Multi-scale loess foundation stability prediction method, system, storage medium and device

Abstract

The present invention discloses a multi-scale loess foundation stability prediction method, system, storage medium and equipment, which belongs to the field of geophysics and hydrogeology technology. The method mainly includes: using high-density electrical method to collect resistivity data of loess foundation in the monitoring area; inverting resistivity data in different time periods to obtain different inversion profiles; gridding the resistivity data of loess foundation in the monitoring area, comparing the difference in resistivity distribution between different inversion profiles, and screening out high-risk areas; performing kernel density estimation and distribution difference analysis based on the screened high-risk areas to obtain the trend of low resistivity data changes, and combining multi-scale breakpoint detection technology to predict collapse risk based on the trend of low resistivity data changes. The present invention is based on high-density electrical method and combines multiple data analysis technologies to achieve efficient, economical and sustainable monitoring and prediction of loess foundation stability.

Description

Technical Field

[0001] The present invention relates to the field of geophysics and hydrogeology technology, and in particular to a multi-scale loess foundation stability prediction method, system, storage medium and equipment. Background Art

[0002] Loess soil is widely distributed in northern China and has unique collapsible properties. Under certain external loads, loess suddenly produces large irreversible settlement due to water absorption and softening. Loess layers often cause ground collapse due to changes in moisture, posing a threat to the safety of buildings and infrastructure. Although traditional monitoring methods such as ground penetrating radar and settlement observation instruments can achieve preliminary monitoring of the collapse of loess layers, there is a time lag in data collection and analysis, which makes it difficult to fully reflect the dynamic changes in the collapsible area and cannot fully adapt to the sudden changes in the collapsible properties of loess. Therefore, it is particularly necessary to develop an efficient, economical and sustainable loess foundation stability monitoring and prediction method that combines multiple data analysis technologies. Summary of the invention

[0003] The purpose of the present invention is to overcome the problems existing in the prior art and to provide a multi-scale loess foundation stability prediction method, system, storage medium and device.

[0004] The objective of the present invention is achieved through the following technical solutions:

[0005] In a first aspect, a multi-scale loess foundation stability prediction method is provided, comprising the following steps:

[0006] Data collection: Use high-density electrical method to collect resistivity data of loess foundation in the monitoring area;

[0007] Data inversion: Invert the resistivity data in different time periods to obtain different inversion profiles;

[0008] Risk area screening: grid the resistivity data of the loess foundation in the monitoring area, compare the differences in resistivity distribution between different inversion profiles, and screen out high-risk areas;

[0009] Stability prediction: Kernel density estimation and distribution difference analysis are performed based on the screened high-risk areas to obtain the trend of low resistivity data changes, and the collapse risk is predicted based on the trend of low resistivity data changes.

[0010] In some embodiments, the high-density electrophysiology method uses a Wenner device or a parallel dipole device.

[0011] In some embodiments, comparing the differences in resistivity distribution between different inversion profiles includes:

[0012] Compute the normalized difference in resistivity distribution between two inversion profiles: ,in, D norm It represents the normalized difference value of the resistivity distribution between the two inversion profiles. R1 is the resistivity value of the inversion profile P1 in the corresponding grid, and R2 is the resistivity value of the inversion profile P2 in the corresponding grid.

[0013] In some embodiments, comparing the differences in resistivity distribution between different inversion profiles further includes:

[0014] On grids of different scale resolutions, the resistivity differences between different inversion sections are calculated and visualized.

[0015] In some embodiments, screening out high-risk areas includes:

[0016] set up D norm When the threshold is exceeded, the existing area is marked as a high-risk area and the results are visualized.

[0017] In some embodiments, performing kernel density estimation and distribution difference analysis based on the screened high-risk areas includes:

[0018] The two-dimensional resistivity histograms of the corresponding depths of different inversion profiles are extracted respectively, and the resistivity distribution in each grid is counted. The difference is quantified by calculating the KL divergence, the distribution overlap is calculated, and the similarity of the resistivity distribution between different inversion profiles is evaluated; the KL divergence is defined as:

[0019] , in the continuous case is defined as: ,in, is the resistivity probability density function / distribution of the inverted profile P1, is the resistivity probability density function / distribution of the inverted profile P2. The closer the KL divergence is to 0, the closer the two distributions are. The larger the KL divergence is, the greater the difference between the two distributions is.

