CN113111573A - Landslide displacement prediction method based on GRU - Google Patents

Landslide displacement prediction method based on GRU Download PDF

Info

Publication number
CN113111573A
CN113111573A CN202110311836.3A CN202110311836A CN113111573A CN 113111573 A CN113111573 A CN 113111573A CN 202110311836 A CN202110311836 A CN 202110311836A CN 113111573 A CN113111573 A CN 113111573A
Authority
CN
China
Prior art keywords
model
gru
landslide
landslide displacement
value
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202110311836.3A
Other languages
Chinese (zh)
Other versions
CN113111573B (en
Inventor
孙希延
林子安
何清
白杨
付文涛
梁维彬
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Guilin University of Electronic Technology
Original Assignee
Guilin University of Electronic Technology
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Guilin University of Electronic Technology filed Critical Guilin University of Electronic Technology
Priority to CN202110311836.3A priority Critical patent/CN113111573B/en
Publication of CN113111573A publication Critical patent/CN113111573A/en
Application granted granted Critical
Publication of CN113111573B publication Critical patent/CN113111573B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
    • G01B21/00—Measuring arrangements or details thereof, where the measuring technique is not covered by the other groups of this subclass, unspecified or not relevant
    • G01B21/02—Measuring arrangements or details thereof, where the measuring technique is not covered by the other groups of this subclass, unspecified or not relevant for measuring length, width, or thickness
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
    • G01B21/00—Measuring arrangements or details thereof, where the measuring technique is not covered by the other groups of this subclass, unspecified or not relevant
    • G01B21/32—Measuring arrangements or details thereof, where the measuring technique is not covered by the other groups of this subclass, unspecified or not relevant for measuring the deformation in a solid
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/048—Activation functions
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/084—Backpropagation, e.g. using gradient descent

Abstract

The invention discloses a landslide displacement prediction method based on GRU, which monitors landslide earth surface displacement deformation, and calculates a replacement value of an abnormal value in obtained landslide displacement data by utilizing a constructed polynomial model; carrying out mean normalization processing on all the landslide displacement data after being replaced by the replacement values, and constructing a GRU model, a training set, a verification set and a test set; performing loop iteration on the GRU model by using the training set, and performing parameter adjustment on the trained GRU model by using the verification set; and testing the GRU model after the parameters are adjusted by using the test set, and performing inverse mean normalization processing on the obtained model prediction result to complete landslide displacement prediction and improve prediction precision.

