Early warning method for loess slope geological disaster prevention and control

Abstract

The invention relates to the technical field of geological disaster prevention and control, and discloses an early warning method for loess slope geological disaster prevention and control, which comprises the following steps: arranging stress sensors and pore water pressure gauges at different depths of a loess slope, and collecting and preprocessing ground stress and pore water pressure data; constructing a ground stress change gradient index and a pore water pressure index of the loess slope; scanning the loess slope by using a laser scanner, generating a three-dimensional point cloud, performing point cloud registration, and then calculating a point cloud gradient to obtain a vegetation point cloud of the loess slope; a Kriging interpolation method is adopted to carry out interpolation calculation on the elevation value of the vegetation point cloud, and the average value, the maximum value and the difference value of the elevation difference are utilized to carry out local correction on the interpolation process to obtain the elevation value of the vegetation point cloud after elevation difference correction so as to construct a vegetation elevation variation coefficient index; and finally, constructing a comprehensive early warning model for geological disaster prevention and control of the loess slope to carry out geological disaster early warning. According to the method, the accuracy of early warning of the loess slope geological disasters is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of geological disaster prevention and control, and particularly relates to a warning method for preventing and controlling geological disasters of loess slopes. Background Art

[0002] Loess slopes are common geological phenomena in loess areas. Their stability is affected by various factors, including geological structures, hydrogeological conditions, climatic conditions, human engineering activities, etc. The physical and mechanical properties of loess are the internal factors determining the slope stability, while hydrogeological conditions and climatic conditions are the external factors affecting the slope stability.

[0003] Geological disasters of loess slopes are unstable loess structures existing on the slopes in the loess-covered area. Under the action of gravity, they are a kind of geological disasters that produce overall sliding along the sliding shear plane. Due to the characteristics of large pores, water sensitivity, poor mechanical properties, and developed joints and fissures of loess itself, loess landslides have become one of the sudden geological disasters with the largest number of occurrences, the most serious damage, and the worst impact in China. They often have characteristics such as multiple occurrences, concealment, disaster, and complexity.

[0004] In addition, due to the particularity of loess mainly reflected in its developed vertical joints, numerous vertical cracks, and easy disintegration when encountering water, these characteristics make loess slopes prone to landslides and collapses under the action of rainfall or groundwater. Therefore, how to monitor loess slopes to prevent geological disasters is an urgent problem to be solved; most of the existing warning methods for preventing and controlling geological disasters of loess slopes use high-precision sensors or monitoring systems for warning, and mostly independently monitor single influencing factors, without establishing the relationship between various influencing factors, resulting in inaccurate warning results and thus unable to fully reflect the actual disaster occurrence situation of the slope. Summary of the Invention

[0005] In view of the above deficiencies in the prior art, the present invention provides a warning method for preventing and controlling geological disasters of loess slopes, which is used to solve the problem that the existing warning methods for preventing and controlling geological disasters of loess slopes only monitor single influencing factors, resulting in inaccurate warning results.

[0006] In order to achieve the above invention purpose, the technical solution adopted by the present invention is as follows:

[0007] A warning method for preventing and controlling geological disasters of loess slopes, comprising the following steps:

[0008] S1. Install stress sensors and pore water pressure gauges at different depths of the loess slope, collect in-situ stress data and pore water pressure data and perform data preprocessing. By calculating the average value and standard deviation of the in-situ stress change gradient index of the loess slope, construct the in-situ stress change gradient index of the loess slope. By calculating the average value and standard deviation of the pore water pressure data, construct the pore water pressure index of the loess slope;

[0009] S2. Use a laser scanner to scan the loess slope, generate a three-dimensional point cloud and perform point cloud registration to obtain an accurate three-dimensional point cloud map of the loess slope. By calculating the point cloud slope of the accurate three-dimensional point cloud map and comparing it with the slope threshold, obtain the vegetation point cloud of the loess slope;

[0010] S3. Use Kriging interpolation method to interpolate and calculate the elevation values of the vegetation point cloud, and use the average value, maximum value and difference value of the elevation difference to locally correct the interpolation process to obtain the elevation values of the vegetation point cloud after elevation difference correction;

[0011] S4. Calculate the average value and standard deviation of the elevation values of the vegetation point cloud after elevation difference correction, and construct the vegetation elevation variation coefficient index;

[0012] S5. Based on the in-situ stress change gradient index of the loess slope, the pore water pressure index of the loess slope and the vegetation elevation variation coefficient index, construct a comprehensive early warning model for the prevention and control of loess slope geological disasters, which is used to early warn of loess slope geological disasters.

