Improved landslide critical sliding time prediction system and method

Abstract

The invention discloses an improved landslide critical sliding time prediction system and method, and the method comprises the steps: S1, obtaining real-time monitoring data, and arranging the real-time monitoring data according to a time sequence; s2, performing calculation and connection drawing on the arranged real-time monitoring data to obtain a landslide deformation rate smooth curve; s3, nondimensionalizing the landslide deformation rate smooth curve through a tangent angle calculation formula alpha = arctan (V / VP) to obtain a plurality of tangent angles; the multiple tangent angles are connected and drawn to obtain a tangent angle change curve; s4, the landslide deformation rate smooth curve is calculated through a velocity reciprocal formula 1 / V = T / S, a plurality of velocity reciprocal values are obtained, V represents the deformation rate after smoothing, T represents time, and S represents displacement; and the multiple speed reciprocal values are connected and drawn to obtain a landslide speed reciprocal change curve.

Description

Technical Field

[0001] The invention relates to the field of geological disaster prevention and control, and in particular to an improved landslide imminent sliding time prediction system and method. Background Art

[0002] The prediction of landslide time has gone through a long process. Since the 1960s, scholars at home and abroad have proposed the Saito model (Saito M. 1969. Forecasting time of slope failure by tertiary creep [C] / Proceedings of the 7th International Conference on Soil Mechanics and Foundation Engineering Mexico city, (2): 677-683.), the graphical extension method model (Hoek E, Bray J. 1977. Rock slope engineering. Institution of Mining and Metallurgy [J], Publication of Institution of Mining & Metallurgy, 14 (4): 492-494.), the Fuluzono slope time prediction method (Fuluzono T. Recent Studies on Time Prediction of Slope Failure [J]. Landslide News, 1990, (4): 9-12.) and other deterministic prediction methods; grey prediction model (Chen Mingdong, Wang Lansheng. Grey prediction method for slope deformation and failure: Proceedings of the Third National Engineering Geology Conference, 1988: 1226-1233), Verhulst model (Yin Kunlong, Yan Tongzhen. Landslide prediction and related models [J]. Chinese Journal of Rock Mechanics and Engineering, 1996, (01): 1-8.) and other statistical prediction models; BP neural network model (Wei Yuming, Dang Xinghai. Application of BP neural network in landslide deformation prediction of unequal time intervals [J]. Science Technology and Engineering, 2010, 10 (12): 2959-2962.), nonlinear dynamic model (Zhou Cuiying, Yan Tongzhen, Tang Liansheng. Research on nonlinear dynamics of landslide hazard system [J]. Journal of Changchun Institute of Geology, 1995, (03): 310-316.) and other nonlinear prediction models. Although there are many types of existing prediction models, the calculation methods are different, resulting in large differences in prediction time, and most of them are retrospective predictions and one-time predictions, that is, after the landslide occurs, the prediction model is used to analyze the data, predict the landslide time, and verify the accuracy of the model, and fails to update the prediction time in real time according to the deformation rate during the landslide process. In addition, the existing monitoring equipment is affected by accidental errors and gross errors, resulting in the monitoring data showing up and down fluctuations. At the same time, most prediction models cannot process the error data in time, resulting in the deviation of the landslide prediction time ranging from a few days to a few months.Especially for some sudden landslides, when entering the mid-acceleration stage, the time from occurrence may range from more than ten minutes to several hours. It is difficult to calculate and estimate the imminent time of the landslide using common prediction models, which will seriously affect subsequent emergency rescue and emergency evacuation of the masses.

[0003] Based on the tangent angle model and the velocity inverse model, an improved landslide prediction system and method are proposed. Through the research and analysis of a large number of scholars at home and abroad, it is found that the landslide pre-slip time predicted by the existing model is generally late. The main reason is that after the landslide enters the acceleration stage, the deformation rate continues to increase, causing the velocity inverse value to accelerate to 0, the tangent angle increases sharply, and continues to approach 90°, and some prediction models cannot recognize this feature in time. In addition, some scholars did not consider the impact of the monitoring data after the landslide enters the acceleration stage and the data of the uniform deformation stage on the prediction time. They only made a one-time prediction based on a certain threshold, ignoring the correlation between the previous and next data, which ultimately led to a large difference between the predicted time and the actual occurrence time. In recent years, multiple refined deformation curves of the whole process of landslides have been obtained through intelligent monitoring equipment, which confirms that the deformation rate of the landslide deformation curve will increase after entering the initial acceleration stage, and the deformation rate will increase again after entering the mid-acceleration and pre-slip stages. The late prediction time is likely to bring potential threats to emergency rescue and other disaster prevention work. Summary of the invention

[0004] The purpose of the present invention is to provide an improved landslide prediction system and method, aiming to overcome the problems of low time accuracy, low stability, limited scope of application and the like of existing landslide prediction.

[0005] In order to achieve the purpose of the present invention, the present invention provides an improved method for predicting the time of landslide, comprising:

[0006] Step S1: acquiring real-time monitoring data, and arranging the real-time monitoring data in time sequence;

[0007] Step S2: Calculating and drawing lines on the arranged real-time monitoring data to obtain a smooth curve of landslide deformation rate;

[0008] Step S3: Calculate the tangent angle by the formula α = arctan (V / V P ) The landslide deformation rate smooth curve is dimensionless to obtain multiple tangent angles, where: V is the smoothed deformation rate, V P is the deformation rate of the uniform deformation stage of the landslide, α is the tangent angle, and the value range is 0 to 90°; a plurality of the tangent angles are connected and drawn to obtain a tangent angle variation curve;

[0009] Step S4: Calculate the landslide deformation rate smooth curve by the velocity inverse formula 1 / V=T / S to obtain a plurality of velocity inverse values, wherein V represents the deformation rate after smoothing, T represents time, and S represents displacement; connect the plurality of velocity inverse values ​​and draw a line to obtain a landslide velocity inverse change curve;

[0010] Step S5: Based on the tangent angle variation curve and the landslide velocity inverse variation curve, with the tangent angle of 80° as the prediction starting point, 100 sets of data are selected for expansion and linear fitting to obtain a landslide velocity inverse linear fitting curve, and the monitoring time when the velocity inverse value in the landslide velocity inverse linear fitting curve is 0 is used as the preliminary prediction time of the landslide, and this step is repeated to obtain multiple preliminary prediction times of the landslide;

[0011] Step S6: Select 8 groups of the latest landslide preliminary prediction times and sort them by size, remove the first, third and second smallest landslide preliminary prediction times, calculate the mean of the remaining landslide preliminary prediction times, and obtain the current landslide preliminary prediction time;

[0012] Step S7: taking the current landslide prediction time as one of the latest eight groups of landslide preliminary prediction times at the next moment, and repeating step S6 to obtain the landslide prediction time at the next moment.

[0013] Furthermore, the arranged real-time monitoring data are calculated and connected to obtain a smooth curve of landslide deformation rate, including:

[0014] The latest 10 sets of real-time monitoring data are used as an interval to calculate multiple deformation rates, and the formula is: V t =ΔS / ΔT, where V t is the deformation rate at the latest moment, ΔS is the difference between the latest monitored displacement and the earliest monitored displacement in the interval, and ΔT is the difference between the end time and the start time of the calculation interval;

[0015] Select the latest first 10 groups of deformation rates in turn, use the mean method to smooth the selected deformation rates to obtain the smoothed deformation rates at multiple latest monitoring moments, the formula is V = (V t-1 +V t-2 +…+V t-N ) / N, where: V t-N is the Nth value of the latest deformation rate, N is the total number of monitoring data in the interval, NN = 10;

[0016] The smoothed deformation rates are connected and drawn to obtain a smooth curve of landslide deformation rate.

[0017] Furthermore, based on the tangent angle change curve and the landslide velocity inverse change curve, with the tangent angle of 80° as the prediction starting point, 100 sets of data were selected for expansion and linear fitting to obtain the landslide velocity inverse linear fitting curve, including:

[0018] The first 100 data sets with tangent angles of 80° and above were selected, the last 10 data sets of the 100 data sets were copied 3 times and added to the 100 data sets to obtain 130 data sets, and the 130 data sets were linearly fitted to obtain the inverse linear fitting curve of landslide velocity.

[0019] The second aspect of the present invention provides an improved landslide imminent sliding time prediction system, which is used in any of the above methods, comprising:

[0020] Data acquisition module: used to acquire real-time monitoring data and arrange the real-time monitoring data in time sequence;

[0021] A landslide deformation rate smooth curve drawing module is used to calculate and draw the arranged real-time monitoring data to obtain a landslide deformation rate smooth curve;

[0022] Tangent angle change curve drawing module: used to calculate the tangent angle through the formula α = arctan (V / V P ) The landslide deformation rate smoothing curve is dimensionless to obtain multiple tangent angles, and the tangent angles are connected and drawn to obtain a tangent angle change curve; wherein: V is the smoothed deformation rate, V P is the deformation rate of the landslide in the uniform deformation stage, α is the tangent angle, and its value range is 0 to 90°;

[0023] The landslide velocity reciprocal change curve drawing module is used to calculate the landslide deformation rate smooth curve through the velocity reciprocal formula 1 / V=T / S to obtain multiple velocity reciprocal values, and connect the multiple velocity reciprocal values ​​to obtain the landslide velocity reciprocal change curve, wherein V represents the deformation rate after smoothing, T represents time, and S represents displacement;

[0024] Landslide Preliminary Time Prediction Module: used to make a preliminary prediction of the landslide occurrence time to obtain the preliminary predicted time of the landslide;

[0025] The current landslide time prediction module is used to calculate the initial landslide time prediction to obtain the current landslide time prediction;

[0026] Landslide next moment prediction module: used to calculate the current landslide prediction time to obtain the next moment landslide prediction time

[0027] The beneficial effects of the present invention are:

[0028] By arranging the real-time monitoring data in time series, the weight of the latest monitoring data is increased. The landslide deformation rate smooth curve is obtained by calculating and drawing the real-time monitoring data. The tangent angle change curve and the landslide velocity inverse change curve are then calculated and drawn. Then, based on the tangent angle change curve and the landslide velocity inverse change curve, with the tangent angle of 80° as the prediction starting point, 100 groups of data are selected for expansion and linear fitting to obtain the landslide velocity inverse linear fitting curve. The monitoring time when the velocity inverse value in the landslide velocity inverse linear fitting curve is 0 is used as the preliminary prediction time of the landslide. In this way, the situation that the landslide enters the acceleration stage, the deformation rate continues to increase, the velocity inverse value accelerates to 0, and the tangent angle increases sharply and approaches 90° can be identified in time. This step reduces the influence of error data on the velocity inverse linear fitting, thereby reducing the difference between the landslide prediction time and the actual occurrence time, making the landslide occurrence time prediction more accurate.

[0029] By selecting 8 groups of the latest preliminary prediction times for landslides and sorting them by size, removing the first, third and second largest preliminary prediction times for landslides, and then calculating the mean of the remaining preliminary prediction times for landslides, the current prediction time for landslides is obtained; then the current prediction time for landslides is used as one of the latest 8 groups of preliminary prediction times for landslides at the next moment, and the prediction time for landslides at the next moment is calculated again. This method realizes the combination of the latest monitoring data and the existing data, so that the prediction time for landslides is constantly close to the actual occurrence time, thereby updating the prediction time for landslide occurrence in real time, which has a certain real-time predictability, provides a reliable time reference for landslide emergency response, and further improves the accuracy of landslide occurrence time prediction.

[0030] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 A flow chart of an improved landslide imminent sliding time prediction method provided by the present invention;

[0032] Figure 2 A schematic diagram of an improved landslide imminent sliding time prediction system provided by the present invention;

[0033] Figure 3 A flow chart for predicting the time of landslide imminent sliding provided by the present invention;

[0034] Figure 4 The landslide imminent sliding time prediction start calculation diagram provided by the present invention;

[0035] Figure 5 A smooth curve diagram of landslide deformation rate provided by the present invention;

[0036] Figure 6 The inverse change curve diagram of the landslide velocity provided by the present invention;

[0037] Figure 7 A comparison chart of the time difference between the common landslide prediction and improved method provided by the present invention and the actual occurrence time of the landslide;

[0038] Figure 8 This is a practical application diagram of an improved landslide imminent sliding time prediction system and method provided by the present invention. DETAILED DESCRIPTION

[0039] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0040] In the description of the present invention, it is necessary to understand that the terms "upper", "lower", "front", "back", "left", "right", "top", "bottom", "inside", "outside", etc., indicating the orientation or position relationship are based on the orientation or position relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention.

[0041] Example 1: Combination Figure 1-Figure 8 This embodiment is described.

[0042] This embodiment is an improved method for predicting the time of landslide, including:

[0043] Step S1: acquiring real-time monitoring data, and arranging the real-time monitoring data in time sequence;

[0044] Step S2: Calculating and drawing lines on the arranged real-time monitoring data to obtain a smooth curve of landslide deformation rate;

[0045] The latest 10 sets of real-time monitoring data are used as an interval to calculate multiple deformation rates, and the formula is: V t =ΔS / ΔT, where V t is the deformation rate at the latest moment, ΔS is the difference between the latest monitored displacement and the earliest monitored displacement in the interval, and ΔT is the difference between the end time and the start time of the calculation interval;

[0046] Select the latest first 10 groups of deformation rates in turn, use the mean method to smooth the selected deformation rates to obtain the smoothed deformation rates at multiple latest monitoring moments, the formula is V = (V t-1 +V t-2 +…+Vt-N ) / V, where: V t-N is the Nth value of the latest deformation rate, N is the total number of monitoring data in the interval, N = 10;

[0047] The smoothed deformation rates are connected and drawn to obtain a smooth curve of landslide deformation rate.

[0048] As the real-time monitoring data is updated, the real-time monitoring data is arranged in chronological order, with the latest 10 sets of real-time monitoring data as an interval, using the formula V t =ΔS / ΔT. Divide the displacement and time of the latest and earliest monitoring data in the interval to calculate the deformation rate at the latest moment. V t is the latest deformation rate, ΔS is the difference between the latest monitored displacement and the earliest monitored displacement in the interval, and ΔT is the difference between the end and start time of the calculation interval.

[0049] Since the monitoring data will be updated in real time, the deformation rate will also change continuously. The latest top 10 groups of deformation rates are selected and the average method V = (V t-1 +V t-2 +…+V t-N ) / N to calculate the mean value, which is used as the smoothed deformation rate at the latest monitoring time. The subsequent smoothed deformation rates are calculated in this way, and the smoothed deformation rates are connected and plotted to obtain the smoothed curve of landslide deformation rate.

[0050] Step S3: Calculate the tangent angle by the formula α = arctan (V / V P ) The landslide deformation rate smooth curve is dimensionless to obtain multiple tangent angles, where: V is the smoothed deformation rate, V P is the deformation rate of the uniform deformation stage of the landslide, α is the tangent angle, and its value range is 0-90°; a plurality of the tangent angles are connected and drawn to obtain a tangent angle variation curve.

[0051] The landslide deformation rate smoothing curve is dimensionless, and the tangent angle calculation formula α=arctan(V / V P ) The tangent angles at different times are obtained by calculating the smooth curve of landslide deformation rate, and the tangent angles at different times are connected and plotted to obtain the tangent angle variation curve, so that the imminent sliding time of various landslides can be predicted under the same characteristic standard.

[0052] Step S4: Calculate the landslide deformation rate smooth curve by the velocity inverse formula 1 / V=T / S to obtain a plurality of velocity inverse values, wherein V represents the deformation rate after smoothing, T represents time, and S represents displacement; connect the plurality of velocity inverse values ​​and draw a line to obtain a landslide velocity inverse change curve.

[0053] The landslide deformation rate smooth curve is calculated according to the velocity inverse formula 1 / V=T / S to obtain the velocity inverse value at the latest monitoring time, and the velocity inverse values ​​at the latest monitoring time are connected and drawn to obtain the landslide velocity inverse change curve.

[0054] Step S5: Based on the tangent angle change curve and the landslide velocity inverse change curve, with the tangent angle of 80° as the prediction starting point, 100 groups of data are selected for expansion and linear fitting to obtain the landslide velocity inverse linear fitting curve, and the monitoring time when the velocity inverse value of the landslide velocity inverse linear fitting curve is 0 is used as the preliminary prediction time of the landslide, and this step is repeated to obtain multiple preliminary prediction times of the landslide.

[0055] When the tangent angle of the landslide exceeds 80°, its deformation rate will be further accelerated, and the landslide will enter an irreversible stage. The closer the tangent angle is to 90°, the greater the probability of landslide occurrence, and the closer the calculated imminent slip time is to the actual time of landslide occurrence. Therefore, based on the tangent angle change curve and the landslide velocity inverse change curve, with the tangent angle of 80° as the prediction starting point, 100 sets of data are selected, including: velocity inverse value and corresponding monitoring time. The last 10 sets of data of the 100 sets of data are copied 3 times and added to the previous 100 sets of data to obtain 130 sets of data. The 130 sets of data are linearly analyzed. The linear fitting curve of the inverse of the landslide velocity is obtained by linear fitting. The value of the linear fitting curve of the inverse of the landslide velocity is 0 as the preliminary prediction time of the landslide, and this step is repeated to obtain multiple preliminary prediction times of the landslide. Through this step, it is possible to timely identify the situation where the landslide enters the acceleration stage, the deformation rate continues to increase, the inverse value of the velocity accelerates to 0, and the tangent angle increases sharply and approaches 90°. This step also reduces the influence of the error data on the linear fitting of the inverse velocity, thereby reducing the difference between the predicted time of the landslide and the actual occurrence time, making the prediction of the landslide occurrence time more accurate.

[0056] Step S6: Select 8 groups of the latest landslide preliminary prediction times and sort them by size, remove the first, third and second smallest landslide preliminary prediction times, calculate the mean of the remaining landslide preliminary prediction times, and obtain the current landslide preliminary prediction time;

[0057] Step S7: taking the current landslide prediction time as one of the latest eight groups of landslide preliminary prediction times at the next moment, and repeating step S6 to obtain the landslide prediction time at the next moment.

[0058] As time goes by, the monitoring data will be continuously updated, and the preliminary prediction time of landslide will also change accordingly. Multiple different preliminary prediction times of landslide will be calculated. First, the latest 8 groups of preliminary prediction times of landslide will be sorted by size. Secondly, in order to further smooth the fitting curve and reduce the problem of large difference between the previous and next prediction times, the first, third and second smallest preliminary prediction times of landslide will be removed. Finally, the mean of the remaining 5 groups of preliminary prediction times of landslide will be calculated to obtain the current landslide prediction time, that is, the landslide prediction time at the latest monitoring moment.

[0059] The landslide prediction time at the latest monitoring moment replaces the latest group of preliminary landslide prediction times as one of the latest 8 groups of preliminary landslide prediction times at the next moment. The latest 8 groups of preliminary landslide prediction times at the next moment include a group of current landslide prediction times and 7 groups of latest preliminary landslide prediction times. The new 8 groups of preliminary landslide prediction times are sorted again, the first, third and second smallest preliminary landslide prediction times are removed, the remaining preliminary landslide prediction times are averaged, and the landslide prediction time at the next moment is obtained. This method realizes the combination of the latest monitoring data and the existing data, thereby improving the accuracy of landslide occurrence time prediction.

[0060] Embodiment 2:

[0061] The embodiment of the present invention further provides an improved landslide imminent sliding time prediction system, which is used in the above method, comprising:

[0062] Data acquisition module 10: used to acquire real-time monitoring data and arrange the real-time monitoring data in time sequence;

[0063] The landslide deformation rate smooth curve drawing module 20 is used to calculate and draw the arranged real-time monitoring data to obtain a landslide deformation rate smooth curve;

[0064] The latest 10 sets of real-time monitoring data are used as an interval to calculate multiple deformation rates, and the formula is: V t =ΔS / ΔT, where V t is the deformation rate at the latest moment, ΔS is the difference between the latest monitored displacement and the earliest monitored displacement in the interval, and ΔT is the difference between the end time and the start time of the calculation interval;

[0065] Select the latest first 10 groups of deformation rates in turn, use the mean method to smooth the selected deformation rates to obtain the smoothed deformation rates at multiple latest monitoring moments, the formula is V = (V t-1 +V t-2 +…+V t-N ) / N, where: V t-Nis the Nth value of the latest deformation rate, N is the total number of monitoring data in the interval, N = 10;

[0066] The smoothed deformation rates are connected and drawn to obtain a smooth curve of landslide deformation rate.

[0067] Tangent angle variation curve drawing module 30: used to calculate the tangent angle by the formula α=arctan(V / V P ) The landslide deformation rate smoothing curve is dimensionless to obtain multiple tangent angles, and the tangent angles are connected and drawn to obtain a tangent angle change curve; wherein: V is the smoothed deformation rate, V P is the deformation rate of the landslide in the uniform deformation stage, α is the tangent angle, and its value range is 0 to 90°;

[0068] The landslide velocity reciprocal change curve drawing module 40 is used to calculate the landslide deformation rate smooth curve by the velocity reciprocal formula 1 / V=T / S to obtain a plurality of velocity reciprocal values, and connect the plurality of velocity reciprocal values ​​to obtain the landslide velocity reciprocal change curve, wherein V represents the deformation rate after smoothing, T represents time, and S represents displacement;

[0069] Landslide preliminary time prediction module 50: used to make a preliminary prediction of the landslide occurrence time to obtain a preliminary predicted time of the landslide;

[0070] Based on the tangent angle change curve and the landslide velocity inverse change curve, with the tangent angle of 80° as the prediction starting point, 100 sets of data are selected, the data including: velocity inverse value and the corresponding monitoring time, the last 10 sets of data of the 100 sets of data are copied 3 times and added to the previous 100 sets of data to obtain 130 sets of data, and the 130 sets of data are linearly fitted to obtain the landslide velocity inverse linear fitting curve, and the value of the landslide velocity inverse linear fitting curve is 0 as the preliminary prediction time of the landslide.

[0071] The current landslide time prediction module 60 is used to calculate the initial landslide time prediction to obtain the current landslide time prediction;

[0072] The eight latest groups of the preliminary prediction times for landslides are selected and sorted by size, the first, third and second smallest preliminary prediction times for landslides are removed, and the mean of the remaining preliminary prediction times for landslides is calculated to obtain the current preliminary prediction time for landslides.

[0073] The next moment prediction module 70 of the landslide is used to calculate the current predicted time of the landslide to obtain the next moment prediction time of the landslide;

[0074] The current landslide prediction time is taken as one of the latest 8 groups of landslide preliminary prediction times at the next moment, and then the landslide current time prediction module is called to calculate the current landslide prediction time again to obtain the landslide prediction time at the next moment.

[0075] The specific application of the present invention is as follows:

[0076] like Figure 5-7 As shown in the figure, at 5:53 am on February 17, 2019, a bedding rock landslide occurred in Group 9, Longjing Village, Xingyi City, Qianxinan Prefecture, Guizhou Province. Through the analysis of the acquired data, it can be seen that from the monitoring on February 1, 2019 to 17:30 on February 7, the tangent angle of the landslide was less than 45°, and it was in the uniform deformation stage. From 18:00 on February 7, the tangent angle exceeded 45° and reached 80° at 11:00 on the 16th, indicating that the landslide entered the medium acceleration stage. The improved landslide imminent sliding time prediction system provided by the present invention began to predict the landslide imminent sliding time. Then, the predicted time of the landslide was compared with the actual time of occurrence, and it was found that the predicted time of the landslide fluctuated greatly in the early prediction process. Before 16:00 on the 16th, the maximum difference reached -2.72 (about 65 hours), but then the difference gradually narrowed. Until 10 minutes before the landslide, the difference fluctuation basically remained between 0.01 and 0.34 (about 15 minutes to 8 hours), which was significantly better than the common prediction methods.

[0077] like Figure 8 As shown, Figure 8 As shown in the figure, a landslide occurred in Heifangtai, Gansu Province. At 4:59 on March 26, 2019, the cumulative displacement-time curve of the whole process of Dangchuan 6# landslide was successfully obtained through the deployed monitoring equipment. It was found that the tangent angle of the landslide reached 80° for the first time at 8:01 on March 24, 2019. Subsequently, an improved landslide prediction method provided by the present invention was used to predict the imminent sliding time of the landslide. The prediction difference gradually decreased from the maximum of -9.96 with the passage of monitoring time and the increase of monitoring data. The final prediction time was only -0.07 different from the actual occurrence of the landslide.

[0078] The above are the preferred implementations expressed in the present invention, but are not limited thereto. For ordinary technicians in this field, based on the above implementations, they can understand the ideas of the present invention and make different extensions and changes. However, as long as they do not deviate from the concept of the present invention, they are all within the protection scope of the present invention.

Claims

1. An improved method for predicting the time of landslide, characterized in that: include: Step S1: acquiring real-time monitoring data, and arranging the real-time monitoring data in time sequence; Step S2: Calculating and drawing lines on the arranged real-time monitoring data to obtain a smooth curve of landslide deformation rate; Step S3: Calculate the tangent angle by the formula α = arctan (V / V P ) The landslide deformation rate smooth curve is dimensionless to obtain multiple tangent angles, where: V is the smoothed deformation rate, V P is the deformation rate of the uniform deformation stage of the landslide, α is the tangent angle, and the value range is 0 to 90°; a plurality of the tangent angles are connected and drawn to obtain a tangent angle variation curve; Step S4: Calculate the landslide deformation rate smooth curve by the velocity inverse formula 1 / V=T / S to obtain a plurality of velocity inverse values, wherein V represents the deformation rate after smoothing, T represents time, and S represents displacement; connect the plurality of velocity inverse values ​​and draw a line to obtain a landslide velocity inverse change curve; Step S5: Based on the tangent angle variation curve and the landslide velocity inverse variation curve, with the tangent angle of 80° as the prediction starting point, 100 sets of data are selected for expansion and linear fitting to obtain a landslide velocity inverse linear fitting curve, and the monitoring time when the velocity inverse value in the landslide velocity inverse linear fitting curve is 0 is used as the preliminary prediction time of the landslide, and this step is repeated to obtain multiple preliminary prediction times of the landslide; Step S6: Select 8 groups of the latest landslide preliminary prediction times and sort them by size, remove the first, third and second smallest landslide preliminary prediction times, calculate the mean of the remaining landslide preliminary prediction times, and obtain the current landslide preliminary prediction time; Step S7: taking the current landslide prediction time as one of the latest eight groups of landslide preliminary prediction times at the next moment, and repeating step S6 to obtain the landslide prediction time at the next moment.

2. The improved landslide imminent sliding time prediction method according to claim 1 is characterized in that: The arranged real-time monitoring data are calculated and connected to obtain a smooth curve of landslide deformation rate, including: The latest 10 sets of real-time monitoring data are used as an interval to calculate multiple deformation rates, and the formula is: V t =ΔS / ΔT, where V t is the deformation rate at the latest moment, ΔS is the difference between the latest monitored displacement and the earliest monitored displacement in the interval, and ΔT is the difference between the end time and the start time of the calculation interval; Select the latest first 10 groups of deformation rates in turn, use the mean method to smooth the selected deformation rates to obtain the smoothed deformation rates at multiple latest monitoring moments, the formula is V = (V t-1 +V t-2 +…+Vx -N ) / N, where: V t-N is the Nth value of the latest deformation rate, N is the total number of monitoring data in the interval, N = 10; The smoothed deformation rates are connected and drawn to obtain a smooth curve of landslide deformation rate.

3. The improved landslide imminent sliding time prediction method according to claim 2 is characterized in that: Based on the tangent angle change curve and the landslide velocity inverse change curve, taking the tangent angle of 80° as the prediction starting point, 100 sets of data were selected for expansion and linear fitting to obtain the landslide velocity inverse linear fitting curve, including: The first 100 data sets with tangent angles of 80° and above were selected, the last 10 data sets of the 100 data sets were copied 3 times and added to the 100 data sets to obtain 130 data sets, and the 130 data sets were linearly fitted to obtain the inverse linear fitting curve of landslide velocity.

4. An improved landslide imminent sliding time prediction system, used in the method according to any one of claims 1 to 3, characterized in that: include: Data acquisition module: used to acquire real-time monitoring data and arrange the real-time monitoring data in time sequence; A landslide deformation rate smooth curve drawing module is used to calculate and draw the arranged real-time monitoring data to obtain a landslide deformation rate smooth curve; Tangent angle change curve drawing module: used to calculate the tangent angle through the formula α = arctan (V / V P ) The landslide deformation rate smoothing curve is dimensionless to obtain multiple tangent angles, and the tangent angles are connected and drawn to obtain a tangent angle change curve; wherein: V is the smoothed deformation rate, V P is the deformation rate of the landslide in the uniform deformation stage, α is the tangent angle, and its value range is 0 to 90°; The landslide velocity reciprocal change curve drawing module is used to calculate the landslide deformation rate smooth curve through the velocity reciprocal formula 1 / V=T / S to obtain multiple velocity reciprocal values, and connect the multiple velocity reciprocal values ​​to obtain the landslide velocity reciprocal change curve, wherein V represents the deformation rate after smoothing, T represents time, and S represents displacement; Landslide Preliminary Time Prediction Module: used to make a preliminary prediction of the landslide occurrence time to obtain the preliminary predicted time of the landslide; The current landslide time prediction module is used to calculate the initial landslide time prediction to obtain the current landslide time prediction; The landslide next moment prediction module is used to calculate the current landslide prediction time to obtain the next moment landslide prediction time.

Images