Real-time signal early warning system based on geotechnical engineering slope monitoring

Abstract

The invention discloses a real-time signal early warning system based on geotechnical engineering slope monitoring, and belongs to the technical field of geotechnical engineering safety monitoring. The system comprises a multi-source data sensing module, a multi-field coupling mechanism analysis module, a critical state early warning judgment module and a multi-mode early warning release module. Four types of slope data are collected and divided into primary data and secondary data, and data validity is guaranteed through standardized preprocessing; effective rainfall, pore pressure response and microseismic damage functions are constructed based on a physical mechanism, and a slope internal disturbance index is generated through coupling of a geometric mean method; constructing a dynamic critical displacement rate function and a ratio type early warning index based on a critical state line theory, and generating a discrete early warning level by combining a time accumulation mechanism after correction and smoothing; outputting graded early warning signals according to the early warning grades, linking emergency equipment, and executing closed-loop management at the same time; according to the method, unbiased coupling and dynamic critical discrimination of slope multi-source data are realized, and reliable technical support is provided for safety prevention and control of slope engineering.

Description

A Real-Time Signal Early Warning System Based on Geotechnical Engineering Slope Monitoring Technical Field

[0001] This invention relates to the field of geotechnical engineering safety monitoring technology, and more specifically, to a real-time signal early warning system based on geotechnical engineering slope monitoring. Background Technology

[0002] Geotechnical engineering slopes are widely distributed in highway, railway, mining, and water conservancy projects. Slope instability and landslides can easily trigger geological disasters, causing damage to engineering facilities, casualties, and significant economic losses. Therefore, real-time monitoring and accurate early warning of slope instability risk have become the core link in the safety control of geotechnical engineering. At present, the industry has developed slope monitoring and early warning technology solutions with varying degrees of maturity. The mainstream approach is to collect basic monitoring data such as deep slope displacement, rainfall intensity, pore water pressure, and microseismic energy through various sensors. After simple outlier screening and data preprocessing, multi-source data are integrated using a weighted method based on human experience, or slope instability risk is determined directly through fixed thresholds of single-type data. Finally, early warning signals are issued to relevant management personnel through conventional means such as on-site audible and visual alarms and SMS notifications. This type of solution has been initially applied in the daily monitoring of various geotechnical engineering slopes.

[0003] Existing slope monitoring and early warning technologies, when applied in practical engineering, exhibit numerous significant shortcomings due to their low alignment between design logic and the physical laws governing slope instability, and a lack of standardized implementation procedures. These shortcomings fail to meet the core requirements of engineering sites for accurate, real-time, and reliable early warning systems. The data acquisition and preprocessing stages lack unified standardization requirements. Monitoring data is not prioritized based on the characteristics of slope instability, and sensor placement and sampling cycles are not specifically tailored to the specific characteristics of the slope engineering. Furthermore, data preprocessing merely involves simple outlier removal, without effectively supplementing missing data, significantly compromising the effectiveness and stability of the monitoring data. The multi-source data fusion process lacks support from the physical mechanisms of the soil and rock mass, relying solely on subjective weighting based on human experience. It fails to construct specific quantification functions for the different disturbance characteristics of rainfall, pore pressure, and microseismic events, thus failing to accurately reflect the synergistic disturbance effects of multiple factors and resulting in significant deviations in the representation of the actual internal state of the slope. The early warning system uses a fixed threshold judgment logic, failing to consider the dynamic impact of internal slope disturbance on the critical instability state. This contradicts the actual physical laws of slope instability and is prone to misjudgment and omission due to instantaneous data fluctuations. The design of the early warning issuance and execution process is also inadequate. A tiered early warning mechanism is lacking, as are corresponding time-accumulation judgment rules. The output format of early warning signals is relatively simple, with poor linkage with on-site drainage, access control, and emergency broadcasting equipment. Furthermore, the lack of a closed-loop management process for early warning record storage and receiver feedback verification leads to untimely early warning responses and compromises the efficiency and effectiveness of on-site emergency response. Summary of the Invention

[0004] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a real-time signal early warning system based on geotechnical engineering slope monitoring. The system addresses the technical problems of existing slope monitoring and early warning systems, such as low data validity, strong subjectivity in multi-source data fusion, large early warning discrimination bias, single early warning release format, and poor emergency response coordination.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a real-time signal early warning system based on geotechnical engineering slope monitoring, comprising: a multi-source data sensing module: which collects deep displacement data, rainfall intensity data, pore water pressure data, and microseismic energy data in real time through different types of sensors deployed at the target location of the slope, and defines deep displacement data as the main data and the other three types as auxiliary data; a multi-field coupling mechanism analysis module: which receives the three types of auxiliary data and constructs an effective rainfall function, a pore pressure response function, and a microseismic damage function based on physical mechanisms, and analyzes and generates an internal disturbance index of the slope based on the effective rainfall function, the pore pressure response function, and the microseismic damage function by constructing a coupling function; a critical state early warning discrimination module: which constructs an exponential decay type critical displacement rate function and a ratio type early warning index function, and generates discrete early warning levels by combining the deep displacement data and the internal disturbance index of the slope; and a multimodal early warning release module: which receives the early warning levels and outputs multimodal early warning signals matching the early warning levels according to a preset correspondence.

