CN117195749A

CN117195749A - Regional highway network risk prediction method by climate change and readable storage medium

Info

Publication number: CN117195749A
Application number: CN202311464140.XA
Authority: CN
Inventors: 张永林; 李悦; 李明君; 吴睿; 宋媛媛; 邢有凯; 李晓易; 李齐丽; 程逸楠; 徐洪磊; 黄全胜
Original assignee: Transport Planning And Research Institute Ministry Of Transport
Current assignee: Transport Planning And Research Institute Ministry Of Transport
Priority date: 2023-11-06
Filing date: 2023-11-06
Publication date: 2023-12-08
Anticipated expiration: 2043-11-06
Also published as: CN117195749B

Abstract

The invention discloses a regional highway network risk prediction method by climate change and a readable storage medium, comprising the following steps: firstly, obtaining highway network information and disaster influence factors of an area to be predicted; collecting reference year observation data; selecting and simulating global climate mode data; determining a predicted scenario and predicting the climate characteristics of a target year; performing correction operation through a space-time sequence prediction model to obtain reliable target annual climate characteristics; determining a climate characteristic change result; dividing the region to be predicted, and interpolating to obtain the climate change risk coefficient of each space unit; and finally, obtaining a regional highway network risk prediction result of the regional highway network to be predicted according to the risk coefficient. By the design, a scientific and effective evaluation mode is provided, and the capability of the highway network for coping with climate change is improved.

Description

Regional highway network risk prediction method by climate change and readable storage medium

Technical Field

The invention relates to the field of highway planning, in particular to a regional highway network risk prediction method by climate change and a readable storage medium.

Background

As climate change problems become more serious, their impact on traffic infrastructure, especially on road networks, becomes greater.

However, the existing climate change has certain limitations on regional highway network risk prediction methods, such as insufficient accuracy, lack of real-time performance and the like.

These methods often fail to accurately predict the impact of future climate changes on the road network, nor provide an effective countermeasure strategy.

Therefore, a new method capable of predicting the risk of the climate change to the regional highway network more accurately is developed, and the method has important practical significance and application value.

Disclosure of Invention

The invention aims to provide a regional highway network risk prediction method by climate change.

In a first aspect, an embodiment of the present invention provides a method for predicting risk of a regional highway network by using climate change, including:

acquiring regional information of a regional highway network to be predicted and disaster influence factors;

collecting reference year observation data of the highway network of the area to be predicted according to the disaster influence factors;

selecting target global climate mode data, and simulating the reference annual climate characteristics of the highway network of the area to be predicted according to the target global climate mode data;

determining a prediction scene of climate change on regional highway network risks corresponding to the regional highway network to be predicted according to the target global climate mode data, and predicting target annual climate characteristics of the regional highway network to be predicted under the prediction scene, wherein the prediction scene comprises the following steps:

Inputting the reference year observation data, the reference year climate characteristics and the target year climate characteristics under the prediction scene into a pre-trained space-time sequence prediction model for correction operation, so as to obtain reliable target year climate characteristics of the highway network of the area to be predicted under the prediction scene;

determining a climate characteristic change result according to the difference value of the reference year climate characteristic and the reliable target year climate characteristic;

determining a climate characteristic space distribution result of the regional highway network to be predicted based on the climate characteristic change result;

dividing the highway network of the area to be predicted to obtain a plurality of space units;

interpolating the climate characteristic space distribution result into the plurality of space units to obtain a climate change risk coefficient of each space unit under the prediction scene;

calculating the risk coefficient of the climate change of each space unit according to the regional information to obtain the risk coefficient of the regional highway network by the climate change of each space unit;

and obtaining a regional highway network risk prediction result of the regional highway network to be predicted under the prediction scene according to the regional highway network risk coefficient of the climate change of each space unit.

In one possible implementation manner, the collecting the reference year observation data of the highway network of the area to be predicted according to the disaster influence factors includes:

acquiring historical disaster record data and field disaster investigation data corresponding to the highway network of the area to be predicted;

analyzing and determining the disaster influence factors aiming at the highway network of the area to be predicted according to the historical disaster record data and the site disaster investigation data;

and collecting reference year observation data which are related based on the disaster influence factors in the highway network of the area to be predicted.

In one possible implementation manner, the selecting the target global climate pattern data and simulating the reference annual climate characteristic of the highway network of the to-be-predicted area according to the target global climate pattern data includes:

acquiring a climate pattern dataset comprising a plurality of global climate pattern data;

simulating the climate mode data set based on the disaster influence factors to obtain simulation scores of each global climate mode data for the disaster influence factors;

selecting global climate mode data with highest simulation score as the target global climate mode data;

And simulating the reference year climate characteristics of the highway network of the area to be predicted according to the target global climate mode data.

In a possible implementation manner, the simulating the reference annual climate characteristic of the highway network of the area to be predicted according to the target global climate pattern data includes:

and inputting the target global climate pattern data, the reference year and the grid information of the regional highway network to be predicted into a WRF-FDDA model, and obtaining the reference year climate characteristics of the regional highway network to be predicted through simulation.

In a possible implementation manner, the prediction scenario of the climate change on the regional highway network risk corresponding to the regional highway network to be predicted is determined according to the target global climate mode data, and the target annual climate characteristic of the regional highway network to be predicted in the prediction scenario is predicted;

determining a prediction scene of the climate change on the regional highway network risk corresponding to the regional highway network to be predicted according to the target global climate mode data;

inputting the target year, the grid information of the highway network of the area to be predicted and the target global climate pattern data of the target year section under the prediction scene into a WRF model, simulating to obtain the climate characteristics of the target year section under the prediction scene, and taking the climate characteristics of the target year section under the prediction scene as the target year climate characteristics, wherein the target year section comprises a target year, the previous year of the target year and the next year of the target year.

