A multi-sensor-based real-time monitoring system for pipeline construction foundation pit landslide

Abstract

The application relates to the technical field of pipeline construction safety monitoring, and particularly discloses a pipeline construction foundation pit landslide real-time monitoring system based on multiple sensors, which collects displacement-depth corresponding data groups; constructs a displacement change curve and solves the distribution of the curvature gradient along the depth according to the data groups; sets a sliding window along the depth direction, marks strain localization candidate areas with fluctuation amplitudes; selects a local extreme point with the maximum curvature gradient absolute value in each candidate area as a preliminary selected depth, extracts a displacement amount central deviation degree and a curve slope change symmetry degree as shape parameters after expanding a preset range, and weightedly calculates a confidence value; takes the preliminary selected depth with a confidence value exceeding a threshold value as a final sliding surface position; comprehensively judges a landslide risk grade according to the position depth proportion of the position depth in the excavation depth and a displacement change rate trend, and outputs; and the application effectively distinguishes overall drift of bottom soft soil and local shear strain.

Description

Technical Field

[0001] This invention relates to the field of pipeline construction safety monitoring technology, specifically to a real-time monitoring system for landslides in pipeline construction pits based on multiple sensors. Background Technology

[0002] During the excavation of deep and long foundation pits in pipeline engineering, landslides are highly likely to occur on the slopes due to stress release in the surrounding soil and disturbances from external factors, posing a threat to the safety of construction personnel and the quality of pipeline laying. To monitor the stability of the slope in real time, existing technologies typically employ the method of burying fixed array inclinometers within the foundation pit slope. By reading the angular deflection values ​​of each measuring point at various depths within the inclinometer relative to the direction of gravity, the horizontal displacement of each measuring point is calculated. Based on the assumption that the bottom of the inclinometer is an absolutely fixed point, the displacement of each measuring point relative to the bottom of the pipe is zeroed out, and a displacement change curve along the depth is plotted. The presence of a slip surface is then determined based on the curve shape.

[0003] When excavating pipeline foundation pits in deep soft soil strata, the bottom end of the fixed array inclinometer tube cannot penetrate the soft soil layer and anchor to the underlying bedrock. It is actually buried in the soft soil, which is still in a creeping state, and undergoes overall horizontal drift under the unloading and subsequent loading during foundation pit excavation. Existing monitoring methods assume the bottom of the inclinometer tube is an absolutely fixed point and force the tube bottom displacement to zero. This approach directly discards the actual physical displacement information at the tube bottom, causing the linear offset component of the displacement curve caused by the overall drift of the bottom soft soil to be misjudged as the nonlinear bending characteristics of the deep shear zone during inversion calculations. This leads to the rigid body translation being mistakenly identified as a through-slip surface, triggering false alarms. Furthermore, existing technology cannot effectively distinguish between the overall drift of the bottom soft soil and local shear strain. Summary of the Invention

[0004] The purpose of this invention is to provide a real-time monitoring system for landslides in pipeline construction pits based on multiple sensors, in order to solve the problems mentioned above.

[0005] The objective of this invention can be achieved through the following technical solutions: A real-time monitoring system for landslides in pipeline construction pits based on multiple sensors, comprising: The displacement-depth correspondence data generation module collects horizontal displacement data of each depth measuring point in the slope of the pipeline foundation pit, and synchronously records the depth position corresponding to each measuring point to form a displacement-depth correspondence data set. The curvature gradient distribution calculation module constructs a continuous displacement change curve along the depth direction based on the displacement-depth corresponding data set, and solves the curvature change rate of the displacement change curve at each depth to obtain the distribution result of the curvature gradient along the depth. The strain localization candidate region identification module sets a sliding window along the depth direction, counts the fluctuation amplitude of the curvature gradient within each sliding window, and marks the depth interval covered by the sliding window whose fluctuation amplitude exceeds the preset benchmark value as the strain localization candidate region. The potential slip surface confidence assessment module selects the location where the curvature gradient reaches its extreme value as the initial depth of the potential slip surface within each strain localization candidate region, and calculates the confidence value of the initial depth based on the inflection point morphology characteristics of the displacement change curve at the initial depth. The sliding surface determination output module determines the initial depth where the confidence value exceeds a preset threshold as the final sliding surface position, and outputs the depth value corresponding to the final sliding surface position and its displacement change rate trend, which is used to determine the landslide risk.

[0006] As a further aspect of the present invention: obtaining the distribution result of the curvature gradient along the depth specifically includes: Arrange the depth measurement points in the displacement-depth corresponding data group in order from shallow to deep, and calculate the first-order rate of change of displacement relative to depth between adjacent depth measurement points to obtain the first-order rate of change value of each depth interval. Arrange the first-order rate of change values ​​of each depth interval in depth order, and calculate the difference between the first-order rate of change values ​​of adjacent depth intervals. Use the difference as the initial value of the rate of change of curvature at the boundary of adjacent depth intervals, and summarize the initial values ​​of the rate of change of curvature at all boundary positions to form the preliminary distribution of curvature gradient along the depth. The initial distribution of curvature gradient along the depth is processed by removing low-amplitude oscillations. The oscillation intervals where the curvature gradient values ​​continuously cross positive and negative values ​​and the amplitude is lower than the preset amplitude threshold are smoothed and corrected. The unidirectional continuous change intervals where the amplitude exceeds the preset amplitude threshold are retained. The corrected curvature gradient values ​​are used as the final distribution result of curvature gradient along the depth.

