Power monitoring method for measuring landslide stability by using rainfall

Abstract

The invention discloses a dynamic monitoring method for measuring landslide stability by using rainfall capacity, and relates to the technical field of landslide monitoring, and the method comprises environmental factor analysis, rainfall landslide analysis and landslide stability early warning: firstly, through analyzing non-rainfall environmental data and non-rainfall landslide data, determining each environmental type and each landslide stability grade; setting a sensor placement scheme according to the data, then collecting rainfall data, rainfall landslide data and rainfall environment data during rainfall, and analyzing the data based on corresponding environmental landslide stability levels to obtain actual landslide data and rainfall landslide stability levels corresponding to respective rainfall types; and then, according to the current landslide data, the current rainfall data and the current environment data, the data are analyzed to obtain a predicted landslide stability level, and according to the actual landslide level and the predicted landslide stability level, early warning processing is performed, so that the early warning accuracy is improved.

Description

Technical Field

[0001] The invention relates to the technical field of landslide monitoring, and in particular to a dynamic monitoring method for measuring landslide stability by utilizing rainfall. Background Art

[0002] Landslide is a very destructive geological disaster. It can destroy buildings, roads, bridges and other infrastructure in a short period of time, causing casualties and huge economic losses. Rainfall is one of the important factors that trigger many landslides. When rainfall reaches a certain level, the water content of the mountain soil increases, the shear strength of the soil decreases, and the mountain is prone to instability and landslides. Therefore, by using the dynamic monitoring method of measuring landslide stability using rainfall, early warning of possible landslides can be given.

[0003] Prior art, such as the invention patent application with announcement number CN103149340B, discloses a dynamic monitoring method for determining landslide stability using rainfall, taking the ratio of the incremental change of rainfall in a certain rainfall process to the initial rainfall as the dynamic loading rate of landslide instability, taking the ratio of the change of the corresponding landslide displacement rate to the initial displacement rate as the displacement dynamic response rate of the sliding body, defining the ratio of the dynamic loading rate of landslide instability to the displacement dynamic response rate as the landslide displacement dynamic loading rate, taking the displacement rainfall dynamic loading rate of the landslide as the criterion of landslide stability, determining the rainfall dynamic loading rate prediction parameters of the landslide, establishing a landslide stability evaluation model based on the quantitative relationship between the rainfall dynamic loading rate prediction parameters and the landslide stability, and evaluating and predicting the stability of the landslide. This method synchronously monitors rainfall and landslide displacement or displacement rate, has an accurate and unified criterion for predicting landslides, and provides an effective basis for landslide prediction, forecasting and early warning management.

[0004] The above scheme has at least the following shortcomings: the above scheme simply analyzes the relationship between rainfall and landslide data, and does not take into account the impact of environmental factors on landslide stability. Under the same rainfall conditions, different groundwater levels, internal friction angles and cohesion will lead to different landslide stabilities. The above scheme does not reduce the impact of environmental factors in the analysis of rainfall and landslide stability, which makes the above scheme more susceptible to interference from environmental factors and reduces the accuracy of the analysis. Summary of the invention

[0005] In view of the above-mentioned technical deficiencies, the object of the present invention is to provide a dynamic monitoring method for measuring landslide stability using rainfall.

[0006] To solve the above technical problems, the present invention adopts the following technical solutions: The present invention provides a safety inspection method based on artificial intelligence, comprising the following steps: Step 1, environmental factor analysis: Collect pre-rainfall environmental data and pre-rainfall landslide data, analyze the pre-rainfall environmental data and pre-rainfall landslide data based on the environmental landslide model, obtain the environmental landslide stability level corresponding to each environmental type, collect the pre-rainfall environmental data corresponding to each collection area in each current slope body, and set a sensor placement plan based on the pre-rainfall environmental data corresponding to each collection area in each current slope body.

[0007] Step 2: Rainfall-landslide analysis: During rainfall, rainfall data, rainfall-landslide data and rainfall environment data are collected. Based on the environmental landslide stability level corresponding to each environmental type, the rainfall environment data and rainfall-landslide data are analyzed to obtain actual landslide data. Based on the rainfall-landslide model, the rainfall data and actual landslide data are analyzed to obtain the rainfall-landslide stability level corresponding to each rainfall type.

