CN117689210A - Intelligent evaluation method for risk of natural disasters of roads - Google Patents
Intelligent evaluation method for risk of natural disasters of roads Download PDFInfo
- Publication number
- CN117689210A CN117689210A CN202410113673.1A CN202410113673A CN117689210A CN 117689210 A CN117689210 A CN 117689210A CN 202410113673 A CN202410113673 A CN 202410113673A CN 117689210 A CN117689210 A CN 117689210A
- Authority
- CN
- China
- Prior art keywords
- road
- natural disaster
- risk
- disaster risk
- natural
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000011156 evaluation Methods 0.000 title claims abstract description 23
- 238000011160 research Methods 0.000 claims abstract description 48
- 238000012502 risk assessment Methods 0.000 claims abstract description 44
- 238000000034 method Methods 0.000 claims abstract description 26
- 238000012549 training Methods 0.000 claims abstract description 25
- 238000013210 evaluation model Methods 0.000 claims abstract description 12
- 230000000694 effects Effects 0.000 claims abstract description 9
- 238000013528 artificial neural network Methods 0.000 claims description 13
- 238000001556 precipitation Methods 0.000 claims description 9
- 238000010606 normalization Methods 0.000 claims description 6
- 238000002790 cross-validation Methods 0.000 claims description 5
- 238000012795 verification Methods 0.000 claims description 5
- 230000002159 abnormal effect Effects 0.000 claims description 3
- 238000004140 cleaning Methods 0.000 claims description 3
- 238000012545 processing Methods 0.000 claims description 3
- 239000011159 matrix material Substances 0.000 claims description 2
- 238000004364 calculation method Methods 0.000 description 6
- 238000011161 development Methods 0.000 description 5
- 238000012417 linear regression Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 230000009471 action Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000010485 coping Effects 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 238000004451 qualitative analysis Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the technical field of road natural disaster risk assessment, and particularly discloses an intelligent road natural disaster risk assessment method, which comprises the following steps: determining the natural disaster type of a road in a research area; determining influence factors of natural disasters of roads; establishing a road natural disaster risk assessment data set based on the data; training and storing an intelligent evaluation model of the risk of the natural disasters of the roads; establishing a road natural disaster prediction data set and calculating a prediction result; cross-verifying the training set and the prediction set; grading the influence factors and the evaluation results, and drawing a road natural disaster risk demarcation graph by using the ArcGIS. The invention establishes a road natural disaster risk assessment index system and a road natural disaster risk intelligent assessment model based on the actual situation of a research area and by combining the relevant influences of weather, geology and human activities. The method has the advantages that the influence factor index is easy to obtain, the method is simple to operate, and new thinking and directions are provided based on the subsequent road natural disaster risk assessment research of the intelligent assessment means.
Description
Technical Field
The invention belongs to the technical field of road natural disaster risk assessment, and particularly relates to natural disaster risk assessment, road natural disaster risk assessment and neural network technology.
Technical Field
The stormwater and the natural disasters derived from the stormwater cause road surface damage, landslide and the like of roadbeds, and seriously obstruct the life and property safety of people endangered by traffic. The road is a life line for evacuating people after disasters, transporting rescue teams and materials, has important strategic positions in national economy development, has close development relationship with other industrial departments of national economy, and the loss caused by the road disasters can be transmitted to other related industries to seriously influence the life and property safety of people and the development of regional socioeconomic performance.
National institutes point out that to improve the anti-risk capability of a traffic network, strengthen the safety risk assessment and the classified management and control of traffic infrastructure, strengthen the identification of major risk sources and the dynamic monitoring analysis, prediction and early warning of the whole process, build a meteorological monitoring and early warning system in important channels, hubs and shipping areas, and improve the capability of the traffic infrastructure for adapting to climate change.
In the Internet era, digitization and informatization are main development trends of various industries and departments, and artificial intelligence such as image recognition and machine learning is vigorously developed to provide a new thought for disaster risk assessment technical research. Developing a natural disaster risk index system, adopting novel technical means, establishing an intelligent evaluation model of the road natural disaster risk, constructing a scientific and reasonable evaluation system, providing scientific basis and risk avoiding means for related departments in pre-disaster prediction, promoting the comprehensive upgrading of a road network management system, and greatly improving the resolution capability of coping with the road natural disaster.
At present, the research on disaster risk is gradually changed from single disaster research to multi-disaster coupling and multi-disaster chain combined action, and the occurrence of natural disasters on roads is the combined action result of multiple factors; the evaluation result can be selected according to different influencing factors and different results are generated by a calculation method, and a unified evaluation standard is difficult to determine; the research on road disaster risk assessment is still in a starting stage, is concentrated on image recognition and remote sensing images, and is difficult to realize real-time prediction of road disasters.