[0020] The distribution overlap is defined as: , if the distribution overlap is close to 1, the distribution difference is small; if the overlap is low, the distribution difference is obvious.

[0021] In some embodiments, the method further comprises introducing a multi-scale breakpoint detection technique into the resistivity data to locate the specific location of the resistivity data change.

[0022] In a second aspect, a multi-scale loess foundation stability prediction system is provided, comprising:

[0023] A data acquisition module configured to acquire resistivity data of loess foundation in the monitoring area using a high-density electrical method;

[0024] A data inversion module configured to invert resistivity data in different time periods to obtain different inversion profiles;

[0025] The risk area screening module is configured to grid the resistivity data of the loess foundation in the monitoring area, compare the differences in resistivity distribution between different inversion profiles, and screen out high-risk areas;

[0026] The stability prediction module is configured to perform kernel density estimation and distribution difference analysis based on the screened high-risk areas to obtain the trend of low resistivity data changes and predict the collapse risk based on the trend of low resistivity data changes.

[0027] In a third aspect, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the multi-scale loess foundation stability prediction method as described in the first aspect is implemented.

[0028] In a fourth aspect, an electronic device is provided, comprising a memory and a processor, wherein the memory stores computer instructions executable on the processor, and when the processor executes the computer instructions, the multi-scale loess foundation stability prediction method as described in the first aspect is implemented.

[0029] It should be further explained that the technical features corresponding to the above aspects can be combined or replaced with each other to form a new technical solution without conflict.

[0030] Compared with the prior art, the present invention has the following beneficial effects:

[0031] 1. The present invention collects the resistivity data of the loess foundation in the monitoring area by high-density electrical method to ensure the real-time and accuracy of data collection; inverts the resistivity data in different time periods to obtain different inversion profiles, performs grid processing on the resistivity data of the loess foundation in the monitoring area, compares the difference in resistivity distribution between different inversion profiles, and screens out high-risk areas, and finally performs kernel density estimation and distribution difference analysis based on the screened high-risk areas to obtain the trend of low resistivity data changes, and predicts the collapse risk based on the trend of low resistivity data changes. It can fully reflect the dynamic changes of the collapsible area, improve the recognition accuracy of the collapse risk, and better adapt to the sudden collapsible changes of loess.

[0032] 2. The present invention introduces multi-scale breakpoint detection technology in resistivity time series data and spatial data to identify change points in time series or spatial series, and detects small and medium changes in the series by accumulating the deviations between data points and the mean, which is suitable for weak and sudden problems in collapse tasks, and further improves the recognition accuracy of potential collapse risks. At the same time, the specific location of data changes can be effectively located to determine the collapse prediction area.

[0033] 3. During the continuous observation process, the present invention analyzes and corrects the errors caused by the time drift of the instrument, changes in geological conditions, and different observation scales to ensure data accuracy.

[0034] 4. The present invention calculates and visualizes the resistivity differences between different inversion profiles on grids of different scale resolutions, which helps to compare the differences between coarse and fine grids. At the same time, the change trends of different scales are observed to help determine which areas have undergone significant changes at different resolutions. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 A flow chart of a multi-scale loess foundation stability prediction method shown in an embodiment of the present invention;

[0036] Figure 2 The resistivity data diagram corresponding to the inversion profile P1 shown in the embodiment of the present invention;

[0037] Figure 3 The resistivity data diagram corresponding to the inversion profile P2 shown in the embodiment of the present invention;

[0038] Figure 4 The resistivity values ​​corresponding to the various color regions in the inversion profile shown in the embodiment of the present invention;

[0039] Figure 5 A heat map showing areas with significant differences according to an embodiment of the present invention;

