Method and device for evaluating influence of avalanche disaster on road based on deep learning

Abstract

The invention provides a deep learning-based method and device for evaluating the influence of an avalanche disaster on a road, and relates to the technical field of disaster coping, and the method comprises the steps: obtaining the historical avalanche monitoring data of a monitoring region; establishing an avalanche disaster influence prediction model according to the topographic parameters and the physical parameters of the snow body before avalanche; according to the avalanche disaster influence characteristics, the road positioning data and the post-avalanche road influence parameters, establishing an avalanche disaster road influence prediction model; acquiring current data of the monitoring area; obtaining current avalanche disaster influence characteristics according to the avalanche disaster influence prediction model; and predicting the influence of the avalanche disaster on each road section according to the avalanche disaster road influence prediction model. According to the method, key parameters of an avalanche occurrence area can be acquired in real time, and the occurrence probability of an avalanche disaster and the influence degree of the avalanche disaster on a road can be accurately predicted and early warned by means of a deep learning algorithm, so that the safety and traffic capacity of the road are effectively improved.

Description

Technical Field

[0001] The present invention relates to the field of disaster response technology, and in particular to a method and device for evaluating the impact of avalanche disasters on roads based on deep learning. Background Art

[0002] Avalanches are one of the major natural disasters in the mountainous areas of western China. They are characterized by suddenness, wide range of disasters, and strong destructiveness. As climate change and human activities continue to expand into high-altitude and cold mountainous areas, the population and infrastructure exposed to the danger of avalanches are increasing, seriously threatening the normal operation of transportation lines, power transmission lines, pipelines and human activities in high-altitude and cold mountainous areas. Avalanche disaster prevention and control is one of the current focuses of disaster management in the western mountainous areas, and it is also an important issue that needs to be urgently addressed in the current development and construction of the western mountainous areas.

[0003] In traditional assessment work, avalanche disaster risk assessment relies heavily on the experience and judgment of monitoring personnel, and predictions are also carried out through expert experience methods and mechanism models, but the accuracy of avalanche forecasts is low and the results are unstable. Different terrains, snowfall, slopes, wind speeds and temperatures also have significant effects on avalanche formation. Achieving a scientific assessment of the impact of avalanche disasters on roads and providing accurate and detailed risk prediction results are not only of great significance to the construction and operation of engineering projects and the safety of life and property, but also have significant social and economic benefits.

[0004] Therefore, how to conduct a safety risk assessment of roads posed by avalanche hazards without relying on researchers' experience is an urgent problem that needs to be solved. Summary of the invention

[0005] In order to improve the above problems, the present invention provides a method and device for evaluating the impact of avalanche disasters on roads based on deep learning.

[0006] A first aspect of an embodiment of the present invention provides a method for evaluating the impact of avalanche disasters on roads based on deep learning, the method comprising: Acquire historical avalanche monitoring data of the monitoring area, wherein the historical avalanche monitoring data includes terrain parameters of snow accumulation before historical avalanches, physical parameters of snow accumulation before historical avalanches, road positioning data, and road impact parameters after historical avalanches; Establishing an avalanche disaster impact prediction model based on the terrain parameters of the snow body before the avalanche and the physical parameters of the snow body before the avalanche, wherein the input of the avalanche disaster impact prediction model is the terrain parameters of the snow body before the avalanche and the physical parameters of the snow body before the avalanche, and the output of the avalanche disaster impact prediction model is the avalanche disaster impact characteristics corresponding to the avalanche disaster impact range; Establishing an avalanche disaster road impact prediction model according to the avalanche disaster impact characteristics, road positioning data and post-avalanche road impact parameters, wherein the input of the avalanche disaster road impact prediction model is the avalanche disaster impact characteristics and road positioning data, and the output of the avalanche disaster road impact prediction model is the post-avalanche road impact parameters; Obtain the current snow mass topographic parameters, current snow mass physical parameters and road positioning data of each road section in the monitoring area; Obtaining the current avalanche disaster impact characteristics according to the avalanche disaster impact prediction model; The impact of avalanche disasters on each road section is predicted according to the avalanche disaster road impact prediction model to obtain the post-avalanche road impact parameters of each road section in the monitoring area.

[0007] Optionally, the step of establishing an avalanche disaster impact prediction model based on the terrain parameters of the snow body before the avalanche and the physical parameters of the snow body before the avalanche specifically includes: Taking the terrain parameters of the snow accumulation body before the avalanche and the physical parameters of the snow accumulation body before the avalanche as known quantities, a preliminary simulation is performed using an avalanche disaster simulation model to obtain a first simulation result; The first simulation result is compared with the actual avalanche result in the avalanche historical monitoring data, and the avalanche disaster simulation model is adjusted according to the comparison result to obtain an avalanche disaster impact prediction model.

