results show that camping activities (including tent bottoms, sleeping mats, and other large-area ground contact equipment) carried the highest amount of equipment, with a normalization coefficient of 1.0; hiking activities carried approximately 60% of the amount carried by the soles of shoes, with a coefficient of 0.6; and mountain biking carried approximately 30% of the amount carried by tires, with a coefficient of 0.3. These values ​​reflect the differences in the contact area and duration of different activity types on the ground—camping activities involve continuous ground contact at the same location for several hours, resulting in a significantly higher carrying amount than the brief contact of shoe soles during walking.

[0045] In the aforementioned example, Route A (Tiger Leaping Gorge High Route) sees approximately 800 visitors during the summer season, Route B (Haba Climbing Route) approximately 150 visitors, and Route C (Cycling Route) approximately 300 visitors. (node To node Taking the first segment of route A as an example: traffic flow The activity is for 800 participants per quarter, the activity type is hiking, and the activity carrying capacity is [missing information]. The propagation flux attribute is Effective reach per quarter. (Based on...) (node To node For example, route B): traffic flow The target is 150 participants per quarter. Campsite C2 is located at the end of Route B. The activity types include hiking and camping. The activity carrying capacity is calculated as a weighted average. The propagation flux attribute is Effective reach per quarter.

[0046] Control and passage coefficients of each side The initial value is set to 1.0, representing that all physical channels are in a state of uninterrupted free passage in the initial state—no routes are closed, no road sections are restricted, and all channels can pass freely. The control passage coefficient will be updated in subsequent step S4 based on the delineation results of the control zones. This update mechanism will be explained in detail in step S5.

[0047] Furthermore, a distance decay function is needed in the calculation of the diffusion equation to characterize the decrease in the amount of invasive progeny carried as a function of distance due to shedding and inactivation during transport. This function takes the form of a negative exponential function:

[0048] ;

[0049] in This represents the actual path distance of the edge. This refers to the attenuation coefficient. The value of is determined by fitting measured data on the survival rate decline of invasive propagules at different transport distances. In the aforementioned example, by collecting the number of residual seeds on the soles of hikers' shoes at different mileage points along route A, it was found that the amount of *Ageratum adenophorum* seeds remaining on the soles of hikers decreased approximately negatively exponentially with walking distance, and the value was obtained by fitting. / km, meaning that after walking 1km, the seed residue is about 86% of the initial value, after walking 5km it is about 47%, and after walking 10km it is about 22%.

[0050] return Figure 1 In step S3, the local risk attributes of each node are used as the initial calculation state. The propagation flux attributes associated with each edge and the current control passage coefficient are used as topology transmission parameters. The preset diffusion equation is iteratively solved on the propagation network graph model to obtain the cascade propagation risk value of each node.

[0051] In this step, the cascading propagation risk value of each node is calculated on the propagation network graph model. Unlike existing techniques that calculate the risk value of each cell independently on the grid cell, in this method, the cascading propagation risk value of each node depends not only on the node's own local risk attribute but also on the topological transmission of the cascading propagation risk values ​​of its upstream neighboring nodes in the propagation network graph model via edges. This means that a high-altitude node with a very low local risk attribute will have its cascading propagation risk value increase due to the influx of upstream risk if there is a high-risk low-altitude node upstream of it and the two are connected by a high-throughput hiking route—this is precisely the mathematical expression of "human activity paths as intrusion propagation channels."

[0052] The specific form of the diffusion equation is:

[0053] ;

[0054] in, For the first Nodes in the next iteration In the quarter The risk value of cascading propagation below; This is the local risk attribute of the node, which is the product of the survival probability and the environmental vulnerability score. The propagation coupling coefficient controls the influence weight of the network propagation term relative to the local term. Its initial value is determined by parameter optimization using known intrusion diffusion records in the study area as labeled data, with the goal of maximizing the spatial fit between the cascade propagation risk value distribution predicted by the model and the actual intrusion distribution. For nodes The set of upstream adjacent nodes in a propagation network graph model; For the edge The propagation flux attribute; It is the distance decay function; For the edge The current traffic control coefficient.

[0055] The iterative solution process is as follows:

[0056] During initialization, the cascading propagation risk value of all nodes is set equal to their local risk attribute, i.e. At this point, we are not considering online dissemination.

[0057] Then, iterations are performed round by round: In each round, the cascading propagation risk value of all nodes is updated synchronously according to the above equation—the new value of each node equals its local risk plus the sum of the risk contributions from all upstream neighboring nodes via edges. Iteration continues until the maximum change in the cascading propagation risk value of all nodes in two adjacent iterations is less than the convergence threshold. (Values ​​such as 0.001) indicate that the iteration has converged.

[0058] The physical meaning of this iterative process is as follows: the invasive progeny spreads downstream from the high-suitability area (low-altitude river valley) along the cultural and tourism route network, hopping from one hop to the next. The propagation amount of each hop is jointly regulated by the flow of people (how many people pass through), the activity carrying coefficient (how many progeny each person carries), distance attenuation (how many are shed along the way), and the control and passage coefficient (whether the passage is restricted or blocked). After multiple rounds of propagation, the cascading propagation risk value of each node no longer changes, and a steady-state propagation pressure distribution is formed on the network.

[0059] In the aforementioned example, taking summer (the season when the peak ripening period of *Aeonium adenophorum* seeds overlaps with the peak season for hiking tourism) as an example, we take... All traffic control coefficients (Initially without intervention), after multiple rounds of iterative convergence, the cascading propagation risk values ​​of each representative node are as follows:

[0060] node (Qiaotou Town Parking Lot): The cascading propagation risk value is 0.65, which is the same as the local risk attribute, because this node is located at the starting point of the route and there is no upstream node to propagate risk to it.

