A slope monitoring system

Abstract

The present invention discloses a slope monitoring system, including a slope natural environment data acquisition unit, a stratum data acquisition unit, a geological data acquisition unit, a data preprocessing unit, a data analysis unit, a slope displacement comprehensive calculation unit, a single data value change prediction unit, a slope displacement trend prediction unit and a comprehensive early warning unit. The present invention performs a series of processing such as preprocessing, identifying abnormal data, data conversion, correlation analysis and data division on the collected data to obtain linear change data, nonlinear change data and periodic change data that are most relevant to the displacement data values ​​of the slopes on both sides of the railway; classifies the linear change data, nonlinear change data and periodic change data in the single data value for prediction; obtains the slope displacement trend prediction data value through comprehensive prediction; and adjusts the range value of the preset threshold of the comprehensive early warning unit through the slope displacement trend prediction data value.

Description

Technical Field

[0001] The invention belongs to the technical field of geological disaster monitoring, and in particular, relates to a slope monitoring system. Background Art

[0002] For railway and road slopes, the factors affecting their stability usually include stratum and lithology characteristics, geological structure characteristics, topographic and geomorphic characteristics, urban road slope structure type characteristics, climate and hydrogeology characteristics and other influencing factors. The slope is affected by natural forces (earthquakes, rainstorms, etc.), which affects the soil and eventually leads to instability. Or the existing support structure (support structure, etc.) ages to the point where it cannot withstand the impact of unstable soil, leading to disasters. If the status of the slope and support structure can be monitored, the health status of the slope and support structure can be evaluated, and an early warning can be issued before the disaster occurs, which will greatly reduce the probability of disasters.

[0003] The defect of the existing technology is that the existing slope monitoring system only determines whether the slope is landslide based on the actual monitoring value, but cannot change the monitoring threshold according to the changes in the natural environment, strata and geology of the slope to achieve better monitoring and risk prediction of slope landslide. Summary of the invention

[0004] Aiming at the problem that the existing slope monitoring system cannot change the monitoring threshold according to the changes in the natural environment, strata and geology of the slope, the present invention provides a slope monitoring system.

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

[0006] A slope monitoring system, comprising a slope natural environment data acquisition unit, a stratum data acquisition unit, a geological data acquisition unit, a data preprocessing unit, a data analysis unit, a slope displacement comprehensive calculation unit, a single data value change prediction unit, a slope displacement trend prediction unit and a comprehensive early warning unit;

[0007] The slope natural environment data acquisition unit is communicatively connected with the data preprocessing unit and is used to collect the natural environment data of the slope;

[0008] The stratum data acquisition unit is communicatively connected with the data preprocessing unit and is used for acquiring stratum data of the slope;

[0009] The geological data acquisition unit is in communication connection with the data preprocessing unit and is used for acquiring geological data of the slope;

[0010] The data preprocessing unit is in communication connection with the data analysis unit and is used to preprocess the collected natural environment data, formation data and geological data, and identify abnormal data and data conversion;

[0011] The data analysis unit is connected to the slope displacement comprehensive calculation unit for screening the data related to the slope displacement, eliminating the data irrelevant to the slope displacement, and dividing the data according to the change trend of the data related to the slope displacement;

[0012] The slope displacement comprehensive calculation unit multiplies the relevant data of the slope displacement by the relevant factor and sums them up to obtain the current slope displacement comprehensive data value;

[0013] The single data value change prediction unit is connected to the data analysis unit for communication, and predicts the data value of the next time period according to the change trend of the relevant data of the slope displacement;

[0014] The slope displacement trend prediction unit is connected to the single data value change prediction unit for communication, and the data value of the next time period is multiplied by the correlation factor and the sum is calculated to obtain the slope displacement comprehensive prediction data value of the next time period;

[0015] The comprehensive early warning unit is communicatively connected with the slope displacement comprehensive calculation unit and the slope displacement trend prediction unit, and is used to comprehensively judge whether the current slope displacement comprehensive data value and the slope displacement comprehensive prediction data value in the next time period exceed a preset threshold.

