Real-time monitoring device and method for seepage of groundwater at different depths in sliding soil layer

Abstract

The present application belongs to the technical field of groundwater monitoring, and particularly relates to a device and method for real-time monitoring of groundwater seepage at different depths in a sliding soil layer, comprising: a retraction structure, which is communicatively connected with a controller; a cable is wound around the retraction structure, and a plurality of monitoring structures are arranged at equal intervals at the movable end of the cable; a sensing mechanism is arranged at the top end of an upper fixing seat of the monitoring structure; a plurality of filamentous flow velocity and direction sensing structures are arranged at equal intervals on the outer side of a connecting cylinder; the outer sides of the upper and lower fixing seats are respectively vertically and slidably connected with sealing cylinders, and a driving mechanism is drivingly connected with the two sealing cylinders; and a sampling mechanism is arranged at the bottom end of the inner side of the lower fixing seat.

Description

Technical Field

[0001] This invention belongs to the field of groundwater monitoring technology, and in particular relates to a device and method for real-time monitoring of groundwater seepage at different depths in landslide soil layers. Background Technology

[0002] A landslide occurs when a portion of a mountain slope slides along a sliding surface. It is typically caused by erosion of the soil and rock at the base of the mountain or by gravity. Landslides are more likely to occur in geological environments where the soil layer is widely distributed. The direction and velocity of groundwater flow can influence the formation of the sliding surface and the sliding speed. There is a close relationship between landslides and groundwater flow direction and velocity. Groundwater flowing through soil and rock erodes and dissolves them. Particularly within mountains, groundwater can significantly impact the stability of the mountain. If the groundwater flow is too rapid, it can lead to erosion of the soil and rock at the base of the mountain, thereby reducing its stability.

[0003] To protect lives and property, landslide prevention and control measures are necessary. Installing groundwater seepage monitoring equipment is an effective way to provide early warnings of landslides. This equipment monitors soil moisture and groundwater flow, thereby predicting the stability of the mountain. Abnormal groundwater flow indicates a potential landslide.

[0004] Landslide early warning also involves monitoring groundwater at different depths. By monitoring groundwater at different depths, we can understand its distribution, including its location, flow direction, and flow velocity. This helps assess the stability of the mountain and predict potential landslide risks, enabling people to take appropriate countermeasures. Summary of the Invention

[0005] The purpose of this invention is to provide a real-time monitoring device for groundwater seepage at different depths in a landslide soil layer, so as to solve the above-mentioned problems and achieve the goal of real-time and accurate monitoring of groundwater at different depths.

[0006] To achieve the above objectives, the present invention provides the following solution:

[0007] A real-time monitoring device for groundwater seepage at different depths in a sliding soil layer includes: a retractable structure, the retractable structure being communicatively connected to a controller, a cable being wound around the retractable structure, and a plurality of monitoring structures being equally spaced at the movable end of the cable.

[0008] The monitoring structure includes a vertically arranged connecting cylinder, with an upper fixed seat shafted to the top of the connecting cylinder. The top of the upper fixed seat is fixedly connected to the cable. A lower fixed seat shafted to the bottom of the connecting cylinder is fixedly connected to the cable. The diameters of the upper and lower fixed seats are larger than the diameter of the connecting cylinder. Adjacent upper and lower fixed seats are fixedly connected by cables.

[0009] A sensing mechanism is provided at the top of the upper fixed base;

[0010] The connecting cylinder is provided with a plurality of filamentous flow velocity and direction sensing structures at equal intervals around its outer circumference. The plurality of filamentous flow velocity and direction sensing structures are parallel to the axis of the connecting cylinder. The two ends of the filamentous flow velocity and direction sensing structures are respectively fixedly connected to the upper fixed seat and the lower fixed seat.

[0011] The upper fixed seat and the lower fixed seat are respectively vertically slidably connected to the outer side of the sealing cylinder, and the inner side of the connecting cylinder is provided with a driving mechanism, which is connected to the two sealing cylinders in a transmission manner.

[0012] A sampling mechanism is provided at the bottom inner side of the lower fixed base;

[0013] A control mechanism is fixedly connected to the top inner side of the upper fixed base. The sensing mechanism, the filamentous flow velocity and direction sensing structure, and the driving mechanism are electrically connected to the control mechanism. The control mechanism is communicatively connected to the controller.