Description

Landslide displacement prediction method based on GRU
Technical Field
The invention relates to the technical field of geological disaster monitoring and prediction, in particular to a landslide displacement prediction method based on GRUs.
Background
Landslide is one of the most devastating geological disasters in many areas of the world, causing a large number of casualties and very important property losses each year. A reliable landslide prediction early warning method is a reasonable approach for reducing landslide risk. If the method capable of accurately predicting the landslide displacement can be successfully implemented, the method has great social value and important economic benefit.
Landslide prediction models can be divided into two broad categories, physical models and data-driven models. Compared with a physical model, the data-driven model is more popular because of simple process, accurate prediction and lower cost, but abnormal values are usually ignored blindly in the prediction process, and the prediction precision is not high due to the complicated structure of the currently commonly used RNN and LSTM models.
Disclosure of Invention
The invention aims to provide a landslide displacement prediction method based on GRU, and the prediction precision is improved.
In order to achieve the above object, the present invention provides a landslide displacement prediction method based on GRU, comprising the following steps:
monitoring landslide surface displacement deformation, and calculating a replacement value of an abnormal value in the obtained landslide displacement data by utilizing a constructed polynomial model;
carrying out mean normalization processing on all the landslide displacement data after being replaced by the replacement values, and constructing a GRU model, a training set, a verification set and a test set;
performing loop iteration on the GRU model by using the training set, and performing parameter adjustment on the trained GRU model by using the verification set;
and testing the GRU model after the parameters are adjusted by using the test set, and performing anti-mean normalization processing on the obtained model prediction result to finish landslide displacement prediction.
The method comprises the following steps of monitoring landslide surface displacement deformation, and calculating a replacement value of an abnormal value in obtained landslide displacement data by utilizing a constructed polynomial model, wherein the replacement value comprises the following steps:
monitoring landslide ground surface displacement deformation to obtain a plurality of landslide displacement data, wherein the plurality of landslide displacement data comprise a plurality of normal values and a plurality of abnormal values;
and taking a plurality of normal value serial numbers except the abnormal values as independent variables, taking a plurality of normal values as dependent variables to construct a polynomial model, inputting the abnormal value serial numbers into the polynomial model, and obtaining an output value of the polynomial model, namely a replacement value of the abnormal values.
Wherein, using the training set to perform loop iteration on the GRU model, and using the validation set to perform parameter adjustment on the trained GRU model comprises:
calculating values of an update gate, a reset gate, a hidden state and a current candidate state in the GRU model by using the landslide displacement data in the training set;
calculating the difference value between the landslide displacement data and the corresponding hidden state, and summing all the difference values to obtain a corresponding error term;
adjusting the corresponding weight value based on the error term until the iteration termination condition is reached, and finishing the training of the GRU model;
and verifying the trained GRU model by using the verification set, and adjusting the model learning rate and the number of corresponding model nodes.
Adjusting a corresponding weight value based on the error term until an iteration termination condition is reached, and completing training of the GRU model, wherein the training comprises the following steps:
respectively deriving the reset gate weight, the updated gate weight and the candidate state weight by using the error term to obtain corresponding derivatives;
subtracting the product of the corresponding derivative and a model learning rate from the reset gate weight, the updated gate weight and the candidate state weight to obtain a new reset gate weight, an updated gate weight and a candidate state weight;
and finishing the training of the GRU model until the current iteration times reach a set iteration threshold or the sum of the error terms is less than a set value.
The method comprises the following steps of testing the GRU model after parameters are adjusted by using the test set, and carrying out anti-mean normalization processing on an obtained model prediction result to complete landslide displacement prediction, wherein the method comprises the following steps:
inputting the test set into the GRU model after parameters are adjusted to obtain a corresponding model prediction result;
and carrying out anti-mean normalization processing on the model prediction result to obtain final measurement data, and completing landslide displacement prediction.
The landslide displacement prediction method based on GRU monitors landslide surface displacement deformation, and calculates a replacement value of an abnormal value in obtained landslide displacement data by utilizing a constructed polynomial model; carrying out mean normalization processing on all the landslide displacement data after being replaced by the replacement values, and constructing a GRU model, a training set, a verification set and a test set; performing loop iteration on the GRU model by using the training set, and performing parameter adjustment on the trained GRU model by using the verification set; and testing the GRU model after the parameters are adjusted by using the test set, and performing inverse mean normalization processing on the obtained model prediction result to complete landslide displacement prediction and improve prediction precision.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