[0006] The technical effects and advantages of this invention are as follows: 1. By classifying slope monitoring data into primary and secondary categories, and combining this invention with standardized sensor selection, deployment specifications, and data preprocessing methods, the invention systematically standardizes the entire process of slope monitoring data collection, clarifies the core data basis for slope instability judgment, and effectively improves the effectiveness, stability, and relevance of monitoring data, laying a reliable data foundation for subsequent early warning and judgment work; 2. Based on the physical mechanism of soil and rock, this invention constructs dedicated quantification functions for different types of auxiliary monitoring data, abandoning the traditional multi-source data fusion method of subjective weighting, and adopts an unbiased coupling method to achieve organic fusion of multi-field data. This can realistically and accurately characterize the synergistic disturbance effect of multiple factors within the slope, allowing the generated slope internal disturbance index to be more accurate. The invention closely matches the actual slope condition, solving the problems of strong subjectivity and large deviations in the internal slope state representation of traditional multi-source data fusion. Furthermore, by constructing a dynamic critical displacement rate function based on critical state line theory, combined with the correction and smoothing of ratio-based early warning indicators, and a time-accumulated early warning level determination mechanism, the invention achieves dynamic critical discrimination of slope instability, effectively avoiding the problems of missed or misjudgment caused by instantaneous data fluctuations or single-dimensional discrimination. Simultaneously, by establishing a hierarchical multimodal early warning release mechanism, the invention enables differentiated linkage between early warning signals and various on-site emergency equipment, and implements closed-loop management of the entire process after early warning release, significantly improving the accuracy, real-time performance, and emergency response capabilities of slope early warning, providing reliable technical support for the safety control of various geotechnical engineering slopes. Attached Figure Description

[0007] Figure 1 is a schematic diagram of the overall system structure of the present invention; Figure 2 is a schematic diagram of the process of collecting main data and auxiliary data of the present invention; Figure 3 is a schematic diagram of the process of obtaining the internal disturbance index of the slope of the present invention; Figure 4 is a schematic diagram of the process of obtaining the early warning level of the present invention; Figure 5 is a schematic diagram of the process of issuing multimodal early warning signals of the present invention. Detailed Implementation

[0008] 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.

[0009] As shown in Figures 1 to 5, a real-time signal early warning system based on geotechnical engineering slope monitoring includes: a multi-source data sensing module: using different types of sensors deployed at the target location of the slope to collect deep displacement data, rainfall intensity data, pore water pressure data, and microseismic energy data in real time, defining deep displacement data as the primary data and the other three types as auxiliary data; it should be specifically explained that the specific process of the multi-source data sensing module collecting the required data in real time is as follows: Deep displacement data: a sliding joint gauge is used as the acquisition sensor. The sensor has an accuracy of not less than ±0.1mm and a collection range covering 0 to 500mm. The sensor is rigidly installed 2 to 5m above the potential sliding surface of the slope, with one sensor every 50m along the slope direction. The installation contact surface is ground flat and then fixed by anchoring to ensure that the sensing direction is consistent with the potential sliding direction of the slope; after the sensor is installed, zero-point calibration is performed to eliminate the initial installation deviation. The deep relative displacement value of the slope soil and rock mass is continuously collected with a sampling period of 10 minutes. Abnormal jumps caused by installation gaps and temperature drift are eliminated to obtain stable deep displacement time series data.

[0010] Rainfall intensity data: A tipping bucket rain gauge was used as the data acquisition sensor, with a measurement accuracy of no less than ±0.01mm and a range of 0–4mm / min. The sensor was installed on the top of the slope in an unobstructed area without water catchment, at least 5m from the edge of the slope, and 1.5m above the ground, ensuring horizontal installation. Real-time rainfall intensity was collected synchronously with a sampling period of 5 minutes. During the collection process, interference from fallen leaves and debris was shielded to obtain rainfall intensity data that truly reflects the rainfall infiltration conditions of the slope area.

[0011] Pore ​​water pressure data: Vibrating wire pore water pressure gauges are used as the data acquisition sensors, with an accuracy of no less than ±1 kPa and a range covering 0–1000 kPa. One gauge is deployed every 2 m along the slope depth direction, buried 1–3 m below the groundwater level and not exceeding the potential sliding surface depth of the slope. After drilling and installation, bentonite is used to seal the boreholes to ensure full coupling between the sensor and the soil mass. After the sensor is allowed to stabilize for 24 hours, zero-point calibration is performed. The sampling period is 15 minutes, and the pore water pressure changes inside the slope are collected in real time to obtain pore water pressure data reflecting the effective stress changes of the soil mass.

[0012] Microseismic energy data: A piezoelectric microseismic sensor was used as the data acquisition sensor, with an energy detection range of 10. -3 ~10 3 J, accuracy not less than ±10 -4 J. One sensor is installed every 30m along the slope direction in areas of stress concentration such as the slope toe and slope waist. The sensor is rigidly attached to the slope surface and the sensing direction is perpendicular to the slope surface to shield against environmental mechanical vibration interference. The energy value released by micro-fractures inside the rock and soil is collected with a sampling period of 20 minutes to form microseismic energy time series data reflecting the degree of damage to the internal structure of the slope.

[0013] After data acquisition is completed, the multi-source data sensing module preprocesses four types of raw data through the configured embedded data processing unit: outliers caused by sensor failure, environmental interference, etc. are removed using the 3σ criterion; missing data are supplemented by linear interpolation of adjacent valid data (if ≥3 consecutive data are missing, sensor failure needs to be investigated and data is re-acquired); and data classification is completed after preprocessing.