[0008] Optionally, the step of establishing an avalanche disaster road impact prediction model based on the avalanche disaster impact characteristics, road positioning data, and post-avalanche road impact parameters specifically includes: Obtaining the road morphological characteristics of each road section according to the road positioning data; Using an avalanche disaster simulation model to superimpose the avalanche disaster impact characteristics with the road morphology characteristics of each road section, to obtain a second simulation result of the avalanche disaster road impact; The second simulation result is compared with the post-avalanche road impact parameter, and the avalanche disaster simulation model is adjusted according to the comparison result to obtain an avalanche disaster road impact prediction model.

[0009] Optionally, the method for calculating the actual avalanche result in the avalanche historical monitoring data specifically includes: Based on the road influence parameters after the historical avalanche and based on the road positioning data, the influence of the road on the avalanche range is removed to obtain the actual avalanche results in the avalanche history monitoring data.

[0010] Optionally, the step of comparing the first simulation result with the actual avalanche result in the avalanche historical monitoring data, and adjusting the avalanche disaster simulation model according to the comparison result to obtain the avalanche disaster impact prediction model specifically includes: Dividing the avalanche historical monitoring data into training set data and test set data; Training the avalanche disaster simulation model that obtains the first simulation result through the training set data; The trained avalanche disaster simulation model is tested and retrained using the test set data to obtain an avalanche disaster impact prediction model.

[0011] Optionally, the step of comparing the second simulation result with the post-avalanche road impact parameter, adjusting the avalanche disaster simulation model according to the comparison result, and obtaining the avalanche disaster road impact prediction model specifically includes: Training the avalanche disaster simulation model that obtains the second simulation result by using training set data; The trained avalanche disaster simulation model is tested and retrained using the test set data to obtain an avalanche disaster road impact prediction model.

[0012] Optionally, the step of predicting the impact of avalanche disasters on each road section according to the avalanche disaster road impact prediction model specifically includes: Obtaining road structure features according to the road positioning data; Matching the current avalanche disaster impact characteristics with the road morphology characteristics, and judging whether the avalanche disaster will have an impact on the road section according to the matching result; For the road section that will be affected, the current avalanche disaster impact characteristics are further superimposed with the road morphology characteristics to obtain the post-avalanche road impact parameters of the road section.

[0013] A second aspect of an embodiment of the present invention provides a device for evaluating the impact of avalanche disasters on roads based on deep learning, comprising: A data acquisition unit, used to acquire historical avalanche monitoring data of the monitoring area, wherein the historical avalanche monitoring data includes terrain parameters of snow accumulation before historical avalanches, physical parameters of snow accumulation before historical avalanches, road positioning data, and road impact parameters after historical avalanches; A first model building unit is used to build an avalanche disaster impact prediction model according to the pre-avalanche snow body terrain parameters and the pre-avalanche snow body physical parameters, wherein the input of the avalanche disaster impact prediction model is the pre-avalanche snow body terrain parameters and the pre-avalanche snow body physical parameters, and the output of the avalanche disaster impact prediction model is the avalanche disaster impact characteristics corresponding to the avalanche disaster impact range; A second model building unit is used to build an avalanche disaster road impact prediction model according to the avalanche disaster impact characteristics, road positioning data and post-avalanche road impact parameters, wherein the input of the avalanche disaster road impact prediction model is the avalanche disaster impact characteristics and road positioning data, and the output of the avalanche disaster road impact prediction model is the post-avalanche road impact parameters; The data acquisition unit is also used to acquire current snow body topographic parameters, current snow body physical parameters and road positioning data of each road section in the monitoring area; A first model prediction unit, configured to obtain current avalanche disaster impact characteristics according to the avalanche disaster impact prediction model; The second model prediction unit is used to obtain the road positioning data of each road section in the monitoring area, predict the impact of the avalanche disaster on each road section according to the avalanche disaster road impact prediction model, and obtain the post-avalanche road impact parameters of each road section in the monitoring area.

[0014] According to a third aspect of an embodiment of the present invention, there is provided an electronic device, including: One or more processors; a memory; one or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the one or more processors, and the one or more applications are configured to execute the method as described in the first aspect.

[0015] According to a fourth aspect of an embodiment of the present invention, a computer-readable storage medium is provided, in which a program code is stored. The program code can be called by a processor to execute the method described in the first aspect.

[0016] In summary, the present invention provides a method and device for evaluating the impact of avalanche disasters on roads based on deep learning, and combines advanced Internet of Things technology, big data analysis and artificial intelligence technology to achieve comprehensive, real-time evaluation and intelligent management of the impact of avalanche disasters on roads. By collecting key parameters of the avalanche occurrence area in real time, and using a cloud computing platform for large-scale data storage and analysis. At the same time, with the help of deep learning algorithms, the probability of avalanche disasters and the degree of their impact on roads can be accurately predicted and warned, and potential risk factors can be discovered in a timely manner, thereby effectively improving the safety and traffic capacity of roads. It not only significantly improves the level of intelligence in avalanche disaster management, but also provides strong technical support for road traffic safety management and disaster prevention. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0018] Figure 1 A method flow chart of a method for evaluating the impact of avalanche disasters on roads based on deep learning according to an embodiment of the present invention; Figure 2 This is a functional module block diagram of a device for evaluating the impact of avalanche disasters on roads based on deep learning according to an embodiment of the present invention; Figure 3 This is a structural block diagram of an electronic device for executing a method for evaluating the impact of avalanche disasters on roads based on deep learning according to an embodiment of the present application.

