CN-122024408-A - Geological disaster space-time distribution probability prediction method and system under different rainfall conditions

CN122024408ACN 122024408 ACN122024408 ACN 122024408ACN-122024408-A

Abstract

The invention discloses a geological disaster space-time distribution probability prediction method under different rainfall conditions, which comprises the steps of obtaining historical geological disaster data and weather rainfall data, constructing a disaster-rainfall sample set, calculating total effective rainfall and sectional accumulated rainfall of each sample based on an exponential decay model, screening out a water-falling disaster sample according to the comparison relation between the total effective rainfall and the sectional accumulated rainfall of each sample and a preset threshold value, judging the water-falling disaster sample based on a judgment coefficient and an actual contribution rate, determining a dominant rainfall type inducing disasters, and carrying out space-time distribution probability analysis according to the dominant rainfall type to generate a geological disaster occurrence prediction result. According to the invention, the accumulated attenuation effect of rainfall is accurately quantified by introducing an exponential attenuation model, the dominant rainfall type is automatically identified based on the theoretical discrimination coefficient and the actual contribution rate, and the precise prediction of geological disasters is realized by combining double threshold screening and disaster type self-adaptive parameter setting.

Inventors

ZHAO XU
LI HAO
YANG NAN
WU XINGBIN
YU HONG
ZI YUNJIANG
Yang Yindan
XU CHENWEI

Assignees

云南电网有限责任公司电力科学研究院

Dates

Publication Date: 20260512
Application Date: 20251218

Claims (10)

1. A method for predicting the probability of a geological disaster space-time distribution under different rainfall conditions, which is characterized by comprising the following steps: acquiring historical geological disaster data and weather precipitation data, and constructing a disaster-rainfall sample set; Calculating the total effective precipitation amount and the sectional accumulated precipitation amount of each sample based on the exponential decay model; Screening out a water-falling disaster sample according to the comparison relation between the total effective precipitation amount of each sample in the disaster-rainfall sample set and the sectional accumulated precipitation amount and a preset threshold value; Based on the discrimination coefficient and the actual contribution rate, discriminating the rainfall type disaster sample, and determining the dominant rainfall type of the induced disaster; and carrying out space-time distribution probability analysis on the geological disasters according to the dominant rainfall type, and generating geological disaster occurrence prediction results.
2. The method for predicting the probability of a geological disaster space-time distribution under different rainfall conditions according to claim 1, wherein the constructing a disaster-rainfall sample set specifically comprises: Acquiring occurrence time and geographic position information of each disaster record in the historical geological disaster database; Based on the geographical position information, adopting a nearby matching algorithm in a weather precipitation database to correlate disaster points with weather sites which are closest in space; And extracting rainfall data of the associated meteorological sites at the disaster occurrence time and earlier stage, and forming a disaster-rainfall sample together with disaster records.
3. The method for predicting the probability of a geological disaster time-space distribution under different rainfall conditions according to claim 2, wherein the calculating the total effective precipitation amount of each sample specifically comprises: The total effective precipitation per sample was calculated by the following formula: Wherein, the When i=1, namely the daily precipitation amount of the current day and the i-1 th day before the disaster, Is a rainfall attenuation coefficient, Total effective precipitation for each sample, N, is calculated as total days.
4. A method for predicting the probability of a geological disaster spatiotemporal distribution under different rainfall conditions according to claim 3, wherein said sectional accumulated precipitation amount includes a short-term accumulated precipitation amount, a medium-term accumulated precipitation amount and a long-term accumulated precipitation amount; the short-term accumulated precipitation is: The medium-term accumulated precipitation amount is as follows: The long-term accumulated precipitation is: Wherein, the When i=1, i.e., the daily precipitation amount of the current day and the i-1 th day before the occurrence of the disaster.
5. The method for predicting the probability of geologic hazard spatial-temporal distribution under different rainfall conditions according to claim 4, wherein the step of screening out the samples of the rainfall type disaster according to the comparison relation between the total effective rainfall and the sectional accumulated rainfall of each sample in the disaster-rainfall sample set and a preset threshold value specifically comprises the following steps: setting a first threshold and a second threshold for each sample; If the total effective precipitation amount of the sample is not less than a first threshold value or the accumulated precipitation amount of any section is not less than a second threshold value, judging that the sample is a precipitation type disaster sample; Otherwise, the sample is removed as an uncorrelated noise sample.
6. The method for predicting the probability of a geological disaster space-time distribution under different rainfall conditions according to claim 5, wherein the determining the dominant rainfall type of the induced disaster by distinguishing the rainfall type disaster sample based on the distinguishing coefficient and the actual contribution rate specifically comprises the following steps: Determining discrimination coefficients according to disaster types : Wherein m is dynamically adjusted according to different disaster types, Is a landslide, Is a mud-rock flow, Collapse and k are rainfall attenuation coefficients; calculating the actual contribution rate of short-term precipitation in the current sample : Wherein, the Short-term accumulated precipitation for the jth precipitation disaster sample, The actual contribution rate of the medium-short-term precipitation, Is the j-th precipitation disaster total effective precipitation of the sample; comparing the actual contribution rate of short-term precipitation in the current sample Coefficient of discrimination : If the actual contribution rate of short-term precipitation in the current sample Discrimination coefficient Judging that the sample is short-term precipitation dominant; If the actual contribution rate of short-term precipitation in the current sample Judging coefficient less than or equal to And judging the sample as the dominant type of long-duration precipitation.
7. The method for predicting the probability of occurrence of a geological disaster in accordance with claim 6, wherein said performing a spatial-temporal distribution probability analysis of the geological disaster according to the dominant rainfall type to generate a prediction result of occurrence of the geological disaster comprises: Dynamically dividing a prediction area into a short-term precipitation dominant region, a long-duration precipitation dominant region and a short-term strong precipitation induction region based on the dominant rainfall type; For each subarea, respectively matching a historical disaster space-time distribution mode corresponding to the dominant rainfall type; the space-time distribution mode comprises a time lag characteristic probability function and a space density distribution characteristic of disaster occurrence; Calculating a future specified time window for different spatial grids within each partition based on the matched spatio-temporal distribution patterns Probability of occurrence of geological disasters in the mouth; Comparing the calculation result with the early warning level threshold value to generate a rainfall pattern containing disaster occurrence probability and dominant rainfall pattern And (5) a grid geological disaster risk prediction map of time-space positions and early warning grades.
8. The geological disaster space-time distribution probability prediction system under different rainfall conditions is characterized by comprising a sample set construction unit, a sample data calculation unit, a sample screening unit, a sample type judging unit and a space-time analysis prediction unit; The sample set construction unit is used for acquiring historical geological disaster data and weather precipitation data and constructing a disaster-rainfall sample set; The sample data calculation unit is used for calculating the total effective precipitation and the sectional accumulated precipitation of each sample based on the exponential decay model; The sample screening unit is used for screening out a water-falling disaster sample according to the comparison relation between the total effective precipitation amount and the sectional accumulated precipitation amount of each sample in the disaster-rainfall sample set and a preset threshold value; The sample type judging unit is used for judging the rainfall type disaster samples based on the judging coefficient and the actual contribution rate and determining the dominant rainfall type of the induced disasters; The space-time analysis prediction unit is used for carrying out space-time distribution probability analysis of the geological disasters according to the dominant rainfall type and generating geological disaster occurrence prediction results.
9. A readable storage medium storing a computer program, which when executed by a processor causes the processor to perform the steps of the method according to any one of claims 1 to 7.
10. A computer device comprising a memory and a processor, wherein the memory stores a computer program which, when executed by the processor, causes the processor to perform the steps of the method as claimed in any one of claims 1 to 7.

