Foundation waterproof early warning and seepage drainage system and method

Abstract

The invention discloses a foundation waterproof early warning and seepage drainage system and method, and relates to the field of seepage prevention and control of constructional engineering. The system is composed of a monitoring platform, a power supply module, communication equipment and a drainage module. The drainage module comprises an electrode array which is uniformly distributed along the periphery of the foundation, a multi-source sensor is integrated on the electrode array, and soil moisture content, temperature, conductivity and underground water level data are collected in real time; the monitoring platform calculates the comprehensive moisture content through a dynamic weight fusion algorithm, predicts the future trend in combination with an LSTM neural network, dynamically corrects an alarm threshold value based on the underground water level rising and falling rate and the meteorological rainfall intensity, triggers multi-stage early warning, improves the monitoring and seepage and drainage control capability of the foundation waterproof system and the early warning accuracy of the system, and improves the water seepage and drainage control capability of the foundation waterproof system. And false alarm or missing alarm is avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of leakage prevention and control in construction engineering, and particularly to a foundation waterproof warning and drainage system and method. Background Art

[0002] With the acceleration of the urbanization process, the scale of underground projects is constantly expanding, and foundation waterproofing and leakage control have become the key to ensuring building safety.

[0003] Traditional foundation waterproofing technologies mainly rely on physical isolation (such as waterproof membranes) or passive drainage (such as blind ditch drainage). The drainage efficiency is low, mainly relying on natural infiltration. The drainage rate is insufficient in cohesive soils or high water table strata, and local water accumulation is likely to occur. In addition, most existing alarm systems use fixed moisture content thresholds to monitor the soil layer on the periphery of the foundation, which is prone to false alarms or missed alarms.

[0004] Therefore, it is necessary to provide a foundation waterproof warning and drainage system and method to solve the problems mentioned in the above background art. Summary of the Invention

[0005] To achieve the above object, the present invention provides the following technical solution: a foundation waterproof warning and drainage system and method, including: a monitoring platform, a power supply module, a communication device, and a drainage module disposed on the outer peripheral side of the foundation;

[0006] A multi-source information acquisition network is arranged on the drainage module, and the multi-source information acquisition network includes:

[0007] A soil parameter acquisition unit for real-time acquisition of the volumetric water content of the target soil layer;

[0008] An auxiliary parameter acquisition unit for synchronously monitoring the soil temperature and conductivity to compensate for the moisture content measurement error;

[0009] An environmental parameter acquisition unit for monitoring the groundwater level and connecting meteorological data;

[0010] The multi-source information acquisition network transmits the acquired multi-source data to the monitoring platform through the communication device after fusion;

[0011] The monitoring platform includes a control module and a warning module. The control module is built-in with a dynamic threshold model, which dynamically calculates the comprehensive moisture content alarm threshold based on the historical moisture content trend and the current soil layer state, as well as the groundwater level, precipitation intensity, and duration in the real-time environment. The control module issues a warning signal to the warning module for multi-level warning based on the comprehensive moisture content alarm threshold, the measured moisture content value, and the preset warning mechanism, and starts the drainage module to perform corresponding drainage operations based on the warning level;

[0012] The power supply module is used to provide a gradient DC voltage to the drainage module.

[0013] Preferably, the drainage module further includes: a plurality of electrode arrays uniformly distributed along the outer contour of the foundation. The electrode arrays are composed of vertically buried cathode conductive rods and a plurality of anode conductive rods. The layout spacing between each anode conductive rod and the cathode conductive rod is equal, and they are connected to the positive and negative electrodes of the power supply module;

[0014] A plurality of anti-interference segments are uniformly distributed along the axial direction of the anode conductive rod and the cathode conductive rod. The soil parameter acquisition unit, the auxiliary parameter acquisition unit, and the environmental parameter acquisition unit are embedded on each anti-interference segment, jointly constituting the multi-source information acquisition network;

[0015] A water collection unit is arranged at the bottom of the cathode conductive rod for collecting and discharging the water body collected by the electrode array.