[0016] Furthermore, the detailed steps of preprocessing natural environment data, stratigraphic data and geological data are as follows:

[0017] Clean the collected natural environment data, stratigraphic data and geological data; fill in the missing data with the median or historical data mean.

[0018] The cleaned data is adjusted to the same measurement unit as the threshold data preset by the system. Each single data is converted into a unified unit for storing the single data in the system, which is convenient for the subsequent automatic calculation of the system.

[0019] Furthermore, the detailed steps for identifying abnormal data include:

[0020] Perform single data threshold comparison on the collected natural environment data, stratigraphic data and geological data;

[0021] If it is not within the preset single data threshold range, it is determined that the acquisition device is abnormal, and the data abnormal value caused by the acquisition device abnormality is deleted and replaced;

[0022] If it is within the preset single data threshold range, the collected data is determined to be normal and enters data conversion.

[0023] Furthermore, data conversion: the collected natural environment data, stratigraphic data and geological data are converted into digital signals that can be recognized by the system, which is convenient for the system to uniformly recognize and calculate the comprehensive data value of slope displacement.

[0024] Furthermore, the detailed steps of screening the data related to the displacement of the slope and eliminating the data irrelevant to the displacement of the slope in the data analysis unit are as follows:

[0025] The maximum mutual information coefficient is used to calculate the correlation degree between the natural environment data, stratigraphic data and geological data and the slope displacement data value, and the relevant data with a low correlation degree with the slope displacement data value are eliminated.

[0026] Furthermore, the detailed steps for calculating the maximum mutual information coefficient and the degree of correlation between the natural environment data, stratigraphic data and geological data and the slope displacement data values ​​are as follows:

[0027] Assume that natural environment data, stratigraphic data and geological data are variables A={a i , i=1,2,…,N} and the slope displacement comprehensive data value is set to B={b i , i=1,2,…,N}, where N is the number of samples, and the MI between A and B, denoted as I(A;B), is defined as follows:

[0028]

[0029] Where P AB (a,b) is the joint probability density function of A and B, P A (a) and P B (b) are the marginal probability density functions of A and B respectively.

[0030] For a finite data set D = {(a i , b i ),i = 1,2,…,N}, we can divide the 2D plane represented by D into x×y grids G. For a fixed D, it can be divided into different numbers of grids G. Therefore, for different grids G, different distributions D|G can be obtained, and we can calculate MI based on a given grid G. Definition I * As the maximum MI obtained for any grid G ​​with x columns and y rows over the data D, it can be written as follows:

[0031]

[0032] where I(D|G) is the MI of D|G.

[0033] Based on I * , the characteristic matrix M(D) of set D x,y It can be expressed as

[0034]

[0035] Then, based on the feature matrix M(D) x,y Get MIC:

[0036]

[0037] Where N is the number of samples, N 0.6 Indicates the upper bound of the grid size.

[0038] The range of MIC value is [0,1]. The larger the MIC value of influencing factor A and displacement rate, the higher the correlation between A and deformation, and vice versa.

[0039] Furthermore, the change trend data of the relevant data of the slope displacement are divided into: linear change data, nonlinear change data and periodic change data.

[0040] Detailed steps for dividing the changing trend data of the slope displacement related data:

[0041] A section of historical data collected from the relevant data on slope displacement;

[0042] The collected data on slope displacement will be arranged in chronological order;

[0043] Subtract or compare the relevant data of the slope displacement generated by the first moment with the relevant data of the slope displacement generated by the next moment, and determine whether the difference or comparison value has any relationship with the relevant data and related items of the slope displacement generated by the first moment;

[0044] If there is a relationship, it is determined that the relevant data for the displacement of the slope is linearly changing data;

[0045] If there is no relationship, a second difference or comparison is performed on the difference or comparison value, and the cycle is repeated N times. If there is still no relationship between the relevant data and related items of the slope displacement, it is determined to be nonlinear change data and the cycle is exited.