Fig. 1 is a schematic step diagram of a method for predicting landslide displacement based on GRU according to the present invention.
Fig. 2 is a schematic flow chart of a method for predicting landslide displacement based on a GRU according to the present invention.
Fig. 3 is a diagram of a GRU model structure provided by the present invention.
FIG. 4 is a graph of actual raw data for a GRU model prediction provided by the present invention.
FIG. 5 is a schematic diagram of a polynomial model provided by the present invention.
FIG. 6 is a graph of normalized actual raw data for GRU model prediction provided by the present invention.
FIG. 7 is a diagram of a GRU model prediction result provided by the present invention.
Detailed Description
Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to the same or similar elements or elements having the same or similar function throughout. The embodiments described below with reference to the drawings are illustrative and intended to be illustrative of the invention and are not to be construed as limiting the invention.
In the description of the present invention, "a plurality" means two or more unless specifically defined otherwise.
Referring to fig. 1 and 2, the present invention provides a method for predicting landslide displacement based on GRU, comprising the following steps:
s101, monitoring displacement deformation of the landslide surface, and calculating a replacement value of an abnormal value in the obtained landslide displacement data by using the constructed polynomial model.
Specifically, a high-precision GPS receiver or a landslide displacement sensor or a remote sensing satellite or manual measurement is used for monitoring the displacement and deformation of the landslide earth surface. And carrying out abnormal value replacement processing on the obtained landslide displacement data, and constructing a polynomial model to calculate a replacement value of the abnormal value.
The process is as follows: let the polynomial model be y ═ anxn+an-1xn-1+...+a1x+a0X is an independent variable, aiConstant but unknown, y is the dependent variable and n is the highest power of the polynomial. Suppose there are m data, x1，x2，x3，...xmLet x be3If the abnormal value is found, the independent variables are numbers 1, 2 and 4 … m, x1，x2，x4，...xmConstructing a polynomial model as a dependent variable to obtain y ═ anxn+an-1xn-1+...+a1x+a0At this time aiThe value of y is calculated by substituting the number 3 into the polynomial model for the known constant, where the value of y is x3The replacement value of (a). All abnormal values are replaced by analogy, and the problem that the prediction precision is low due to the fact that the abnormal values in the landslide displacement data are not processed or directly eliminated is solved.
S102, performing mean normalization processing on all the landslide displacement data after replacement by using the replacement values, and constructing a GRU model, a training set, a verification set and a test set.
Specifically, the mean normalization processing is performed on all the landslide displacement data after the abnormal value is replaced by the replacement value, so that the data is stable. Mean normalization procedure: y isi＝(xi-μ)/σ，xiFor the ith raw data, μ is the mean of the x data set and σ is the standard deviation of the x data set.
The GRU model is built as shown in fig. 3.
The GRU is a typical neural network model and comprises an input layer, a hidden layer and an output layer, wherein the input layer is connected with the hidden layer, the hidden layer is connected with the output layer, n nodes are arranged in the hidden layer and are also connected with each other, and each connection has the weight; the landslide displacement data are divided into a training set, a verification set and a test set according to the proportion of 0.6, 0.2 and 0.2 (the proportion can be other proportions, and is not necessarily limited to the three proportions), the training set is input from an input layer and is used for training a GRU (generalized regression unit) model, the model is better, the verification set is used for performing cross verification after the model is trained, the model is optimal, a prediction result is output by an output layer and is compared with the test set to obtain the estimation accuracy of the model, and the problem that the prediction accuracy is not high because a machine learning algorithm only divides the data into the training set and the test set is solved.
S103, performing loop iteration on the GRU model by using the training set, and performing parameter adjustment on the trained GRU model by using the verification set.
Specifically, the landslide displacement data in the training set is used for calculating values of an update gate, a reset gate, a hidden state and a current candidate state in the GRU model, calculating differences between the landslide displacement data and the corresponding hidden state, summing all the differences to obtain corresponding error terms, deriving a reset gate weight, an update gate weight and a candidate state weight respectively by using the error terms to obtain corresponding derivatives, and subtracting a product of the corresponding derivatives and a model learning rate from the reset gate weight, the update gate weight and the candidate state weight to obtain new reset gate weight, the update gate weight and the candidate state weight; and completing the training of the GRU model until the current iteration times reach a set iteration threshold or the sum of the error terms is less than a set value, adjusting part of parameters of the model in each iteration, and comprising the following detailed procedures:
xtdenotes the t-th data, ztUpdate gate, r, indicating time ttReset gate, h, indicating time ttShowing a hidden state at time t,