[0040] Figure 6 Comparison of resistivity distribution of two profiles in the depth range of -6 to -12 meters for an embodiment of the present invention;

[0041] Figure 7 Comparison of resistivity distribution of two profiles in the depth range of -12 to -18 meters for an embodiment of the present invention;

[0042] Figure 8 This is a schematic diagram of breakpoint analysis according to an embodiment of the present invention. DETAILED DESCRIPTION

[0043] The technical solution of the present invention is clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. The components of the embodiments of the present application described and shown in the accompanying drawings can be arranged and designed in various configurations. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0044] It should be noted that the defects existing in the solutions in the above-mentioned prior art are the results obtained by the inventor after practice and careful research. Therefore, the discovery process of the above-mentioned problems and the solutions proposed in the embodiments of the present application for the above-mentioned problems below should all be the contributions made by the inventor to the present application in the process of invention and creation, and should not be understood as technical contents known to technical personnel in this field.

[0045] In view of the technical problems pointed out in the background technology, the embodiments provided by the present invention are as follows:

[0046] In an exemplary embodiment, a multi-scale loess foundation stability prediction method is provided, such as Figure 1 As shown, the following steps are included:

[0047] Data collection: Use high-density electrical method to collect resistivity data of loess foundation in the monitoring area;

[0048] Data inversion: Invert the resistivity data in different time periods to obtain different inversion profiles; the different time periods can be a combination of multiple continuous monitoring time periods, or a monitoring time period / cycle can be divided into multiple time periods as required, or the corresponding monitoring time periods can be set according to the monitoring requirements for before and after comparison;

[0049] Risk area screening: grid the resistivity data of the loess foundation in the monitoring area, compare the differences in resistivity distribution between different inversion profiles, and screen out high-risk areas;

[0050] Stability prediction: Kernel density estimation and distribution difference analysis are performed based on the screened high-risk areas to obtain the trend of low resistivity data changes, and the collapse risk is predicted based on the trend of low resistivity data changes.

[0051] This method also requires corresponding preparation before prediction, including:

[0052] Collection of geological data: Utilize regional geological maps and existing geological data to clarify the distribution range of collapsible loess, identify possible weak zones (such as water-rich areas, fracture zones, etc.) within the survey area, and clarify the monitoring scope.

[0053] Equipment preparation and setup: Prepare and check high-density electrical instruments, electrodes, wires and other equipment to ensure that all equipment is in good condition and that the electrodes are in good contact. Set instrument parameters, including current intensity, electrode spacing, etc., to suit specific geological conditions and detection depth.

[0054] Determination of acquisition method: There are many electrode layout methods for high-density electrical method, such as the Wenner Array. The electrodes of the Wenner Array are arranged at equal intervals, the two in the middle are current electrodes (A, B), and the two on the outside are potential electrodes (M, N). This device is sensitive to changes in resistivity and has a high vertical resolution. It is suitable for layered structure detection and distribution detection of loess collapsible layers. If there is a large-scale monitoring setting, a parallel dipole device can be used according to the actual situation, and the acquisition method can be confirmed according to the actual work area situation.

[0055] Furthermore, the density of measuring points is adjusted according to the monitoring target, and the electrodes are evenly arranged on the measuring line. The electrode spacing is adjusted according to the detection target depth and accuracy requirements.

[0056] In addition, according to the needs of the target monitoring area, it is divided into short-term observation and long-term monitoring. Among them, short-term observation is to conduct intensive continuous high-density electrical observation in the determined low-resistivity area, collect and process data once every hour or as needed to monitor the dynamic changes of resistivity over time. Long-term monitoring is to implement a continuous monitoring plan and analyze long-term series data to facilitate early identification of changes in foundation stability and ensure that repeated observations are performed before testing to reduce the impact of data errors. The probability formula for measuring whether there is a collapse should be adjusted according to different monitoring targets.

[0057] Before observation, it is necessary to confirm the XY coordinate range of the target monitoring area, among which the horizontal coordinate X range of the profile measurement is preliminarily determined, usually based on the length along the profile, such as the range from the minimum X coordinate to the maximum X coordinate. Determine the vertical depth Y range of the profile, for example, from the surface (0 meters) to the maximum detection depth. The electrode spacing is the distance between each electrode, and the electrode distribution density is confirmed according to the target.

