阅读说明：本技术 一种滑坡预报的滞后非线性时间预报模型 (Hysteresis nonlinear time forecasting model for landslide forecasting ) 是由 谢婉丽 杨惠 于 2019-10-16 设计创作，主要内容包括：本发明公开了一种滑坡预报的滞后非线性时间预报模型,现阶段滑坡的预报重心偏向即时预报,但实际上浅层土壤含水率和地下水位也对滑坡的发生有一定的影响作用。该模型实现了同时分析地下水位、土壤浅层含水率随降雨量及时间变化的关系；表现了地下水位、土壤浅层含水率随降雨量及时间变化产生滞后效应和非线性效应,从而应用于滑坡的有效预报,完善目前滑坡监测预报系统。(The invention discloses a lag nonlinear time forecasting model for landslide forecasting, which is used for forecasting the deviation of the forecasting gravity center of landslide at the present stage in real time, but actually, the water content of shallow soil and the underground water level also have certain influence on the occurrence of landslide. The model realizes the simultaneous analysis of the relation between the underground water level and the soil shallow layer water content along with the rainfall and the time change; the method expresses that the underground water level and the soil shallow layer water content generate a hysteresis effect and a nonlinear effect along with the change of rainfall and time, thereby being applied to the effective forecasting of landslide and perfecting the existing landslide monitoring and forecasting system.)

1. A model for lag nonlinear time prediction for landslide prediction, comprising the steps of:

the method comprises the following steps: establishing matrix group data of rainfall and time based on a hysteresis and continuous nonlinear relation between landslide occurrence and rainfall, and introducing the matrix group data, shallow soil water content and underground water bit data into calculation software at the same time to obtain a rainfall-time-underground water level and rainfall-time-soil water content three-dimensional curved surface model diagram;

step two: based on a three-dimensional curved surface model diagram, a distributed lag nonlinear basic model is obtained by using a cross basis function, lag time is added into the basic model, a lag effect relational expression is established, and finally a lag nonlinear time forecasting model for landslide forecasting is established;

step three: the model is applied to an actual landslide monitoring and early warning system, the hysteresis effect and the nonlinear effect of the underground water level and the soil moisture content are evaluated at the same time, and the change of the geological disaster within 24 hours in the future can be predicted within a time period with large rainfall.

2. The model of claim 1, wherein the distributed lag nonlinear time prediction model obtained by using the cross basis function in the second step is:

in the formula: g-family of chaining functions;

f (Y) -a dependent variable function of the groundwater level and the soil moisture content;

oc (rainfall, time) -a matrix of rainfall and time;

x_ijindependent variables- -i.e. the amount of rainfall and the time of analysis that varies day by day;

f_j-independent variable X_ijA basis function;

μ_k-functions of influencing factors such as water evaporation and infiltration in the landslide body;

β_j、γ_k-the corresponding influence coefficients in the function.

3. The model of claim 1, wherein the steps are performed in a manner that the model is based on a model of non-linear time-of-flight predictionThe method for adding the lag time into the basic model in the second step is as follows: taking the change of underground water as an example, expressing the rainfall-underground water level relation of independent variable and dependent variable by using a function v (x), expressing the time-underground water level relation by using a function u (l), combining the two functions to obtain a two-dimensional rainfall-time-underground water relation function v, u (x, l), and simplifying and expressing the function as delta in the calculation and analysis process_i(x,l)。

4. The model of claim 1, wherein the relationship of the hysteresis effect established in the second step is:

in the formula: theta-coefficient of hysteresis

L-maximum lag time, maximum lag time less than or equal to 24h

Chi-rainfall varying day by day

l-time of analysis

δ_i(x, l) is simplification of a v.u (x, l) function, the relation takes the hysteresis effect into consideration, and means that the change of the underground water level along with the rainfall and the time is analyzed, the process of rainfall penetrating into a slip slope body is taken into consideration, the underground water level is judged to be changed from the day, and the hysteresis time is obtained. And similarly, obtaining a hysteresis effect change function of the soil moisture content of the underground shallow layer.

5. The model of claim 1, wherein the distributed lag nonlinear time prediction model finally constructed in the second step adds lag time to the rainfall-groundwater level and rainfall-soil moisture content functional relationship through a cross basis function, so as to realize a distribution process simultaneously showing a lag effect generated by a dependent variable along with an independent variable.

6. The model for predicting the non-linear time lag in landslide according to claim 1, wherein the model can simultaneously evaluate the non-linear effect and the lag effect of rainfall affecting the groundwater level and the soil water content in the third step.

7. The model of claim 1, wherein when the model is applied to an actual monitoring and early warning system in the third step, the model analyzes the acquired data to obtain a real-time lag relationship, and predicts the change of the geological disaster point within 24 hours in the future within a time period with a large rainfall, so as to achieve the purpose of accurate prediction and effective prevention and ensure the safety of life and property of personnel in the geological disaster threat area.

Technical Field

The invention relates to the field of geological disaster monitoring and early warning, in particular to a hysteresis nonlinear time forecasting model for landslide forecasting.

Background

The landslide hazard is the most developed geological hazard type in the Qinba mountain area and has the characteristics of wide distribution, large quantity and strong activity.

The landslide in the Qinba mountainous area is divided into a buildup layer landslide, a loess landslide and a rock landslide according to the composition of the substances. The accumulation layer landslide body substance consists of a fourth series of slope deposit, residual deposit, silt, silty clay and detritus, and is a landslide type with the widest distribution, the largest quantity and the highest occurrence frequency in the Qinba mountainous area. The landslide is obvious and complete in form, most of sliding surfaces are located at the interface of a stacking layer and an underlying bedrock or inside the stacking layer, most of the sliding surfaces are in the initial creep deformation stage, the inducing factor is heavy rain or continuous overcast rain, and the catastrophe is mainly characterized by strong outburst.