[0013] The present invention has the following beneficial effects:

[0014] An early warning method for the prevention and control of loess slope geological disasters proposed by the present invention comprehensively considers three factors: in-situ stress, pore water pressure and vegetation, covering all key elements affecting the stability of the loess slope. Among them, in-situ stress and pore water pressure directly affect the mechanical properties of the soil mass, while vegetation enhances the slope stability. Therefore, this method of comprehensive multi-factor analysis can more accurately evaluate the slope stability and improve the reliability of early warning. At the same time, when analyzing the vegetation-covered area of the loess slope, point cloud data is obtained by a three-dimensional laser scanner and registered, and an improved Kriging interpolation method is introduced to calculate the elevation values of the vegetation point cloud, improving the accuracy of the elevation values and the accuracy of the constructed vegetation elevation variation coefficient index. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a schematic flow chart of an early warning method for the prevention and control of loess slope geological disasters proposed by the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0016] The following describes the specific embodiments of the present invention to facilitate those skilled in the art of this technology to understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific embodiments. For those of ordinary skill in the art of this technology, as long as various changes are within the spirit and scope of the present invention defined and determined by the appended claims, these changes are obvious, and all inventions created using the concept of the present invention are within the scope of protection.

[0017] As Figure 1 shown, a warning method for preventing and controlling geological disasters of loess slopes includes the following steps S1 - S5:

[0018] S1. Stress sensors and pore water pressure gauges are arranged at different depths of the loess slope to collect in - situ stress data and pore water pressure data and perform data pre - processing. By calculating the average value and standard deviation of the in - situ stress change gradient index of the loess slope, the in - situ stress change gradient index of the loess slope is constructed. By calculating the average value and standard deviation of the pore water pressure data, the pore water pressure index of the loess slope is constructed.

[0019] In this embodiment, since the in - situ stress change gradient and pore water pressure are important factors affecting the stability of loess slopes, by constructing the in - situ stress change gradient index of the loess slope and the pore water pressure index of the loess slope, it provides an important basis for warning of geological disasters of loess slopes, helps to timely detect potential hidden dangers of geological disasters, and at the same time avoids the limitations of data from a single monitoring point. Among them, the in - situ stress change gradient index of the loess slope comprehensively reflects the overall characteristics and dispersion degree of the in - situ stress change gradient of the entire loess slope, and the pore water pressure index of the loess slope comprehensively reflects the overall magnitude and fluctuation of the pore water pressure of the entire loess slope.

[0020] Specifically, step S1 specifically includes S11 - S15:

[0021] S11. Uniform grid division is performed according to the topography of the loess slope, and the center point of the grid is used as the depth monitoring point. Stress sensors and pore water pressure gauges are arranged at each depth monitoring point.

[0022] S12. The in - situ stress data is collected by using stress sensors and the pore water pressure data is collected by using pore water pressure gauges, and pre - processing such as filtering and denoising is performed on the in - situ stress data and pore water pressure data to obtain the pre - processed in - situ stress data and pore water pressure data.

[0023] In this embodiment, the purpose of filtering and denoising is to improve the quality of the collected in - situ stress data and pore water pressure data for subsequent index construction.

[0024] S13. According to the pre - processed in - situ stress data, calculate the in - situ stress change and depth difference between adjacent depth monitoring points to generate the in - situ stress change gradient index of adjacent depth monitoring points.

[0025]

[0026] Among them, G i,i+1 represents the in-situ stress change gradient index of adjacent depth monitoring points, Δσ i,i+1 represents the in-situ stress change between adjacent depth monitoring points, Δz i,i+1 represents the depth difference between adjacent depth monitoring points, σ i+1 represents the in-situ stress at the (i + 1)-th depth monitoring point, σ i represents the in-situ stress at the i-th depth monitoring point, z i+1 represents the depth at the (i + 1)-th depth monitoring point, z i represents the depth at the i-th depth monitoring point.