Figure BDA0002990092680000051
Represents the current candidate state at time t, WrWeight, W, representing reset gatezA weight representing an update gate,
Figure BDA00029900926800000512
Weight representing candidate state, sigma representing sigmoid function, tanh representing tanh function,
Figure BDA0002990092680000052
Representing the model learning rate. Wherein]Representing the concatenation of two vectors, representing the product of the matrices.
rt＝σ(Wr·[ht-1,xt])
zt＝σ(Wz·[ht-1,xt])
Figure BDA0002990092680000053
Figure BDA0002990092680000054
For each iteration, the GRU has two directions of computation propagation, forward propagation and backward propagation, respectively. Forward propagation: calculating rt、zt、ht、
Figure BDA0002990092680000055
The value of (c), back propagation: calculating xtAnd htAre summed to obtain error terms, which are respectively coupled to Wr、Wz、
Figure BDA0002990092680000056
Derivation to obtain the derivative of the error term to each weight, Wr、Wz、
Figure BDA0002990092680000057
Subtract the product of the derivative and the learning rate to obtain the new Wr、Wz、
Figure BDA0002990092680000058
And (5) circularly iterating until the set times or the sum of the error terms is less than the set value, and finishing the model training.
Validating the GRU model using the validation set data after training is complete, and adjusting includes
Figure BDA0002990092680000059
Parameters including (adjustment learning rate)
Figure BDA00029900926800000510
And node number n), last W of the usage modelr、Wz、
Figure BDA00029900926800000511
And keeping the same, the learning rate is increased from small to 1 (after the learning rate is changed every time, the number of nodes is from 2 to n2Increasing), the error term and the learning rate (and the number of nodes) at each time are recorded, and the learning rate at the time with the minimum error term is adopted as the final learning rate (and the number of final nodes n).
And S104, testing the GRU model after the parameters are adjusted by using the test set, and performing anti-mean normalization processing on the obtained model prediction result to finish landslide displacement prediction.
Specifically, after verification is completed, the test set is used for testing the GRU model to obtain a model prediction result, the model prediction result is subjected to inverse mean normalization processing to obtain final landslide prediction displacement data, namely final test data, of the GRU model, and the inverse mean normalization process comprises the following steps: y isi＝xiσ + μ, μ is the mean of the original x data set, σ is the standard deviation in the original x data set, yiFor final prediction of data, xiIs the ith data calculated in the model.
Referring to fig. 4 to 7, fig. 4 is a graph of actual raw data of GRU model prediction provided by the present invention, which is a monthly displacement data value from 12 months of 06 to 12 months of 12 of the monitoring point of the white water river landslide ZG 118.
Fig. 5 is a schematic diagram of a polynomial model constructed using 72 normal values, where the polynomial model of the power of 5 is used in the current model.
Fig. 6 is a data diagram of normalized actual original data predicted by the GRU model provided by the present invention, and the original data is normalized by 73 landslide displacement data after replacing an abnormal value.
FIG. 7 is a diagram of a GRU model prediction result provided by the present invention. And selecting the last 12 data of the data as a test set, wherein the curve part in the graph is actual landslide displacement data, and the x part in the graph is predicted displacement predicted by the model. As can be seen from the figure, the prediction accuracy of the GRU is very high.
Advantageous effects
1. The statistical characteristic and the development regularity of the time sequence data are considered, and the accuracy of landslide displacement prediction is improved.
2. The landslide displacement prediction method can predict landslide displacement under the conditions of less actual landslide displacement data and less kinds of influence factor data, and reduces the prediction difficulty.
3. Compared with the traditional static landslide displacement prediction algorithm, the GRU model is a dynamic model prediction algorithm and is more suitable for practical needs.
4. The method solves the problems of gradient explosion and gradient disappearance which can occur in a neural network algorithm.
5. The data are divided into a training set, a verification set and a test set, so that the landslide displacement prediction model is more accurate.
6. The problem of low efficiency due to complex structure in a neural network algorithm is solved.
7. And considering the continuity of the landslide data, and solving the problem of abnormal values by constructing a polynomial model.
The landslide displacement prediction method based on GRU monitors landslide surface displacement deformation, and calculates a replacement value of an abnormal value in obtained landslide displacement data by utilizing a constructed polynomial model; carrying out mean normalization processing on all the landslide displacement data after being replaced by the replacement values, and constructing a GRU model, a training set, a verification set and a test set; performing loop iteration on the GRU model by using the training set, and performing parameter adjustment on the trained GRU model by using the verification set; and testing the GRU model after the parameters are adjusted by using the test set, and performing inverse mean normalization processing on the obtained model prediction result to complete landslide displacement prediction and improve prediction precision.
While the invention has been described with reference to a preferred embodiment, it will be understood by those skilled in the art that various changes in form and detail may be made therein without departing from the spirit and scope of the invention as defined by the appended claims.

