Method and system for evaluating network risk of rural road in mountainous area

Abstract

The invention provides a mountainous area and rural area road network risk evaluation method and system, and relates to the technical field of geological disasters, and the method comprises the steps: obtaining the distance from a fault, the rock property and the landform type of a mountainous area and rural area as evaluation factors and evaluation factor weights; obtaining a road network risk classification result, a road network vulnerability classification result and a road network importance classification result; and obtaining a risk evaluation model, and combining the road network vulnerability classification result and the road network exposure degree classification result to obtain a risk evaluation map of the to-be-evaluated mountain road after heavy rainfall.

Description

Technical Field

[0001] The present invention relates to the technical field of geological disasters, and particularly relates to a method and system for risk assessment of mountain rural road networks. Background Art

[0002] Regarding mountain geological disasters caused by extreme rainfall, the current responses mainly focus on the microscopic site level, and are mostly engineering and structural risk assessment technologies, which cannot effectively evaluate the post-disaster risks of the macroscopic and holistic rural road networks. Summary of the Invention

[0003] In order to overcome the shortcomings of the above-mentioned prior art that cannot realize the risk assessment such as road interruption and delay of rescue caused by the occurrence of a large number of risk points along mountain rural roads after heavy rainfall, the main object of the present invention is to provide a method and system for risk assessment of mountain rural road networks, which can effectively evaluate the risks such as road interruption and delay of rescue caused by the occurrence of a large number of risk points along mountain rural roads after heavy rainfall.

[0004] To achieve the above object, the present invention adopts the following technical solutions. A method for risk assessment of mountain rural road networks includes the following steps: Obtain the distance from the fault, rock properties, and landform types of mountain villages as evaluation factors. The evaluation factors are combined with corresponding weights, subjected to spatial weighting processing, and the weighted processing results are assigned to the regional road network and classified to obtain the classified results of the road network hazard after assignment; Obtain the distance from the fault, the distance from water bodies, and the distance from artificial structures of mountain villages as geographical feature elements. Obtain the geological disaster data caused by rainfall in the mountain villages to be evaluated in history, and train a random forest model to obtain a trained random forest model for predicting geological disasters after rainfall; Obtain the geographical feature elements and rainfall data of the mountain villages to be evaluated, input them into the trained random forest model to obtain a probability value grid of the occurrence of geological disasters after rainfall, assign the probability value grid within the buffer range on both sides of the regional road network to the regional road network, and classify the assigned regional road network according to the natural break method to obtain the classified results of the road network vulnerability; Map the regional road network of mountain villages into an undirected topological network, where the nodes represent road intersections, endpoints, and the connection points between villages and roads, and the links represent the roads between adjacent nodes; Construct an adjacency matrix for the connection relationships of all node pairs in the undirected topological network, use the road length and village population as the weights of the adjacency matrix to obtain the importance of the road network, and classify the importance of the road network according to the natural break method to obtain the classified results of the road network importance; Introduce a risk assessment model, and combine the classification results of the vulnerability, hazard, and importance of the road network to obtain the risk assessment result of the road network.

[0005] The obtaining of the road network hazard classification result includes the following steps: Obtain the weights corresponding to the evaluation factors through the entropy weight TOPSIS method; Perform spatial weighting processing according to the evaluation factors and the corresponding weights to obtain the grid values within the buffer range on both sides of the regional road network, and assign the grid values to the regional road network to obtain the assigned regional road network; Based on the assigned regional road network, classify the geological hazard risk of the road network into multiple categories to obtain the road network hazard classification result.

[0006] The trained random forest model for predicting geological disasters after rainfall includes the following steps: Perform disaster annotation on the geological disaster data caused by historical rainfall in the mountainous rural areas to be evaluated to obtain historical disaster annotation data; Based on the historical disaster annotation data, combine the obtained multiple geographical elements as independent variables. The independent variables include categorical variables and continuous variables. Perform coding processing on the categorical variables to obtain the coding processing result; the continuous variables are standardized using min-max normalization to obtain the standard processing result. Use the coding processing result and the standard processing result as the preprocessed historical disaster annotation data; Use the stepwise feature elimination method to perform feature selection on the preprocessed historical disaster annotation data to obtain historical disaster features, and divide them into a training set and a validation set. Train and validate the random forest model to obtain the trained random forest model for predicting geological disasters after rainfall.