Therefore, an intelligent evaluation method for the risk of the natural disasters of the roads is established, the selection of risk evaluation indexes is divided into three categories, namely meteorological conditions, geological conditions and human activity factors, and ten influencing factors, namely temperature, relative humidity, precipitation, elevation, gradient, slope direction, geological lithology, vegetation coverage rate, historical influencing factors and road density, are selected from the three categories, and the influence of multiple disasters on the natural disasters of the roads is comprehensively considered; meanwhile, the road natural disaster risk intelligent assessment model is adopted to carry out road natural disaster risk assessment, a quantitative mode is used for predicting disaster risk, human intervention is reduced, accuracy is improved, and the selected influence factor indexes are easy to acquire and determine, so that the result is basically reliable, and the method is suitable for road natural disaster risk assessment.
Disclosure of Invention
The invention aims to establish an intelligent road natural disaster risk assessment method to solve the problems of qualitative assessment and artificial subjective factor interference in natural disaster risk assessment, and realize comprehensive assessment and analysis of multi-disaster road natural disasters.
In order to achieve the above purpose, the invention adopts the following technical scheme:
analyzing historical disaster conditions of a research area and road natural disaster influence factors, determining main natural disaster types and influence factors of roads of the research area, selecting ten influence factor indexes of temperature, relative humidity, precipitation, elevation, gradient, slope direction, geological lithology, vegetation coverage rate, historical influence factors and road density from three major factors of meteorological conditions, geological conditions and human activity factors, and establishing a road natural disaster risk intelligent assessment model; determining an optimal road natural disaster risk intelligent evaluation model, and corresponding weight values and threshold values of all influence factors through a training set; calculating an influence factor input value of a road network disaster point through a constructed road natural disaster risk intelligent evaluation model to output a prediction result; establishing a road natural disaster risk assessment system of a research area, and carrying out assessment grade division according to the values of all influence factors and the prediction result of the neural network linear return model; and drawing a regional graph for natural disaster risk assessment of the road in the research region by adopting ArcGIS according to the prediction result of the neural network linear return model.
The method comprises the following steps:
analyzing the natural disaster history condition of the research area and the road natural disaster influence factors to determine the common types of the natural disasters of the research area, including: collapse, landslide, debris flow, subsidence, collapse, and flood;
comprehensively analyzing natural influence factors of roads in a research area, and selecting ten indexes including temperature, relative humidity, precipitation, elevation, gradient, slope direction, geological lithology, vegetation coverage, historical influence factors and road density by taking meteorological conditions, geological conditions and human activity factors into consideration;
establishing a natural disaster risk assessment data set of a research area based on the influence factors, and performing data cleaning, abnormal value removal and data normalization processing on the constructed data set;
inputting the processed data set data into an intelligent evaluation model of the risk of the natural disasters for training, and storing the trained model and the optimal weight value when the output data reach the optimal training result;
establishing a research area road disaster point prediction set, calling a trained model and calculating optimal weights, and calculating a research area road natural disaster risk assessment prediction result;
cross-verifying the obtained training set and the prediction set, and verifying the accuracy of the evaluation result of the model;
constructing a road natural disaster risk assessment system of a research area, dividing influence factor classification and road natural disaster risk assessment results, and dividing a prediction result into a low-level road risk area, a medium-level road risk area and a high-level road risk area;
and carrying out risk classification according to the prediction result of the road natural disaster risk intelligent assessment model, and drawing a road natural disaster risk assessment regional map of the research region through ArcGIS software.
Preferably, in the step S2, ten indexes of the influence factors, including the selected temperature, the relative humidity, the precipitation, the elevation, the gradient, the slope direction, the geological lithology, the vegetation coverage, the history influence factors and the road density, may be obtained and stored through a system server of a corresponding department, and the corresponding real-time data may be evaluated.
Preferably, the training set data is determined according to a fuzzy comprehensive evaluation method, an influence factor set is constructed, the weight of each factor is determined through an entropy value method, single-factor fuzzy evaluation is carried out to obtain an evaluation matrix, and a final score is obtained to form the training set.
Preferably, in S4, the neural network linear model is trained by the training set to achieve the best model accuracy, and the model accuracy verification index selects the mean absolute error, the root mean square error and the correlation coefficient.
Preferably, in the step S6, the prediction set in step S5 and the data set in step S3 are cross-validated, and the cross-validation method selects k-fold cross-validation.