In a possible implementation manner, the inputting the reference year observation data, the reference year climate feature and the target year climate feature in the prediction scenario into a pre-trained space-time sequence prediction model for correction operation, so as to obtain the reliable target year climate feature of the highway network of the area to be predicted in the prediction scenario, which includes:

inputting the reference year observation data, the reference year climate characteristics and the target year climate characteristics under the prediction scene into the space-time sequence prediction model constructed by the convolutional neural network and the long-short-time neural network for correction operation, so as to obtain the climate characteristics respectively corresponding to the target year, the previous year of the target year and the next year of the target year;

and calculating to obtain the average values of the climate characteristics corresponding to the target year, the previous year of the target year and the next year of the target year respectively, and taking the average values as the reliable target year climate characteristics of the highway network of the area to be predicted under the prediction scene.

In a possible implementation manner, the obtaining the regional highway network risk prediction result of the regional highway network to be predicted in the prediction scenario according to the regional highway network risk coefficient according to the climate change of each space unit includes:

Taking the space units with the regional highway network risk coefficient larger than a preset regional highway network risk coefficient threshold value as important attention space units according to the regional highway network risk coefficient of the climate change of each space unit;

accumulating the risk coefficients of the regional highway network by the climate change of each space unit to obtain the overall risk coefficient of the regional highway network to be predicted;

and determining the road network risk prediction result under the prediction scene according to the important attention space unit and the overall risk coefficient.

In a second aspect, an embodiment of the present invention provides a risk prediction apparatus for a regional highway network by climate change, where the apparatus includes:

the acquisition module is used for acquiring the regional information of the regional highway network to be predicted and disaster influence factors;

the simulation module is used for collecting the reference year observation data of the highway network of the area to be predicted according to the disaster influence factors; selecting target global climate mode data, and simulating the reference annual climate characteristics of the highway network of the area to be predicted according to the target global climate mode data; determining a prediction scene of climate change on regional highway network risks corresponding to the regional highway network to be predicted according to the target global climate mode data, and predicting target annual climate characteristics of the regional highway network to be predicted under the prediction scene; inputting the reference year observation data, the reference year climate characteristics and the target year climate characteristics under the prediction scene into a pre-trained space-time sequence prediction model for correction operation, so as to obtain reliable target year climate characteristics of the highway network of the area to be predicted under the prediction scene; determining a climate characteristic change result according to the difference value of the reference year climate characteristic and the reliable target year climate characteristic; determining a climate characteristic space distribution result of the regional highway network to be predicted based on the climate characteristic change result;

The prediction module is used for dividing the highway network of the area to be predicted to obtain a plurality of space units; interpolating the climate characteristic space distribution result into the plurality of space units to obtain a climate change risk coefficient of each space unit under the prediction scene; calculating the risk coefficient of the climate change of each space unit according to the regional information to obtain the risk coefficient of the regional highway network by the climate change of each space unit; and obtaining a regional highway network risk prediction result of the regional highway network to be predicted under the prediction scene according to the regional highway network risk coefficient of the climate change of each space unit.

In a third aspect, an embodiment of the present invention provides a computer device, where the computer device includes a processor and a nonvolatile memory storing computer instructions, where the computer instructions, when executed by the processor, perform a method for predicting risk of a regional highway network by climate change as described in at least one possible implementation manner of the first aspect.

In a fourth aspect, an embodiment of the present invention provides a readable storage medium, where the readable storage medium includes a computer program, where the computer program controls a computer device where the readable storage medium is located to execute a method for predicting risk of a regional highway network by using climate change as described in at least one possible implementation manner of the first aspect.

Compared with the prior art, the invention has the beneficial effects that: by adopting the regional highway network risk prediction method by the climate change and the readable storage medium, regional highway network information to be predicted and disaster influence factors are obtained; collecting reference year observation data; selecting and simulating global climate mode data; determining a predicted scenario and predicting the climate characteristics of a target year; performing correction operation through a space-time sequence prediction model to obtain reliable target annual climate characteristics; determining a climate characteristic change result; dividing the region to be predicted, and interpolating to obtain the climate change risk coefficient of each space unit; and finally, obtaining a regional highway network risk prediction result of the regional highway network to be predicted according to the risk coefficient.

By the design, a scientific and effective evaluation mode is provided, and the capability of the highway network for coping with climate change is improved.

Drawings

In order to more clearly illustrate the technical solution of the embodiments of the present invention, the drawings that are required to be used in the embodiments will be briefly described.

It is appreciated that the following drawings depict only certain embodiments of the invention and are therefore not to be considered limiting of its scope.

Other relevant drawings may be made by those of ordinary skill in the art without undue burden from these drawings.

FIG. 1 is a schematic flow chart of steps of a method for predicting regional highway network risk by climate change according to an embodiment of the present invention;

FIG. 2 is a schematic block diagram of a climate change versus regional highway network risk prediction device according to an embodiment of the present invention;

fig. 3 is a schematic block diagram of a computer device according to an embodiment of the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the embodiments of the present invention more clear, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

It will be apparent that the described embodiments are some, but not all, embodiments of the invention.

The components of the embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations.

Thus, the following detailed description of the embodiments of the invention, as presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention.

All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.

It should be noted that: like reference numerals and letters denote like items in the following figures, and thus once an item is defined in one figure, no further definition or explanation thereof is necessary in the following figures.

In the description of the present invention, it should be understood that the directions or positional relationships indicated by the terms "upper", "lower", "inner", "outer", "left", "right", etc. are based on the directions or positional relationships shown in the drawings, or the directions or positional relationships conventionally put in place when the product of the application is used, or the directions or positional relationships conventionally understood by those skilled in the art are merely for convenience of describing the present invention and simplifying the description, and do not indicate or imply that the apparatus or elements to be referred to must have a specific direction, be constructed and operated in a specific direction, and thus should not be construed as limiting the present invention.

Furthermore, the terms "first," "second," and the like, are used merely to distinguish between descriptions and should not be construed as indicating or implying relative importance.

