A method and system for analyzing the sensitivity of highway collapse under heavy rainfall
By using an interactive piecewise weighted regression model and adaptive identification technology, the lack of accuracy in the analysis of road landslide factors in existing technologies has been solved, enabling the quantification and differentiated prevention and control of road landslide risks.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- UNIV OF SCI & TECH BEIJING
- Filing Date
- 2026-03-04
- Publication Date
- 2026-06-19
AI Technical Summary
Existing technologies are insufficient for systematic and precise sensitivity analysis of factors influencing road landslides under heavy rainfall conditions, resulting in a lack of targeted prevention and control measures and suboptimal resource allocation.
An interactive piecewise weighted regression model was adopted, and a database was established using measured data. An adaptive sensitivity index for identifying factors influencing highway landslides was constructed, and deviation standardization was performed to classify sensitivity levels and formulate differentiated prevention and control strategies.
It enables quantitative characterization and dynamic risk assessment of factors sensitive to highway landslides, improves the pertinence and accuracy of risk management, and provides a theoretical basis for differentiated prevention and control strategies.
Smart Images
Figure CN122241638A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of highway landslide factor identification technology, specifically to a sensitivity tendency analysis method and system for highway landslide influencing factors under heavy rainfall. Background Technology
[0002] In recent years, with the intensification of global climate change and the frequent occurrence of extreme heavy rainfall events, landslides induced by heavy rainfall have been increasing, seriously threatening road traffic safety and operational efficiency. Slope landslides are a complex nonlinear process, the result of the interaction of multiple factors such as hydrological seepage, soil properties, and service damage. Heavy rainfall, as an external triggering factor, significantly reduces slope stability by altering the internal seepage field, softening the mechanical parameters of the soil and rock, and exacerbating damage to existing structures.
[0003] Currently, after identifying potential landslide hazard areas, a systematic and quantitative assessment of the sensitive factors affecting highway slope stability and their dominant tendencies is often not conducted. Instead, a "one-size-fits-all" or overly conservative prevention and control measures are adopted based solely on experience. This approach not only lacks precision and specificity but also makes it difficult to achieve optimal resource allocation and refined risk management.
[0004] Therefore, it is necessary to study and develop sensitivity analysis methods for highway landslides in order to accurately assess the sensitivity and dominant tendency of various influencing factors under heavy rainfall, provide reliable risk identification basis and engineering treatment priority classification support for highway management departments, and thus improve the accuracy and economy of slope disaster prevention and mitigation. Summary of the Invention
[0005] The purpose of this invention is to supplement the shortcomings of the prior art. This invention aims to provide a method and system for analyzing the sensitivity tendency of highway landslides under heavy rainfall, which can adaptively identify the sensitivity tendency state of highway slopes and realize the unified and quantitative characterization of highway landslide sensitive elements.
[0006] On the one hand, this invention provides a sensitivity propensity analysis method for factors influencing road landslides under heavy rainfall, comprising the following steps: (1) Obtain measured data of disaster-causing factors of two types of factors affecting road landslides under heavy rainfall and establish a database; (2) Construct and calibrate an interactive piecewise weighted regression model based on the database; (3) Sensitivity indices were calculated using a calibrated interactive piecewise weighted regression model; (4) Perform deviation standardization on the sensitivity indicators of each disaster-causing factor, and classify the sensitivity level and determine the sensitivity tendency; (5) Extract elements and formulate differentiated prevention and control strategies based on different sensitivity levels and sensitivity tendencies.
[0007] In addition to the aspects and any possible implementations described above, a further implementation is provided in which the two types of influencing factors are roadbed landslide factors and pavement collapse factors; The factors contributing to roadbed landslides include: rainfall intensity, rainfall duration, groundwater level, hydraulic gradient, groundwater recharge, surface runoff recharge, soil moisture content, and soil pore water pressure. The factors contributing to pavement collapses include: vehicle load, cumulative number of standard load applications, roadbed settlement, pavement structural layer strength, fine particle loss rate, soil cohesion, soil internal friction angle, and soil weight.