[0061] node (Naxi Yage): The cascading propagation risk value is 0.37, which is higher than its local risk attribute of 0.25. Of this, 0.25 is the local risk contribution, and the remaining 0.12 comes from upstream nodes. Jingbian The risk of intrusion – the high risk of intrusion into Qiaotou Town is spread along Route A towards Naxi Yage through the summer passenger flow of 800 people on the Tiger Leaping Gorge Expressway.

[0062] node (Midway Inn): The cascading propagation risk value is 0.25, higher than its local risk attribute of 0.20. Upstream node The risk contribution from the incoming data is 0.05. Although Because of itself The introduction of this virus already carries a high risk of cascading propagation, but this risk continues to increase. The propagation is reduced by distance attenuation, therefore the actual amount of data entering is limited. This demonstrates the distance attenuation function. Its role in controlling the range of cascading propagation.

[0063] node (Haba Village): The cascading propagation risk value is 0.18, higher than its local risk attribute of 0.15. The incoming risk is relatively low because this node is close to the high-risk source ( The distance has exceeded 15 kilometers, and the distance decay means that the risk contribution transmitted through multiple hops is reduced upon arrival. The time has already decreased significantly.

[0064] node (Haba Base Camp): The cascade propagation risk value is 0.04, higher than its local risk attribute of 0.03. Although the absolute increase is small, this increase comes entirely from human propagation channels—Haba Base Camp is almost impossible to be invaded by Eupatorium adenophorum under natural conditions (local risk is only 0.03), and it is the propagules carried by 150 climbers passing through Haba Village that give this location an additional risk of invasion.

[0065] node (Walnut Orchard): The risk value of cascading transmission is 0.59, which is higher than its local risk attribute of 0.58. A small risk is introduced through cycling route C, but the risk is low due to the low activity carrying capacity of cycling ( The contribution from input is limited.

[0066] node (Daju Township): The risk value of cascading transmission is 0.56, which is close to its local risk attribute of 0.55.

[0067] Two key characteristics can be observed from the above results: First, the distribution of cascading propagation risk values ​​is no longer simply equivalent to the distribution of local risk attributes—nodes Because it received the upstream The risk of large-scale inflow is significantly higher than the risk of local spread (0.25) due to its cascading propagation risk value (0.37). If only local risk is considered, It will not be identified as a node requiring special attention, but considering network propagation, In reality, it is a key relay node in the cascading propagation chain of risk from low-altitude valleys to high-altitude areas. Second, the intensity of the cascading propagation effect decreases along the path—the further downstream the node is from the high-risk source, the more hops it passes through, and the fewer people passing through, the less incoming risk it receives. This attenuation characteristic ensures the convergence of the equation iteration.

[0068] return Figure 1 In step S4, the target area is divided into control zones with different control levels based on the cascading propagation risk value, and the control access coefficient associated with the edges contained in the control zone is updated according to the control level to characterize the actual connectivity changes after implementing flow restriction or blocking intervention on the corresponding physical channels.

[0069] In a preferred embodiment, the delineation of the control zone specifically includes the following process:

[0070] A high-risk threshold is determined based on the cascading propagation risk value of each node, and nodes with cascading propagation risk values ​​greater than or equal to the high-risk threshold are extracted as a high-risk node set.

[0071] Using the shortest path distance on the graph accumulated along the edges between nodes in the propagation network graph model as the topological distance metric, spatial clustering is performed on the high-risk node set to obtain at least one risk cluster; where the topological distance metric characterizes the actual reachable path length of the physical channel;

[0072] For each risk cluster, generate an envelope boundary that surrounds each node within it to define the control zone;

[0073] Based on the statistical values ​​of the cascading risk values ​​of each node within the risk cluster, the control level of the corresponding control zone is determined.

[0074] It should be noted that, in the above process, the high-risk threshold... The threshold is determined using the Jenks Natural Breaks method. This method automatically divides the cascading propagation risk values ​​of all nodes in the entire domain into multiple natural levels by maximizing inter-class variance and minimizing intra-class variance, and takes the lower bound of the highest level as the high-risk threshold. This approach makes the determination of the threshold dependent on the distribution characteristics of the data itself, rather than a fixed value preset by humans.

[0075] The reason for using the shortest path distance on the graph as the topological distance metric, rather than Euclidean distance, when performing spatial clustering on a high-risk node set is due to the terrain characteristics of the Greater Shangri-La region. Taking the nodes in the aforementioned example... (Midway Inn, 2600 meters above sea level) and nodes Taking Walnut Garden (at an altitude of 1900 meters) as an example: the straight-line distance between the two locations may be only about 3 kilometers (belonging to the slopes on either side of Tiger Leaping Gorge), but they are completely separated by the deep canyon of Tiger Leaping Gorge, and there are no directly connected edges in the propagation network graph model—to get from... arrive You must return to Qiaotou Town via Route A. Then follow route C to the walnut orchard ( The shortest path distance on the graph exceeds 30 kilometers. If Euclidean distance clustering is used, and Because of their shortest straight-line distance, they might be grouped into the same cluster, causing the control zone to cross a practically impassable valley—making unified control within this zone neither feasible nor effective. However, by using the shortest path distance on the graph, and In a topological sense, these are distant nodes that will not be grouped into the same cluster, and the boundaries of the controlled partitions are faithful to the connectivity of the actual passable paths.

[0076] The control level for each risk cluster is determined by ranking the risk values ​​propagated through cascading nodes within the cluster from highest to lowest, resulting in three levels. The update rule for the control access coefficient is as follows: within the control partition of the highest control level, the control access coefficient of all edges... A value set to 0 indicates that the corresponding physical passage is completely blocked (corresponding to road closure or access restriction in actual operation); within the control zone of the second highest control level, the control passage coefficient for all edges. Set to the flow restriction ratio value set for this zone (e.g., 0.3, indicating that a flow restriction is implemented, allowing only 30% of the original traffic volume to pass); within the control zone of the third control level, the control passage coefficient of the edge. It remains at 1.0, but nodes within this partition are marked as priority monitoring targets.