[0046] Detailed steps for dividing periodic change data into trend data:

[0047] A section of historical data collected from the relevant data on slope displacement;

[0048] The collected data on slope displacement will be arranged in chronological order;

[0049] The relevant data of slope displacement collected at different times are selected one by one as matching items, and whether there is the same data in a section of historically collected data is traversed. If the same data appears, the number of times or items of the same data appear is calculated, and the position where the first same data appears is intercepted; if the same data cannot be found, it is determined whether there is a data excerpt missing after looping through all data items, and the same data is matched accordingly;

[0050] When the matching item of the relevant data of the slope displacement generated from the first acquisition is cycled to the previous item of data at the first position where the same data appears, the whole cycle is jumped out;

[0051] Determine whether all the relevant data on slope displacement collected at different times match the same data, and whether the number of times or items of the same data appear are the same;

[0052] If they are all the same, it is determined to be periodically changing data; if one is different, it is determined not to be periodically changing data.

[0053] Furthermore, the slope displacement comprehensive calculation unit calculates the current slope displacement comprehensive data value using the formula:

[0054] C=T*R T +P*R P +N*R N ;

[0055] Where C is the current slope displacement comprehensive data value, T, P and N represent linear change data, nonlinear change data and periodic change data respectively; R T , R P and R N Represent the correlation factors respectively.

[0056] Furthermore, in the single data value change prediction unit, the double exponential smoothing method is used to predict the single data value for the linear change data in the natural environment data, the formation data and the geological data;

[0057] The single hidden layer feedforward neural network (SLFN) learning algorithm is used to predict single data values ​​for nonlinear variation data in natural environment data, stratigraphic data and geological data;

[0058] For periodically changing data in natural environment data, stratigraphic data and geological data, the current data value is used as the item in the historical data value, and the data value of the next item is used as the predicted value at the next moment.

[0059] Furthermore, the double exponential smoothing method predicts a single data value formula:

[0060] ;

[0061] F t+m represents the predicted value at time t+m, {s t} represents the smoothed value at time t, {b t} represents the best estimate of the trend at time t; m is the prediction item. If it is a displacement prediction one step ahead, m is equal to 1.

[0062] For {x t}, using t=0 to represent the observed original data sequence, and predicting the x value for m>0 based on the original data before time t:

[0063] ;

[0064] For t>1, the prediction of x value is:

[0065]

[0066] Where α is the data smoothing factor (0<α<1) and β is the trend smoothing factor (0<β<1).

[0067] Furthermore, the single hidden layer feedforward neural network learning algorithm predicts a single data value formula:

[0068]

[0069] For N arbitrarily distinct samples (x i , y i ), where x i =[x i1 , x i2 , …, x in ] T ∈R n Represents the input variable, y i =[y i1 , y i2 , …, y im ] T ∈R m represents the corresponding output variable, with K hidden nodes (K≤N) and activation function g(x); w i =[w i1 , w i2 , …, w in ] T is the weight vector connecting the i-th hidden layer neuron and the input node, β i =[β i1 , β i2 , …, β im ] T is the weight vector connecting the i-th hidden layer neuron and the output node, o j =[o j1 , o j2 , …,o jm ] T is the jth output vector of the single hidden layer feedforward neural network learning algorithm, b i is the bias of the neuron in the ith hidden layer.

[0070] Furthermore, the preset lower limit value of the threshold is the minimum value of the comprehensive calculation of historical slope displacement, and the upper limit value is the maximum value of the comprehensive calculation of historical slope displacement; the preset lower limit value and upper limit value of the threshold are replaced by the slope displacement trend prediction value.