[0027] S14. Calculate the average value and standard deviation of the in-situ stress change gradient indices of all adjacent depth monitoring points, and construct the in-situ stress change gradient index of the loess slope, that is:

[0028]

[0029] Among them, G σ represents the in-situ stress change gradient index of the loess slope, G s represents the standard deviation of the in-situ stress change gradient indices of the depth monitoring points, G avg represents the average value of the in-situ stress change gradient indices of all adjacent depth monitoring points, and N represents the total number of depth monitoring points.

[0030] In this embodiment, the average value and standard deviation of the in-situ stress change gradient indices of all adjacent depth monitoring points are calculated to construct the in-situ stress change gradient index of the loess slope. The average value can reflect the overall level of the in-situ stress change gradient, and the standard deviation can reflect the dispersion degree of the in-situ stress change gradient. The combination of the two can more comprehensively describe the characteristics of the in-situ stress change gradient.

[0031] S15. According to the preprocessed pore water pressure data, calculate the average value and standard deviation of the pore water pressure data of all depth monitoring points, and construct the pore water pressure index of the loess slope, that is:

[0032]

[0033] Among them, U u represents the pore water pressure index of the loess slope, U s represents the standard deviation of the pore water pressures of all depth monitoring points, u i represents the pore water pressure at the i-th depth monitoring point, U avg represents the average value of the pore water pressures of all depth monitoring points.

[0034] In this embodiment, the pore water pressure index of the loess slope is constructed by calculating the average value and standard deviation of the pore water pressure data at all depth monitoring points. The average value can reflect the overall magnitude of the pore water pressure, and the standard deviation can reflect the fluctuation of the pore water pressure among different depth monitoring points. Therefore, constructing the pore water pressure index of the loess slope using the average value and standard deviation helps to analyze the distribution characteristics of the pore water pressure.

[0035] S2. Scan the loess slope using a laser scanner, generate a three-dimensional point cloud and perform point cloud registration to obtain an accurate three-dimensional point cloud map of the loess slope. By calculating the point cloud slope of the accurate three-dimensional point cloud map and comparing it with the slope threshold, the vegetation point cloud of the loess slope is obtained.

[0036] Specifically, step S2 specifically includes S21 - S26:

[0037] S21. Scan the vegetation-covered area of the loess slope using a laser scanner, generate a three-dimensional point cloud and perform preprocessing of denoising and stitching to generate a three-dimensional point cloud dataset, that is:

[0038] A = {Q 1 , Q 2 , …, Q n}

[0039] where A represents the three-dimensional point cloud dataset, n represents the number of the three-dimensional point cloud datasets, Q 1 represents the first three-dimensional point cloud dataset, Q 2 represents the second three-dimensional point cloud dataset, and Q n represents the nth three-dimensional point cloud dataset.

[0040] In this embodiment, the purpose of denoising and stitching is to improve the quality and integrity of the point cloud data.

[0041] S22. Screen two most similar point cloud groups using the coordinate characteristics of the point clouds in the three-dimensional point cloud dataset, that is:

[0042]

[0043] where Q a , Q b respectively represent any two point cloud groups in the preprocessed three-dimensional point cloud dataset A, Q a represents the reference point cloud group, Q b represents the target point cloud group, I represents the number of point clouds in the target point cloud group, T a represents the coordinate characteristics of the point clouds in the reference point cloud group, T b represents the coordinate characteristics of the point clouds in the target point cloud group, a represents the number of the point cloud in the reference point cloud group, and b represents the number of the point cloud in the target point cloud group.

[0044] S23. After grouping the most similar point clouds (Q e , Q c ) selected through step S22, perform point cloud matching based on the coordinate characteristics of the point clouds within the point cloud group (Q e , Q c ) to generate the point cloud set Q 0 after point cloud matching. Specifically:

[0045] Based on the coordinate characteristics T e of any point cloud e within the point cloud set Q e , screen out the point cloud c closest to the point cloud e within the point cloud set Q c and continue the closest point cloud matching, that is:

[0046]

[0047] where T c represents the coordinate characteristics of the point cloud c.

[0048] When the point clouds within the point cloud set Q e and the point clouds within the point cloud set Q c complete the closest point cloud matching, calculate the intermediate coordinate characteristics after the closest point cloud matching is successful, that is:

[0049]

[0050] where T e～c represents the intermediate coordinate characteristics after the closest point cloud matching is successful.

[0051] Take the intermediate coordinate characteristics T e～c as the point cloud characteristics after the two closest point cloud matchings, and generate the point cloud set Q e and the point cloud set Q c after point cloud matching between the point cloud set Q 0 .