[0019] Figure 4 The present invention is a structural block diagram of a computer-readable storage medium for storing or carrying a program code for implementing the abnormal vessel identification method according to an embodiment of the present application.

[0020] Reference numerals: Data acquisition unit 110; first model building unit 120; second model building unit 130; first model prediction unit 140; second model prediction unit 150; electronic device 300; processor 310; memory 320; computer readable storage medium 400; program code 410. DETAILED DESCRIPTION

[0021] Avalanches are one of the major natural disasters in the mountainous areas of western China. They are characterized by suddenness, wide range of disasters, and strong destructiveness. As climate change and human activities continue to expand into high-altitude and cold mountainous areas, the population and infrastructure exposed to the danger of avalanches are increasing, seriously threatening the normal operation of transportation lines, power transmission lines, pipelines and human activities in high-altitude and cold mountainous areas. Avalanche disaster prevention and control is one of the current focuses of disaster management in the western mountainous areas, and it is also an important issue that needs to be urgently addressed in the current development and construction of the western mountainous areas.

[0022] In traditional assessment work, avalanche disaster risk assessment relies heavily on the experience and judgment of monitoring personnel, and predictions are also carried out through expert experience methods and mechanism models, but the accuracy of avalanche forecasts is low and the results are unstable. Different terrains, snowfall, slopes, wind speeds and temperatures also have significant effects on avalanche formation. Achieving a scientific assessment of the impact of avalanche disasters on roads and providing accurate and detailed risk prediction results are not only of great significance to the construction and operation of engineering projects and the safety of life and property, but also have significant social and economic benefits.

[0023] Therefore, how to conduct a safety risk assessment of roads posed by avalanche hazards without relying on researchers' experience is an urgent problem that needs to be solved.

[0024] In view of this, the designers of the present invention designed a method and device for evaluating the impact of avalanche disasters on roads based on deep learning.

[0025] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0026] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0027] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0028] In the description of the present invention, it should be noted that the terms "top", "bottom", "inside", "outside", etc. indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, or the orientation or positional relationship in which the invented product is usually placed when in use, which is only for the convenience of describing the present invention and simplifying the description, and does not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", etc. are only used to distinguish the description, and cannot be understood as indicating or implying relative importance.

[0029] In the description of the present invention, it is also necessary to explain that, unless otherwise clearly specified and limited, the terms "set", "install", "connect", and "connect" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0030] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other.

[0031] The following is a detailed description of the deep learning-based method for evaluating the impact of avalanche disasters on roads provided in this embodiment.

[0032] See also Figure 1 , the present embodiment provides a method for evaluating the impact of avalanche disasters on roads based on deep learning, the method comprising: Step S101, obtaining historical avalanche monitoring data of the monitoring area, wherein the historical avalanche monitoring data includes snow body terrain parameters before historical avalanches, snow accumulation physical parameters before historical avalanches, road positioning data, and road impact parameters after historical avalanches.

[0033] The historical avalanche monitoring data is the relevant monitoring data before and after the avalanches occurred in the monitoring area. Among them, the terrain parameters of the snow body before the historical avalanche are the terrain parameters of the high snow body on the slope at the corresponding position before the avalanche, including height, slope, distribution of avalanche gullies, etc.; the physical parameters of the snow body before the historical avalanche are the relevant parameters of the snow accumulation at the corresponding position before the avalanche, including snowfall, snow thickness, snow density, snow layer moisture content and other parameters that can reflect the state of snow accumulation; the road positioning data are the specific parameters of different sections in the monitoring area, including road location coordinates, road length, road width, and road direction; the road impact parameters after the historical avalanche are the relevant parameters that reflect the impact of the road after the avalanche, including the affected range of the road after the avalanche and the thickness of the snow accumulation on the road after the avalanche.

[0034] By collecting historical avalanche monitoring data in the monitoring area and performing a certain degree of preprocessing, it is possible to provide data support for the subsequent establishment of a deep learning-based prediction model.

[0035] Step S102, establishing an avalanche disaster impact prediction model based on the terrain parameters of the snow body before the avalanche and the physical parameters of the snow body before the avalanche, wherein the input of the avalanche disaster impact prediction model is the terrain parameters of the snow body before the avalanche and the physical parameters of the snow body before the avalanche, and the output of the avalanche disaster impact prediction model is the avalanche disaster impact characteristics corresponding to the avalanche disaster impact range.

[0036] The inventors found in the application of actual scenarios that for areas where avalanches may occur, the terrain of high snow bodies on the slopes is usually relatively fixed, but the shape and position of the roads in the area may change over time, such as the renovation and maintenance of the original roads, road widening, and the addition of new roads. Therefore, when predicting the impact of avalanche disasters on roads by establishing a model, the snow body terrain parameters can be regarded as fixed parameters, and the road positioning data can be regarded as dynamic variables. In order to improve the accuracy and practicality of the prediction model, in the scheme of the present invention, the prediction of avalanches and the prediction of the impact of avalanches on roads are divided into two stages. The first stage is the prediction of avalanches, predicting the occurrence of avalanches. The second stage predicts the impact of avalanche disasters on roads.