In the description of the present invention, it should also be noted that, unless explicitly specified and limited otherwise, terms such as "disposed," "connected," and the like are to be construed broadly, and for example, "connected" may be either fixedly connected, detachably connected, or integrally connected; can be mechanically or electrically connected; can be directly connected or indirectly connected through an intermediate medium, and can be communication between two elements.

The specific meaning of the above terms in the present invention can be understood by those of ordinary skill in the art according to the specific circumstances.

The following describes specific embodiments of the present invention in detail with reference to the drawings.

In order to solve the foregoing technical problems in the background art, fig. 1 is a schematic flow chart of a method for predicting regional highway network risk by using climate change according to an embodiment of the present disclosure, and the method for predicting regional highway network risk by using climate change is described in detail below.

Step S201, obtaining regional information of a regional highway network to be predicted and disaster influence factors;

step S202, collecting reference year observation data of the highway network of the area to be predicted according to the disaster influence factors;

Step S203, selecting target global climate mode data, and simulating the reference annual climate characteristics of the highway network of the area to be predicted according to the target global climate mode data;

step S204, determining a prediction scene of the climate change on the regional highway network risk corresponding to the regional highway network to be predicted according to the target global climate mode data, and predicting the target annual climate characteristics of the regional highway network to be predicted under the prediction scene;

step S205, inputting the reference year observation data, the reference year climate characteristics and the target year climate characteristics under the predicted scene into a pre-trained space-time sequence prediction model for correction operation, and obtaining reliable target year climate characteristics of the highway network of the area to be predicted under the predicted scene;

step S206, determining a climate characteristic change result according to the difference value of the reference year climate characteristic and the reliable target year climate characteristic;

step S207, determining a climate characteristic space distribution result of the highway network of the area to be predicted based on the climate characteristic change result;

step S208, dividing the highway network of the area to be predicted to obtain a plurality of space units; interpolating the climate characteristic space distribution result into the plurality of space units to obtain a climate change risk coefficient of each space unit under the prediction scene; calculating the risk coefficient of the climate change of each space unit according to the regional information to obtain the risk coefficient of the regional highway network by the climate change of each space unit;

Step S209, obtaining a regional highway network risk prediction result of the regional highway network to be predicted under the prediction scenario according to the regional highway network risk coefficient of the climate change of each space unit.

In the embodiment of the invention, it is assumed that the risk of highway networks in a certain island city is predicted, and first, regional information of the city and factors possibly affecting the highway networks need to be collected.

The regional information may include the geographic location of the city, road network structure, altitude, etc.

And disaster influencing factors may include typhoon frequency, rainfall, sea level rise, etc.

Based on previously collected disaster impact factors, during the baseline year we need to acquire observations related to these factors.

Taking rainfall as an example, we will collect rainfall data, including rainfall size, duration, etc., for each season, month or specific time period in the reference year for the island city.

A suitable global climate pattern dataset, such as CMIP (Coupled Model Intercomparison Project) data, is selected and simulated based on the data to obtain climate characteristics of the island in the sea city over the reference year.

For example, the distribution of meteorological elements such as rainfall, air temperature, and wind speed in a reference year is simulated.

Using the selected global climate pattern data, different climate change scenarios are determined, such as RCP4.5 (Representative Concentration Pathway 4.5) and RCP8.5, etc.

Then, future climate change trends of island cities, such as increase or decrease of rainfall, sea level rise, etc., are predicted from these scenes, and target annual climate characteristics of the urban road network, such as road wet skid degree, road surface breakage, etc., are predicted in these scenes.

Let us assume that we use a neural network as a spatiotemporal sequence prediction model and take as input the reference year observation, the reference year climate characteristics, and the target year climate characteristics in the predicted scenario.

By training and correcting the data, the model can learn the reliable target annual climate characteristics of the regional highway network to be predicted under the prediction scene, such as the variation trend of average rainfall in a month in the future.

By calculating the difference between the baseline annual climate characteristic and the reliable target annual climate characteristic, the outcome of the change in the different climate characteristics may be determined.

For example, if the reference annual rainfall is 100 mm and the reliable target annual rainfall is predicted to be 120 mm, a change result of the increase in rainfall can be obtained.

And according to the change result of the climate characteristics, the climate characteristic distribution condition of each region of the highway network of the region to be predicted can be determined.

For example, from the result of the change in the increase in rainfall, it can be inferred that certain areas may experience more rainfall, while certain areas may be less affected.

Dividing the regional highway network to be predicted according to geographic positions or other suitable standards to form a plurality of space units.

For example, a highway network in an island city is divided into multiple areas or segments, north, south, coastal, inland, etc.

It is assumed that we have obtained the spatial distribution of the climate characteristics of the regional highway network to be predicted and divided into four spatial units, north, south, coastal and inland.

And applying the spatial distribution result of the climate characteristics to each space unit by an interpolation method (such as Kriging interpolation), thereby obtaining the climate change risk coefficient of each space unit under the predicted scene.

For example, a risk factor for increased rainfall in north areas, or for increased sea level in coastal areas, may be calculated.

By combining the regional information and the climate change risk coefficient, calculation or model derivation can be performed to determine the influence degree of the climate change of each space unit on the regional highway network risk.

For example, the influence degree of rainfall increase on the road network risk of the region is calculated according to the risk coefficient of rainfall increase of the region in North and the factors such as road drainage capacity, soil stability and the like of the region.

According to the climate change of each space unit, the risk coefficient of the regional highway network can be comprehensively calculated or statistically analyzed to obtain the risk level prediction result of the regional highway network to be predicted under the prediction scene.

For example, by combining the risk coefficient of increased rainfall in the north region, the risk coefficient of reduced rainfall in the south region and the like, the risk level prediction result of the whole island city highway network under the prediction scene can be obtained.

In one possible implementation, the aforementioned step S202 may be performed by the following steps.