[0077] In the aforementioned example, based on the classification of natural breakpoints, the high-risk threshold is assumed to be determined as follows: The high-risk node set then includes nodes. (Cascade propagation risk value 0.65) and nodes (Cascade propagation risk value: 0.37). These two nodes are directly connected on the graph (through an edge). The shortest path distance on the graph is approximately 6 kilometers, satisfying the proximity condition for clustering, and thus it is grouped into the same risk cluster. After generating the envelope boundary of this cluster, a control zone is formed. The cluster mean was approximately 0.51, which was designated as the highest control level. Accordingly, the control zone... inner edge The traffic control coefficient was updated from 1.0 to 0, indicating that Route A from Qiaotou Town to Naxi Yage was closed.

[0078] return Figure 1 In step S5, the updated control passage coefficient is used as a dynamic boundary condition and substituted back into the preset diffusion equation to re-execute the iterative solution and delineation of the control partition until the boundary of the control partition generated in the adjacent rounds reaches the preset topological stability condition.

[0079] This step is the core difference between this method and existing technologies. In existing technologies, the control partition output by step S4 is the final result—it doesn't concern itself with "what changes would occur in the risk distribution if control were implemented according to this partition." In this method, however, the control partition output by step S4 is not the final result, but an intermediate state. Its corresponding control passability coefficient will be substituted back into the diffusion equation of step S3 to drive a new round of iterative solution.

[0080] Its physical meaning is: when the section of Route A from Qiaotou Town to Naxi Yage is closed (control traffic coefficient) (From 1.0 to 0), from through Towards The channels through which the risk of transmission is spread have been physically blocked. This means... No longer receiving from Ingress risk contribution — The cascading propagation risk value will drop from 0.37 to close to its local risk attribute of 0.25, while the downstream nodes... , , The risk of cascading propagation will also be reduced accordingly.

[0081] In the aforementioned example, the specific execution process of step S5 is as follows:

[0082] First generation – Control zone inner edge Traffic control coefficient The value has been set to 0, and the diffusion equation iteration in step S3 is re-executed with the updated control passage coefficient. The changes in the cascading propagation risk value of each node after convergence are as follows: Node The value remains at 0.65 (as the source node, it is not affected by downstream control), but the node It dropped from 0.37 to 0.25 (lost) The inbound contribution, falling back to the local risk attribute), node From 0.25 to 0.22, node From 0.18 to 0.16, node It decreased from 0.04 to 0.03.

[0083] The key change lies in: nodes The cascading propagation risk value decreased from 0.37 to 0.25, below the high-risk threshold. The node no longer belongs to the high-risk node set. The control partitioning in step S4 is re-executed with the new cascading propagation risk value—at this point, the high-risk node set only contains nodes... (0.65), the risk cluster is reduced to only containing A single-node cluster. Controlled partition. The boundary from the original containing and The region shrinks to contain only The area.

[0084] Second cycle generation - partitioning After contraction, the edge No longer located within the controlled zone ( (No longer within the partition). But because Still within the highest control level partition, from The departure side still applies the control passage coefficient. —That is, the section of Route A starting from Qiaotou Town remains closed. The diffusion equation iteration is re-executed. After convergence, the cascading propagation risk value of each node is consistent with the result of the first round. The control zone delineation is re-executed; the high-risk node set remains the one containing only… Control zone The boundary no longer changes.

[0085] At this point, the boundaries and control levels of the control partitions generated in the two adjacent rounds (the first round and the second round) have not changed, satisfying the preset topological stability conditions, and the iteration converges.

[0086] The above process demonstrates the practical effect of back-substitution iteration of the control passage coefficient—the initially defined control zone. Include and Two nodes mean that the entire route A from Qiaotou Town to Naxi Yage needs to be closed. However, after back-substitution iteration, it was automatically discovered that: The reason it became a high-risk node was not due to its local environmental conditions, but entirely because... The risk is transmitted via route A. Once... The transmission route in that section of the road was blocked. The risk naturally subsides to a safe level, eliminating the need for closure. The location of the road segment. The final converged control zone only includes... This means that simply closing the entrance to the road section near the Qiaotou Town parking lot is sufficient to effectively block the spread of risk, while Route A further away from Naxi Yage can remain open. Compared to the initial plan, the controlled area has been reduced by approximately 6 kilometers, avoiding unnecessary control over the popular hiking section from Naxi Yage to the midway inn. This ensures effective control of the intrusion risk while preserving as much usable space as possible for cultural, sports, and tourism activities.

[0087] return Figure 1 In step S6, a risk space map is rendered and output. The risk space map includes the cascading propagation risk value of each node and the boundary of the control partition that has reached the topological stability condition.

[0088] In this step, the calculation results converged in step S5 are overlaid and rendered onto the geographic base map to form the final risk spatial map. The map contains the following information layers: the cascading propagation risk value of each node is represented by a circular marker, the marker radius of which is proportional to the cascading propagation risk value, and the color is mapped by a continuous color band from green (low risk) to yellow (medium risk) to red (high risk); each edge is represented by a line segment, the line width of which is proportional to the propagation flux attribute, and the color corresponds to the average cascading propagation risk value of the two end nodes; control zones are displayed as overlaid semi-transparent polygons, and different control levels are distinguished by different fill colors.

[0089] In the aforementioned example, the final output risk space map visually presents the following information: Qiaotou Town parking area ( The control zone marked as having the highest control level is covered by a dark red semi-transparent polygon; Naxi Yage ( ), Midway Inn ( Nodes located outside the controlled zone, such as those marked with yellow-green circles, are indicated to be in [the controlled zone]. With the road sections closed under the control plan, the risk of intrusion in these areas has been reduced to an acceptable level; each section of Route A is connected to each node with lines of different thicknesses and colors, which intuitively reflects the propagation flux and risk level of each path.

