A method for visualizing and evaluating ground subsidence risk of urban underground road shield excavation based on GIS and InSAR

Abstract

The application discloses a kind of based on GIS and InSAR's city underground road shield excavation Ground subsidence risk visualization assessment and prediction method, and city ground object grading and intelligent annotation are realized by the multi-source data of GIS and InSAR, using the improved filtering method to extract local subsidence and combine related subsidence influence factor to build susceptibility multi-classification evaluation dataset, in combination with artificial intelligence algorithm, according to the established dataset, realize city subsidence susceptibility multi-classification evaluation, and then based on risk matrix, with susceptibility multi-classification as possibility, city ground object importance classification as severity, realize the visualization assessment and prediction of settlement risk for city underground road construction.The advantage of the present application is to realize the visualization assessment and prediction of settlement risk for underground road construction, improve the rationality of line planning and the safety of construction period.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of civil engineering, in particular to a ground subsidence risk visualization evaluation and prediction method for urban underground road shield excavation based on GIS and InSAR. BACKGROUND

[0002] With the rapid development of cities, the ground space of cities is increasingly scarce, and the development of underground space has become a hot spot for future urban development. Among them, underground roads, as the core component connecting various underground spaces, have developed rapidly in recent years. However, in the high-intensity development of underground roads, the subsidence risk problem in construction has not been given enough attention. The urban surface environment is complex, and there are many underground buried pipelines. Once subsidence disasters occur in the construction of underground roads, it will bring huge economic losses and livelihood problems. Therefore, the ground subsidence evaluation for urban underground road construction, especially the prediction in the planning stage, is of great significance for the safe construction of future urban underground space.

[0003] The current problem is that: at present, there is a lack of consideration of construction period subsidence risk in the planning stage of underground roads. Although the city buildings along the line will be detected during construction, it relies on manual detection, which is not only time-consuming and laborious in the context of a large number of buildings in the city, but also lacks focus, and often cannot be warned until the subsidence disaster occurs initially, which is seriously lagging. The existing method of filtering and extracting local subsidence has certain limitations. When the concave-convex nature of the radial average power curve is not significant, there are difficulties in determining the cut-off frequency. In addition, the conventional binary classification subsidence susceptibility evaluation is not conducive to reflecting the susceptibility difference under different subsidence rates. This is the place that needs to be improved in the present application. SUMMARY

[0004] The technical problem to be solved by the present application is to provide a ground subsidence risk visualization evaluation and prediction method for urban underground road shield excavation based on GIS and InSAR, to realize the visualization evaluation and prediction of subsidence risk in underground road construction, and to improve the rationality of line planning and the safety during construction.

[0005] In order to solve the above technical problems, the present application provides a ground subsidence risk visualization evaluation and prediction method for urban underground road shield excavation based on GIS and InSAR, the specific steps are as follows:

[0006] Step S1, a city feature grading and intelligent labeling method fusing GIS and InSAR multi-source data is obtained to obtain the city feature grading result;

[0007] Step S2, an improved filtering method based on InSAR is used to obtain high-precision InSAR subsidence data of the city area, and an improved band-pass filtering method is used to extract local subsidence caused by underground road excavation;

[0008] Step S3, combine the shield construction data, and fuse the ground object classification results, high-precision settlement data and other settlement-related influencing factors to form a city settlement risk data set;

[0009] Step S4, build a city settlement susceptibility multi-classification machine learning algorithm, train according to the city settlement risk data set, realize the visual evaluation of the ground object settlement risk along the shield excavation line, and predict the risk of the excavation section.

[0010] The step S1, the specific steps are as follows:

[0011] Step S11, according to the importance of ground objects in urban development and the compensation ability of other facilities in the city after the occurrence of settlement disasters, the grades of urban ground objects are divided;

[0012] Step S12, fuse GIS satellite image data, InSAR scattering point distribution data and other multi-source data, and use machine vision algorithm to label the ground object data according to the divided urban ground object grades.