[0058] After acquiring the collected resistivity data, the resistivity data is preliminarily cleaned, and the resistivity data within a reasonable range is screened, such as setting the resistivity value between 0 and 5000Ωm, eliminating abnormal values ​​and jump points (such as extremely large or extremely small noise data), and performing appropriate smoothing according to the data situation. Log conversion is performed on the data to enhance the difference comparison effect and reduce the impact of extreme values ​​on the analysis. The formula is as follows: ,in, R is the original resistivity value.

[0059] The high-density resistivity method can be inverted according to the algorithms available, such as the least squares method, quasi-Newton method, Gauss-Newton method, etc., which can be used to invert the collected data by relying on various inversion methods or using professional resistivity inversion software. The inversion results of the high-density resistivity method are debugged according to the inversion effect to generate underground resistivity profiles. However, it is necessary to ensure that the inversion methods and inversion parameters of each profile are consistent to reduce the multiple solutions caused by inconsistent inversion method parameters.

[0060] During the acquisition, it is necessary to perform timed acquisition and data imaging according to the needs of the target monitoring area and the actual task requirements. Therefore, the collapse detection methods analyze and predict the inversion results of each time profile. The collapse task is mainly due to the change in water content caused by wetting, and the resistivity attribute is extremely sensitive to water content. Therefore, the low-resistance data change area is mainly detected to predict and detect the collapse area. The following takes two time periods as an example for collapse detection and stability prediction, as follows:

[0061] Use high-density electrical equipment to collect underground resistivity data at the first time T1 and obtain the inversion profile P1, such as Figure 2 As shown; use high-density electrical equipment to collect underground resistivity data at the second time T2 to obtain the inversion profile P2, as shown Figure 3 As shown, Figure 2 , Figure 3 The resistivity values ​​corresponding to the color areas are as follows: Figure 4 As shown. Since the detection depth of the high-density resistivity method is determined by the electrode device and the image is an inverted triangular trapezoid, the result of the differential thermal map analysis is the full-depth result, and the inversion result profile is only performed on the target depth. Therefore, the target depth data can be intercepted for easy display, so that it can be mapped into a trapezoid. First, perform a differential thermal map analysis on the data to obtain a large range of data difference areas. The steps to be completed are as follows:

[0062] Data gridding: According to the X and Y coordinate ranges, the study area is divided into fixed-size grids. The size of each grid is determined by the detection target and data resolution (such as 10 partitions in the X direction and 5 partitions in the Y direction).

[0063] Normalized Difference Calculation: Compare the normalized difference in resistivity distribution between two inversion profiles: , where D norm It represents the normalized difference value of the resistivity distribution between the two inversion profiles. R1 is the resistivity value of the inversion profile P1 in the corresponding grid, and R2 is the resistivity value of the inversion profile P2 in the corresponding grid.

[0064] Screening of significant change areas: Setting thresholds (such as D norm The absolute value of is greater than 0.2) is marked as a high-risk area. And the results are visualized: the heat map is used to show the areas with significant differences, the possible collapse depth location yi is identified, and the color scale is used to show the areas with obvious differences, such as Figure 5 shown.

[0065] For the selected high-risk areas, kernel density estimation and distribution difference analysis are performed. The two-dimensional resistivity histograms of the corresponding depths of different inversion profiles are extracted respectively. The resistivity distribution in each grid is statistically analyzed. The difference is quantified by calculating the KL divergence, the distribution overlap is calculated, and the similarity of the resistivity distribution between different inversion profiles is evaluated. The KL divergence is defined as: , in the continuous case is defined as: ,in, is the resistivity probability density function / distribution of the inverted profile P1, is the resistivity probability density function / distribution of the inverted profile P2. The closer the KL divergence is to 0, the closer the two distributions are. The larger the KL divergence is, the greater the difference between the two distributions is. For example, if the KL divergence is equal to 0, the two distributions are almost the same. When the KL divergence is greater than a certain threshold, it indicates that there may be significant differences between the two distributions. The threshold of the KL divergence can be determined according to the actual situation.