[0052] S24. Use the three-dimensional point cloud map formed by the point cloud set Q 0 as the precise three-dimensional point cloud map of the loess slope.

[0053] S25. According to the precise three-dimensional point cloud map of the loess slope, obtain the slope of each point cloud. Specifically:

[0054] For point P, select the point set M(P) within its neighborhood, and fit a plane by the least squares method to calculate the normal vector m of this plane, and obtain the slope s of point P, that is:

[0055] s = arccos(m · z axis )

[0056] where arccos represents the inverse cosine function, and zaxis It represents the unit vector in the z-axis direction.

[0057] S26. Set the slope threshold, and classify the point cloud with a slope greater than the slope threshold as the vegetation point cloud.

[0058] In this embodiment, vegetation usually has an irregular shape and a complex surface structure, and its slope is generally larger than that of the bare soil part of the loess slope. Therefore, by setting an appropriate slope threshold, the point cloud with a larger slope can be classified as the vegetation point cloud to achieve the extraction of the vegetation point cloud.

[0059] S3. Use the Kriging interpolation method to perform interpolation calculation on the elevation value of the vegetation point cloud, and use the average value, maximum value, and difference value of the elevation difference to locally correct the interpolation process to obtain the elevation value of the vegetation point cloud after elevation difference correction.

[0060] In this embodiment, the Kriging interpolation method is used to perform interpolation calculation on the elevation value of the vegetation point cloud. Due to problems such as uneven distribution and measurement errors in the point cloud data, the interpolation result may have deviations in local areas. The average value, maximum value, median, and difference value of the elevation difference reflect the change characteristics of the elevation in the local area. Therefore, these statistics are introduced, and the interpolation result is adjusted using the adjustment coefficient to correct the local deviation, making the corrected elevation value more in line with the actual situation.

[0061] Specifically, in step S3, the average value, maximum value, median, and difference value of the elevation difference are used to locally correct the interpolation process to obtain the formula for the elevation value of the vegetation point cloud after elevation difference correction:

[0062]

[0063] where H k ′ represents the elevation value of the vegetation point cloud k after elevation difference correction, α, β, γ, and δ all represent adjustment coefficients, and H0 i represents the original elevation value of the vegetation point cloud, Avg d represents the average value of the elevation difference, Max d represents the maximum value of the elevation difference, Median d represents the median of the elevation difference, and Percentile d represents the percentile of the elevation difference.

[0064] In this embodiment, according to the interpolation situation, the average value, maximum value, and difference value of the elevation difference are used to locally correct the interpolation process, that is, elevation correction. Among them, the adjustment coefficients α, β, γ, and δ are 0.1, 0.01, 0.1, and 0.01 respectively. The specific setting ratio can be set according to the actual interpolation situation.

[0065] S4. Calculate the mean and standard deviation of the elevation values of the vegetation point cloud after elevation difference correction, and construct an index of the coefficient of variation of vegetation elevation.

[0066] In this embodiment, since the growth condition of vegetation is closely related to the stability of the loess slope, an index of the coefficient of variation of vegetation elevation is introduced and incorporated as an important reference index into the comprehensive early warning model for preventing and controlling geological disasters of the loess slope. When this index changes abnormally, it may indicate that the reinforcement effect of vegetation on the slope has changed, thereby affecting the stability of the slope, which helps to detect potential hidden dangers of geological disasters in advance. This index can comprehensively reflect the degree of dispersion of the elevation of the vegetation point cloud. In addition, if the value of this index is large, it indicates that there is a large elevation difference in the vegetation in space, which may mean that the growth conditions of the vegetation are uneven, and there are different growth stages, different species, or the influence of different environmental factors; on the contrary, if the index value is small, it indicates that the vegetation grows relatively neatly and may be in a similar growth environment and growth stage.

[0067] Specifically, step S4 specifically includes S41 - S42:

[0068] S41. Calculate the mean and standard deviation of the elevation values of the vegetation point cloud after elevation difference correction, that is:

[0069]

[0070] Among them, V h respectively represent the mean and standard deviation of the elevation values of the vegetation point cloud after elevation difference correction, and K represents the number of vegetation point clouds after elevation difference correction.

[0071] S42. According to the mean and standard deviation of the elevation values of the vegetation point cloud after elevation difference correction, construct an index of the coefficient of variation of vegetation elevation, that is:

[0072]

[0073] Among them, C h represents the index of the coefficient of variation of vegetation elevation.