[0016] Preferably, the water collection unit includes a water collection bin and a drain pipe. The water collection bin is a cylindrical bin. The top of the cylindrical bin is an inwardly concave frustum structure. The bottom of the frustum structure is fixedly connected to the cathode conductive rod. Water diversion ports are uniformly arranged on the conical surface of the frustum structure, and a filtering structure is arranged at the water diversion ports;

[0017] One end of the drain pipe is fixedly arranged at the bottom of the water collection bin, and the other end is connected to an external suction device. A liquid level sensor is arranged in the water collection bin.

[0018] Preferably, the soil parameter acquisition unit includes a time domain reflectometry sensor, and the time domain reflectometry sensor is hermetically embedded on the anti-interference segment.

[0019] Preferably, the auxiliary parameter acquisition unit includes a thermistor and a four-electrode sensor embedded on the anti-interference segment. The thermistor is used to collect the soil temperature, and the four-electrode sensor is used to collect the soil ion concentration.

[0020] Preferably, the environmental parameter acquisition unit includes a piezoresistive water level sensor and a meteorological data interface embedded on the anti-interference segment. The water level sensor intermittently collects the groundwater level at a set frequency;

[0021] The meteorological data interface is connected to a regional meteorological station for real-time acquisition of future precipitation intensity prediction data.

[0022] Preferably, the dynamic threshold model is configured as:

[0023] Continuously collect historical moisture content data, and synchronously record the soil temperature, conductivity, groundwater level, and meteorological precipitation intensity at the corresponding time points;

[0024] Analyze the fluctuation characteristics of the historical moisture content data and its correlation with the groundwater level, and extract trend parameters;

[0025] Perform time series modeling on historical data through a long short-term memory neural network to predict the water content change curve in the future period;

[0026] Combine the real-time groundwater level rise and fall rate with the predicted value of meteorological precipitation to dynamically adjust the boundary of the water content alarm threshold.

[0027] Preferably, the data fusion method includes:

[0028] Filter the original data collected by the water level sensor and retain the stable data reflecting the real water level change;

[0029] Integrate the measurement results of the time domain reflectometer, conductivity sensor and water level sensor, and obtain the comprehensive soil water content through weighted calculation.

[0030] Preferably, the configuration relationship between the warning mechanism of the warning module and the drainage module is:

[0031] When it is detected that the real-time water content exceeds the comprehensive water content threshold and the duration is ≥ 15 minutes, trigger a primary warning, and control the power module to apply a first gradient voltage to the electrode array where the warning point is located for spot drainage;

[0032] When it is detected that the groundwater level rise rate > 5 cm / h and the comprehensive soil water content exceeds the threshold by > 30%, trigger an intermediate warning, and control the power module to apply a second gradient voltage to at least three groups of circumferentially adjacent electrode arrays where the warning point is located for regional drainage;

[0033] When it is detected that the rainstorm intensity > 50 mm / h and the groundwater level rise rate > 10 cm / h, trigger an emergency warning, and control the power module to apply the maximum gradient voltage to all electrode arrays for comprehensive drainage.

[0034] The foundation waterproof warning and infiltration drainage method includes the following steps:

[0035] S1. Arrange electrode arrays along the outer contour of the foundation, and set up a multi-source information acquisition network on the electrode arrays to collect the volumetric water content, soil temperature, conductivity and groundwater level data of the target soil layer in real time, and synchronously obtain the predicted information of meteorological precipitation;

[0036] S2. Filter the original data collected by the water level sensor, extract the effective water level change data, dynamically adjust the weights based on the sensor type accuracy and environmental conditions, and perform weighted fusion calculation on the water content related data;

[0037] S3. Analyze the correlation between historical water content data and groundwater level, construct a time series prediction model, and dynamically correct the water content alarm threshold in combination with the real-time groundwater level change rate and the predicted precipitation intensity;

[0038] S4. Based on the corrected moisture content alarm threshold, the measured moisture content value, and the preset early warning mechanism, send an early warning signal to the early warning module for multi-level early warning, and activate the drainage module to perform corresponding drainage operations based on the early warning level.