[0066] The distribution overlap is defined as: , if the distribution overlap is close to 1, the distribution difference is small; if the overlap is low, the distribution difference is obvious.

[0067] like Figure 6 As shown in the figure, in the depth range of -6 to -12 meters, the KL divergence is 0.0877, indicating that there is a certain difference in the resistivity distribution of the two profiles in this depth range. Figure 7 As shown, in the depth range of -12 to -18 m, the KL divergence is 0.0438, indicating that at this deeper level, the resistivity distribution difference between the two profiles is smaller than that in the shallower layer, which may be caused by the small change in moisture in the deeper layer.

[0068] The purpose of this step is to test whether there is a significant increase in low resistivity data, because a significant increase in low resistivity data may indicate an increase in underground water content. By obtaining clear resistivity trend change characteristics for data in a specific depth range, the data distribution is discovered. The stability of the loess foundation and the risk of collapse are predicted based on the resistivity change trend.

[0069] In the screening of high-risk areas, the coarse grid of data is screened to determine the high-risk areas. The KL divergence calculation can obtain the amplitude of the resistivity distribution change in the high-risk area, but this parameter can only locate whether the degree of resistivity change is significant, and cannot determine and locate the specific monitoring range. Therefore, it is necessary to introduce multi-scale breakpoint detection technology to locate the specific depth coordinates of the data to improve the accuracy of identifying potential collapse risks, so as to guide the subsequent excavation and collapse treatment. Multi-scale breakpoint detection technology is used to identify change points in time series or spatial series, and detect small and medium changes in the series by accumulating the deviation of data points from the mean, which is suitable for weak and sudden problems in collapse tasks.

[0070] The multi-scale breakpoint calculation formula and process are as follows:

[0071] Initialization: Set the initial CUSUM (cumulative error) value to 0. By default, no cumulative deviation is accumulated initially.

[0072] Iteration: For each data point, calculate its deviation and accumulate these deviations to update the positive and negative CUSUM values. If the deviation is not large enough to make the CUSUM value zero, the corresponding CUSUM value will be reset to 0, which helps ignore small, meaningless fluctuations. Specifically, calculate the mean (μ):

[0073]

[0074] Calculate the standard deviation (σ):

[0075]

[0076] Iteratively calculate CUSUM: For each point in the data series , perform the following steps:

[0077] Calculate Deviation :

[0078]

[0079] Update Positive CUSUM :

[0080]

[0081] Here, k is a small positive number called the “reference value” or “drift” that is used to adjust the sensitivity.

[0082] Update Negative CUSUM :

[0083]

[0084] Change point determination: By setting a threshold, it is determined when the cumulative deviation of the CUSUM is large enough to be considered a substantial change. If it exceeds the set threshold (e.g. 5σ), it is considered that there is a positive change point at position i.

[0085] if If the value exceeds the set threshold (e.g. -5σ), it is considered that there is a negative change point at position i.

[0086] This method can effectively locate the specific location of data changes, determine the lateral range xi, and obtain the final collapse prediction area. The results of breakpoint analysis using multi-scale breakpoint detection technology are as follows: Figure 8 shown.

[0087] Furthermore, in order to ensure data accuracy during continuous observation, it is necessary to analyze and correct possible errors. Common sources of error include instrument time drift, changes in geological conditions, and different observation scales. :

[0088]

[0089] in, Calibrate the instrument to a percentage error (eg 5% or 0.05).

[0090] Environmental Error :

[0091]

[0092] in, Indicates the proportion of measurement error caused by geological conditions, humidity, temperature, etc.