[0007] As a further aspect of the present invention: obtaining the first-order rate of change values ​​for each depth interval specifically includes: Obtain the current depth measurement point and its adjacent previous and next depth measurement points, record the first depth distance between the current measurement point and the previous measurement point, and the second depth distance between the current measurement point and the next measurement point; Based on the first depth interval and the displacement of the current measuring point and the previous measuring point, calculate the first rate of change component of the displacement with depth; based on the second depth interval and the displacement of the current measuring point and the next measuring point, calculate the second rate of change component of the displacement with depth. The sum of the first depth spacing and the second depth spacing is taken as the total spacing. The proportions of the first depth spacing and the second depth spacing to the total spacing are calculated respectively, and the proportions are used as the weighted values ​​of the corresponding rate of change components. The product of the first rate of change component and its weighted value is added to the product of the second rate of change component and its weighted value, and the sum is output as the first-order rate of change value at the current depth measurement point.

[0008] As a further aspect of the present invention: marking the depth range covered by the sliding window whose fluctuation amplitude exceeds a preset benchmark value as a strain localization candidate region specifically includes: Obtain the final distribution result of the curvature gradient along the depth, and sequentially extract several continuous depth intervals along the depth direction at preset depth intervals as sliding windows, and record the upper and lower limits of the depth covered by each sliding window. For each sliding window, extract the curvature gradient values ​​at all depth positions within the sliding window, and select the maximum and minimum values ​​from them. Use the absolute difference between the maximum and minimum values ​​as the fluctuation amplitude of the sliding window. The fluctuation amplitude of each sliding window is compared with the preset benchmark value one by one. If the fluctuation amplitude of the current sliding window is greater than or equal to the preset benchmark value, the depth range covered by the sliding window is marked as a strain localization candidate region. If it is less than the preset benchmark value, the depth range is marked as a strain inactive region. All depth intervals marked as candidate regions for strain localization are merged, and the total depth range covered by the merged interval is used as the marking result output for subsequent steps to extract the initial depth of potential sliding surfaces.

[0009] As a further aspect of the present invention: the calculation of the confidence value of the initial selected depth specifically includes: Within each strain localization candidate region, all local extrema points in the curvature gradient distribution along the depth are extracted. The absolute values ​​of the curvature gradients of each local extrema point are compared, and the depth position corresponding to the local extrema point with the largest absolute value is selected as the initial depth of the potential sliding surface. Centered on the initial depth, the preset depth range is expanded to the shallow and deep directions respectively. Within the expanded range, the morphological parameters of the displacement change curve in the interval are extracted. The morphological parameters include the degree of deviation of the displacement at the initial depth from the displacement at both ends of the expanded range, and the degree of symmetry of the change in the slope of the displacement change curve on both sides of the initial depth. Based on the degree of deviation from centering and the degree of symmetry of change, deviation factor and symmetry factor are generated respectively. The deviation factor and symmetry factor are weighted and summed, and the sum is used as the confidence value of the initial selection depth.

[0010] As a further aspect of the present invention: the extraction of morphological parameters of the displacement change curve within the extended range specifically includes: Obtain the displacement corresponding to the initial depth as the center displacement value, obtain the displacement corresponding to the shallow end and deep end of the extended range as the shallow end displacement value and deep end displacement value respectively, calculate the average displacement value of the shallow end displacement value and the deep end displacement value, and then calculate the absolute value of the difference between the center displacement value and the average displacement value, and use the absolute value as the degree of centering deviation. The slope of the displacement change curve at the adjacent position on the shallow side of the initially selected depth is obtained as the shallow slope value, and the slope of the displacement change curve at the adjacent position on the deep side of the initially selected depth is obtained as the deep slope value. The absolute value of the difference between the shallow slope value and the deep slope value is calculated, and then the ratio of the absolute value of the difference to the larger of the shallow slope value and the deep slope value is calculated. The ratio is used as the degree of symmetry of the change. The degree of deviation from the center and the degree of symmetry of change are used as the morphological parameters of the displacement change curve at the initial selected depth.

[0011] As a further aspect of the present invention: the depth value corresponding to the final sliding surface position and its displacement change rate trend are used to determine the landslide risk, specifically including: Obtain the confidence value corresponding to each initial selection depth, take the initial selection depth with the confidence value exceeding the preset threshold as the valid initial selection depth, compare the size of the confidence values ​​corresponding to all valid initial selection depths, and select the valid initial selection depth with the largest confidence value as the final sliding surface position. Obtain the displacement at the final sliding surface position at multiple consecutive historical moments, calculate the displacement change rate corresponding to each historical moment, and determine whether the displacement change rate shows an increasing trend, a decreasing trend, or a constant trend based on the relationship between the displacement change rate and time at each historical moment. The determination result is used as the trend of the displacement change rate. Based on the depth value corresponding to the final sliding surface location and the displacement change rate trend, combined with the preset depth risk classification standard and rate trend risk classification standard, the landslide risk level of the current foundation pit is determined and output.