Claims (5)

1. A landslide displacement prediction method based on GRU is characterized by comprising the following steps:
monitoring landslide surface displacement deformation, and calculating a replacement value of an abnormal value in the obtained landslide displacement data by utilizing a constructed polynomial model;
carrying out mean normalization processing on all the landslide displacement data after being replaced by the replacement values, and constructing a GRU model, a training set, a verification set and a test set;
performing loop iteration on the GRU model by using the training set, and performing parameter adjustment on the trained GRU model by using the verification set;
and testing the GRU model after the parameters are adjusted by using the test set, and performing anti-mean normalization processing on the obtained model prediction result to finish landslide displacement prediction.
2. The GRU-based landslide displacement prediction method of claim 1 wherein monitoring landslide surface displacement deformation and using a constructed polynomial model to calculate a replacement value for the outlier in the resulting landslide displacement data comprises:
monitoring landslide ground surface displacement deformation to obtain a plurality of landslide displacement data, wherein the plurality of landslide displacement data comprise a plurality of normal values and a plurality of abnormal values;
and taking a plurality of normal value serial numbers except the abnormal values as independent variables, taking a plurality of normal values as dependent variables to construct a polynomial model, inputting the abnormal value serial numbers into the polynomial model, and obtaining an output value of the polynomial model, namely a replacement value of the abnormal values.
3. The GRU-based landslide displacement prediction method of claim 1 wherein performing a loop iteration on the GRU model using the training set and performing a parameter adjustment on the trained GRU model using the validation set comprises:
calculating values of an update gate, a reset gate, a hidden state and a current candidate state in the GRU model by using the landslide displacement data in the training set;
calculating the difference value between the landslide displacement data and the corresponding hidden state, and summing all the difference values to obtain a corresponding error term;
adjusting the corresponding weight value based on the error term until the iteration termination condition is reached, and finishing the training of the GRU model;
and verifying the trained GRU model by using the verification set, and adjusting the model learning rate and the number of corresponding model nodes.
4. The GRU-based landslide displacement prediction method of claim 3 wherein adjusting the corresponding weight value based on the error term until completion of training of the GRU model after reaching an iteration termination condition comprises:
respectively deriving the reset gate weight, the updated gate weight and the candidate state weight by using the error term to obtain corresponding derivatives;
subtracting the product of the corresponding derivative and a model learning rate from the reset gate weight, the updated gate weight and the candidate state weight to obtain a new reset gate weight, an updated gate weight and a candidate state weight;
and finishing the training of the GRU model until the current iteration times reach a set iteration threshold or the sum of the error terms is less than a set value.
5. The GRU-based landslide displacement prediction method of claim 1 wherein testing the GRU model after adjusting parameters using the test set and performing an inverse mean normalization of the obtained model prediction results to complete landslide displacement prediction comprises:
inputting the test set into the GRU model after parameters are adjusted to obtain a corresponding model prediction result;
and carrying out anti-mean normalization processing on the model prediction result to obtain final measurement data, and completing landslide displacement prediction.
CN202110311836.3A 2021-03-24 2021-03-24 Landslide displacement prediction method based on GRU Active CN113111573B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110311836.3A CN113111573B (en) 2021-03-24 2021-03-24 Landslide displacement prediction method based on GRU

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110311836.3A CN113111573B (en) 2021-03-24 2021-03-24 Landslide displacement prediction method based on GRU

Publications (2)

Publication Number Publication Date
CN113111573A true CN113111573A (en) 2021-07-13
CN113111573B CN113111573B (en) 2022-09-23