[0093] Total error :

[0094]

[0095] Finally, the collapse prediction probability is calculated:

[0096]

[0097] is the gradient value of grid i, indicating the drastic change of resistivity in the XY plane: , w1 and w2 are weight values; is the average resistivity of the inversion profile P1, is the average resistivity of the inversion profile P2. In actual monitoring, if long-term collapse observation is set up, considering the environmental error and instrument error, it is necessary to measure the Pcollapse Mean, if the currently calculated P collapse The value is greater than P in the total observation time collapse When the mean is 0.04, it indicates that the current change is not within a reasonable range, and it is judged that there is a probability of collapse.

[0098] It should be noted that the previous step of marking high-risk areas by setting a threshold of normalized difference is to confirm the resistivity range under the coarse grid. This method is relatively rough and cannot determine the quantitative probability of collapse. The collapse prediction possibility calculation takes into account external factors such as the intensity and time of resistivity change and the error of the instrument itself to comprehensively and quantitatively judge the probability of collapse. For example, if the collapse possibility threshold is set, >3 is marked as a high-risk area, indicating that the probability of collapse is high, otherwise it is marked as a low-risk area, indicating that the probability of collapse is low. The specific threshold is set according to the actual situation (for example, according to the P collapse The mean value is set).

[0099] The present invention also calculates and visualizes the resistivity differences between different inversion profiles on grids of different scale resolutions. Specifically, the steps include:

[0100] Coarse grid difference calculation: Use a smaller grid resolution for preliminary calculations. Generally, a coarser grid will contain more regional averages and be able to capture large-scale variation trends.

[0101] Gradually refine the mesh: As the mesh resolution increases, the differences in the refined area are further calculated. The refined mesh can reveal more local details.

[0102] Compute the difference at each scale resolution: Compute and visualize the difference in resistivity at different scale resolutions. This helps compare the difference between coarse and fine grids.

[0103] Analyze resistivity differences at different scales: Observe the changing trends at different scales to help determine which areas have undergone significant changes at different resolutions.

[0104] Results Summary: Graphically display the resistivity differences at each scale resolution and compare the results for coarse and fine grids to help identify areas where differences are most significant in the multiscale analysis.

[0105] In another exemplary embodiment, based on the same inventive concept as the method, a multi-scale loess foundation stability prediction system is provided, comprising:

[0106] A data acquisition module configured to acquire resistivity data of loess foundation in the monitoring area using a high-density electrical method;

[0107] A data inversion module configured to invert resistivity data in different time periods to obtain different inversion profiles;

[0108] The risk area screening module is configured to grid the resistivity data of the loess foundation in the monitoring area, compare the differences in resistivity distribution between different inversion profiles, and screen out high-risk areas;

[0109] The stability prediction module is configured to perform kernel density estimation and distribution difference analysis based on the screened high-risk areas to obtain the trend of low resistivity data changes and predict the collapse risk based on the trend of low resistivity data changes.

[0110] In another exemplary embodiment, based on the same inventive concept as the method, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the multi-scale loess foundation stability prediction method provided by the embodiment of the present invention is implemented. Based on such an understanding, the technical solution of this embodiment is essentially or partly contributed to the prior art or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0111] In another exemplary embodiment, based on the same inventive concept as the method, an electronic device is provided, including a memory and a processor, wherein the memory stores computer instructions executable on the processor, and when the processor executes the computer instructions, the multi-scale loess foundation stability prediction method provided in an embodiment of the present invention is executed.

[0112] The processor may be a single-core or multi-core central processing unit or a specific integrated circuit, or one or more integrated circuits configured to implement the present invention.

[0113] Embodiments of the subject matter and functional operations described in this specification may be implemented in: tangibly embodied computer software or firmware, computer hardware including the structures disclosed in this specification and their structural equivalents, or a combination of one or more of them. Embodiments of the subject matter described in this specification may be implemented as one or more computer programs, i.e., one or more modules of computer program instructions encoded on a tangible, non-transitory program carrier to be executed by a data processing device or to control the operation of the data processing device. Alternatively or additionally, the program instructions may be encoded on an artificially generated propagated signal, such as a machine-generated electrical, optical, or electromagnetic signal, which is generated to encode and transmit information to a suitable receiver device for execution by the data processing device.