[0090] In actual management, the distribution of invasive species is not static, and on-site monitoring may uncover new invasive distribution points not recorded in the original data. When such monitoring data is fed back, the local risk attributes of the corresponding nodes in the propagation network graph model will change—for example, patrol personnel at a roadside inn ( The discovery of new *Ageratum adenophora* seedlings nearby indicates that the susceptibility probability of this node needs to be adjusted from the original raster value to reflect the actual invasion. To efficiently respond to such local updates without recalculating the entire graph, in a preferred implementation, when the local risk attribute of a node changes, the iterative solution employs an incremental calculation strategy, specifically including:

[0091] The nodes whose local risk attributes have changed and their neighboring nodes within a preset hop count range on the propagation network graph model constitute the affected node set;

[0092] Use the current cascade propagation risk value of the boundary node of the affected node set as a fixed boundary condition;

[0093] The iterative solution of the preset diffusion equation is re-executed only within the affected node set to obtain the updated cascade propagation risk value of each node within the affected node set;

[0094] Determine whether there are nodes in the affected node set whose cascading propagation risk value is greater than the high-risk threshold. If so, re-define the control zone and update the control passage coefficient locally in the area where the affected node set is located, and use the updated control passage coefficient as a dynamic boundary condition to substitute back into the preset diffusion equation.

[0095] In the example above, assume the patrol personnel are at the node New seedlings of *Eupatorium adenophorum* have been discovered near the (Midway Inn). Probability of survival The increase from 0.35 to 0.60 resulted in The local risk attribute increased from 0.20 to 0.45. Centered on the graph, expand outwards within a 3-hop radius, and label the affected node set as... ( and (Outside the 3-hop range or a boundary node). and The current cascading propagation risk value at the boundary is a fixed value, only when... The diffusion equation iteration is re-executed internally. This local recalculation involves only 3 nodes instead of the full 108 nodes, resulting in improved response speed. If the local recalculation... The cascading propagation risk value exceeds the high-risk threshold. Then in The surrounding areas will undergo localized re-delineation of control zones and iterative back-substitution of control and traffic coefficients.

[0096] The significance of incremental computation strategy lies in the fact that in real-world application scenarios where the propagation network graph model contains hundreds or even thousands of nodes, triggering a full graph recalculation with each monitoring data update would result in excessive computational latency, failing to meet the requirement of rapid response to on-site data updates. Incremental computation limits the recalculation scope to the affected local subgraph, making the computational load proportional to the size of the affected area rather than the size of the entire graph. Simultaneously, fixed boundary conditions ensure consistency between the local recalculation results and the overall graph state.

[0097] Furthermore, in actual management, cultural, sports, and tourism routes are not fixed—management departments may plan new hiking routes, tourists may trample and create informal trails, and event organizers may propose new trail running routes. Once these new physical pathways are incorporated, they will alter the topological connectivity of the propagation network model, potentially providing previously isolated, highly fertile areas with propagation paths to densely populated areas, leading to a structural change in the intrusion risk pattern. To assess the impact of these new physical pathways on the risk pattern, in a preferred embodiment, the method further includes the following steps:

[0098] Receive spatial data of newly added physical channels and their corresponding traffic flow data and activity carrying coefficient;

[0099] The spatial data of the newly added physical channels are discretized at preset intervals to generate new node sequences and new edges connecting adjacent new nodes. Environmental variable data are extracted from each node in the new node sequence to determine its local risk attributes. The propagation flux attributes of each new edge are determined based on traffic flow data and activity carrying coefficient. The control passage coefficient associated with each new edge is initially set to a value that represents the free passage state without intervention.

[0100] Insert the new node sequence and newly added edges into the propagation network graph model to identify the spatial proximity relationship between the new physical channels and existing physical channels at the traffic connection points and establish cross-route connection edges.

[0101] Using the propagation network graph model after inserting the new edge as the basis, the iterative solution of the preset diffusion equation and the delineation of the control partition are re-executed until the boundary of the control partition generated in the adjacent rounds reaches the preset topological stability condition.

[0102] The risk space map is rendered and output using a difference overlay view. The difference overlay view includes the changes in the cascading propagation risk values ​​of each node before and after the addition of a new physical channel, as well as the changes in the control zone boundaries.

[0103] The following examples illustrate the cascading impact of new routes on the risk landscape.

[0104] Suppose an event organizer proposes to open a route from Qiaotou Town ( Direct access to Haba Village ( The trail running shortcut (hereinafter referred to as Route D) is approximately 10 kilometers long. This route traverses a previously untouched alpine meadow area, and is expected to be used by about 600 people per season in the summer. The activity type is trail running (activity carrying capacity). ).

[0105] After receiving the GPS trajectory data for route D, approximately 20 new nodes and corresponding new edges are generated by discretizing the data at 500-meter intervals. The propagation flux attribute of the new edges is... Effective reach per quarter. The origin and connecting station T1 of route D were identified. The space is adjacent to the terminal and the connecting station T2. (Spatial proximity, establish cross-route connection edges)

[0106] After inserting route D into the propagation network graph model, the diffusion equation is solved iteratively again. The key change is: before inserting route D, The high risk of intrusion needs to spread to It is necessary to follow route A. , The data was passed hop-by-hop, covering a distance of over 22 kilometers, and was transmitted after multiple distance attenuations. The risk contribution is already very small. Route D provides a path from... arrive The direct route is only 10 kilometers long, with minimal distance attenuation. Much greater than the cumulative decay along the multi-hop path of route A. After re-iteration and convergence, The risk value of cascading propagation increased significantly from 0.18 to 0.33—exceeding the high-risk threshold. .

[0107] More importantly, The risk of increased risk will spread along route B to downstream nodes. (Haba Base Camp) Cascaded Propagation— The cascading transmission risk value increased from 0.04 to 0.06. Despite The absolute risk remains low, but this increase stems entirely from the indirect effects of Route D—Route D itself does not pass through Habakkuk's base camp, but it shortens the route. arrive The distance of transmission indirectly exacerbated the spread to even more distant areas. The risk of intrusion. This cascading effect across routes cannot be captured at all in the traditional independent grid evaluation mode—in the grid model, adding a new route only affects the grid cells along the route, and has no effect on grid cells not on the route.