[0008] Step 3: Landslide stability warning: Collect current landslide data, current rainfall data and current environmental data, analyze the current landslide data to obtain the actual landslide grade, analyze the current rainfall data and current environmental data based on the environmental landslide stability grade of each environmental type and the rainfall landslide stability grade corresponding to each rainfall type, obtain the predicted landslide stability grade, and perform warning processing based on the actual landslide grade and the predicted landslide stability grade.

[0009] Preferably, the environmental data and pre-rainfall landslide data are analyzed, and the specific analysis process is as follows: the pre-rainfall environmental data include the groundwater level, internal friction angle and cohesion corresponding to each collection area in each slope collected each time without rainfall, and the pre-rainfall landslide data include the inclination change and displacement change corresponding to each collection area in each slope collected each time without rainfall. The groundwater level, internal friction angle and cohesion corresponding to each collection area in each slope collected each time without rainfall are respectively calculated with the inclination change and displacement change to obtain the weight factor of the groundwater level, the weight factor of the internal friction angle and the weight factor of the cohesion.

[0010] Based on the weight factors of groundwater level height, internal friction angle and cohesion, the groundwater level, internal friction angle and cohesion corresponding to each collection area in each slope collected without rainfall are weightedly calculated to obtain the environmental coefficient of each collection area in each slope collected without rainfall. If the environmental coefficient of a collection area in a slope collected without rainfall belongs to the environmental coefficient interval corresponding to a certain environmental type, it indicates that the environmental type of the collection area in the slope collected without rainfall is the environmental type. In this way, the environmental type of each collection area in each slope collected without rainfall is obtained. The inclination change and displacement change corresponding to each collection area in each slope collected without rainfall are summarized according to the environmental type corresponding to each collection area in each slope collected without rainfall to obtain each inclination change and each displacement change corresponding to each environmental type.

[0011] The average values ​​of the inclination changes and displacement changes corresponding to each environmental type are calculated respectively to obtain the average inclination changes and average displacement changes corresponding to each environmental type. The maximum inclination change, average inclination change, maximum displacement change and average displacement change corresponding to each environmental type are input into the environmental landslide model to obtain the output results of each environmental type. The value i of the output result is the environmental landslide stability grade, i=1,2...u, u>2, and u is the maximum environmental landslide stability grade.

[0012] The beneficial effects of the present invention are as follows: 1. First, by analyzing the pre-rainfall environment data and the pre-rainfall landslide data, each environment type and each landslide stability level are determined, and a sensor placement plan is set accordingly; then, during the rainfall period, rainfall data, rainfall landslide data and rainfall environment data are collected; based on the corresponding environment landslide stability level, the data are analyzed to obtain actual landslide data and rainfall landslide stability levels corresponding to the respective rainfall types; then, based on the current landslide data, the current rainfall data and the current environment data, the data are analyzed to obtain the predicted landslide stability level; and based on the actual landslide level and the predicted landslide stability level, early warning processing is performed, thereby improving the accuracy of the early warning.

[0013] 2. The present invention first analyzes the influence of environmental data on landslide stability when there is no rainfall, obtains the influence of various environmental types on landslide stability, and then analyzes the influence of rainfall data on landslide stability when it rains. At the same time, in the process of rainfall data analysis, the influence of environmental type on landslide stability is eliminated, which increases the effectiveness of data analysis. At the same time, in the early warning analysis process of landslide early warning, both rainfall data and environmental data are analyzed, which increases the reliability of early warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0015] Figure 1 The present invention is a schematic flow chart of the steps for implementing the method. DETAILED DESCRIPTION

[0016] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0017] Step 1: Environmental factor analysis: Collect pre-rainfall environmental data and pre-rainfall landslide data. Based on the environmental landslide model, analyze the pre-rainfall environmental data and pre-rainfall landslide data to obtain the environmental landslide stability level corresponding to each environmental type. Collect the pre-rainfall environmental data corresponding to each collection area in each current slope body. Based on the pre-rainfall environmental data corresponding to each collection area in each current slope body, set up a sensor placement plan.

[0018] In a specific embodiment, the data on the environment and landslides before rainfall are collected, and the specific collection process is as follows: the groundwater level corresponding to each collection area in each slope is collected by a pressure water level sensor, the internal friction angle corresponding to each collection area in each slope is collected by a direct shear test, and the cohesion corresponding to each collection area in each slope is collected by a cross plate shear test, so as to obtain the groundwater level, internal friction angle and cohesion corresponding to each collection area in each slope collected before rainfall, collect the current inclination of each collector by an inclinometer, obtain the inclination of each collector collected last time in history from a database, and the absolute value of the value obtained by subtracting the inclination of the last collection in history from the current inclination of each collector is the inclination change of each collector, and the current position of each marker point is collected by a total station measurement method using a total station, and a three-dimensional model is established according to the last historical position of each marker point in the database, the current position and the last historical position of each marker point, and the displacement change of each marker point is obtained, so as to obtain the inclination change and displacement change corresponding to each collection area in each slope collected before rainfall.