Description

Geological disaster space-time distribution probability prediction method and system under different rainfall conditions Technical Field The invention relates to the technical field of geological disaster prediction, in particular to a geological disaster space-time distribution probability prediction method and a geological disaster space-time distribution probability prediction system under different rainfall conditions. Background Geological disasters are an important natural disaster type threatening the life and property safety of people, wherein rainfall is a main factor for inducing landslide, collapse, debris flow and other geological disasters. The traditional rainfall type geological disaster prediction method mainly adopts a critical rainfall method or a statistical regression model, and has obvious limitations that firstly, only total rainfall or single-day rainfall is considered, the accumulation effect and the attenuation process of early rainfall cannot be fully reflected, secondly, the type of rainfall process is not accurately judged, different induction mechanisms of short-term strong rainfall and long-duration rainfall cannot be distinguished, thirdly, the prediction result is mostly qualitative or semi-quantitative, and the latticed space-time probability prediction is difficult to realize. In the prior art, although partial researches try to introduce an effective precipitation concept, a simple linear accumulation or fixed weight coefficient is mostly adopted, and the physical process of soil moisture infiltration and evaporation cannot be accurately reflected. Meanwhile, the rainfall response characteristic difference of different disaster types (such as landslide, debris flow and collapse) is not considered enough, so that the prediction accuracy is limited. In addition, when the conventional method is used for constructing a disaster-rainfall sample set, the problem that disaster points are not matched with the space positions of meteorological sites is often ignored, and the data of the nearest site is directly adopted, so that data association errors are caused. Therefore, there is a need for a geological disaster space-time distribution probability prediction method capable of comprehensively considering early rainfall accumulation effect, rainfall process type discrimination and response characteristics of different disaster types, so as to improve prediction accuracy and practicality. Disclosure of Invention Based on the above, it is necessary to provide a method for predicting the probability of the spatial and temporal distribution of geological disasters under different rainfall conditions. A method for predicting probability of geologic hazard spatial-temporal distribution under different rainfall conditions, the method comprising the steps of: acquiring historical geological disaster data and weather precipitation data, and constructing a disaster-rainfall sample set; Calculating the total effective precipitation amount and the sectional accumulated precipitation amount of each sample based on the exponential decay model; Screening out a water-falling disaster sample according to the comparison relation between the total effective precipitation amount of each sample in the disaster-rainfall sample set and the sectional accumulated precipitation amount and a preset threshold value; Based on the discrimination coefficient and the actual contribution rate, discriminating the rainfall type disaster sample, and determining the dominant rainfall type of the induced disaster; and carrying out space-time distribution probability analysis on the geological disasters according to the dominant rainfall type, and generating geological disaster occurrence prediction results. In the above scheme, the constructing the disaster-rainfall sample set specifically includes: Acquiring occurrence time and geographic position information of each disaster record in the historical geological disaster database; Based on the geographical position information, adopting a nearby matching algorithm in a weather precipitation database to correlate disaster points with weather sites which are closest in space; And extracting rainfall data of the associated meteorological sites at the disaster occurrence time and earlier stage, and forming a disaster-rainfall sample together with disaster records. In the above solution, the calculating the total effective precipitation amount of each sample specifically includes: The total effective precipitation per sample was calculated by the following formula: Wherein, the When i=1, namely the daily precipitation amount of the current day and the i-1 th day before the disaster,Is a rainfall attenuation coefficient,Total effective precipitation for each sample, N, is calculated as total days. In the above scheme, the sectional accumulated precipitation includes short-term accumulated precipitation, medium-term accumulated precipitation and long-term accumulated precip