[0108] Present the changes before and after route D is accessed using a difference overlay view: The area changed from green to orange-red, and a new control zone was added that includes... The risk area, while the original The control zones remain unchanged. Event organizers can visually see the impact of this shortcut route on the overall risk profile, and assess whether route design adjustments or additional control measures are needed. If event organizers adjust the projected number of passengers for route D on the visualization interface (e.g., reducing it from 600 to 200), the diffusion equation iteration and control zone iteration will be re-executed immediately with the new traffic flow data, allowing them to see the impact of changes in passenger flow on the risk profile in real time.

[0109] After a period of operation, the model's prediction accuracy may deviate due to the actual complexity of environmental conditions—for example, the model may overestimate the impact of the slope factor and underestimate the impact of the water source distance factor, leading to an underestimation of the intrusion risk in gentle slope areas near streams. To enable the model to continuously improve its accuracy during operation, in a preferred embodiment, the method further includes the following steps:

[0110] Obtain measured score data for the risk status of some nodes in the propagation network graph model, and use the nodes with measured score data as observation nodes;

[0111] Once the number of observation nodes accumulated within the same control zone reaches a preset threshold, the residual between the cascading propagation risk value of each observation node within the control zone and the corresponding measured score data is calculated.

[0112] Calculate the correlation coefficient between the residual and each model parameter in the preset diffusion equation, and determine the model parameter whose correlation coefficient is greater than the preset correlation threshold as the target parameter;

[0113] With minimizing the sum of squared residuals as the optimization objective, constrained optimization calculations are performed on the objective parameters to obtain the corrected parameter values.

[0114] Replace the target parameters in the preset diffusion equation with the corrected parameter values, and re-execute the iterative solution of the preset diffusion equation and the delineation of the control partitions until the boundaries of the control partitions generated in adjacent rounds reach the preset topological stability conditions.

[0115] It should be noted that the measured scoring data was collected by on-site inspectors at specific locations according to standardized scoring specifications. The scoring indicators include invasive species ground cover, species quantity, reproductive density (such as the number of seeds per unit area), and reproductive activity (such as the flowering / fruiting ratio). These are comprehensively quantified into a scoring value on the same scale as the cascade spread risk value for direct comparison.

[0116] In the aforementioned example, assume that after one quarter of operation, inspectors collected measured scoring data at multiple nodes along Route A. Among these, the midway inn ( Multiple nodes near the model generally exhibit the phenomenon that "the measured score is higher than the model prediction value," i.e., the residual... In the controlled zone The parameter correction process was triggered after 12 observation nodes were accumulated in the area. Attribution analysis revealed residuals. Distance factor from water source A significant positive correlation was observed (correlation coefficient 0.72), indicating that the model underestimated the intrusion risk near the stream area. The distance from the water source was weighted... The target parameter was determined and adjusted from 0.20 to 0.35 through constrained optimization. (The corrected parameter is then used.) Re-execute the diffusion equation iteration and control zone iteration. The converged graph shows that the cascading propagation risk value of multiple nodes near the stream near the inn has increased, and the control zone boundary has been slightly extended towards the area closer to the stream.

[0117] The visualization interface presents the differences before and after the correction using a version comparison view: the graphs before and after the correction are displayed side by side using a sliding mask, with the areas of difference highlighted by blue (decreased after correction) to red (increased after correction) color bands, and the parameter changes involved in this correction are noted on one side of the graph ("water source distance weight"). : The triggering reason is that the residuals of the 12 observation nodes in the intermediate inn area are significantly positively correlated with the water source distance factor. Based on this, the manager can trace the cause and scope of each correction and judge whether the correction is reasonable.

[0118] It should be noted that the invasive species suitability probability data and cultural and tourism activity traffic data involved in steps S1 to S6 above both exhibit seasonal temporal variation characteristics. The spread activity of invasive species is controlled by phenological cycles—the seeds of *Ageratum adenophorum* mature and are released in autumn, with summer and autumn being the peak spread periods, and the spread activity being lowest during the winter seed dormancy period. Cultural and tourism activities are driven by peak tourist seasons and event schedules—outdoor activities in the Greater Shangri-La region are highly concentrated from May to October. When switching to different quarterly time slices, the suitability probability data for the corresponding quarter is read. Traffic flow data Then, repeat all calculations from steps S3 to S6. Because species activity and human activity levels vary across seasons, the same propagation network model will present different cascading propagation risk patterns and control zoning schemes in different seasons—for example, the summer tourist season overlaps with the seed maturation period, resulting in the highest propagation risk and the largest control area; in winter, with fewer tourists and seed dormancy, the propagation risk is lowest, and most routes require no control. This seasonal risk map switching provides managers with a basis for formulating seasonal control strategies.

[0119] In summary, the spatial map visualization generation method for outdoor intrusion risk proposed in this application achieves the following through the complete process of steps S1 to S6:

[0120] First, by abstracting the cultural, sports and tourism route network into a propagation network graph model and establishing a diffusion equation on it, the risk assessment of invasion can be transformed from "where the species is" to "where humans bring the species". This can capture the cascading propagation effect along human activity channels that traditional grid models cannot express. The risk impact of adding a new route or discovering an informal channel on downstream areas far from the route can be quantitatively assessed.

[0121] Second, through the back-substitution mechanism of the control access coefficient and the diffusion-partitioning iterative convergence, the control partitions are not statically delineated based on the original risk distribution, but rather inherently contain the equilibrium result of "to what extent the risk will be suppressed after control according to this scheme". The boundaries of the control partitions adaptively adjust with the feedback of the control effect, avoiding unnecessary restrictions on cultural, sports and tourism activities due to over-control, and preserving the activity space to the maximum extent while ensuring the effectiveness of intrusion risk control.