[0007] The obtaining of the importance of the road network includes the following steps: Take the road length and village population as the weights of the adjacency matrix to obtain the initial efficiency of the network; By removing a single regional road network, obtain the change in the overall regional road network efficiency after losing this road, which represents the importance of the road network.

[0008] A risk assessment system for mountainous rural road networks includes: A hazard analysis module, which is used to obtain the distance to the fault, rock properties, and landform types in mountainous rural areas as evaluation factors. The evaluation factors are combined with the corresponding weights, perform spatial weighting processing and assign them to the regional road network, and classify the assigned regional road network to obtain the road network hazard classification result; Vulnerability analysis module, which is used to obtain the distances from mountain villages to faults, to water bodies, and to artificial structures as geographical feature elements, obtain the historical geological disaster data caused by rainfall in the mountain villages to be evaluated, train a random forest model, and obtain a trained random forest model for predicting geological disasters after rainfall; obtain the geographical feature elements and rainfall data of the mountain villages to be evaluated, input them into the trained random forest model for predicting geological disasters after rainfall, obtain a probability value grid of the occurrence of geological disasters after rainfall, assign the probability value grid within the buffer range on both sides of the regional road network to the regional road network, classify the assigned regional road network according to the natural break method, and obtain the road network vulnerability classification result; Importance analysis module, which is used to map the regional road network of mountain villages into an undirected topological network, where nodes represent road intersections, endpoints, and connection points between villages and roads, and links represent roads between adjacent nodes; construct an adjacency matrix for the connection relationships of all node pairs in the undirected topological network, and use the road length and village population as the weights of the adjacency matrix to obtain the importance of the road network; classify the importance of the road network according to the natural break method to obtain the road network importance classification result; Risk analysis module, which is used to introduce a risk assessment model, and combine the classification results of the vulnerability, hazard, and importance of the road network to obtain the risk assessment result of the road network.

[0009] Compared with the prior art, the beneficial effects of the present invention are as follows: The present invention uses comprehensive multi-source data to evaluate the risk degree of the road network in mountain villages after heavy rainfall, providing an effective way for the optimization of the road network in mountain villages, the planning of the village system, and post-disaster emergency rescue. Compared with the traditional verification of geological disaster risk ledger points for rural roads, this method helps to quickly and comprehensively identify high-risk areas from a macroscopic regional level, contributing to the rapid formulation of pre-disaster protection plans, the rapid decision-making of emergency rescue plans during disasters, and the rapid development of post-disaster recovery plans. Brief Description of the Drawings

[0010] The drawings described herein are used to provide a further understanding of the present application and form a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation to the present application.

[0011] Figure 1 It is a schematic structural diagram of the process of the present invention. Detailed Embodiments

[0012] Regarding mountain geological disasters caused by extreme rainfall, the current responses mainly focus on the microscopic site level, mostly engineering and structural risk assessment technologies, and there is no mature risk assessment method for the macroscopic and overall rural road network and village system.

[0013] The existing risk assessment of geological hazard points focuses on the direct impacts brought by geological hazard points, mainly focusing on the engineering and structural risk assessment techniques at the micro-level site.

