A method and system for dynamically setting the warning threshold of a single landslide

Abstract

The present invention is applicable to the technical field of electronic digital data processing, and particularly relates to a method and system for dynamically setting the warning threshold of a single landslide. The method includes: performing data cleaning according to the data characteristics in a large number of cross-regional single landslide-related data; making a preliminary judgment on the influencing factors of ontology relevance according to the cleaned historical landslide data; calculating and judging the key influencing factors according to the cleaned landslide surface deformation data; classifying different landslides stage by stage according to the key influencing factors; comparing the results of the preliminary judgment of the influencing factors with the structure of the calculated and judged key influencing factors, and calculating the threshold; collecting data, and performing dynamic threshold correction. The present invention makes a judgment on the relevance influence and makes a judgment on the key influencing factors, thereby determining the threshold, and then continuously performing dynamic correction on the threshold according to the collected data during the subsequent data collection process, so as to obtain the most accurate and reliable warning threshold.

Description

Technical Field

[0001] The present invention belongs to the technical field of electronic digital data processing, and particularly relates to a method and system for dynamically setting the warning threshold of a single landslide. Background Art

[0002] A landslide refers to the natural phenomenon that soil or rock mass on a slope, under the influence of factors such as river scouring, groundwater activity, rainwater soaking, earthquake, and artificial slope cutting, slides down the slope as a whole or dispersedly along a certain weak surface or weak zone under the action of gravity. The moving rock (soil) mass is called the displaced body or sliding body, and the underlying rock (soil) mass that has not moved is called the sliding bed.

[0003] The basic conditions for generating a landslide are that there is a sliding space in front of the slope body and cutting surfaces on both sides. For example, in the southwestern region of China, especially in the hilly mountainous areas of the southwest, the most basic topographical and geomorphic features are numerous mountains, steep mountain slopes, loose soil structures, easy water accumulation, and ravine rivers distributed among the mountains and cutting each other, thus forming numerous slope bodies and cutting surfaces with sufficient sliding space. The basic conditions for landslides are widely present, and landslide disasters are quite frequent.

[0004] For slope bodies under different conditions, the thresholds for generating landslides are also different, and the prior art cannot achieve dynamic adjustment of the warning threshold. Summary of the Invention

[0005] An object of an embodiment of the present invention is to provide a method for dynamically setting the warning threshold of a single landslide, aiming to solve the problem that the prior art cannot achieve dynamic adjustment of the warning threshold.

[0006] An embodiment of the present invention is implemented as follows. A method for dynamically setting the warning threshold of a single landslide, the method includes:

[0007] Performing data cleaning according to the data characteristics in the relevant data of numerous single landslides across regions;

[0008] Performing a preliminary judgment on the influencing factors of ontology relevance according to the historical landslide data after cleaning;

[0009] Performing a calculation and judgment on the key influencing factors according to the landslide surface deformation data after cleaning;

[0010] Classifying different landslides stage by stage according to the key influencing factors;

[0011] Comparing the results of the preliminary judgment of the influencing factors with the results of the calculation and judgment of the key influencing factors, and calculating the threshold;

[0012] Collecting data and performing dynamic threshold correction.

[0013] Preferably, the data cleaning step specifically includes collecting and aggregating a large number of individual landslide data and cleaning and refining dirty and waste landslide data.

[0014] Preferably, the step of making a preliminary judgment on the influencing factors of the ontology correlation based on the cleaned historical landslide data specifically includes:

[0015] Screening of single landslides where major landslide events have occurred;

[0016] The screened single landslide data are sorted out according to the known main influencing factors;

[0017] The collated data were correlated with the data of the landslide when it had a significant displacement, and the main factors affecting the significant displacement of the single landslide were extracted.

[0018] The single landslides that have undergone significant displacement are preliminarily classified according to the main factors, and the thresholds of different deformation stages are recorded.

[0019] Preferably, the step of calculating and judging key influencing factors based on the cleaned landslide surface deformation data specifically includes:

[0020] Screen individual landslides that have not experienced major landslide events, and extract the time points when they experienced relatively drastic deformation based on their historical data;

[0021] Record the time points of relatively drastic deformation, and extract and organize the contextual data respectively;

[0022] A combined algorithm is used to calculate key influencing factors of context data and severe deformation data;

[0023] Summarize the key influencing factors and changing patterns of key influencing factors when clustered deformation occurs in different single landslides.