[0071] Compared with the prior art, the present invention has the following beneficial effects:

[0072] After effectively collecting the natural environment data, stratum data and geological data of the slopes on both sides of the railway, the collected data is preprocessed, abnormal data is identified, data is converted, correlation analysis is performed and data is divided to obtain the linear change data, nonlinear change data and periodic change data that are most relevant to the displacement data values ​​of the slopes on both sides of the railway; the slope displacement comprehensive calculation unit is then used to calculate the current slope displacement comprehensive data value; then the single data value is classified into linear change data, nonlinear change data and periodic change data for prediction; finally, the slope displacement trend prediction data value is obtained through comprehensive prediction; at the same time, the preset threshold of the comprehensive early warning unit is adjusted through the slope displacement trend prediction data value. The effect of changing the monitoring threshold value due to changes in the natural environment, stratum and geology of the slope is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] Figure 1 It is an overall structural block diagram of a slope monitoring system in an embodiment of the present invention;

[0074] Figure 2 A detailed flow chart of steps for dividing the change trend data of the relevant data of the slope displacement in an embodiment of the present invention;

[0075] Figure 3 It is a detailed step flow chart of dividing the periodically changing data and the changing trend data in an embodiment of the present invention.

[0076] Explanation of the marks in the figure: 10 - slope natural environment data acquisition unit, 20 - stratigraphic data acquisition unit, 30 - geological data acquisition unit, 40 - data preprocessing unit, 50 - data analysis unit, 60 - slope displacement comprehensive calculation unit, 70 - single data value change prediction unit, 80 - slope displacement trend prediction unit, 90 - comprehensive early warning unit. DETAILED DESCRIPTION

[0077] In order to facilitate the understanding of those skilled in the art, the present invention is further described below in conjunction with embodiments and drawings. The contents mentioned in the implementation modes are not intended to limit the present invention.

[0078] like Figure 1As shown, this embodiment provides a slope monitoring system, including a slope natural environment data acquisition unit 10, a stratum data acquisition unit 20, a geological data acquisition unit 30, a data preprocessing unit 40, a data analysis unit 50, a slope displacement comprehensive calculation unit 60, a single data value change prediction unit 70, a slope displacement trend prediction unit 80 and a comprehensive early warning unit 90;

[0079] The slope natural environment data acquisition unit 10 is in communication connection with the data preprocessing unit 40 and is used to collect the natural environment data of the slope;

[0080] The formation data acquisition unit 20 is in communication connection with the data preprocessing unit 40 and is used to acquire the formation data of the slope;

[0081] The geological data acquisition unit 30 is in communication connection with the data preprocessing unit 40 and is used to acquire geological data of the slope;

[0082] The data preprocessing unit 40 is in communication with the data analysis unit 50 and is used to preprocess the collected natural environment data, formation data and geological data, and identify abnormal data and data conversion;

[0083] The data analysis unit 50 is in communication connection with the slope displacement comprehensive calculation unit 60, and is used to screen the data related to the slope displacement, remove the data irrelevant to the slope displacement, and divide the data according to the change trend of the data related to the slope displacement;

[0084] The slope displacement comprehensive calculation unit 60 multiplies the relevant data of the slope displacement by the relevant factor, and sums them up to obtain the current slope displacement comprehensive data value;

[0085] The single data value change prediction unit 70 is in communication connection with the data analysis unit 50, and predicts the data value of the next time period according to the change trend of the relevant data of the slope displacement;

[0086] The slope displacement trend prediction unit 80 is in communication connection with the single data value change prediction unit 70, and multiplies the data value of the next time period by the correlation factor and sums them up to obtain the slope displacement comprehensive prediction data value of the slope displacement in the next time period;

[0087] The comprehensive warning unit 90 is in communication with the slope displacement comprehensive calculation unit 60 and the slope displacement trend prediction unit 80 to comprehensively determine whether the current slope displacement comprehensive data value and the slope displacement comprehensive prediction data value for the next time period exceed a preset threshold.

[0088] Detailed steps for preprocessing natural environment data, stratigraphic data and geological data:

[0089] Clean the collected natural environment data, stratigraphic data and geological data; fill in the missing data with the median or historical data mean.

