Geological monitoring and early warning method and system for geological investigation

Abstract

The invention discloses a geological monitoring and early warning method and system for geological exploration, and relates to the technical field of land exploration, and the method comprises the steps: monitoring the displacement at different depths in a sliding rock-soil body; carrying out real-time data acquisition on factors influencing the displacement of the rock-soil body; preprocessing the deep displacement time sequence data of the unstable slope slip plane; decomposing the displacement time sequence into a trend sample and a fluctuation sample by using a function fitting method; predicting a displacement time sequence trend sample by using an artificial neural network model; predicting a displacement time sequence fluctuation sample by using the long-short time memory network model; obtaining a displacement time sequence predicted value through a superposition summation method; and judging the deformation stage of the unstable slope according to the displacement time sequence predicted value, and outputting a corresponding early warning signal. By arranging the data monitoring module and the data prediction module, advanced perception of slope deformation trend development is realized in real time.

Description

Technical Field

[0001] The present invention relates to the field of land exploration technology, and in particular to a geological monitoring and early warning method and system for geological exploration. Background Art

[0003] Landslide is one of the common geological disasters. Under the influence of external factors such as rainfall or groundwater uplift, the rock and soil on the slope suffers overall shear failure due to excessive self-weight and weakened shear strength, causing the rock and soil on the slope to become unstable from top to bottom along the through slip surface. At present, a large number of studies have been conducted at home and abroad on the monitoring and early warning of slope landslide disasters. The selection of early warning models and the determination of early warning index thresholds have always been the focus of research. Due to different research backgrounds, there is no common standard for the selection of early warning index thresholds when using similar early warning models for research. Basically, they are selected empirically using statistical methods combined with engineering practice. Summary of the invention

[0004] In order to solve the above technical problems, a geological monitoring and early warning method and system for geological survey are provided. This technical solution solves the problems raised in the above background technology.

[0005] In order to achieve the above purpose, the technical solution adopted by the present invention is:

[0006] A geological monitoring and early warning method for geological survey, comprising:

[0007] The unstable slopes are surveyed by installing a fixed pulley inclinometer to monitor the displacement at different depths in the sliding rock and soil.

[0008] By installing at least one data monitoring device, real-time data collection is performed on factors affecting the displacement of the rock and soil mass, wherein the factors include soil water pressure, soil moisture content and rainfall;

[0009] The deep displacement time series data of the unstable slope slip surface are preprocessed to generate samples with equal time intervals;

[0010] The displacement time series is decomposed into trend samples and fluctuation samples using the function fitting method;

[0011] Based on the influence of factors affecting the displacement of rock and soil mass on the displacement time series trend samples, the artificial neural network model is used for prediction;

[0012] Based on the influence of factors affecting the displacement of rock and soil mass on the fluctuation samples of displacement time series, the long and short time memory network model is used for prediction.

[0013] The predicted values ​​of the fluctuation samples and trend samples are respectively superimposed and summed to obtain the predicted value of the displacement time series;

[0014] The deformation stage of the unstable slope is determined by the displacement time series prediction value, and a corresponding warning signal is output. The deformation stage includes the undeformed stage, the initial deformation stage, the uniform deformation stage and the accelerated deformation stage.

[0015] Preferably, the step of installing a pulley-type fixed inclinometer to perform geological survey on the unstable slope and monitoring the displacement at different depths in the sliding rock and soil body specifically includes:

[0016] A vertical guide groove is arranged in the inclinometer tube, and at least two sensors are connected in series at the same interval in the guide groove. The sensor with an accelerometer controls its offset direction in the guide groove through a guide wheel;

[0017] When the slope slides, the rock mass drives the inclinometer tube fixed at the bottom on the bedrock to deviate. At the same time, due to the effect of the gravity field, the sensor with accelerometer can obtain the gravitational components in the horizontal and vertical directions.