[0074] S5. Based on the index of the change gradient of ground stress of the loess slope, the index of pore water pressure of the loess slope, and the index of the coefficient of variation of vegetation elevation, construct a comprehensive early warning model for preventing and controlling geological disasters of the loess slope, which is used to give early warnings of geological disasters of the loess slope.

[0075] Specifically, step S5 specifically includes:

[0076] S51. Use the entropy weight method to assign weights to the ground stress change gradient index of the loess slope, the pore water pressure index of the loess slope, and the vegetation elevation variation coefficient index, and construct a comprehensive early warning model for the prevention and control of loess slope geological disasters, that is:

[0077]

[0078] Among them, y represents the comprehensive early warning value for the prevention and control of loess slope geological disasters, ω σ 、ω u 、ω u respectively represent the weight coefficients of the ground stress change gradient index of the loess slope, the pore water pressure index of the loess slope, and the vegetation elevation variation coefficient index.

[0079] In this embodiment, the entropy weight method is an objective weight assignment method. It determines the weight based on the dispersion degree of the index data. The greater the dispersion degree of the data, the smaller the entropy value, the greater the amount of information provided by the index, and the greater the weight. Therefore, the weights determined in this way can avoid the interference of subjective factors and make the weight distribution of each index more reasonable.

[0080] S52. Set the three-level early warning threshold, and judge whether the comprehensive early warning value predicted by the comprehensive early warning model for the prevention and control of loess slope geological disasters is greater than or equal to the first-level early warning threshold. If so, issue a red early warning, indicating that the loess slope is unstable and a geological disaster is about to occur. Otherwise, judge whether the comprehensive early warning value predicted by the comprehensive early warning model for the prevention and control of loess slope geological disasters is greater than or equal to the second-level early warning threshold and less than the first-level early warning threshold. If so, issue a yellow early warning, indicating that the loess slope is about to reach the unstable state and the possibility of a geological disaster is high. Otherwise, the comprehensive early warning value predicted by the comprehensive early warning model for the prevention and control of loess slope geological disasters is less than the third-level early warning threshold, and issue a blue early warning, indicating that the loess slope is stable.

[0081] Among them, the three-level early warning threshold is set based on the historical data of loess slope geological disasters and actual monitoring experience, that is, by analyzing the historical data of loess slope geological disasters in the research area, finding the critical values of each index when a geological disaster occurs, and determining the current early warning threshold based on this.

[0082] In summary, an early warning method for preventing and controlling geological disasters of loess slopes proposed by the present invention comprehensively considers three factors: ground stress, pore water pressure, and vegetation, covering all the key elements affecting the stability of loess slopes. Among them, ground stress and pore water pressure directly affect the mechanical properties of the soil mass, while vegetation enhances the slope stability. Therefore, this method of comprehensive multi-factor analysis can more accurately evaluate the slope stability and improve the reliability of early warning. At the same time, when analyzing the vegetation-covered area of loess slopes, point cloud data is obtained through a three-dimensional laser scanner and registered, and an improved Kriging interpolation method is introduced to calculate the elevation values of vegetation point clouds, improving the accuracy of elevation values and the accuracy of the constructed vegetation elevation variation coefficient index.

[0083] In the present invention, specific embodiments are used to elaborate on the principle and implementation manner of the present invention. The description of the above embodiments is only used to help understand the method and its core idea of the present invention. At the same time, for those of ordinary skill in the art, based on the idea of the present invention, there will be changes in the specific implementation manner and application scope. In summary, the content of this specification should not be construed as a limitation to the present invention.

[0084] Those of ordinary skill in the art will realize that the embodiments described herein are for helping the reader understand the principle of the present invention, and it should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those of ordinary skill in the art can make various other specific deformations and combinations that do not depart from the essence of the present invention based on the technical revelations disclosed in the present invention, and these deformations and combinations are still within the protection scope of the present invention.