[0014] Preferably, the sensing mechanism includes a pH value sensing head and a pressure sensing head, which are respectively fixedly connected to the top of the upper fixed base, and are electrically connected to the control mechanism.

[0015] Preferably, the filamentous flow velocity and direction sensing structure includes a flow velocity sensing filament, which is parallel to the axis of the connecting cylinder. Tension sensing elements are fixedly connected to both ends of the flow velocity sensing filament, and the two tension sensing elements are fixedly connected to the upper fixed seat and the lower fixed seat, respectively. The tension sensing elements are electrically connected to the control mechanism.

[0016] Preferably, the driving mechanism includes two telescopic motors arranged opposite each other. The telescopic motors are vertically arranged and fixedly connected to the inner side of the connecting cylinder. The output shaft of the telescopic motor is axially connected to a connecting frame. The upper fixed seat and the lower fixed seat are both provided with sliding grooves on their periphery. The two connecting frames are slidably connected to the corresponding sliding grooves. The periphery of the two connecting frames is fixedly connected to the inner side of the corresponding sealing cylinder.

[0017] Preferably, the sampling mechanism includes a liquid storage chamber located inside the lower fixed base, a connecting pipe fixedly connected to the top of the liquid storage chamber, an inlet located at the top of the lower fixed base of the connecting pipe, the connecting pipe being fixedly connected to the inlet, and an outlet located on the side of the lower fixed base being fixedly connected to the liquid storage chamber.

[0018] Preferably, the control mechanism includes a battery and a data processing chip electrically connected to the battery. The battery and the data processing chip are fixedly connected to the inner side of the upper fixed base. The sensing mechanism, the filamentary flow velocity and direction sensing structure, and the driving mechanism are electrically connected to the data processing chip.

[0019] Preferably, the take-up and release structure includes a roller base, a cable take-up and release roller is rotatably connected to the top of the roller base, a control motor is shaft-connected to one side of the cable take-up and release roller, the control motor is fixedly connected to the roller base, the cable is wound around the cable take-up and release roller, and a communicator is fixedly connected to the roller base, the communicator is communicatively connected to the controller.

[0020] The method of using a real-time monitoring device for groundwater seepage at different depths in a landslide soil layer is as follows:

[0021] S1. Determine the location to be monitored and drill to a suitable depth;

[0022] S2. Deploy several monitoring structures into the borehole to the specified depth;

[0023] S3. The controller controls the drive mechanism to move the two sealing cylinders away from each other, so that the filamentous flow velocity flows towards the sensing structure and contacts the water.

[0024] S4. Monitor the tensile data of multiple filamentous flow velocity and direction sensing structures;

[0025] S5. Use a BP neural network model to monitor the flow velocity and flow rate of groundwater at different depths in real time.

[0026] Preferably, step S5 includes:

[0027] S5.1 Define network structure, data collection, and preprocessing;

[0028] S5.2 Create a dataset to build a recurrent neural network model;

[0029] S5.3 Create a model, define the loss function and optimizer, and reduce errors caused by complex underground environments;

[0030] S5.4 Define the loss function and optimizer to train the model;

[0031] S5.5 The tensile data of multiple filamentary flow velocity and direction sensing structures monitored in real time are imported into the training model for evaluation;

[0032] S5.6. Import the tensile data of multiple filamentary flow velocity and direction sensing structures monitored in real time into the model for evaluation and adjust the model according to the error.

[0033] S5.7. The groundwater flow direction and velocity are calculated using the tensile data from multiple filamentary flow velocity and direction sensing structures monitored in real time. Compared with the prior art, the present invention has the following advantages and technical effects:

[0034] By setting up a sensing mechanism, it is possible to monitor the presence of groundwater, the pH value of groundwater, pressure, etc. By setting up a sampling mechanism, it is also possible to sample water bodies at different depths. By setting up two sealing cylinders that are connected to the drive mechanism, the drive mechanism can drive the two sealing cylinders to fit tightly together, thereby protecting the filamentous flow velocity to the sensing structure, and making it smoother when pulling the entire monitoring structure up and down.