[0090] The cleaned data is adjusted to the same measurement unit as the threshold data preset by the system. Each single data is converted into a unified unit for storing the single data in the system, which is convenient for the subsequent automatic calculation of the system.

[0091] The detailed steps to identify abnormal data include:

[0092] Perform single data threshold comparison on the collected natural environment data, stratigraphic data and geological data;

[0093] If it is not within the preset single data threshold range, it is determined that the acquisition device is abnormal, and the data abnormal value caused by the acquisition device abnormality is deleted and replaced;

[0094] If it is within the preset single data threshold range, the collected data is determined to be normal and enters data conversion.

[0095] Data conversion: The collected natural environment data, stratigraphic data and geological data are converted into digital signals that can be recognized by the system. This facilitates the subsequent unified recognition and calculation of the comprehensive data value of slope displacement by the system.

[0096] The data analysis unit 50 selects data related to the displacement of the slope, calculates the correlation degree of the natural environment data, stratum data and geological data with the slope displacement data value using the maximum mutual information coefficient, and removes data with low correlation degree with the slope displacement data value.

[0097] Calculation of the maximum mutual information coefficient Calculation of the correlation between natural environment data, stratigraphic data and geological data and slope displacement data values;

[0098] Assume that natural environment data, stratigraphic data and geological data are variables A={a i , i=1,2,…,N} and the slope displacement comprehensive data value is set to B={b i , i=1,2,…,N}, where N is the number of samples, and the MI between A and B, denoted as I(A;B), is defined as follows:

[0099]

[0100] Where P AB (a,b) is the joint probability density function of A and B, P A (a) and P B (b) are the marginal probability density functions of A and B respectively.

[0101] For a finite data set D = {(a i, b i ),i = 1,2,…,N}, we can divide the 2D plane represented by D into x×y grids G. For a fixed D, it can be divided into different numbers of grids G. Therefore, for different grids G, different distributions D|G can be obtained, and we can calculate MI based on a given grid G. Definition I * As the maximum MI obtained for any grid G ​​with x columns and y rows over the data D, it can be written as follows:

[0102]

[0103] where I(D|G) is the MI of D|G.

[0104] Based on I * , the characteristic matrix M(D) of set D x,y It can be expressed as

[0105]

[0106] Then, based on the feature matrix M(D) x,y Get MIC:

[0107]

[0108] Where N is the number of samples, N 0.6 Indicates the upper bound of the grid size.

[0109] The range of MIC value is [0,1]. The larger the MIC value of influencing factor A and displacement rate, the higher the correlation between A and deformation, and vice versa. In the training of the prediction model, influencing factors that have little relationship with deformation should be eliminated.

[0110] like Figure 2 As shown in the figure, the change trend data of the relevant data that causes the displacement of the slope is divided into: linear change data, nonlinear change data and periodic change data. Linear change data: For example, the linear component of the cumulative displacement controlled by topography, geological structure and physical and mechanical properties. Nonlinear change data: Sensor aging, electromagnetic interference and irregular data acquisition operations will cause data noise; the plant growth process leads to an increase in the geological soil solidification area and water absorption performance, and the slope displacement value decreases accordingly, effectively protecting the slope from landslide risks. Periodic change data is controlled by external factors, such as rainfall and snowmelt. The time when the train passes the current slope every day, month or year is the periodic change data that brings vibration to the slope and affects the displacement value.

[0111] Detailed steps for dividing the changing trend data of the slope displacement related data:

[0112] S101, selecting a section of historically collected data related to the displacement of the slope;

[0113] S102, arranging the collected data related to the displacement of the slope in chronological order;

[0114] S103, subtract or compare the relevant data of the slope displacement generated by the first moment with the relevant data of the slope displacement generated by the next moment, and determine whether the value after the difference or comparison is related to the relevant data of the slope displacement generated by the slope and the related items; the relationship is that the value after the difference or comparison is the same as the relevant data of the slope displacement generated by the original item collected at the initial moment or forms an arithmetic progression or a geometric progression.

