Search

CN-122024057-A - Slope stability monitoring method based on sky-ground integration

CN122024057ACN 122024057 ACN122024057 ACN 122024057ACN-122024057-A

Abstract

The invention discloses a sky-ground integrated slope stability monitoring method, which belongs to the field of slope stability monitoring and comprises the steps of obtaining earth surface displacement and crack observation data from an earth surface layer, removing noise and abnormal values by adopting a data cleaning algorithm to obtain a high-precision earth surface local change data set, obtaining drilling stress and underground water level data from the earth surface layer, extracting deep features by adopting a signal processing algorithm to obtain an underground stability feature data set, extracting dynamic change trend of a slope by adopting a time sequence analysis algorithm according to a multi-dimensional slope feature data set to obtain a slope stability change trend, updating the multi-dimensional slope feature data set by adopting a real-time data flow processing algorithm according to a slope stability risk level to obtain a dynamically updated slope stability evaluation result, and generating a real-time slope stability early warning signal by adopting a warning signal generation algorithm aiming at the slope stability evaluation result. The method and the device remarkably improve the accuracy and timeliness of slope stability monitoring.

Inventors

  • WANG WEN
  • XU MINGXIA

Assignees

  • 山东城市建设职业学院

Dates

Publication Date
20260512
Application Date
20260130

Claims (10)