[0035] By setting up filamentary flow velocity and direction sensing structures, when water flows past several of these structures, the connecting cylinder obstructs the flow, causing the filamentary flow velocity and direction sensing structures at different locations around the connecting cylinder to be impacted by water flow of varying strengths. The resulting data changes are then analyzed using a BP neural network model to determine the flow direction and velocity of the groundwater at the current location. Compared to traditional image-based monitoring, this monitoring method can achieve stable and accurate monitoring results. It is also suitable for relatively turbid water quality, offering higher stability. Furthermore, by transmitting only floating-point data, this device saves more power compared to directly transmitting images, making it suitable for long-term, real-time monitoring.

[0036] Using these structures, a real-time monitoring device for groundwater seepage at different depths in landslide soil layers was developed, which is suitable for real-time monitoring of groundwater conditions at different depths. Attached Figure Description

[0037] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly described below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0038] Figure 1 This is a schematic diagram of the structure of the present invention;

[0039] Figure 2 This is a schematic diagram illustrating the usage status of the monitoring structure.

[0040] Figure 3 This is a schematic diagram of a partial cross-sectional structure of the monitoring structure;

[0041] Figure 4 This is a frontal cross-sectional view of the monitored structure;

[0042] Reference numerals: 1. Controller; 2. Retraction / Extraction Structure; 3. Monitoring Structure; 201. Cable Retraction / Extraction Roller; 202. Control Motor; 203. Communicator; 204. Cable; 301. Upper Fixing Base; 302. Sealing Cylinder; 303. Connecting Cylinder; 304. Flow Direction and Velocity Sensing Filament; 305. Lower Fixing Base; 306. Liquid Storage Tank; 307. Connecting Pipe; 308. Tension Sensing Element; 309. Connecting Frame; 310. Slide Groove; 311. pH Value Sensing Head; 312. Pressure Sensing Head; 313. Battery; 314. Data Processing Chip; 315. Liquid Outlet; 316. Telescopic Motor. Detailed Implementation

[0043] 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.

[0044] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0045] Reference Figures 1-4 As shown, the present invention provides a real-time monitoring device for groundwater seepage at different depths in a sliding soil layer, comprising: a retractable structure 2, the retractable structure 2 being communicatively connected to a controller 1, a cable 204 being wound around the retractable structure 2, and a plurality of monitoring structures 3 being provided at equal intervals at the movable end of the cable 204.

[0046] The monitoring structure 3 includes a vertically arranged connecting cylinder 303. The top end of the connecting cylinder 303 is axially connected to an upper fixed seat 301, and the bottom end of the connecting cylinder 303 is axially connected to a lower fixed seat 305. The diameters of the upper fixed seat 301 and the lower fixed seat 305 are larger than the diameter of the connecting cylinder 303. Adjacent upper fixed seats 301 and lower fixed seats 305 are fixedly connected by a cable 204.

[0047] A sensing mechanism is provided at the top of the upper fixing base 301;

[0048] The outer side of the connecting cylinder 303 is provided with several filamentous flow velocity and direction sensing structures at equal intervals. The several filamentous flow velocity and direction sensing structures are parallel to the axis of the connecting cylinder 303. The two ends of the filamentous flow velocity and direction sensing structures are fixedly connected to the upper fixed seat 301 and the lower fixed seat 305, respectively.

[0049] The upper fixed seat 301 and the lower fixed seat 305 are vertically and slidably connected to the outer sides of the sealing cylinder 302, and the inner side of the connecting cylinder 303 is provided with a driving mechanism, which is connected to the two sealing cylinders 302 in a transmission manner.

[0050] A sampling mechanism is provided at the bottom inner side of the lower fixed base 305;

[0051] A control mechanism is fixedly connected to the top inner side of the upper fixed base 301. The sensing mechanism, the filamentous flow velocity and direction sensing structure, the drive mechanism and the control mechanism are electrically connected. The control mechanism is communicatively connected to the controller 1.

[0052] By setting up a sensing mechanism, it is possible to monitor the presence of groundwater, the pH value of groundwater, pressure, etc. By setting up a sampling mechanism, it is also possible to sample water bodies at different depths. By setting up two sealing cylinders 302 that are connected to the drive mechanism, the drive mechanism can drive the two sealing cylinders 302 to fit tightly together, thereby protecting the filamentous flow velocity towards the sensing structure, and making it smoother when pulling the entire monitoring structure 3 up and down.