[0018] The deflection angle of the sensor in the inclinometer casing is calculated by the gravity components in the horizontal and vertical directions;

[0019] The horizontal displacement between the two sensors is calculated using the inclinometer formula;

[0020] The inclinometer formula is:

[0021]

[0022] Where S is the horizontal displacement between the two sensors, H is the depth difference between the two sensors, θ is the deflection angle between the two sensors, and α is the initial angle value when the fixed inclinometer is installed.

[0023] Preferably, the real-time data collection of factors affecting the displacement of the rock and soil mass by installing at least one data monitoring device specifically includes:

[0024] The soil water pressure is monitored in real time through the pore water pressure gauge. The pressure water acts on the pressure-bearing membrane through the permeable stone at the bottom of the sensor. The pressure-bearing membrane deforms and generates stress, which is automatically converted into pressure through the internal sensor.

[0025] The soil moisture content is monitored in real time through a soil moisture meter, and the change of moisture content is reflected by measuring the dielectric constant of the soil;

[0026] The rainfall is monitored in real time through a tipping bucket rain gauge. As the rainfall continues, the volume of water in the tipping bucket increases. When it reaches the preset height, the tipping bucket loses balance due to its excessive weight and flips left to right. The sensor records the rainfall value by recording the number of times the tipping bucket pendulum swings.

[0027] Preferably, the prediction using an artificial neural network model based on the influence of factors affecting the displacement of the rock and soil mass on the displacement time series trend samples specifically includes:

[0028] S101: Determine the number of layers of the artificial neural network model and the number of neurons in each layer;

[0029] S102: Select a continuous and smooth S-shaped growth curve function as the neuron activation function;

[0030] S103: Randomly initialize the weights and biases of the network;

[0031] S104: forward propagating the input data through the network to obtain a predicted output;

[0032] S105: Calculate the loss function value according to the difference between the predicted output and the actual data;

[0033] S106: Update the weights and biases of the network according to the gradient of the loss function;

[0034] S107: Determine whether the preset number of iterations has been reached. If yes, output the current prediction value as the prediction value of the fluctuation sample. If no, return to step S104.

[0035] The S-shaped growth curve function is:

[0036]

[0037] Where f(x) is the S-shaped growth curve function, and x is the input data of the neuron.

[0038] Preferably, the prediction using the long-short time memory network model based on the influence of the factors affecting the displacement of the rock and soil mass on the displacement time series fluctuation samples specifically includes:

[0039] According to the gated recurrent unit formula, the input gate, forget gate and output gate functions are calculated respectively;

[0040] Construct a long-short time memory network model based on input gate, forget gate and output gate functions;

[0041] The factors affecting the displacement of the rock and soil mass and the fluctuation samples of the displacement time series are input into the long and short time memory network model to obtain the predicted value of the trend sample;

[0042] The gated recurrent unit formula is:

[0043]

[0044] In the formula, α t , β t , γt are the input gate, forget gate and output gate functions respectively, W α , W β , W γ are the parameter matrices of the input gate, forget gate, and output gate from time t-1 to time t, U α , U β , U γ are the parameter matrices from the input layer to the current layer of the input gate, forget gate, and output gate, respectively, x' t-1 , x' t are the data to be learned at time t-1 and time t respectively, and tanh(y) is the hyperbolic tangent function;

[0045] The hyperbolic tangent function is:

[0046]

[0047] Where y is the input data of the hyperbolic tangent function, and tanh(y) is the output data of the hyperbolic tangent function.

[0048] Preferably, judging the deformation stage of the unstable slope by the displacement time series prediction value and outputting the corresponding warning signal specifically includes:

[0049] Determine whether the increasing rate of the displacement in the displacement time series prediction value is higher than the preset rate threshold. If so, no output is made. If not, determine whether the displacement data in the displacement time series prediction value is higher than the first preset displacement threshold. If so, output that the deformation stage of the unstable slope is the uniform deformation stage and output a mild warning signal. If not, output that the deformation stage of the unstable slope is the undeformed stage and no warning signal needs to be output.