[0039] Compared with the prior art, the present invention provides a foundation waterproof early warning and infiltration drainage system and method, which has the following beneficial effects:

[0040] In the present invention, through the synergistic effect of multi-source sensor fusion, dynamic threshold early warning, and gradient electroosmotic drainage, the monitoring and infiltration drainage control capabilities of the foundation waterproof system are improved; at the same time, the combination of intelligent weight allocation and the LSTM prediction model is used to dynamically adjust the comprehensive moisture content threshold in real time, improve the early warning accuracy of the system, and avoid false alarms or missed alarms. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a schematic diagram of the overall structure of the foundation waterproof early warning and infiltration drainage system and method;

[0042] Figure 2 It is a schematic diagram of the structure of the foundation waterproof early warning and infiltration drainage system and method;

[0043] In the figure: 1. Monitoring platform; 2. Communication device; 3. Drainage module; 4. Multi-source information acquisition network; 5. Power supply module; 6. Cathode conductive rod; 7. Anode-cathode conductive rod; 8. Anti-interference section; 9. Water collection bin; 10. Water diversion port; 11. Filter structure; 12. Drain pipe. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0044] Please refer to Figure 1-2 , the present invention provides a foundation waterproof early warning and infiltration drainage system and method, including:

[0045] A monitoring platform 1, a power supply module 5, a communication device 2, and a drainage module 3 arranged on the outer peripheral side of the foundation;

[0046] A multi-source information acquisition network 4 is arranged on the drainage module 3, and the multi-source information acquisition network 4 includes:

[0047] A soil parameter acquisition unit for real-time acquisition of the volumetric moisture content of the target soil layer;

[0048] An auxiliary parameter acquisition unit for synchronously monitoring soil temperature and conductivity to compensate for moisture content measurement errors;

[0049] An environmental parameter acquisition unit for monitoring the groundwater level and connecting meteorological data;

[0050] The multi-source information acquisition network 4 transmits the acquired multi-source data to the monitoring platform 1 through the communication device 2 in a fused manner;

[0051] The monitoring platform 1 includes a control module and an early warning module. The control module is built with a dynamic threshold model, which dynamically calculates the comprehensive moisture content alarm threshold based on the historical moisture content trend, the current soil layer state, as well as the groundwater level, precipitation intensity, and duration in the real-time environment. The control module issues an early warning signal to the early warning module for multi-level early warning based on the comprehensive moisture content alarm threshold, the measured moisture content value, and the preset early warning mechanism, and activates the drainage module 3 to perform corresponding drainage operations based on the early warning level.

[0052] Specifically, when applying the foundation waterproof early warning and infiltration drainage system, the drainage module is set at the position to be detected outside the foundation. The multi-source information acquisition network collects the soil parameters (volume moisture content), auxiliary parameters (soil temperature, conductivity), and environmental parameters (groundwater level, meteorological data) of the soil layer on the outer periphery of the foundation, and fuses and transmits the collected multi-source data to the monitoring platform through communication equipment. The dynamic threshold model in the control module dynamically calculates the comprehensive moisture content alarm threshold based on the historical moisture content trend, the current soil layer state, as well as the groundwater level, precipitation intensity, and duration in the real-time environment. Through multi-dimensional data cross-verification, single data errors are eliminated, and the moisture content detection accuracy and early warning accuracy are improved.

[0053] The power supply module 5 is used to provide a gradient DC voltage to the drainage module 3, facilitating the adoption of corresponding drainage modes with different soil moisture contents and improving the adaptability between modules.

[0054] Furthermore, the drainage module 3 further includes: a plurality of electrode arrays evenly distributed along the outer contour of the foundation. The electrode arrays are composed of vertically buried cathode conductive rods 6 and a plurality of anode conductive rods 7. The layout spacing between each anode conductive rod 7 and the cathode conductive rod 6 is equal, and they are connected to the positive and negative poles of the power supply module 5.