[0014] In order to overcome the disadvantages in the above-mentioned existing technologies, such as the inability to conduct risk assessments on risks such as road interruption and rescue delay caused by the occurrence of a large number of risk points along mountain rural roads after heavy rainfall, the main purpose of the present invention is to provide a risk assessment method and system for mountain rural road networks, which can effectively conduct risk assessments on risks such as road interruption and rescue delay caused by the occurrence of a large number of risk points along mountain rural roads after heavy rainfall. Refer to Figure 1 , the present invention adopts the following technical solutions. A risk assessment method for mountain rural road networks includes the following steps: Obtain the distance from faults, rock properties, and landform types of mountain villages as evaluation factors. The evaluation factors are combined with corresponding weights, subjected to spatial weighting processing, and the weighted processing results are assigned to the regional road network and classified to obtain the classified results of the road network hazard after assignment; Obtain the distance from faults, the distance from water bodies, and the distance from artificial structures of mountain villages as geographical feature elements. Obtain the geological disaster data caused by historical rainfall in the mountain villages to be evaluated, train the random forest model, and obtain the trained random forest model for predicting geological disasters after rainfall; Obtain the geographical feature elements and rainfall data of the mountain villages to be evaluated, input them into the trained random forest model, obtain the probability value grid of geological disasters occurring after rainfall, assign the probability value grid within the buffer range on both sides of the regional road network to the regional road network, and classify the assigned regional road network according to the natural break method to obtain the classified results of the road network vulnerability; Map the regional road network of mountain villages into an undirected topological network, where nodes represent road intersections, endpoints, and the connection points between villages and roads, and links represent the roads between adjacent nodes; Construct an adjacency matrix for the connection relationships of all node pairs in the undirected topological network, use the road length and village population as the weights of the adjacency matrix to obtain the importance of the road network, and classify the importance of the road network according to the natural break method to obtain the classified results of the road network importance; Introduce a risk assessment model, and combine the classified results of the vulnerability, hazard, and importance of the road network to obtain the risk assessment results of the road network.

[0015] The present invention will be further described below in conjunction with the drawings and embodiments.

[0016] Example 1: Select the distance from the fault, rock properties, and landform type as evaluation factors, use the entropy weight TOPSIS method to calculate the weights of each factor, and conduct spatial weighting processing based on the weights. Calculate the raster values within the buffer areas on both sides of the road network and assign them to the road network. Divide the geological disaster risk of the road network into five categories according to the calculation results to obtain the geological disaster risk of the road network.

[0017] For obtaining the vulnerability of the road network under heavy rainfall, it specifically includes: First is the preparation of relevant basic data.

[0018] In the first step, obtain the historical geological disaster data caused by rainfall, and make labels indicating whether geological disasters occurred.

[0019] In the second step, select 17 geographical elements of 6 categories as independent variables.

[0020] (1) For terrain factors, based on digital elevation model (DEM) data, use ArcGIS software to extract terrain features such as elevation, slope, aspect, curvature, plane curvature, and profile curvature. These factors affect the stability of the geological environment.

[0021] (2) The disturbance of human factors is mainly reflected in the destructive impact of road construction and urban built-up area development on the surface structure. Therefore, the distances from roads and built-up areas are selected as quantification indicators to evaluate the impact of human activities on geological stability.

[0022] (3) Geological factors reflect the stability of the geological structure within the region. Select the distance from the fault and lithology indicators as representatives.

[0023] (4) For surface cover factors, consider the stabilizing or promoting effects of vegetation and soil on geological disasters such as landslides. Use the normalized difference vegetation index (NDVI) and land use type to characterize the influence of different surface covers on the formation of geological disasters.

[0024] (5) In terms of hydrological factors, the aggravating effect of water bodies on rock and soil erosion and the loosening and deformation effects on soil structure are significant. Therefore, select the distance from rivers, terrain wetness index, and river power index to evaluate hydrological conditions.

[0025] (6) In terms of environmental factors, the persistence of rainfall or instantaneous heavy rainfall will significantly affect geological stability, especially in the induction of disasters such as landslides. Select the 5-day average rainfall intensity and 5-day maximum rainfall intensity in extreme rainfall events as environmental factors.

[0026] That is, in the calculation of vulnerability, i.e., geological disaster susceptibility, it includes: In the calculation of vulnerability, i.e., geological disaster susceptibility, the distances from faults, from water bodies, and from artificial structures are selected, and the entropy weight TOPSIS method is used to calculate the weights. First, a data matrix is established based on the indicator factors covering three categories. The data is normalized (min-max normalization) to generate a normalized data matrix. Then, the entropy weight method is used to calculate the entropy value and determine the weights of the indicator factors, calculate the positive and negative ideal solutions, and finally calculate the distances from the positive and negative ideal solutions and then calculate the comprehensive score for grid assignment. Next, spatial overlay is performed. The grid values within the buffer range on both sides of the road network are counted and assigned to the road network, and reclassification is carried out using the natural breaks method to divide the road vulnerability into five categories.