[0024] Preferably, in the data cleaning step, the collection method adopted is sensors and monitoring equipment installed on the single landslide, and the collected data content is environmental related data of the area that has been judged to be a single landslide. The environmental related data at least includes rainfall forecast data, surface deformation data, earthquake data, lake and river groundwater level data, cracks and longitude and latitude data.

[0025] Preferably, the major landslide event refers to a landslide that has caused an alarm event in history or an event in which a single landslide undergoes a staged change in landslide according to internationally accepted rainfall, displacement, displacement acceleration, and displacement deformation.

[0026] Preferably, the association analysis refers to the relevant algorithms in data mining for calculating the correlation between different landslide factor data and displacement changes in the historical data of major landslide events that have occurred, and classifying individual landslides in multiple different regions uniformly through factor similarity. The classification basis is the factor similarity of the association analysis.

[0027] Another object of the embodiments of the present invention is to provide a system for dynamically setting the warning threshold of individual landslides, and the system includes:

[0028] A data cleaning module for cleaning data according to the data characteristics in the relevant data of numerous individual landslides across regions;

[0029] A relevance impact judgment module for preliminarily judging the ontology relevance impact factors according to the cleaned historical landslide data;

[0030] A key impact judgment module for calculating and judging key impact factors according to the cleaned landslide surface deformation data;

[0031] A landslide classification module for classifying different landslides stage by stage according to key impact factors;

[0032] A threshold calculation module for comparing the results of the preliminary judgment of impact factors with the results of the calculation and judgment of key impact factors and calculating the threshold;

[0033] A threshold correction module for collecting data and performing dynamic threshold correction.

[0034] A method for dynamically setting the warning threshold of individual landslides provided by the embodiments of the present invention cleans the data of individual landslides, then makes a relevance impact judgment according to the cleaned historical landslide data and makes a key impact factor judgment according to the landslide surface deformation data, thereby determining the threshold. Then, in the subsequent data collection process, the threshold is continuously corrected dynamically according to the collected data, so as to obtain the most accurate and reliable warning threshold. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is the main flowchart of a method for dynamically setting the warning threshold of individual landslides provided by the embodiments of the present invention;

[0036] Figure 2 It is the architecture diagram of a system for dynamically setting the warning threshold of individual landslides provided by the embodiments of the present invention;

[0037] Figure 3 It is the whole process flowchart of a method for dynamically setting the warning threshold of individual landslides provided by the embodiments of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0038] To make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0039] It can be understood that the terms "first", "second", etc. used in this application may be used herein to describe various elements, but unless otherwise specified, these elements are not limited by these terms. These terms are only used to distinguish one element from another. For example, without departing from the scope of this application, the first xx script may be referred to as the second xx script, and similarly, the second xx script may be referred to as the first xx script.

[0040] As Figure 3 shown, it is the whole process flowchart of a method for dynamically setting the warning threshold of a single landslide provided by an embodiment of the present invention;

[0041] As Figure 1 shown, it is the main flowchart of a method for dynamically setting the warning threshold of a single landslide provided by an embodiment of the present invention, and the method includes:

[0042] S100, perform data cleaning according to the data characteristics in the data related to numerous single landslides across regions.

[0043] In this step, the data cleaning step includes two operations: collecting and summarizing data of numerous single landslides, and cleaning and refining landslide dirty data, waste data, etc.; in the data cleaning step:

[0044] Collection method: Monitoring data of sensors and monitoring devices deployed on single landslides, and meteorological data transmitted through the meteorological interface;

[0045] Data content: Environment-related data in areas that have been determined to be single landslide areas, including but not limited to rainfall forecast data, surface deformation data, earthquake data, lake, river and groundwater level data, cracks, longitude and latitude data, etc.;

[0046] Data storage: Stored in databases and data warehouses.

[0047] S200, perform a preliminary judgment on the influencing factors of ontology relevance according to the historical landslide data after cleaning.