[0115] S104, if there is a relationship, determining that the relevant data of the slope displacement is linear change data;

[0116] S105, if there is no relationship, then perform a second difference or comparison on the difference or comparison value, and repeat N times in sequence. If there is still no relationship between the relevant data and related items of the slope displacement, it is determined to be nonlinear change data and the loop is exited.

[0117] like Figure 3 As shown in the figure, the detailed steps of dividing the periodic change data and the trend data are as follows:

[0118] S201, selecting a section of historically collected data related to the displacement of the slope;

[0119] S202, arranging the collected data related to the displacement of the slope in chronological order;

[0120] S203, select the relevant data of the slope displacement collected at different times as matching items one by one, and traverse whether there is the same data in a section of historically collected data. If the same data appears, calculate the number of times or items of the same data appear, and intercept the first position of the same data; if the same data cannot be found, determine whether the data excerpts are missing after traversing all data items in a loop, and match the same data; prevent the second periodic appearance of the same data from having a long time interval, resulting in the historically collected section of data for the relevant data of the slope displacement not being able to fully accommodate the data values ​​within two cycles. The relevant data of the entire slope displacement can be determined as periodically changing data by the appearance of N identical data items.

[0121] S204, when the matching item of the relevant data of the slope displacement generated from the first collected data is cycled to the previous item of data at the first position where the same data appears, the whole cycle is jumped out;

[0122] S205, judging whether all the relevant data on slope displacement collected at different times are matched to the same data, and whether the number of times or items of the same data are the same;

[0123] S206: If they are all the same, it is determined to be periodically changing data; if there is one difference, it is determined not to be periodically changing data.

[0124] The slope displacement comprehensive calculation unit 60 calculates the current slope displacement comprehensive data value using the formula:

[0125] C=T*R T +P*R P +N*R N ;

[0126] Where C is the current slope displacement comprehensive data value, T, P and N represent linear change data, nonlinear change data and periodic change data respectively; R T , R P and R N Represent the correlation factors respectively.

[0127] In the single data value change prediction unit 70, the double exponential smoothing method is used to predict the single data value for the linear change data in the natural environment data, the formation data and the geological data;

[0128] The single hidden layer feedforward neural network (SLFN) learning algorithm is used to predict single data values ​​for nonlinear variation data in natural environment data, stratigraphic data and geological data;

[0129] For periodically changing data in natural environment data, stratigraphic data and geological data, the current data value is used as the item in the historical data value, and the data value of the next item is used as the predicted value at the next moment.

[0130] The formula for predicting a single data value using the double exponential smoothing method is:

[0131] ;

[0132] F t+m represents the predicted value at time t+m, {s t} represents the smoothed value at time t, {b t} represents the best estimate of the trend at time t; m is the prediction item. If it is a displacement prediction one step ahead, m is equal to 1.

[0133] For {x t}, using t=0 to represent the observed original data sequence, and predicting the x value for m>0 based on the original data before time t:

[0134] ;

[0135] For t>1, the prediction of x value is:

[0136]

[0137] Where α is the data smoothing factor (0<α<1) and β is the trend smoothing factor (0<β<1).

[0138] The single hidden layer feedforward neural network learning algorithm predicts a single data value formula:

[0139]

[0140] For N arbitrarily distinct samples (x i , y i ), where x i =[x i1 , x i2 , …, x in ] T ∈R n Represents the input variable, y i =[y i1 , y i2 , …, y im ] T ∈R m represents the corresponding output variable, with K hidden nodes (K≤N) and activation function g(x); w i =[w i1 , w i2 , …, w in ] T is the weight vector connecting the i-th hidden layer neuron and the input node, β i =[β i1 , β i2 , …, β im ] T is the weight vector connecting the i-th hidden layer neuron and the output node, o j =[o j1 , o j2 , …,o jm ] T is the jth output vector of the single hidden layer feedforward neural network learning algorithm, b i is the bias of the neuron in the ith hidden layer.