[0050] Determine whether the displacement data in the displacement time series prediction value is higher than the second preset displacement threshold. If so, output that the deformation stage of the unstable slope is the accelerated deformation stage, output a high warning signal, and evacuate nearby residents. If not, output that the deformation stage of the unstable slope is the initial deformation stage, and no warning signal needs to be output.

[0051] Furthermore, an intelligent operation and maintenance management planning system for an air compressor unit is proposed, which is used to implement the intelligent operation and maintenance management planning method for an air compressor unit as described above, including:

[0052] A data monitoring module, the data monitoring module is used to perform geological surveys on unstable slopes by installing a pulley-type fixed inclinometer, monitor the displacement at different depths in the sliding rock and soil body, and collect real-time data on factors affecting the displacement of the rock and soil body by installing at least one data monitoring device;

[0053] A data prediction module, the data prediction module is used to pre-process the deep displacement time series data of the unstable slope sliding surface, generate samples with equal time intervals, decompose the displacement time series into trend samples and fluctuation samples by using a function fitting method, predict using an artificial neural network model based on the influence of factors affecting the displacement of the rock and soil body on the displacement time series trend samples, predict using a long-short time memory network model based on the influence of factors affecting the displacement of the rock and soil body on the displacement time series fluctuation samples, and obtain the displacement time series prediction value by superimposing and summing the predicted values ​​of the fluctuation samples and the trend samples;

[0054] A signal warning module is used to determine the deformation stage of the unstable slope through the displacement time series prediction value and output a corresponding warning signal.

[0055] Optionally, the data monitoring module specifically includes:

[0056] A displacement monitoring unit, which is used to perform geological surveys on unstable slopes by installing a pulley-type fixed inclinometer to monitor displacements at different depths in the sliding rock and soil body;

[0057] The inducement monitoring unit is used to collect real-time data on factors affecting the displacement of the rock and soil mass by installing at least one data monitoring device.

[0058] Optionally, the data prediction module specifically includes:

[0059] A data preprocessing unit, the data preprocessing unit is used to preprocess the deep displacement time series data of the unstable slope slip surface to generate samples with equal time intervals;

[0060] A data decomposition unit, wherein the data decomposition unit is used to decompose the displacement time series into trend samples and fluctuation samples by using a function fitting method;

[0061] A first model building unit, the first model building unit is used to use an artificial neural network model to predict based on the influence of factors affecting the displacement of the rock and soil body on the displacement time series trend sample;

[0062] A second model building unit, the second model building unit is used to predict using a long short-term memory network model based on the influence of factors affecting the displacement of the rock and soil body on the displacement time series fluctuation samples;

[0063] The data superposition unit is used to obtain the displacement time series prediction value by superimposing and summing the prediction values ​​of the fluctuation sample and the trend sample respectively.

[0064] Compared with the prior art, the present invention has the following beneficial effects:

[0065] Based on the machine learning method, a hybrid prediction model for landslide displacement was studied, which realized the real-time advance perception of the slope deformation trend. Based on the idea of ​​deformation process warning, a multi-layer and multi-level warning criterion warning model was constructed according to the predicted displacement time series samples, which realized the purpose of real-time landslide hazard level classification and identification, and finally achieved the purpose of high slope displacement prediction accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 The flowchart of the geological monitoring and early warning method for geological survey of the present invention;

[0067] Figure 2 It is a flow chart of the method for monitoring displacement at different depths in a sliding rock and soil body of the present invention;

[0068] Figure 3 This is a flow chart of a method for real-time data collection of factors affecting the displacement of rock and soil bodies according to the present invention;

[0069] Figure 4 This is a flow chart of the prediction method using an artificial neural network model of the present invention;

[0070] Figure 5 This is a flow chart of the prediction method using the long short-term memory network model of the present invention;

[0071] Figure 6 The present invention is a flow chart of a method for determining the deformation stage of an unstable slope by using displacement time series prediction values. DETAILED DESCRIPTION

[0072] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are only examples, and those skilled in the art may think of other obvious variations.

[0073] Reference Figure 1 As shown, a geological monitoring and early warning method for geological exploration includes:

[0074] The unstable slopes are surveyed by installing a fixed pulley inclinometer to monitor the displacement at different depths in the sliding rock and soil.