[0037] As a preferred embodiment of the present invention, step S102 specifically includes: Taking the terrain parameters of the snow accumulation body before the avalanche and the physical parameters of the snow accumulation body before the avalanche as known quantities, a preliminary simulation is performed using an avalanche disaster simulation model to obtain a first simulation result; The first simulation result is compared with the actual avalanche result in the avalanche historical monitoring data, and the avalanche disaster simulation model is adjusted according to the comparison result to obtain an avalanche disaster impact prediction model.

[0038] Since dynamic road positioning data will not affect the probability and state of avalanche occurrence, and considering that the snow mass terrain parameters are regarded as fixed parameters, a relatively accurate prediction model for avalanche occurrence can be established based on the obtained historical snow mass terrain parameters before avalanches and historical snow mass physical parameters before avalanches, namely, the avalanche disaster impact prediction model. The input of this model is the snow mass terrain parameters before avalanches and the snow mass physical parameters before avalanches. Through the snow mass physical parameters before avalanches during each avalanche, the avalanche effect can be predicted.

[0039] It should be noted that the historical avalanche monitoring data obtained does not contain direct data on the impact of avalanches. Instead, it records the impact parameters of the road after the avalanche occurs. Therefore, it is necessary to preprocess the historical avalanche monitoring data to obtain direct data on the impact of avalanches after the avalanche occurs. The specific calculation method is as follows.

[0040] As a preferred embodiment of the present invention, the method for calculating the actual avalanche result in the avalanche history monitoring data specifically includes: Based on the road influence parameters after the historical avalanche and based on the road positioning data, the influence of the road on the avalanche range is removed to obtain the actual avalanche results in the avalanche history monitoring data.

[0041] The road impact parameters after historical avalanches are the data of the affected roads after an avalanche occurs. When the road location data is known, by removing the influence of the road part on the avalanche range based on the road impact parameters after historical avalanches, the direct data of the avalanche impact, that is, the actual avalanche results in the avalanche historical monitoring data, can be obtained.

[0042] The specific calculation process needs to rely on a basic avalanche disaster simulation model.

[0043] In an embodiment of the present invention, the avalanche disaster simulation model can adopt the currently publicly used model, such as the RAMMS (Rapid Mass Movement Simulation) model. RAMMS is a widely used avalanche simulation tool that can simulate the entire process of avalanche from initiation to deposition. For road impact assessment, RAMMS can be used to predict the range, speed and volume of avalanche coverage after deposition, so as to estimate the possibility and degree of road blockage. In this basic model, the terrain parameters of the snow body before the historical avalanche, the physical parameters of the snow body before the historical avalanche, the road positioning data and the road impact parameters after the historical avalanche are all input as known quantities, and the actual avalanche results without considering the road can be obtained.

[0044] After obtaining relevant data on actual avalanche results through the basic avalanche disaster simulation model, the avalanche disaster impact prediction model can be established on the basis of the basic avalanche disaster simulation model. Through the historical avalanche monitoring data obtained, the mapping relationship between the terrain parameters of the snow body before the historical avalanche, the physical parameters of the snow body before the historical avalanche, and the road positioning data and the avalanche disaster impact characteristics is established. Then, through the deep learning algorithm, the avalanche disaster impact prediction model is optimized based on the large amount of historical avalanche monitoring data obtained.

[0045] It should be noted that for the selection of basic avalanche disaster simulation models, existing mature models that have been made public, such as RAMMS, can be used. It can also be achieved by establishing a rule-based model (RBR). Through predefined rules, the probability and impact of avalanches under different conditions can be evaluated. Specifically, when using the RBR model, the model input is the terrain parameters of the high snow body on the slope (including: the height H of the high snow body on the slope, the slope P, the distribution form of the avalanche gully F), the physical parameters of the snow body (including: snowfall Q, snow thickness T, snow density D, snow layer moisture content M) and road positioning data (road location coordinates C, road length L, road width W, road direction Z), and the output of the model is the impact characteristics of avalanche disasters (impact range S, snow accumulation thickness A). In the calculation process, the boundary conditions of the avalanche numerical simulation process are set to perform numerical analysis and calculation, and the impact range of different avalanches is calculated in combination with the avalanche numerical calculation method.

[0046] By preprocessing the historical avalanche monitoring data, a data set is formed for input into the machine learning model for training and verification, and then the basic avalanche disaster simulation model used is trained. The specific model optimization method is as follows: As a preferred embodiment of the present invention, the step of comparing the first simulation result with the actual avalanche result in the avalanche historical monitoring data, adjusting the avalanche disaster simulation model according to the comparison result, and obtaining the avalanche disaster impact prediction model specifically includes: Dividing the avalanche historical monitoring data into training set data and test set data; Training the avalanche disaster simulation model that obtains the first simulation result through the training set data; The trained avalanche disaster simulation model is tested and retrained using the test set data to obtain an avalanche disaster impact prediction model.