[0048] In this step, it specifically includes:

[0049] Screen single landslides where major landslide events have occurred;

[0050] Sort out the single landslide data that has been screened according to known main influencing factors such as longitude and latitude, distance from water area, precipitation, earthquake, etc.;

[0051] Perform a correlation analysis on the sorted data and the data at the time of significant displacement of the landslide, and extract 1 - 3 main factors affecting the significant displacement of individual landslides;

[0052] Based on the main factors, conduct a preliminary classification of these individual landslides that have experienced significant displacement, and record the thresholds at different stages of their deformation;

[0053] Major landslide events refer to: landslides that have issued warning events in history or events determined as stage changes of individual landslides based on internationally common rainfall, displacement, displacement acceleration, and displacement deformation;

[0054] Distance to water area: The distance to water area refers to the effective water area distance data that can affect individual landslides, such as relevant data of rivers, lakes, dams, etc., including data such as water area, flow velocity, depth, and straight-line distance to the landslide;

[0055] Correlation analysis: Refers to the correlation analysis related algorithms in data mining, aiming to calculate the correlation between different landslide-related factor data such as rainfall, displacement acceleration, earthquake, hydrology, and vegetation in the historical data of major landslide events that have occurred and the displacement change, and then conduct a unified classification of individual landslides in multiple different regions based on the factor similarity. The classification basis is the factor similarity of the correlation analysis.

[0056] S300, calculate and judge the key influencing factors based on the cleaned landslide surface deformation data.

[0057] In this step, the steps for calculating and judging the key influencing factors specifically include:

[0058] Screen individual landslides that have not experienced major landslide events, and extract the time points of relatively severe deformation of themselves based on their historical data;

[0059] Record these time points of relatively severe deformation, and perform separate context data extraction and sorting. The sorted data includes processed data;

[0060] Use classification / clustering / regression combination algorithms to calculate the key influencing factors for the context data and the severe deformation data;

[0061] Sort out the key influencing factors and the change rules of these key influencing factors when different individual landslides undergo clustering deformation;

[0062] In this step, not having experienced a major landslide event means: The data related to individual landslides without warnings selected in this step, that is, individual landslides that have been determined but have not generated warnings or events determined as not having stage changes of landslides for individual landslides based on internationally common rainfall, displacement, displacement acceleration, and displacement deformation;

[0063] Relatively severe deformation: It refers to the moment in the historical data of a single landslide where the deformation of the surface displacement data relative to itself is relatively large;

[0064] Context data: In this step, the context data refers to a series of data where the values of the influencing factors of the single landslide tend to change from stability before and after the moment of relatively severe deformation;

[0065] Processed data: In this step, the processed data is relative to the collected data. The collected data is the data collected by the collection devices deployed at the single landslide site in the first step. The processed data refers to the secondary or multiple data obtained by re - processing the collected data through manual means. Generally, it refers to three types of data: statistical data, such as the weekly rainfall; analysis process data, such as the entropy value or gain in the calculation process of the vertical displacement acceleration classification algorithm; calculated data, such as the rainfall acceleration at each moment;

[0066] Combined algorithm: In this step, three data mining algorithms, namely classification, clustering, and regression, are mainly used. The combined algorithm refers to the scenario where, when calculating the key influencing factors, for different influencing factors or numerical combinations of influencing factors, two or three algorithms are used for hybrid calculation. The purpose is to find out the main influencing factors of the single landslide for surface displacement at this moment. For example, the random forest algorithm is used to obtain the ranking of the influencing factors at a time point, and then the regression algorithm is used to verify the effectiveness of this ranking;

[0067] Influencing factors: In this step, the influencing factors refer to different collected data that can affect the surface displacement of a single landslide, such as the commonly used collected data internationally, including vertical displacement, horizontal displacement, rainfall, displacement acceleration, earthquake, groundwater level, etc., and also include the influencing factors that are currently considered to have relatively weak influence internationally, such as temperature, elevation, etc.;

[0068] Variation law: In this step, the variation law refers to the law between the surface displacement change and the key factor change when the single landslide undergoes relatively severe deformation.

[0069] S400, classify different landslides stage - by - stage according to the key influencing factors;

[0070] In this step, classification operations are performed on the single landslide according to the generated key influencing factors (independent data and processed data); Classifying the single landslide: In this step, classifying the single landslide means classifying the single landslides with similar influencing factors affecting relatively severe deformation, similar importance rankings of influencing factors, and similar change amplitudes of influencing factors into one category.

[0071] S500, compare the preliminary judgment result of the influencing factors with the calculation and judgment result of the key influencing factors, and calculate the threshold.