[0053] By setting up filamentary flow velocity and direction sensing structures, when water flows past several of these structures, the connecting cylinder 303 acts as a barrier, causing the filamentary flow velocity and direction sensing structures at different positions around the connecting cylinder 303 to be impacted by water flow of varying strengths. The resulting data changes are then analyzed using a BP neural network model to determine the flow direction and velocity of the groundwater at the current location. Compared to traditional image-based monitoring, this monitoring method can achieve stable and accurate monitoring results. It is also suitable for relatively turbid water quality, offering higher stability. Furthermore, by transmitting only floating-point data, this device saves more power than directly transmitting images, making it suitable for long-term, real-time monitoring.

[0054] The solution is further optimized so that the sensing mechanism includes a pH value sensing head 311 and a pressure sensing head 312. The pH value sensing head 311 and the pressure sensing head 312 are fixedly connected to the top of the upper fixed base 301, and the pH value sensing head 311 and the pressure sensing head 312 are electrically connected to the control mechanism.

[0055] The pH sensor 311 is used to monitor the pH value of groundwater, and the pressure sensor 312 is used to monitor water pressure.

[0056] The scheme is further optimized. The filamentous flow velocity and direction sensing structure includes a flow velocity sensing filament 304, which is parallel to the axis of the connecting cylinder 303. Tension sensing elements 308 are fixedly connected to both ends of the flow velocity sensing filament 304. The two tension sensing elements 308 are fixedly connected to the upper fixed seat 301 and the lower fixed seat 305, respectively. The tension sensing elements 308 are electrically connected to the control mechanism.

[0057] When groundwater impacts the flow velocity sensing filament 304 in one direction, the filaments 304 located around the connecting cylinder 303 will generate different readings due to the impact of the water flow. The filaments 304 located on the sides of the connecting cylinder 303 facing the water flow direction will experience a greater impact and more pronounced fluctuations in their readings, while the filaments 304 blocked by the connecting cylinder 303 will have smaller readings. By observing these changes in values, a BP neural network model can accurately determine the direction and velocity of the water flow at the current location.

[0058] The scheme is further optimized. The drive mechanism includes two telescopic motors 316 arranged opposite each other. The telescopic motors 316 are vertically arranged and fixedly connected to the inner side of the connecting cylinder 303. The output shaft of the telescopic motor 316 is axially connected to the connecting frame 309. The upper fixed seat 301 and the lower fixed seat 305 are both provided with sliding grooves 310 on their periphery. The two connecting frames 309 are slidably connected to the corresponding sliding grooves 310. The periphery of the two connecting frames 309 is fixedly connected to the inner side of the corresponding sealing cylinder 302.

[0059] During deployment, the two sealed cylinders 302 are tightly pressed together, protecting the tension sensing element 308 connected to the flow direction and velocity sensing filament 304 on the inner side, preventing damage to the tension sensing element 308. The tension sensing element 308 uses a miniature electronic force gauge. If the flow direction and velocity sensing filament 304 is directly exposed during deployment, it may damage the tension sensing element 308, compromising the accuracy of the measurement.

[0060] The sampling mechanism is further optimized by including a liquid storage chamber 306 located inside the lower fixed base 305. A connecting pipe 307 is fixedly connected to the top of the liquid storage chamber 306. An inlet is provided at the top of the lower fixed base 305 for the connecting pipe 307. The connecting pipe 307 is fixedly connected to the inlet. An outlet 315 is provided on the side of the lower fixed base 305. The outlet 315 is fixedly connected to the liquid storage chamber 306.

[0061] The scheme is further optimized. The control mechanism includes a battery 313 and a data processing chip 314 electrically connected to the battery 313. The battery 313, the data processing chip 314 and the upper fixed base 301 are fixedly connected to the inside. The sensing mechanism, the filamentary flow velocity and direction sensing structure, the drive mechanism and the data processing chip 314 are electrically connected.

[0062] The battery 313 supplies power to the entire device, while the data processing chip 314 is equipped with a data processing unit and a communication unit. The data processing unit organizes and summarizes the read data in floating-point mode, packages the data within a certain period of time, and sends it to the communication unit. The communication unit then communicates with the external controller 1 at fixed time intervals, so that the data can be stably and accurately transmitted to the controller 1.