Preferably, the risk and the like are divided according to the values of the prediction results of the neural network, wherein the prediction results Y <0.2 are extremely low risk level road sections, Y <0.2 is less than or equal to 0.4 and is a lower risk level road section, Y <0.6 is less than or equal to 0.4 and is an intermediate risk level road section, Y <0.8 is a high risk level road section, and Y >0.8 is an extremely high risk level road section.
Compared with the prior art, the method and the system fully combine the basic environment condition, the historical disaster condition and the road natural disaster condition of the research area, analyze the road natural disaster type and the characteristics of the research area, the climate condition, the geological condition and the human activity condition of the research area, select main influencing factors, respond to the dangerousness of disaster causing factors through the temperature, the relative humidity and the precipitation in the meteorological condition, respond to the sensitivity of disaster bearing bodies through the elevation, the gradient, the slope direction, the geological lithology and the vegetation coverage rate in the geological condition, superimpose the index historical influencing factors and the road density which represent the human activity factors, establish a road natural disaster risk assessment model which can be used for multiple disasters through a road natural disaster risk intelligent assessment model, calculate the road natural disaster risk prediction value of the research area, establish a road natural disaster risk assessment system, divide the road natural disaster risk assessment level, wherein the acquisition of influencing factor data can be acquired through corresponding official websites and adopt qualitative analysis, reduce the scientificity of subjective judgment and enhance the assessment result, and if the method and the system can be connected with a related department database, realize real-time prediction of the road natural disaster risk, real-time early warning and control of the road natural disaster risk, enhance the trip safety index and improve the capability of the disaster prevention.
Drawings
FIG. 1 is a diagram of a model structure of a fuzzy neural network in an embodiment of the present invention.
FIG. 2 is a block diagram of a neural network linear regression model, in accordance with an embodiment of the present invention.
Fig. 3 is a schematic flow chart of the intelligent evaluation method for risk of natural disasters of the road.
Detailed Description
The invention will now be described in detail with reference to specific examples, wherein the invention is illustrated by way of example and not by way of limitation.
The technical scheme adopted by the invention is an intelligent evaluation method for the risk of the natural disasters of the research area, which comprises the steps of analyzing the historical disaster situation of the research area and the influence factors of the natural disasters of the road, fully combining the basic environment situation of the research area and the natural disasters of the road, analyzing the occurrence type and the characteristics of the natural disasters of the research area, the climate conditions, the geological conditions and the human activities of the research area, and obtaining the influence factor indexes of the natural disaster risk points of the research area to construct a natural disaster risk evaluation data set of the research area; determining a correlation coefficient through training a training set model to enable the road natural disaster risk intelligent evaluation model and the weight value to reach an optimal result and storing the model; establishing a road disaster point prediction set of a research area, inputting relevant values of influence factors into a trained model, and outputting a prediction result; cross-verifying the training set and the prediction set, and verifying the accuracy of the model; establishing a road natural disaster risk assessment system, and carrying out risk classification on the index values of the influencing factors and the prediction results of the intelligent road natural disaster risk assessment model; and (3) carrying out natural disaster risk assessment regional graphs on the road of the research region by drawing the prediction result by adopting ArcGIS. The method is simple, the influence factor index data is easy to obtain, and if the method can be connected with a related department database, the data can be obtained in real time and predicted, so that new thinking and technical means are provided for the development of a follow-up road natural disaster risk assessment and early warning system.
According to the invention, the main influencing factors of the natural disasters of the road are determined by counting and analyzing the disaster loss of the natural disasters of the road, ten indexes including temperature, relative humidity, precipitation, elevation, gradient, slope direction, geological lithology, vegetation coverage rate, history influencing factors and road density are determined, and the corresponding weights of the indexes are determined to form a risk evaluation index system of the natural disasters of the road; aiming at different disaster factors, an intelligent road natural disaster risk assessment model is adopted, the index weight of the influence factors is determined, the index of the road natural disaster risk is calculated, the risk level of the road natural disaster risk point is determined by combining data analysis, a road natural disaster risk assessment system is established, a road natural disaster risk area map is drawn, an intelligent road natural disaster risk assessment method is formed, and a flow chart of the assessment method is shown in fig. 3.