[0014] It needs further explanation that deep displacement data is defined as primary data, while the other three categories are defined as auxiliary data. The specific reasons are as follows: Deep displacement is the most direct and essential macroscopic manifestation of slope instability and failure. Its magnitude and rate of change directly correspond to the severity of slope slippage and deformation, and it is the core basis for judging whether a slope has entered a critical state of instability. It can independently reflect the current stability state of the slope, and therefore is considered primary data. Rainfall intensity is an external excitation factor that induces an increase in slope moisture content and softens the soil and rock. Pore water pressure is a key intrinsic indicator that reduces the effective stress of the soil and rock and weakens its shear strength. Microseismic energy is an early damage signal that characterizes the initiation and expansion of internal cracks in the slope. All three are indirect causes that trigger changes in deep displacement and induce slope instability. They cannot be used alone to determine whether a slope is on the verge of instability. They can only be used to form disturbance indicators through mechanism coupling to assist deep displacement data in completing early warning judgment. Therefore, they are classified as auxiliary data.

[0015] Multi-field coupling mechanism analysis module: This module receives three types of auxiliary data and constructs effective rainfall functions, pore pressure response functions, and microseismic damage functions based on physical mechanisms. It then analyzes and generates the slope internal disturbance index through coupling functions based on these functions. Specifically, when receiving the three types of auxiliary data, the time series of all auxiliary data is first aligned to a 10-minute sampling period for deep displacement. Then, effective rainfall functions, pore pressure response functions, and microseismic damage functions are constructed based on physical mechanisms, and the slope internal disturbance index is generated through coupling analysis. The specific process is as follows: Effective rainfall function construction: Considering the physical characteristic that rainfall disturbance to slope soil and rock is a nonlinear infiltration saturation effect, and combining the soil and rock permeability with the cumulative rainfall infiltration law, the saturated infiltration depth of the soil and rock is introduced for dimensionless processing, constructing an exponential effective rainfall function E(t). This quantifies the equivalent rainfall intensity of the hydrodynamic disturbance actually generated by the infiltration into the soil and rock. The function expression is: In the formula, I(t) represents the real-time rainfall intensity at time t after preprocessing, in mm / min; Δt = 5 min is the fixed sampling period for rainfall data; n t The number of rainfall sampling points included in the cumulative calculation before time t is determined based on the saturation infiltration time (24 hours) of the slope soil and rock mass. (Fixed value); The cumulative rainfall intensity before time t reflects the cumulative effect of historical rainfall infiltration, in mm; H is the saturated infiltration depth of the soil and rock mass, measured by indoor infiltration tests, in mm, used for dimensionless exponential term; k is the permeability coefficient of the soil and rock mass, determined by indoor constant head permeability tests, in mm / min, characterizing the permeability of the soil and rock mass; E(t) is the effective rainfall function value at time t, in mm. 2 / min, the greater the rainfall intensity, the more cumulative the rainfall, and the stronger the permeability of the soil and rock, the larger the function value, and the more significant the hydrodynamic softening disturbance.

[0016] Construction of the pore pressure response function: Based on the effective stress principle, an increase in pore water pressure nonlinearly weakens the effective stress of soil and rock, and the closer the pore pressure is to the limit value, the more severe the impact on stability. Combining the porosity of soil and rock with the ultimate pore water pressure, a logarithmic pore pressure response function U(t) is constructed to quantify the degree to which pore water pressure weakens the effective stress of soil and rock. The function expression is as follows: In the formula, u(t) is the real-time pore water pressure at time t after pretreatment, in kPa; umax is the ultimate pore water pressure of the slope, which is determined by the saturated unit weight of the soil and rock and the slope height, in kPa, representing the upper limit of pore pressure before slope instability. U(t) represents the porosity of the soil and rock mass, which is determined by laboratory geotechnical tests. It is dimensionless and reflects the degree of porosity development in the soil and rock mass. U(t) is the pore pressure response function value at time t, in kPa. The closer the pore pressure is to the limit value, the faster the function value increases, and the more severe the effective stress loss in the soil and rock mass.

[0017] Construction of Microseismic Damage Function: Based on the cumulative damage mechanism of microfractures in rock mass, the greater the release of microseismic energy, the more intense the internal crack propagation and the more severe the structural damage. An exponential microseismic damage function D(t) is constructed to quantify the degree of structural damage to rock mass caused by microfractures. The function expression is as follows: In the formula, E m (t) represents the real-time microseismic energy at time t after preprocessing, in J; Δt'=20min is the fixed sampling period for microseismic data; m t The number of microseismic sampling points participating in the cumulative calculation before time t is determined according to the cumulative damage characteristic duration of slope microfractures (48h). (Fixed value); The accumulated microseismic energy before time t reflects the cumulative damage degree of microfractures in the rock mass, expressed in J; σ c V represents the uniaxial compressive strength of the soil and rock mass, determined by laboratory mechanical tests, in kPa, characterizing the rock mass's resistance to damage; V is the effective volume of the soil and rock mass in the monitoring area, calculated from the slope dimensions, in m³. 3 ;σ c V represents the characteristic resistance energy of the rock mass, used to normalize the accumulated microseismic energy; D(t) is the microseismic damage coefficient at time t, which is dimensionless and ranges from [0,1). The larger the accumulated microseismic energy, the closer the coefficient is to 1, and the worse the integrity of the rock mass structure.