(1) Acquiring historical disaster record data and field disaster investigation data corresponding to the highway network of the area to be predicted;

(2) Analyzing and determining the disaster influence factors aiming at the highway network of the area to be predicted according to the historical disaster record data and the site disaster investigation data;

(3) And collecting reference year observation data which are related based on the disaster influence factors in the highway network of the area to be predicted.

In the embodiment of the invention, it is assumed that we want to predict disasters on a highway network of a certain city, and first need to collect recorded data of past disaster events of the city, such as earthquake, flood, landslide, etc.

These records may be from disaster resistant institutions or other reliable disaster databases.

Meanwhile, on-site investigation is required, and professional teams are dispatched to investigate the road conditions of the region in the field, including road quality, traffic flow, existing risk points and the like.

Based on the acquired historical disaster record data and on-site disaster investigation data, data analysis and statistics are performed to determine factors related to the road network disasters.

For example, by analysis of historical flood events and field investigation data, it may be found that factors such as rainfall, drainage system conditions, river distance, etc. are related to highway flood disasters.

After determining factors related to road network disasters, acquisition of observation data is required for the factors.

For example, if rainfall is determined to be one of the key factors, rainfall data for many years in the relevant region needs to be collected.

Such data may come from weather stations, water departments, or other reliable sources of data.

Reference year observations refer to reference data used to build models and compare future observations.

In one possible implementation, the aforementioned step S203 may be implemented by performing the following steps.

(1) Acquiring a climate pattern dataset comprising a plurality of global climate pattern data;

(2) Simulating the climate mode data set based on the disaster influence factors to obtain simulation scores of each global climate mode data for the disaster influence factors;

(3) Selecting global climate mode data with highest simulation score as the target global climate mode data;

(4) And simulating the reference year climate characteristics of the highway network of the area to be predicted according to the target global climate mode data.

In embodiments of the present invention, a data set comprising a plurality of global climate pattern data may be obtained by cooperation with a weather research institute or an international weather organization.

These data sets may include meteorological data from different regions and for different time periods, covering various meteorological indicators such as temperature, rainfall, wind speed, etc.

The acquired climate pattern dataset is simulated according to the determined disaster influencing factors (such as rainfall, temperature and the like).

For example, using numerical models and statistical methods, the behavior of different global climate pattern data on specific disaster impact factors is simulated, and a corresponding simulation score is generated for each model to evaluate its simulation capabilities.

And selecting a model with the best performance from the plurality of global climate mode data as target global climate mode data according to the simulation score.

In general, models with higher simulation scores are considered more accurate and reliable in predicting specific disaster impact factors.

And simulating the reference annual climate characteristics of the highway network of the area to be predicted by using the selected target global climate mode data.

For example, according to the mode data, weather indexes such as average temperature, rainfall and the like in a future period of time of the area are simulated, and a corresponding weather characteristic model is established.

In one possible embodiment, the step of simulating the reference annual climate characteristic of the regional highway network to be predicted according to the target global climate pattern data may be implemented as follows.

In the embodiment of the invention, the target global climate pattern data, the reference year and the grid information of the regional highway network to be predicted are input into a WRF-FDDA (Weather Research and Forecasting-Forward Domain Data Assimilation) model for simulation, so that the reference year climate characteristics of the regional highway network to be predicted are obtained.

Suppose we are making disaster predictions for a highway network of a city and have selected a target global climate pattern data.

We also have grid information of the road network of the urban area to be predicted, including the position, shape, attribute, etc. of the roads.

First, we input the selected target global climate pattern data into the WRF-FDDA model along with mesh information for the regional highway network to be predicted.

WRF-FDDA is a weather research and forecasting model that may be used to simulate weather phenomena in a particular area.

Then we specify the year of the reference year, in order to ensure that the climate characteristics we simulate are based on a specific period of time.

For example, we can choose a certain year in the past as the reference year.

Next, the WRF-FDDA model will perform a simulation calculation using the target global climate pattern data and the reference year information, as well as the grid information of the regional highway network to be predicted.

The model may consider a number of factors such as geography, weather, and road grids to generate reference year climate characteristics corresponding to the urban road network.

Finally, we can obtain simulation results from the WRF-FDDA model, including various meteorological indicators such as temperature, rainfall, wind speed, etc. during the reference year.

These simulated baseline annual climate characteristics will provide an important reference for subsequent road network disaster predictions and risk assessment.

In one possible implementation, the aforementioned step S204 may be implemented by the following detailed implementation.

(1) Determining a prediction scene of the climate change on the regional highway network risk corresponding to the regional highway network to be predicted according to the target global climate mode data;

(2) Inputting the target year, the grid information of the highway network of the area to be predicted and the target global climate pattern data of the target year section under the prediction scene into a WRF model, simulating to obtain the climate characteristics of the target year section under the prediction scene, and taking the climate characteristics of the target year section under the prediction scene as the target year climate characteristics, wherein the target year section comprises a target year, the previous year of the target year and the next year of the target year.

In an embodiment of the invention, for example, assume we are researching a highway network for a city and selecting specific target global climate pattern data.

By analyzing the pattern data and associated climate indicators (e.g., temperature, rainfall, etc.), we can determine different climate change scenarios such as elevated temperature, increased rainfall, increased frequency of extreme weather events, etc.

Inputting WRF (Weather Research and Forecasting) model into the grid information of the road network of the area to be predicted and the global climate pattern data of the target year section under the predicted scene, simulating to obtain the climate characteristics of the target year section under the predicted scene, and taking the climate characteristics of the target year section under the predicted scene as the climate characteristics of the target year, wherein the target year section comprises the target year, the previous year of the target year and the next year of the target year.

Suppose we are researching a highway network for a city and wish to predict the impact of the climate for the next few years (2024 to 2026) on highway risk.

We have selected target global climate pattern data that provides global weather predictions.

First, based on the target global climate pattern data and the geographic location and characteristics of the city, we have determined possible climate change scenarios, such as predicted temperature increases and changes in rainfall.