[0012] As a further aspect of the present invention: the determination and output of the current landslide risk level of the foundation pit specifically includes: Obtain the depth value corresponding to the final sliding surface position, compare the depth value with the excavation depth of the foundation pit, calculate the proportion of the depth value to the excavation depth of the foundation pit, and determine the depth risk level based on the preset proportion range into which the proportion falls. The preset proportion range is divided into low depth risk range, medium depth risk range and high depth risk range in ascending order. The displacement change rate trend is obtained. If the displacement change rate trend is decreasing or constant and the current displacement change rate is lower than the preset rate threshold, it is determined to be a low rate risk level. If the displacement change rate trend is increasing or constant and the current displacement change rate is higher than the preset rate threshold but lower than twice the preset rate threshold, it is determined to be a medium rate risk level. If the displacement change rate trend is increasing and the current displacement change rate is higher than twice the preset rate threshold, it is determined to be a high rate risk level. Based on the corresponding combination of depth risk level and rate risk level, the comprehensive risk level is determined according to the principle that depth risk takes precedence over rate risk, and the comprehensive risk level is output as the landslide risk level of the current foundation pit.

[0013] The beneficial effects of this invention are: (1) This invention constructs a displacement change curve and solves the distribution of curvature gradient along the depth. Combined with the comparison of sliding window fluctuation amplitude and the extraction of local extreme points, it directly locks the strain localization depth range based on the numerical characteristics of the curvature gradient, and then selects the final sliding surface position through confidence evaluation. This process does not rely on the preset condition that the bottom of the inclinometer tube is an absolutely fixed point. Even if the bottom of the inclinometer tube is located in the soft soil creep layer and the whole body drifts, the linear offset component of the displacement curve caused by the drift is automatically filtered out in the curvature gradient calculation and will not form strain localization characteristics. This avoids the situation where rigid body translation is misjudged as shear slip due to the failure of the reference point, and reduces the frequency of invalid shutdown and investigation caused by slip surface alarm.

[0014] (2) In the confidence assessment, this invention expands a preset range centered on the initially selected depth, extracting the degree of deviation of the displacement from the center and the degree of symmetry of the slope change of the displacement curve within the expanded range. The weighted sum of these two values ​​is then used as the output confidence value. This confidence value comprehensively reflects the centering of the displacement curve at the initially selected depth within the expanded range and the symmetry of the slopes on both sides. Only when the displacement at the initially selected depth is centered within the expanded range and the symmetry of the slopes on both sides is high can the confidence value reach the preset threshold, ensuring that the final output of the sliding surface position conforms to the inflection point characteristics of the displacement curve corresponding to a typical shear zone in terms of spatial morphology. The quantitative output of the confidence value provides a repeatable basis for determining the sliding surface position, making the determination process of the monitoring results traceable. Attached Figure Description

[0015] The invention will now be further described with reference to the accompanying drawings.

[0016] Figure 1 This is a system block diagram of the present invention; Figure 2 This is a flowchart of the process for obtaining the distribution of curvature gradient along depth in this invention; Figure 3This is a flowchart of the process for calculating the confidence value of the initial selection depth in this invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Please see Figure 1 As shown, this invention is a real-time monitoring system for landslides in pipeline construction pits based on multiple sensors, comprising: The displacement-depth correspondence data generation module collects horizontal displacement data of each depth measuring point in the slope of the pipeline foundation pit, and synchronously records the depth position corresponding to each measuring point to form a displacement-depth correspondence data set. The curvature gradient distribution calculation module constructs a continuous displacement change curve along the depth direction based on the displacement-depth corresponding data set, and solves the curvature change rate of the displacement change curve at each depth to obtain the distribution result of the curvature gradient along the depth. The strain localization candidate region identification module sets a sliding window along the depth direction, counts the fluctuation amplitude of the curvature gradient within each sliding window, and marks the depth interval covered by the sliding window whose fluctuation amplitude exceeds the preset benchmark value as the strain localization candidate region. The potential slip surface confidence assessment module selects the location where the curvature gradient reaches its extreme value as the initial depth of the potential slip surface within each strain localization candidate region, and calculates the confidence value of the initial depth based on the inflection point morphology characteristics of the displacement change curve at the initial depth. The sliding surface determination output module determines the initial depth where the confidence value exceeds a preset threshold as the final sliding surface position, and outputs the depth value corresponding to the final sliding surface position and its displacement change rate trend, which is used to determine the landslide risk.

[0019] In the displacement-depth correspondence data generation module, horizontal displacement data of each depth measuring point within the pipeline foundation pit slope are collected, and the corresponding depth position of each measuring point is recorded simultaneously to form a displacement-depth correspondence data set, specifically including: On the predetermined monitoring section of the pipeline pit slope, fixed array inclinometer tubes are buried at intervals along the slope depth direction. Several measuring points are arranged at equal intervals along the length of the tube inside the inclinometer tube, and each measuring point has a built-in gravity acceleration sensor element.

[0020] The bottom end of the inclinometer tube is buried in the stable stratum below the excavation surface of the slope, and the top end of the inclinometer tube is fixed at the slope surface to ensure that the inclinometer tube deforms synchronously with the slope. During data acquisition, a matching readout instrument is connected to the output cable at the top of the inclinometer tube, and the angular deflection value of each measuring point relative to the direction of gravity is read sequentially. Based on the angular deflection value of each measuring point and the distance between measuring points, the horizontal displacement of each measuring point relative to the bottom end of the inclinometer tube is calculated as the horizontal displacement data of the slope at that measuring point.

[0021] Based on the depth markings of the measuring points calibrated at the factory of the inclinometer tube, record the depth positions corresponding to each horizontal displacement data, and match the horizontal displacement data with the depth positions one by one, forming a displacement-depth corresponding data group with the depth position as the index.