[0122] Third, by dynamically inserting new physical channels and overlaying differential views, managers can assess the cascading impact of new routes on the overall risk landscape during the route planning stage, transforming risk assessment from post-event analysis to pre-event simulation, thus supporting the scientific formulation and dynamic optimization of control plans.

[0123] Exemplary system:

[0124] Figure 2 The illustration shows a spatial map visualization generation system for outdoor intrusion risk according to an embodiment of this application, including: a data acquisition module for acquiring spatial data of physical channels within a target area, the physical channels including routes for outdoor activities and associated transportation connections; a network modeling module for discretizing the spatial data and constructing a propagation network graph model; wherein each node in the propagation network graph model is associated with a local risk attribute representing the vulnerability of the environment to intrusion, each edge is associated with a propagation flux attribute representing the transfer capability of the physical channel, and a control passage coefficient representing the connectivity state of the physical channel, the initial value of the control passage coefficient representing that the corresponding physical channel is in a free passage state without intervention; and a risk solution module for using the local risk attribute of each node as the initial calculation state, and using the propagation flux attribute associated with each edge and the current control passage coefficient as topology transfer. The system employs several modules: a parameter module, a risk space module, and a graph module. The parameter module iteratively solves a pre-defined diffusion equation on the propagation network graph model to obtain the cascading propagation risk value of each node. The partitioning module delineates the target area into control zones with different control levels based on the cascading propagation risk values. It updates the control access coefficients associated with the edges within each control zone according to the control level, representing the actual connectivity changes after implementing flow restriction or blocking interventions on the corresponding physical channels. The iterative optimization module uses the updated control access coefficients as dynamic boundary conditions, substituting them back into the pre-defined diffusion equation to re-execute the iterative solution and control zone delineation until the boundaries of the control zones generated in adjacent rounds reach the pre-defined topological stability condition. The graph output module renders and outputs a risk space graph, which includes the cascading propagation risk values ​​of each node and the boundaries of the control zones that have reached the topological stability condition.

[0125] In one example, the delineation of control zones with different control levels specifically includes: determining a high-risk threshold based on the cascading propagation risk value of each node; extracting nodes with cascading propagation risk values ​​greater than or equal to the high-risk threshold as a high-risk node set; using the shortest path distance along the edges between nodes in the propagation network graph model as a topological distance metric, spatially clustering the high-risk node set to obtain at least one risk cluster; wherein, the topological distance metric characterizes the actual reachable path length of the physical channel; generating an envelope boundary surrounding each node within each risk cluster to delineate it as a control zone; and determining the control level of the corresponding control zone based on the statistics of the cascading propagation risk values ​​of each node contained within the risk cluster.

[0126] In one example, when the local risk attribute of a node changes, the iterative solution includes: marking the node whose local risk attribute has changed and its neighboring nodes within a preset hop count range on the propagation network graph model to form an affected node set; using the current cascading propagation risk value of the boundary nodes of the affected node set as a fixed boundary condition; re-executing the iterative solution of the preset diffusion equation only within the affected node set to obtain the updated cascading propagation risk value of each node in the affected node set; determining whether there are nodes in the affected node set whose cascading propagation risk value is greater than the high-risk threshold; if so, re-executing the delineation of the control zone and the update of the control passage coefficient locally in the region where the affected node set is located, and substituting the updated control passage coefficient back into the preset diffusion equation as a dynamic boundary condition.

[0127] In one example, the propagation flux attribute associated with each edge is determined as follows: the traffic flow data of the physical channel corresponding to each edge and the activity carrying coefficient corresponding to each physical channel are obtained; wherein, the activity carrying coefficient is determined based on the measured carrying amount of the target species' reproductive bodies in the surface attachment samples of the outdoor activity carriers traveling through the physical channel; the traffic flow data is multiplied by the activity carrying coefficient to obtain the propagation flux attribute associated with each edge.

[0128] In one example, the system may further include an extension module for: receiving spatial data of newly added physical channels and their corresponding traffic flow data and activity carrying coefficients; discretizing the spatial data of newly added physical channels at preset intervals to generate new node sequences and new edges connecting adjacent new nodes; extracting environmental variable data for each node in the new node sequence to determine its local risk attributes; determining the propagation flux attributes of each new edge based on traffic flow data and activity carrying coefficients; initially setting the control traffic coefficient associated with each new edge to a value representing a free-traffic state without intervention; inserting the new node sequence and new edges into the propagation network graph model; identifying the spatial proximity relationship between the newly added physical channels and existing physical channels at traffic junctions and establishing cross-route connection edges; using the propagation network graph model after inserting new edges as a basis to re-execute the iterative solution of the preset diffusion equation and the delineation of control zones until the boundaries of control zones generated in adjacent rounds reach the preset topological stability conditions; rendering and outputting a risk spatial map with a difference overlay view, the difference overlay view including the change in the cascade propagation risk value of each node before and after the access of the new physical channel, as well as the change in the boundary of the control zone.

[0129] In one example, the local risk attribute is determined as follows: Environmental variable data and the suitability probability data of the target species at the node's location are obtained. The environmental variable data includes land cover type, terrain slope, and distance to water sources. Based on land cover type, terrain slope, and distance to water sources, corresponding environmental factor sub-values ​​are calculated. These environmental factor sub-values ​​characterize the degree to which environmental variables promote the colonization and spread of the target species. The environmental factor sub-values ​​are then combined using a weighted summation method to form an environmental vulnerability score. The weighting coefficients for each environmental factor sub-value in the weighted summation method are determined through parameter search, using known risk distribution record points as positive samples and random background points as negative samples, with the goal of maximizing classification accuracy. The local risk attribute is obtained by multiplying the environmental vulnerability score by the sampled value of the suitability probability data at the node's location.