[0027] 。

[0028] 。

[0029] 。

[0030] Among them, is the proportion of the i th project in the j th category, represents the data of the i th project in the j th category after normalization.

[0031] 。

[0032] Among them, 。

[0033] Among them, m is the number of regions, is the information entropy of the j th category.

[0034] 。

[0035] Among them, is the weight of the j th category, n is the number of indicators.

[0036] 。

[0037] 。

[0038] Among them, is the positive ideal solution, is the negative ideal solution.

[0039] 。

[0040] 。

[0041] Among them, is the Euclidean distance to the positive ideal solution, is the Euclidean distance to the negative ideal solution.

[0042] 。

[0043] Among them, is the comprehensive score of the i th region, and the larger the value, the higher the risk.

[0044] Data preprocessing and random forest model training in the present invention. First, categorical variables are encoded, and continuous variables are standardized using min-max normalization. If the number of disaster-occurring points and non-disaster-occurring points is unbalanced, random resampling is used to balance the samples.

[0045] In the model construction stage, stepwise feature elimination is used for feature selection, and the sample data is divided into a training set and a test set in a ratio of 8:2 for model training and validation.

[0046] Through the trained random forest model, input the geographical feature elements of the evaluation area and the rainfall data of a certain extreme emergency response event, obtain the probability value raster of geological disaster points after rainfall, count the raster values within the buffer range on both sides of the road network and assign them to the road network, establish natural breakpoint reclassification, divide the road network vulnerability into five categories, and obtain the road network vulnerability classification result.

[0047] Exposure calculation is based on complex network theory to evaluate the performance of the road transportation system. This method considers the connectivity of the road and the impact on the rescue demand of village population, and uses two weight factors of road length and village population in the road network efficiency formula.

[0048] First, map the regional road network into an undirected topological network, where nodes represent road intersections, endpoints, and connection points between villages and roads, and links represent road segments between adjacent nodes.

[0049] Subsequently, construct an adjacency matrix to represent the connection relationship of all node pairs in the network. This adjacency matrix uses road length as the weight. Among them, d i and d j represent adjacent nodes i and jThe actual length of the road segment in between is used as the weight of the road connection. Then, the population weight is added to calculate the rescue intensity. Villages with a larger population usually have more demand for emergency rescue. Therefore, the service capacity of the road system is measured based on the weight of the affected population. Increase the population weight W ij , comprehensively considering the structural and functional characteristics of the road network, that is, the road length and the number of people who may need rescue.

[0050] .

[0051] Among them, N represents the number of nodes in the network.

[0052] The importance of a single road network is calculated by removing each road network and calculating the change in the overall network efficiency. The network efficiency of a single road can be expressed as: .

[0053] Among them, E ij is the change in the overall road network efficiency after the loss of road ij , representing the road importance, E is the initial network efficiency, E’ is the network efficiency after removing d ij .

[0054] Through the natural break point method for classification, the road importance is divided into five categories.

[0055] Introduce the risk assessment model formula. The riskiness (R) can be expressed as: R = f(H, V, E), where H represents the hazard, V represents the vulnerability, and E represents the exposure degree, that is, the importance. The riskiness is a function of H, V, and E. In the risk calculation framework of the mountainous rural road network caused by strong rainfall-triggered geological disasters, H is the hazard of the road network geological disasters, reflecting the geological environmental conditions, which is the internal factor for the occurrence and development evolution of geological disasters; V is the vulnerability of the road network under strong rainfall, reflecting the possibility of geological disasters triggered by rainfall under strong rainfall conditions; E is the road network importance, reflecting the topological connectivity of the road network and the difference in village population accessibility. Use the risk assessment model formula to draw the risk assessment map of the mountainous roads after strong rainfall.

[0056] Example 2: A certain district in a certain city is located in the transitional zone between the mountainous area and the plain of a certain water system. It is located at the junction of Mountain A and Mountain B, with an unstable geological structure and is vulnerable to rainstorm floods and secondary geological disasters. The settlements in this district are scattered in the river valley area, and rescue responses need to be carried out in a timely manner after disasters occur. Therefore, this invention takes this district as an example to demonstrate the application method of the invention. The scope selects the small-scale inhabited mountain village area in a certain district excluding the urban cluster within the administrative division.