The influence of rainfall on the soil moisture content and the underground water level has certain hysteresis and persistence, is related to the rainfall on the same day and can be influenced by the previous day or even the previous days, and in order to better reflect the relation among the three, a hysteresis nonlinear time forecasting model for landslide forecasting is established for typical geological disaster points in the Qinba mountainous area.

At present, landslide prediction mainly takes immediate prediction as the center of gravity, but actually landslide occurs not only in relation to the immediate rainfall, but also in relation to the influence of the accumulated rainfall on the soil moisture content and the underground water level, and the establishment of the model is helpful for perfecting the current landslide monitoring and predicting system.

Disclosure of Invention

The invention aims to solve the technical problem that the landslide monitoring and forecasting technology is insufficient at present, and the rainfall on the soil moisture content, the underground water level lag and the nonlinear effect are not considered in a forecasting system.

In order to solve the technical problems, the invention aims to: the change situation of the change of the underground water level and the soil moisture content along with the increase and the decrease of the rainfall is obtained through the model, the hysteresis effect and the nonlinear effect of the change situation are discussed, and the landslide monitoring and forecasting system is perfected.

In order to achieve the purpose, the invention provides the following technical scheme:

a model for lag nonlinear time prediction for landslide prediction comprising the steps of:

the method comprises the following steps: establishing matrix group data of rainfall and time based on a hysteresis and continuous nonlinear relation between landslide occurrence and rainfall, and introducing the matrix group data, shallow soil water content and underground water bit data into calculation software at the same time to obtain a rainfall-time-underground water level and rainfall-time-soil water content three-dimensional curved surface model diagram;

step two: based on a three-dimensional curved surface model diagram, a distributed lag nonlinear basic model is obtained by using a cross basis function, lag time is added into the basic model, a lag effect relational expression is established, and finally a lag nonlinear time forecasting model for landslide forecasting is established;

step three: the model is applied to an actual landslide monitoring and early warning system, the hysteresis effect and the nonlinear effect of the underground water level and the soil moisture content are evaluated at the same time, and the change of the geological disaster within 24 hours in the future can be predicted within a time period with large rainfall.

As a further scheme of the invention: the distribution lag nonlinear basic model obtained by using the cross basis function in the second step is as follows:

in the formula: g-family of chaining functions;

f (Y) -a dependent variable function of the groundwater level and the soil moisture content;

oc (rainfall, time) -a matrix of rainfall and time;

x_ijindependent variables- -i.e. the amount of rainfall and the time of analysis that varies day by day;

f_j-independent variable X_ijA basis function;

μ_k-functions of influencing factors such as water evaporation and infiltration in the landslide body;

β_j、γ_k-the corresponding influence coefficients in the function.

As a still further scheme of the invention: the method for adding the lag time into the basic model in the step two comprises the following steps: taking the change of underground water as an example, expressing the rainfall-underground water level relation of independent variable and dependent variable by using a function v (x), expressing the time-underground water level relation by using a function u (l), combining the two functions to obtain a two-dimensional rainfall-time-underground water relation function v, u (x, l), and simplifying and expressing the function as delta in the calculation and analysis process_i(x,l)。

As a still further scheme of the invention: the hysteresis effect relational expression established in the second step is as follows:

in the formula: theta-coefficient of hysteresis

L-maximum lag time, maximum lag time less than or equal to 24h

Chi-rainfall varying day by day

l-time of analysis

δ_i(x, l) is simplification of a v.u (x, l) function, the relation takes the hysteresis effect into consideration, and means that the change of the underground water level along with the rainfall and the time is analyzed, the process of rainfall penetrating into a slip slope body is taken into consideration, the underground water level is judged to be changed from the day, and the hysteresis time is obtained. And similarly, obtaining a hysteresis effect change function of the soil moisture content of the underground shallow layer.

As a still further scheme of the invention: and the distribution lag nonlinear model finally established in the step two adds lag time into the functional relations of rainfall-underground water level and rainfall-soil moisture content through a cross basis function, so that the distribution process that the dependent variable generates a lag effect along with the independent variable is realized.

As a still further scheme of the invention: and in the third step, the model can simultaneously evaluate the hysteresis effect and the nonlinear effect of rainfall on the underground water level and the soil moisture content.

As a still further scheme of the invention: when the model in the third step is applied to an actual monitoring and early warning system, the model obtains a real-time lag relationship according to the analysis of the acquired data, and predicts the change of the geological disaster point within 24 hours in the future within a time period with large rainfall, so that the aim of accurately predicting and effectively preventing the change is fulfilled, and the safety of the life and property of personnel in the geological disaster threatening area is ensured.

Compared with the prior art, the invention has the beneficial effects that: according to the invention, monitoring, forecasting and analysis of different landslides can be carried out by collecting monitoring data of different monitoring points, prediction of geological disaster points within 24h in the future in a time period with larger rainfall is realized, and the hysteresis effect and the nonlinear effect of the underground water level and the soil moisture content are evaluated at the same time, so that the aims of accurately predicting and effectively preventing landslides are achieved.

Drawings

FIG. 1 is a three-dimensional curve model diagram of the groundwater level as a function of rainfall and time according to the present invention,

FIG. 2 is a three-dimensional curve model diagram of the variation of soil moisture content with rainfall and time according to the invention.

Detailed Description

The technical solution of the present invention will be described in further detail with reference to specific embodiments.