[0008] In addition to the aspects and any possible implementations described above, a further implementation is provided, wherein step (2) includes: (21). Establishing a roadbed landslide function based on the rainfall intensity, rainfall duration, groundwater level, hydraulic gradient, groundwater recharge, surface runoff recharge, soil moisture content, and soil pore water pressure; (22). Establish a road collapse function based on vehicle load, cumulative standard load application times, subgrade settlement, pavement structure layer strength, fine particle loss rate, soil cohesion, soil internal friction angle and soil weight; (23). Establish a roadbed landslide weighting function based on the sample mean values of rainfall intensity, rainfall duration, groundwater level, hydraulic gradient, groundwater recharge, surface runoff recharge, soil moisture content, and soil pore water pressure; (24). Establish an adaptive function for roadbed landslides based on the roadbed landslide weight function; (25). Establish a road collapse weighting function based on the sample mean values of vehicle load, cumulative standard load application times, subgrade settlement, pavement structure layer strength, fine particle loss rate, soil cohesion, soil internal friction angle and soil weight. (26). Establish an adaptive function for road collapse based on the road collapse weight function; (27). An interactive piecewise weighted regression model is established based on the roadbed landslide function, the roadbed landslide adaptive function, the pavement collapse function, and the pavement collapse adaptive function.
[0009] In addition to the aspects and any possible implementations described above, a further implementation is provided in which the sensitivity indicators include a roadbed landslide tendency indicator and a pavement collapse tendency indicator.
[0010] In addition to the aspects and any possible implementations described above, a further implementation is provided in which the roadbed landslide tendency index is obtained by multiplying the partial derivative of the roadbed landslide function and the roadbed landslide adaptive function, and the pavement collapse tendency index is calculated by the partial derivative of the pavement collapse function and the pavement collapse adaptive function.
[0011] In addition to the aspects and any possible implementations described above, a further implementation is provided, wherein the deviation standardization processing of the sensitivity indicators of each disaster-causing factor and the classification of sensitivity levels specifically includes: comparing the standardized sensitivity indicators with a set threshold to determine the sensitivity level of highway landslides.
[0012] In addition to the aspects described above and any possible implementation, an implementation is further provided in which the threshold includes five values: 0, 0.25, 0.5, 0.75, and 1.
[0013] In addition to the aspects and any possible implementations described above, a further implementation is provided, wherein the sensitivity index of each disaster-causing element is subjected to deviation standardization processing to determine the sensitivity tendency, specifically including: when the roadbed landslide tendency index is less than the pavement collapse tendency index, then there is a pavement collapse tendency; when the roadbed landslide tendency index is greater than the pavement collapse tendency index, then there is a roadbed landslide tendency.
[0014] In addition to the aspects described above and any possible implementation, a further implementation is provided, wherein the interactive piecewise weighted regression model is the sum of the product of the roadbed landslide adaptive function and the roadbed landslide function, the product of the pavement collapse adaptive function and the pavement collapse function, and the model error term; wherein the sum of the roadbed landslide adaptive function value and the pavement collapse adaptive function value is 1.
[0015] In addition to the aspects and any possible implementations described above, a further implementation is provided in which the measured data of rainfall intensity and duration are obtained by continuously monitoring by deploying tipping bucket rain gauges around the slope.
[0016] The measured data of groundwater level were obtained by installing conductivity level gauges at the toe, middle, and top of the slope. The measured data of the hydraulic gradient were obtained by using multiple piezometers along the same streamline. The measured data on groundwater recharge and surface runoff recharge were obtained by installing weir and flume flowmeters along the main runoff paths. The measured data of surface runoff recharge were obtained by installing weir and flume flow meters at groundwater outcrops. The measured data of soil cohesion and soil internal friction angle were obtained by obtaining triaxial compression tests on standard soil core specimens using drilling detection equipment. The measured data of soil weight were obtained by weighing standard soil core samples obtained through drilling detection equipment. The measured data of soil moisture content were obtained by first using drilling equipment to obtain standard soil core samples, and then by drying method. The measured data of soil pore water pressure is obtained by embedding a piezoresistive pore water pressure sensor in the soil. The measured data of vehicle load and cumulative standard load application times were obtained through traffic monitoring data. The measured data of roadbed settlement were obtained by using a magnetic ring-type settlement sensor. The measured data of the pavement structure layer strength are obtained by first using drilling detection equipment to obtain standard core specimens of the pavement structure layer, and then by uniaxial compression test. The measured data on the loss rate of fine particles were obtained by collecting particles by installing a filter screen at the seepage outlet, drying them, and then weighing them.