[0072] In this step, the steps for comparison include:

[0073] Use algorithms such as classification and correlation to calculate the correlation between the landslide classification in S400 and the landslide classification in S200;

[0074] Assign the threshold of S200 with a relatively large positive correlation to the landslide threshold in S400. Since there are many key factors in S400 and it includes processing data factors, it is necessary to calculate the key factor thresholds in S400 for the corresponding points of the threshold in S200;

[0075] In this step, correlation: In this step, it refers to the degree of correlation between the influencing factors that affect large-scale surface deformation of a single landslide in Step 2 and Step 4, that is, the degree of correlation with similar influencing factors, similar importance rankings of influencing factors, and similar change ranges of influencing factors. Since the influencing factors in Step 4 include processing data, the original data before processing is used as the division of this influencing factor when calculating the correlation;

[0076] Calculate the key factor thresholds in Step 4 for the corresponding points of the threshold in Step 2: Here, it is to infer the possible warning threshold in Step 4 through the threshold of the alarm that has occurred in Step 2, where the processing data threshold in Step 4 is calculated through the threshold of its original data.

[0077] S600, collect data and perform dynamic threshold correction.

[0078] In this step, continuously collect data and continuously correct the threshold operation of the corresponding single landslide in S500 according to the classification changes caused by the changes of influencing factors in S300; Continuously correct the threshold: Continuously correcting the threshold in this step means that since the deformation stage of a single landslide will change according to its natural environment, and there are differences in influencing factors in different deformation stages, here, by continuously collecting data, the deformation stage of a single landslide is continuously corrected, and then the influencing factors are continuously corrected to continuously correct the threshold.

[0079] Such as Figure 2 shown, is the architecture diagram of a system for dynamically setting the warning threshold of a single landslide provided by an embodiment of the present invention. The system includes:

[0080] Data cleaning module 100, used to clean data according to data characteristics in the relevant data of numerous single landslides across regions.

[0081] Relevance influence judgment module 200, used to make a preliminary judgment on the influencing factors of ontology relevance according to the cleaned historical landslide data.

[0082] Key influence judgment module 300, used to calculate and judge key influencing factors according to the cleaned landslide surface deformation data.

[0083] The landslide classification module 400 is used to classify different landslides stage by stage according to key influencing factors.

[0084] The threshold calculation module 500 is used to compare the results of the preliminary judgment of influencing factors with the results of the calculation and judgment of key influencing factors, and calculate the threshold.

[0085] The threshold correction module 600 is used to collect data and perform dynamic threshold correction.

[0086] It should be understood that although the steps in the flowcharts of the embodiments of the present invention are shown in sequence according to the indications of the arrows, these steps do not necessarily have to be executed in the order indicated by the arrows. Unless there is a clear indication in this article, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, at least a part of the steps in each embodiment may include multiple sub-steps or multiple stages. These sub-steps or stages do not necessarily have to be executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages does not necessarily have to be sequential, but can be executed alternately or alternately with at least a part of other steps or sub-steps or stages of other steps.

[0087] Those of ordinary skill in the art can understand that all or part of the processes of implementing the methods in the above embodiments can be completed by instructing relevant hardware through a computer program. The program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, storage, database or other medium used in the embodiments provided in the present application can include non-volatile and / or volatile memories. Non-volatile memories can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memories can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0088] The technical features of the above-described embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as falling within the scope described in this specification.

[0089] The above-described embodiments merely represent several implementation manners of the present invention. The description thereof is relatively specific and detailed, but it should not be construed as a limitation to the scope of the patent for the present invention. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can still be made, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent for the present invention shall be subject to the appended claims.

[0090] The foregoing is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent replacements, and improvements made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for dynamically setting the warning threshold of a single landslide, characterized in that, the method includes: S100, performing data cleaning according to the data characteristics in the data related to numerous single landslides across regions; S200, making a preliminary judgment on the influencing factors of ontological relevance according to the historical landslide data after cleaning, specifically including: screening single landslides where major landslide events have occurred; sorting out the data of the screened single landslide data according to the known main influencing factors such as longitude and latitude, distance from water area, precipitation, and earthquake; performing correlation analysis on the sorted data and the data when the landslide has a major displacement, and extracting 1 to 3 main factors affecting the major displacement of the single landslide; making a preliminary classification of these single landslides with major displacements according to the main factors, and recording the thresholds at different deformation stages; S300, calculating and judging the key influencing factors according to the landslide surface deformation data after cleaning, specifically including: screening single landslides where major landslide events have not occurred, and extracting the time points when relatively severe deformation occurs according to their historical data; recording these time points of relatively severe deformation and extracting and sorting the context data respectively, and the sorted data includes processed data; using a classification / clustering / regression combination algorithm to calculate the key influencing factors of the context data and the severe deformation data; sorting out the key influencing factors and the change rules of these key influencing factors when different single landslides have clustering deformation; S400, classifying different landslides in stages according to the key influencing factors, specifically including: performing classification operations on single landslides according to the generated key influencing factors; classifying single landslides, and classifying single landslides means classifying single landslides with similar influencing factors affecting relatively severe deformation of single landslides, similar importance rankings of influencing factors, and similar change ranges of influencing factors into one category; S500, comparing the results of the preliminary judgment of influencing factors and the results of calculating and judging key influencing factors, and calculating the threshold, specifically including: calculating the correlation between the landslide classification in S400 and the landslide classification in the historical landslide data after cleaning by using a classification and correlation algorithm; assigning the threshold in S200 with a larger positive correlation to the landslide threshold in S400; S600, collecting data and performing dynamic threshold correction.