[0063] The scheme is further optimized. The take-up and release structure 2 includes a roller seat. A cable take-up and release roller 201 is rotatably connected to the top of the roller seat. A control motor 202 is shaft-connected to one side of the cable take-up and release roller 201. The control motor 202 is fixedly connected to the roller seat. The cable 204 is wound around the cable take-up and release roller 201. A communicator 203 is fixedly connected to the roller seat. The communicator 203 is communicatively connected to the controller 1.

[0064] The communicator 203 is configured so that the controller 1 can easily control the operation of the motor 202.

[0065] The method of using a real-time monitoring device for groundwater seepage at different depths in a landslide soil layer is as follows:

[0066] S1. Determine the location to be monitored and drill to a suitable depth;

[0067] S2. Deploy several monitoring structures 3 into the borehole to the specified depth;

[0068] S3. The controller 1 controls the drive mechanism to move the two sealing cylinders 302 away from each other, so that the filamentous flow velocity flows towards the sensing structure and contacts the water.

[0069] S4. Monitor the tensile data of multiple filamentous flow velocity and direction sensing structures;

[0070] S5. Use a BP neural network model to monitor the flow velocity and flow rate of groundwater at different depths in real time.

[0071] Further optimization of the solution, the S5 steps include:

[0072] S5.1 Define the network structure;

[0073] Layer 1 (fc1): This is the part of the network from the input layer to the hidden layer. The input size of this layer is 1. Several tension sensing elements 308 are paired and formed into several arrays F in the form of (Index, Timescale, A, B). Here, Index represents the current sequence number of the flow direction and velocity sensing filament 304, used to distinguish different flow direction and velocity sensing filaments 304. Timescale represents the time scale. A represents the reading of the tension sensing element 308 located above the same flow direction and velocity sensing filament 304, and B represents the reading of the tension sensing element 308 located below the same flow direction and velocity sensing filament 304. Several arrays F (the number of arrays F is equal to the number of flow direction and velocity sensing filaments 304) form a matrix M. Matrix M serves as the input feature to calculate the force on the flow direction and velocity sensing filament 304 at different positions under the same time scale and the rate of change of tension value under adjacent time scales, thereby calculating the water flow direction and velocity.

[0074] The matrix M has 10 output nodes, which is the number of nodes in the hidden layer. The size of the weight matrix for this layer is 10x1, and the size of the bias vector is 10x1. The ReLU activation function sets all negative outputs to zero during the forward propagation process and retains all positive outputs.

[0075] The second layer (fc2): This is the part of the network from the hidden layer to the output layer. The input size of this layer is 10 because our hidden layer has 10 nodes. It has 2 output nodes because we are predicting two features, the water flow direction and velocity. The size of the weight matrix for this layer is 2x10, and the size of the bias vector is 2x1. The ReLU activation function sets all negative outputs to zero during the forward propagation process and retains all positive outputs.

[0076] S5.2, Create a dataset;

[0077] Specifically, it includes data collection: Collect the tensile data of the tensile sensing element 308, calculate the actual water flow impact force data using the formula, and statistically organize the tensile data of multiple groups of tensile sensing elements 308 at the same time into several arrays F, and form a matrix M from this array F. These data can be collected by a detector.

[0078] Data preprocessing: Perform data preprocessing, including standardization, normalization, or outlier removal operations. After processing the matrix M, a more standardized matrix N is obtained.

[0079] Create a dataset: Divide the sample data obtained after testing into input data and output data. The input data is the water flow impact force received by several flow velocity and direction sensing filaments 304, and the output data is the water flow direction and velocity, which are known data. Convert these data into PyTorch tensors and then input them into the BP neural network for training.

[0080] S5.3, Create a model;

[0081] S5.4, Define the loss function and optimizer; <00><000>[

[0082] In the neural network model for water flow velocity and direction, the loss function and optimizer are both very important parts, which are responsible for measuring the performance of the model and adjusting the parameters of the model respectively:

[0083] Specifically, for each observation (i.e., a data point), the loss function calculates the difference between the model's predicted flow direction and velocity and the actual flow. This difference is probability-based (cross-entropy loss), a function used for binary classification problems. In this recurrent backpropagation (BP) neural model, it is used to evaluate the difference between the model's predicted probabilities and the actual labels. This function strives to make the predicted probabilities as close as possible to the actual labels, thereby optimizing the model's performance. Ideally, the lower the value of the loss function, the more accurate the model's predictions.