Then, we set the target year to 2025 and input mesh information of the regional highway network to be predicted (including the position, shape, attribute, etc. of the road) and target global climate pattern data of the target year in the predicted scenario into the WRF model for simulation.

The WRF model takes the input climate data and road network information into consideration, and simulates and derives climate characteristics, such as average temperature, rainfall distribution, and the like, of the target annual segments (2024, 2025, and 2026) in the predicted scenario.

These simulation results provide the climate conditions of the regional highway network to be predicted in the prediction scenario.

Finally, we will get the climate characteristics of the target year segment under the predicted scenario as the climate characteristics of the target year (2025).

These features can be used to further analyze highway risks, such as assessing the probability of occurrence and extent of impact of events such as floods, road freezes, or extreme temperatures.

In summary, in step S204, the target year is set to 2025 according to the target global climate pattern data and the grid information of the highway network of the area to be predicted, and the climate characteristics of the target years (2024, 2025, and 2026) under the predicted scenario are obtained by using WRF model simulation.

These features provide a detailed understanding of the climate change in the coming years in the context of road network risk prediction, supporting the performance of road network disaster prediction and risk assessment.

In one possible implementation, the aforementioned step S205 may be implemented by the following detailed steps.

(1) Inputting the reference year observation data, the reference year climate characteristics and the target year climate characteristics under the prediction scene into the space-time sequence prediction model constructed by the convolutional neural network and the long-short-time neural network for correction operation, so as to obtain the climate characteristics respectively corresponding to the target year, the previous year of the target year and the next year of the target year;

(2) And calculating to obtain the average values of the climate characteristics corresponding to the target year, the previous year of the target year and the next year of the target year respectively, and taking the average values as the reliable target year climate characteristics of the highway network of the area to be predicted under the prediction scene.

In the embodiments of the present invention, it is assumed that we have collected baseline annual observation data and climate characteristics of the regional highway network to be predicted.

At the same time, using the WRF model in step S204, we also obtain the target annual climate characteristics in the predicted scenario.

Now, we input these data into a spatio-temporal sequence prediction model constructed from Convolutional Neural Networks (CNNs) and long-short-time neural networks (LSTM).

This model is designed to process the time series data and to be able to learn and predict the trend of the change in the climate characteristics.

Through correction operation, the model can utilize reference year observation data and reference year climate characteristics, and combine the target year climate characteristics under the prediction scene to correct and predict the climate characteristics of the target year, the previous year of the target year and the next year of the target year.

In this way, we obtain corrected climate characteristic data corresponding to three years for further analysis and calculation.

Based on the corrected climate characteristic data of the target year, the previous year of the target year and the next year of the target year, we can obtain a more stable and reliable target year climate characteristic by calculating the average value of the climate characteristics corresponding to the years.

The climate characteristics of three years are averaged, so that the influence of possible abnormal values or noise can be reduced, and the reliability of the result is improved.

This average value reflects the average climate conditions in which the regional road network to be predicted is located in the prediction scenario.

For example, assume that the climate characteristics of the corrected target year, the previous year of the target year, and the next year of the target year are X1, X2, and X3, respectively.

By calculating (x1+x2+x3)/3, we obtain the result as a reliable target annual climate characteristic of the regional highway network to be predicted in the predicted scenario.

In summary, according to this possible embodiment, in step S205, we input the reference year observation data, the reference year climate feature, and the target year climate feature in the predicted scenario into the spatio-temporal sequence prediction model to perform the correction operation, so as to obtain the climate features corresponding to the target year, the previous year of the target year, and the next year of the target year, respectively.

And then, calculating the average value of the weather characteristics corresponding to the years to obtain the reliable target year weather characteristics of the road network of the area to be predicted under the prediction scene.

In one possible implementation, the aforementioned step S209 may be performed by the following steps.

(1) Taking the space units with the regional highway network risk coefficient larger than a preset regional highway network risk coefficient threshold value as important attention space units according to the regional highway network risk coefficient of the climate change of each space unit;

(2) Accumulating the risk coefficients of the regional highway network by the climate change of each space unit to obtain the overall risk coefficient of the regional highway network to be predicted;

(3) And determining the road network risk prediction result under the prediction scene according to the important attention space unit and the overall risk coefficient.

In the embodiment of the invention, it is assumed that we are analyzing the road network of a city and have calculated the risk factors of the climate change of different space units for the regional road network.

In this step, we will screen out the spatial units with risk coefficients higher than the threshold according to the preset regional highway network risk coefficient threshold, and these spatial units are considered as the regions needing important attention.

For example, we divide a space cell into grids, each representing a space cell.

If the risk factor caused by the climate change of a certain grid exceeds a preset risk factor threshold, the grid is marked as a space unit of great interest.

The risk coefficients of the regional highway network are calculated in an accumulated mode according to the climate change of each space unit, so that the overall risk coefficient of the regional highway network to be predicted is obtained.

For example, for each space unit of great interest, we will add risk coefficients corresponding to its climate change to obtain an accumulated value.

In this way, a risk coefficient comprehensively considering each space unit can be obtained, and the overall risk level of the whole highway network in the area to be predicted under the prediction scene is reflected.

For space units of great interest, we can further analyze the specific values and the trend of the risk coefficients thereof to determine the risk level of these regions in the predicted scenario.

This provides detailed risk assessment information for a particular region.

Secondly, through the overall risk coefficient, the overall risk level of the whole regional highway network to be predicted under the prediction scene can be estimated.

This result can be used as a reference for decision making and planning, helping the relevant departments to take appropriate action to address the potential risk.

Illustratively, in the predictive scenario we focus on the risk of locating a major highway intersection area and consider that this area may be subject to a higher climate-related risk.

Meanwhile, the overall risk coefficient is 0.75, which indicates that the overall regional highway network to be predicted has a higher comprehensive risk level under the prediction scene.