[0013] The step S2, the specific steps are as follows:

[0014] The improved band-pass filtering calculation method for extracting local settlement caused by underground road excavation: the frequency spectrum and power spectrum of the settlement data are calculated by two-dimensional Fourier transform; the radial average data of the power spectrum is intercepted, and the cut-off frequency is determined by using nonlinear least squares method according to formula (1); the frequency spectrum is cropped according to the cut-off frequency, and the local settlement caused by underground road excavation is obtained by inverse Fourier transform;

[0015] ;

[0016] In the formula:

[0017] is the function to be fitted,

[0018] , , , , , , , , , is the parameter to be fitted, wherein , The cut-off frequencies of low-pass and high-pass filtering are respectively.

[0019] Step S3 also includes: finding the existing urban underground road excavation cases in the region, and counting the settlement influence factor data along the line in the project as the features of subsequent model training. The settlement influence factor data mainly includes InSAR settlement data, natural influence factors, construction influence factors and urban influence factors. The finally established urban settlement risk data set is saved in the form of GIS grid data, mainly including two categories of feature data and label data.

[0020] The step S4 also includes: the susceptibility model adopts common machine learning models such as random forest and XGBoost to train on the constructed data set. In the label setting, the settlement rate after excavation is divided into multiple settlement levels to constitute the susceptibility label, and the features adopt related settlement influence factors. After completing the model training according to the label, the feature data of the region along the underground road is input, and the risk matrix is constituted by the model output end and the importance classification of urban features, so as to realize the evaluation or prediction of the settlement risk along the underground road under construction.

[0021] The superior efficacy of the present application is that:

[0022] 1) The present application realizes the route safety evaluation in the planning stage and the settlement risk early warning in the construction stage, and improves the rationality of the line design and the safety in the construction stage;

[0023] 2) The improved band-pass filtering calculation method for extracting the local settlement caused by underground road excavation can fit the function with insignificant concave-convex, and obtain the cut-off frequency in one step, realizing the local settlement extraction of underground road construction;

[0024] 3) The improved band-pass filtering calculation method for extracting the local settlement caused by underground road excavation can fit the function with insignificant concave-convex, and obtain the cut-off frequency in one step, realizing the local settlement extraction of underground road construction; BRIEF DESCRIPTION OF DRAWINGS

[0025] The drawings accompanying the specification of the present application serve to provide a further understanding of the present application, and the illustrative embodiments of the present application and their descriptions serve to explain the present application, and do not constitute an improper limitation on the present application. In the drawings:

[0026] Figure 1 is a flowchart of a specific embodiment of the present application;

[0027] Figure 2 is a flowchart of an intelligent feature labeling method combining GIS and InSAR of a specific embodiment of the present application;

[0028] Figure 3This is a flowchart of a method for calculating and extracting local settlement caused by underground road excavation using bandpass filtering, according to an embodiment of the present invention. Detailed Implementation

[0029] The embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0030] like Figure 1 As shown, this invention provides a method for visual assessment and prediction of ground settlement risk during shield tunneling excavation of urban underground roads based on GIS and InSAR. The specific steps are as follows:

[0031] Step S1: The method of urban land cover classification and intelligent annotation is integrated with GIS and InSAR multi-source data to obtain urban land cover classification results; in the process, urban land cover is annotated with multi-source information for subsequent construction of susceptibility multi-level dataset and subsidence risk assessment based on risk matrix.

[0032] Step S11: Classify urban features according to their importance in urban development and the compensatory capacity of other urban facilities after a subsidence disaster.

[0033] The importance of urban features in urban development, from a people-centered perspective, is categorized based on their indispensability to people's lives; the more inaccessible a feature, the higher its ranking. The city's compensatory capacity refers to the ability of other urban facilities to support the function of a feature in the event of subsidence or other disasters; the better the compensatory capacity, the lower the feature's ranking. A combination of these two factors is used to classify urban features.