The method comprises the following steps:
analyzing the natural disaster history condition and the road natural disaster type of the research area to determine the common road natural disaster type of the research area, comprising the following steps: collapse, landslide, debris flow, subsidence, collapse, and flood;
comprehensively analyzing and researching natural characteristics of a regional road to select influence factors, and selecting ten indexes including temperature, relative humidity, precipitation, elevation, gradient, slope direction, geological lithology, vegetation coverage, historical influence factors and road density by considering meteorological conditions, geological conditions and human activity factors;
based on the influence factors, a natural disaster risk assessment data set of a research area is established, and data cleaning, abnormal value removal and data normalization processing are carried out on training set data;
normalization calculation formula:
inverse normalization formula:
inputting the processed training set data into an intelligent evaluation model of the risk of the natural disasters for training, and storing the trained model and the optimal weight value when the output data reach the optimal training result;
the intelligent evaluation model 1 for the risk of the natural disasters of the roads is as shown in fig. 1, the fuzzy neural network structure inputs influencing factors, membership calculation is performed, normalization calculation is performed through judgment of a rule layer, a prediction result is output, and the fuzzy neural network model relates to the formula:
average absolute error formula:
root mean square error formula:
membership function calculation, gaussian formula:
the road natural disaster risk intelligent evaluation model 2 is shown in fig. 2, a neural network linear regression structure diagram is shown in fig. 2, influence factors are input, a predicted value is output through hidden layer calculation, and the neural network linear regression model relates to the formula:
mean square loss function formula:
the optimizer formula:
building a road natural disaster prediction set, calling a trained model of a training set and calculating optimal weights, and calculating a road natural disaster risk assessment prediction result of a research area;
performing cross verification on the obtained training set and the prediction set, verifying the accuracy of the evaluation result of the model, selecting a k-fold cross verification method, dividing the data set for k times, performing training and evaluation once for each division, obtaining the evaluation result after k times of division, and averaging to obtain the final score;
constructing a road natural disaster risk assessment system, dividing influence factor classification and road natural disaster risk assessment results, and dividing the prediction results into a low-level road risk area, a medium-level road risk area and a high-level road risk area;
and (3) carrying out road natural disaster risk classification according to the neural network linear regression prediction result, and drawing a road natural disaster risk assessment regional map of the research region through ArcGIS software.
Claims (7)
1. An intelligent assessment method for risk of natural disasters of roads is characterized by comprising the following steps:
analyzing natural disaster history conditions and disaster influence factors of a research area;
selecting proper natural disaster influence factors, and constructing a natural disaster influence factor data set of a research area;
training the data set through an intelligent evaluation model, and storing the trained model and an optimal weight value;
establishing a road natural disaster prediction set, inputting prediction set data into a trained model, and outputting a prediction result;
constructing a road natural disaster risk assessment system, classifying risk grades of prediction results, and drawing a road natural disaster risk assessment regional graph of a research region.
2. The road natural disaster risk intelligent assessment method according to claim 1, comprising the steps of:
s1, analyzing the historical condition of the natural disasters of the road in the research area, and determining the common types of the natural disasters of the road in the research area, wherein the types comprise: collapse, landslide, debris flow, subsidence, collapse, and flood;
s2, comprehensively analyzing natural influence factors of a research area, and selecting ten indexes including temperature, relative humidity, precipitation, elevation, gradient, slope direction, geological lithology, vegetation coverage, history influence factors and road density by considering meteorological conditions, geological conditions and human activity factors;
s3, a research area natural disaster influence factor data set is established according to the natural influence factors and the research area natural disaster history conditions, and then data cleaning, abnormal value removal and data normalization processing are carried out on the data of the data set;
s4, inputting the processed data set into an intelligent evaluation model of the natural disaster risk of the road for training, and storing the trained model and the optimal weight value when the output data reaches the optimal training result;
s5, establishing a prediction set of the road network in the research area, and calculating a natural disaster risk assessment prediction result of the road network in the research area through the model and the optimal weight value to obtain the prediction set;
s6, carrying out cross verification on the prediction set in the S5 and the data set in the S3, and verifying the accuracy of the evaluation result of the intelligent evaluation model of the road natural disaster risk;
s7, constructing a road network natural disaster risk assessment system, dividing influence factor classification and road network natural disaster risk assessment results, and dividing the prediction results into a low-level risk road section area, a medium-level risk road section area and a high-level risk road section area;
and S8, carrying out road network natural disaster risk classification according to the prediction result obtained in the S7, and drawing a road network natural disaster risk assessment regional graph of the research region through ArcGIS software.
3. The supervised learning based road natural disaster risk assessment method of claim 2, wherein: in the step S2, ten indexes of the influence factors, namely, the selected temperature, the relative humidity, the precipitation, the elevation, the gradient, the slope direction, the geological lithology, the vegetation coverage, the history influence factors and the road density, can be acquired and stored through a system server of a corresponding department, and the corresponding real-time data can be evaluated.