[0018] Generation of the internal disturbance index of the slope: Considering that the disturbances of the slope caused by rainfall hydrodynamic field, pore pressure stress field, and microseismic damage field are synergistic and mutually reinforcing, to avoid the subjectivity of artificial weighting, a geometric mean coupling method of cube root of product is adopted to unbiasedly fuse the three types of mechanism functions, generating the internal disturbance index D of the slope that characterizes the overall disturbance degree of the slope. dist (t), the coupling formula is: In the formula, The dimensionless effective rainfall function value (E) max (This represents the measured maximum value of E(t) during the monitoring period). The dimensionless pore pressure response function value (U) max For u max The calculated value is obtained by substituting it into U(t), and the specific calculation formula is as follows: D(t) is the microseismic damage coefficient; D dist(t) is the internal disturbance index of the slope at time t, which is dimensionless and ranges from [0,1]. The larger the index value, the more severe the combined disturbance of the slope soil and rock mass caused by rainfall infiltration, pore pressure increase and micro-seismic damage, the more deteriorated the internal structure and stress state, and the higher the risk of slope instability.

[0019] Critical State Early Warning Judgment Module: Constructs an exponentially decaying critical displacement rate function and a ratio-based early warning index function, and generates discrete early warning levels by combining the deep displacement data and the slope internal disturbance index. Specifically, the construction of the exponentially decaying critical displacement rate function is as follows: First, the deep displacement data and the slope internal disturbance index are received, and an exponentially decaying critical displacement rate function is constructed based on the critical state line theory. After receiving the data, preprocessing and verification are performed to ensure that the data meets the judgment requirements: For deep displacement time series data x(t), its continuity (no more than 3 consecutive missing data points) and stability (no residual outliers exceeding the 3σ criterion) are verified. If an anomaly exists, it is immediately fed back to the multi-source data sensing module for re-collection or supplementary preprocessing; for the slope internal disturbance index D... dist (t), verify its value range (ensure D) dist (t)≥0, since all three disturbances are non-negative effects, the exponent has no negative value). If a negative value appears, it is judged as an abnormal coupling calculation, and the multi-field coupling mechanism analysis module is called again to perform coupling calculation until the data verification is qualified; Actual displacement rate calculation and smoothing: Based on the verified deep displacement time series data x(t), the actual displacement rate v(t) at time t is calculated using the "adjacent time displacement difference method". The core purpose is to quantify the instantaneous intensity of the current sliding deformation of the slope. The specific calculation process is as follows: Basic rate calculation: The displacement difference between two adjacent sampling times is divided by the sampling period to obtain the original actual displacement rate. The calculation expression is: In the formula, v raw (t) represents the original actual displacement rate at time t, in mm / min; x(t) represents the preprocessed deep displacement value at time t, in mm; The deep displacement value in mm is the value of the sampled layer before time t. A fixed sampling period is used for deep displacement data; rate smoothing is applied: the original actual displacement rate may be affected by minute fluctuations in the sensor and instantaneous environmental interference, resulting in instantaneous jumps. Therefore, a "3-point moving average method" is used to smooth the velocity. raw (t) is smoothed to eliminate instantaneous disturbances and obtain a stable actual displacement rate v(t). The smoothing formula is: when t=1, v(1)=v raw (1); When t=2, When t≥3, In the formula, v(t) is the actual displacement rate at time t after smoothing, in mm / min; These are the original actual displacement rates of the two sampling periods before time t, respectively; outlier rate removal: after smoothing, the 3σ criterion is used again to identify outliers in v(t). If v(t) at a certain time satisfies |v(t)-μ v |>3σ v (where μ) v σ is the mean of the smoothed rates of the most recent 10 sets. v If the corresponding standard deviation is used, it is determined to be an abnormal rate. The linear interpolation method of two adjacent effective rates is used to supplement it to ensure the stability of the actual displacement rate time series data. The magnitude of v(t) directly corresponds to the severity of slope sliding deformation. The larger v(t) is, the faster the slope rock and soil sliding speed is, the more severe the internal structural deformation is, and the closer it is to the critical state of instability. When v(t) tends to be stable and the value is small, it means that the slope is in a stable state.

[0020] It should be further explained that, based on the critical state line theory (the core idea of ​​this theory is that the essence of slope instability is that "the actual displacement rate exceeds the critical displacement rate under the current state," and the critical displacement rate is not a fixed value, but decreases as the degree of internal disturbance of the slope increases), combined with the slope internal disturbance index D... dist The physical meaning of (t) is used to construct the exponentially decaying critical displacement rate function v. c (t), serving as the core benchmark for determining whether a slope is approaching a critical instability state, is constructed as follows: Function construction idea: Slope internal disturbance index D dist (t) comprehensively reflects the synergistic disturbance effect of rainfall infiltration, pore water pressure increase, and microseismic damage. dist The larger (t) is, the more deteriorated the soil and rock structure and the weaker the resistance to sliding. The corresponding critical displacement rate (i.e., the maximum allowable displacement rate for the slope to remain stable) is lower. The two exhibit a non-linear exponential decay relationship—when there is no disturbance (D dist (t) = 0), the critical displacement rate is at its maximum; when the disturbance intensifies, the critical displacement rate decays rapidly. Therefore, an exponential decay form is used to construct the function to ensure that it conforms to the physical mechanism of slope instability; function expression: Detailed parameter calibration and explanation: v c (t): Critical displacement rate of the slope at time t, in mm / min. Its core physical meaning is "the maximum sliding rate that the soil and rock mass can withstand under the current level of internal disturbance of the slope; exceeding this rate will cause the slope to enter a critical state of instability"; v0: Initial critical displacement rate under undisturbed conditions, in mm / min, i.e., D. distThe critical displacement rate at t=0 represents the ultimate slip resistance rate of a slope in a healthy, undisturbed state. The calibration method is as follows: the shear strength τ of the slope's soil and rock mass is determined through indoor direct shear tests. Combined with slope geometric parameters (slope θ, sliding surface area A) and soil and rock weight γ, the rate is derived using mechanical equilibrium equations and the creep time effect. The derived formula is: (where ξ is the rate correction coefficient, ranging from 0.01 to 0.03, and is calibrated using a combination of engineering analogy and indoor tests: the initial value is first determined by measured values ​​of slopes of the same type and geotechnical mass in the same region, and the final value is obtained by back-calculation through indoor slope model tests; L is the characteristic displacement of the sliding body, determined by the arc length of the potential sliding surface of the slope, in mm; V is the volume of the sliding body; T is the characteristic coefficient of the sliding time of the geotechnical mass, in min, measured by indoor creep tests).

