Intelligent grading early warning system for outdoor safety accident risk

Abstract

The invention relates to the technical field of data processing, in particular to an intelligent grading early warning system for outdoor safety accident risks, which comprises the following steps: acquiring all natural disaster point locations and travel routes; obtaining an anti-risk factor of the outdoor team; acquiring the environmental convergence of the same kind of disaster point locations; obtaining the disaster type convergence degree of the same kind of disaster point locations; obtaining a plurality of clusters and cluster areas thereof; obtaining the disaster occurrence probability of each cluster; obtaining the external factor influence degree of each disaster type; acquiring a cluster to which each marching point belongs in the travel route of the outdoor team, and obtaining a risk assessment value of the travel plan; and screening an optimal travel plan according to the risk assessment value. According to the method, risk factor mining is carried out by utilizing historical data, so that the risk assessment result of the real-time travel plan is more credible, the travel plan can be adjusted by utilizing the risk assessment model, and the practicability and the flexibility are higher.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and particularly to an intelligent hierarchical early warning system for outdoor safety accident risks. Background Art

[0002] An intelligent hierarchical early warning system for outdoor safety accident risks is a system that comprehensively utilizes data analysis, sensing technology, artificial intelligence, and big data processing to identify and evaluate potential safety risks in outdoor activities and provide hierarchical early warning information to help people take appropriate measures to reduce risks. Such a system can provide useful information and suggestions in various outdoor activities, such as mountain climbing, hiking, camping, water activities, etc., to ensure the safety of participants.

[0003] Compared with real-time risk early warning, the risk early warning for travel plans is equally crucial, which can significantly reduce the probability of accidents for outdoor teams before traveling. However, due to the complex data dimensions of natural disaster influencing factors, the existing accident risk assessment and early warning methods for outdoor travel plans are relatively rough, lacking specific and scientific accident risk data analysis methods, and unable to meet the accuracy requirements of outdoor risk early warning and risk avoidance. Summary of the Invention

[0004] The present invention provides an intelligent hierarchical early warning system for outdoor safety accident risks to solve the problems of complex data dimensions of natural disaster influencing factors and poor accuracy of existing outdoor travel plan risk assessment and early warning models.

[0005] The following technical solutions are adopted for an intelligent hierarchical early warning system for outdoor safety accident risks of the present invention: An embodiment of the present invention provides an intelligent hierarchical early warning system for outdoor safety accident risks, and the system includes: Data acquisition and preprocessing module: statistically obtain all natural disaster points and their disaster types, disaster levels, and meteorological monitoring data; obtain the travel route of the outdoor team and several camping points therein; obtain the risk resistance factors of the outdoor team; Risk influencing factor analysis module: take natural disaster points with the same disaster type as the same type of disaster points; obtain the environmental convergence of the same type of disaster points according to the distance between the same type of disaster points and the similarity of meteorological monitoring data; adjust the environmental convergence according to the difference in the disaster levels of the same type of disaster points to obtain the disaster type convergence degree of the same type of disaster points; cluster all natural disaster points using the disaster type convergence degree to obtain several clusters and their cluster regions; analyze the density of natural disaster points in each cluster to obtain the disaster occurrence probability of each cluster; analyze the influence of the meteorological monitoring data of natural disaster points in each cluster under the same disaster type on the cluster region of the cluster to obtain the external factor influence degree of each disaster type; Risk assessment module: Obtain the cluster to which each camping point belongs in the travel route of the outdoor team, and based on the disaster occurrence probability of the cluster to which it belongs and the influence degree of external factors of the disaster type corresponding to the cluster to which it belongs, combined with the risk resistance factor of the outdoor team, obtain the risk assessment value of the travel plan; Screen the optimal travel plan according to the risk assessment value.

[0006] Further, the specific steps for obtaining the risk resistance factor include: By collecting the physical health data of each traveler in the outdoor team and the skill tags of each traveler, using a neural network to evaluate the physical health data of all travelers in the outdoor team, obtaining the physical health index of each traveler in the outdoor team, and then combining the skill tags included in the outdoor team to obtain the risk resistance factor of the outdoor team.

[0007] Further, the specific steps for obtaining the environmental convergence include: Project all natural disaster points onto a topographic map of a two-dimensional plane to obtain a projection map; For any two similar disaster points a and b in the projection map, the calculation method of the environmental convergence of the similar disaster points a and b is: ; Among them, represents the mean square error of all meteorological type data in the meteorological monitoring data of the two similar disaster points a and b, represents the spatial distance of the similar disaster points a and b on the projection map, represents the environmental convergence between the similar disaster points a and b.

[0008] Further, the specific steps for obtaining the disaster type similarity degree include: The degree of disaster difference between the similar disaster points a and b is calculated as: ; Among them, represents the disaster level of the a-th natural disaster point, represents the disaster level of the b-th natural disaster point, represents the maximum disaster level in the disaster type to which the similar disaster points a and b belong on the projection map, represents the minimum disaster level in the disaster type to which the similar disaster points a and b belong on the projection map; Multiply the environmental convergence of the similar disaster points a and b by the degree of disaster difference to obtain the disaster type similarity degree of the similar disaster points a and b.

[0009] Further, the specific steps for obtaining several clusters and their cluster regions include: Using the disaster type similarity degree between any two disaster points of the same type to perform k-means clustering on all disaster points of the same type, several clusters are obtained. Each cluster contains several disaster points of the same type. The convex hull area of the natural disaster points in each cluster is denoted as the cluster area of each cluster.

[0010] Furthermore, the specific steps for obtaining the disaster occurrence probability of the cluster include: Denote the disaster types of all natural disaster points in each cluster as the disaster type of each cluster; For the v-th cluster, the disaster occurrence probability of the v-th cluster is calculated as: ; where represents the number of natural disaster points in the v-th cluster, represents the area of the cluster area of the v-th cluster.

[0011] Furthermore, the specific steps for obtaining the influence degree of external factors of the disaster type include: Obtain the influence degree of external factors of the o-th disaster type: ; where is the variance of the number of natural disaster points in all clusters belonging to the o-th disaster type, is the meteorological influence factor of the o-th disaster type.

[0012] Furthermore, the specific steps for obtaining the meteorological influence factor of the disaster type include: Pairwise combine any two natural disaster points in each cluster to obtain several pairs of disaster points, and calculate the mean value of the mean square error of the meteorological monitoring data in all the pairs of disaster points, which is denoted as the meteorological characteristic parameter of each cluster; Among all clusters, arrange all clusters of the same disaster type in descending order according to the size of the cluster to obtain the cluster sequence of each disaster type. Sort the meteorological characteristic parameters and disaster occurrence probabilities of each cluster in the order of the cluster sequence to obtain the meteorological characteristic parameter sequence and disaster occurrence probability sequence of each disaster type. Denote the absolute value of the Pearson correlation coefficient of the meteorological characteristic parameter sequence and disaster occurrence probability sequence of each disaster type as the meteorological influence factor of each disaster type.

[0013] Furthermore, the specific steps for obtaining the risk assessment value of the travel plan include: Obtain the risk assessment value of the travel plan. Denote the difference between the risk assessment coefficient of the travel route of the outdoor team and the risk resistance factor of the outdoor team as the risk assessment value of the travel plan.

[0014] Furthermore, the specific steps for obtaining the risk assessment value of the travel plan include: Obtain the camping time and meteorological monitoring data of each camping site; Map the travel route of the outdoor team onto a projection map, and obtain the cluster to which each camping site in the travel route of the outdoor travel belongs in the projection map, denoted as the camping cluster of each camping site; Denote the camping time of the outdoor team at the p-th camping site as ; Denote the influence degree of external factors of the disaster type to which the n-th camping cluster of the p-th camping site belongs as ; Denote the disaster occurrence probability of the n-th camping cluster of the p-th camping site as ; Obtain the risk assessment coefficient of the travel route of the outdoor team : ; where L represents the number of camping sites in the travel route of the outdoor team, represents the number of camping clusters of the p-th camping site in the travel route of the outdoor team; is the sequence composed of all meteorological type data in the meteorological monitoring data of the p-th camping site in the travel route of the outdoor team; is the sequence composed of all meteorological type data in the meteorological monitoring data of the natural disaster point closest to the p-th camping site in the travel route of the outdoor team; represents the calculation of the mean square error function; is the hyperbolic tangent function.

[0015] The beneficial effects of the technical solution of the present invention are as follows: The present invention conducts risk early warning for camping and outdoor activity plans. First, natural disaster points are obtained by dividing geological regions and projecting historical disaster record data, and then a disaster type similarity model is constructed to obtain the clustering results of similar disaster points, and the disaster occurrence probability of each cluster region and the degree of external influence on the cluster region are obtained; then, according to the risk resistance ability of the participants, combined with the difference between the weather conditions at the path stopping points of the participants and the weather influencing factors of natural disasters in the regions where the stopping points are located, a risk early warning model for the travel plan is constructed to evaluate the risk of this travel plan and improve the safety of outdoor travel plans. Compared with the traditional travel plan risk model, the present invention uses historical data to mine risk factors, so that the risk assessment results for real-time travel plans are more credible, and the risk assessment model of the present invention can be used to adjust the travel plan, with higher practicality and flexibility. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0017] Figure 1 It is a structural block diagram of an intelligent hierarchical early warning system for outdoor safety accident risks of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the intended invention purpose, the following, in combination with the accompanying drawings and preferred embodiments, details the specific implementation manners, structures, features, and effects of an intelligent hierarchical early warning system for outdoor safety accident risks proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0019] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which the present invention belongs.

[0020] The following specifically describes the specific technical solutions of an intelligent hierarchical early warning system for outdoor safety accident risks provided by the present invention with reference to the accompanying drawings.

[0021] Please refer to Figure 1, which shows a structural block diagram of an intelligent classification and early warning system for outdoor safety accident risks provided by an embodiment of the present invention. The system includes: Data acquisition and preprocessing module 101: Obtain the topographic map of the outdoor team's travel area, divide it into multiple geological regions, then obtain the historical disaster records of the local area, and overlay the historical disaster record data on the topographic map to obtain a projection map.

[0022] Divide the outdoor travel topographic map to obtain several geological regions; count all natural disaster points in each geological region, and obtain the disaster type, disaster level, and meteorological monitoring data of each natural disaster point in each geological region. Use satellite images or remote sensing images to obtain the topographic map of the outdoor team's travel area. Since different types of natural disasters occur in different topographical features, it is necessary to artificially divide the topographic map into multiple geological regions and assign classification labels to different geological regions, such as river regions, jungle regions, valley regions, etc., to obtain multiple geological regions.

[0023] Then obtain the natural disaster record data in the past ten years of the local area, and input the natural disaster record data into QGIS (Quantum GIS) software; the QGIS is a free and open-source GIS software with a wide range of plugins and extensions, which can be used to create maps, process geographical data, and overlay natural disaster data on the topographic map. Record the coordinate positions of each natural disaster that occurred in each geological region in the natural disaster record data as a natural disaster point in each geological region, and obtain the disaster type, disaster level, and meteorological monitoring data of each natural disaster point in each geological region from the natural disaster record data, where the meteorological monitoring data includes meteorological types such as rainfall, temperature, wind force, etc. Obtain all the camping points passed by the outdoor team. The connection lines between all the camping points form the travel route of the outdoor team, and obtain the meteorological monitoring data of all the camping points and the camping time at each camping point. It should be noted that the impact of bad weather may increase the risk level of outdoor safety accidents, and people's anti-risk ability to outdoor risks can be manifested in multiple aspects, such as physical health level, relevant outdoor experience and training, etc. Outdoor experience and training can cope with situations that may occur at any time, and a healthy body can provide more energy and stronger resistance, thus enhancing the ability to face risks in the outdoor environment; therefore, in this embodiment, by collecting and integrating the outdoor experience and training of the outdoor team's travel personnel, a skill label is established for each person, such as: rock climbing, swimming, making fire, cardiopulmonary resuscitation, trap making, outdoor cooking, etc. The same labels are of the same type. Then, according to the physical index values of each travel personnel in the outdoor team, a trained neural network is used to evaluate the health level, and then the anti-risk factor of the outdoor team is obtained.

[0024] By collecting the physical health data of each traveler in an outdoor team and the skill tags of each traveler, using a neural network to evaluate the physical health data of all travelers in the outdoor team, obtaining the physical health index of each traveler in the outdoor team, and then combining the skill tags included in the outdoor team to obtain the risk resistance factor of the outdoor team.

[0025] The specific steps are as follows: 1. Use wearable health monitoring devices to obtain the physical health data of each traveler in the outdoor team every day in the past month. The physical health data includes but is not limited to blood sugar, blood pressure, body temperature, heart rate, etc.; 2. Query existing relevant materials to obtain the normal range of each item of physical health data. According to the difference between each item of physical health data and the normal range, obtain the score of each item of physical health data, and use the score as the labeled data set. The score is divided into levels at intervals of 0.1 between 0 and 1. The higher the score, the healthier the traveler. As an example, in this embodiment, taking the body temperature in the physical health data as an example, the normal range of body temperature is between 36 degrees and 37.4 degrees, with 36.7 degrees as the anchor value. For every 0.1 degree difference between the traveler's body temperature and the anchor value, the score is reduced by one level; 3. Use the scores of each item of physical health data to construct a classification network to achieve inputting the physical health data of travelers and outputting the physical health index of travelers; the classification network structure is a 5-layer fully connected neural network. In other embodiments, it can be set as a neural network with other structures, which is not specifically limited in this embodiment. The loss function of the classification network is the cross-entropy function; divide the labeled data set into a training set and a validation set according to a ratio of 7:3, input the training set data into the classification network for training, and use the gradient descent method for training until the loss function converges to complete the training of the classification network; 4. Input the physical health data of each traveler in the outdoor team every day in the past month into the classification network to obtain the physical health index of each traveler in the outdoor team every day.

[0026] Further, the specific steps to obtain the risk resistance factor of the outdoor team are as follows: Statistically obtain the skill tags of each traveler in the outdoor team. The tags indicate the skills familiar to the traveler, such as rock climbing, swimming, making fire, cardiopulmonary resuscitation, trap making, wild cooking, etc. Among them, the same traveler can hold multiple skill tags; Obtain the physical health index of each traveler in the outdoor team every day; ; Wherein, X represents the total number of days, A represents the number of people in the outdoor team going out, Q represents the risk resistance factor of the outdoor team, and M represents the total number of types of skill tags held in the outdoor team. is the hyperbolic tangent function. represents the physical health index of the j-th person going out on the r-th day; It should be noted that represents the historical average health level of all members of the team; the greater the average health level and the more types of skill tags, the greater the Q value, indicating that the outdoor team has a higher risk resistance ability and is more capable of coping with possible risks outdoors.

[0027] Risk impact factor analysis module 102: Calculate the environmental convergence and disaster-affected difference degree between any two similar disaster points in the same geological area, and then obtain the disaster type convergence degree of any two similar disaster points. Use the disaster type convergence degree to cluster all similar disaster points in the same geological area to obtain the clustering result; According to the clustering result, obtain the disaster occurrence probability in different cluster areas, and then obtain the meteorological characteristic parameters of each cluster area. According to the number of occurrences of different types of natural disasters in each geological area and the correlation coefficient between the disaster occurrence probability and meteorological characteristic parameters in each cluster area, obtain the influence degree of meteorological and geological factors when different types of natural disasters occur; Then obtain the personnel skill tags and physical health index, and then obtain the risk resistance factor of the outdoor team.