[0019] In a specific embodiment, the environmental data and pre-rainfall landslide data are analyzed, and the specific analysis process is as follows: the pre-rainfall environmental data include the groundwater level, internal friction angle and cohesion corresponding to each collection area in each slope collected without rainfall, and the pre-rainfall landslide data include the inclination change and displacement change corresponding to each collection area in each slope collected without rainfall. The groundwater level, internal friction angle and cohesion corresponding to each collection area in each slope collected without rainfall are respectively calculated with the inclination change and displacement change to obtain the weight factor of the groundwater level, the weight factor of the internal friction angle and the weight factor of the cohesion.

[0020] Based on the weight factors of groundwater level height, internal friction angle and cohesion, the groundwater level, internal friction angle and cohesion corresponding to each collection area in each slope collected without rainfall are weightedly calculated to obtain the environmental coefficient of each collection area in each slope collected without rainfall. If the environmental coefficient of a collection area in a slope collected without rainfall belongs to the environmental coefficient interval corresponding to a certain environmental type, it indicates that the environmental type of the collection area in the slope collected without rainfall is the environmental type. In this way, the environmental type of each collection area in each slope collected without rainfall is obtained. The inclination change and displacement change corresponding to each collection area in each slope collected without rainfall are summarized according to the environmental type corresponding to each collection area in each slope collected without rainfall to obtain each inclination change and each displacement change corresponding to each environmental type.

[0021] The average values ​​of the inclination changes and displacement changes corresponding to each environment type are calculated respectively to obtain the average inclination change and average displacement change corresponding to each environment type. The maximum inclination change, average inclination change, maximum displacement change and average displacement change corresponding to each environment type are input into the environmental landslide model to obtain the output results of each environment type. The value i of the output result is the environmental landslide stability grade, i=1,2...u, u>2, and u is the maximum environmental landslide stability grade.

[0022] In a specific embodiment, the environmental landslide model expression is:

[0023] Among them, α a is the output result of environment type a, where a is the number of the environment type, a=1,2......m, m>2, A″ a , A′ a , B″ a and B′ aare the maximum tilt change, average tilt change, maximum displacement change and average displacement change corresponding to environment type a, respectively; A″, A′, B″ and B′ are the preset standard maximum tilt change, standard tilt change, standard maximum displacement change and standard displacement change, respectively; ε1 and ε2 are the preset tilt weight factor and displacement weight factor, respectively; ε1>0, ε2>0, ε1+ε2=1, N1, N2, N i-1 、N i and N u They are respectively the preset first environmental landslide stability index, the second environmental landslide stability index, the i-1th environmental landslide stability index, the i-th environmental landslide stability index and the u-th environmental landslide stability index.

[0024] It should be noted that the standard parameters A″, A′, B″, B′, N1, N2, N i-1 、N i and N u They are all set by the staff, for example, A″ is 1.2, A′ is 0.7, B″ is 1.8, B′ is 1.1, N1 is 0.7, N2 is 2.1, N i-1 6.1, N i 7.3 and N u is 9.4, and the weight factors ε1 and ε2 are set by the staff, for example, ε1 is 0.6 and ε2 is 0.4.

[0025] In a specific embodiment, the sensor placement scheme is set up, and the specific setting process is as follows: the environmental data corresponding to each collection area in each current slope body is converted to obtain the environmental type of each collection area in each current slope body, and then the landslide stability corresponding to the environmental type of each collection area in each current slope body is obtained, and each collection area in the current slope body whose landslide stability is less than the preset landslide stability is recorded as a dangerous collection area, and a rainfall sensor and a landslide data collector are set in each dangerous collection area in each slope body.

[0026] It should be noted that the environmental data corresponding to each collection area in each current slope body include the groundwater level, internal friction angle and cohesion corresponding to each collection area in each current slope body. The "conversion" is to perform weighted calculation on the groundwater level, internal friction angle and cohesion corresponding to each collection area in each current slope body through the weight factor of the groundwater level, the weight factor of the internal friction angle and the weight factor of the cohesion to obtain the environmental coefficient corresponding to each collection area in each current slope body. If the environmental coefficient of a collection area in a certain slope body belongs to the environmental coefficient interval corresponding to a certain environmental type, it indicates that the environmental type of the collection area in the current slope body is the environmental type. In this way, the environmental data corresponding to each collection area in each current slope body are converted to obtain the environmental type of each collection area in the current slope body.