[0021] α: Rate decay coefficient, dimensionless, characterizing the internal disturbance index D of the slope. dist (t) The degree of attenuation of the critical displacement rate; the larger α is, the more significant the attenuation effect of the disturbance on the critical rate. The calibration method is as follows: through indoor slope model tests, simulate the critical displacement rate under different disturbance levels (different rainfall intensities, pore pressures, and microseismic energy) to establish v c With D dist By fitting the relationship, the optimal value of α is obtained by reverse calculation, with a range of 0.8 to 1.2, ensuring that the function can accurately reflect the nonlinear relationship between the disturbance and the critical rate; D dist (t): The internal disturbance index of the slope at time t; it should be further explained that, in order to intuitively and quantitatively reflect the distance between the current state of the slope and the critical state of instability, and to avoid the bias caused by the single displacement rate or disturbance index, the actual displacement rate v(t) and the critical displacement rate v(t) are used as the basis for the determination. c (t), construct a ratio-based early warning index R(t). The specific construction process is as follows: Function construction idea: The early warning index should be able to directly reflect the "degree of deviation of the actual state from the critical state". Therefore, the "ratio of the actual displacement rate to the critical displacement rate" is used as the core index. The larger the ratio, the closer the actual rate is to or exceeds the critical rate, and the higher the risk of slope instability; the smaller the ratio, the more stable the slope. Basic function expression: In the formula, R raw v(t) represents the original warning index at time t, which is dimensionless and its core physical meaning is "the multiple of the actual displacement rate relative to the critical displacement rate"; v(t) represents the smoothed actual displacement rate. c (t) represents the critical displacement rate at time t.

[0022] Indicator Correction Processing: To avoid distortion of the indicator due to extreme cases, the original early warning indicator Rraw(t) needs to be corrected: When D distWhen (t) → 0 (undisturbed state), v c If v(t) = v0, and v(t) is extremely small (approaching 0) at this point, it may lead to R raw (t) approaches 0, requiring no correction, and is directly determined to be a stable state; when D dist (t) is relatively large (D) dist (t)≥0.8), and v c When (t) approaches 0, to avoid R raw (t) approaches infinity, so set an upper limit R for the index. max =2.0, that is, when R raw When R(t) ≥ 2.0, we uniformly take R(t) = 2.0, corresponding to the highest warning level; the expression for the corrected warning index R(t) is: Indicator smoothing: Consistent with the actual displacement rate processing, a 3-point moving average method is used to smooth R(t), eliminating misjudgments caused by instantaneous fluctuations and ensuring the stability of the early warning indicator. The smoothing formula is as follows: , (R s (t) represents the smoothed early warning indicator; subsequent early warning level determinations are all based on R. s (based on t); it should be further explained that R s (t) takes values ​​in the range [0, 2.0], is dimensionless, and its value is positively correlated with the slope instability risk: R s The smaller (t) is, the lower the actual displacement rate is compared to the critical displacement rate, indicating weak internal disturbances and better stability of the slope; R s The larger (t) is, the closer the actual displacement rate is to or exceeds the critical displacement rate, the more severe the internal disturbance of the slope, and the higher the risk of instability; when R s When (t) = 1.0, the actual displacement rate is equal to the critical displacement rate, and the slope enters the critical state of instability, requiring an immediate early warning response.

[0023] Multimodal early warning release module: used to receive the early warning level and output a multimodal early warning signal that matches the early warning level according to a preset correspondence.

[0024] It should be specifically explained that the specific implementation process of the multimodal early warning release module outputting the multimodal early warning signal is as follows: The module first synchronously receives the corrected and smoothed ratio-type early warning index R output by the critical state early warning discrimination module. s (t), and simultaneously receive supporting core monitoring data, including the actual displacement rate v(t) at time t and the critical displacement rate v. c (t) Slope internal disturbance index D dist (t) and deep displacement time series data x(t), all data carrying a unified timestamp t; after receiving the data, the module first performs time series alignment and validity verification, matching the timestamp of each data with R. sThe timestamps of (t) are precisely matched to ensure that the data are synchronized data from the same monitoring time; at the same time, R is verified. s Does the range of values ​​for (t) meet the correction result requirement of [0, 2.0]? If R... s An outlier (t) < 0 is determined to be an error in the upstream indicator calculation. It is immediately fed back to the critical state early warning judgment module for recalculation. If the data timing is misaligned or the supporting data is missing, the early warning process is suspended and a data retransmission instruction is issued until complete and valid synchronization data is received.