[0028] It should be noted that the occurrence of natural disasters is affected by two main factors: meteorology and geology. Different meteorological data and geological data lead to different natural disasters. According to this characteristic, first classify all natural disaster points according to the disaster type. In order to reflect the two influencing factors of meteorology and geology, in this embodiment, similar disaster points are classified in each geological area. The disaster type convergence degree is specifically: Natural disaster points with the same disaster type are called similar disaster points; among them, the meteorological monitoring data of each natural disaster point includes several meteorological types. For any two similar disaster points a and b in the projection map, the calculation method of the environmental convergence of similar disaster points a and b is: ; Wherein, represents the mean square error of all meteorological type data in the meteorological monitoring data of similar disaster points a and b, represents the spatial distance between similar disaster points a and b on the projection map, represents the environmental convergence between similar disaster points a and b. The smaller the environmental convergence, the closer the meteorological monitoring data and geological factors of similar disaster points a and b are during the occurrence of disasters; Represent the mean square error of the meteorological monitoring data of the same type of disaster points a and b, as well as the Euclidean norm of the coordinate distance between the disaster points. Furthermore, similar geology and meteorology may also lead to natural disasters of different levels. Therefore, according to the disaster level differences of natural disaster points, obtain the disaster-affected difference degrees of the same type of disaster points a and b: The disaster-affected difference degrees of the same type of disaster points a and b The calculation method is as follows: ; Among them, represents the disaster level of the a-th natural disaster point, represents the disaster level of the b-th natural disaster point, represents the maximum disaster level in the disaster type to which the same type of disaster points a and b belong on the projection map, represents the minimum disaster level in the disaster type to which the same type of disaster points a and b belong on the projection map; represents the absolute value of the disaster level difference between the same type of disaster points a and b, represents the normalization of the absolute value of the disaster level difference between the same type of disaster points a and b; Multiply the environmental similarity of the same type of disaster points a and b by the disaster-affected difference degree to obtain the disaster type similarity degree of the same type of disaster points a and b; Similarly, obtain the disaster type similarity degrees of any two same type of disaster points.

[0029] Use the disaster type similarity degrees of any two same type of disaster points to perform k-means clustering on all same type of disaster points, obtaining several clusters. Each cluster contains several same type of disaster points, and record the convex hull area of the natural disaster points in each cluster as the cluster area of each cluster.

[0030] It should be noted that for all natural disaster points of any disaster type in the same geological area, since the clustering is based on the disaster type of the natural disaster points, each cluster represents the distribution of a disaster type in this geological area. Then, the more the number of natural disaster points in a cluster and the smaller the distribution range of the cluster area, the more it can indicate that the disaster occurrence probability of the disaster type corresponding to this cluster is greater.

[0031] Specifically, record the disaster type of all natural disaster points in each cluster as the disaster type of each cluster; For the v-th cluster, the disaster occurrence probability of the v-th cluster The calculation method is as follows: ; Among them, represents the number of natural disaster points within the v-th cluster, represents the area of the cluster region of the v-th cluster; represents the ratio of the number of natural disaster points within the v-th cluster to the area of the cluster region of the v-th cluster.

[0032] It should be noted that since there is a relationship between geological disasters and meteorology for some disaster types, it is necessary to analyze the correlation between the disaster occurrence probability of each disaster type and meteorology. Therefore, first, the meteorological monitoring data of each cluster is quantified to obtain the meteorological characteristic parameters of each cluster.

[0033] Specifically, any two natural disaster points within each cluster are combined pairwise to obtain a number of pairs of disaster points, and the mean of the mean square errors of the meteorological monitoring data among all the pairs of disaster points is calculated and recorded as the meteorological characteristic parameter of each cluster.

[0034] It should be noted that if there are natural disaster points in the cluster that cannot be paired, the meteorological characteristic parameter does not consider this natural disaster point.