[0047] After continuous training, verification and testing, we obtained an avalanche disaster impact prediction model that can predict the impact characteristics of avalanche disasters based on snow mass terrain parameters, snow mass physical parameters and road positioning data.

[0048] It should be noted that in this solution, there is no specific restriction on the deep learning algorithm used for the training model, and any algorithm that is applicable to the current scenario can be used. In this embodiment, as a preferred implementation, the support vector machine (SVM) algorithm is used to perform regression prediction on the output avalanche disaster impact characteristics (impact range S, snow accumulation thickness A); the loss function uses the Huber loss function, and the formula is as follows: ; Where f(x) is the model prediction value, y is the true value of the sample, and δ is a hyperparameter. The value of δ is the focus of the Huber loss function. When |y−f(x)|≤δ, it is the mean square error loss function; when |y−f(x)|>δ, it is the mean absolute error loss function. By setting the hyperparameter δ, the sensitivity to outliers is reduced.

[0049] Step S103, establishing an avalanche disaster road impact prediction model based on the avalanche disaster impact characteristics, road positioning data and post-avalanche road impact parameters, wherein the input of the avalanche disaster road impact prediction model is the avalanche disaster impact characteristics and road positioning data, and the output of the avalanche disaster road impact prediction model is the post-avalanche road impact parameters.

[0050] After completing the establishment of the avalanche disaster impact prediction model, it is necessary to establish a second-stage prediction model based on the prediction results of the impact of avalanche disasters on roads, namely the avalanche disaster road impact prediction model. The input of this model is the avalanche disaster impact characteristics and road positioning data, and the output is the post-avalanche road impact parameters, that is, the impact of the avalanche on the road after it occurs.

[0051] In the above theoretical analysis, it is mentioned that the road positioning data changes accordingly based on the actual situation of the road section in the monitoring area. This model establishes the mapping relationship between the impact characteristics of avalanche disasters and the road positioning data and the road impact parameters after avalanches. The establishment of the model also relies on the basic avalanche disaster simulation model. The specific model selection is the same as the avalanche disaster impact prediction model, which will not be repeated here.

[0052] Specifically, the established avalanche disaster road impact prediction model takes as input the avalanche disaster impact characteristics (impact range S, snow accumulation thickness A), road positioning data (road position coordinates Cn, road length Ln, road width Wn, road direction Zn), and outputs the post-avalanche road impact parameters (road impact range Sn, snow accumulation thickness An), where n is 1~i; i is the number of input road sections.

[0053] The training method of the avalanche hazard road impact prediction model is as follows. Similar to the process of establishing an avalanche hazard impact prediction model, it is trained by using a dataset that has been preprocessed with historical avalanche monitoring data.

[0054] As a preferred embodiment of the present invention, step S103 specifically includes: Obtaining the road morphological characteristics of each road section according to the road positioning data; Using an avalanche disaster simulation model to superimpose the avalanche disaster impact characteristics with the road morphology characteristics of each road section, to obtain a second simulation result of the avalanche disaster road impact; The second simulation result is compared with the post-avalanche road impact parameter, and the avalanche disaster simulation model is adjusted according to the comparison result to obtain an avalanche disaster road impact prediction model.

[0055] As a preferred embodiment of the present invention, the step of comparing the second simulation result with the post-avalanche road impact parameter, adjusting the avalanche disaster simulation model according to the comparison result, and obtaining the avalanche disaster road impact prediction model specifically includes: Training the avalanche disaster simulation model that obtains the second simulation result by using training set data; The trained avalanche disaster simulation model is tested and retrained using the test set data to obtain an avalanche disaster road impact prediction model.

[0056] It should be noted that when training the avalanche disaster road impact prediction model, the deep learning algorithm used can be the same as the avalanche disaster impact prediction model trained above, or a different deep learning algorithm can be used. In actual use, the appropriate basic model and deep learning algorithm can be selected according to the collected data.

[0057] Through the modeling of the above two steps, the model establishment step for predicting the impact of avalanche disasters on roads is completed, and the actual impact of avalanche disasters on roads can be predicted through the established model.

[0058] Step S104, obtaining the current snow body topographic parameters, current snow body physical parameters and road positioning data of each road section in the monitoring area.

[0059] When a prediction is needed, various data at the current time point are obtained through drone photography and avalanche monitoring equipment.

[0060] Step S105, obtaining the current avalanche disaster impact characteristics according to the avalanche disaster impact prediction model.

[0061] Firstly, the avalanche disaster impact prediction model is used to obtain the current avalanche disaster impact characteristics that can directly reflect the impact of avalanche disasters. This result can more intuitively see the scope and extent of the current avalanche disaster impact.

[0062] Step S106, predicting the impact of avalanche disasters on each road section according to the avalanche disaster road impact prediction model, and obtaining post-avalanche road impact parameters of each road section in the monitoring area.

[0063] On this basis, the road location data of the specific road sections that need to be predicted are brought in, and the relevant parameters of the road parameters that may be affected by the avalanche can be obtained through the model. The relevant staff can issue corresponding avalanche disaster risk warning notices to the managers of the corresponding road sections according to different prediction results.