[0057] Grid data such as elevation, slope, and aspect are extracted from a 30m×30m resolution digital elevation model, and the data is sourced from the Geospatial Data Cloud; engineering geological rock group and fault data are sourced from a 1:50,000 geological map; road and water system data are sourced from online geographical data; the normalized difference vegetation index is calculated from the bands of 30m×30m resolution Landsat8 image data, and the data is sourced from the Geospatial Data Cloud.

[0058] In the calculation of vulnerability, i.e., the susceptibility to geological disasters, the distance from faults and water bodies is selected as the key factor, that is, the higher the vulnerability of the road network closer to the river and the fault. First, the entropy weight method is used to calculate the weights of each factor, and then the TOPSIS method is used to evaluate the risk. The grid values within the buffer range on both sides of the road network are statistically analyzed and assigned to the road network; then, the natural break point method is used to reclassify the data, and the road vulnerability is divided into five categories.

[0059] In the calculation of hazard, i.e., geological stability, the dependent variable is defined as the risk classification of geological disaster points; in the calculation of hazard, i.e., rainstorm flood risk, the dependent variable is defined as the classification of whether geological disasters occur under specific rainfall conditions. Since there is a slight imbalance in both positive and negative samples of the two dependent variables, the ADASYN method is used to balance the sample size. This method enhances the number of minority class samples by generating synthetic samples and adaptively selects the minority class samples for which synthetic samples are to be generated according to the distribution characteristics of the samples.

[0060] In terms of selecting independent variables, a total of six types of factors and 17 elements are selected for analysis, including the distance from faults, the distance from water bodies, and the distance from artificial structures in mountain villages. In terms of environmental factors, the persistence of rainfall or instantaneous heavy rainfall will significantly affect geological stability, especially in the induction of disasters such as landslides. Therefore, the 5-day average rainfall intensity and the 5-day maximum rainfall intensity in a certain rainfall event are selected as environmental factors. Based on the digital elevation model data for topographic factors, topographic features such as elevation, slope, aspect, plane curvature, and profile curvature are extracted using ArcGIS software. These factors not only affect the stability of the geological environment but also play an important role in the formation of regional climate. In terms of hydrological factors, the intensifying effect of water bodies on rock and soil erosion and the loosening and deformation of soil structure are significant. Therefore, the distance from the river, the terrain humidity index, and the river power index are selected for hydrological condition assessment.

[0061] The disturbances of human factors are mainly reflected in the destructive effects of road construction and urban built-up area development on the surface structure. Therefore, the distances from roads and built-up areas are selected as quantitative indicators to evaluate the impact of human activities on geological stability. Geological factors reflect the stability of the geological structure within the region, and the distances from faults and lithology indicators are selected as representatives. The surface cover factor takes into account the stabilizing or promoting effects of vegetation and soil on geological disasters such as landslides, and the normalized difference vegetation index and land use type are used to characterize the impact of different surface covers on geological disasters.

[0062] In the data preprocessing process, categorical variables are encoded, and continuous variables are standardized using min-max normalization. In the model construction stage, stepwise feature elimination is used for feature selection, and the sample data is divided into a training set and a test set in a ratio of 8:2 for model training and validation.

[0063] The accuracy of the disaster-induced factor model on the test set is 0.80, the positive class precision is 0.83, the recall rate is 0.74, and the F1 score is 0.78; the negative class precision is 0.77, the recall rate is 0.85, and the F1 score is 0.81. The macro-average and weighted-average F1 scores are both 0.79, the AUC value of the ROC curve is 0.87, and NDVI, faults, and the distance from roads are the more important features.

[0064] From the calculation results, it can be seen that the main roads running through a certain area are generally less threatened by geological disasters, while the branch roads connecting each village show higher risks.

[0065] The exposure calculation is based on complex network theory to evaluate the performance of the road transport system. The evaluation includes the connectivity of roads and the rescue needs of the village population, and two weight factors, road length and village population, are used in the road network efficiency formula.