[0022] To cover the overall deformation characteristics of the pipeline pit slope, at least one monitoring section is set up at the top, middle, and toe of each slope, with at least one array inclinometer tube installed on each monitoring section. The data acquisition cycle is dynamically set according to the excavation progress and weather conditions: during normal excavation, the angle deflection value of each measuring point is collected every 12 hours; during rainy season construction or when the excavation depth exceeds 2 / 3 of the design depth, the acquisition interval is shortened to once every 6 hours; when the slope deformation rate increases abnormally, it is further shortened to once every 1 hour, until the deformation tends to stabilize and then the original acquisition frequency is restored.

[0023] After each data collection, the horizontal displacement data and corresponding depth position of each measuring point are synchronously stored in the monitoring database. Data from different collection times on the same monitoring section are merged according to depth position to form a displacement-depth corresponding data group sequence sorted by depth, which is used for subsequent displacement change curve construction and analysis.

[0024] Please see Figure 2 As shown, in the curvature gradient distribution calculation module, based on the displacement-depth corresponding data set, a continuous displacement change curve along the depth direction is constructed, and the rate of curvature change of the displacement change curve at each depth is solved to obtain the distribution result of the curvature gradient along the depth, specifically including: The depth measurement points in the displacement-depth corresponding data group are arranged in order of depth value from shallow to deep, and the depth value and displacement of each adjacent depth measurement point are obtained in turn.

[0025] For every two adjacent depth measurement points, calculate the difference between their depth values ​​as the depth interval, calculate the difference between their displacements as the displacement change in the depth interval, and then calculate the ratio of the displacement change to the depth interval. This ratio is used as the first-order rate of change of displacement with depth in the depth interval.

[0026] A depth interval is formed between adjacent depth measurement points, and the corresponding first-order rate of change value is obtained. The first-order rate of change values ​​of all depth intervals are summarized to form an initial sequence of first-order rates of change along the depth.

[0027] Based on this, for any current depth measurement point (excluding the shallowest and deepest measurement points), obtain its previous and next depth measurement points, record the depth distance between the current measurement point and the previous measurement point as the first depth distance, and record the depth distance between the current measurement point and the next measurement point as the second depth distance.

[0028] Calculate the difference between the displacement of the current measuring point and the displacement of the previous measuring point, divide it by the first depth interval, and use the resulting ratio as the first rate of change component; calculate the difference between the displacement of the next measuring point and the displacement of the current measuring point, divide it by the second depth interval, and use the resulting ratio as the second rate of change component.

[0029] The sum of the first depth spacing and the second depth spacing is taken as the total spacing. The ratio of the first depth spacing to the total spacing is taken as the first weighted value, and the ratio of the second depth spacing to the total spacing is taken as the second weighted value. The product of the first rate of change component multiplied by the first weighted value is added to the product of the second rate of change component multiplied by the second weighted value, and the sum is taken as the first-order rate of change value at the current depth measuring point. The first-order rate of change values ​​at each depth measuring point except for the two end measuring points are calculated in this way and added to the corresponding depth positions in the aforementioned initial sequence of first-order rate of change along the depth to form a complete sequence of first-order rate of change along the depth.

[0030] Arrange the first-order rate of change values ​​in the complete sequence along the depth according to their corresponding depth positions from shallow to deep. Calculate the difference between the first-order rate of change values ​​at two adjacent depth positions, and use this difference as the initial value of the rate of change of curvature at the boundary position between the two adjacent depth positions. Traverse all adjacent depth positions, and summarize the initial value of the rate of change of curvature and its depth position corresponding to each boundary position to form the preliminary distribution result of the curvature gradient along the depth.

[0031] The initial distribution of curvature gradient along the depth is processed by low-amplitude oscillation elimination.

[0032] The specific method is as follows: set the amplitude threshold to 0.3 mm per square meter, check the curvature gradient value at each depth position point by point along the depth direction, and when the curvature gradient value at three or more consecutive depth positions alternates between positive and negative values ​​and the absolute value of the curvature gradient at each depth position is less than 0.3 mm per square meter, the continuous depth interval is determined to be a low amplitude oscillation interval.

[0033] For depth ranges identified as low-amplitude oscillation ranges, the average value of the curvature gradient at both ends of the range is taken, and this average value is used to replace the original curvature gradient values ​​at each depth position within the range to complete the smoothing correction.

[0034] For continuous depth intervals where the absolute value of the curvature gradient exceeds 0.3 mm per square meter and maintains a consistent sign, the original curvature gradient values ​​are retained without further processing and treated as unidirectional continuous variation intervals. After processing all depth locations, the corrected curvature gradient values ​​at each depth location are output as the final distribution result of the curvature gradient along the depth, for use in subsequent steps to extract candidate regions for strain localization.

[0035] In the strain localization candidate region identification module, a sliding window is set along the depth direction, and the fluctuation amplitude of the curvature gradient within each sliding window is counted. The depth interval covered by the sliding window whose fluctuation amplitude exceeds a preset benchmark value is marked as a strain localization candidate region, specifically including: Obtain the final distribution of the curvature gradient along the depth, which includes the curvature gradient values ​​at each depth location. Along the depth direction, from shallow to deep, sequentially extract continuous depth intervals at preset depth intervals, each interval serving as a sliding window. The preset depth interval for the sliding windows is set to 0.5 meters, meaning each sliding window covers a depth range of 0.5 meters in thickness. During extraction, the starting depth positions between two adjacent sliding windows differ by 0.25 meters, ensuring a 0.25-meter overlap between adjacent sliding windows. For each extracted sliding window, record the starting and ending depth positions covered by that window, which serve as the upper and lower depth limits for that sliding window.