This means that in a predictive scenario, the relevant departments need to enhance the monitoring and maintenance work of the intersection area, take appropriate measures to reduce the potential risk impact, and should make finer planning and improvement of the road network for the whole area to increase the overall coping capacity.

In a word, by determining the road network risk prediction result under the prediction scene according to the important attention space unit and the overall risk coefficient, targeted information can be provided for decision making and planning, and the relevant departments can be helped to make effective coping strategies.

The following provides an overall implementation of an embodiment of the invention under another implementation.

(1) The area of investigation is determined.

The area of investigation (i.e., the area highway network to be predicted) is determined and area highway network information (i.e., the area information, which may include highway location, length, grade, etc.) is collected.

The research area needs to cover highway network and meteorological stations, and can be China or a certain area (Ji, chang triangle, zhu triangle and the like) or a certain province (Jiangsu, shanxi and the like).

(2) And determining the main influence factor weight of the climate change on the regional highway network.

Collecting main disaster condition data (i.e. historical disaster record data) of a highway network in a research area, combing main influence factors (i.e. disaster influence factors such as temperature, precipitation, wind speed and the like) of climate change on the highway network in the area by utilizing a literature analysis and field investigation method (i.e. field disaster investigation data), and respectively determining weights Y by adopting an expert scoring method _T 、Y _R 、Y _W Etc.

Wherein, the temperature is represented by T, the precipitation is represented by R, the wind speed is represented by W, and the temperature can be correspondingly expanded on the basis of the temperature, Y _T +Y _R +Y _W +Y _… =1。

(3) And constructing a climate model of the research area, and predicting climate change characteristics of the area.

(3.1) screening global climate pattern data.

Collecting daily weather station observation data (namely reference year observation data such as temperature, precipitation, wind speed and the like) of a reference year (2014) of a research area by combining the main influence factors (such as temperature, precipitation, wind speed and the like) of the step (2); the CMIP6 dataset (i.e., the climate pattern dataset including 26 climate patterns) was obtained, the simulation data of the weather station for 26 climate patterns in 2014 was interpolated and evaluated for 26 using the quantitative index S.

The modeling capability of the climate mode is that the index S is normalized and calculated to obtain each mode rank, and global climate mode data (data of 1 climate mode are selected) with the best performance, such as EC-Earth3, is selected.

The method comprises the steps of carrying out a first treatment on the surface of the Wherein, R is a correlation coefficient between an analog value (temperature, precipitation, wind speed) and an observed value (temperature, precipitation, wind speed), note that the meaning of letters related to the related indexes of R and subsequent precipitation is different herein, so as to be interpreted as follows; r is R ₀ Is the maximum correlation coefficient that can be achieved, i.e., the maximum of the correlation coefficients in all modes;is the standard deviation of analog value (STD _m ) And the standard deviation of observed values (STD _o ) Is a ratio of (2).

The closer the analog value is to the observed value data, the closer S is to 1, and the better the mode simulation capability.

The CMIP6 data set comprises 26 weather pattern simulated daily precipitation and air temperature data, wherein the data comprises 1960-2014 standard period simulation data and 4 different future prediction scenes to 2100 years future simulation data, and the 4 future scenes are SSP 1-2.6, SSP 2-4.5, SSP 3-7.0 and SSP 5-8.5 respectively.

(3.2) simulating the climate characteristics of the reference annual region based on the WRF model.

Based on the global climate pattern data EC-Earth3 determined in step 3.1, reference year zone climate characteristics (i.e. reference year climate characteristics) are simulated.

Reference years (2014) and study area grid information (for example, 121×151 of warp×weft grid number and 3km of warp×weft spatial resolution) are determined, 2014 EC-Earth3 global climate pattern data is used as analysis data for driving a WRF model, and study area reference years (2014 years) day-by-day weather station observation data (such as temperature, precipitation, wind speed and the like) are used as input of an FDDA assimilation system in a WRF model, the information such as WRF mode time, nesting, spatial resolution, terrain data category, regional center longitude and latitude, projection mode, regional grid range, physical process parameters and the like are set, environment variables and various parameters required by the FDDA assimilation system, and the day-by-day temperature T of a research region 2014 is simulated _{2014, day} Precipitation R _{2014, day} Wind speed W _{2014, day} As a result.

Calculating the average value of the temperature, precipitation and wind speed day by day in 2014 as the average temperature T in the reference year ₂₀₁₄ Average precipitation R ₂₀₁₄ Average wind speed W ₂₀₁₄ As a result.

The FDDA is to add an additional term, namely a relaxation term of the difference between the simulation value and the observed value, in one or more prediction equations at the appointed moment of numerical mode integration, and enable the solution of the equation to approach the observed value in the time period of the observed data, so that the dynamic balance among the physical quantities in each integration step is ensured, and the obtained mode solution is used as the initial value of simulation, thereby improving the simulation effect of the mode.

(3.3) determining a predicted scenario.

Based on the global climate pattern data EC-Earth3 determined in step 3.1, a future scenario (i.e. predicted scenario) j (e.g. SSP 1-2.6, SSP 2-4.5, SSP 3-7.0) is determined, which predicts the risk of climate change to the regional highway network.

3.4 The regional climate characteristics of different scenarios j are predicted based on the WRF model.

Determining a staged target year N (for example, 2050), researching regional grid information (for example, 121 x 151 of grid number in warp direction and weft direction, 3km of spatial resolution in warp direction and weft direction, 3km of grid number in weft direction and weft direction), respectively using global climate mode data of N-1, N, N +1 years EC-Earth3 in SSP 1-2.6, SSP 2-4.5 and SSP 3-7.0 scenes as analysis data for driving a WRF model, setting WRF mode time, nesting, spatial resolution, terrain data category, regional center longitude and latitude, projection mode, regional grid range, physical process parameter and other information, and simulating N-1, N, N +1 year day-by-day temperature (T) of different scenes j of a research region _{N-1, day, j} 、T _{N, day, j} 、T _{N+1, day, j} ) Precipitation (R) _{N-1, day, j} 、R _{N, day, j} 、R _{N+1, day, j} ) Wind speed (W) _{N-1, day, j} 、W _{N, day, j} 、W _{N+1, day, j} ) As a result.