[0066] First, the regional road network is mapped into an undirected topological network, where nodes represent road intersections, endpoints, and the connection points between villages and roads, and links represent the road segments between adjacent nodes. Subsequently, an adjacency matrix is constructed to represent the connection relationships of all node pairs in the network. This adjacency matrix uses road length as the weight. Among them, d i and d j represent adjacent nodes i and jThe actual length of the road segment in between is used as the weight of the road connection. The population weight is then added to calculate the rescue intensity. Villages with a larger population usually have a greater demand for emergency rescue. Therefore, the service capacity of the road system is measured based on the weight of the affected population. The population weight Wij is increased, comprehensively considering the structural and functional characteristics of the road network, namely the road length and the number of people who may need rescue. The modified network efficiency becomes a comprehensive indicator for measuring network connectivity and the population transportation capacity of villages. The system integrates the network topology, system redundancy, and network function. A shorter path length and a larger passenger flow mean a higher network efficiency.

[0067] 。

[0068] Among them, N is the number of nodes in the network; d ij is the node i and the node j the shortest path length between. W ij is the weight factor, calculated by the total population of the villages passing through the node i and the node j in between, representing the connection strength between nodes, that is, the number of village populations is proportional to the total population distributed at the node i and the node j .

[0069] The occurrence of geological disasters will cause some links in the road network to break, thus affecting the overall network efficiency. First of all, it will lead to the loss of connections between some nodes and the population transportation volume of villages, that is, the path length d ij becomes infinite, and the transportation weight factor W ij becomes zero. Secondly, the shortest path length between some node pairs will increase, that is, d ij becomes larger. Assuming that a landslide or other disaster causes a complete interruption of a road connection, the connection index between the corresponding nodes in the adjacency matrix will become infinite. After updating the matrix, the performance of the road network after failure is measured by the modified network efficiency.

[0070] To quantitatively evaluate the impact of landslide disasters on the entire road network, an impact assessment is carried out based on the reduction of network efficiency caused by the interruption of a single road connection due to a landslide. The formula is as follows: 。

[0071] The middle section of the road between Village A and Village B entering a certain mountainous area has a high exposure degree when facing disasters due to its critical location and the large number of people it carries.

[0072] Introduce a risk assessment model to calculate the road risk. Use the formula R = H × E × V, and combine the classification results of the vulnerability, hazard, and importance of the road network to obtain the risk assessment result of the road network.

[0073] It should be noted that in the present invention, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article or device.

[0074] The above embodiments are only illustrative examples of the present invention and do not constitute a limitation on the protection scope of the present invention. Any design identical or similar to the present invention falls within the protection scope of the present invention.

Claims

1. A risk assessment method for mountain rural road networks, characterized in that Including the following steps: Obtain the distance from faults, rock properties, and geomorphic types in mountainous rural areas as evaluation factors. Combine the evaluation factors with corresponding weights, perform spatial weighting processing, assign the weighted processing results to the regional road network and classify it to obtain the classified result of the risk of the assigned road network; Obtain the distance from faults, distance from water bodies, and distance from artificial structures in mountainous rural areas as geographical feature elements. Obtain the historical geological disaster data caused by rainfall in the mountainous rural area to be evaluated, and train the random forest model to obtain the trained random forest model for predicting geological disasters after rainfall; Obtain the geographical feature elements and rainfall data of the mountainous rural area to be evaluated, input them into the trained random forest model to obtain the probability value grid of geological disasters occurring after rainfall, assign the probability value grid within the buffer range on both sides of the regional road network to the regional road network, and classify the assigned regional road network according to the natural breakpoint method to obtain the classified result of the vulnerability of the road network; Map the regional road network of the mountainous rural area into an undirected topological network, where the nodes represent road intersections, endpoints, and connection points between villages and roads, and the links represent the roads between adjacent nodes; Construct an adjacency matrix for the connection relationships of all node pairs in the undirected topological network, use the road length and village population as the weights of the adjacency matrix to obtain the importance of the road network, and classify the importance of the road network by the natural breakpoint method to obtain the classified result of the importance of the road network; Introduce a risk assessment model, and combine the classified results of the vulnerability, risk, and importance of the road network to obtain the risk assessment result of the road network.