[0017] This invention also provides a sensitivity propensity analysis system for factors influencing road landslides under heavy rainfall, the system being used to implement the method, including... The acquisition module is used to acquire measured data of disaster-causing factors affecting road landslides under heavy rainfall and establish a database; The building module is used to construct and calibrate an interactive piecewise weighted regression model based on the database. The calculation module is used to calculate sensitivity indicators and sensitivity propensity indicators using a calibrated interactive piecewise weighted regression model; The segmentation module is used to perform deviation standardization on the sensitivity indicators of each disaster-causing factor, and to classify the sensitivity level and determine the sensitivity tendency. The prevention and control module is used to extract elements and formulate differentiated prevention and control strategies based on different sensitivity levels and sensitivity tendencies.
[0018] This invention presents a method and system for analyzing the sensitivity of highways to landslides under heavy rainfall. By introducing an adaptive weighting mechanism to identify the sensitivity of influencing factors, it can dynamically characterize the synergistic effect of two types of influencing factors—subgrade landslides and pavement collapses—on displacement response. This achieves unified processing and quantitative characterization of sensitive elements for highway landslides. The evaluation results can provide guidance and basis for identifying the potential dominant risks of subgrade landslides and pavement collapses and prioritizing engineering remediation. Due to the adoption of the above technical solution, this invention has the following advantages compared with existing technologies: (1) Quantitative characterization of sensitivity response characteristics under the synergistic effect of multiple factors Currently, existing technologies mostly use simple parameter perturbations or the first second moment method for sensitivity analysis. These methods are difficult to reflect the complex interactions between various influencing factors. This invention is based on an interactive piecewise weighted regression model. By introducing a sub-function structure containing explicit interaction terms, it quantitatively characterizes the sensitivity response features under the synergistic effect of multiple factors.
[0019] (2) An adaptive sensitivity tendency identification mechanism is proposed. Unlike the static and fixed parameter weight system in traditional sensitivity analysis, this invention responds to changes in the external environment and sensitive elements in real time through an adaptive weight function. It can dynamically adjust the weights of two types of influencing factors, namely roadbed landslide and pavement collapse, according to different working conditions, thereby realizing the identification of sensitivity tendency and improving the pertinence and accuracy of risk control.
[0020] (3) The risk level of highway landslides was quantitatively evaluated. This invention combines quantitative analysis results with engineering prevention and control, and constructs a differentiated prevention and control strategy library based on sensitivity level and tendency type, providing a theoretical basis for targeted prevention and control and refined management of highway landslide risks.
[0021] Of course, any product implementing this invention does not necessarily need to achieve all of the technical effects described above at the same time. Attached Figure Description
[0022] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0023] Figure 1 This is a flowchart of the highway landslide sensitivity analysis method in an embodiment of the present invention. Detailed Implementation
[0024] To better understand the technical solution of the present invention, the embodiments of the present invention will be described in detail below with reference to the accompanying drawings.
[0025] It should be understood that the described embodiments are merely some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.
[0026] In view of the shortcomings of the existing technology, the present invention provides a method for analyzing the sensitivity tendency of road landslides under heavy rainfall, including: (1) obtaining measured data of disaster-causing factors of two types of influencing factors of road landslides under heavy rainfall and establishing a database; (2) Construct and calibrate an interactive piecewise weighted regression model based on the database; (3) Sensitivity indices were calculated using a calibrated interactive piecewise weighted regression model; (4) Perform deviation standardization on the sensitivity indicators of each disaster-causing factor, and classify the sensitivity level and determine the sensitivity tendency; (5) Extract elements and formulate differentiated prevention and control strategies based on different sensitivity levels and sensitivity tendencies.
[0027] Preferably, the two types of influencing factors are roadbed landslide factors and pavement collapse factors; The factors contributing to roadbed landslides include: rainfall intensity, rainfall duration, groundwater level, hydraulic gradient, groundwater recharge, surface runoff recharge, soil moisture content, and soil pore water pressure. The factors contributing to pavement collapses include: vehicle load, cumulative number of standard load applications, roadbed settlement, pavement structural layer strength, fine particle loss rate, soil cohesion, soil internal friction angle, and soil weight.
[0028] Preferably, step (2) includes: (21). Establishing a roadbed landslide function based on the rainfall intensity, rainfall duration, groundwater level, hydraulic gradient, groundwater recharge, surface runoff recharge, soil moisture content, and soil pore water pressure; (22). Establish a road collapse function based on vehicle load, cumulative standard load application times, subgrade settlement, pavement structure layer strength, fine particle loss rate, soil cohesion, soil internal friction angle and soil weight; (23). Establish a roadbed landslide weighting function based on the sample mean values of rainfall intensity, rainfall duration, groundwater level, hydraulic gradient, groundwater recharge, surface runoff recharge, soil moisture content, and soil pore water pressure; (24). Establish an adaptive function for roadbed landslides based on the roadbed landslide weight function; (25). Establish a road collapse weighting function based on the sample mean values of vehicle load, cumulative standard load application times, subgrade settlement, pavement structure layer strength, fine particle loss rate, soil cohesion, soil internal friction angle and soil weight. (26). Establish an adaptive function for road collapse based on the road collapse weight function; (27). An interactive piecewise weighted regression model is established based on the roadbed landslide function, the roadbed landslide adaptive function, the pavement collapse function, and the pavement collapse adaptive function.
[0029] Preferably, the sensitivity indicators include roadbed landslide tendency indicators and pavement collapse tendency indicators.
[0030] Preferably, the roadbed landslide tendency index is obtained by multiplying the partial derivative of the roadbed landslide function and the roadbed landslide adaptive function, and the pavement collapse tendency index is calculated by the partial derivative of the pavement collapse function and the pavement collapse adaptive function.
[0031] Preferably, the step of performing deviation standardization processing on the sensitivity indicators of each disaster-causing factor and classifying the sensitivity levels specifically includes: comparing the standardized sensitivity indicators with a set threshold to determine the sensitivity level of highway landslides.
[0032] Preferably, the threshold includes five values: 0, 0.25, 0.5, 0.75, and 1.
[0033] Preferably, the step of performing deviation standardization processing on the sensitivity indicators of each disaster-causing factor to determine the sensitivity tendency specifically includes: when the roadbed landslide tendency index is less than the pavement collapse tendency index, then there is a pavement collapse tendency; when the roadbed landslide tendency index is greater than the pavement collapse tendency index, then there is a roadbed landslide tendency.
[0034] Preferably, the interactive piecewise weighted regression model is the sum of the product of the roadbed landslide adaptive function and the roadbed landslide function, the product of the pavement collapse adaptive function and the pavement collapse function, and the model error term; wherein the sum of the roadbed landslide adaptive function value and the pavement collapse adaptive function value is 1.
[0035] Specifically, such as Figure 1 As shown, the entire method process of the present invention is as follows: Step 1: Identify the disaster-causing factors of the two types of factors affecting road landslides under heavy rainfall; The two categories of influencing factors include: roadbed landslide factors and pavement collapse factors. Specifically, the elements of roadbed landslide factors include: rainfall intensity, rainfall duration, groundwater level, hydraulic gradient, groundwater recharge, surface runoff recharge, soil moisture content, and soil pore water pressure. The elements of pavement collapse factors include: vehicle load, cumulative number of standard load applications, roadbed settlement, pavement structural layer strength, fine particle loss rate, soil cohesion, soil internal friction angle, and soil weight.
[0036] Step 2: Obtain measured data of disaster-causing factors and establish a disaster-causing factor database; Measured data on rainfall intensity and duration were obtained through continuous monitoring using tipping bucket rain gauges deployed around the slope perimeter; measured data on groundwater level were obtained through conductivity level gauges deployed at the slope toe, middle, and top; measured data on hydraulic gradient were obtained through multiple piezometers along the same streamline; measured data on groundwater recharge and surface runoff recharge were obtained through weir flowmeters deployed along the main runoff paths; measured data on surface runoff recharge were obtained through weir flowmeters deployed at groundwater outcrops; measured data on soil cohesion and internal friction angle were obtained through triaxial compression tests on standard soil core samples obtained using drilling-while-exploration equipment; measured data on soil weight were obtained through drilling-while-exploration equipment. The measured data for soil moisture content were obtained by weighing standard core specimens; the measured data for soil pore water pressure were obtained by first obtaining standard core specimens using drilling equipment, and then obtaining the moisture content through drying; the measured data for soil pore water pressure were obtained by embedding piezoresistive pore water pressure sensors in the soil; the measured data for vehicle load and cumulative standard load cycles were obtained from traffic monitoring data; the measured data for subgrade settlement were obtained from magnetic ring settlement sensors; the measured data for pavement structural layer strength were obtained by first obtaining standard core specimens of the pavement structural layer using drilling equipment, and then obtaining the strength through uniaxial compression tests; the measured data for fine particle loss rate were obtained by collecting particles by installing filters at the seepage outlet, drying them, and then weighing them.
[0037] Step 3: Construct an interactive piecewise weighted regression model; The independent variables used in the model are rainfall intensity, rainfall duration, groundwater level, hydraulic gradient, groundwater recharge, surface runoff recharge, soil moisture content, soil pore water pressure, vehicle load, cumulative number of standard load applications, subgrade settlement, pavement structure layer strength, fine particle loss rate, soil cohesion, soil internal friction angle, and soil weight; the dependent variable is the horizontal displacement at the toe of the slope. The constructed interactive piecewise weighted regression model is as follows: (1) In the above formula, Y is the dependent variable, namely the horizontal displacement at the toe of the slope, and X1 is the set of independent variables of the disaster-causing factors of roadbed landslide, including rainfall intensity x. 11 Rainfall duration x 12 Groundwater level height x 13 Hydraulic gradient x 14 Groundwater recharge x 15 Surface runoff replenishment x 16 Soil moisture content x 17 Soil pore water pressure x 18 These are the eight disaster-causing factors. X2 represents the set of independent variables for road collapse disaster-causing factors, including vehicle load x. 21 Cumulative number of standard load applications x22 Roadbed settlement x 23 , pavement structure layer strength x 24 Fine particle loss rate x 25 Soil cohesion x 26 , soil internal friction angle x 27 Soil weight x 28 These eight disaster-causing factors, totaling 16, are used to ensure the reliability and effectiveness of the interactive piecewise weighted regression model.
[0038] ω1 is the adaptive function for roadbed landslide, ω2 is the adaptive function for pavement collapse, f1(X1) is the roadbed landslide function, f2(X2) is the pavement collapse function, and ɛ is the model error; where ω1 and ω2 satisfy ω1+ω2=1.
[0039] Step 4: Determine the sub-function structure and adaptive weight function; The sub-function structure in the interactive piecewise weighted regression model includes a roadbed landslide function and a pavement collapse function, where the roadbed landslide function f1(X1) is: (2) In the above formula, α 11 α 12 α 13 α 14 α 15 α 16 α 17 α 18 The coefficients, in order, are rainfall intensity, rainfall duration, groundwater level, hydraulic gradient, groundwater recharge, surface runoff recharge, soil moisture content, and soil pore water pressure, β. 11 β is the regression coefficient of the interaction term between rainfall duration and groundwater level. 12 β is the regression coefficient of the interaction term between rainfall intensity and surface runoff recharge. 13 β is the regression coefficient of the interaction term between hydraulic gradient and groundwater recharge. 14 This represents the regression coefficient of the interaction term between soil moisture content and soil pore water pressure.
[0040] The road surface collapse function f2(X2) is: (3) In the above formula, α 20 α is the constant term of the road surface collapse function. 21 α 22 α 23 α 24 α 25 α 26 α 27 α 28The regression coefficients, in order, are vehicle load, cumulative standard load application times, subgrade settlement, pavement structure layer strength, fine particle loss rate, soil cohesion, soil internal friction angle, and soil weight, β. 21 β is the regression coefficient of the interaction term between vehicle load and pavement structure layer strength. 22 β is the regression coefficient of the interaction term between the cumulative number of standard load applications and the roadbed settlement. 23 is the regression coefficient of the interaction term between soil cohesion and soil internal friction angle.
[0041] The adaptive weighting functions of the interactive piecewise weighted regression model include an adaptive function for roadbed landslides and an adaptive function for pavement collapses. The adaptive function for roadbed landslides is expressed as follows: (4) In the above formula, D1 is the weight function for roadbed landslides, and D2 is the weight function for road surface collapses.
[0042] The adaptive function for road surface collapse is expressed as: (5) In the two formulas above, the roadbed landslide weight function D1 is expressed as: (6) Where, m 11 m 12 m 13 m 14 m 15 m 16 m 17 m 18 The values, in order, are the sample means of rainfall intensity, rainfall duration, groundwater level, hydraulic gradient, groundwater recharge, surface runoff recharge, soil moisture content, and soil pore water pressure among the independent variables, θ. 10 θ is the constant term of the roadbed landslide weight function. 11 θ 12 θ 13 θ 14 θ 15 θ 16 θ 17 θ 18 The weighted parameters are, in order: rainfall intensity, rainfall duration, groundwater level, hydraulic gradient, groundwater recharge, surface runoff recharge, soil moisture content, and soil pore water pressure.
[0043] The road collapse weighting function D2 is expressed as: (7) In the above formula, m 21 m 22 m23 m 24 m 25 m 26 m 27 m 28 The sample mean values, in order, are: vehicle load, cumulative number of standard load applications, subgrade settlement, pavement structure layer strength, fine particle loss rate, soil cohesion, soil internal friction angle, and soil weight. 20 θ is the constant term of the road surface collapse weighting function. 21 θ 22 θ 23 θ 24 θ 25 θ 26 θ 27 θ 28 The weighted parameters are, in order: vehicle load, cumulative number of standard load applications, subgrade settlement, pavement structure layer strength, fine particle loss rate, soil cohesion, soil internal friction angle, and soil weight.
[0044] Step 5: Perform two-stage alternating model parameter calibration and complete the overall solution; Build a calibration platform using PYTHON, MATLAB, or R language, and import the disaster-causing element database from step 2 (containing measured data of 16 disaster-causing elements and corresponding horizontal displacement of the slope toe), the model framework from step 3, and the sub-functions and adaptive weight functions from step 4. Within the overall framework of the interactive piecewise weighted regression model, with fixed initial values for weights ω1 and ω2, the constant term α of the roadbed landslide function f1(X1) in the model is fitted using the least squares method. 10 The coefficient of a single term α 11 α 12 α 13 α 14 α 15 α 16 α 17 α 18 Interaction term coefficient β 11 β 12 β 13 β 14 And the constant term α of the road surface collapse function f2(X2). 20 The coefficient of a single term α 21 α 22 α 23 α 24 α 25 α 26 α 27 α 28 Interaction term coefficient β 21 β 22 β 23This yields the initial sub-function model of the model. Based on the initial sub-function model, the adaptive weight function is substituted, the sub-function coefficients are fixed, and the constant term θ of the roadbed landslide weight function D1 is solved by minimizing the sum of squared residuals. 10 Weighting parameter θ 11 θ 12 θ 13 θ 14 θ 15 θ 16 θ 17 θ 18 And the constant term θ of the road surface collapse weighting function D2. 20 Weighting parameter θ 21 θ 22 θ 23 θ 24 θ 25 θ 26 θ 27 θ 28 Then, fix the weight parameters, update the coefficients of the sub-functions, and repeat the iteration until the sum of squared residuals converges, at which point the calibration process ends. The normality of the residual sequence of the calibrated model is tested, and a significance level is set, preferably 0.05 in this invention. The mean of the residuals is taken as a fixed value of the error term α. ; Substituting the final calibrated sub-function coefficients, weight parameters, and error terms into the original model yields a complete calibrated interactive piecewise weighted regression model. , These represent the horizontal displacement at the toe of the slope after calibration, the adaptive function of the roadbed landslide, the roadbed landslide function, the adaptive function of the pavement collapse, the pavement collapse function, and the model error, respectively, thus providing a definite functional form and parameter value for the calculation of the sensitivity index in step 6 below; After calibration with measured data, the above polynomial regression model characterizes the unified quantitative relationship between highway landslide elements and the horizontal displacement at the toe of the slope.
[0045] Step 6: Calculate the sensitivity index and sensitivity propensity index based on the calibrated interactive piecewise weighted regression model; The formula for calculating the sensitivity index is: (8) In the above formula, S a As sensitivity indicators, X1, X2, All values are known.
[0046] Sensitive tendency indicators include roadbed landslide tendency indicators and pavement collapse tendency indicators. The calculation formula for the roadbed landslide tendency indicator is as follows: (9) In the above formula, I1 is the roadbed landslide tendency index.
[0047] The formula for calculating the road surface collapse tendency index is as follows: (10) In the above formula, I2 is the road surface collapse tendency index.
[0048] Step 7: Perform deviation standardization on the sensitivity indices of each factor, and classify the sensitivity levels and determine the sensitivity tendency; The specific implementation steps of step 7 include: calculating S. a The discrete standardized value, the deviation standardization calculation formula is: (11) In the above formula, S b For S a The discrete standardized value, S amin For S a The minimum value of S amax For S a The maximum value, S amin and S amax All of these are preset values.
[0049] The landslide sensitivity level of highways was classified using the equidistant quartering method, with five threshold values set: 0, 0.25, 0.5, 0.75, and 1. The specific classification of highway landslide sensitivity levels is as follows: Grade A sensitivity (extremely high sensitivity): 0.75 < S b ≤1, Grade B sensitivity (high sensitivity): 0.5 < S b ≤0.75, C-level sensitivity (weak sensitivity): 0.25 < S b ≤0.5, Grade D sensitivity (very weak sensitivity): 0 ≤ S b ≤0.25.
[0050] The dominant discrimination method is used to determine the sensitivity tendency of highways to landslides. The specific method for determining the sensitivity tendency of highways to landslides is as follows: Roadbed landslide tendency: I2 < I1, Road surface collapse tendency: I1 < I2.
[0051] Step 8: Extract elements and formulate differentiated prevention and control strategies based on different sensitivity levels and sensitivity tendencies.
[0052] The specific implementation steps of step 8 include: extracting sensitivity level and sensitivity tendency elements according to different sensitivity levels and sensitivity tendencies, and formulating differentiated prevention and control strategies for highway slopes; The specific differentiated prevention and control strategies for highway slopes are as follows: When a highway slope is prone to subgrade landslides and is sensitive at level A, intercepting ditches and blind ditches should be added at the top and bottom of the slope, water pumps should be installed to achieve real-time drainage, impermeable membranes should be laid to block rainwater infiltration, grid beams should be laid at the bottom of the slope, grouting reinforcement should be carried out in areas with weak soil, traffic speed should be restricted, and weighing monitoring points should be added to provide early warning for situations where the cumulative load exceeds the standard; When a highway slope is prone to subgrade landslides and is sensitive at level B, existing drainage ditches should be dredged, temporary drainage channels should be added, and pre-drainage should be carried out before rain. Water-filled puddles should be removed, and the monitoring frequency of surface and deep displacement of the slope should be increased to assess the risk of instability. Preventative maintenance measures such as sealing and crack sealing should be implemented. When the highway slope is prone to landslides and the sensitivity level is C, drainage facilities should be maintained before rainfall to ensure they are unobstructed, and silt should be cleared immediately after rainfall. Regular inspections and periodic assessments of the highway subgrade soil and pavement should also be conducted. When the highway slope is prone to landslides and the sensitivity level is D, drainage facilities should be routinely inspected and periodically cleaned. Regular inspections and periodic assessments of the highway subgrade soil and pavement should also be conducted. Regular inspections and periodic assessments will be conducted. When a highway slope is prone to road collapse and is classified as sensitive at Level A, strict traffic diversion will be implemented, and severely damaged sections will be closed. Epoxy resin grouting will be used to repair road surface cracks. Existing drainage ditches will be cleared, and temporary drainage channels will be added. Pre-rain drainage and water storage pits will be prepared before rain. Grid beams will be laid at the toe of the slope, and weak areas will be reinforced with grouting. When a highway slope is prone to road collapse and is classified as sensitive at Level B, traffic speed will be restricted, additional weighing monitoring points will be installed, and overloading will be promptly warned. Pre-rain maintenance and drainage will also be carried out. Water facilities should be cleaned of silt after rain, and the frequency of monitoring surface and deep displacement of slopes should be increased to dynamically assess risks. When a highway slope is prone to road collapse and is sensitive at level C, preventive maintenance plans such as sealing and crack sealing should be implemented. At the same time, drainage facilities should be regularly inspected and cleaned, and the highway subgrade soil should be regularly inspected and assessed. When a highway slope is prone to road collapse and is sensitive at level D, drainage facilities should be regularly inspected and cleaned, and the highway subgrade soil and pavement should be regularly inspected and assessed.
[0053] The foregoing description illustrates and describes several preferred embodiments of the present invention. However, as previously stated, it should be understood that the present invention is not limited to the forms disclosed herein and should not be construed as excluding other embodiments. It can be used in various other combinations, modifications, and environments, and can be altered within the scope of the inventive concept described herein through the foregoing teachings or techniques or knowledge in related fields. Any modifications and variations made by those skilled in the art that do not depart from the spirit and scope of the present invention should be within the protection scope of the appended claims.
Claims
1. A sensitivity trend analysis method for factors influencing road landslides under heavy rainfall, characterized in that, Includes the following steps: (1) Obtain measured data of disaster-causing factors of two types of factors affecting road landslides under heavy rainfall and establish a database; (2) Construct and calibrate an interactive piecewise weighted regression model based on the database; (3) Sensitivity indices were calculated using a calibrated interactive piecewise weighted regression model; (4) Perform deviation standardization on the sensitivity indicators of each disaster-causing factor, and classify the sensitivity level and determine the sensitivity tendency; (5) Extract elements and formulate differentiated prevention and control strategies based on different sensitivity levels and sensitivity tendencies.
2. The method according to claim 1, characterized in that, The two types of influencing factors are roadbed landslide factors and pavement collapse factors; The factors contributing to roadbed landslides include: rainfall intensity, rainfall duration, groundwater level, hydraulic gradient, groundwater recharge, surface runoff recharge, soil moisture content, and soil pore water pressure. The factors contributing to pavement collapses include: vehicle load, cumulative number of standard load applications, roadbed settlement, pavement structural layer strength, fine particle loss rate, soil cohesion, soil internal friction angle, and soil weight.
3. The method according to claim 2, characterized in that, The step (2) includes: (21). Establishing a roadbed landslide function based on the rainfall intensity, rainfall duration, groundwater level, hydraulic gradient, groundwater recharge, surface runoff recharge, soil moisture content and soil pore water pressure; (22). Establish a road collapse function based on vehicle load, cumulative standard load application times, subgrade settlement, pavement structure layer strength, fine particle loss rate, soil cohesion, soil internal friction angle and soil weight; (23). Establish a roadbed landslide weighting function based on the sample mean values of rainfall intensity, rainfall duration, groundwater level, hydraulic gradient, groundwater recharge, surface runoff recharge, soil moisture content, and soil pore water pressure; (24). Establish an adaptive function for roadbed landslides based on the roadbed landslide weight function; (25). Establish a road collapse weighting function based on the sample mean values of vehicle load, cumulative standard load application times, subgrade settlement, pavement structure layer strength, fine particle loss rate, soil cohesion, soil internal friction angle and soil weight. (26). Establish an adaptive function for road collapse based on the road collapse weight function; (27). An interactive piecewise weighted regression model is established based on the roadbed landslide function, the roadbed landslide adaptive function, the pavement collapse function, and the pavement collapse adaptive function.
4. The method according to claim 3, characterized in that, The sensitivity indicators include roadbed landslide tendency indicators and pavement collapse tendency indicators.
5. The method according to claim 4, characterized in that, The roadbed landslide tendency index is obtained by multiplying the partial derivative of the roadbed landslide function and the roadbed landslide adaptive function, and the pavement collapse tendency index is calculated by the partial derivative of the pavement collapse function and the pavement collapse adaptive function.
6. The method according to claim 1, characterized in that, The process of standardizing the sensitivity indicators of each disaster-causing factor and classifying the sensitivity levels specifically includes comparing the standardized sensitivity indicators with a set threshold to determine the sensitivity level of highway landslides.
7. The method according to claim 6, characterized in that, The threshold values include five values: 0, 0.25, 0.5, 0.75, and 1.
8. The method according to claim 6, characterized in that, The process of standardizing the sensitivity indices of each disaster-causing factor to determine the sensitivity tendency specifically includes: when the roadbed landslide tendency index is less than the pavement collapse tendency index, there is a pavement collapse tendency; when the roadbed landslide tendency index is greater than the pavement collapse tendency index, there is a roadbed landslide tendency.
9. The method according to claim 3, characterized in that, The interactive piecewise weighted regression model is the sum of the product of the roadbed landslide adaptive function and the roadbed landslide function, the product of the pavement collapse adaptive function and the pavement collapse function, and the model error term. The sum of the adaptive function values for roadbed landslides and road surface collapse is 1.
10. A sensitivity trend analysis system for factors influencing road landslides under heavy rainfall, characterized in that, The system is used to implement the method according to any one of claims 1-9, including The acquisition module is used to acquire measured data of disaster-causing factors affecting road landslides under heavy rainfall and establish a database; The building module is used to construct and calibrate an interactive piecewise weighted regression model based on the database. The calculation module is used to calculate sensitivity indicators and sensitivity propensity indicators using a calibrated interactive piecewise weighted regression model; The segmentation module is used to perform deviation standardization on the sensitivity indicators of each disaster-causing factor, and to classify the sensitivity level and determine the sensitivity tendency. The prevention and control module is used to extract elements based on different sensitivity levels and sensitivity tendencies. And formulate differentiated prevention and control strategies.