[0075] By installing at least one data monitoring device, real-time data collection is performed on factors affecting the displacement of the rock and soil mass, wherein the factors include soil water pressure, soil moisture content and rainfall;

[0076] The deep displacement time series data of the unstable slope slip surface are preprocessed to generate samples with equal time intervals;

[0077] The displacement time series is decomposed into trend samples and fluctuation samples using the function fitting method;

[0078] Based on the influence of factors affecting the displacement of rock and soil mass on the displacement time series trend samples, the artificial neural network model is used for prediction;

[0079] Based on the influence of factors affecting the displacement of rock and soil mass on the fluctuation samples of displacement time series, the long and short time memory network model is used for prediction.

[0080] The predicted values ​​of the fluctuation samples and trend samples are respectively superimposed and summed to obtain the predicted value of the displacement time series;

[0081] The deformation stage of the unstable slope is determined by the displacement time series prediction value, and a corresponding warning signal is output. The deformation stage includes the undeformed stage, the initial deformation stage, the uniform deformation stage and the accelerated deformation stage.

[0082] Reference Figure 2 As shown in the figure, by installing a pulley-type fixed inclinometer to conduct geological survey on unstable slopes, the displacement at different depths in the sliding rock and soil is monitored, including:

[0083] A vertical guide groove is arranged in the inclinometer tube, and at least two sensors are connected in series at the same interval in the guide groove. The sensor with an accelerometer controls its offset direction in the guide groove through a guide wheel;

[0084] When the slope slides, the rock mass drives the inclinometer tube fixed at the bottom on the bedrock to deviate. At the same time, due to the effect of the gravity field, the sensor with accelerometer can obtain the gravitational components in the horizontal and vertical directions.

[0085] The deflection angle of the sensor in the inclinometer casing is calculated by the horizontal and vertical gravity components;

[0086] The horizontal displacement between the two sensors is calculated using the inclinometer formula;

[0087] The inclinometer formula is:

[0088]

[0089] Where S is the horizontal displacement between the two sensors, H is the depth difference between the two sensors, θ is the deflection angle between the two sensors, and α is the initial angle value when the fixed inclinometer is installed.

[0090] The pulley-type fixed inclinometer can not only directly monitor the deformation at different depths in the sliding rock and soil, but also observe the deformation between different soil layers, and judge the stability of the slope and the deformation law of the sliding surface from the overall perspective.

[0091] Reference Figure 3 As shown, by installing at least one data monitoring device, real-time data collection of factors affecting the displacement of the rock and soil mass specifically includes:

[0092] The soil water pressure is monitored in real time through the pore water pressure gauge. The pressure water acts on the pressure-bearing membrane through the permeable stone at the bottom of the sensor. The pressure-bearing membrane deforms and generates stress, which is automatically converted into pressure through the internal sensor.

[0093] The soil moisture content is monitored in real time through a soil moisture meter, and the change of moisture content is reflected by measuring the dielectric constant of the soil;

[0094] The rainfall is monitored in real time through a tipping bucket rain gauge. As the rainfall continues, the volume of water in the tipping bucket increases. When it reaches the preset height, the tipping bucket loses balance due to its excessive weight and flips left to right. The sensor records the rainfall value by recording the number of times the tipping bucket pendulum swings.

[0095] ZigBee communication technology has been widely used in landslide, mine, water conservancy and other monitoring projects in recent years due to its advantages such as low energy consumption, wide coverage, low cost and strong anti-interference. It also has two-way communication function, which can not only transmit sensor monitoring data to the upper structure in real time, but also allow staff to control the sensor data collection frequency in real time from the upper structure according to system requirements.

