A landslide displacement prediction method and system based on logistic regression optimized SVR-LSTM
Patent Information
- Application Number
- CN202310396175.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-04-14
- Publication Date
- 2026-08-21
- Estimated Expiration
- 2043-04-14
AI Technical Summary
但传统的预测模型往往存在预测精度不高，泛化能力较弱的问题；单一算法模型其自身可能存在易陷入局部最优值、收敛精度不高的问题
[0039]本发明的有益效果：本发明提供一种逻辑回归优化的SVR-LSTM滑坡位移预测方法及系统，使用两种不同的算法对数据集和测试集分别进行了预测，在得到两种不同算法的总位移预测结果之后，通过逻辑回归算法计算各预测值权重后得到的最终滑坡位移预测值，有效地构建了滑坡位移与因素之间的响应关系，有助于提高预测结果的准确性及稳定性，体现了该方法建立下的集成模型在滑坡易发区的滑坡位移预测中具有应用潜力。
Smart Images
Figure CN116522286B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of landslide displacement prediction technology, specifically to a logistic regression-optimized SVR-LSTM landslide displacement prediction method and system. Background Technology
[0002] Landslides are a common type of sudden geological disaster.
[0003] Currently, landslide displacement prediction has received significant attention from scholars both domestically and internationally, leading to its steady development. However, traditional prediction models often suffer from low prediction accuracy and weak generalization ability; single algorithm models may also be prone to getting trapped in local optima and exhibit low convergence accuracy. Summary of the Invention
[0004] The purpose of this section is to outline some aspects of embodiments of the present invention and to briefly describe some preferred embodiments. Simplifications or omissions may be made in this section, as well as in the abstract and title of this application, to avoid obscuring the purpose of these documents; however, such simplifications or omissions should not be construed as limiting the scope of the invention.
[0005] In view of the above-mentioned problems, the present invention is proposed.
[0006] A first aspect of this invention provides a logistic regression-optimized SVR-LSTM landslide displacement prediction method, comprising: collecting landslide displacement data from various monitoring points; decomposing the landslide displacement data into trend data and periodic data using the average moving average method; constructing a dataset based on the landslide displacement data and initially screening candidate input factors; performing Pearson correlation coefficient analysis on the candidate input factors and the periodic data to obtain model input factors; using support vector regression and LSTM algorithms to predict the trend data and periodic data respectively, and obtaining the total displacement prediction results corresponding to the two algorithms respectively; using the model input factors and the total displacement prediction results as candidate factors for a logistic regression classification algorithm; calculating the weights of the support vector regression model and the LSTM model using the logistic regression classification algorithm; and processing the data to obtain a logistic regression-optimized total displacement prediction value.
[0007] As a preferred embodiment of the logistic regression-optimized SVR-LSTM landslide displacement prediction method described in this invention, the landslide displacement data is decomposed into trend term data and periodic term data, including:
[0008] The fluctuations in the landslide displacement data are smoothed using the moving average method, and the trend data is extracted. The periodic data is extracted based on the moving average period.
[0009] The calculation of the average moving average method includes,
[0010]
[0011] in, X represents the value of the trend term displacement at time step t. t This represents the cumulative displacement at time step t, where n represents the moving average period.
[0012] As a preferred embodiment of the logistic regression-optimized SVR-LSTM landslide displacement prediction method described in this invention, wherein: the acquisition of the model input factors includes,
[0013] Pearson correlation coefficient analysis was performed on the candidate input factors and the periodic term data, and several candidate input factors corresponding to the maximum values of the Pearson correlation coefficients were selected as input factors in the periodic term displacement prediction model.
[0014] The calculation of the Pearson correlation coefficient includes,
[0015]
[0016] Where X and Y represent two vectors of equal length, and N represents the number of vector elements.