[0035] Furthermore, among all the clusters, all the clusters of the same disaster type are sorted in descending order according to the size of the cluster to obtain the cluster sequence of each disaster type. According to the order of the cluster sequence, the meteorological characteristic parameters and disaster occurrence probabilities of each cluster are sorted to obtain the meteorological characteristic parameter sequence and disaster occurrence probability sequence of each disaster type. The absolute value of the Pearson correlation coefficient of the meteorological characteristic parameter sequence and disaster occurrence probability sequence of each disaster type is recorded as the meteorological influence factor of each disaster type.

[0036] Furthermore, obtain the degree of influence of external factors for the o-th disaster type: ; Among them, is the variance of the number of natural disaster points in all the clusters belonging to the o-th disaster type, is the meteorological influence factor of the o-th disaster type; It should be noted that since different geographical regions are different, the influence of external factors is also different. When the o-th disaster type occurs in all the clusters in different geological regions, and the variance of the number of natural disasters that occur is larger, it proves that this disaster type is affected by the geological and meteorological environments in different geological regions, that is, it represents that the o-th natural disaster is more affected by the geological environment. Therefore, the degree of influence of external factors of the o-th disaster type is greater.

[0037] Risk assessment module 103: Assess the risk of this trip according to the planned travel route of the outdoor team.

[0038] Since outdoor trips require advance planning of the travel route, the travel route will pass through different geological regions. Based on the length of stay in different regions and the current meteorological monitoring information during the trip, a hierarchical early warning of outdoor accident risks is carried out to take preventive measures against risks in advance and plan the route; First, overlap the travel route of the outdoor team on the projection map that has completed clustering processing. When the travel route passes through each cluster area, use the historical disaster record data of the cluster area to evaluate the risk of each position on the travel route. It should be noted that each route position may pass through multiple different cluster areas simultaneously. Specifically: Map the travel route of the outdoor team to the projection map to obtain the cluster to which each campsite in the travel route of the outdoor trip belongs in the projection map, denoted as the stationed cluster of each campsite; it should be noted that since the cluster area of each cluster is the convex hull area of the cluster, there is an intersection between clusters, so each campsite has more than one stationed cluster; Denote the camping time of the outdoor team at the p-th campsite as ; Denote the influence degree of external factors of the disaster type to which the n-th stationed cluster of the p-th campsite belongs as ; Denote the disaster occurrence probability of the n-th stationed cluster of the p-th campsite as ; The way to obtain the risk assessment value of the travel plan is as follows: Obtain the risk assessment coefficient of the travel route of the outdoor team : ; Among them, L represents the number of campsites in the travel route of the outdoor team, represents the number of stationed clusters of the p-th campsite in the travel route of the outdoor team; is the sequence composed of all meteorological type data in the meteorological monitoring data of the p-th campsite in the travel route of the outdoor team; is the sequence composed of all meteorological type data in the meteorological monitoring data of the natural disaster point closest to the p-th campsite in the travel route of the outdoor team; represents obtaining the mean square error function; is the hyperbolic tangent function; It should be noted that It represents the residence time at position p in the outdoor team travel plan, divided by the average of the mean square errors between the meteorological monitoring data at the p-th position on the travel day and the meteorological monitoring data of all natural disaster points in the n-th cluster area. That is, when the residence time of the outdoor team at the p-th position is longer and the average of the mean square errors between the meteorological monitoring data on the travel day and the historical meteorological monitoring data is smaller, the output value of this fraction is larger.

[0039] Furthermore, the difference between the risk assessment coefficient of the travel route of the outdoor team and the risk resistance factor of the outdoor team is recorded as the risk assessment value of the travel plan.

[0040] The preset risk warning threshold is 0.5. When the risk assessment value of the travel plan is greater than the risk warning threshold, a warning is issued for this travel plan, and then the travel plan is re-planned. The travel plan includes the travel date and the travel route. Several re-planned travel plans are obtained, and each re-planned travel plan gets a risk assessment value. The travel plan with the smallest risk assessment value is taken as the optimal travel plan.