[0027] Step 2: Rainfall-landslide analysis: During rainfall, rainfall data, rainfall-landslide data and rainfall environment data are collected. Based on the environmental landslide stability level corresponding to each environmental type, the rainfall environment data and rainfall-landslide data are analyzed to obtain actual landslide data. Based on the rainfall-landslide model, the rainfall data and actual landslide data are analyzed to obtain the rainfall-landslide stability level corresponding to each rainfall type.

[0028] In a specific embodiment, the rainfall data, rainfall-landslide data and rainfall-environmental data are collected, and the specific collection process is as follows: the collection area with the minimum landslide stability in each slope is recorded as the main collection area, and when collecting data on each slope, the rainfall data, rainfall-landslide data and rainfall-environmental data of each dangerous collection area are collected, the initial rainfall, the average rainfall increase and the maximum rainfall increase are collected by a tipping bucket rain gauge, and the rainfall-landslide data and rainfall-environmental data are collected by a landslide data and environmental data collection method, so as to obtain the rainfall data, rainfall-landslide data and rainfall-environmental data corresponding to each dangerous collection area in each slope, and weights are assigned according to the regional distance between each dangerous collection area and the main collection area in each slope, and the collected data of each collection area in each slope is weightedly calculated to obtain the collected data of each slope, so as to obtain the initial rainfall, the average rainfall increase and the maximum rainfall increase corresponding to each slope in each rainfall, and also obtain the inclination change and the displacement change, and also obtain the groundwater level height, internal friction angle and cohesion corresponding to each slope in each rainfall.

[0029] It should be noted that when weighting is performed according to the distance between each dangerous collection area and the main collection area in each slope, the farther the distance between the dangerous collection area and the main collection area, the smaller the weight factor.

[0030] In a specific embodiment, the rainfall environment data and rainfall landslide data are analyzed, and the specific analysis process is as follows: the rainfall environment data includes the groundwater level, internal friction angle and cohesion corresponding to each slope body in each rainfall, and the groundwater level, internal friction angle and cohesion corresponding to each slope body in each rainfall are weightedly calculated to obtain the environmental coefficient corresponding to each slope body in each rainfall. If the environmental coefficient corresponding to a slope body in a certain rainfall belongs to the environmental coefficient interval corresponding to a certain environmental type, it indicates that the environmental type corresponding to the slope body in this rainfall is the environmental type. According to the environmental landslide stability grade corresponding to each environmental type, the environmental landslide stability corresponding to each slope body in each rainfall is obtained, and the environmental landslide stability of each environmental type is obtained from the database. The slope stability corresponding to each inclination change correction factor and each displacement change correction factor are used to obtain the inclination change correction factor and displacement change correction factor corresponding to each slope body for each rainfall. The rainfall landslide data include the inclination change and displacement change corresponding to each slope body for each rainfall. The inclination change corresponding to each slope body for each rainfall is multiplied by the inclination change correction factor to obtain the actual inclination change corresponding to each slope body for each rainfall. The displacement change corresponding to each slope body for each rainfall is multiplied by the displacement change correction factor to obtain the actual displacement change corresponding to each slope body for each rainfall. The actual rainfall landslide data include the actual inclination change and actual displacement change corresponding to each slope body for each rainfall.

[0031] In a specific embodiment, the rainfall data and actual landslide data are analyzed, and the specific analysis process is as follows: the rainfall data includes the initial rainfall, average rainfall increase and maximum rainfall increase corresponding to each slope body in each rainfall, and the initial rainfall, average rainfall increase and maximum rainfall increase corresponding to each slope body in each rainfall are respectively calculated with the actual inclination change and actual displacement change corresponding to each slope body by Pearson correlation coefficient to obtain the weight factor of the initial rainfall, the weight factor of the average rainfall increase and the weight factor of the maximum rainfall increase, and the initial rainfall, average rainfall increase and maximum rainfall increase corresponding to each slope body in each rainfall are weightedly calculated to obtain the rainfall of each slope body in each rainfall. Rain coefficient: if the rainfall coefficient of a slope in a certain rainfall belongs to the rainfall coefficient interval corresponding to a certain rainfall type, it indicates that the rainfall type of the slope in this rainfall is this rainfall type. In this way, the rainfall type corresponding to each slope in each rainfall is obtained, and then the actual inclination change and the actual displacement change corresponding to each rainfall type are obtained. At the same time, if the actual inclination change of a slope in a certain rainfall is less than the preset inclination change, and the actual displacement change of the slope collected in this rainfall is less than the preset displacement change, the state of the slope collected in this rainfall is recorded as the non-landslide state, and the number of occurrences of the non-landslide state in each rainfall is obtained by statistics. According to the rainfall type corresponding to each rainfall, the number of non-landslide collections of each rainfall type is obtained.

[0032] The average values ​​of the actual inclination changes and actual displacement changes corresponding to each rainfall type are calculated respectively to obtain the average actual inclination changes and average actual displacement changes corresponding to each rainfall type. The average actual inclination changes, average actual displacement changes, maximum actual inclination changes and maximum actual displacement changes corresponding to each rainfall type are input into the rainfall-landslide model to obtain the output results of each rainfall type. The value j of the output result is the landslide stability grade, j=1,2......v, v>2, and v is the maximum rainfall-landslide stability grade.

[0033] In a specific embodiment, the rainfall landslide model expression is:

[0034] Where βb is the output result of rainfall type b, b is the number of rainfall type, b=1,2......n, n>2, E″ b , E′ b , F″ b and F′ b are the maximum actual tilt change, average actual tilt change, maximum actual displacement change and average actual displacement change corresponding to environment type b, E″, E′, F″ and F′ are the preset standard maximum actual tilt change, standard actual tilt change, standard maximum actual displacement change and standard actual displacement change, respectively; C b is the number of non-landslide acquisitions of the preset rainfall type b, C′ is the preset standard non-landslide acquisition number, φ1 and φ2 are the preset weight factors of rainfall inclination and rainfall displacement, φ1>0, φ2>0, φ1+φ2=1, M1, M2, M j-1 、M j and M v They are respectively the preset first rainfall landslide stability index, the second rainfall landslide stability index, the j-1th rainfall landslide stability index, the jth rainfall landslide stability index and the vth rainfall landslide stability index.

[0035] It should be noted that the standard parameters E″, E′, F″, F′, C′, M1, M2, M j-1 、M j and M v The setting process is the same as the setting process of the standard parameter A″, for example, E″ is 1.9, E′ is 0.7, F″ is 2.1, F′ is 0.9, C′ is 8, M1 is 1.2, M2 is 1.3, M j-1 5.6, M j 5.9 and M vis 8.2, and the setting process of the weight factors φ1 and φ2 is the same as the setting process of the above-mentioned weight factor ε1, for example, φ1 is 0.34 and φ2 is 0.66.

[0036] Step 3: Landslide stability warning: Collect current landslide data, current rainfall data and current environmental data, analyze the current landslide data to obtain the actual landslide grade, analyze the current rainfall data and current environmental data based on the environmental landslide stability grade of each environmental type and the rainfall landslide stability grade corresponding to each rainfall type, obtain the predicted landslide stability grade, and perform warning processing based on the actual landslide grade and the predicted landslide stability grade.

[0037] In a specific embodiment, the current landslide data, current rainfall data and current environmental data are collected, and the specific collection process is as follows: the current rainfall data is collected by a rainfall data collection method, the current rainfall data is converted to obtain the current rainfall type, and the landslide data collection period and environmental data collection period corresponding to the current rainfall type are obtained from a database; according to the current landslide data collection period, the current landslide data is collected by a landslide data collection method; according to the current environmental data collection period, the current environmental data is collected by an environmental data collection method.

[0038] In a specific embodiment, the landslide data collected each time within a preset time length are analyzed, and the specific analysis process is as follows: the landslide data collected each time within the preset time length include the inclination change collected each time corresponding to each current slope body, the displacement change collected each time, the total inclination change and the total displacement change, the inclination change collected each time corresponding to each current slope body, the displacement change collected each time, the total inclination change and the total displacement change are input into the actual landslide model to obtain the output result of the actual landslide model, and the value k of the output result is the actual landslide grade, k=1,2...w, w>2, w is the maximum actual landslide grade.

[0039] In a specific embodiment, the actual landslide model expression is:

[0040] Among them, γ is the output result of the actual landslide model, ΔS′ cd and ΔL′ cd are the slope change and displacement change corresponding to the current slope c for the d times of acquisition, ΔS″ c and ΔL″ care the total changes in inclination and displacement corresponding to the current slope c, ΔS′, ΔL′, ΔS″ and ΔL″ are the preset single standard inclination changes, single standard displacement changes, total standard inclination changes and total standard displacement changes, φ1 and φ2 are the preset rainfall inclination weight factors and rainfall displacement weight factors, respectively. is the preset weight factor of the current slope c, Q1, Q2, Q k-1 , Q k and Q w They are respectively the preset first landslide assessment index, the second landslide assessment index, the k-1th landslide assessment index, the kth landslide assessment index and the wth landslide assessment index.

[0041] It should be noted that the standard parameters ΔS′, ΔL′, ΔS″, ΔL″, Q1, Q2, Q k-1 , Q k and Q w The setting process is the same as the setting process of the standard parameter A″, for example, ΔS′ is 0.7, ΔL′ is 0.8, ΔS″ is 1.2, ΔL″ is 2.3, Q1 is 0.9, Q2 is 1.1, Q k-1 3.4, Q k 3.5 and Q w is 8.2, the weight factor The setting process of is the same as the setting process of the weight factor ε1 above, for example is 0.12.

[0042] In a specific embodiment, the current rainfall data and the current environmental data are analyzed, and the specific analysis process is as follows: the current rainfall data is converted to obtain the current rainfall type, based on the rainfall landslide stability level corresponding to each rainfall type, the current rainfall landslide stability level corresponding to the current rainfall type is obtained, the current environmental data is converted to obtain the current environmental type, based on the environmental landslide stability level of each environmental type, the current environmental landslide stability level corresponding to the current environmental type is obtained, the current rainfall landslide stability level and the current environmental landslide stability level are weightedly calculated to obtain a preset landslide stability level, and the preset landslide level corresponding to the preset landslide stability level is obtained from the database.

[0043] In a specific embodiment, the early warning processing is performed as follows: if the actual landslide level is greater than a preset standard actual landslide level or the preset landslide level is greater than a preset standard preset landslide level, an early warning prompt is issued, the standard actual landslide level and the preset landslide level are substituted into the early warning prompt index calculation formula to obtain the current early warning index, the early warning index interval corresponding to each early warning level is obtained from the database, the early warning level in the early warning index interval in which the current early warning index is located is recorded as the current early warning level, the early warning plan corresponding to the current early warning level is obtained from the database, recorded as the current early warning plan, and an early warning is issued according to the current early warning plan.

[0044] It should be noted that the early warning plan, for example: issuing the highest level of early warning information through multiple channels, notifying relevant personnel to immediately evacuate surrounding residents, block dangerous areas and temporarily protect important facilities.

[0045] The above contents are merely examples and explanations of the concept of the present invention. Those skilled in the art may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the concept of the invention or exceed the scope defined in this specification, they shall all fall within the protection scope of the present invention.

Claims

1. A dynamic monitoring method for determining landslide stability using rainfall, characterized in that: The steps include: Step 1: Environmental factor analysis: Collect pre-rainfall environmental data and pre-rainfall landslide data, analyze the pre-rainfall environmental data and pre-rainfall landslide data based on the environmental landslide model, obtain the environmental landslide stability level corresponding to each environmental type, collect the pre-rainfall environmental data corresponding to each collection area in each current slope body, and set the sensor placement plan based on the pre-rainfall environmental data corresponding to each collection area in each current slope body; Step 2: Rainfall landslide analysis: during rainfall, rainfall data, rainfall landslide data and rainfall environment data are collected. Based on the environmental landslide stability level corresponding to each environmental type, the rainfall environment data and rainfall landslide data are analyzed to obtain actual landslide data. Based on the rainfall landslide model, the rainfall data and actual landslide data are analyzed to obtain the rainfall landslide stability level corresponding to each rainfall type. Step 3: Landslide stability warning: Collect current landslide data, current rainfall data and current environmental data, analyze the current landslide data to obtain the actual landslide grade, analyze the current rainfall data and current environmental data based on the environmental landslide stability grade of each environmental type and the rainfall landslide stability grade corresponding to each rainfall type, obtain the predicted landslide stability grade, and perform warning processing based on the actual landslide grade and the predicted landslide stability grade.

2. The artificial intelligence-based safety inspection method according to claim 1 is characterized in that: The environmental data and pre-rainfall landslide data are analyzed, and the specific analysis process is as follows: The pre-rainfall environmental data include the groundwater level, internal friction angle and cohesion corresponding to each sampling area in each slope collected at each pre-rainfall period; the pre-rainfall landslide data include the inclination change and displacement change corresponding to each sampling area in each slope collected at each pre-rainfall period. The groundwater level, internal friction angle and cohesion corresponding to each sampling area in each slope collected at each pre-rainfall period are respectively calculated with the inclination change and displacement change for the Pearson correlation coefficient, and the weight factor of the groundwater level, the weight factor of the internal friction angle and the weight factor of the cohesion are obtained; Based on the weight factor of the groundwater level, the weight factor of the internal friction angle and the weight factor of the cohesion, the groundwater level, the internal friction angle and the cohesion corresponding to each collection area in each slope collected without rainfall are weighted and calculated to obtain the environmental coefficient of each collection area in each slope collected without rainfall. If the environmental coefficient of a collection area in a slope collected without rainfall belongs to the environmental coefficient interval corresponding to a certain environmental type, it indicates that the environmental type of the collection area in the slope collected without rainfall is the environmental type. In this way, the environmental type of each collection area in each slope collected without rainfall is obtained. The inclination change and displacement change corresponding to each collection area in each slope collected without rainfall are summarized according to the environmental type corresponding to each collection area in each slope collected without rainfall to obtain each inclination change and each displacement change corresponding to each environmental type. The average values ​​of the inclination changes and displacement changes corresponding to each environment type are calculated respectively to obtain the average inclination change and average displacement change corresponding to each environment type. The maximum inclination change, average inclination change, maximum displacement change and average displacement change corresponding to each environment type are input into the environmental landslide model to obtain the output results of each environment type. The value i of the output result is the environmental landslide stability grade, i=1,2...u, u>2, and u is the maximum environmental landslide stability grade.

3. The artificial intelligence-based safety inspection method according to claim 2 is characterized in that: The environmental landslide model expression is: Among them, α a is the output result of environment type a, where a is the number of the environment type, a=1,2......m, m>2, A″ a , A′ a , B″ a and B′ a are the maximum tilt change, average tilt change, maximum displacement change and average displacement change corresponding to environment type a, respectively; A″, A′, B″ and B′ are the preset standard maximum tilt change, standard tilt change, standard maximum displacement change and standard displacement change, respectively; ε1 and ε2 are the preset tilt weight factor and displacement weight factor, respectively; ε1>0, ε2>0, ε1+ε2=1, N1, N2, N i-1 、N i and N u They are respectively the preset first environmental landslide stability index, the second environmental landslide stability index, the i-1th environmental landslide stability index, the i-th environmental landslide stability index and the u-th environmental landslide stability index.

4. The artificial intelligence-based safety inspection method according to claim 2 is characterized in that: The sensor placement scheme is set up, and the specific setting process is as follows: The environmental data corresponding to each collection area in each current slope body are transformed to obtain the environmental type of each collection area in each current slope body, and then the landslide stability corresponding to the environmental type of each collection area in each current slope body is obtained. Each collection area in each current slope body whose landslide stability is less than the preset landslide stability is recorded as a dangerous collection area, and rainfall sensors and landslide data collectors are set in each dangerous collection area in each slope body.

5. The artificial intelligence-based safety inspection method according to claim 2 is characterized in that: The rainfall environment data and rainfall landslide data are analyzed, and the specific analysis process is as follows: According to the analysis process of the environmental type, the rainfall environmental data is analyzed to obtain the environmental type corresponding to each slope body in each rainfall, and the environmental landslide stability corresponding to each slope body in each rainfall is obtained according to the environmental landslide stability grade corresponding to each environmental type. The correction factors of each slope change and each displacement change corresponding to each environmental landslide stability are obtained from the database to obtain the correction factors of each slope change and the displacement change corresponding to each slope body in each rainfall. The rainfall landslide data includes the slope change and the displacement change corresponding to each slope body in each rainfall. The slope change corresponding to each slope body in each rainfall is multiplied by the slope change correction factor to obtain the actual slope change corresponding to each slope body in each rainfall. The displacement change corresponding to each slope body in each rainfall is multiplied by the displacement change correction factor to obtain the actual displacement change corresponding to each slope body in each rainfall. The actual rainfall landslide data includes the actual slope change and the actual displacement change corresponding to each slope body in each rainfall.

