Landslide early warning method and device based on comprehensive standard deformation index and storage medium

Abstract

The application provides a landslide early warning method and device based on a comprehensive standard deformation index and a storage medium, and comprises the following steps: S1, data acquisition: selecting a representative deformation key position of a slope surface to arrange a deformation monitoring device, and continuously collecting accumulated displacement time series and slope deformation speed time data of the position; S2, maximum normalization or mean normalization to normalize the displacement speed; S3, normality test of the speed displacement probability density distribution: calculating the normalized speed displacement probability density distribution, checking whether it meets the normality distribution, if yes, considering that it is in a stable state, and if not, entering the next step; S4, cumulative probability fitting: obtaining a total cumulative probability distribution function of the landslide displacement speed; S5, standard normalization conversion and comprehensive standard deformation index output; and S6, early warning grade division and output. The analysis process does not consider the volume and material properties of the landslide, has strong generalization, and can be applied to various landslide deformation early warning.

Description

Technical Field

[0001] This invention relates to the technical field of geological disaster early warning, and in particular to a landslide early warning method, equipment and storage medium based on a comprehensive standard deformation index. Background Technology

[0002] my country has a complex topography with numerous mountains, leading to frequent natural geological disasters, especially landslides. Thousands of landslides and collapses occur annually, causing significant loss of life and economic damage, and seriously threatening human production and daily life. Prediction and forecasting are crucial for landslide disaster prevention and control. Achieving landslide deformation prediction and early warning based on reliable monitoring data is an important future development direction.

[0003] Displacement is the most representative parameter reflecting the evolutionary state of a landslide during its spatiotemporal evolution, and based on this, numerous phenomenological early warning models have been established. Following the initial Saito model, Fukuzono proposed the Inverse Velocity (INV) method, which predicts landslide collapse time by linearly fitting the reciprocal of the average velocity through analysis of large-scale slope failure tests. Since then, this method has been continuously developed in case studies, with improvements in the definition of early warning velocity thresholds and prediction accuracy, but it remains limited by measurement noise. Furthermore, in the past few decades, many other displacement-related models, including grey system models, catastrophe models, Verhulst models, and machine learning models, have been reported for landslide early warning and have proven effective. Due to the inherent characteristics of landslides and the complexity of their triggering factors, early warning models and corresponding thresholds derived for specific landslides are difficult to extend to other landslides. To reduce false alarms caused by this limitation, researchers have proposed dimensionless indices, such as dimensionless displacement trends, loading / unloading response ratios, normalized velocity trends, and displacement tangential angles. However, the past is key to the future. This approach focuses on the accelerated deformation phase while ignoring historical information from the uniform deformation phase. This amplifies some features of the displacement curve, leading to possible misjudgments.

[0004] Therefore, proposing an accurate and reliable early warning method for landslide stability is of great practical significance. Summary of the Invention

[0005] In view of this, embodiments of the present invention provide a landslide early warning method based on a comprehensive standard deformation index, which aims to solve the problems of the limited scope and limited application of existing landslide stability early warning methods.

[0006] Embodiments of the present invention provide a landslide early warning method based on a comprehensive standard deformation index, comprising the following steps: S1, Data Acquisition: Based on the landslide site reconnaissance data, deformation monitoring devices were deployed at representative key deformation locations on the slope surface to continuously collect cumulative displacement time series data at these locations. The landslide deformation velocity time series is obtained by dividing the cumulative difference between two adjacent monitoring points by time: ; S2, Displacement velocity normalization: Through extreme value normalization or mean normalization, wherein the extreme value normalization can map the displacement velocity to the range [0-1], and the mean normalization can map the displacement velocity to the range [-1-1]. S3, Normality test of velocity-displacement probability density distribution: First, calculate the velocity-displacement probability density distribution after the extreme value normalization and the mean normalization, and check whether it satisfies the normality distribution. If it does, it is considered to be in a stable state. If it does not, proceed to the next step. S4, Cumulative Probability Fitting: First, all negative displacement velocities are assigned a value of 0. Then, various skewed probability distribution functions are used to fit the normalized velocity probability distributions greater than 0. The fitting effect of each distribution is compared using the correlation coefficient R² as the evaluation index. The distribution with the largest R² is selected as the optimal fitting method, and the corresponding probability density function is obtained. Then, the cumulative probability function of the landslide displacement velocity with a velocity greater than 0 is obtained. The cumulative probability P-value of the data with a velocity of 0 corresponding to the extreme value normalization and mean normalization methods is the number of data with a velocity of 0 divided by the total number of input data. The overall cumulative probability distribution function of the landslide displacement velocity is obtained as follows: G(v)=P+(1-P)F(v) (1; S5, Standard Normalization Transformation and Output of Comprehensive Standard Deformation Index: Based on the equiprobability transformation of the cumulative probability function, the cumulative probability distribution function G(v) of the landslide displacement velocity is standardized into a standard normal distribution. At this point, the random variable of the standard normal function obtained after the cumulative probability transformation for each landslide displacement velocity value is the comprehensive standard deformation index value. The comprehensive standard deformation index values ​​corresponding to the extreme value normalization and the mean normalization methods are respectively CSDI. （最） and CSDI （均） ; S6, Warning Level Classification and Output: Based on the comprehensive standard deformation index value of the current monitoring point and its danger zone, the warning level is output; the danger zone classification of the comprehensive standard deformation index value is based on the probability distribution characteristics of the standard normal function.