[0055] A plurality of anti-interference segments 8 are evenly distributed along the axial direction of the anode conductive rod 7 and the cathode conductive rod 6. The soil parameter acquisition unit, the auxiliary parameter acquisition unit, and the environmental parameter acquisition unit are embedded on each anti-interference segment 8, jointly constituting the multi-source information acquisition network 4.

[0056] A water collection unit is arranged at the bottom of the cathode conductive rod 6 for collecting and discharging the water collected by the electrode array.

[0057] It can be understood that by evenly distributing a plurality of electrode arrays along the outer contour of the foundation to form a continuous electric field, the electroosmotic drainage principle is utilized to improve the soil moisture migration efficiency around the foundation. By setting anti-interference segments on the electrode conductive rods, the sensors are isolated from the electrode electric field, ensuring the detection accuracy of each sensor. At the same time, by arranging the sensors along the electrode array, the multi-source information acquisition network is quickly formed, improving the layout efficiency.

[0058] In addition, by arranging a water collection unit at the bottom of the cathode conductive rod, it is equivalent to having a self - contained collection function, which improves the water collection speed. Specifically, the arrangement mode of the conductive rods can be plum - blossom - shaped, radial or mesh - shaped, etc.

[0059] Furthermore, the water collection unit includes a water collection bin 9 and a drain pipe 12. The water collection bin 9 is a cylindrical bin, the top of the cylindrical bin is an inward - concave frustum structure, the bottom of the frustum structure is fixedly connected to the cathode conductive rod 6, the conical surface of the frustum structure is evenly provided with water diversion ports 10, and a filtering structure 11 is arranged at the water diversion ports 10;

[0060] One end of the drain pipe 12 is fixedly arranged at the bottom of the water collection bin 9, and the other end is connected to an external suction device. A liquid level sensor is arranged in the water collection bin 9.

[0061] It should be explained that through the combination of the water diversion ports and the filtering structure arranged on the conical surface, particulate matters in the soil are effectively intercepted to avoid the blockage of the drain pipe, and the liquid level sensor monitors the liquid level in the water collection bin. When the liquid level reaches the set threshold, the suction device starts to suck and discharge the water collected in the water collection bin. Specifically, the inclination angle of the conical surface is 45 - 60°, the suction device is a water pump, electromagnetic valves are installed on each drain pipe 12, and the electromagnetic valves are independently controlled by a control module, which is convenient to adopt corresponding drainage modes according to different soil moisture contents.

[0062] Furthermore, the soil parameter acquisition unit includes a time - domain reflectometry sensor, and the time - domain reflectometry sensor is hermetically embedded on the anti - interference section 8.

[0063] Furthermore, the auxiliary parameter acquisition unit includes a thermistor and a four - electrode sensor embedded on the anti - interference section 8. The thermistor is used to collect the soil temperature, and the four - electrode sensor is used to collect the soil ion concentration.

[0064] Furthermore, the environmental parameter acquisition unit includes a piezoresistive water level sensor and a meteorological data interface embedded on the anti - interference section 8. The water level sensor intermittently collects the groundwater level according to a set frequency;

[0065] The meteorological data interface is connected to a regional meteorological station for real - time acquisition of future precipitation intensity prediction data.

[0066] Furthermore, the construction steps of the dynamic threshold model include:

[0067] Step 1: Continuously collect historical data for at least 12 months, including hourly records of soil moisture content, temperature, conductivity, groundwater level, and meteorological precipitation intensity;

[0068] Step 2: Analyze the fluctuation characteristics of historical water content data and its correlation with the groundwater level, and extract trend parameters;

[0069] Specifically, analyze the daily / weekly fluctuation patterns of water content through statistical methods, and extract key trend parameters, which at least include: the threshold of water content rising rate, the duration of the stable period, etc.; establish a correlation matrix between the groundwater level and water content to identify the impact of the groundwater level rising rate on the change of water content;

[0070] Step 3: Perform time series modeling on historical data through a long short-term memory neural network to predict the water content change curve in the future period;

[0071] Specifically, adopt a long short-term memory neural network (LSTM), input parameters such as historical water content sequences, real-time soil temperature, conductivity, groundwater level, and predicted values of precipitation in the next 6 hours, and then output the water content change curve within the next 12 hours through the model and mark the confidence interval. During training, an adversarial sample enhancement strategy can be introduced to simulate abnormal conditions such as extreme precipitation (e.g., heavy rain) and underground pipe bursts to improve the robustness of the model.