[0084] Optimizer: In a neural network, the optimizer adjusts model parameters, including weights and biases, to minimize the loss function. The optimizer determines how to change the parameters based on the current loss function value of the model, aiming to reduce the loss function as much as possible. This recurrent model uses a stochastic gradient descent (SGD) optimizer.

[0085] The basic idea of ​​stochastic gradient descent (SGD) is to randomly select a sample in each iteration to calculate the gradient of the loss function (that is, the derivative of the loss function with respect to the model parameters), and then update the parameters according to this gradient.

[0086] Specifically, training this recurrent backpropagation (BP) neural network using the stochastic gradient descent algorithm requires the following steps:

[0087] Initialization: Initialize some values ​​for the weights and biases of the neural network. These values ​​are usually chosen randomly to break symmetry and help the network learn.

[0088] Forward propagation: Using these initial parameter values, the network is used to compute its predicted output for each input. This involves a series of mathematical operations performed according to the network's structure and parameters (weights and biases).

[0089] Loss calculation: After obtaining the predicted output, calculate the cross-entropy loss of the model, which is the difference between the predicted output and the actual output.

[0090] Backpropagation: Calculates the gradient of the loss function with respect to each parameter. This is accomplished using the backpropagation algorithm, which starts from the output layer and progressively calculates the gradients back towards the input layer.

[0091] Update parameters: Use these gradients to update the network's parameters. This is done by subtracting the learning rate multiplied by the gradient. The learning rate is a hyperparameter that determines the step size for parameter updates. In SGD, parameters are updated using only one sample at a time, instead of the entire dataset.

[0092] Iteration: The above steps are repeated on a single data batch until a preset number of iterations is reached, or other stopping criteria are met. In each iteration, the network parameters are updated using a new data batch.

[0093] S5.5 Training the Model: When training the model, the above steps can be repeated based on existing data to correct errors and biases, resulting in a relatively better model and improving the overall accuracy. Alternatively, fluid simulation software can be used for simulation and prediction, and the simulation results can be imported into the model for training.

[0094] S5.6. Import the tensile data of multiple filamentary flow velocity and direction sensing structures monitored in real time into the model for evaluation;

[0095] S5.7 Calculate the groundwater flow direction and velocity using the tensile data from multiple filamentous flow velocity and direction sensing structures monitored in real time.

[0096] In the description of this invention, it should be understood that the terms "longitudinal", "lateral", "up", "down", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing this invention, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this invention.

[0097] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made by those skilled in the art to the technical solutions of the present invention without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A real-time monitoring device for groundwater seepage at different depths in a landslide soil layer, characterized in that, include: The retractable structure (2) is communicatively connected to a controller (1). A cable (204) is wound around the retractable structure (2). Several monitoring structures (3) are provided at equal intervals at the movable end of the cable (204). The monitoring structure (3) includes a vertically arranged connecting cylinder (303), with an upper fixed seat (301) axially connected to the top of the connecting cylinder (303) and a lower fixed seat (305) axially connected to the bottom of the connecting cylinder (303). The diameters of the upper fixed seat (301) and the lower fixed seat (305) are larger than the diameter of the connecting cylinder (303). Adjacent upper fixed seats (301) and lower fixed seats (305) are fixedly connected by a cable (204). The upper fixing base (301) is provided with a sensing mechanism at its top; The connecting cylinder (303) is provided with a plurality of filamentous flow velocity and direction sensing structures at equal intervals around its outer circumference. The plurality of filamentous flow velocity and direction sensing structures are parallel to the axis of the connecting cylinder (303). The two ends of the filamentous flow velocity and direction sensing structures are fixedly connected to the upper fixed seat (301) and the lower fixed seat (305) respectively. The upper fixed seat (301) and the lower fixed seat (305) are respectively vertically slidably connected to sealing cylinders (302). The inner side of the connecting cylinder (303) is provided with a driving mechanism, which is connected to the two sealing cylinders (302) in a transmission manner. A sampling mechanism is provided on the inner bottom end of the lower fixing seat (305); The upper fixed base (301) is fixedly connected to the top of the inner side of the control mechanism. The sensing mechanism, the filamentous flow velocity and direction sensing structure, and the driving mechanism are electrically connected to the control mechanism. The control mechanism is communicatively connected to the controller (1). The take-up and release structure (2) includes a roller seat, and a cable take-up and release roller (201) is rotatably connected to the top of the roller seat. A control motor (202) is shaft-connected to one side of the cable take-up and release roller (201). The control motor (202) is fixedly connected to the roller seat. The cable (204) is wound around the cable take-up and release roller (201). A communicator (203) is fixedly connected to the roller seat. The communicator (203) is communicatively connected to the controller (1).