[0064] It should be noted that when predicting the impact of avalanche disasters on each road section through the avalanche disaster road impact prediction model, if there are multiple road sections within the range, the secondary impact caused by the avalanche affecting the road must be considered.

[0065] Therefore, as a preferred implementation, step S106 specifically includes: Obtaining road structure features according to the road positioning data; Matching the current avalanche disaster impact characteristics with the road morphology characteristics, and judging whether the avalanche disaster will have an impact on the road section according to the matching result; For the road section that will be affected, the current avalanche disaster impact characteristics are further superimposed with the road morphology characteristics to obtain the post-avalanche road impact parameters of the road section.

[0066] When matching the current avalanche disaster impact characteristics with the road morphology characteristics, first determine which areas of the road will be affected by the avalanche disaster, namely Sn, by matching the impact range S in the avalanche disaster impact characteristics with the road positioning data of a certain section (road position coordinates C, road length L, road width W, road direction Z). Then calculate the snow accumulation thickness An in the affected area based on the affected area of ​​the road and the snow accumulation thickness A in the avalanche disaster impact characteristics.

[0067] When there are multiple sections that need to be predicted at the same time, according to the diffusion direction of the impact range S, after calculating the post-avalanche road impact parameters (S1, A1) of the first affected section, it is necessary to consider the superposition of the road morphological characteristics of the first section on the basis of the avalanche disaster impact characteristics (S, A), that is, the impact of the first section on the avalanche disaster impact characteristics after being affected by the avalanche. At this time, the avalanche disaster impact characteristics are (S (1), A (1)), S (1) and A (1) are the impact range and snow accumulation thickness after being affected by the road morphological characteristics of the first section, respectively. The values ​​of S (1) and A (1) are related to the post-avalanche road impact parameters (S1, A1), and the specific correlation factors are related to the road morphological characteristics of the first section, that is, the road positioning data of the first section and the road morphological characteristics. Some road morphologies will have an incremental impact on the avalanche disaster impact characteristics (S, A), while others will have a decremental impact. After obtaining (S(1), A(1)), when predicting the second road section, use (S(1), A(1)) instead of (S, A) as the avalanche disaster impact feature to calculate the post-avalanche road impact parameter (S2, A2) of the first affected road section. After the calculation is completed, (S(2), A(2) is obtained based on (S2, A2), and so on, until the post-avalanche road impact parameter calculation of all road sections is completed.

[0068] In summary, the deep learning-based assessment method for the impact of avalanche disasters on roads provided by the present invention, combined with advanced Internet of Things technology, big data analysis and artificial intelligence technology, realizes a comprehensive, real-time assessment and intelligent management of the impact of avalanche disasters on roads. By deploying a high-precision sensor network, the system can collect key parameters of the avalanche occurrence area in real time, such as snow thickness, temperature, humidity, wind speed, etc., and use cloud computing platforms for large-scale data storage and analysis. At the same time, with the help of deep learning algorithms, the system can accurately predict and warn the probability of avalanche disasters and the degree of their impact on roads, and promptly discover potential risk factors, thereby effectively improving the safety and traffic capacity of roads. This method not only significantly improves the level of intelligence in avalanche disaster management, but also provides strong technical support for road traffic safety management and disaster prevention.

[0069] like Figure 2 As shown, the present invention implements a device for evaluating the impact of avalanche disasters on roads based on deep learning, and the system includes: The data acquisition unit 110 is used to acquire historical avalanche monitoring data of the monitoring area, wherein the historical avalanche monitoring data includes terrain parameters of snow accumulation before historical avalanches, physical parameters of snow accumulation before historical avalanches, road positioning data, and road impact parameters after historical avalanches; The first model building unit 120 is used to establish an avalanche disaster impact prediction model based on the pre-avalanche snow body terrain parameters and the pre-avalanche snow body physical parameters, wherein the input of the avalanche disaster impact prediction model is the pre-avalanche snow body terrain parameters and the pre-avalanche snow body physical parameters, and the output of the avalanche disaster impact prediction model is the avalanche disaster impact characteristics corresponding to the avalanche disaster impact range; A second model building unit 130 is used to build an avalanche disaster road impact prediction model according to the avalanche disaster impact characteristics, the road positioning data and the post-avalanche road impact parameters, wherein the input of the avalanche disaster road impact prediction model is the avalanche disaster impact characteristics and the road positioning data, and the output of the avalanche disaster road impact prediction model is the post-avalanche road impact parameters; The data acquisition unit 110 is also used to acquire the current snow body topographic parameters, the current snow body physical parameters and the road positioning data of each road section in the monitoring area; A first model prediction unit 140 is used to obtain the current avalanche disaster impact characteristics according to the avalanche disaster impact prediction model; The second model prediction unit 150 is used to obtain the road positioning data of each road section in the monitoring area, predict the impact of the avalanche disaster on each road section according to the avalanche disaster road impact prediction model, and obtain the post-avalanche road impact parameters of each road section in the monitoring area.