[0007] Furthermore, in step S2, the method for normalizing the maximum and minimum values ​​is as follows: (2); The method for mean normalization is as follows: (3); in, The landslide deformation rate, For average speed, For the minimum speed, This represents the maximum speed.

[0008] Further, in step S4, the skewed probability distribution function includes: (1) GAMMA distribution: ( (4) (2) distributed: ( (5) (3) Lognormal distribution: ( (6) (4) Log-Logistic distribution: ( (7) (5) Logistic distribution: (8) (6) Weibull distribution: ( (9) in, , , , , , These are all parameters corresponding to each probability distribution function.

[0009] Further, in step S4, when extreme value normalization is used and the displacement velocity value is the maximum value of 1, the... The distribution cannot fit it, so the comprehensive standard deformation index value CSDI_(maximum value) corresponding to the maximum speed is set to a strong warning value greater than 2 by default.

[0010] Further, in step S6, the CSDI （最） The warning levels are divided as follows: less than 0 is safe; 0 to 1 is a level 1 warning; 1 to 2 is a level 2 warning; and 2 and above is a strong warning.

[0011] Further, in step S6, the CSDI （均） The warning levels are divided as follows: 0 to 1.5 is safe; 1.5 to 2.5 is weak warning; and 2.5 and above is strong warning.

[0012] The present invention also provides a landslide deformation early warning device, comprising: a memory, a processor, and a landslide deformation early warning program stored in the memory and executable on the processor. When the landslide deformation early warning program is executed by the processor, it implements the steps of the landslide early warning method based on the comprehensive standard deformation index as described above.

[0013] The present invention also provides a storage medium storing a landslide deformation early warning program, wherein when the landslide deformation early warning program is executed by a processor, the program implements the steps of the landslide early warning method based on the comprehensive standard deformation index as described above.

[0014] The beneficial effects of the technical solutions provided by the embodiments of the present invention are as follows: This invention uses representative slope displacements of landslides as the primary data basis, ensuring reliable and stable data sources and resulting in highly reliable early warning results. By defining steps such as displacement-velocity normalization, normality testing of velocity-displacement probability density distribution, frequency analysis, and cumulative probability standardization transformation, the method predicts landslide deformation patterns based on the probability of displacement-velocity occurrence. The calculation process is simple, the mathematical principles are accurate, and it demonstrates strong integration of data and physical mechanisms. Furthermore, the entire analysis process does not consider landslide volume or material properties, exhibiting strong generalization capabilities and wide applicability to deformation early warning for various types of landslides. Attached Figure Description

[0015] Figure 1 This is a flowchart illustrating the calculation principle of a landslide early warning method based on a comprehensive standard deformation index, provided by the present invention. Figure 2 This is a schematic diagram illustrating the application of the method of this invention to the displacement curve of the Xinwolongsi landslide; Figure 3 This is a schematic diagram illustrating the application of the method of the present invention to the displacement curve of the Xintan landslide. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0017] Please refer to Figure 1 As shown, Figure 1 The following is a flowchart illustrating the calculation principle of the method of the present invention, which introduces a landslide early warning method based on a comprehensive standard deformation index, including the following steps: S1, Data Acquisition: Based on the landslide site reconnaissance data, deformation monitoring devices, such as Global Navigation Satellite System (GNSS), inclinometers, and ground-based InSAR, are deployed at representative key deformation locations on the slope surface to continuously collect cumulative displacement time series data at these locations. The landslide deformation velocity time series is obtained by dividing the cumulative difference between two adjacent monitoring points by time: .