[0096] Reference Figure 4 As shown in the figure, based on the influence of factors affecting the displacement of rock and soil mass on the displacement time series trend samples, the prediction using artificial neural network model specifically includes:

[0097] S101: Determine the number of layers of the artificial neural network model and the number of neurons in each layer;

[0098] S102: Select a continuous and smooth S-shaped growth curve function as the neuron activation function;

[0099] S103: Randomly initialize the weights and biases of the network;

[0100] S104: forward propagating the input data through the network to obtain a predicted output;

[0101] S105: Calculate the loss function value according to the difference between the predicted output and the actual data;

[0102] S106: Update the weights and biases of the network according to the gradient of the loss function;

[0103] S107: Determine whether the preset number of iterations has been reached. If yes, output the current prediction value as the prediction value of the fluctuation sample. If no, return to step S104.

[0104] The S-shaped growth curve function is:

[0105]

[0106] Where f(x) is the S-shaped growth curve function, and x is the input data of the neuron.

[0107] An artificial neural network is an algorithmic mathematical model that imitates the behavioral characteristics of animal neural networks and performs distributed information processing. It abstracts the human brain neural network from the perspective of information processing, establishes a simple model, and forms different networks according to different connection methods.

[0108] Reference Figure 5 As shown in the figure, based on the influence of factors affecting the displacement of rock and soil mass on the fluctuation samples of displacement time series, the prediction using the long and short time memory network model specifically includes:

[0109] According to the gated recurrent unit formula, the input gate, forget gate and output gate functions are calculated respectively;

[0110] Construct a long-short time memory network model based on input gate, forget gate and output gate functions;

[0111] The factors affecting the displacement of the rock and soil mass and the fluctuation samples of the displacement time series are input into the long and short time memory network model to obtain the predicted value of the trend sample;

[0112] The gated recurrent unit formula is:

[0113]

[0114] In the formula, α t , β t , γ t are the input gate, forget gate and output gate functions respectively, W α , W β , W γ are the parameter matrices of the input gate, forget gate, and output gate from time t-1 to time t, U α , U β , U γ are the parameter matrices from the input layer to the current layer of the input gate, forget gate, and output gate, respectively, x' t-1 , x' t are the data to be learned at time t-1 and time t respectively, and tanh(y) is the hyperbolic tangent function;

[0115] The hyperbolic tangent function is:

[0116]

[0117] Where y is the input data of the hyperbolic tangent function, and tanh(y) is the output data of the hyperbolic tangent function.

[0118] Recurrent neural networks are often used in time series related tasks. Common variants include long short-term memory network models and gated recurrent units. The traditional neural network structure is feedforward, and there is no connection between the nodes of each layer. The parameters of each layer of the recurrent neural network are shared. The output of the current layer must consider not only the output of the previous layer, but also the output of the hidden layer at the previous moment.

[0119] Reference Figure 6 As shown in the figure, the deformation stage of the unstable slope is judged by the displacement time series prediction value, and the corresponding warning signal is output, including:

[0120] Determine whether the increasing rate of the displacement in the displacement time series prediction value is higher than the preset rate threshold. If so, no output is made. If not, determine whether the displacement data in the displacement time series prediction value is higher than the first preset displacement threshold. If so, output that the deformation stage of the unstable slope is the uniform deformation stage and output a mild warning signal. If not, output that the deformation stage of the unstable slope is the undeformed stage and no warning signal needs to be output.

[0121] Determine whether the displacement data in the displacement time series prediction value is higher than the second preset displacement threshold. If so, output that the deformation stage of the unstable slope is the accelerated deformation stage, output a high warning signal, and evacuate nearby residents. If not, output that the deformation stage of the unstable slope is the initial deformation stage, and no warning signal needs to be output.

[0122] The slope deformation is in the initial, uniform deformation stage. Generally, the evolution time period is relatively long, and it basically maintains the form of creep slip as a whole. If there is no influence of external triggering slip conditions such as heavy rainfall and earthquakes, the probability of slope instability at this stage is relatively low. This phenomenon is more obvious in soil slopes. When the slope deformation evolves to the accelerated deformation stage, the deformation will increase sharply in a very short period of time. The probability of overall instability and failure of the slope at this stage is very high. This phenomenon is more obvious in rock slopes.