(3.5) correcting the regional climate characteristics of different scenes j (average temperature, average precipitation, average wind speed results of the target year N of the research region) by using a deep learning algorithm.

The observation data of the temperature, precipitation and wind speed of a weather station day by day in a reference year (2014) in a research area collected in the step (3.1), the simulation data in the step (3.2) and the simulation result in the step (3.4) are taken as input, a space-time sequence prediction model is built by combining a Convolutional Neural Network (CNN) (space module) and a long-short-time neural network (LSTM) (time module) algorithm, the day-by-day temperature, precipitation and wind speed correction values of different scenes j in N-1 and N, N +1 years are output, the day-by-day temperature, precipitation and wind speed average value of the different scenes j in N-1 and N, N +1 years after correction is calculated and taken as the average temperature T of different scenes j in the target year N _CN，j Average precipitation R _CN，j Average wind speed W _CN，j As a result.

(3.6) predicting the regional climate change characteristics of different scenarios j.

Correcting the result (T) of the target year N of the different scenes j in the step (3.5) _CN，j 、R _CN，j 、W _CN，j ) Results of the reference year simulation (T) with (3.2) ₂₀₁₄ 、R ₂₀₁₄ 、W ₂₀₁₄ ) Taking difference to obtain the temperature, precipitation and wind speed change results from the reference year to the target year of different scenes j of the research area, and outputting the temperature, precipitation and wind speed change (T) _j 、R _j 、W _j ) Is a result of the spatial distribution of (a).

The global climate mode is the only tool for simulating historical climate states and predicting future climate changes, and can well simulate and predict climate states with a spatial scale of 100-500 km and a time scale of seasons-years, but cannot well represent climate characteristics and dynamics with finer spatial and temporal scales.

In order to obtain finer regional scale climate change characteristics, a dynamic downscaling method is needed to construct a WRF numerical model, the regional mode is driven by taking output or analysis data of the global climate mode as an initial field and boundary conditions, and climate simulation of a research region is carried out to obtain more accurate and detailed climate characteristics of the research region.

(4) And determining the climate change risk spatial distribution of different scenes j of the research area.

Dividing a research area into space units i by adopting a GIS space analysis method based on the research area grid information determined in the step (3.2);

Changing the temperature, precipitation and wind speed (T) from the reference year to the target year N of different scenes j in the step (3.6) _j 、R _j 、W _j ) Is interpolated to a spatial unit T _i，j 、R _i，j 、W _i，j ；

Will T _i，j 、R _i，j 、W _i，j Performing positive normalization calculation to obtain TA _i，j 、RA _i，j 、WA _i，j The method comprises the steps of carrying out a first treatment on the surface of the It should be appreciated that changes in temperature, precipitation, wind speed (T _j 、R _j 、W _j ) The magnitude and unit of the factors are different, for example, the temperature is-1-10 degrees, the precipitation is 100mm, and the wind speed is 10m/s, so parameter normalization processing is recommended.

In the invention, the larger the temperature, precipitation, wind speed and the like, the larger the influence on the highway infrastructure is, and therefore, the invention is just normalized.

Firstly dividing space units, and carrying out normalization processing on parameter values on the space units i.

Determining the climate change risk coefficient C of each space unit of different scenes j of the research area by combining the weight coefficients determined in the step (2) _i,j 。

C _i,j =TA _i，j *Y _T +RA _i，j *Y _R +WA _i，j *Y _W

The larger the statistical index is by adopting the method of 'normal one', the labelThe larger the index value after normalization is, the following formula is:；/>；/>the method comprises the steps of carrying out a first treatment on the surface of the Wherein TA _i，j The temperature value after the standardization of the i space unit under the j scene is represented; t (T) _i，j The temperature value of the i space unit in the j scene is represented; t (T) _min,j Representing T in j scenario _j A minimum value; t (T) _max,j Representing T in j scenario _j Maximum value.

The formulas RA and WA may refer to the TA-related formulas, but the symbols are not identical, and will not be described in detail herein.

(5) And predicting the regional highway network risk caused by climate change of different scenes j.

Interpolating the road network information of step 1 into space units, and calculating the road area S of each space unit _i Road grade P _i ；

Calculating risk coefficient B of climate change of each space unit of different scenes j on regional highway network _i,j ；

B _i,j =S _i *C _i，j / P _i The method comprises the steps of carrying out a first treatment on the surface of the And identifying key areas of the regional highway network risk caused by different scene climate changes.

Or B is to _i,j And the overall risk coefficient of the regional highway network caused by the climate change of different scenes j of the research region is obtained by superposition, so that scientific guidance is provided for improving the capability and the level of the highway network for coping with the climate change, promoting the construction of the highway network and constructing a flexible sustainable traffic system.

In summary, the method for predicting the regional highway network risk by the climate change provided by the invention adopts methods such as statistical analysis, numerical simulation, deep learning algorithm, space recognition and the like to predict the regional highway network risk by the climate change, fills up the technical blank of predicting the regional highway network risk based on future climate change scenes, and provides scientific guidance for improving the capability and the level of the highway network for coping with the climate change, promoting the construction of the highway network and constructing a flexible sustainable traffic system.