[0141] The preset threshold lower limit is the minimum value of the historical slope displacement comprehensive calculation, and the upper limit is the maximum value of the historical slope displacement comprehensive calculation; the preset threshold lower limit and upper limit are replaced by the slope displacement trend prediction value. The preset threshold is updated in real time to ensure that the slope displacement value changes with the change of the natural environment data, formation data and geological data of the slope. It is also possible to predict a single data value and send the single data value prediction to the front-end acquisition device to realize the threshold setting of the front-end acquisition value, and then set a simple data comparison and judgment size single chip at the front end to realize the front-end threshold determination.

[0142] Compared with the prior art, the present invention has the following beneficial effects:

[0143] After effectively collecting the natural environment data, stratum data and geological data of the slopes on both sides of the railway, the collected data is preprocessed, abnormal data is identified, data is converted, correlation analysis is performed and data is divided to obtain the linear change data, nonlinear change data and periodic change data that are most relevant to the displacement data values ​​of the slopes on both sides of the railway; the slope displacement comprehensive calculation unit is then used to calculate the current slope displacement comprehensive data value; then the single data value is classified into linear change data, nonlinear change data and periodic change data for prediction; finally, the slope displacement trend prediction data value is obtained through comprehensive prediction; at the same time, the preset threshold of the comprehensive early warning unit is adjusted through the slope displacement trend prediction data value. The effect of changing the monitoring threshold value due to changes in the natural environment, stratum and geology of the slope is achieved.

[0144] The above is a detailed introduction to a slope monitoring system provided by the present application. The description of the specific embodiments is only used to help understand the method and its core idea of ​​the present application. It should be pointed out that for ordinary technicians in this technical field, without departing from the principles of the present application, several improvements and modifications can be made to the present application, and these improvements and modifications also fall within the scope of protection of the claims of the present application.

Claims

1. A slope monitoring system, characterized in that: It comprises a slope natural environment data acquisition unit (10), a stratum data acquisition unit (20), a geological data acquisition unit (30), a data preprocessing unit (40), a data analysis unit (50), a slope displacement comprehensive calculation unit (60), a single data value change prediction unit (70), a slope displacement trend prediction unit (80) and a comprehensive early warning unit (90); The slope natural environment data collection unit (10) is communicatively connected to the data preprocessing unit (40) and is used to collect the natural environment data of the slope; The stratum data acquisition unit (20) is communicatively connected to the data preprocessing unit (40) and is used to acquire stratum data of the slope; The geological data acquisition unit (30) is communicatively connected with the data preprocessing unit (40) and is used to acquire geological data of the slope; The data preprocessing unit (40) is in communication connection with the data analysis unit (50) and is used to preprocess the collected natural environment data, stratum data and geological data, and to identify abnormal data and data conversion; The data analysis unit (50) is in communication connection with the slope displacement comprehensive calculation unit (60) and is used to screen the data related to the slope displacement, remove the data irrelevant to the slope displacement, and divide the data according to the change trend of the data related to the slope displacement; A slope displacement comprehensive calculation unit (60) multiplies the relevant data of the slope displacement by a relevant factor and sums them up to obtain a current slope displacement comprehensive data value; The single data value change prediction unit (70) is in communication connection with the data analysis unit (50) and predicts the data value of the next time period according to the change trend of the relevant data of the slope displacement; The slope displacement trend prediction unit (80) is in communication connection with the single data value change prediction unit (70), and multiplies the data value of the next time period by the relevant factor and sums them up to obtain the slope displacement comprehensive prediction data value of the slope displacement in the next time period; The comprehensive early warning unit (90) is in communication connection with the slope displacement comprehensive calculation unit (60) and the slope displacement trend prediction unit (80), and is used to comprehensively judge whether the current slope displacement comprehensive data value and the slope displacement comprehensive prediction data value in the next time period exceed a preset threshold value; The data analysis unit (50) selects data related to the displacement of the slope, calculates the correlation degree between the natural environment data, the formation data and the geological data and the slope displacement data value by using the maximum mutual information coefficient, and removes the data related to the slope displacement data value with a low correlation degree; The maximum mutual information coefficient is calculated by calculating the correlation degree between the natural environment data, the formation data and the geological data and the slope displacement data value: Assume that natural environment data, stratigraphic data and geological data are variables A={a i },i=1,2,…,N and the slope displacement comprehensive data value is set to B={bi},i=1,2,…,N, where N is the number of samples, and the mutual information between A and B, denoted as I(A;B), is defined as follows: Where P AB (a,b) is the joint probability density function of A and B, P A (a) and P B (b) are the marginal probability density functions of A and B respectively.