[0025] Based on the verified early warning indicator R s (t), the module strictly follows the quantitative judgment logic of slope instability risk in geotechnical engineering, and constructs a system with R... s (t) A discrete early warning level determination rule with one-to-one correspondence is used to complete the transformation from continuous indicators to discrete levels. This determination rule is consistent with R. s The physical meaning of (t) is deeply intertwined, serving as a core link between critical state judgment and early warning issuance. The judgment logic is as follows: when R s When (t) < 0.5, it is determined to be a no-warning level; when 0.5 ≤ R s When (t) < 1.0, it is determined to be a blue alert level; when 1.0 ≤ R s When (t) < 1.5, it is judged as a yellow warning level; when R s When (t) ≥ 1.5, it is judged as a red alert level; to avoid R at a single moment s (t) Fluctuations lead to false alarms. The module introduces a time accumulation judgment mechanism, that is, the judgment condition of a certain warning level must be met for 3 consecutive deep displacement sampling cycles before the warning level can be finally confirmed. The sampling cycle is adjusted according to the warning level: 10 min for no warning, 5 min for blue warning, and 2 min for yellow / red warning. The judgment cycle is adjusted to 3 consecutive new sampling cycles (30 min for no warning, 15 min for blue warning, and 6 min for yellow / red warning). When the warning level drops, the sampling cycle is restored step by step (red→yellow→blue: 2 min→5 min; blue→no warning: 5 min→10 min). Each adjustment is executed after the data of 3 consecutive new sampling cycles is stable. If only one or two sampling cycles meet the warning condition, the warning level of the previous moment is maintained. This mechanism is consistent with the basic sampling cycle of the critical state warning judgment module.

[0026] After confirming the warning level, the module calls the built-in warning signal matching database. This database has a pre-defined correspondence between "warning level - multimodal signal". The core matching rule is: no warning level corresponds to basic status prompt signal, blue warning level corresponds to mild warning signal, yellow warning level corresponds to moderate warning and initial linkage signal, and red warning level corresponds to severe warning and full emergency linkage signal. Based on the finally confirmed warning level, the module retrieves the corresponding signal parameters from the database, including the working mode of the on-site audio-visual equipment, the prompt format of the remote terminal, the push range of the mobile terminal, and the linkage instructions of the emergency equipment. At the same time, it also sets the current R... s (t), v(t), D dist (t) is encapsulated as early warning details data and embedded into various early warning signals to ensure that the receiving end can obtain complete judgment basis.

[0027] It should be further explained that the generation and output of on-site audible and visual warning signals rely on outdoor waterproof audible and visual alarms and high-brightness LED warning lights pre-deployed on the slope. All devices are wired to the module via the Modbus communication protocol and are equipped with backup batteries to ensure operation during power outages. The module sends precise control commands to the audible and visual devices based on the matched signal parameters: when there is no warning, only a solid green command is sent to the warning lights, while the audible and visual alarms remain silent, indicating that the on-site system is operating normally; when a blue warning occurs, a blue intermittent flashing (1 second on, 2 seconds off) and low-frequency buzzer (1 time / 3 seconds) command is sent, with the volume controlled at 60dB, and simultaneously a command containing R is sent to the on-site data display screen. s (t) Text prompts for value acquisition; During a yellow alert, a command is sent to flash yellow rapidly (on for 0.5s, off for 0.5s) and sound a medium-frequency buzzer (1 time / 1s), with the volume increased to 75dB, and the display screen scrolls through risk warnings and core monitoring data; During a red alert, a command is sent to keep red on and sound a high-frequency buzzer (2 times / 1s), while simultaneously triggering a continuous sounding of the on-site emergency siren, with the volume locked at 85dB, and the display screen displays an emergency evacuation prompt in full screen, with all sound and light signals continuously output until the alert level is reduced or lifted; The generation and output of the warning signal from the remote monitoring terminal are achieved through an Ethernet link with the terminal system of the slope monitoring and control center, with data transmission delay controlled within 1s; The module sends the encapsulated warning information to the monitoring terminal in the form of data frames: When there is no warning, only the real-time data bar on the terminal interface is updated with R. s (t) and associated monitoring data, without pop-ups or sound prompts; when a blue alert is issued, a blue-bordered alert pop-up window appears on the terminal, accompanied by a 3-second low-frequency sound. The pop-up window displays the alert level, timestamp, and R. s (t) and the judgment criteria, the terminal warning status bar displays a blue indicator; when a yellow warning is issued, the terminal pops up a yellow-bordered pop-up window, accompanied by a 10-second mid-frequency prompt tone, and the pop-up window automatically loads nearly 1 hour of R... sThe terminal interface automatically switches to the real-time monitoring screen of the slope monitoring area when the curves of (t) and v(t) change are displayed. When a red alert is issued, a red full-screen pop-up window appears on the terminal, accompanied by a continuous high-frequency prompt sound. The pop-up window is locked and cannot be manually closed. At the same time, the terminal's recording function is triggered to synchronously store the monitoring screen and monitoring data during the alert period. The recording retention time is no less than 72 hours. The generation and output of the mobile terminal's alert signal adopts a dual-channel mode of APP push and SMS push. The module has a built-in personnel group management list, and the list is bound to the alert level to set the push range: the blue alert is only pushed to the APP terminal of the on-site monitoring personnel and the construction team leader; the yellow alert is pushed to the APP terminal of the above personnel and simultaneously sends an SMS to the project management personnel; the red alert is pushed to the APP terminal and SMS terminal of all on-site workers, project management personnel, and superior competent personnel. The pushed information content is based on the encapsulated alert details data and adopts a standardized text format. The APP push includes a link to view real-time data, while the SMS push extracts the core information. For example, the content of the red alert SMS is "[Slope Emergency Warning] At time t, R..." s "(t) = 1.8, the slope is on the verge of instability, evacuate the danger zone immediately." The module enables batch and rapid sending of SMS messages through an interface with telecommunications operators, with a push delay of no more than 30 seconds. The APP push supports offline caching, ensuring that personnel can receive messages synchronously after going offline and connecting to the internet. The generation and output of emergency equipment linkage signals is the core execution link of multimodal early warning. The module achieves linkage with drainage equipment, emergency broadcasting, access control systems, and temporary anchoring equipment at the slope site through standardized control interfaces, with a linkage response time of no more than 2 seconds. According to the matching rules of the early warning level, the module sends differentiated linkage instructions. Commands: During a blue alert, send a sampling period adjustment command to the deep displacement sensor, shortening the 10-minute sampling period to 5 minutes. Simultaneously, send a low-power operation command to the drainage equipment to slowly reduce the pore water pressure inside the slope. During a yellow alert, send a full-power operation command to the drainage equipment, activate the emergency broadcast system to continuously play patrol and evacuation preparation reminders, and send a start command to the slope sprinkler system to assist in reducing the water content of the soil and rock. During a red alert, send an emergency start command to all emergency equipment, turn on emergency lighting to ensure evacuation safety, and close the access control system to prevent personnel from entering the danger zone. When R... s (t) ≥ 1.8 and v (t) ≥ 1.2v c (t) At this time, temporary anchoring equipment is activated to mitigate slope slippage, and the equipment continues to operate until R. s (t) < 1.0 and D dist When (t) < 0.5, stop, activate temporary anchoring equipment to slow down slope slippage, and at the same time send emergency support request data including warning level, monitoring data and site location to the local emergency management department's linkage platform.