Referring to fig. 2 in combination, an embodiment of the present invention provides a climate change versus regional highway network risk prediction apparatus 110, which includes: the obtaining module 1101 is configured to obtain regional information of a highway network in a region to be predicted and disaster influencing factors;

the simulation module 1102 is configured to collect reference year observation data of the highway network in the area to be predicted according to the disaster influence factor; selecting target global climate mode data, and simulating the reference annual climate characteristics of the highway network of the area to be predicted according to the target global climate mode data; determining a prediction scene of climate change on regional highway network risks corresponding to the regional highway network to be predicted according to the target global climate mode data, and predicting target annual climate characteristics of the regional highway network to be predicted under the prediction scene; inputting the reference year observation data, the reference year climate characteristics and the target year climate characteristics under the prediction scene into a pre-trained space-time sequence prediction model for correction operation, so as to obtain reliable target year climate characteristics of the highway network of the area to be predicted under the prediction scene; determining a climate characteristic change result according to the difference value of the reference year climate characteristic and the reliable target year climate characteristic; determining a climate characteristic space distribution result of the regional highway network to be predicted based on the climate characteristic change result;

The prediction module 1103 divides the highway network of the area to be predicted to obtain a plurality of space units; interpolating the climate characteristic space distribution result into the plurality of space units to obtain a climate change risk coefficient of each space unit under the prediction scene; calculating the risk coefficient of the climate change of each space unit according to the regional information to obtain the risk coefficient of the regional highway network by the climate change of each space unit; and obtaining a regional highway network risk prediction result of the regional highway network to be predicted under the prediction scene according to the regional highway network risk coefficient of the climate change of each space unit.

It should be noted that, the implementation principle of the regional highway network risk prediction device 110 by the foregoing climate change may refer to the implementation principle of the regional highway network risk prediction method by the foregoing climate change, which is not described herein again.

It should be understood that the division of the modules of the above apparatus is merely a division of a logic function, and may be fully or partially integrated into a physical entity or may be physically separated when actually implemented.

And these modules may all be implemented in software in the form of calls by the processing element; or can be realized in hardware; the method can also be realized in a form of calling software by a processing element, and the method can be realized in a form of hardware by a part of modules.

For example, the modules above may be one or more integrated circuits configured to implement the methods above, such as: one or more specific integrated circuits (application specific integrated circuit, ASIC), or one or more microprocessors (digital signal processor, DSP), or one or more field programmable gate arrays (field programmable gate array, FPGA), or the like.

For another example, when a module above is implemented in the form of a processing element scheduler code, the processing element may be a general purpose processor, such as a central processing unit (centralprocessing unit, CPU) or other processor that may invoke the program code.

For another example, the modules may be integrated together and implemented in the form of a system-on-a-chip (SOC).

The embodiment of the invention provides a computer device 100, wherein the computer device 100 comprises a processor and a nonvolatile memory storing computer instructions, and when the computer instructions are executed by the processor, the computer device 100 executes the method for predicting the risk of the regional highway network by the climate change.

As shown in fig. 3, fig. 3 is a block diagram of a computer device 100 according to an embodiment of the present invention.

The computer device 100 comprises a climate change versus regional highway network risk prediction means 110, a memory 111, a processor 112 and a communication unit 113.

For data transmission or interaction, the memory 111, the processor 112 and the communication unit 113 are electrically connected to each other directly or indirectly.

For example, the elements may be electrically connected to each other via one or more communication buses or signal lines.

The climate change versus regional highway network risk prediction apparatus 110 includes at least one software function module that may be stored in the memory 111 in the form of software or firmware (firmware) or cured in an Operating System (OS) of the computer device 100.

The processor 112 is configured to execute the climate change versus regional highway network risk prediction device 110 stored in the memory 111, for example, a software function module and a computer program included in the climate change versus regional highway network risk prediction device 110.

The embodiment of the invention provides a readable storage medium, which comprises a computer program, wherein when the computer program runs, computer equipment in which the readable storage medium is controlled to execute the method for predicting the risk of the regional highway network by the climate change.

The foregoing description, for purpose of explanation, has been presented with reference to particular embodiments.

However, the illustrative discussions above are not intended to be exhaustive or to limit the disclosure to the precise forms disclosed.

Many modifications and variations are possible in light of the above teaching.

The embodiments were chosen and described in order to best explain the principles of the disclosure and its practical application, to thereby enable others skilled in the art to best utilize the disclosure and various embodiments with various modifications as are suited to the particular use contemplated.

Many modifications and variations are possible in light of the above teaching.

Claims

1. A method for predicting regional highway network risk by climate change, comprising the steps of:

determining a prediction scene of climate change on regional highway network risks corresponding to the regional highway network to be predicted according to the target global climate mode data, and predicting target annual climate characteristics of the regional highway network to be predicted under the prediction scene;

2. The method according to claim 1, wherein the collecting the reference year observation data of the highway network of the area to be predicted according to the disaster influence factors comprises:

3. The method of claim 1, wherein the selecting the target global climate pattern data and simulating the reference annual climate characteristics of the regional highway network to be predicted based on the target global climate pattern data comprises:

4. A method according to claim 3, wherein said simulating the reference annual climate characteristics of the regional highway network to be predicted from the target global climate pattern data comprises:

5. The method according to claim 4, wherein determining a prediction scenario of a risk of a regional highway network corresponding to the regional highway network to be predicted from the target global climate pattern data, and predicting a target annual climate characteristic of the regional highway network to be predicted in the prediction scenario, comprises:

6. The method according to claim 5, wherein said inputting the reference year observation data, the reference year climate characteristics and the target year climate characteristics in the predicted scenario into a pre-trained spatiotemporal sequence prediction model for correction operation, to obtain reliable target year climate characteristics of the highway network in the predicted scenario in the area to be predicted, comprises:

7. The method according to claim 1, wherein the obtaining the regional highway network risk prediction result for the regional highway network to be predicted in the prediction scenario according to the climate change of each spatial unit includes:

8. A climate change versus regional highway network risk prediction apparatus, the apparatus comprising:

9. A computer device comprising a processor and a non-volatile memory storing computer instructions which, when executed by the processor, perform the climate change versus regional highway network risk prediction method of any of claims 1-7.

10. A readable storage medium, characterized in that the readable storage medium comprises a computer program, which when run controls a computer device in which the readable storage medium is located to perform the climate change versus regional highway network risk prediction method according to any of claims 1-7.