Family

ID=76710839

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110311836.3A Active CN113111573B (en) 2021-03-24 2021-03-24 Landslide displacement prediction method based on GRU

Country Status (1)

Country Link
CN (1) CN113111573B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115511191A (en) * 2022-09-30 2022-12-23 固远晨通科技发展有限公司 High and steep slope displacement point prediction method

Citations (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107193782A (en) * 2017-04-18 2017-09-22 中国船舶重工集团公司第七〇九研究所 A kind of method of abnormal value removing and correction fitted based on multinomial
CN107463633A (en) * 2017-07-17 2017-12-12 中国航天系统科学与工程研究院 A kind of real time data rejecting outliers method based on EEMD neutral nets
CN107578124A (en) * 2017-08-28 2018-01-12 国网山东省电力公司电力科学研究院 The Short-Term Load Forecasting Method of GRU neutral nets is improved based on multilayer
CN108205889A (en) * 2017-12-29 2018-06-26 长春理工大学 Freeway traffic flow Forecasting Methodology based on convolutional neural networks
CN110119845A (en) * 2019-05-11 2019-08-13 北京京投亿雅捷交通科技有限公司 A kind of application method of track traffic for passenger flow prediction
CN110363344A (en) * 2019-07-11 2019-10-22 安徽理工大学 Probability integral parameter prediction method based on MIV-GP algorithm optimization BP neural network
US20190325328A1 (en) * 2018-04-19 2019-10-24 Ptc Inc. Detection and use of anomalies in an industrial environment
CN110457655A (en) * 2019-08-12 2019-11-15 长安大学 Slope Deformation Prediction method
CN110470259A (en) * 2019-07-09 2019-11-19 西安工程大学 Landslide displacement dynamic prediction method based on LSTM
CN111274738A (en) * 2020-02-25 2020-06-12 东南大学 Elevation abnormal value calculation method
US20200210264A1 (en) * 2018-12-28 2020-07-02 AO Kaspersky Lab System and method of generating data for monitoring of a cyber-physical system for early determination of anomalies
CN111444168A (en) * 2020-03-26 2020-07-24 易电务(北京)科技有限公司 Distribution room transformer daily maximum load abnormal data detection processing method
CN111784044A (en) * 2020-06-29 2020-10-16 杭州鲁尔物联科技有限公司 Landslide prediction method, device, equipment and storage medium
CN112150209A (en) * 2020-06-19 2020-12-29 南京理工大学 Construction method of CNN-LSTM time sequence prediction model based on clustering center

Patent Citations (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107193782A (en) * 2017-04-18 2017-09-22 中国船舶重工集团公司第七〇九研究所 A kind of method of abnormal value removing and correction fitted based on multinomial
CN107463633A (en) * 2017-07-17 2017-12-12 中国航天系统科学与工程研究院 A kind of real time data rejecting outliers method based on EEMD neutral nets
CN107578124A (en) * 2017-08-28 2018-01-12 国网山东省电力公司电力科学研究院 The Short-Term Load Forecasting Method of GRU neutral nets is improved based on multilayer
CN108205889A (en) * 2017-12-29 2018-06-26 长春理工大学 Freeway traffic flow Forecasting Methodology based on convolutional neural networks
US20190325328A1 (en) * 2018-04-19 2019-10-24 Ptc Inc. Detection and use of anomalies in an industrial environment
US20200210264A1 (en) * 2018-12-28 2020-07-02 AO Kaspersky Lab System and method of generating data for monitoring of a cyber-physical system for early determination of anomalies
CN110119845A (en) * 2019-05-11 2019-08-13 北京京投亿雅捷交通科技有限公司 A kind of application method of track traffic for passenger flow prediction
CN110470259A (en) * 2019-07-09 2019-11-19 西安工程大学 Landslide displacement dynamic prediction method based on LSTM
CN110363344A (en) * 2019-07-11 2019-10-22 安徽理工大学 Probability integral parameter prediction method based on MIV-GP algorithm optimization BP neural network
CN110457655A (en) * 2019-08-12 2019-11-15 长安大学 Slope Deformation Prediction method
CN111274738A (en) * 2020-02-25 2020-06-12 东南大学 Elevation abnormal value calculation method
CN111444168A (en) * 2020-03-26 2020-07-24 易电务(北京)科技有限公司 Distribution room transformer daily maximum load abnormal data detection processing method
CN112150209A (en) * 2020-06-19 2020-12-29 南京理工大学 Construction method of CNN-LSTM time sequence prediction model based on clustering center
CN111784044A (en) * 2020-06-29 2020-10-16 杭州鲁尔物联科技有限公司 Landslide prediction method, device, equipment and storage medium