[0036] For each sliding window, extract all depth positions and their corresponding curvature gradient values ​​within the depth range covered by the sliding window. Select the maximum and minimum values ​​from all extracted curvature gradient values, calculate the difference between the maximum and minimum values, and take the absolute value of this difference as the fluctuation amplitude of the sliding window. Thus, each sliding window obtains a corresponding fluctuation amplitude.

[0037] The fluctuation amplitude of each sliding window is compared one by one. A preset reference value is set to 0.25 m -1. The relationship between the fluctuation amplitude of each sliding window and the preset reference value is determined one by one: if the fluctuation amplitude of the current sliding window is greater than or equal to 0.25 m -1, the depth range covered by the sliding window is marked as a candidate region for strain localization; if the fluctuation amplitude of the current sliding window is less than 0.25 m -1, the depth range covered by the sliding window is marked as a strain inactive region.

[0038] The depth intervals covered by all sliding windows marked as strain localization candidate regions are merged. During merging, if the depth gap between two adjacent strain localization candidate regions is less than or equal to 0.25 meters, they are merged into a single continuous depth interval to fill the gap; if the depth gap between two adjacent strain localization candidate regions is greater than 0.25 meters, they remain independent and are not merged. After merging, the total depth range covered by the merged region is output as the marking result. This marking result contains one or more continuous depth intervals, each of which is a candidate region with relatively severe curvature gradient fluctuations, for use in subsequent steps to extract the initial depth of potential sliding surfaces.

[0039] Please see Figure 3 As shown, in the potential slip surface confidence assessment module, within each strain localization candidate region, the location where the curvature gradient reaches its extreme value is selected as the initial depth of the potential slip surface. Based on the inflection point morphology of the displacement change curve at the initial depth, the confidence value of the initial depth is calculated, specifically including: Within each strain localization candidate region, the final distribution result of the curvature gradient along the depth is obtained, which includes all curvature gradient values ​​within the depth range covered by the candidate region.

[0040] The curvature gradient values ​​at each depth location are checked point by point along the depth direction from shallow to deep. The curvature gradient value at a certain depth location is compared with the curvature gradient values ​​at its shallow and deep adjacent locations. If the value is greater than or less than both its shallow and deep adjacent values, then the depth location is determined to be a local extreme point.

[0041] A strain localization candidate region may contain multiple local extrema. After extracting all local extrema within the candidate region, the absolute value of the curvature gradient corresponding to each local extrema is taken. The magnitudes of the absolute values ​​are compared, and the depth position corresponding to the local extrema with the largest absolute value is selected as the initial depth of the potential sliding surface within the candidate region. For cases containing multiple strain localization candidate regions, an initial depth is extracted independently for each candidate region.

[0042] Centered on the initially selected depth, a preset depth range is extended in both the shallow and deep directions. This preset depth range is set to 0.5 meters on the shallow side and 0.5 meters on the deep side of the initially selected depth, meaning the total depth range covered after expansion is a depth interval centered on the initially selected depth with a total thickness of 1 meter. Within the expanded depth range, the displacement corresponding to each depth position within this range is extracted from the displacement-depth correspondence data set, and the slope value of the curve at each depth position within this expanded range is obtained based on the displacement change curve.

[0043] The specific method for extracting morphological parameters is as follows: Obtain the displacement corresponding to the initially selected depth as the center displacement value; obtain the displacement at the shallow end of the extended range as the shallow end displacement value; and obtain the displacement at the deep end of the extended range as the deep end displacement value. Calculate the sum of the shallow end displacement value and the deep end displacement value, divide by two, and use the result as the average displacement value.

[0044] The absolute value of the difference between the center displacement value and the average displacement value is calculated. This absolute value is used as the degree of centering deviation. The smaller the value, the closer the displacement at the initial depth is to the average displacement at both ends of the extended range, indicating that the displacement at the initial depth is well centered within the extended range. Simultaneously, the slope value of the displacement change curve at the adjacent position on the shallow side of the initial depth is obtained as the shallow slope value, and the slope value of the displacement change curve at the adjacent position on the deep side of the initial depth is obtained as the deep slope value. The absolute value of the difference between the shallow slope value and the deep slope value is calculated. Then, the ratio of this absolute value to the larger of the shallow slope value and the deep slope value is calculated. This ratio is used as the degree of symmetry of change. The smaller the value, the closer the curve slopes on both sides of the initial depth are to symmetry.

[0045] A deviation factor is generated based on the degree of centering deviation, and a symmetry factor is generated based on the degree of change symmetry. The deviation factor is generated by dividing the degree of centering deviation by the difference between the maximum and minimum displacement values ​​within the extended range; the resulting ratio is used as the deviation factor. The deviation factor ranges from 0 to 1. A deviation factor closer to 0 indicates a more concentrated displacement at the initial depth within the extended range, while a deviation factor closer to 1 indicates a displacement at the initial depth is more biased towards one end of the extended range. The symmetry factor is generated by directly using the degree of change symmetry as the symmetry factor; that is, the symmetry factor equals the degree of change symmetry. The symmetry factor also ranges from 0 to 1. A symmetry factor closer to 0 indicates a more symmetrical slope of the displacement curves on both sides of the initial depth, while a symmetry factor closer to 1 indicates a higher degree of asymmetry in the slope of the displacement curves on both sides of the initial depth.

[0046] The formula used to calculate the deviation factor is as follows: ; in, This represents the displacement at the initially selected depth. This indicates the displacement at the shallow end of the extended range. This indicates the displacement at the deep endpoint of the extended range. This represents the maximum displacement across all depth locations within the extended range. This represents the minimum displacement across all depth locations within the extended range. The ratio calculated using the above formula is the deviation factor.