[0028] It should be further explained that after all warning signals are output, the module initiates a closed-loop management process, receiving feedback signals from each receiving end and linked equipment in real time: on-site audio-visual equipment and emergency equipment report their operating status, mobile terminal receivers report "received" confirmation messages, and the monitoring center reports warning handling instructions; the module associates and stores all feedback information with warning release records, level determination records, and data verification records for a storage period of no less than one year, facilitating subsequent project review; when the critical state warning judgment module outputs R... s (t) When the sampling period is below 0.5 for 5 consecutive sampling periods, the module determines that the warning is lifted, immediately restores the sampling period of the deep displacement sensor to the original 10 minutes, sends a warning lifting instruction to all terminals and devices, stops all warning signals and emergency equipment linkage, and pushes the warning lifting notification in sync, thus completing the closed loop of the entire warning process.

[0029] Secondly: The accompanying drawings of the embodiments disclosed in this invention only involve the structures involved in the embodiments disclosed in this invention. Other structures can refer to the general design. In the absence of conflict, the same embodiment and different embodiments of this invention can be combined with each other.

[0030] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A real-time signal early warning system based on geotechnical engineering slope monitoring, characterized in that, include: Multi-source data sensing module: This module collects real-time deep displacement data, rainfall intensity data, pore water pressure data, and microseismic energy data using different types of sensors deployed at the target location on the slope. Deep displacement data is defined as the primary data, while the other three types are defined as auxiliary data. Multi-field coupling mechanism analysis module: This module receives the three auxiliary data types and constructs effective rainfall functions, pore pressure response functions, and microseismic damage functions based on physical mechanisms. Based on these functions, it analyzes and generates the internal disturbance index of the slope by constructing coupling functions. The effective rainfall function is: In the formula, I(t) is the real-time rainfall intensity at time t after preprocessing; Δt=5min is the fixed sampling period for rainfall data; n t This represents the number of rainfall sampling points that participated in the cumulative calculation before time t; The cumulative rainfall intensity before time t; H is the saturated infiltration depth of the soil and rock mass; k is the permeability coefficient of the soil and rock mass; E(t) is the effective rainfall function value at time t; the pore pressure response function is: In the formula, u(t) is the real-time pore water pressure at time t after pretreatment; u max The ultimate pore water pressure of the slope is φ; the porosity of the soil and rock mass is φ; U(t) is the pore pressure response function value at time t; the microseismic damage function is: In the formula, E m (t) represents the real-time microseismic energy at time t after preprocessing; Δt'=20min is the fixed sampling period for microseismic data; m t This represents the number of microseismic sampling points that participated in the cumulative calculation before time t; σ represents the accumulated microseismic energy before time t; c V represents the uniaxial compressive strength of the soil and rock mass; V represents the effective volume of the soil and rock mass in the monitoring area; D(t) represents the microseismic damage coefficient at time t; the coupling function is: In the formula, The dimensionless effective rainfall function value. Here, D(t) represents the dimensionless pore pressure response function value, and D(t) represents the micro-vibration damage coefficient. dist (t) represents the internal disturbance index of the slope at time t; Critical state early warning discrimination module: constructs an exponential decay type critical displacement rate function and a ratio type early warning index function, and generates discrete early warning levels by combining the deep displacement data and the internal disturbance index of the slope; Multimodal early warning release module: used to receive the early warning level and output a multimodal early warning signal that matches the early warning level according to a preset correspondence.