[0070] The device for evaluating the impact of avalanche disasters on roads based on deep learning provided in an embodiment of the present invention is used to implement the above-mentioned method for evaluating the impact of avalanche disasters on roads based on deep learning. Therefore, the specific implementation method is the same as the above-mentioned method and will not be repeated here.

[0071] like Figure 3 As shown, a structural block diagram of an electronic device 300 provided in an embodiment of the present invention. The electronic device 300 may be an electronic device 300 capable of running applications, such as a smart phone, a tablet computer, an e-book, etc. The electronic device 300 in the present application may include one or more of the following components: a processor 310, a memory 320, and one or more applications, wherein the one or more applications may be stored in the memory 320 and configured to be executed by one or more processors 310, and one or more programs are configured to execute the method described in the aforementioned method embodiment.

[0072] The processor 310 may include one or more processing cores. The processor 310 uses various interfaces and lines to connect various parts of the entire electronic device 300, and executes various functions and processes data of the electronic device 300 by running or executing instructions, programs, code sets or instruction sets stored in the memory 320, and calling data stored in the memory 320. Optionally, the processor 310 can be implemented in at least one hardware form of digital signal processing (DSP), field-programmable gate array (FPGA), and programmable logic array (PLA). The processor 310 can integrate one or a combination of a central processing unit (CPU), a graphics processing unit (GPU), and a modem. Among them, the CPU mainly processes the operating system, user interface, and application programs; the GPU is responsible for rendering and drawing display content; and the modem is used to process wireless communications. It can be understood that the above-mentioned modem may not be integrated into the processor 310, but may be implemented separately through a communication chip.

[0073] The memory 320 may include a random access memory (RAM) or a read-only memory (ROM). The memory 320 may be used to store instructions, programs, codes, code sets or instruction sets. The memory 320 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for implementing at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the following various method embodiments, etc. The data storage area may also store data created by the terminal during use (such as a phone book, audio and video data, chat record data), etc.

[0074] like Figure 4 FIG. 4 is a block diagram of a computer-readable storage medium 400 provided in an embodiment of the present invention. The computer-readable storage medium stores program code 410, which can be called by a processor to execute the method described in the above method embodiment.

[0075] The computer-readable storage medium 400 may be an electronic memory such as a flash memory, an EEPROM (electrically erasable programmable read-only memory), an EPROM, a hard disk, or a ROM. Optionally, the computer-readable storage medium 400 includes a non-transitory computer-readable storage medium. The computer-readable storage medium 400 has storage space for program codes 410 for executing any method steps in the above method. These program codes 410 may be read from or written to one or more computer program products. The program codes 410 may be compressed, for example, in a suitable form.

[0076] In summary, the present invention provides a method and device for evaluating the impact of avalanche disasters on roads based on deep learning. It can collect key parameters of the avalanche occurrence area in real time, and with the help of deep learning algorithms, accurately predict and warn the probability of avalanche disasters and the degree of their impact on roads, thereby effectively improving the safety and traffic capacity of roads.

[0077] In several embodiments disclosed in the present application, it should be understood that the disclosed devices and methods can also be implemented in other ways. The device embodiments described above are merely schematic. For example, the flowcharts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the devices, methods and computer program products according to multiple embodiments of the present application. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of the code, and a module, a program segment or a part of the code contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of boxes in the block diagram and / or flowchart can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0078] In addition, the functional modules in the various embodiments of the present application may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part.

[0079] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

Claims

1. A method for evaluating the impact of avalanche disasters on roads based on deep learning, characterized in that: The method comprises: Acquire historical avalanche monitoring data of the monitoring area, wherein the historical avalanche monitoring data includes terrain parameters of snow accumulation before historical avalanches, physical parameters of snow accumulation before historical avalanches, road positioning data, and road impact parameters after historical avalanches; Establishing an avalanche disaster impact prediction model based on the terrain parameters of the snow body before the avalanche and the physical parameters of the snow body before the avalanche, wherein the input of the avalanche disaster impact prediction model is the terrain parameters of the snow body before the avalanche and the physical parameters of the snow body before the avalanche, and the output of the avalanche disaster impact prediction model is the avalanche disaster impact characteristics corresponding to the avalanche disaster impact range; Establishing an avalanche disaster road impact prediction model according to the avalanche disaster impact characteristics, road positioning data and post-avalanche road impact parameters, wherein the input of the avalanche disaster road impact prediction model is the avalanche disaster impact characteristics and road positioning data, and the output of the avalanche disaster road impact prediction model is the post-avalanche road impact parameters; Obtain the current snow mass topographic parameters, current snow mass physical parameters and road positioning data of each road section in the monitoring area; Obtaining the current avalanche disaster impact characteristics according to the avalanche disaster impact prediction model; The impact of avalanche disasters on each road section is predicted according to the avalanche disaster road impact prediction model to obtain the post-avalanche road impact parameters of each road section in the monitoring area.