  1. 1. The slope stability monitoring method based on sky and ground integration is characterized by comprising the following steps of: Acquiring satellite remote sensing and surface deformation data observed by an unmanned aerial vehicle from a sky layer, and unifying a time reference system and a space reference system of the data by adopting a space-time calibration algorithm to obtain a standardized surface deformation data set; Obtaining earth surface displacement and crack observation data from the earth surface layer, and removing noise and abnormal values by adopting a data cleaning algorithm to obtain a high-precision earth surface local change data set; Acquiring drilling stress and groundwater level data from an underground layer, and extracting deep features by adopting a signal processing algorithm to obtain an underground stability feature data set; Aiming at the surface deformation data set, the ground local change data set and the underground stability characteristic data set, adopting a data fusion algorithm to integrate multi-source data so as to obtain a uniform multi-dimensional slope characteristic data set; according to the multidimensional slope characteristic data set, extracting the dynamic change trend of the slope by adopting a time sequence analysis algorithm to obtain the slope stability change trend; If the change trend of the slope stability exceeds a preset threshold system, classifying the change trend by adopting a machine learning classification algorithm, and judging the slope stability risk level; According to the slope stability risk level, updating a multidimensional slope characteristic data set by adopting a real-time data stream processing algorithm to obtain a dynamically updated slope stability evaluation result; And aiming at the dynamically updated slope stability evaluation result, generating a real-time slope stability early warning signal by adopting an early warning signal generation algorithm.
  2. 2. The method of claim 1, wherein the step of determining the position of the substrate comprises, Obtaining a normalized surface deformation dataset comprising: obtaining surface deformation data from satellite remote sensing and unmanned aerial vehicle observation, and cleaning and format converting the data by adopting a preprocessing method to obtain an initial surface deformation data set; if the initial data set has time or space deviation, unifying a time reference system and a space reference system by adopting a space-time calibration algorithm to obtain a calibrated surface deformation data set; According to the calibrated data set, integrating satellite remote sensing and unmanned aerial vehicle observation data by using a data fusion technology to obtain a high-precision fusion surface deformation data set; And analyzing the surface deformation characteristics by adopting a differential interferometry algorithm aiming at the fusion data set, and determining the deformation area and the deformation magnitude to obtain a standardized surface deformation data set.
  3. 3. The method of claim 1, wherein the step of determining the position of the substrate comprises, Obtaining a high-precision ground local variation data set, comprising: Acquiring earth surface displacement data and crack observation data, and acquiring an original data set through a sensor to obtain an initial data set containing displacement and crack information; carrying out format standardization and missing value filling on the initial data set by adopting a data preprocessing technology to obtain a preprocessed data set with uniform format; noise removing processing is carried out on the preprocessed data set through a median filtering algorithm, and a denoising data set is obtained; if abnormal points deviating from a preset threshold exist in the denoising data set, detecting and eliminating abnormal values by adopting an isolated forest algorithm to obtain an abnormal-free data set; according to the abnormal-free data set, calculating the surface displacement change trend and crack distribution characteristics to obtain a local ground change characteristic set; classifying the local change feature set of the ground by adopting a K-means clustering algorithm to obtain a regional data set of the ground change; and generating a high-precision ground local change data set through the ground change partition data set.
  4. 4. The method of claim 1, wherein the step of determining the position of the substrate comprises, Obtaining a subsurface stability feature dataset comprising: acquiring drilling stress and groundwater level data from an underground layer through a sensor, and storing the drilling stress and groundwater level data as an original geological data set; Performing signal processing on the original geological data set by adopting a wavelet transformation algorithm, extracting deep features, and generating a first feature data set; if the deep feature variance in the first feature data set is larger than a preset threshold, performing dimension reduction on the features through a principal component analysis algorithm to obtain a second feature data set; classifying the underground stability by adopting a support vector machine algorithm according to the second characteristic data set, and determining a stability class; the time sequence analysis is carried out on the stress distribution and the water level change in the second characteristic data set, so that a change trend is obtained; according to the change trend, calculating a dynamic correlation coefficient of stress distribution and water level change by adopting a sliding window method to obtain a correlation characteristic; And generating a subsurface stability feature data set by performing cluster analysis on the correlation features.
  5. 5. The method of claim 1, wherein the step of determining the position of the substrate comprises, Obtaining a unified multidimensional slope characteristic dataset comprising: Acquiring surface deformation data, a ground local change data set and an underground stability characteristic data set, and removing noise and redundancy through preprocessing to obtain a standardized multi-source data set; Carrying out feature extraction on the standardized multi-source data set by adopting a principal component analysis algorithm to generate a feature vector set after dimension reduction; and fusing the surface deformation data, the ground local change and the underground stability characteristics in the characteristic vector set by using a Kalman filtering algorithm to obtain a unified multidimensional slope characteristic data set.
  6. 6. The method of claim 1, wherein the step of determining the position of the substrate comprises, Obtaining a slope stability change trend, comprising: Acquiring original data from a multidimensional slope characteristic data set, and cleaning the data set by adopting a data preprocessing method to obtain standardized time sequence data; according to the standardized time sequence data, extracting the dynamic change characteristics of the side slope by adopting an ARIMA algorithm to obtain the dynamic change trend of the side slope; Extracting key time point characteristics from the dynamic change trend of the side slope, calculating the trend change rate by adopting a sliding window method, and determining the dynamic characteristics of the side slope; if the change rate of the slope dynamic characteristics exceeds a preset threshold value, clustering the change trend by adopting a K-means algorithm, and judging the slope stability grade; According to the slope stability grade, acquiring historical time series data, and calculating a stability fluctuation range by adopting a sliding window method to obtain a stability evaluation result; And extracting abnormal point characteristics from the stability evaluation result, and verifying by adopting an abnormal detection method to determine the change trend of the slope stability.
  7. 7. The method of claim 1, wherein the step of determining the position of the substrate comprises, Judging the slope stability risk level, comprising: acquiring slope stability related data, and acquiring displacement, stress and vibration frequency data from a sensor to obtain an original data set; extracting displacement change rate, stress distribution and vibration frequency change characteristics from the original data set by adopting a characteristic extraction technology to obtain a characteristic data set; Comparing the change rate, distribution and frequency change in the feature data set with a threshold value through a preset threshold system, and judging whether the change rate, distribution and frequency change exceed the threshold value or not to obtain a super-threshold feature set; If the super-threshold feature set is not empty, carrying out classification training on the super-threshold feature set by adopting a support vector machine algorithm to obtain a classification model; and predicting the super-threshold feature set through a classification model to obtain the slope stability risk level.
  8. 8. The method of claim 1, wherein obtaining a dynamically updated slope stability assessment comprises: according to the slope stability risk level, updating the key slope characteristic set by combining the real-time data stream to obtain a dynamically updated characteristic data set; Carrying out trend prediction on the dynamically updated characteristic data set by adopting a time sequence analysis method to obtain a slope stability change trend; And converting the change trend of the slope stability into a dynamic evaluation result through a preset mapping rule, and acquiring real-time updating of the slope stability risk level.
  9. 9. The method of claim 1, wherein generating a real-time slope stability warning signal comprises: generating a real-time early warning signal by adopting a conditional random field model according to the stability change trend; acquiring the latest sensor data through a dynamic updating mechanism to obtain updated geological environment parameters; If the deviation between the updated geological environment parameter and the previous parameter exceeds a preset threshold value, re-executing the support vector machine algorithm, and judging a new stability classification result; and generating slope stability real-time early warning output according to the new stability classification result and the real-time early warning signal.
  10. 10. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method according to any one of claims 1-9.