2. The real-time signal early warning system based on geotechnical engineering slope monitoring according to claim 1, characterized in that: The process of the multi-source sensing module collecting the required data is as follows: Deep displacement data: A sliding joint gauge is used as the data acquisition sensor. The sensor has an accuracy of no less than ±0.1mm and a data acquisition range covering 0–500mm. The sensor is rigidly installed 2–5m above the potential sliding surface of the slope, with one station every 50m along the slope direction. The deep relative displacement values ​​of the slope's soil and rock mass are continuously collected at a sampling period of 10min. Abnormal jumps are eliminated to obtain the deep displacement data. Rainfall intensity data: A tipping bucket rain gauge is used as the data acquisition sensor. The measurement accuracy is no less than ±0.01mm and the data acquisition range covers 0–4mm / min. The sensor is installed on the top of the slope where there is no obstruction or obstruction. In the catchment area, at least 5m from the edge of the slope, the installation height is 1.5m above the ground and horizontally positioned. Real-time rainfall intensity is collected at a sampling interval of 5 minutes. Pore water pressure data: Vibrating wire pore water pressure gauges are used as the acquisition sensors, with an accuracy of at least ±1kPa and a range covering 0–1000kPa. One gauge is deployed every 2m along the slope depth, buried 1–3m below the groundwater level and not exceeding the potential sliding surface depth of the slope. The pore water pressure changes within the slope are collected in real-time at a sampling interval of 15 minutes to obtain pore water pressure data. Microseismic energy data: Piezoelectric microseismic sensors are used as the acquisition sensors, with an energy detection range of 10... -3 ~10 3 J, accuracy not less than ±10 -4 J. In the stress concentration area of ​​the slope, one sampling point is set up every 30m along the slope direction, and the energy value released by the micro-fractures inside the rock and soil is collected with a sampling period of 20min to form microseismic energy data.

3. The real-time signal early warning system based on geotechnical engineering slope monitoring according to claim 1, characterized in that: The reasons for defining the main data and auxiliary data are as follows: Deep displacement data is the main data and is a macroscopic representation of slope instability and failure. The magnitude and rate of change of the values ​​correspond to the severity of slope slippage and deformation, reflecting the current stability state of the slope. Rainfall intensity data, pore water pressure data, and microseismic energy data are auxiliary data. All three are indirect causes that trigger deep displacement changes and induce slope instability. They form disturbance indicators through mechanism coupling, which assist deep displacement data in completing early warning judgment.

4. The real-time signal early warning system based on geotechnical engineering slope monitoring according to claim 1, characterized in that: The process of constructing the multi-field coupling mechanism analysis module based on auxiliary data is as follows: First, the time series of the three auxiliary data are aligned to the sampling period of the deep displacement data. Then, an effective rainfall function is constructed based on the physical characteristics of the nonlinear infiltration saturation effect of rainfall. Based on the effective stress principle, the porosity of the rock and soil and the ultimate pore water pressure are combined to construct the pore pressure response function. Based on the cumulative damage mechanism of rock micro-fracture, a micro-seismic damage function is constructed.

5. A real-time signal early warning system based on geotechnical engineering slope monitoring according to claim 1, characterized in that: The method for obtaining the internal disturbance index of the slope is as follows: First, the effective rainfall function and pore pressure response function are dimensionless, and the microseismic damage function is a dimensionless coefficient. Then, the geometric mean coupling method of taking the cube root of the product is used to unbiasedly fuse the dimensionless effective rainfall function, pore pressure response function and microseismic damage function, and the internal disturbance index of the slope is obtained after calculation.

6. A real-time signal early warning system based on geotechnical engineering slope monitoring according to claim 1, characterized in that: The exponentially decaying critical displacement rate function is constructed based on the critical state line theory. The function reflects the nonlinear exponential decay relationship between the internal disturbance index and the critical displacement rate of the slope. The larger the internal disturbance index of the slope, the lower the corresponding critical displacement rate. In the undisturbed state, the internal disturbance index of the slope is zero, and the critical displacement rate at this time is the initial critical displacement rate.

7. A real-time signal early warning system based on geotechnical engineering slope monitoring according to claim 1, characterized in that: The discrete warning levels are divided into four levels: no warning, blue warning, yellow warning, and red warning, based on the range of values ​​of the smoothed ratio-type warning index. A time accumulation judgment mechanism is introduced, and the judgment condition of a certain warning level must be met for three consecutive corresponding sampling periods and finally confirmed. When the warning level falls, the sampling period is restored level by level. Each adjustment is performed after the data of three consecutive new sampling periods has stabilized.

8. A real-time signal early warning system based on geotechnical engineering slope monitoring according to claim 1, characterized in that: The multimodal early warning signals include on-site audible and visual early warning signals, remote monitoring terminal early warning signals, mobile terminal early warning signals, and emergency equipment linkage signals. Each type of early warning signal adopts a differentiated output form according to the early warning level. On-site audible and visual early warning signals are realized by outdoor waterproof audible and visual alarms and high-brightness LED warning lights. Remote monitoring terminal early warning signals are output to the monitoring center terminal system through an Ethernet link. Mobile terminal early warning signals adopt a dual-channel mode of application push and SMS push. Emergency equipment linkage signals are sent to drainage equipment, emergency broadcast, access control system, and temporary anchoring equipment at the slope site. After all early warning signals are output, the closed-loop management process is initiated.

Images