Claims

1. An early warning method for preventing and controlling geological disasters on loess slopes, characterized in that: The following steps are involved: S1. Stress sensors and pore water pressure gauges are arranged at different depths of the loess slope to collect geostress data and pore water pressure data and perform data preprocessing. The geostress change gradient index of the loess slope is constructed by calculating the average value and standard deviation of the geostress change gradient index of the loess slope. The pore water pressure index of the loess slope is constructed by calculating the average value and standard deviation of the pore water pressure data. S2. Scan the loess slope with a laser scanner to generate a three-dimensional point cloud and perform point cloud registration to obtain an accurate three-dimensional point cloud map of the loess slope. Calculate the point cloud slope of the accurate three-dimensional point cloud map and compare it with the slope threshold to obtain the vegetation point cloud of the loess slope. S3, using the Kriging interpolation method to interpolate and calculate the elevation value of the vegetation point cloud, and using the average value, maximum value and difference value of the elevation difference to locally correct the interpolation process, so as to obtain the elevation value of the vegetation point cloud after the elevation difference correction; S4, calculating the mean and standard deviation of the elevation values ​​of the vegetation point cloud after the elevation difference correction, and constructing the vegetation elevation variation coefficient index; S5. Based on the loess slope ground stress change gradient index, loess slope pore water pressure index and vegetation elevation variation coefficient index, a comprehensive early warning model for loess slope geological disaster prevention and control is constructed to provide early warning for loess slope geological disasters.

2. The early warning method for preventing and controlling geological disasters on loess slopes according to claim 1, characterized in that: Step S1 specifically includes: S11. Perform uniform grid division according to the loess slope terrain, use the center point of the grid as the depth monitoring point, and deploy stress sensors and pore water pressure gauges at each depth monitoring point; S12, collecting ground stress data using a stress sensor and collecting pore water pressure data using a pore water pressure gauge, and pre-processing the ground stress data and pore water pressure data by filtering and denoising to obtain pre-processed ground stress data and pore water pressure data; S13, based on the pre-processed geostress data, calculating the geostress change and depth difference between adjacent depth monitoring points, generating a geostress change gradient index for adjacent depth monitoring points, Among them, G i,i+1 Indicates the gradient index of the ground stress change at adjacent depth monitoring points, Δσ i,i+1 represents the change in ground stress between adjacent depth monitoring points, Δz i,i+1 represents the depth difference between adjacent depth monitoring points, σ i+1 represents the ground stress at the i+1th depth monitoring point, σ i represents the ground stress at the i-th depth monitoring point, z i+1 Indicates the depth of the i+1th depth monitoring point, z i Indicates the depth of the i-th depth monitoring point; S14. Calculate the average value and standard deviation of the geostress change gradient index of all adjacent depth monitoring points, and construct the geostress change gradient index of the loess slope, that is: Among them, G σ G represents the gradient index of the ground stress change of the loess slope. s represents the standard deviation of the in-situ stress gradient index at the depth monitoring point, G avg It represents the average value of the in-situ stress gradient index of all adjacent deep monitoring points, and N represents the total number of deep monitoring points; S15. Based on the pre-processed pore water pressure data, the average value and standard deviation of the pore water pressure data of all depth monitoring points are calculated to construct the pore water pressure index of the loess slope, namely: Among them, U u Indicates the pore water pressure index of loess slope, U s represents the standard deviation of pore water pressure at all depth monitoring points, u i represents the pore water pressure at the i-th depth monitoring point, U avg Represents the average value of pore water pressure at all depth monitoring points.