[0047] The formula used to calculate the symmetry factor is as follows: ; in, This represents the slope of the displacement curve at the location immediately adjacent to the shallow side of the initial depth. This represents the slope of the displacement curve at the location immediately adjacent to the deep side of the initially selected depth. Indicates taking and The larger of the two values. The ratio calculated using the above formula is the symmetry factor.

[0048] The deviation factor and the symmetry factor are weighted and summed, and the sum is used as the confidence value for the initial selected depth. The weight of the deviation factor is set to 0.4, and the weight of the symmetry factor is set to 0.6, with a sum of 1. The confidence value is calculated by adding the product of the deviation factor multiplied by 0.4 and the product of the symmetry factor multiplied by 0.6; the sum is the confidence value. The confidence value ranges from 0 to 1; a smaller value indicates lower confidence in the initial selected depth as the true sliding surface, and a larger value indicates higher confidence.

[0049] The calculated confidence value is output as the confidence value of the initial selected depth, which is then compared with a preset threshold in subsequent steps to determine whether the initial selected depth can be used as the final sliding surface position. For cases involving multiple strain localization candidate regions, the confidence value of each candidate region's initial selected depth is calculated independently, and each initial selected depth and its corresponding confidence value are output separately.

[0050] In the sliding surface determination output module, the initial depth where the confidence value exceeds a preset threshold is determined as the final sliding surface location. The module outputs the depth value corresponding to the final sliding surface location and its displacement rate trend to determine landslide risk. Specifically, this includes: Obtain all initial selection depths and their corresponding confidence values. Set the confidence threshold to 0.60. For each initial selection depth, determine if its confidence value is greater than or equal to 0.60. Select depths with confidence values ​​greater than or equal to 0.60 as valid initial selection depths, and discard those with confidence values ​​lower than 0.60, not using them for subsequent decisions. If there is only one valid initial selection depth, use it directly as the final sliding surface position. If there are multiple valid initial selection depths, compare the confidence values ​​of each and select the one with the highest confidence value as the final sliding surface position. If the confidence values ​​of all initial selection depths are lower than 0.60, determine that there is no reliable sliding surface position in the current monitoring data, terminate the subsequent risk assessment process, and output a "No Sliding Surface" message.

[0051] After determining the final sliding surface position, the displacement values ​​at multiple consecutive historical moments at that position are extracted from the monitoring database. The number of historical moments is set to six, corresponding to the current acquisition moment and the five acquisition cycles preceding it, with equal time intervals between each acquisition cycle. After extracting the displacement values ​​for each of these six historical moments, the displacement change between adjacent historical moments is calculated sequentially. Each displacement change is divided by its corresponding time interval, and the resulting ratio is taken as the displacement change rate for each historical moment, thus obtaining the displacement change rate values ​​for each of the six historical moments.

[0052] The trend is determined based on the relationship between these six displacement rate values ​​and time: if the displacement rate value at the current acquisition time is greater than the displacement rate value at the previous acquisition time, and the displacement rate value at the previous acquisition time is greater than the displacement rate value at the acquisition time before that, that is, the displacement rate value increases for three or more consecutive acquisition times, then the displacement rate trend is determined to be an increasing trend; if the displacement rate value at the current acquisition time is less than the displacement rate value at the previous acquisition time, and the displacement rate value at the previous acquisition time is less than the displacement rate value at the acquisition time before that, that is, the displacement rate value decreases for three or more consecutive acquisition times, then the displacement rate trend is determined to be a decreasing trend; if the absolute value of the difference between the displacement rate values ​​at adjacent acquisition times is less than 0.01 mm per hour, then the displacement rate trend is determined to be a constant trend. If it does not fall into any of the above three categories, the displacement rate trend is determined to be a fluctuating trend. The increasing trend, decreasing trend, constant trend, or fluctuating trend obtained from the determination is output as the displacement rate trend.

[0053] Based on the depth value corresponding to the final sliding surface location and the displacement change rate trend, combined with the preset depth risk classification standard and rate trend risk classification standard, the landslide risk level of the current foundation pit is determined. The specific classification method is as follows: obtain the depth value corresponding to the final sliding surface location, obtain the current excavation depth value of the foundation pit, and calculate the ratio of this depth value to the total excavation depth of the foundation pit, that is, the ratio obtained by dividing the final sliding surface location depth value by the foundation pit excavation depth value.

[0054] Three preset ratio ranges are defined, from smallest to largest: less than or equal to 0.5 is the low-risk range, greater than 0.5 and less than or equal to 0.8 is the medium-risk range, and greater than 0.8 is the high-risk range. The level of risk is determined based on the range in which the ratio falls.

[0055] The criteria for rate trend risk classification are as follows: If the displacement change rate trend is decreasing or constant, and the current displacement change rate is less than 0.05 mm per hour, it is classified as a low rate risk level; if the displacement change rate trend is increasing or constant, and the current displacement change rate is greater than or equal to 0.05 mm per hour but less than 0.10 mm per hour, it is classified as a medium rate risk level; if the displacement change rate trend is increasing, and the current displacement change rate is greater than or equal to 0.10 mm per hour, it is classified as a high rate risk level. Fluctuation trends are uniformly classified as medium rate risk levels.

[0056] After the depth risk level and rate risk level are determined, the comprehensive risk level is determined according to the principle that depth risk takes precedence over rate risk.