4. The road natural disaster risk intelligent assessment method according to claim 2, characterized in that: the training set data is determined according to a fuzzy comprehensive evaluation method, an influence factor set is constructed, the weight of each factor is determined through an entropy value method, single-factor fuzzy evaluation is carried out to obtain an evaluation matrix, and a final score is obtained to form a training set.
5. The road natural disaster risk intelligent assessment method according to claim 2, characterized in that: in the step S4, the neural network linear model is trained through the training set to achieve the optimal model precision, and the model precision verification index selects the average absolute error, the root mean square error and the correlation coefficient.
6. The road natural disaster risk intelligent assessment method according to claim 2, characterized in that: in the step S6, the prediction set in the step S5 and the data set in the step S3 are subjected to cross validation, and the cross validation method selects k-fold cross validation.
7. The road natural disaster risk intelligent assessment method according to claim 2, characterized in that: and dividing risks and the like according to the values of the predicted results of the neural network, wherein the predicted results Y <0.2 are extremely low risk level road sections, Y <0.4 which is more than or equal to 0.2 is relatively low risk level road sections, Y <0.6 which is more than or equal to 0.4 is medium risk level road sections, Y <0.8 which is more than or equal to 0.6 is high risk level road sections, and Y >0.8 is extremely high risk level road sections.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410113673.1A CN117689210A (en) | 2024-01-26 | 2024-01-26 | Intelligent evaluation method for risk of natural disasters of roads |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410113673.1A CN117689210A (en) | 2024-01-26 | 2024-01-26 | Intelligent evaluation method for risk of natural disasters of roads |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117689210A true CN117689210A (en) | 2024-03-12 |
Family
ID=90130369
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410113673.1A Pending CN117689210A (en) | 2024-01-26 | 2024-01-26 | Intelligent evaluation method for risk of natural disasters of roads |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117689210A (en) |
-
2024
- 2024-01-26 CN CN202410113673.1A patent/CN117689210A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108280553B (en) | Mountain torrent disaster risk zoning and prediction method based on GIS-neural network integration | |
| CN109508360B (en) | Geographical multivariate stream data space-time autocorrelation analysis method based on cellular automaton | |
| Janalipour et al. | Building change detection after earthquake using multi-criteria decision analysis based on extracted information from high spatial resolution satellite images | |
| Sun et al. | Spatio‐Temporal Segmented Traffic Flow Prediction with ANPRS Data Based on Improved XGBoost | |
| CN113496314B (en) | Method for predicting road traffic flow by neural network model | |
| CN115730684A (en) | Air quality detection system based on LSTM-CNN model | |
| CN117494034A (en) | Air quality prediction method based on traffic congestion index and multi-source data fusion | |
| KR20150131801A (en) | Landslide Prediction System using Geographic Information System and NeuroFuzzy techniques and Landslide Prediction Method using Thereof | |
| CN116610816A (en) | Personnel portrait knowledge graph analysis method and system based on graph convolution neural network | |
| CN116258241A (en) | Space-time correlation-based supervision environment risk prediction method and system | |
| CN116777079A (en) | Desertification intrusion disaster prediction method based on Bayesian interlayer structure model | |
| CN115629160A (en) | Air pollutant concentration prediction method and system based on space-time diagram | |
| CN116662468A (en) | Urban functional area identification method and system based on geographic object space mode characteristics | |
| CN116702090A (en) | Multi-mode data fusion and uncertain estimation water level prediction method and system | |
| CN116738192A (en) | Digital twinning-based security data evaluation method and system | |
| CN117332909B (en) | Multi-scale urban waterlogging road traffic exposure prediction method based on intelligent agent | |
| CN117808214A (en) | Hydraulic engineering data analysis system | |
| CN116110210B (en) | Data-driven landslide hazard auxiliary decision-making method in complex environment | |
| CN117933701A (en) | Rail transit engineering construction safety risk monitoring method and system | |
| CN117558124A (en) | Channel running state deduction prediction method based on multi-source information depth fusion | |
| CN116933947A (en) | Landslide susceptibility prediction method based on soft voting integrated classifier | |
| CN109636194B (en) | Multi-source cooperative detection method and system for major change of power transmission and transformation project | |
| Zhang et al. | ARIMA Model‐Based Fire Rescue Prediction | |
| CN117689210A (en) | Intelligent evaluation method for risk of natural disasters of roads | |
| CN113191568B (en) | Meteorological-based urban operation management big data analysis and prediction method and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |