Intelligent landslide early warning method based on aging damage mechanism

Abstract

The invention relates to a landslide intelligent early-warning method based on an aging damage mechanism, and the method comprises the steps: building a landslide numerical model, and carrying out the assignment of the mechanical parameters of a rock-soil body; setting a numerical calculation initial condition and a boundary condition for elastic calculation, and calculating and generating a landslide initial ground stress model; calling the dynamic constitutive model, and performing numerical calculation through a creep command; embedding an improved tangent angle threshold algorithm; the system collects monitoring data in real time and automatically completes damage accumulation quantification based on the aging damage dynamic constitutive model; embedding the aging damage early warning model identification system into a computer system; and calculating and identifying the aging damage early warning model, and outputting an early warning result. According to the invention, through the dynamic damage model, the damage evolution rule of the rock-soil body under the long-term natural environment effect is accurately reflected, and the timeliness and accuracy of the early warning system are ensured; the corrected tangent angle lower limit value is adopted, so that the reliability of landslide early warning is improved, the landslide can be found and early warned earlier, and disaster prevention and reduction of the landslide are facilitated.

Description

Technical Field

[0001] The present invention relates to the technical field of geotechnical engineering, and in particular to an intelligent landslide warning method based on the time-dependent damage mechanism. Background Art

[0002] Existing landslide warning methods mainly include the monitoring of real-time displacement, stress or hydrological parameters, and rely on static mechanical models or data-driven algorithms for stability assessment. However, these methods generally ignore the time-dependent damage effect of rock and soil masses under long-term natural environment. Time-dependent damage is manifested as the progressive deterioration of strength and damage accumulation of rock and soil materials due to time-dependent factors such as weathering, wet-dry cycles, freeze-thaw cycles and creep. Its process has the characteristics of concealment, irreversibility and nonlinearity, which may accelerate the landslide instability process and induce sudden failure. For example, in cold-region landslides, freeze-thaw cycles will cause the pores in the rock mass to expand and close repeatedly, resulting in the gradual penetration of the fracture network; in reservoir bank landslides, the periodic rise and fall of the water level accelerates the softening of rock and soil masses and the dissolution of cementing substances, causing the shear strength to continuously decay.

[0003] Existing warning systems only judge risks through instantaneous displacement thresholds or short-term data trends, which can neither quantify the spatio-temporal evolution law of damage accumulation nor capture the "silent instability" (i.e., sudden failure without obvious surface signs) caused by long-term damage of deep rock and soil masses. More seriously, traditional numerical models mostly calculate stability based on initial mechanical parameters, ignoring the time-dependent damage characteristics of material parameters; while machine learning models can fit historical data, but due to the lack of physical mechanism embedding of damage evolution, the reliability of extrapolation prediction under extreme climate or complex geological conditions is significantly reduced. Summary of the Invention

[0004] The technical problem to be solved by the present invention is: to provide an intelligent landslide warning method based on the time-dependent damage mechanism, which realizes the leap from "instantaneous state monitoring" to "whole-process failure damage prediction" of landslide stability by constructing a dynamic constitutive model considering the time-dependent damage of rock and soil masses, combining damage accumulation quantification monitoring technology, and an intelligent warning method for damage evolution, and improves the forward-looking and accuracy of the warning system.

[0005] The technical solution adopted by the present invention to solve its technical problems is: an intelligent landslide warning method based on the time-dependent damage mechanism, including the following steps,

[0006] S1. Establish a landslide numerical model, and obtain relevant mechanical parameters of the landslide rock and soil masses through test data and exploration data and assign values;

[0007] S2. Set the initial conditions and boundary conditions for numerical calculation, and then perform elastic calculation to calculate and generate the initial in-situ stress model of the landslide;

[0008] S3. Call the dynamic constitutive model considering the time-dependent damage of rock and soil masses and perform numerical calculations through the creep command;

[0009] S4. Embed the improved tangent angle threshold algorithm and automatically calculate in combination with the real-time monitoring data;

[0010] S5. The system collects the monitoring data in real time and automatically completes the quantification of damage accumulation based on the dynamic constitutive model of time-dependent damage;

[0011] S6. Embed the identification system of the time-dependent damage warning model into the computer system;

[0012] S7. Calculate and identify the time-dependent damage warning model and output the warning result.

[0013] Furthermore, in step S3 of the present invention, the construction process of the dynamic constitutive model includes the following steps:

[0014] S31. Express the deterioration process of the strength parameters of rock and soil masses through the damage function D(t), that is

[0015]

[0016] where c0 is the initial cohesion, c (t) is the cohesion at time t, c f is the residual cohesion;

[0017] S32. Dynamically adjust the damage function D(t) according to the FLAC3D numerical calculation. The formula is:

[0018]

[0019] where is the damage increment, Δσ(t) is the stress change amount of the damage model, and σ 0 is the initial stress;

[0020] S33. Based on the Burgers model, construct a constitutive model considering time-dependent damage, and combine the damage function D(t) with the Burgers model. The formula is:

[0021]

[0022] where is the initial stress tensor, and D(t) is the damage function at time t;

[0023] S34. Solve the stress-strain relationship of the constitutive model considering time-dependent damage through the finite difference method. The formula is

[0024]

[0025] where is the damage increment, Δσ(t) is the stress change of the damage model, and σ 0 is the initial stress.

[0026] Furthermore, in step S4 of the present invention, the improved tangent angle threshold algorithm includes the following steps:

[0027] S41. Process the displacement-time dimension change into a Ti-t curve to make the horizontal and vertical coordinate dimensions consistent.

[0028]

[0029] In the formula, T i is the vertical coordinate dimension same as the horizontal coordinate after transformation; ε i is the cumulative deformation within time t i ; v i is the average rate within time t i ; m is the number of deformation monitoring times in the constant-speed deformation stage.

[0030] S42. Perform non-linear fitting through a regression model to obtain the improved tangent angle formula:

[0031]

[0032] The beneficial effects of the present invention are as follows: It solves the defects in the background technology.

[0033] 1. Full-process coverage. For the first time, a dynamic constitutive model considering the time-dependent damage of rock and soil masses is constructed. By introducing a damage factor, the full-process evolution simulation from initial damage accumulation to critical failure is realized.

[0034] 2. Based on the global geological disaster database and multi-source remote sensing monitoring data, 327 slope cases with typical time-dependent damage characteristics at home and abroad are integrated. By performing non-linear fitting with a regression model, the tangent angle is corrected to improve reliability.

[0035] 3. The model can replace the deployment of some high-cost sensors and realize low-cost and high-precision slope monitoring through a data-driven method.

[0036] 4. By introducing a time-dependent damage mechanism, the model can be applied to typical scenarios such as progressive creep-type slopes. Brief Description of the Drawings

[0037] Figure 1 is the schematic diagram of the method flow of the present invention;

[0038] Figure 2 is the comparison diagram between the actual monitoring points and the model monitoring points of the Majiagou landslide J03 in the embodiment of the present invention. Detailed Embodiments

[0039] The present invention will now be further described in detail with reference to the accompanying drawings and preferred embodiments. These drawings are all simplified schematic diagrams, only illustrating the basic structure of the present invention in a schematic manner, and therefore only showing the components related to the present invention.

[0040] As Figure 1 shown, a landslide intelligent early warning method based on the aging damage mechanism includes the following steps:

[0041] (1) First, establish a landslide numerical model on the FLAC3D numerical analysis software, and obtain relevant landslide rock and soil mechanics parameters through test data and exploration data and assign values.

[0042] (2) Set the initial conditions and boundary conditions for numerical calculation, then turn on the elastic calculation, calculate and generate the initial in-situ stress model of the landslide.

[0043] (3) Call the dynamic constitutive model considering the aging damage of rock and soil, and turn on the creep command for numerical calculation.

[0044] (4) Embed the improved tangent angle threshold algorithm, and automatically calculate in combination with real-time monitoring data.

[0045] (5) The system collects monitoring data in real time, and automatically completes the quantification of damage accumulation based on the aging damage dynamic constitutive model.

[0046] (6) Through FISH language and software programming control, embed the aging damage early warning model recognition system into the computer system.

[0047] (7) Calculate and identify the aging damage early warning model, and output the early warning result.

[0048] The present invention first constructs a dynamic constitutive model considering the aging damage of rock and soil. By introducing aging damage factors (such as weathering rate, freeze-thaw cycle times, and creep effect), the whole process evolution simulation from initial damage accumulation to critical failure is realized. Based on the aging damage characteristics of rock and soil strength parameters, a UDM interface program is written in C++ language and embedded into the FLAC3D command stream file. Combined with FISH language programming control, a dynamic constitutive model that can accurately describe the aging damage of landslide rock and soil is established.

[0049] Specifically, the deterioration process of rock and soil strength parameters is expressed by the damage function D(t): that is

[0050]

[0051] In the formula, c0 is the initial cohesion, c (t) is the cohesion at time t, c f is the residual cohesion;

[0052] Dynamically adjust the damage function D(t) according to FLAC3D numerical calculation, and the formula is:

[0053]

[0054] Among them, is the damage increment, Δσ(t) is the stress change of the damage model, and σ 0 is the initial stress;

[0055] Based on the Burgers model, construct a constitutive model considering time-dependent damage, and combine the damage function D(t) with the Burgers model. The formula is:

[0056]

[0057] In the formula, is the initial stress tensor, and D(t) is the damage function at time t;

[0058] Solve the stress-strain relationship of the constitutive model considering time-dependent damage by the finite difference method. The formula is

[0059]

[0060] Among them, is the damage increment, Δσ(t) is the stress change of the damage model, and σ 0 is the initial stress.

[0061] Through the dynamic damage model, accurately reflect the damage evolution law of rock and soil masses under the long-term natural environment, and ensure the timeliness and accuracy of the early warning system.

[0062] In addition, previous studies have shown that for different landslides, the dimensions of the cumulative deformation-time curve are different. If the time dimension of the abscissa or the deformation dimension of the ordinate is randomly switched, the displacement tangent angle is very different. Therefore, the previous displacement tangent angle criterion cannot be used for effective early warning.

[0063] In response to this, the present invention proposes an improved tangent angle threshold calculation method considering the time-dependent damage characteristics of landslides for landslide approaching warning. This method constructs a non-linear relationship model between the constant velocity deformation rate of the landslide and the approaching tangent angle by fusing multi-source monitoring data (such as InSAR deformation data, microseismic monitoring information, and damage inversion results) and combining the regression model algorithm.

[0064] Based on the global geological disaster database and multi-source remote sensing monitoring data, this invention integrates 327 slope cases with typical time-dependent damage characteristics at home and abroad, covering types such as rock landslides, soil landslides, and composite landslides, and intelligently divides the landslide evolution stages. The research finds that there is a significant negative correlation between the constant velocity deformation rate and the tangent angle at the verge of sliding, and a non-linear fitting is carried out through a regression model.

[0065] First, the displacement-time dimension change is processed into a Ti-t curve to make the dimensions of the horizontal and vertical coordinates consistent.

[0066]

[0067] In the formula, T i is the vertical coordinate dimension that is the same as the horizontal coordinate after transformation; ε i is the cumulative deformation within time t i ; v i is the average rate within time t i ; m is the number of deformation monitoring times during the constant velocity deformation stage.

[0068] Through non-linear fitting with a regression model, an improved tangent angle formula is obtained:

[0069]

[0070] The model prediction R 2 reaches 0.91, significantly better than the traditional empirical formula. In view of this, adopting the modified lower limit value of the tangent angle not only improves the reliability of landslide early warning, but also enables earlier detection and warning of landslides, which helps in disaster prevention and mitigation of landslides.

[0071] A specific application case is provided below.

[0072] The Majiagou landslide is a typical creeping landslide in the Three Gorges Reservoir area. Under the influence of factors such as intermittent rainfall, groundwater fluctuations, and reservoir water fluctuations, local soil first undergoes creep deformation, resulting in continuous redistribution of internal stress in the rock and soil mass, and finally time-dependent damage occurs.

[0073] Since the first impoundment of the Three Gorges Reservoir area to 135 m in 2003, the Majiagou landslide has started to deform, and has been in a creep deformation state ever since. According to the occurrence conditions and hydrodynamic characteristics of groundwater, it can be known that the rock mass fissures in the landslide area are relatively developed and the water-richness is relatively strong. In order to eliminate the influence of the later anti-slide piles on the overall deformation of the Majiagou landslide, the surface displacement monitoring deformation data from February 2007 to December 2009 were analyzed, and it was found that the Majiagou landslide is in the stage of constant velocity deformation. The surface displacement monitoring points are arranged from the lower part to the upper part of the landslide as J01, J02, J03, J04 and J05 respectively, and the elevations of the monitoring points are 203.1 m, 222.6 m, 247.7 m, 276.3 m and 283.1 m respectively. The cumulative displacement amounts of the monitoring points in the lower part of the landslide are 490, 477 and 470 mm respectively, and the constant velocity deformation rates at this time are 0.46 mm / d, 0.45 mm / d and 0.44 mm / d respectively: the cumulative displacement amounts in the upper part of the landslide are 242 mm and 160 mm respectively, and the constant velocity deformation rates at this time are 0.23 mm / d and 0.15 mm / d respectively.

[0074] Introduce the time-dependent damage model for calculation. The calculation results are as Figure 2 shown in Table 1. The velocity of the J03 monitoring point in the stage of constant velocity deformation is 0.44 mm / d, which is consistent with the actual monitoring velocity of J03, indicating that the landslide is in the stage of constant velocity deformation.

[0075] Table 1 Calculation results of the J03 monitoring point of the Majiagou landslide

[0076]

[0077] Table 1 gives the calculation results of the J03 monitoring point during the damage inversion of the Majiagou landslide. It can be obtained that 0 - 6.05 years is the initial deformation stage of the landslide, and the initial deformation amount is 1908.46 mm. 6.05 - 33.15 is the stage of constant velocity deformation, and the deformation amount is 6068.86 mm, and the constant velocity deformation rate is 0.44 mm / d. Comparing the monitoring values in the stage of constant velocity deformation with the monitoring values from February 2007 to December 2009, it is found through comparison that the displacement amount of this monitoring point is consistent with the time displacement in the stage of constant velocity deformation of the landslide, and the time-dependent damage model can well reflect the creep deformation of the Majiagou landslide. 33.15 - 35.42 years is the stage of accelerated deformation, and the deformation amount reaches 7361.24 mm. 35.42 years is the stage of impending slide warning. According to the improved impending slide tangent angle criterion, take the reciprocal of the rate 1 / ε = 2.27 (mm / d) in the stage of constant velocity deformation of the landslide -1 , when 1 / ε ≥ 1.5, the impending slide tangent angle needs to satisfy the following formula Finally, it is concluded that the critical sliding tangent angle needs to satisfy α≥79.66°. After conversion, the deformation rate ε≥2.41mm / d in the critical sliding stage is obtained. According to the damage inversion results, when the deformation rate ε≥2.41mm / d, the critical sliding tangent angle α≥79.66°. When the landslide deforms for 35.42 years, the landslide will enter the critical sliding warning stage.

[0078] By introducing the time-dependent damage mechanism and the comprehensive warning model and applying them to typical scenarios such as the progressive creep landslide in Majiagou, the whole-process dynamic monitoring and warning from damage accumulation to critical failure are realized. In the case of the Majiagou landslide, the evolution characteristics of the landslide from the constant-speed deformation stage to the accelerated deformation stage are described through the time-dependent damage model. In addition, the model effectively solves the problem of insufficient adaptability of the traditional tangent angle method in complex geological environments by improving the calculation method of the tangent angle threshold, providing a scientific basis and technical support for the disaster prevention and control of similar progressive creep landslides.

[0079] What is described in the above specification is only the specific implementation manners of the present invention. Various examples do not constitute a limitation to the essence of the present invention. Those of ordinary skill in the art can make modifications or variations to the previously described specific implementation manners after reading the specification without departing from the essence and scope of the invention.

Claims

1. A landslide intelligent early warning method based on the aging damage mechanism, characterized in that: including the following steps S1. Establish a landslide numerical model, obtain relevant landslide rock and soil mechanics parameters through experimental data and investigation materials, and assign values; S2. Set the initial conditions and boundary conditions for numerical calculation, and then perform elastic calculation to calculate and generate the initial in-situ stress model of the landslide; S3. Invoke the dynamic constitutive model considering the time-dependent damage of rock and soil masses, and perform numerical calculation through the creep command; S4. Embed the improved tangent angle threshold algorithm and automatically calculate in combination with real-time monitoring data; S5. The system collects monitoring data in real time and automatically completes the quantification of damage accumulation based on the time-dependent damage dynamic constitutive model; S6. Embed the time-dependent damage early warning model recognition system into the computer system; S7. Calculate and recognize the time-dependent damage early warning model and output the early warning result.

2. The landslide intelligent early warning method based on the aging damage mechanism according to claim 1, characterized in that: In the step S3, the construction process of the dynamic constitutive model includes the following steps: S31. Express the deterioration process of the rock and soil strength parameters through the damage function D(t), that is Wherein, c0 is the initial cohesive force, and c (t) is the cohesive force at time t, and c f is the residual cohesive force; S32. Dynamically adjust the damage function D(t) according to the FLAC3D numerical calculation. The formula is: Among them, is the damage increment, Δσ(t) is the stress change of the damage model, and σ 0 is the initial stress; S33. Based on the Burgers model, construct a constitutive model considering time-dependent damage, and combine the damage function D(t) with the Burgers model. The formula is: In the formula, is the initial stress tensor, and D(t) is the damage function at time t; S34. Solve the stress-strain relationship of the constitutive model considering time-dependent damage by the finite difference method. The formula is Among them, is the damage increment, Δσ(t) is the stress change of the damage model, and σ 0 is the initial stress.

3. The intelligent landslide early warning method based on the aging damage mechanism according to claim 1, characterized in that: In the step S4, the improved tangent angle threshold algorithm includes the following steps S41. Process the displacement time dimension change into a Ti-t curve to make the dimensions of the horizontal and vertical coordinates consistent; where T i is the ordinate dimension that is the same as the abscissa after transformation; ε i is the cumulative deformation within time t i ; v i is the average rate within time t i ; m is the number of deformation monitoring times during the constant-rate deformation stage S42. Perform non-linear fitting through the regression model to obtain the improved tangent angle formula:

Images