[0034] Step S12 involves fusing GIS satellite imagery data, InSAR scattering point distribution data, and other multi-source data, and then using machine vision algorithms to label the land cover data according to the classified urban land cover levels. For example... Figure 2 As shown, the map includes a coarse classification of land feature labels and a fine classification of land feature levels. The coarse classification first filters out green areas and river areas in satellite imagery that lack scattering points by using InSAR scattering point distribution, and then labels urban areas using a pre-labeled machine vision model. The fine classification mainly labels land features with indistinct appearances that are difficult for machine vision to distinguish, and finally merges land features of the same level to obtain a land feature level map along the route.

[0035] Step S2: Based on the improved InSAR filtering method, high-precision InSAR settlement data of urban areas are obtained, and the local settlement caused by underground road excavation is extracted by the improved bandpass filtering method.

[0036] like Figure 3The improved bandpass filtering method shown extracts the local settlement caused by underground road excavation, and performs Kriging interpolation on the settlement data of InSAR to complete the missing settlement data; calculates the spectrum and power spectrum of the settlement data by two-dimensional Fourier transform; extracts the radial average data of the power spectrum, and fits it using the nonlinear least squares method according to formula (1) to determine the cutoff frequency; cuts the frequency spectrum according to the cutoff frequency, and obtains the local settlement caused by underground road excavation by inverse Fourier transform.

[0037] ;

[0038] In the formula:

[0039] Let be the function to be fitted.

[0040] , , , , , , , , , Let be the parameters to be fitted, where , These are the cutoff frequencies for the low-pass and high-pass filters, respectively.

[0041] The improvement lies in the use of a new fitting formula (1), which employs a nonlinear least squares method to directly fit and obtain the cutoff frequency. The improved bandpass filtering method for extracting local settlement caused by underground road excavation fits functions with insignificant concavity and convexity, obtaining the cutoff frequency in one step, thus realizing the extraction of local settlement during underground road construction.

[0042] Step S3: Combine shield tunneling construction data with ground feature classification results, high-precision settlement data and other settlement-related influencing factors to form an urban settlement risk dataset;

[0043] We will identify existing urban underground road excavation cases in the region, collect data on settlement influencing factors along the project route, and use these data as features for subsequent model training.

[0044] The settlement influencing factor data includes:

[0045] InSAR settlement data: Settlement data from the year prior to excavation;

[0046] Natural influencing factors: elevation data, geological data, ground slope, etc.;

[0047] Factors affecting construction: the depth, curvature, diameter, and distance of the shield tunnel;

[0048] Urban influencing factors: land cover type, distance of underground pipelines, road density, etc.;

[0049] To predict line settlement risk during the planning phase, settlement influencing factors must be selected from elements related to settlement and acquired during the planning stage. The final dataset is stored in GIS raster data format, with standardized or finer raster sizes to facilitate dataset creation. The susceptibility multi-level dataset includes two main categories: feature data and label data. Feature data consists of collected settlement influencing factors, while label data is high-precision settlement data calculated using improved filtering methods during the construction phase.

[0050] Step S4: Construct a multi-level machine learning algorithm for urban subsidence susceptibility, train it based on the urban subsidence risk dataset, and realize the visual assessment of the subsidence risk of ground features along the shield tunneling excavation line and the risk prediction for the excavation section.

[0051] The susceptibility model is trained on a constructed dataset using common machine learning models such as Random Forest and XGBoost. For labeling, the settlement rate after excavation is divided into multiple settlement levels to form susceptibility labels, and relevant settlement influencing factors are used as features. After training the model based on these labels, feature data of the area along the underground road is input, and a risk matrix is ​​constructed at the model output along with the importance classification of urban features. This enables the assessment of settlement risk along underground roads under construction or the prediction of settlement risk along the design of underground roads.