[0130] In one example, the system may also include a self-correcting module, used for: acquiring measured score data of risk status at some nodes in the propagation network graph model, and designating nodes with measured score data as observation nodes; when the number of observation nodes accumulated in the same control zone reaches a preset threshold, calculating the residual between the cascade propagation risk value of each observation node in the control zone and the corresponding measured score data; calculating the correlation coefficient between the residual and each model parameter in the preset diffusion equation, and determining the model parameter with a correlation coefficient greater than the preset correlation threshold as the target parameter; performing constrained optimization calculation on the target parameter with minimizing the sum of squares of the residual as the optimization objective, and obtaining the corrected parameter value of the target parameter; replacing the target parameter in the preset diffusion equation with the corrected parameter value, and re-executing the iterative solution of the preset diffusion equation and the delineation of the control zone until the boundary of the control zone generated in adjacent rounds reaches the preset topological stability condition.

[0131] Exemplary electronic device:

[0132] Figure 3 A block diagram of an electronic device according to an embodiment of this application is illustrated.

[0133] like Figure 3 As shown, the electronic device includes one or more processors and memory.

[0134] A processor can be a central processing unit (CPU) or other form of processing unit with data processing and / or instruction execution capabilities, and can control other components in an electronic device to perform desired functions.

[0135] The memory may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and / or cache memory. Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc.

[0136] In one example, the electronic device may also include input devices and output devices, which are interconnected via a bus system and / or other forms of connection mechanism (not shown).

[0137] Of course, for the sake of simplicity, Figure 3 Only some of the components of the electronic device relevant to this application are shown, omitting components such as buses, input / output interfaces, etc. In addition, the electronic device may include any other suitable components depending on the specific application.

[0138] Exemplary computer-readable media:

[0139] Embodiments of this application may also be computer-readable storage media storing computer program instructions thereon, which, when executed by a processor, cause the processor to perform the steps described in the "Exemplary Methods" section above according to the various embodiments of this application.

[0140] Computer-readable storage media may take the form of any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may, for example, include, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatuses, or devices, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: electrical connections having one or more wires, portable disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0141] The basic principles of this application have been described above with reference to specific embodiments. However, it should be noted that the advantages, benefits, and effects mentioned in this application are merely examples and not limitations, and should not be considered as essential features of each embodiment of this application. Furthermore, the specific details disclosed above are for illustrative and facilitative purposes only, and are not limitations. These details do not restrict this application from being implemented using the specific details described above.

[0142] The block diagrams of devices, apparatuses, devices, and systems involved in this application are merely illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. As those skilled in the art will recognize, these devices, apparatuses, devices, and systems can be connected, arranged, and configured in any manner. Words such as “comprising,” “including,” “having,” etc., are open-ended terms meaning “including but not limited to,” and are used interchangeably with them. The terms “or” and “and” as used herein refer to the terms “and / or,” and are used interchangeably with them unless the context clearly indicates otherwise. The term “such as” as used herein refers to the phrase “such as but not limited to,” and is used interchangeably with it.

[0143] It should also be noted that in the apparatus, equipment, and methods of this application, the components or steps can be disassembled and / or recombined. These disassemblies and / or recombinations should be considered as equivalent solutions of this application.

[0144] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use this application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein can be applied to other aspects without departing from the scope of this application. Therefore, this application is not intended to be limited to the aspects shown herein, but rather to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0145] The above description has been given for purposes of illustration and description. Furthermore, this description is not intended to limit the embodiments of this application to the forms disclosed herein. Although numerous exemplary aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations thereof.

Claims

1. A method for generating a spatial map visualization of outdoor intrusion risk, characterized in that, include: Acquire spatial data of physical pathways within the target area, including routes for outdoor activities and associated transportation connections; The spatial data is discretized and constructed into a propagation network graph model; wherein, each node in the propagation network graph model is associated with a local risk attribute that characterizes the vulnerability of the environment to intrusion, each edge is associated with a propagation flux attribute that characterizes the physical channel transfer capability, and a control passage coefficient that characterizes the physical channel connectivity state, and the initial value of the control passage coefficient indicates that the corresponding physical channel is in a free passage state without intervention. Using the local risk attributes of each node as the initial calculation state, the propagation flux attributes associated with each edge and the current control passage coefficient are used as topology transmission parameters. The preset diffusion equation is iteratively solved on the propagation network graph model to obtain the cascade propagation risk value of each node. Based on the cascading propagation risk value, the target area is divided into control zones with different control levels, and the control passage coefficient associated with the edges contained in the control zone is updated according to the control level to characterize the actual connectivity changes after implementing flow restriction or blocking intervention on the corresponding physical channel. The updated control passage coefficient is used as a dynamic boundary condition and substituted back into the preset diffusion equation to re-execute the iterative solution and delineation of the control partition until the boundary of the control partition generated in the adjacent rounds reaches the preset topological stability condition. Render and output a risk space map, which includes the cascading propagation risk value of each node and the boundary of the control partition that has reached the topological stability condition.

2. The method for generating a spatial map visualization of outdoor intrusion risk according to claim 1, characterized in that, The delineation of control zones with different control levels specifically includes: A high-risk threshold is determined based on the cascading propagation risk value of each node, and nodes whose cascading propagation risk value is greater than or equal to the high-risk threshold are extracted as a high-risk node set. Using the shortest path distance along the edges between nodes in the propagation network graph model as the topological distance metric, spatial clustering is performed on the high-risk node set to obtain at least one risk cluster; wherein, the topological distance metric characterizes the actual reachable path length of the physical channel; For each of the aforementioned risk clusters, an envelope boundary is generated that surrounds each node within it, to define the control zone; Based on the statistical values ​​of the cascading risk values ​​of each node within the risk cluster, the control level of the corresponding control zone is determined.