[0123] Further, based on the same inventive concept as the above-mentioned geological monitoring and early warning method for geological exploration, the present solution also proposes a geological monitoring and early warning system for geological exploration, comprising:

[0124] A data monitoring module, the data monitoring module is used to perform geological surveys on unstable slopes by installing a pulley-type fixed inclinometer, monitor the displacement at different depths in the sliding rock and soil body, and collect real-time data on factors affecting the displacement of the rock and soil body by installing at least one data monitoring device;

[0125] A data prediction module, the data prediction module is used to pre-process the deep displacement time series data of the unstable slope sliding surface, generate samples with equal time intervals, decompose the displacement time series into trend samples and fluctuation samples by using a function fitting method, predict using an artificial neural network model based on the influence of factors affecting the displacement of the rock and soil body on the displacement time series trend samples, predict using a long-short time memory network model based on the influence of factors affecting the displacement of the rock and soil body on the displacement time series fluctuation samples, and obtain the displacement time series prediction value by superimposing and summing the predicted values ​​of the fluctuation samples and the trend samples;

[0126] A signal warning module is used to determine the deformation stage of the unstable slope through the displacement time series prediction value and output a corresponding warning signal.

[0127] The data monitoring module specifically includes:

[0128] A displacement monitoring unit, which is used to perform geological surveys on unstable slopes by installing a pulley-type fixed inclinometer to monitor displacements at different depths in the sliding rock and soil body;

[0129] The inducement monitoring unit is used to collect real-time data on factors affecting the displacement of the rock and soil mass by installing at least one data monitoring device.

[0130] The data prediction module specifically includes:

[0131] A data preprocessing unit, the data preprocessing unit is used to preprocess the deep displacement time series data of the unstable slope slip surface to generate samples with equal time intervals;

[0132] A data decomposition unit, wherein the data decomposition unit is used to decompose the displacement time series into trend samples and fluctuation samples by using a function fitting method;

[0133] A first model building unit, the first model building unit is used to use an artificial neural network model to predict based on the influence of factors affecting the displacement of the rock and soil body on the displacement time series trend sample;

[0134] A second model building unit, the second model building unit is used to predict using a long short-term memory network model based on the influence of factors affecting the displacement of the rock and soil body on the displacement time series fluctuation samples;

[0135] The data superposition unit is used to obtain the displacement time series prediction value by superimposing and summing the prediction values ​​of the fluctuation sample and the trend sample respectively.

[0136] Furthermore, the present solution also proposes a computer-readable storage medium on which a computer-readable program is stored. When the computer-readable program is called, the above-mentioned geological monitoring and early warning method for geological survey is executed.

[0137] It is understandable that the storage medium may be a magnetic medium, such as a floppy disk, a hard disk, or a magnetic tape; an optical medium, such as a DVD; or a semiconductor medium, such as a solid state drive (SSD).

[0138] In summary, the advantages of the present invention are: based on the machine learning method, a hybrid prediction model for landslide displacement is studied, which realizes real-time advance perception of the development trend of slope deformation; based on the idea of ​​deformation process warning, a multi-layer and multi-level warning criterion warning model is constructed according to the predicted displacement time series samples, which realizes the purpose of dividing and identifying the real-time hazard level of landslides, and finally achieves the purpose of high accuracy in slope displacement prediction.

[0139] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions only describe the principles of the present invention. The present invention may be subject to various changes and improvements without departing from the spirit and scope of the present invention. These changes and improvements fall within the scope of the present invention. The scope of protection claimed by the present invention is defined by the attached claims and their equivalents.