[0018] S2, Displacement velocity normalization: Through extreme value normalization or mean normalization, wherein the extreme value normalization can map the displacement velocity to the range [0-1], and the mean normalization can map the displacement velocity to the range [-1-1].

[0019] The method for normalizing the extreme values ​​is as follows: (2); The method for mean normalization is as follows: (3); in, The landslide deformation rate, For average speed, For the minimum speed, This represents the maximum speed.

[0020] S3, Normality test of velocity-displacement probability density distribution: First, calculate the velocity-displacement probability density distribution after the extreme value normalization and the mean normalization, and check whether it satisfies the normality distribution. If it does, it is considered to be in a stable state. If it does not, proceed to the next step.

[0021] It is worth noting that, due to the characteristics of the normalization function, the shape of the displacement velocity probability density distribution remains unchanged, while the horizontal coordinates of the extreme value normalization and the mean normalization are different. For details, please refer to formula (1) and formula (2).

[0022] S4, Cumulative Probability Fitting: First, assign a value of 0 to all negative displacement velocities. Then, fit the normalized velocity probability distribution with values ​​greater than 0 using a skewed probability distribution function. To evaluate the fitting effect of the index comparison, the optimal fitting method and its probability density function for the corresponding displacement velocity can be obtained. Then, the cumulative probability function of the landslide displacement velocity with a velocity greater than 0 is obtained. The cumulative probability P-value of the data with a velocity of 0 corresponding to the extreme value normalization and mean normalization methods is the number of data with a velocity of 0 divided by the total number of input data. The overall cumulative probability distribution function of the landslide displacement velocity is obtained as follows: G(v)=P+(1-P)F(v) (1)。

[0023] Here, due to the characteristics of the normalization method, the displacement velocity after extreme value normalization contains a zero, while the result after mean normalization contains a large number of negative numbers. For ease of subsequent calculations, all negative displacement velocities are initially assigned a value of 0. Furthermore, extreme value normalization and mean normalization methods preserve different parts of the velocity data characteristics; therefore, their optimal fitting methods are not necessarily the same. Thus, the cumulative probability functions for the positive velocity data corresponding to the two normalization methods are output here. The optimal cumulative probability function .

[0024] Specifically, the skewed probability distribution function includes: (1) GAMMA distribution: ( (4) (2) distributed: ( (5) (3) Lognormal distribution: ( (6) (4) Log-Logistic distribution: ( (7) (5) Logistic distribution: (8) (6) Weibull distribution: ( (9) in, , , , , , These are the corresponding parameters of each probability distribution function. Here, a normalized velocity probability distribution greater than 0 is fitted using multiple skewed probability distribution functions, so that... To evaluate the fit of the indicators, the correlation coefficient is used. To determine the goodness of fit, The larger the value, the better the fit. This step is to compare which fitting function performs best; the function with the best fit will be selected for the next calculation.

[0025] S5, Standard Normalization Transformation and Output of Comprehensive Standard Deformation Index: Based on the equiprobability transformation of the cumulative probability function, the cumulative probability distribution function G(v) of the landslide displacement velocity is standardized into a standard normal distribution. At this point, the random variable of the standard normal function obtained after the cumulative probability transformation for each landslide displacement velocity value is the comprehensive standard deformation index value. The comprehensive standard deformation index values ​​corresponding to the extreme value normalization and the mean normalization methods are respectively CSDI. （最） and CSDI （均） ; S6, Warning Level Classification and Output: Based on the comprehensive standard deformation index value of the current monitoring point and its dangerous range, the warning level is output.

[0026] The dangerous range of the comprehensive standard deformation index value is determined based on the probability distribution characteristics of the standard normal function.

[0027] Specifically, the CSDI （最） The warning levels are divided as follows: less than 0 is safe; 0 to 1 is a level 1 warning; 1 to 2 is a level 2 warning; and 2 and above is a strong warning.

[0028] Specifically, in step S6, the CSDI （均） The warning levels are divided as follows: 0 to 1.5 is safe; 1.5 to 2.5 is weak warning; and 2.5 and above is strong warning.

[0029] It is worth noting that in step S4, since the maximum speed will definitely be 1 after normalization of the extreme values, the... The distribution cannot fit it, therefore the CSDI corresponding to the maximum speed is... （最） This is considered abnormal; by default, any value greater than 2 is considered a strong warning, such as 3.