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
何高清 等: "轴承尺寸检测数据的异常值检测与数据处理研究", 《机电工程》 *
刘振 等: "稳健回归分析在GPS高程拟合中的应用", 《矿山测量》 *
方海泉 等: "基于EEMD的水资源监测数据异常值检测与校正", 《农业机械学报》 *
鄢好 等: "基于时间序列和GRU的滑坡位移预测", 《人民长江》 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115511191A (en) * 2022-09-30 2022-12-23 固远晨通科技发展有限公司 High and steep slope displacement point prediction method

Also Published As

Publication number Publication date
CN113111573B (en) 2022-09-23

Similar Documents

Publication Publication Date Title
CN111008502B (en) Fault prediction method for complex equipment driven by digital twin
CN111563706A (en) Multivariable logistics freight volume prediction method based on LSTM network
CN111539132B (en) Dynamic load time domain identification method based on convolutional neural network
CN113256005B (en) Power station water level process prediction method and equipment based on neural network model
CN113537469B (en) Urban water demand prediction method based on LSTM network and Attention mechanism
CN114897277B (en) Burst-type landslide displacement prediction method based on LSTM
CN113111573B (en) Landslide displacement prediction method based on GRU
Samantaray et al. Sediment assessment for a watershed in arid region via neural networks
CN113592144A (en) Medium-and-long-term runoff probability forecasting method and system
CN111680398B (en) Single machine performance degradation prediction method based on Holt-windows model
CN113505534A (en) Load prediction method considering demand response
CN114707712A (en) Method for predicting requirement of generator set spare parts
CN116451514A (en) Bridge structure state evaluation method and device and electronic equipment
CN116432543A (en) Method for predicting remaining life of power semiconductor module, terminal device and storage medium
CN109540089B (en) Bridge deck elevation fitting method based on Bayes-Kriging model
CN113033091A (en) Landslide displacement prediction method
CN112381591A (en) Sales prediction optimization method based on LSTM deep learning model
CN112307536A (en) Dam seepage parameter inversion method
CN116562454A (en) Manufacturing cost prediction method applied to BIM long-short-time attention mechanism network
CN116910534A (en) Space-time intelligent prediction method and device for ocean environmental elements in different sea areas
CN115510748A (en) Landslide displacement prediction method based on variational modal decomposition and CNN-GRU
Matsoukas et al. Fusion of radar and rain gage measurements for an accurate estimation of rainfall
CN111641471B (en) Weight design method for prediction in atomic clock signal combination control
Xu et al. Variation of dominant discharge along the riverbed based on numerical and deep-learning models: A case study in the Middle Huaihe River, China
CN114638421A (en) Method for predicting requirement of generator set spare parts

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant
EE01 Entry into force of recordation of patent licensing contract

Application publication date: 20210713

Assignee: Guangxi Xinyang Shiji Information Technology Co.,Ltd.

Assignor: GUILIN University OF ELECTRONIC TECHNOLOGY

Contract record no.: X2022450000415

Denomination of invention: Prediction Method of Landslide Displacement Based on GRU

Granted publication date: 20220923

License type: Common License

Record date: 20221227

EE01 Entry into force of recordation of patent licensing contract