[0114] The processes and logic flows described in this specification can be performed by one or more programmable computers executing one or more computer programs to perform corresponding functions by operating on input data and generating output. The processes and logic flows can also be performed by special purpose logic circuits, such as FPGAs (field programmable gate arrays) or ASICs (application-specific integrated circuits), and the apparatus can also be implemented as special purpose logic circuits.

[0115] Processors suitable for executing computer programs include, for example, general and / or special microprocessors, or any other type of central processing unit. Typically, the central processing unit will receive instructions and data from a read-only memory and / or a random access memory. The basic components of a computer include a central processing unit for implementing or executing instructions and one or more memory devices for storing instructions and data. Typically, the computer will also include one or more large-capacity storage devices for storing data, such as magnetic disks, magneto-optical disks, or optical disks, or the computer will be operably coupled to this large-capacity storage device to receive data from it or to transmit data to it, or both. However, the computer does not necessarily have such a device. In addition, the computer can be embedded in another device, such as a mobile phone, a personal digital assistant (PDA), a mobile audio or video player, a game console, a global positioning system (GPS) receiver, or a portable storage device such as a universal serial bus (USB) flash drive, just to name a few.

[0116] It should be understood that each box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, a program segment or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a sequence different from that marked in the accompanying drawings. For example, two continuous boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flow chart, and the combination of the boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs the specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0117] The above specific implementation methods are detailed descriptions of the present invention. It cannot be determined that the specific implementation methods of the present invention are limited to these descriptions. For ordinary technicians in the technical field to which the present invention belongs, several simple deductions and substitutions can be made without departing from the concept of the present invention, which should be regarded as belonging to the protection scope of the present invention.

Claims

1. A multi-scale loess foundation stability prediction method, characterized in that: The method includes the following steps: data acquisition: using high-density electrical method to acquire resistivity data of loess foundation in the monitoring area; data inversion: inverting resistivity data in different time periods to obtain different inversion profiles; Risk area screening: grid the resistivity data of the loess foundation in the monitoring area, compare the differences in resistivity distribution between different inversion profiles, and screen out high-risk areas; The comparing the differences in resistivity distribution between different inversion profiles also includes: On grids of different scale resolutions, the resistivity differences between different inversion profiles are calculated and visualized. The specific steps include the following: Coarse grid difference calculation: Use a smaller grid resolution for preliminary calculations. The coarse grid will contain more regional averages and can capture large-scale variation trends. Gradually refine the grid: As the grid resolution increases, the differences in the refined area are further calculated, and the refined grid can reveal more local detail differences; Calculate the difference at each scale resolution: Calculate and visualize the difference in resistivity on grids of different scale resolutions, and compare the difference on coarse and fine grids; Analyze resistivity differences at different scales: Observe the trend of changes at different scales to help determine which areas have undergone significant changes at different resolutions; Summarize results: Display resistivity differences at each scale through graphs and compare the results of coarse and fine grids to help identify which areas have the most significant differences in multi-scale analysis; Stability prediction: Kernel density estimation and distribution difference analysis are performed based on the selected high-risk areas to obtain the trend of low resistivity data changes, and the collapse risk is predicted based on the trend of low resistivity data changes; The predicted collapse risk includes: The probability of collapse is calculated by the following formula: Among them, P collapse represents the probability of collapse, G i is the gradient value of grid i, indicating the drastic change of resistivity in the XY plane: w1 and w2 are weight values; R 1，i is the average resistivity of the inversion profile P1, R 2，i is the average resistivity of the inversion profile P2, δ instrument Indicates the instrument calibration error ratio, δ environment Indicates the proportion of measurement errors caused by geological conditions, humidity, and temperature; Predict the collapse risk based on the probability of collapse.

2. The multi-scale loess foundation stability prediction method according to claim 1, characterized in that: The high-density electrical method adopts a Wenner device or a parallel dipole device.