Claims

1. A geological monitoring and early warning method for geological survey, characterized in that: include: The unstable slopes are surveyed by installing a fixed pulley inclinometer to monitor the displacement at different depths in the sliding rock and soil. By installing at least one data monitoring device, real-time data collection is performed on factors affecting the displacement of the rock and soil mass, wherein the factors include soil water pressure, soil moisture content and rainfall; The deep displacement time series data of the unstable slope slip surface are preprocessed to generate samples with equal time intervals; The displacement time series is decomposed into trend samples and fluctuation samples using the function fitting method; Based on the influence of factors affecting the displacement of rock and soil mass on the displacement time series trend samples, the artificial neural network model is used for prediction; Based on the influence of factors affecting the displacement of rock and soil mass on the fluctuation samples of displacement time series, the long and short time memory network model is used for prediction. The predicted values ​​of the fluctuation samples and trend samples are respectively superimposed and summed to obtain the predicted value of the displacement time series; The deformation stage of the unstable slope is determined by the displacement time series prediction value, and a corresponding warning signal is output. The deformation stage includes the undeformed stage, the initial deformation stage, the uniform deformation stage and the accelerated deformation stage.

2. A geological monitoring and early warning method for geological survey according to claim 1, characterized in that: The geological survey of the unstable slope by installing a pulley-type fixed inclinometer to monitor the displacement at different depths in the sliding rock and soil body specifically includes: A vertical guide groove is arranged in the inclinometer tube, and at least two sensors are connected in series at the same interval in the guide groove. The sensor with an accelerometer controls its offset direction in the guide groove through a guide wheel; When the slope slides, the rock mass drives the inclinometer tube fixed at the bottom on the bedrock to deviate. At the same time, due to the effect of the gravity field, the sensor with accelerometer can obtain the gravitational components in the horizontal and vertical directions. The deflection angle of the sensor in the inclinometer casing is calculated by the horizontal and vertical gravity components; The horizontal displacement between the two sensors is calculated using the inclinometer formula; The inclinometer formula is: Where S is the horizontal displacement between the two sensors, H is the depth difference between the two sensors, θ is the deflection angle between the two sensors, and α is the initial angle value when the fixed inclinometer is installed.

3. A geological monitoring and early warning method for geological survey according to claim 2, characterized in that: The real-time data collection of factors affecting the displacement of the rock and soil mass by installing at least one data monitoring device specifically includes: The soil water pressure is monitored in real time through the pore water pressure gauge. The pressure water acts on the pressure-bearing membrane through the permeable stone at the bottom of the sensor. The pressure-bearing membrane deforms and generates stress, which is automatically converted into pressure through the internal sensor. The soil moisture content is monitored in real time through a soil moisture meter, and the change of moisture content is reflected by measuring the dielectric constant of the soil; The rainfall is monitored in real time through a tipping bucket rain gauge. As the rainfall continues, the volume of water in the tipping bucket increases. When it reaches the preset height, the tipping bucket loses balance due to its excessive weight and flips left to right. The sensor records the rainfall value by recording the number of times the tipping bucket pendulum swings.

4. A geological monitoring and early warning method for geological survey according to claim 3, characterized in that: The prediction based on the influence of the factors affecting the displacement of the rock and soil mass on the displacement time series trend samples using the artificial neural network model specifically includes: S101: Determine the number of layers of the artificial neural network model and the number of neurons in each layer; S102: Select a continuous and smooth S-shaped growth curve function as the neuron activation function; S103: Randomly initialize the weights and biases of the network; S104: forward propagating the input data through the network to obtain a predicted output; S105: Calculate the loss function value according to the difference between the predicted output and the actual data; S106: Update the weights and biases of the network according to the gradient of the loss function; S107: Determine whether the preset number of iterations has been reached. If yes, output the current prediction value as the prediction value of the fluctuation sample. If no, return to step S104. The S-shaped growth curve function is: Where f(x) is the S-shaped growth curve function, and x is the input data of the neuron.