[0041] The risk assessment module of the risk warning system can re-plan the travel plan and use big data to avoid travel risks in advance. After the actual travel, the risk warning system also needs to use some auxiliary devices to monitor the travel process in real time. The risk warning system will be connected to the pan-tilt camera near the personnel activity area and the portable earthquake monitoring device carried by the outdoor team, communicate bidirectionally and interact information in real time. The image processing system carried by the infrared device on the pan-tilt camera can accurately identify the living animals near the personnel activity area, and the earthquake monitoring device is used to monitor the ground vibration frequency of the personnel activity area. After connecting the pan-tilt camera near the personnel activity area and the portable seismograph carried by the outdoor team, the real-time monitoring data of the living animals and the ground vibration frequency in the personnel activity area can be obtained in real time, and the real-time monitoring information is sent to the mobile terminal device of the outdoor team for the outdoor team to view conveniently. And when a wild animal appears within the field of view of the pan-tilt camera or the ground vibration frequency exceeds the normal value, the risk warning system will send a voice alarm to the outdoor team in time to further ensure the safety of the outdoor team during travel.

[0042] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the principles of the present invention shall be included in the protection scope of the present invention.

Claims

1. An intelligent hierarchical early warning system for outdoor safety accident risks, characterized in that: The system includes: Data acquisition and preprocessing module: Statistically obtain all natural disaster locations and their disaster types, disaster levels, and meteorological monitoring data; obtain the travel routes of outdoor teams and several camping sites; obtain the risk resistance factors of outdoor teams; Risk influencing factor analysis module: natural disaster points with the same disaster type are regarded as similar disaster points; according to the distance of similar disaster points and the similarity of meteorological monitoring data, the environmental convergence of similar disaster points is obtained; according to the difference in disaster levels of similar disaster points, the environmental convergence is adjusted to obtain the disaster type convergence of similar disaster points; all natural disaster points are clustered using the disaster type convergence to obtain several clusters and their cluster areas; the density of natural disaster points in each cluster is analyzed to obtain the probability of disaster occurrence in each cluster; the influence of meteorological monitoring data of natural disaster points in each cluster under the same disaster type on the cluster area of ​​the cluster is analyzed to obtain the influence degree of external factors of each disaster type; Risk assessment module: obtain the cluster to which each camping point in the outdoor team's travel route belongs, and obtain the risk assessment value of the travel plan based on the probability of disaster occurrence in the cluster and the degree of influence of external factors of the disaster type corresponding to the cluster, combined with the risk resistance factor of the outdoor team; select the optimal travel plan based on the risk assessment value.

2. According to claim 1, an intelligent hierarchical early warning system for outdoor safety accident risks is characterized in that: The specific steps of obtaining the anti-risk factor include: By collecting the physical health data and skill labels of each traveler in the outdoor team, the physical health data of all travelers in the outdoor team are evaluated using a neural network to obtain the physical health index of each traveler in the outdoor team, and then combined with the skill labels contained in the outdoor team to obtain the anti-risk factor of the outdoor team.

3. According to claim 1, an intelligent hierarchical early warning system for outdoor safety accident risks is characterized in that: The specific steps of obtaining the environmental convergence include: Project all natural disaster points onto a two-dimensional topographic map to obtain a projection map; For any similar disaster points a and b in the projection map, the calculation method of the environmental convergence of similar disaster points a and b is: ； in, Represents the mean square error of all meteorological data types in the meteorological monitoring data of the same disaster points a and b, Represents the spatial distance between similar disaster points a and b on the projection map, Represents the environmental convergence between similar disaster locations a and b.

4. According to claim 1, an intelligent hierarchical early warning system for outdoor safety accident risks is characterized in that: The specific steps for obtaining the disaster type convergence degree include: the difference in the degree of disaster damage between the same disaster points a and b The calculation method is: ； in, represents the disaster level of the ath natural disaster point, represents the disaster level of the bth natural disaster point, Represents the maximum disaster level of the disaster type to which the same disaster points a and b on the projection map belong. Represents the minimum disaster level among the disaster types to which the same disaster points a and b on the projection map belong; Multiply the environmental similarity of similar disaster sites a and b by the degree of disaster difference to obtain the disaster type similarity of similar disaster sites a and b.