2. The method for dynamically setting the warning threshold of a single landslide according to claim 1, characterized in that, the steps of the data cleaning specifically include collecting and summarizing data of numerous single landslides and cleaning and refining the dirty data and waste data of the landslides.

3. The method for dynamically setting the warning threshold of a single landslide according to claim 2, characterized in that, in the steps of the data cleaning, the acquisition method used is sensors and monitoring devices set on single landslides, and the data content acquired is environmental-related data of the area judged to be a single landslide area, and the environmental-related data at least includes rainfall forecast data, surface deformation data, earthquake data, lake, river and groundwater level data, crack and longitude and latitude data.

4. The method for dynamically setting the warning threshold of a single landslide according to claim 1, characterized in that, The major landslide event refers to a landslide that has issued an alarm event in history or an event in which a single landslide undergoes a phased change in landslide as determined by internationally common rainfall, displacement, displacement acceleration, and displacement deformation.

5. The method for dynamically setting the warning threshold of a single landslide according to claim 1, wherein, the correlation analysis refers to the correlation analysis-related algorithms in data mining, which is used to calculate the correlation between different landslide factor data and displacement changes in the historical data of major landslide events that have occurred, and to uniformly classify single landslides in multiple different regions through factor similarity, and the classification basis is the influence factor similarity of the correlation analysis.

6. A system for dynamically setting the warning threshold of a single landslide, wherein, the system includes: a data cleaning module, which is used to clean data according to the data characteristics in the relevant data of numerous single landslides across regions; a relevance influence judgment module, which is used to preliminarily judge the ontology relevance influence factors according to the cleaned historical landslide data, specifically including: screening single landslides where major landslide events have occurred; sorting out the data of the screened single landslides according to the known main influence factors such as longitude and latitude, distance from water area, precipitation, and earthquake; performing correlation analysis on the sorted data and the data when the landslide undergoes a major displacement, and extracting 1-3 main factors that affect the major displacement of the single landslide; preliminarily classifying these single landslides that have undergone major displacements according to the main factors, and recording the thresholds at different deformation stages; a key influence judgment module, which is used to calculate and judge the key influence factors according to the cleaned landslide surface deformation data, specifically including: screening single landslides where major landslide events have not occurred, and extracting the time points when relatively severe deformations occur in themselves according to their historical data; recording these time points of relatively severe deformations and separately extracting and sorting the context data, and the sorted data includes processed data; using a classification / clustering / regression combination algorithm to calculate the key influence factors of the context data and the severe deformation data; sorting out the key influence factors and their change rules when different single landslides undergo clustering deformations; a landslide classification module, which is used to classify different landslides in stages according to the key influence factors, specifically including: classifying single landslides according to the generated key influence factors; classifying single landslides, and classifying single landslides means classifying single landslides with similar influence factors that affect relatively severe deformations of single landslides, similar importance rankings of influence factors, and similar change ranges of influence factors into one category; a threshold calculation module, which is used to compare the results of the preliminary judgment of influence factors and the results of the calculation and judgment of key influence factors, and calculate the threshold, specifically including: using a classification and correlation algorithm to calculate the correlation between the landslide classification in the landslide classification module and the landslide classification in the relevance influence judgment module; assigning the threshold of the relevance influence judgment module with a larger positive correlation to the landslide threshold in the landslide classification module; a threshold correction module, which is used to collect data and perform dynamic threshold correction.

Images