[0052] The multi-level machine learning algorithm for urban subsidence susceptibility is trained by setting multiple subsidence levels. It calculates the subsidence susceptibility level based on the expected value of each subsidence level, reflecting the differences in susceptibility under different subsidence rates, making it more suitable for subsidence risk assessment in urban environments. Ultimately, it generates visualized assessments or predictions of urban underground roads, assisting in current route selection and providing early warnings for tunnels under construction.

[0053] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for visual assessment and prediction of ground settlement risk during shield tunneling excavation of urban underground roads based on GIS and InSAR, comprising the following steps: Step S1: The method of urban land cover classification and intelligent annotation, which integrates GIS and InSAR multi-source data, is used to obtain the urban land cover classification results. Step S2: Based on the improved InSAR filtering method, high-precision InSAR settlement data of urban areas are obtained, and the local settlement caused by underground road excavation is extracted by the improved bandpass filtering method. An improved method for calculating and extracting local settlement caused by underground road excavation using bandpass filtering: The spectrum and power spectrum of settlement data are calculated by two-dimensional Fourier transform; the radial average data of the power spectrum is extracted, and the cutoff frequency is determined by fitting the nonlinear least squares method according to formula (1); the frequency spectrum is clipped according to the cutoff frequency, and the local settlement caused by underground road excavation is obtained by inverse Fourier transform. ； In the formula: Let be the function to be fitted. , , , , , , , , , Let be the parameters to be fitted, where , These are the cutoff frequencies for the low-pass and high-pass filters, respectively. Step S3: Combine shield tunneling construction data with ground feature classification results, high-precision settlement data and other settlement-related influencing factors to form an urban settlement risk dataset; Step S4: Construct a multi-level machine learning algorithm for urban subsidence susceptibility, train it based on the urban subsidence risk dataset, and realize the visual assessment of the subsidence risk of ground features along the shield tunneling excavation line and the risk prediction for the excavation section. The settlement rate after excavation is divided into multiple settlement levels to form susceptibility labels, and relevant settlement influencing factors are used as features. After the model is trained based on the labels, the feature data of the area along the underground road is input, and a risk matrix is ​​formed at the model output and the urban land feature classification results to realize the assessment of settlement risk along the underground road under construction or the prediction of settlement risk along the underground road under design.

2. The method for visual assessment and prediction of ground settlement risk in shield tunneling of urban underground roads based on GIS and InSAR as described in claim 1, characterized in that: Step S1 includes the following steps: Step S11: Classify urban features according to their importance in urban development and the compensatory capacity of other urban facilities after a subsidence disaster. Step S12: Integrate GIS satellite imagery data, InSAR scattering point distribution data, and other multi-source data, and use machine vision algorithms to label the land cover data according to the classified urban land cover levels.

3. The method for visual assessment and prediction of ground settlement risk in shield tunneling of urban underground roads based on GIS and InSAR as described in claim 1, characterized in that: Step S3 further includes: finding existing urban underground road excavation cases in the region, and statistically analyzing the settlement impact factor data along the project route as features for subsequent model training.

4. The method for visual assessment and prediction of ground settlement risk in shield tunneling of urban underground roads based on GIS and InSAR as described in claim 1, characterized in that: The urban subsidence risk dataset in step S3 is stored in the form of GIS raster data, which includes two main categories: feature data and label data.

5. The method for visual assessment and prediction of ground settlement risk in shield tunneling of urban underground roads based on GIS and InSAR as described in claim 1, characterized in that: The settlement influencing factor data in step S3 includes InSAR settlement data, natural influencing factors, construction influencing factors, and urban influencing factors.

6. A computer system comprising a memory, a processor, and a computer program stored in the memory, characterized in that: The processor executes the computer program to implement the steps of the method of claim 1.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, it implements the steps of the method of claim 1.

8. A computer program product, comprising a computer program, characterized in that: When the computer program is executed by a processor, it implements the steps of the method of claim 1.

Images