6. The artificial intelligence-based safety inspection method according to claim 1 is characterized in that: The rainfall data and actual landslide data are analyzed, and the specific analysis process is as follows: According to the analysis process of the pre-rainfall environmental data and the pre-rainfall landslide data, the rainfall data and the actual landslide data are analyzed to obtain the rainfall type corresponding to each slope body in each rainfall, and then obtain the actual inclination change amount and the actual displacement change amount corresponding to each rainfall type. At the same time, if the actual inclination change amount of a slope body in a certain rainfall is less than the preset inclination change amount, and the actual displacement change amount of the slope body collected in this rainfall is less than the preset displacement change amount, the state of the slope body collected in this rainfall is recorded as the non-landslide state, and the number of occurrences of the non-landslide state in each rainfall is obtained by statistics, and the number of non-landslide collections of each rainfall type is obtained according to the rainfall type corresponding to each rainfall; The average values ​​of the actual inclination changes and actual displacement changes corresponding to each rainfall type are calculated respectively to obtain the average actual inclination changes and average actual displacement changes corresponding to each rainfall type. The average actual inclination changes, average actual displacement changes, maximum actual inclination changes and maximum actual displacement changes corresponding to each rainfall type are input into the rainfall-landslide model to obtain the output results of each rainfall type. The value j of the output result is the landslide stability grade, j=1,2......v, v>2, and v is the maximum rainfall-landslide stability grade.

7. The artificial intelligence-based safety inspection method according to claim 6 is characterized in that: The rainfall landslide model expression is: Among them, β b is the output result of rainfall type b, b is the number of rainfall type, b=1,2……n, n>2, E″ b , E′ b , F″ b and F′ b are the maximum actual tilt change, average actual tilt change, maximum actual displacement change and average actual displacement change corresponding to environment type b, E″, E′, F″ and F′ are the preset standard maximum actual tilt change, standard actual tilt change, standard maximum actual displacement change and standard actual displacement change, respectively; C b is the number of non-landslide acquisitions of the preset rainfall type b, C′ is the preset standard non-landslide acquisition number, φ1 and φ2 are the preset weight factors of rainfall inclination and rainfall displacement, φ1>0, φ2>0, φ1+φ2=1, M1, M2, M j-1 、M j and M v They are respectively the preset first rainfall landslide stability index, the second rainfall landslide stability index, the j-1th rainfall landslide stability index, the jth rainfall landslide stability index and the vth rainfall landslide stability index.

8. The artificial intelligence-based safety inspection method according to claim 1 is characterized in that: The landslide data collected each time within the preset time period are analyzed, and the specific analysis process is as follows: The landslide data collected each time within the preset time include the inclination change collected each time corresponding to each current slope body, the displacement change collected each time, the total inclination change and the total displacement change. The inclination change collected each time corresponding to each current slope body, the displacement change collected each time, the total inclination change and the total displacement change are input into the actual landslide model to obtain the output result of the actual landslide model. The value k of the output result is the actual landslide grade, k=1,2...w, w>2, and w is the maximum actual landslide grade.

9. The artificial intelligence-based safety inspection method according to claim 1 is characterized in that: The actual landslide model expression is: ， Among them, γ is the output result of the actual landslide model, ΔS′ cd and ΔL′ cd are the slope change and displacement change corresponding to the current slope c for the d times of acquisition, ΔS″ c and ΔL″ c are the total changes in inclination and displacement corresponding to the current slope c, ΔS′, ΔL′, ΔS″ and ΔL″ are the preset single standard inclination change, single standard displacement change, total standard inclination change and total standard displacement change, φ1 and φ2 are the preset rainfall inclination weight factor and rainfall displacement weight factor, respectively. is the preset weight factor of the current slope c, Q1, Q2, Q k-1 , Q k and Q w They are respectively the preset first landslide assessment index, the second landslide assessment index, the k-1th landslide assessment index, the kth landslide assessment index and the wth landslide assessment index.

10. The artificial intelligence-based safety inspection method according to claim 1, characterized in that: The aforementioned early warning processing: If the actual landslide level is greater than the preset standard actual landslide level or the preset landslide level is greater than the preset standard preset landslide level, an early warning will be issued, the standard actual landslide level and the preset landslide level will be substituted into the early warning index calculation formula to obtain the current early warning index, the early warning index interval corresponding to each early warning level is obtained from the database, the early warning level in the early warning index interval where the current early warning index is located is recorded as the current early warning level, the early warning plan corresponding to the current early warning level is obtained from the database, recorded as the current early warning plan, and an early warning is issued according to the current early warning plan.

Images