[0030] Different response measures should be taken based on the different warning levels, as shown in the table below: Table 1 CSDI (最) Warning level

[0031] Table 2 CSDI (均) Warning level classification

[0032] The calculation of CSDI values ​​in this method is based on landslide deformation monitoring data, and therefore needs to be based on long-term deformation monitoring. The longer the monitoring time, the more accurate the result.

[0033] Figure 2-3 This document describes the application effect of the method of the present invention in the early warning embodiments of deformation of two unstable landslides: the Xinwolongsi landslide and the Xintan landslide.

[0034] Figure 2 This is a schematic diagram illustrating the application of the method of this invention to the displacement curve of the Xinwolongsi landslide. The cumulative displacement curve of the Xinwolongsi landslide is exponential, characterized by a long-term stable and slow growth trend followed by an accelerated deformation stage, where deformation increases sharply until failure occurs. The application of this patented early warning method to the landslide shows that the CSDI of the landslide displacement velocity... (最) The optimal fitting method is the β function, while CSDI (均) The optimal fitting method is the Logistic function. The CSDI of this landslide... (最) and CSDI (均) All of them exhibit a characteristic of fluctuating at low values ​​for a long period before increasing sharply. Among them, CSDI (最) The value has fluctuated within a relatively small range over a long period and remains within a safe range. Since May 3, 1971, the CSDI... (最) The value increased sharply, quickly entering the Level 1 warning range, and then, a month before the slide, it entered the strong warning range. CSDI (均) The value has remained in a safe range for a long period without significant fluctuations. Since May 3, 1971, the CSDI... (均) The value increased sharply, quickly entering the weak warning zone, until it entered the strong warning zone a month before the slide. CSDI (最) and CSDI (均) The values ​​and corresponding early warning results are consistent with the actual landslides, demonstrating a good early warning effect.

[0035] Figure 3 This diagram illustrates the application of the method of this invention to the displacement curve of the Xintan landslide. The Xintan landslide, located in Zigui County, Three Gorges Reservoir area, China, failed and became unstable in June 1985. Its cumulative displacement curve exhibits a strong step-like characteristic. The Comprehensive Standard Deformation Index (CSDI) was calculated from deformation monitoring data from December 1977 to May 1985. The optimal fitting method for the landslide displacement velocity was the β function, with goodness of fits of 0.7689 and 0.7616, respectively. The corresponding CSDI values ​​are... (最) and CSDI (均) All exhibit characteristics consistent with cumulative displacement fluctuations. Among them, CSDI (最) The value remains within the safe range for a long period. At locations where landslide deformation exhibits significant step changes, the CSDI... (最) The value has risen to the highest warning level. This occurred in the months leading up to the impending instability of the landslide. (最) The value increased sharply from the Level 1 warning range to the strong warning range, which is consistent with the actual situation of landslides. Correspondingly, CSDI... (均)In contrast, it exhibits long-term stability but is sensitive to several monitoring data points with large displacement velocities, causing a sharp increase in the corresponding values. Therefore, the value of this index remains within the safe range for a long period. In the months leading up to the impending instability of the landslide, the CSDI... (均) The value increased sharply from a weak warning range to a strong warning range, which is consistent with the actual situation of landslides. The Comprehensive Standard Deformation Index (CSDI) in the method of this invention provides a good early warning of the failure of stepped landslides.

[0036] The present invention also provides a landslide deformation early warning device, comprising: a memory, a processor, and a landslide deformation early warning program stored in the memory and executable on the processor. When the landslide deformation early warning program is executed by the processor, it implements the steps of the landslide early warning method based on the comprehensive standard deformation index as described above.

[0037] In addition, the present invention also provides a storage medium storing a landslide deformation early warning program, which, when executed by a processor, implements the steps of the landslide early warning method based on the comprehensive standard deformation index as described above.

[0038] In this document, the directional terms such as front, back, top, and bottom are defined based on the location of the components in the accompanying drawings and their relative positions to each other, solely for the purpose of clarity and convenience in expressing the technical solution. It should be understood that the use of these directional terms should not limit the scope of protection claimed in this application.

[0039] Where there is no conflict, the above embodiments and features described herein can be combined with each other.