[0057] The specific combination method is as follows: If the depth risk level is high depth risk, then the comprehensive risk level is directly determined to be high risk level, and the rate risk level is no longer considered. If the depth risk level is medium-deep risk and the rate risk level is high rate risk, then the overall risk level is determined to be high risk level. If the rate risk level is medium rate risk or low rate risk, then the overall risk level is determined to be medium risk level. If the depth risk level is low and the rate risk level is high, the overall risk level is determined to be medium. If the rate risk level is medium or low, the overall risk level is determined to be low.

[0058] The final determined comprehensive risk level will be used as the current landslide risk level output for the foundation pit.

[0059] The working principle of this invention is as follows: Horizontal displacement data at various depths is collected using fixed array inclinometers embedded in the slope of the foundation pit, and the corresponding depth positions are recorded, forming a displacement-depth data set. Based on this data set, a displacement variation curve along the depth direction is constructed, and the rate of curvature change of this curve at each depth is calculated to obtain the distribution of the curvature gradient along the depth. A sliding window is set along the depth direction, and the fluctuation amplitude of the curvature gradient within each window is statistically analyzed. The depth interval covered by the window whose fluctuation amplitude exceeds a preset benchmark value is marked as a candidate region for strain localization. Within each candidate region for strain localization, [the following steps are taken]. The depth corresponding to the local extreme point with the largest absolute value of curvature gradient is taken as the initial depth of the potential sliding surface. A preset range is expanded with this initial depth as the center, and the degree of deviation from the center and the degree of change symmetry within the expanded range are extracted as morphological parameters. The confidence value of the initial depth is calculated based on the morphological parameters. The initial depth with a confidence value exceeding a preset threshold is determined as the final sliding surface location. The displacement change rate trend is determined based on the historical displacement data of this location. The landslide risk level is comprehensively determined and output by combining the proportion of the final sliding surface depth to the excavation depth of the foundation pit and the displacement change rate trend.

[0060] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A real-time monitoring system for landslides in pipeline construction pits based on multiple sensors, characterized in that, include: The displacement-depth correspondence data generation module collects horizontal displacement data of each depth measuring point in the slope of the pipeline foundation pit, and synchronously records the depth position corresponding to each measuring point to form a displacement-depth correspondence data set. The curvature gradient distribution calculation module constructs a continuous displacement change curve along the depth direction based on the displacement-depth corresponding data set, and solves the curvature change rate of the displacement change curve at each depth to obtain the distribution result of the curvature gradient along the depth. The strain localization candidate region identification module sets a sliding window along the depth direction, counts the fluctuation amplitude of the curvature gradient within each sliding window, and marks the depth interval covered by the sliding window whose fluctuation amplitude exceeds the preset benchmark value as the strain localization candidate region. The potential slip surface confidence assessment module selects the location where the curvature gradient reaches its extreme value as the initial depth of the potential slip surface within each strain localization candidate region, and calculates the confidence value of the initial depth based on the inflection point morphology characteristics of the displacement change curve at the initial depth. The sliding surface determination output module determines the initial depth where the confidence value exceeds a preset threshold as the final sliding surface position, and outputs the depth value corresponding to the final sliding surface position and its displacement change rate trend, which is used to determine the landslide risk.

2. The real-time monitoring system for landslides in pipeline construction pits based on multiple sensors according to claim 1, characterized in that, The method for obtaining the distribution of curvature gradient along depth specifically includes: Arrange the depth measurement points in the displacement-depth corresponding data group in order from shallow to deep, and calculate the first-order rate of change of displacement relative to depth between adjacent depth measurement points to obtain the first-order rate of change value of each depth interval. Arrange the first-order rate of change values ​​of each depth interval in depth order, and calculate the difference between the first-order rate of change values ​​of adjacent depth intervals. Use the difference as the initial value of the rate of change of curvature at the boundary of adjacent depth intervals, and summarize the initial values ​​of the rate of change of curvature at all boundary positions to form the preliminary distribution of curvature gradient along the depth. The initial distribution of curvature gradient along the depth is processed by removing low-amplitude oscillations. The oscillation intervals where the curvature gradient values ​​continuously cross positive and negative values ​​and the amplitude is lower than the preset amplitude threshold are smoothed and corrected. The unidirectional continuous change intervals where the amplitude exceeds the preset amplitude threshold are retained. The corrected curvature gradient values ​​are used as the final distribution result of curvature gradient along the depth.

3. The real-time monitoring system for landslides in pipeline construction pits based on multiple sensors according to claim 2, characterized in that, The process of obtaining the first-order rate of change values ​​for each depth interval specifically includes: Obtain the current depth measurement point and its adjacent previous and next depth measurement points, record the first depth distance between the current measurement point and the previous measurement point, and the second depth distance between the current measurement point and the next measurement point; Based on the first depth interval and the displacement of the current measuring point and the previous measuring point, calculate the first rate of change component of the displacement with depth; based on the second depth interval and the displacement of the current measuring point and the next measuring point, calculate the second rate of change component of the displacement with depth. The sum of the first depth spacing and the second depth spacing is taken as the total spacing. The proportions of the first depth spacing and the second depth spacing to the total spacing are calculated respectively, and the proportions are used as the weighted values ​​of the corresponding rate of change components. The product of the first rate of change component and its weighted value is added to the product of the second rate of change component and its weighted value, and the sum is output as the first-order rate of change value at the current depth measurement point.