[0072] Step 4: Combine the real-time rising and falling rate of the groundwater level and the predicted value of meteorological precipitation to dynamically adjust the boundary of the water content alarm threshold.

[0073] Specifically, the method for dynamically adjusting the water content alarm threshold: set the initial alarm threshold according to the soil layer type (e.g., the threshold for sandy soil is 25%, and for clay is 35%); when the meteorological interface obtains a strong precipitation forecast in the next 3 hours (e.g., rainfall > 50 mm / h), temporarily lower the threshold by 20%-30% based on the predicted precipitation intensity; if the groundwater level is monitored to rise by more than 5 cm per hour, automatically increase the current threshold by 5%-10% according to the water level-water content correlation matrix to avoid false alarms caused by drainage lag. When the soil temperature is below 5°C, suspend the threshold adjustment due to freezing factors and directly adopt the reference threshold.

[0074] Furthermore, the data fusion method includes:

[0075] Filter the original data collected by the water level sensor and retain the stable data reflecting the real water level change;

[0076] Integrate the measurement results of the time domain reflectometer, conductivity sensor, and water level sensor, and obtain the comprehensive soil water content through weighted calculation.

[0077] Specifically, in the system initialization stage, the monitoring platform assigns initial weights to various sensors: the time domain reflectometer sensor is directly measuring soil water content and is given the highest weight of 50%; the conductivity sensor is inversing water content through ion concentration and is given a weight of 30%; the groundwater level sensor is relying on the water level-water content correlation model for calculation and is given a weight of 20%;

[0078] Environmental monitoring and weight correction method: When the soil temperature sensor detects that the temperature exceeds 35°C, the weight of the conductivity sensor is reduced to 20%, the weight of the time domain reflectometer sensor is increased to 55%, and the weight of the water level sensor is increased to 25%;

[0079] When the groundwater level sensor monitors that the water level rises by more than 5 cm per hour, the weight of the water level sensor is increased to 35%, and the weights of the time domain reflectometer sensor and the conductivity sensor are reduced by 7.5%;

[0080] When a heavy precipitation forecast for the next 2 hours (such as rainfall > 30 mm / h) is obtained through the meteorological interface, the weight of the time domain reflectometer sensor is temporarily increased by 10% to enhance the decision-making priority of direct measurement data.

[0081] Comprehensive moisture content θ 综合 The threshold calculation formula is:

[0082] θ 综合 = θ1×W1 + θ2×W2 + θ3×W3

[0083] Where θ1 is the measurement value of the time domain reflectometer sensor, θ2 is the conductivity measurement value, θ3 is the water level measurement value, and W1, W2, and W3 are the weights of the corresponding parameters respectively.

[0084] Furthermore, the configuration relationship between the warning mechanism of the warning module and the drainage module 3 is as follows:

[0085] When it is detected that the real-time moisture content exceeds the comprehensive moisture content threshold and the duration is ≥ 15 minutes, a primary warning is triggered, and the power supply module 5 is controlled to apply a first gradient voltage to the electrode array where the warning point is located for dot-shaped drainage;

[0086] When it is detected that the groundwater level rising rate > 5 cm / h and the soil comprehensive moisture content exceeds the threshold > 30%, a medium-level warning is triggered, and the power supply module 5 is controlled to apply a second gradient voltage to at least three groups of circumferentially adjacent electrode arrays where the warning point is located for regional drainage;

[0087] When it is detected that the rainstorm intensity > 50 mm / h and the groundwater level rising rate > 10 cm / h, an emergency warning is triggered, and the power supply module 5 is controlled to apply the maximum gradient voltage to all electrode arrays for comprehensive drainage.