2. The risk assessment method for mountain rural road networks according to claim 1, characterized in that, The obtaining of the classified result of the risk of the road network includes the following steps: Obtain the weights corresponding to the evaluation factors by the entropy weight TOPSIS method; Define the buffer of the regional road network and perform gridification processing to obtain the gridified buffer. Assign the data values of the evaluation factors to the gridified buffer to obtain the assigned grid buffer; According to the evaluation factors and corresponding weights, perform spatial weighting processing on the evaluation factors of each grid cell on the assigned grid buffer to obtain the grid values within the buffer range on both sides of the regional road network, assign the grid values to the regional road network to obtain the assigned regional road network; Based on the assigned regional road network, divide the geological disaster risk of the road network into multiple categories to obtain the classified result of the risk of the road network.

3. The method for evaluating the risk of the mountainous rural road network according to claim 1, characterized in that The trained random forest model for predicting geological disasters after rainfall includes the following steps: Perform disaster annotation on the historical geological disaster data caused by rainfall in the mountainous rural area to be evaluated to obtain the historical disaster annotation data; Based on historical disaster annotation data, combining the obtained multiple geographical elements as independent variables, where the independent variables include categorical variables and continuous variables, encoding the categorical variables to obtain an encoding result; normalizing the continuous variables using min-max normalization to obtain a normalized result, and taking the encoding result and the normalized result as the preprocessed historical disaster annotation data; Using the stepwise feature elimination method to perform feature selection on the preprocessed historical disaster annotation data to obtain historical disaster features, and dividing them into a training set and a validation set, training and validating a random forest model to obtain a trained random forest model for predicting geological disasters after rainfall.

4. The risk assessment method for mountain rural road networks according to claim 1, characterized in that The obtaining of the importance of the road network includes the following steps: Taking the road length and the village population as the weights of the adjacency matrix to obtain the initial efficiency of the network; By removing a single regional road network, obtaining the change in the overall regional road network efficiency after losing this road, which represents the importance of the road network.

5. A risk assessment system for mountain rural road networks, characterized in that, Including: A hazard analysis module for obtaining the distance from faults, rock properties, and landform types in mountainous rural areas as evaluation factors, performing spatial weighted processing on the evaluation factors combined with corresponding weights and assigning them to the regional road network, and classifying the assigned regional road network to obtain the road network hazard classification result; A vulnerability analysis module for obtaining the distance from faults, the distance from water bodies, and the distance from artificial structures in mountainous rural areas as geographical feature elements, obtaining the historical geological disaster data caused by rainfall in the mountainous rural areas to be evaluated, training a random forest model to obtain a trained random forest model for predicting geological disasters after rainfall; obtaining the geographical feature elements and rainfall data of the mountainous rural areas to be evaluated, inputting them into the trained random forest model for predicting geological disasters after rainfall to obtain a probability value grid of geological disasters after rainfall, and assigning the probability value grid within the buffer range on both sides of the regional road network to the regional road network, and classifying the assigned regional road network according to the natural break method to obtain the road network vulnerability classification result; An importance analysis module for mapping the regional road network in mountainous rural areas into an undirected topological network, where the nodes represent road intersections, endpoints, and the connection points between villages and roads, and the links represent the roads between adjacent nodes; constructing an adjacency matrix for the connection relationships of all node pairs in the undirected topological network, and taking the road length and the village population as the weights of the adjacency matrix to obtain the importance of the road network; Classifying the importance of the road network according to the natural break method to obtain the road network importance classification result; A risk analysis module for introducing a risk assessment model and combining the classification results of the vulnerability, hazard, and importance of the road network to obtain the risk assessment result of the road network.

6. A computer-readable storage medium, characterized in that, A storage medium stores a computer program, and when the computer program is executed by a processor, the method described in any one of claims 1 to 4 above is implemented.

7. A computer device, characterized in that, Including a memory, a processor, and a computer program stored on the memory and executable on the processor, and when the processor executes the program, the method described in any one of claims 1 to 4 above is implemented.

Images