3. The early warning method for preventing and controlling geological disasters on loess slopes according to claim 2 is characterized in that: Step S2 specifically includes: S21. Use a laser scanner to scan the vegetation-covered area of ​​the loess slope, generate a three-dimensional point cloud, and perform denoising and splicing preprocessing to generate a three-dimensional point cloud data set, namely: A={Q1,Q2,…,Q n } Where A represents a 3D point cloud dataset, n represents the number of 3D point cloud datasets, Q1 represents the first 3D point cloud dataset, Q2 represents the second 3D point cloud dataset, and Q n Represents the nth 3D point cloud dataset; S22. Using the coordinate features of the point clouds in the three-dimensional point cloud data set, two most similar point cloud set groups are selected, namely: Among them, Q a , Q b Respectively represent any two point cloud sets in the preprocessed 3D point cloud dataset A, q a represents the reference point cloud, Q b represents the target point cloud set, I represents the number of point clouds in the target point cloud set, T a Represents the coordinate characteristics of the point cloud in the reference point cloud set, T b Represents the coordinate features of the point cloud of the target point cloud set, a represents the number of the point cloud in the reference point cloud set, and b represents the number of the point cloud in the target point cloud set; S23, the most similar point cloud set group (Q e ,Q c ) then, according to the point cloud group (Q e ,Q c ) to perform point cloud matching based on the coordinate features of the point cloud in the image, and generate a point cloud set Q0 after point cloud matching, specifically: According to the point cloud Q e The coordinate feature T of any point cloud e in e , in the point cloud Q c The point cloud c that is closest to the point cloud e is selected internally, and the nearest point cloud matching is continued, that is: Among them, T c Represents the coordinate features of point cloud c; When the cloud gathers Q e Point cloud and point cloud set Q c After the point cloud in the nearest point cloud is matched, the intermediate coordinate features after the nearest point cloud matching is successfully calculated, that is: Among them, T e～c Indicates the intermediate coordinate features after the nearest point cloud is successfully matched; The intermediate coordinate feature T e～c As the point cloud features after the two closest point clouds are matched, the point cloud set Q is generated e Q c Point cloud set Q0 after point cloud matching; S24, using the three-dimensional point cloud image formed by the point cloud set q0 as an accurate three-dimensional point cloud image of the loess slope; S25. According to the accurate three-dimensional point cloud map of the loess slope, the slope of each point cloud is obtained, specifically: For point P, select the point set M(P) in its neighborhood, fit the plane by the least squares method, calculate the normal vector m of the plane, and get the slope s of point P, that is: s=arccos(m·z axis ) Where arccos represents the inverse cosine function, z axis Represents the unit vector in the z-axis direction; S26. Set a slope threshold, and classify the point cloud with a slope greater than the slope threshold as a vegetation point cloud.

4. The early warning method for preventing and controlling geological disasters on loess slopes according to claim 3 is characterized in that: In step S3, the mean, maximum, median and difference values ​​of the elevation difference are used to locally correct the interpolation process, and the formula for obtaining the elevation value of the vegetation point cloud after elevation difference correction is: Among them, H k ′ represents the elevation value of the vegetation point cloud k after elevation difference correction, α, β, γ, δ all represent adjustment coefficients, H0 i Represents the original elevation value of the vegetation point cloud, Avg d Indicates the average value of elevation difference, Max d Indicates the maximum value of the elevation difference, Median d Indicates the median of the elevation difference, Percentile d Represents the percentile quantile of the elevation difference.

5. The early warning method for preventing and controlling geological disasters on loess slopes according to claim 4, characterized in that: Step S4 specifically includes: S41. Calculate the mean and standard deviation of the elevation values ​​of the vegetation point cloud after elevation difference correction, that is: in, V h They represent the mean and standard deviation of the elevation values ​​of the vegetation point cloud after elevation difference correction, and K represents the number of vegetation point clouds after elevation difference correction; S42. According to the mean value and standard deviation of the elevation values ​​of the vegetation point cloud after the elevation difference correction, the vegetation elevation variation coefficient index is constructed, that is: Among them, C h It represents the coefficient of variation of vegetation elevation.

6. The early warning method for preventing and controlling geological disasters on loess slopes according to claim 5, characterized in that: Step S5 specifically includes: S51. The entropy weight method is used to assign weights to the geostress change gradient index of the loess slope, the pore water pressure index of the loess slope, and the vegetation elevation variation coefficient index, and a comprehensive early warning model for the prevention and control of geological disasters on the loess slope is constructed, namely: Among them, y represents the comprehensive warning value of loess slope geological disaster prevention and control, ω σ ,ω u ,ω u They represent the weight coefficients of the geostress variation gradient index of the loess slope, the pore water pressure index of the loess slope, and the vegetation elevation variation coefficient index respectively; S52. Set the third-level warning threshold to determine whether the comprehensive warning value predicted by the comprehensive warning model for prevention and control of geological disasters on loess slopes is greater than or equal to the first-level warning threshold. If so, a red warning is issued, indicating that the loess slope is unstable and a geological disaster is about to occur. Otherwise, determine whether the comprehensive warning value predicted by the comprehensive warning model for prevention and control of geological disasters on loess slopes is greater than or equal to the second-level warning threshold and less than the first-level warning threshold. If so, a yellow warning is issued, indicating that the loess slope is about to reach an unstable state and a geological disaster is likely to occur. Otherwise, the comprehensive warning value predicted by the comprehensive warning model for prevention and control of geological disasters on loess slopes is less than the third-level warning threshold, and a blue warning is issued, indicating that the loess slope is stable.

Images