[0088] Foundation waterproof warning and seepage drainage method, including the following steps:

[0089] S1. Arrange electrode arrays along the outer contour of the foundation. By setting up a multi-source information acquisition network 4 on the electrode arrays, the volumetric moisture content, soil temperature, conductivity, and groundwater level data of the target soil layer are collected in real time, and meteorological precipitation prediction information is obtained synchronously;

[0090] S2. Filter the raw data collected by the water level sensor, extract the effective water level change data, dynamically adjust the weights based on the sensor type accuracy and environmental conditions, and perform weighted fusion calculation on the water content related data;

[0091] S3. Analyze the correlation between historical water content data and the groundwater level, construct a time series prediction model, and dynamically correct the water content alarm threshold in combination with the real-time groundwater level change rate and the predicted precipitation intensity;

[0092] S4. Based on the corrected water content alarm threshold, the measured water content value and the preset early warning mechanism, send an early warning signal to the early warning module for multi-level early warning, and start the drainage module 3 to perform corresponding drainage operations based on the early warning level.

[0093] The above is only the preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, makes equivalent substitutions or changes, and should be covered by the protection scope of the present invention.

Claims

1. A foundation waterproof warning and seepage drainage system, characterized in that, It includes a monitoring platform (1), a power module (5), a communication device (2), and a drainage module (3) arranged on the outer periphery of the foundation; A multi-source information acquisition network (4) is arranged on the drainage module (3), and the multi-source information acquisition network (4) includes: A soil parameter acquisition unit for real-time acquisition of the volumetric water content of the target soil layer; An auxiliary parameter acquisition unit for synchronously monitoring soil temperature and conductivity to compensate for the moisture content measurement error; An environmental parameter acquisition unit for monitoring the groundwater level and connecting meteorological data; The multi-source information acquisition network (4) transmits the acquired multi-source data to the monitoring platform (1) through the communication device (2) in a fused manner; The monitoring platform (1) includes a control module and an early warning module. The control module is built-in with a dynamic threshold model. The dynamic threshold model dynamically calculates the comprehensive moisture content alarm threshold based on the historical moisture content trend and the current soil layer state, as well as the groundwater level, precipitation intensity, and duration in the real-time environment. The control module issues an early warning signal to the early warning module for multi-level early warning based on the comprehensive moisture content alarm threshold, the measured moisture content value, and a preset early warning mechanism, and starts the drainage module (3) to perform corresponding drainage operations based on the early warning level; The power module (5) is used to provide a gradient DC voltage to the drainage module (3).

2. The foundation waterproof warning and seepage drainage system according to claim 1, wherein The drainage module (3) further includes: a plurality of electrode arrays evenly distributed along the outer contour of the foundation. The electrode arrays are composed of vertically buried cathode conductive rods (6) and a plurality of anode conductive rods (7). The layout spacing between each anode conductive rod (7) and the cathode conductive rod (6) is equal, and they are connected to the positive and negative poles of the power module (5); A plurality of anti-interference segments (8) are evenly distributed along the axial direction of the anode conductive rod (7) and the cathode conductive rod (6). The soil parameter acquisition unit, the auxiliary parameter acquisition unit, and the environmental parameter acquisition unit are embedded on each anti-interference segment (8), jointly constituting the multi-source information acquisition network (4); A water collection unit is arranged at the bottom of the cathode conductive rod (6) for collecting and discharging the water collected by the electrode array.

3. The foundation waterproof warning and seepage drainage system according to claim 2, characterized in that, The water collection unit includes a water collection bin (9) and a drain pipe (12). The water collection bin (9) is a cylindrical bin. The top of the cylindrical bin is an inwardly concave frustum structure. The bottom of the frustum structure is fixedly connected to the cathode conductive rod (6). Water inlet openings (10) are evenly arranged on the conical surface of the frustum structure, and a filtering structure (11) is arranged at the water inlet openings (10); One end of the drain pipe (12) is fixedly arranged at the bottom of the water collection bin (9), and the other end is connected to an external suction device. A liquid level sensor is arranged in the water collection bin (9).