5. A geological monitoring and early warning method for geological survey according to claim 4, characterized in that: The prediction based on the influence of the factors affecting the displacement of the rock and soil body on the displacement time series fluctuation samples using the long and short time memory network model specifically includes: According to the gated recurrent unit formula, the input gate, forget gate and output gate functions are calculated respectively; Construct a long-short time memory network model based on input gate, forget gate and output gate functions; The factors affecting the displacement of the rock and soil mass and the fluctuation samples of the displacement time series are input into the long and short time memory network model to obtain the predicted value of the trend sample; The gated recurrent unit formula is: In the formula, α t , β t , γ t are the input gate, forget gate and output gate functions respectively, W α , W β , W γ are the parameter matrices of the input gate, forget gate, and output gate from time t-1 to time t, U α , U β , U γ are the parameter matrices from the input layer to the current layer of the input gate, forget gate, and output gate, respectively, x' t-1 , x' t are the data to be learned at time t-1 and time t respectively, and tanh(y) is the hyperbolic tangent function; The hyperbolic tangent function is: Where y is the input data of the hyperbolic tangent function, and tanh(y) is the output data of the hyperbolic tangent function.

6. A geological monitoring and early warning method for geological survey according to claim 5, characterized in that: The method of judging the deformation stage of the unstable slope by the displacement time series prediction value and outputting the corresponding warning signal specifically includes: Determine whether the increasing rate of the displacement in the displacement time series prediction value is higher than the preset rate threshold. If so, no output is made. If not, determine whether the displacement data in the displacement time series prediction value is higher than the first preset displacement threshold. If so, output that the deformation stage of the unstable slope is the uniform deformation stage and output a mild warning signal. If not, output that the deformation stage of the unstable slope is the undeformed stage and no warning signal needs to be output. Determine whether the displacement data in the displacement time series prediction value is higher than the second preset displacement threshold. If so, output that the deformation stage of the unstable slope is the accelerated deformation stage, output a high warning signal, and evacuate nearby residents. If not, output that the deformation stage of the unstable slope is the initial deformation stage, and no warning signal needs to be output.

7. A geological monitoring and early warning system for geological exploration, used to implement the geological monitoring and early warning method for geological exploration as claimed in any one of claims 1 to 6, characterized in that: include: A data monitoring module, the data monitoring module is used to perform geological surveys on unstable slopes by installing a pulley-type fixed inclinometer, monitor the displacement at different depths in the sliding rock and soil body, and collect real-time data on factors affecting the displacement of the rock and soil body by installing at least one data monitoring device; A data prediction module, the data prediction module is used to pre-process the deep displacement time series data of the unstable slope sliding surface, generate samples with equal time intervals, decompose the displacement time series into trend samples and fluctuation samples by using a function fitting method, predict using an artificial neural network model based on the influence of factors affecting the displacement of the rock and soil body on the displacement time series trend samples, predict using a long-short time memory network model based on the influence of factors affecting the displacement of the rock and soil body on the displacement time series fluctuation samples, and obtain the displacement time series prediction value by superimposing and summing the predicted values ​​of the fluctuation samples and the trend samples; A signal warning module is used to determine the deformation stage of the unstable slope through the displacement time series prediction value and output a corresponding warning signal.

8. A geological monitoring and early warning system for geological survey according to claim 7, characterized in that: The data monitoring module specifically includes: A displacement monitoring unit, which is used to perform geological surveys on unstable slopes by installing a pulley-type fixed inclinometer to monitor displacements at different depths in the sliding rock and soil body; The inducement monitoring unit is used to collect real-time data on factors affecting the displacement of the rock and soil mass by installing at least one data monitoring device.

9. A geological monitoring and early warning system for geological survey according to claim 8, characterized in that: The data prediction module specifically includes: A data preprocessing unit, the data preprocessing unit is used to preprocess the deep displacement time series data of the unstable slope slip surface to generate samples with equal time intervals; A data decomposition unit, wherein the data decomposition unit is used to decompose the displacement time series into trend samples and fluctuation samples by using a function fitting method; A first model building unit, the first model building unit is used to use an artificial neural network model to predict based on the influence of factors affecting the displacement of the rock and soil body on the displacement time series trend sample; A second model building unit, the second model building unit is used to predict using a long short-term memory network model based on the influence of factors affecting the displacement of the rock and soil body on the displacement time series fluctuation samples; The data superposition unit is used to obtain the displacement time series prediction value by superimposing and summing the prediction values ​​of the fluctuation sample and the trend sample respectively.

Images