[0040] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A landslide early warning method based on a comprehensive standard deformation index, characterized in that, Includes the following steps: S1, Data Acquisition: Based on the landslide site reconnaissance data, deformation monitoring devices were deployed at representative key deformation locations on the slope surface to continuously collect cumulative displacement time series data at these locations. The landslide deformation velocity time series is obtained by dividing the cumulative difference between two adjacent monitoring points by time: ; S2, Displacement-velocity normalization: Through extreme value normalization and mean value normalization, wherein the extreme value normalization can map the displacement-velocity to the range [0-1], and the mean value normalization can map the displacement-velocity to the range [-1-1]. S3, Normality test of velocity-displacement probability density distribution: First, calculate the velocity-displacement probability density distribution after normalization of the extreme values ​​and the mean, and check whether it satisfies the normality distribution. If it does, the monitoring point is determined to be in a stable state; if it does not, proceed to the next step. S4, Cumulative Probability Fitting: First, all negative displacement velocities are assigned a value of 0. Then, various skewed probability distribution functions are used to fit the normalized velocity probability distributions greater than 0. The fitting effect of each distribution is compared using the correlation coefficient R² as the evaluation index. The distribution with the largest R² is selected as the optimal fitting method, and the corresponding probability density function is obtained. Then, the cumulative probability function of the landslide displacement velocity with a velocity greater than 0 is obtained. The cumulative probability P-value of data with a velocity of 0 corresponding to the extreme value normalization and mean normalization methods is the number of data with a velocity of 0 divided by the total number of input data. The overall cumulative probability distribution function of the landslide displacement velocity is obtained as follows: G(v)=P+(1-P)F(v) (1; S5, Standard Normalization Transformation and Output of Comprehensive Standard Deformation Index: Based on the equiprobability transformation of the cumulative probability function, the cumulative probability distribution function G(v) of the landslide displacement velocity is standardized into a standard normal distribution. At this point, the random variable of the standard normal function obtained after the cumulative probability transformation for each landslide displacement velocity value is the comprehensive standard deformation index value. The comprehensive standard deformation index values ​​corresponding to the extreme value normalization and the mean normalization methods are respectively CSDI. （最） and CSDI （均） ; S6 Warning level classification and output: Based on the comprehensive standard deformation index value of the current monitoring point and its dangerous zone, the warning level is output; The dangerous range of the comprehensive standard deformation index value is based on the probability distribution characteristics of the standard normal function.

2. The landslide early warning method based on the comprehensive standard deformation index as described in claim 1, characterized in that, In step S2, the method for normalizing the maximum and minimum values ​​is as follows: （2）； The method for mean normalization is as follows: （3）； in, The landslide deformation rate, For average speed, For the minimum speed, This represents the maximum speed.

3. The landslide early warning method based on the comprehensive standard deformation index as described in claim 1, characterized in that, In step S4, the skewed probability distribution function includes: (1) GAMMA distribution: ( )（4） (2) distributed: ( )（5） (3) Lognormal distribution: ( )（6） (4) Log-Logistic distribution: ( )（7） (5) Logistic distribution: （8） (6) Weibull distribution: ( )（9） in, , , , , , These are all parameters corresponding to each probability distribution function.

4. The landslide early warning method based on the comprehensive standard deformation index as described in claim 3, characterized in that, In step S4, when extreme value normalization is used and the maximum velocity of the displacement velocity value is 1, the... The distribution cannot fit it, so the comprehensive standard deformation index value CSDI_(maximum value) corresponding to the maximum speed is set to a strong warning value greater than 2 by default.

5. The landslide early warning method based on the comprehensive standard deformation index as described in claim 1 or 4, characterized in that, In step S6, the CSDI （最） The warning levels are divided as follows: less than 0 is safe; 0 to 1 is a level 1 warning; 1 to 2 is a level 2 warning; and 2 and above is a strong warning.

6. The landslide early warning method based on the comprehensive standard deformation index as described in claim 1 or 4, characterized in that, In step S6, the CSDI （均） The warning levels are divided into: 0 to 1.5 is considered safe; A value of 1.5 to 2.5 indicates a weak warning; a value of 2.5 or higher indicates a strong warning.

7. A landslide deformation early warning device, characterized in that, include: The system includes a memory, a processor, and a landslide deformation early warning program stored in the memory and executable on the processor, wherein the landslide deformation early warning program, when executed by the processor, implements the steps of the landslide early warning method based on a comprehensive standard deformation index as described in any one of claims 1 to 6.

8. A storage medium, characterized in that, The storage medium stores a landslide deformation early warning program, which, when executed by a processor, implements the steps of the landslide early warning method based on a comprehensive standard deformation index as described in any one of claims 1 to 6.

Images