Description

Slope stability monitoring method based on sky-ground integration Technical Field The invention belongs to the technical field of slope stability monitoring, and particularly relates to a slope stability monitoring method based on sky-ground integration. Background Slope stability monitoring is a key field in civil engineering and geological disaster prevention and control, and is directly related to the security of infrastructure and the guarantee of lives and properties of people. The instability of the side slope can cause disasters such as landslide, mud-rock flow and the like, so that great economic loss and social influence are caused, and therefore, the development of efficient and accurate monitoring technology is very important. Currently, slope monitoring methods mostly rely on single observation means or local data analysis, such as measuring displacement only by ground instruments or obtaining surface information only by satellite remote sensing. The methods have the defects in data integration and dynamic response, and complex changes inside and outside the slope are difficult to comprehensively capture, so that timeliness and accuracy of early warning are limited. In particular under complex geological conditions, the prior art has difficulty in achieving stability assessment of multidimensional, synergistic, and other factors. Disclosure of Invention In order to solve the technical problems, the invention provides a slope stability monitoring method based on sky-ground integration, which aims to solve the problems existing in the prior art. In order to achieve the above object, in a first aspect, the present invention provides a slope stability monitoring method based on sky-ground integration, including: Acquiring satellite remote sensing and surface deformation data observed by an unmanned aerial vehicle from a sky layer, and unifying a time reference system and a space reference system of the data by adopting a space-time calibration algorithm to obtain a standardized surface deformation data set; Obtaining earth surface displacement and crack observation data from the earth surface layer, and removing noise and abnormal values by adopting a data cleaning algorithm to obtain a high-precision earth surface local change data set; Acquiring drilling stress and groundwater level data from an underground layer, and extracting deep features by adopting a signal processing algorithm to obtain an underground stability feature data set; Aiming at the surface deformation data set, the ground local change data set and the underground stability characteristic data set, adopting a data fusion algorithm to integrate multi-source data so as to obtain a uniform multi-dimensional slope characteristic data set; according to the multidimensional slope characteristic data set, extracting the dynamic change trend of the slope by adopting a time sequence analysis algorithm to obtain the slope stability change trend; If the change trend of the slope stability exceeds a preset threshold system, classifying the change trend by adopting a machine learning classification algorithm, and judging the slope stability risk level; According to the slope stability risk level, updating a multidimensional slope characteristic data set by adopting a real-time data stream processing algorithm to obtain a dynamically updated slope stability evaluation result; And aiming at the dynamically updated slope stability evaluation result, generating a real-time slope stability early warning signal by adopting an early warning signal generation algorithm. Preferably, a normalized surface deformation dataset is obtained, comprising: obtaining surface deformation data from satellite remote sensing and unmanned aerial vehicle observation, and cleaning and format converting the data by adopting a preprocessing method to obtain an initial surface deformation data set; if the initial data set has time or space deviation, unifying a time reference system and a space reference system by adopting a space-time calibration algorithm to obtain a calibrated surface deformation data set; According to the calibrated data set, integrating satellite remote sensing and unmanned aerial vehicle observation data by using a data fusion technology to obtain a high-precision fusion surface deformation data set; And analyzing the surface deformation characteristics by adopting a differential interferometry algorithm aiming at the fusion data set, and determining the deformation area and the deformation magnitude to obtain a standardized surface deformation data set. Preferably, obtaining a high precision ground local variation dataset comprises: Acquiring earth surface displacement data and crack observation data, and acquiring an original data set through a sensor to obtain an initial data set containing displacement and crack information; carrying out format standardization and missing value filling on the initial data set by adopting a data preprocessing tec