4. The real-time monitoring system for landslides in pipeline construction pits based on multiple sensors according to claim 1, characterized in that, The step of marking the depth range covered by the sliding window whose fluctuation amplitude exceeds a preset benchmark value as a strain localization candidate region specifically includes: Obtain the final distribution result of the curvature gradient along the depth, and sequentially extract several continuous depth intervals along the depth direction at preset depth intervals as sliding windows, and record the upper and lower limits of the depth covered by each sliding window. For each sliding window, extract the curvature gradient values ​​at all depth positions within the sliding window, and select the maximum and minimum values ​​from them. Use the absolute difference between the maximum and minimum values ​​as the fluctuation amplitude of the sliding window. The fluctuation amplitude of each sliding window is compared with the preset benchmark value one by one. If the fluctuation amplitude of the current sliding window is greater than or equal to the preset benchmark value, the depth range covered by the sliding window is marked as a strain localization candidate region. If it is less than the preset benchmark value, the depth range is marked as a strain inactive region. All depth intervals marked as candidate regions for strain localization are merged, and the total depth range covered by the merged interval is used as the marking result output for subsequent steps to extract the initial depth of potential sliding surfaces.

5. The real-time monitoring system for landslides in pipeline construction pits based on multiple sensors according to claim 1, characterized in that, The calculation of the confidence value for the initial selected depth specifically includes: Within each strain localization candidate region, all local extrema points in the curvature gradient distribution along the depth are extracted. The absolute values ​​of the curvature gradients of each local extrema point are compared, and the depth position corresponding to the local extrema point with the largest absolute value is selected as the initial depth of the potential sliding surface. Centered on the initial depth, the preset depth range is expanded to the shallow and deep directions respectively. Within the expanded range, the morphological parameters of the displacement change curve in the interval are extracted. The morphological parameters include the degree of deviation of the displacement at the initial depth from the displacement at both ends of the expanded range, and the degree of symmetry of the change in the slope of the displacement change curve on both sides of the initial depth. Based on the degree of deviation from centering and the degree of symmetry of change, deviation factor and symmetry factor are generated respectively. The deviation factor and symmetry factor are weighted and summed, and the sum is used as the confidence value of the initial selection depth.

6. The real-time monitoring system for landslides in pipeline construction pits based on multiple sensors according to claim 5, characterized in that, The extraction of morphological parameters of the displacement change curve within the extended range specifically includes: Obtain the displacement corresponding to the initial depth as the center displacement value, obtain the displacement corresponding to the shallow end and deep end of the extended range as the shallow end displacement value and deep end displacement value respectively, calculate the average displacement value of the shallow end displacement value and the deep end displacement value, and then calculate the absolute value of the difference between the center displacement value and the average displacement value, and use the absolute value as the degree of centering deviation. The slope of the displacement change curve at the adjacent position on the shallow side of the initially selected depth is obtained as the shallow slope value, and the slope of the displacement change curve at the adjacent position on the deep side of the initially selected depth is obtained as the deep slope value. The absolute value of the difference between the shallow slope value and the deep slope value is calculated, and then the ratio of the absolute value of the difference to the larger of the shallow slope value and the deep slope value is calculated. The ratio is used as the degree of symmetry of the change. The degree of deviation from the center and the degree of symmetry of change are used as the morphological parameters of the displacement change curve at the initial selected depth.

7. A real-time monitoring system for landslides in pipeline construction pits based on multiple sensors, as described in claim 1, is characterized in that... The depth value corresponding to the final sliding surface position and its displacement change rate trend are used to determine the landslide risk, specifically including: Obtain the confidence value corresponding to each initial selection depth, take the initial selection depth with the confidence value exceeding the preset threshold as the valid initial selection depth, compare the size of the confidence values ​​corresponding to all valid initial selection depths, and select the valid initial selection depth with the largest confidence value as the final sliding surface position. Obtain the displacement at the final sliding surface position at multiple consecutive historical moments, calculate the displacement change rate corresponding to each historical moment, and determine whether the displacement change rate shows an increasing trend, a decreasing trend, or a constant trend based on the relationship between the displacement change rate and time at each historical moment. The determination result is used as the trend of the displacement change rate. Based on the depth value corresponding to the final sliding surface location and the displacement change rate trend, combined with the preset depth risk classification standard and rate trend risk classification standard, the landslide risk level of the current foundation pit is determined and output.

8. A real-time monitoring system for landslides in pipeline construction pits based on multiple sensors, as described in claim 7, is characterized in that... The process of determining and outputting the current landslide risk level of the foundation pit specifically includes: Obtain the depth value corresponding to the final sliding surface position, compare the depth value with the excavation depth of the foundation pit, calculate the proportion of the depth value to the excavation depth of the foundation pit, and determine the depth risk level based on the preset proportion range into which the proportion falls. The preset proportion range is divided into low depth risk range, medium depth risk range and high depth risk range in ascending order. The displacement change rate trend is obtained. If the displacement change rate trend is decreasing or constant and the current displacement change rate is lower than the preset rate threshold, it is determined to be a low rate risk level. If the displacement change rate trend is increasing or constant and the current displacement change rate is higher than the preset rate threshold but lower than twice the preset rate threshold, it is determined to be a medium rate risk level. If the displacement change rate trend is increasing and the current displacement change rate is higher than twice the preset rate threshold, it is determined to be a high rate risk level. Based on the corresponding combination of depth risk level and rate risk level, the comprehensive risk level is determined according to the principle that depth risk takes precedence over rate risk, and the comprehensive risk level is output as the landslide risk level of the current foundation pit.

Images