3. The multi-scale loess foundation stability prediction method according to claim 1, characterized in that: The comparison of resistivity distribution differences between different inversion profiles includes: Compute the normalized difference in resistivity distribution between two inversion profiles: Among them, D norm It represents the normalized difference value of the resistivity distribution between the two inversion profiles. R1 is the resistivity value of the inversion profile P1 in the corresponding grid, and R2 is the resistivity value of the inversion profile P2 in the corresponding grid.

4. The multi-scale loess foundation stability prediction method according to claim 3, characterized in that: The high-risk areas screened include: Setting D norm When the threshold is exceeded, the existing area is marked as a high-risk area and the results are visualized.

5. The multi-scale loess foundation stability prediction method according to claim 3, characterized in that: The kernel density estimation and distribution difference analysis are performed based on the screened high-risk areas, including: The two-dimensional resistivity histograms of the corresponding depths of different inversion profiles are extracted respectively, and the resistivity distribution in each grid is statistically analyzed to perform kernel density estimation. The similarity of resistivity distribution between different inversion profiles is evaluated by calculating the KL divergence to quantify the difference or calculate the distribution overlap. The KL divergence is defined as: In the continuous case it is defined as: Where P(x) is the resistivity probability density function / distribution of the inversion profile P1, Q(x) is the resistivity probability density function / distribution of the inversion profile P2, the closer the KL divergence is to 0, the closer the two distributions are, and the larger the KL divergence is, the greater the difference between the two distributions; The distribution overlap is defined as: Xf = ∫min(P(x), Q(x))dx. If the distribution overlap is close to 1, the distribution difference is small; if the overlap is low, the distribution difference is obvious.

6. The multi-scale loess foundation stability prediction method according to claim 1, characterized in that: The method also includes introducing a multi-scale breakpoint detection technique into the resistivity data to locate the specific location of the resistivity data change.

7. A multi-scale loess foundation stability prediction system, characterized in that: include: A data acquisition module configured to acquire resistivity data of loess foundation in the monitoring area using a high-density electrical method; A data inversion module configured to invert resistivity data in different time periods to obtain different inversion profiles; The risk area screening module is configured to grid the resistivity data of the loess foundation in the monitoring area, compare the differences in resistivity distribution between different inversion profiles, and screen out high-risk areas; The comparing the differences in resistivity distribution between different inversion profiles also includes: calculating and visualizing the differences in resistivity between different inversion profiles on grids of different scale resolutions; specifically including the following steps: Coarse grid difference calculation: Use a smaller grid resolution for preliminary calculations. The coarse grid will contain more regional averages and can capture large-scale variation trends. Gradually refine the grid: As the grid resolution increases, the differences in the refined area are further calculated, and the refined grid can reveal more local detail differences; Calculate the difference at each scale resolution: Calculate and visualize the difference in resistivity on grids of different scale resolutions, and compare the difference on coarse and fine grids; Analyze resistivity differences at different scales: Observe the trend of changes at different scales to help determine which areas have undergone significant changes at different resolutions; Results Summary: Graphically display the resistivity differences at each scale resolution and compare the results for the coarse and fine grids to help identify areas where differences are most significant in the multiscale analysis; The stability prediction module is configured to perform kernel density estimation and distribution difference analysis based on the screened high-risk areas to obtain the trend of low resistivity data changes, and predict the collapse risk based on the trend of low resistivity data changes; The predicted collapse risk includes: The probability of collapse is calculated by the following formula: Among them, P collapse represents the probability of collapse, G i is the gradient value of grid i, indicating the drastic change of resistivity in the XY plane: w1 and w2 are weight values; R 1，i is the average resistivity of the inversion profile P1, R 2，i is the average resistivity of the inversion profile P2, δ instrument Indicates the instrument calibration error ratio, δ environment Indicates the proportion of measurement errors caused by geological conditions, humidity, and temperature; Predict the collapse risk based on the probability of collapse.

8. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the multi-scale loess foundation stability prediction method described in any one of claims 1 to 6 is implemented.

9. An electronic device comprising a memory and a processor, wherein the memory stores computer instructions that can be executed on the processor, wherein: When the processor runs the computer instructions, it executes the multi-scale loess foundation stability prediction method described in any one of claims 1-6.

Images