2. A slope monitoring system according to claim 1, characterized in that: Detailed steps for preprocessing natural environment data, stratigraphic data and geological data: Clean the collected natural environment data, stratigraphic data and geological data; Adjust the measurement units of the cleaned data to be consistent with the threshold data preset by the system.

3. A slope monitoring system according to claim 2, characterized in that: The detailed steps to identify abnormal data include: Perform single data threshold comparison on the collected natural environment data, stratigraphic data and geological data; If it is not within the preset single data threshold range, it is determined that the acquisition device is abnormal, and the data abnormal value caused by the acquisition device abnormality is deleted and replaced; If it is within the preset single data threshold range, the collected data is determined to be normal and enters data conversion.

4. A slope monitoring system according to claim 3, characterized in that: Data conversion: Convert the collected natural environment data, formation data and geological data into digital signals that can be recognized by the system.

5. A slope monitoring system according to claim 4, characterized in that: The change trend data of the relevant data of slope displacement are divided into: linear change data, nonlinear change data and periodic change data; Detailed steps for dividing the changing trend data of the slope displacement related data: S101, selecting a section of historically collected data related to the displacement of the slope; S102, arranging the collected data related to the displacement of the slope in chronological order; S103, subtracting or comparing the relevant data of the slope displacement generated at the first moment with the relevant data of the slope displacement generated at the next moment, and determining whether the difference or comparison value has any relationship with the relevant data and related items of the slope displacement generated; S104, if there is a relationship, determining that the relevant data of the slope displacement is linear change data; S105, if there is no relationship, then perform a second difference or comparison on the difference or comparison value, and repeat N times in sequence. If there is still no relationship between the relevant data and related items of the slope displacement, it is determined to be nonlinear change data and the loop is exited.

6. A slope monitoring system according to claim 5, characterized in that: Detailed steps for dividing periodic change data into trend data: S201, selecting a section of historically collected data related to the displacement of the slope; S202, arranging the collected data related to the displacement of the slope in chronological order; S203, selecting the relevant data of the slope displacement collected at different times as matching items one by one, and traversing whether there is the same data in a section of historically collected data. If the same data appears, the number of times or items of the same data appear, and the position where the first same data appears is intercepted; if the same data cannot be found, it is determined whether there is a data excerpt missing corresponding to the same matching after traversing all data items in a loop; S204, when the matching item of the relevant data of the slope displacement generated from the first collected data is cycled to the previous item of data at the first position where the same data appears, the whole cycle is jumped out; S205, judging whether all the relevant data on slope displacement collected at different times are matched to the same data, and whether the number of times or items of the same data appear are the same; S206: If they are all the same, it is determined to be periodically changing data; if there is one difference, it is determined not to be periodically changing data.

7. A slope monitoring system according to claim 6, characterized in that: The slope displacement comprehensive calculation unit (60) calculates the current slope displacement comprehensive data value using the formula: C=T*R T +P*R P +N*R N ; Where C is the current slope displacement comprehensive data value, T, P and N represent linear change data, nonlinear change data and periodic change data respectively; R T , R P and R N Represent the correlation factors respectively.

Images