5. According to claim 1, an intelligent hierarchical early warning system for outdoor safety accident risks is characterized in that: The specific steps of obtaining a plurality of clusters and their cluster regions include: All similar disaster points are clustered using k-means using the disaster type similarity between any two similar disaster points to obtain several clusters. Each cluster contains several similar disaster points. The convex hull area of ​​the natural disaster points in each cluster is recorded as the cluster area of ​​each cluster.

6. According to claim 1, an intelligent hierarchical early warning system for outdoor safety accident risks is characterized in that: The specific steps of obtaining the cluster disaster occurrence probability include: The disaster types of all natural disaster points in each cluster are recorded as the disaster type of each cluster; For the vth cluster, the probability of disaster occurrence in the vth cluster is The calculation method is: ； in, represents the number of natural disaster points in the vth cluster, Represents the area of ​​the cluster region of the vth cluster.

7. According to claim 1, an intelligent hierarchical early warning system for outdoor safety accident risks is characterized in that: The specific steps for obtaining the impact degree of external factors of the disaster type include: Get the external factors impact degree of the oth disaster type : ； in, is the variance of the number of natural disaster locations in all clusters belonging to the oth disaster type, is the meteorological influencing factor of the oth disaster type.

8. According to claim 7, an intelligent graded early warning system for outdoor safety accident risks is characterized in that: The specific steps of obtaining the meteorological influencing factors of the disaster type include: Combining any two natural disaster points in each cluster in pairs to obtain several disaster point pairs, calculating the mean of the mean square error of the meteorological monitoring data in all the disaster point pairs, and recording it as the meteorological characteristic parameter of each cluster; Among all clusters, all clusters of the same disaster type are arranged in descending order according to the size of the clusters to obtain the cluster sequence of each disaster type. The meteorological characteristic parameters and the disaster occurrence probability of each cluster are sorted according to the order of the cluster sequence to obtain the meteorological characteristic parameter sequence and the disaster occurrence probability sequence of each disaster type. The absolute value of the Pearson correlation coefficient of the meteorological characteristic parameter sequence and the disaster occurrence probability sequence of each disaster type is recorded as the meteorological influencing factor of each disaster type.

9. According to claim 1, an intelligent hierarchical early warning system for outdoor safety accident risks is characterized in that: The specific steps of obtaining the risk assessment value of the travel plan include: Obtain the risk assessment value of the travel plan, and record the difference between the risk assessment coefficient of the outdoor team's travel route and the anti-risk factor of the outdoor team as the risk assessment value of the travel plan.

10. The intelligent hierarchical early warning system for outdoor safety accident risks according to claim 9, characterized in that: The specific steps of obtaining the risk assessment value of the travel plan include: Obtain the camping time and meteorological monitoring data of each camping point; Map the travel route of the outdoor team into the projection map, obtain the cluster to which each camping point in the travel route of the outdoor trip belongs in the projection map, and record it as the stationing cluster of each camping point; The camping time of the outdoor team at the pth camping point is recorded as ; The external factor influence degree of the disaster type of the nth camping cluster at the pth camping point is recorded as ; The probability of disaster occurrence of the nth cluster at the pth camping point is recorded as ; Get the risk assessment factor of the outdoor team's travel route : ； Among them, L represents the number of camping points in the outdoor team's travel route. represents the number of camping clusters at the p-th camping point in the travel route of the outdoor team; It is a sequence consisting of all meteorological type data in the meteorological monitoring data of the p-th camping point in the travel route of the outdoor team; is a sequence consisting of all meteorological type data in the meteorological monitoring data of the natural disaster point closest to the pth camping point in the travel route of the outdoor team; Indicates the mean square error function; is the hyperbolic tangent function.

Images