3. The method for generating a spatial map visualization of outdoor intrusion risk according to claim 2, characterized in that, When the local risk attribute of a node changes, the iterative solution includes: The nodes whose local risk attributes have changed and their neighboring nodes within a preset hop count range on the propagation network graph model are marked to form an affected node set; The current cascade propagation risk value of the boundary node of the affected node set is used as a fixed boundary condition. The iterative solution of the preset diffusion equation is re-executed only within the affected node set to obtain the updated cascade propagation risk value for each node within the affected node set; Determine whether there are any nodes in the affected node set whose cascading propagation risk value is greater than the high-risk threshold. If so, re-execute the delineation of the control partition and the update of the control passage coefficient in the local area where the affected node set is located, and substitute the updated control passage coefficient back into the preset diffusion equation as a dynamic boundary condition.

4. The method for generating a spatial map visualization of outdoor intrusion risk according to claim 1, characterized in that, The propagation flux attributes associated with each edge are determined in the following way: Obtain the traffic flow data of the physical channels corresponding to each side, and the activity carrying coefficient of each physical channel; wherein, the activity carrying coefficient is determined based on the measured carrying amount of the target species' reproductive bodies in the surface attachment samples of outdoor activity carriers traveling through the physical channels. Multiplying the traffic flow data by the activity carrying coefficient yields the propagation flux attribute associated with each edge.

5. The method for generating a spatial map visualization of outdoor intrusion risk according to claim 4, characterized in that, The method further includes: Receive spatial data of newly added physical channels and their corresponding traffic flow data and activity carrying coefficient; The spatial data of the newly added physical channel is discretized at a preset interval to generate a new node sequence and new edges connecting adjacent new nodes. Environmental variable data is extracted from each node in the new node sequence to determine its local risk attribute. The propagation flux attribute of each new edge is determined according to the traffic flow data and the activity carrying coefficient. The control passage coefficient associated with each new edge is initially set to a value that represents the free passage state without intervention. Insert the new node sequence and the newly added edge into the propagation network graph model, identify the spatial proximity relationship between the newly added physical channel and the existing physical channel at the traffic connection point, and establish cross-route connection edges; Using the propagation network graph model after inserting the new edge as the basis, the iterative solution of the preset diffusion equation and the delineation of the control partition are re-executed until the boundary of the control partition generated in adjacent rounds reaches the preset topological stability condition. The risk space map is rendered and output using a difference overlay view. The difference overlay view includes the change in the cascading propagation risk value of each node before and after the access of the new physical channel, as well as the change in the boundary of the control zone.

6. The method for generating a spatial map visualization of outdoor intrusion risk according to claim 1, characterized in that, The local risk attribute is determined in the following way: The environmental variable data of the node's location and the survival probability data of the target species are obtained. The environmental variable data includes land cover type, terrain slope and distance to water source. Based on the land cover type, the terrain slope, and the distance to the water source, the corresponding environmental factor sub-values ​​are calculated respectively; wherein, the environmental factor sub-values ​​characterize the degree to which environmental variables promote the colonization and spread of the target species; The environmental vulnerability score is synthesized by weighted summation of the values ​​of each environmental factor component; wherein, the weight coefficients corresponding to each environmental factor component value in the weighted summation method are determined by parameter search with known risk distribution record points as positive samples and random background points as negative samples, with the goal of maximizing classification accuracy; The local risk attribute is obtained by multiplying the environmental vulnerability score by the sampled value of the adaptability probability data at that node.

7. The method for generating a spatial map visualization of outdoor intrusion risk according to claim 1, characterized in that, The method further includes: Obtain measured score data for the risk status of some nodes in the propagation network graph model, and use the nodes with the measured score data as observation nodes; When the number of observation nodes accumulated within the same control zone reaches a preset threshold, the residual between the cascading propagation risk value of each observation node within the control zone and the corresponding measured score data is calculated. Calculate the correlation coefficient between the residual and each model parameter in the preset diffusion equation, and determine the model parameter whose correlation coefficient is greater than the preset correlation threshold as the target parameter; With minimizing the sum of squares of the residuals as the optimization objective, constrained optimization calculations are performed on the objective parameters to obtain the corrected parameter values. Replace the target parameter in the preset diffusion equation with the corrected parameter value, and re-execute the iterative solution of the preset diffusion equation and the delineation of the control partition until the boundary of the control partition generated in adjacent rounds reaches the preset topological stability condition.

8. A spatial map visualization generation system for outdoor intrusion risk, characterized in that, include: The data acquisition module is used to acquire spatial data of physical channels within the target area, including routes for outdoor activities and associated transportation connections. The network modeling module is used to discretize the spatial data and construct a propagation network graph model. Each node in the propagation network graph model is associated with a local risk attribute that represents the vulnerability of the environment to intrusion. Each edge is associated with a propagation flux attribute that represents the physical channel transfer capability and a control passage coefficient that represents the physical channel connectivity state. The initial value of the control passage coefficient represents that the corresponding physical channel is in a free passage state without intervention. The risk solution module is used to perform iterative solution of the preset diffusion equation on the propagation network graph model, taking the local risk attributes of each node as the initial calculation state, the propagation flux attributes associated with each edge and the current control passage coefficient as topology transfer parameters, to obtain the cascade propagation risk value of each node. The partitioning module is used to delineate the target area into control partitions with different control levels based on the cascading propagation risk value, and update the control access coefficient associated with the edges contained in the control partition according to the control level, so as to characterize the actual connectivity status change after implementing flow restriction or blocking intervention on the corresponding physical channel. The iterative optimization module is used to take the updated control passage coefficient as a dynamic boundary condition, substitute it back into the preset diffusion equation, and re-execute the iterative solution and control partition delineation until the boundary of the control partition generated in the adjacent rounds reaches the preset topological stability condition. The graph output module is used to render and output a risk space graph, which includes the cascading propagation risk value of each node and the boundary of the control partition that has reached the topological stability condition.

9. An electronic device comprising a memory and a processor, characterized in that: The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions, which, when executed by the processor, implement the steps of the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium having computer-executable instructions stored thereon, characterized in that: When the computer-executable instructions are executed by a processor, they implement the steps of the method as described in any one of claims 1 to 7.

Images