2. The method for evaluating the impact of avalanche disasters on roads based on deep learning according to claim 1, characterized in that: The step of establishing an avalanche disaster impact prediction model based on the terrain parameters of the snow body before the avalanche and the physical parameters of the snow body before the avalanche specifically includes: Taking the terrain parameters of the snow accumulation body before the avalanche and the physical parameters of the snow accumulation body before the avalanche as known quantities, a preliminary simulation is performed using an avalanche disaster simulation model to obtain a first simulation result; The first simulation result is compared with the actual avalanche result in the avalanche historical monitoring data, and the avalanche disaster simulation model is adjusted according to the comparison result to obtain an avalanche disaster impact prediction model.

3. The method for evaluating the impact of avalanche disasters on roads based on deep learning according to claim 2, characterized in that: The step of establishing an avalanche disaster road impact prediction model based on the avalanche disaster impact characteristics, road positioning data, and post-avalanche road impact parameters specifically includes: Obtaining the road morphological characteristics of each road section according to the road positioning data; Using an avalanche disaster simulation model to superimpose the avalanche disaster impact characteristics with the road morphology characteristics of each road section, to obtain a second simulation result of the avalanche disaster road impact; The second simulation result is compared with the post-avalanche road impact parameter, and the avalanche disaster simulation model is adjusted according to the comparison result to obtain an avalanche disaster road impact prediction model.

4. The method for evaluating the impact of avalanche disasters on roads based on deep learning according to claim 2 or 3, characterized in that: The method for calculating the actual avalanche results in the avalanche historical monitoring data specifically includes: Based on the road influence parameters after the historical avalanche and based on the road positioning data, the influence of the road on the avalanche range is removed to obtain the actual avalanche results in the avalanche history monitoring data.

5. The method for evaluating the impact of avalanche disasters on roads based on deep learning according to claim 4, characterized in that: The step of comparing the first simulation result with the actual avalanche result in the avalanche historical monitoring data, and adjusting the avalanche disaster simulation model according to the comparison result to obtain the avalanche disaster impact prediction model specifically includes: Dividing the avalanche historical monitoring data into training set data and test set data; Training the avalanche disaster simulation model that obtains the first simulation result through the training set data; The trained avalanche disaster simulation model is tested and retrained using the test set data to obtain an avalanche disaster impact prediction model.

6. The method for evaluating the impact of avalanche disasters on roads based on deep learning according to claim 5, characterized in that: The step of comparing the second simulation result with the post-avalanche road impact parameter, and adjusting the avalanche disaster simulation model according to the comparison result to obtain the avalanche disaster road impact prediction model specifically includes: Training the avalanche disaster simulation model that obtains the second simulation result by using training set data; The trained avalanche disaster simulation model is tested and retrained using the test set data to obtain an avalanche disaster road impact prediction model.

7. The method for evaluating the impact of avalanche disasters on roads based on deep learning according to claim 4, characterized in that: The step of predicting the impact of avalanche disasters on each road section according to the avalanche disaster road impact prediction model specifically includes: Obtaining road structure features according to the road positioning data; Matching the current avalanche disaster impact characteristics with the road morphology characteristics, and judging whether the avalanche disaster will affect the road section according to the matching result; For the road section that will be affected, the current avalanche disaster impact characteristics are further superimposed with the road morphology characteristics to obtain the post-avalanche road impact parameters of the road section.

8. A device for evaluating the impact of avalanche disasters on roads based on deep learning, characterized in that: include: A data acquisition unit, used to acquire historical avalanche monitoring data of the monitoring area, wherein the historical avalanche monitoring data includes terrain parameters of snow accumulation before historical avalanches, physical parameters of snow accumulation before historical avalanches, road positioning data, and road impact parameters after historical avalanches; A first model building unit is used to build an avalanche disaster impact prediction model according to the pre-avalanche snow body terrain parameters and the pre-avalanche snow body physical parameters, wherein the input of the avalanche disaster impact prediction model is the pre-avalanche snow body terrain parameters and the pre-avalanche snow body physical parameters, and the output of the avalanche disaster impact prediction model is the avalanche disaster impact characteristics corresponding to the avalanche disaster impact range; A second model building unit is used to build an avalanche disaster road impact prediction model according to the avalanche disaster impact characteristics, road positioning data and post-avalanche road impact parameters, wherein the input of the avalanche disaster road impact prediction model is the avalanche disaster impact characteristics and road positioning data, and the output of the avalanche disaster road impact prediction model is the post-avalanche road impact parameters; The data acquisition unit is also used to acquire current snow terrain parameters, current snow accumulation physical parameters and road positioning data of each road section in the monitoring area; A first model prediction unit, configured to obtain current avalanche disaster impact characteristics according to the avalanche disaster impact prediction model; The second model prediction unit is used to obtain the road positioning data of each road section in the monitoring area, predict the impact of the avalanche disaster on each road section according to the avalanche disaster road impact prediction model, and obtain the post-avalanche road impact parameters of each road section in the monitoring area.

9. An electronic device, characterized in that: include: one or more processors; Memory; One or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the one or more processors, and the one or more applications are configured to execute the method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores program codes, which can be called by a processor to execute the method according to any one of claims 1 to 7.

Images