4. The foundation waterproof warning and seepage drainage system according to claim 2, characterized in that, The soil parameter acquisition unit includes a time domain reflectometer sensor, and the time domain reflectometer sensor is hermetically embedded on the anti-interference segment (8).

5. The foundation waterproof warning and seepage drainage system according to claim 2, characterized in that, The auxiliary parameter acquisition unit includes a thermistor and a four-electrode sensor embedded on the anti-interference segment (8). The thermistor is used for collecting the soil temperature, and the four-electrode sensor is used for collecting the soil ion concentration.

6. The foundation waterproof warning and seepage drainage system according to claim 2, characterized in that, The environmental parameter acquisition unit includes a piezoresistive water level sensor and a meteorological data interface embedded in the anti-interference section (8). The water level sensor intermittently collects the groundwater level at a set frequency. The meteorological data interface is connected to a regional meteorological station for real-time acquisition of future precipitation intensity prediction data.

7. The foundation waterproof warning and seepage drainage system according to claim 1, characterized in that, The dynamic threshold model is configured as follows: Continuously collect historical water content data, and simultaneously record the soil temperature, conductivity, groundwater level, and meteorological precipitation intensity at the corresponding time points. Analyze the fluctuation characteristics of historical water content data and its correlation with the groundwater level, and extract trend parameters. Perform time series modeling on historical data through a long short-term memory neural network to predict the water content change curve in the future period. Combine the real-time groundwater level rise and fall rate and the meteorological precipitation prediction value to dynamically adjust the boundary of the water content alarm threshold.

8. The foundation waterproof warning and seepage drainage system according to claim 1, characterized in that, The data fusion method includes: Filter the original data collected by the water level sensor and retain the stable data reflecting the real water level change. Integrate the measurement results of the time domain reflectometer, conductivity sensor, and water level sensor, and obtain the comprehensive soil water content through weighted calculation.

9. The foundation waterproof warning and seepage drainage system according to claim 3, characterized in that, The configuration relationship between the warning mechanism of the warning module and the drainage module (3) is: When it is detected that the real-time water content exceeds the comprehensive water content threshold and the duration is ≥ 15 minutes, trigger a primary warning, and control the power supply module (5) to apply a first gradient voltage to the electrode array where the warning point is located for dot-shaped drainage. When it is detected that the groundwater level rise rate > 5 cm / h and the comprehensive soil water content exceeds the threshold by > 30%, trigger a secondary warning, and control the power supply module (5) to apply a second gradient voltage to at least three groups of circumferentially adjacent electrode arrays where the warning point is located for regional drainage. When it is detected that the rainstorm intensity > 50 mm / h and the groundwater level rise rate > 10 cm / h, trigger an emergency warning, and control the power supply module (5) to apply the maximum gradient voltage to all electrode arrays for comprehensive drainage.

10. Foundation waterproof warning and seepage drainage method, characterized in that, For implementing the foundation waterproof warning and infiltration drainage system according to any one of claims 1 to 9, the method includes the following steps: S1. Layout the electrode array along the outer contour of the foundation. By setting a multi-source information acquisition network (4) on the electrode array, collect the volumetric water content, soil temperature, conductivity, and groundwater level data of the target soil layer in real time, and simultaneously obtain meteorological precipitation prediction information. S2. Filter the original data collected by the water level sensor, extract the effective water level change data, and dynamically adjust the weight based on the sensor type accuracy and environmental conditions to perform weighted fusion calculation on the water content-related data. S3. Analyze the correlation between historical water content data and the groundwater level, construct a time series prediction model, and dynamically correct the water content alarm threshold in combination with the real-time groundwater level change rate and precipitation intensity prediction value. S4. Based on the corrected water content alarm threshold, measured water content value, and preset warning mechanism, send a warning signal to the warning module for multi-level warning, and start the drainage module (3) to perform corresponding drainage operations based on the warning level.

Images