2. The real-time monitoring device for groundwater seepage at different depths in a landslide soil layer according to claim 1, characterized in that, The sensing mechanism includes a pH value sensing head (311) and a pressure sensing head (312). The pH value sensing head (311) and the pressure sensing head (312) are respectively fixedly connected to the top of the upper fixed base (301). The pH value sensing head (311) and the pressure sensing head (312) are electrically connected to the control mechanism.

3. The real-time monitoring device for groundwater seepage at different depths in a landslide soil layer according to claim 1, characterized in that, The filamentary flow velocity and direction sensing structure includes a flow velocity sensing filament (304), which is parallel to the axis of the connecting cylinder (303). Tension sensing elements (308) are fixedly connected to both ends of the flow velocity sensing filament (304). The two tension sensing elements (308) are fixedly connected to the upper fixed seat (301) and the lower fixed seat (305) respectively. The tension sensing elements (308) are electrically connected to the control mechanism.

4. The real-time monitoring device for groundwater seepage at different depths in a landslide soil layer according to claim 1, characterized in that, The driving mechanism includes two telescopic motors (316) arranged opposite to each other. The telescopic motors (316) are arranged vertically and are fixedly connected to the inner side of the connecting cylinder (303). The output shaft of the telescopic motors (316) is axially connected to the connecting frame (309). The upper fixed seat (301) and the lower fixed seat (305) are both provided with sliding grooves (310) on their periphery. The two connecting frames (309) are slidably connected to the corresponding sliding grooves (310). The periphery of the two connecting frames (309) is fixedly connected to the inner side of the corresponding sealing cylinder (302).

5. The real-time monitoring device for groundwater seepage at different depths in a landslide soil layer according to claim 1, characterized in that, The sampling mechanism includes a liquid storage chamber (306) located inside the lower fixed base (305). A connecting pipe (307) is fixedly connected to the top of the liquid storage chamber (306). An inlet is provided at the top of the lower fixed base (305). The connecting pipe (307) is fixedly connected to the inlet. An outlet (315) is provided on the side of the lower fixed base (305). The outlet (315) is fixedly connected to the liquid storage chamber (306).

6. The real-time monitoring device for groundwater seepage at different depths in a landslide soil layer according to claim 1, characterized in that, The control mechanism includes a battery (313) and a data processing chip (314) electrically connected to the battery (313). The battery (313) and the data processing chip (314) are fixedly connected to the inner side of the upper fixed base (301). The sensing mechanism, the filamentary flow velocity and direction sensing structure, and the driving mechanism are electrically connected to the data processing chip (314).

7. The method of using the real-time monitoring device for groundwater seepage at different depths in a landslide soil layer according to any one of claims 1-6 is as follows: S1. Determine the location to be monitored and drill to a suitable depth; S2. Deploy several monitoring structures (3) into the borehole to the specified depth; S3. The controller (1) controls the drive mechanism to drive the two sealing cylinders (302) away from each other, so that the filamentous flow velocity flows to the sensing structure and contacts the water. S4. Monitor the tensile data of multiple filamentous flow velocity and direction sensing structures; S5. Use a BP neural network model to monitor the flow velocity and flow rate of groundwater at different depths in real time.

8. The method of using the real-time monitoring device for groundwater seepage at different depths in a landslide soil layer according to claim 7, characterized in that, Step S5 includes: S5.1 Define the network structure; S5.2, Create a dataset; S5.3, Create the model; S5.4 Define the loss function and optimizer; S5.5, Training the model; S5.

6. Import the tensile data of multiple filamentary flow velocity and direction sensing structures monitored in real time into the model for evaluation; S5.7 Calculate the groundwater flow direction and velocity using the tensile data from multiple filamentous flow velocity and direction sensing structures monitored in real time.

Images