[0017] As a preferred embodiment of the logistic regression-optimized SVR-LSTM landslide displacement prediction method described in this invention, the prediction of trend term data and periodic term data includes:
[0018] The trend data and periodic data are predicted using support vector regression and LSTM algorithms, respectively. The trend data is predicted using a univariate prediction model, and the periodic data is predicted using a multivariate prediction model.
[0019] Based on the trend data, 55% to 65% of the data is used as the training set for the univariate prediction model, and 10% to 20% of the data is used as the validation set for the univariate prediction model to adjust the model's hyperparameters. The optimal prediction model for trend data based on support vector regression and LSTM algorithms is obtained. Based on the optimal prediction model for trend data, the prediction results for trend data based on support vector regression and LSTM algorithms are obtained respectively.
[0020] The above prediction method is used to obtain the prediction results of periodic terms based on the support vector regression algorithm and the LSTM algorithm.
[0021] As a preferred embodiment of the logistic regression-optimized SVR-LSTM landslide displacement prediction method described in this invention, the candidate factors for the logistic regression classification algorithm include:
[0022] The trend data prediction results and period data prediction results based on the support vector regression algorithm and LSTM algorithm are added together to obtain the total displacement prediction result based on the support vector regression algorithm and LSTM algorithm.
[0023] The model input factors, the total displacement prediction results based on the support vector regression algorithm, the total displacement prediction results based on the LSTM algorithm, and the difference between the two total displacement prediction results are used as candidate factors for the logistic regression classification algorithm.
[0024] As a preferred embodiment of the logistic regression-optimized SVR-LSTM landslide displacement prediction method described in this invention, the acquisition of the total displacement prediction value includes:
[0025] Pearson correlation coefficient analysis was performed on the candidate factors and the candidate input factors, and the factors with high correlation in the Pearson correlation coefficient analysis were selected as the input factors of the logistic regression classification algorithm model.
[0026] The first 60 sets of input factors are used as the training set for the logistic regression classification algorithm model to obtain the optimal classification prediction model. The remaining 12 sets of data are used as the prediction dataset. The weights of the support vector regression model and the LSTM model are calculated by the logistic regression classification algorithm. After processing the data, the total displacement prediction value optimized by logistic regression is obtained.
[0027] As a preferred embodiment of the logistic regression-optimized SVR-LSTM landslide displacement prediction method described in this invention, it further includes:
[0028] The displacement monitoring values and logistic regression-optimized total displacement prediction values in the predicted dataset are evaluated, and the root mean square error and mean absolute percentage error are calculated to achieve the prediction and evaluation of landslide displacement.
[0029] A second aspect of the present invention provides a logistic regression-optimized SVR-LSTM landslide displacement prediction system, comprising:
[0030] The data acquisition unit is used to collect landslide displacement data from each monitoring point and uses the average moving average method to decompose the landslide displacement data into trend data and periodic data.
[0031] The data processing unit is used to construct a dataset based on the landslide displacement data, preliminarily screen candidate input factors, perform Pearson correlation coefficient analysis on the candidate input factors and the periodic data to obtain model input factors, use support vector regression algorithm and LSTM algorithm to predict the trend data and periodic data respectively, and obtain the total displacement prediction results corresponding to the two algorithms respectively.
[0032] The displacement prediction unit is used to take the model input factors and the total displacement prediction results as candidate factors for the logistic regression classification algorithm, calculate the weights of the support vector regression model and the LSTM model through the logistic regression classification algorithm, and obtain the total displacement prediction value optimized by logistic regression after processing the data.
[0033] A third aspect of the present invention provides an apparatus, the apparatus comprising,
[0034] processor;
[0035] Memory used to store processor-executable instructions;
[0036] The processor is configured to invoke instructions stored in the memory to execute the method described in any embodiment of the present invention.
[0037] A fourth aspect of the present invention provides a computer-readable storage medium having stored thereon computer program instructions, including:
[0038] When the computer program instructions are executed by the processor, they implement the method as described in any embodiment of the present invention.
[0039] The beneficial effects of this invention are as follows: This invention provides a logistic regression-optimized SVR-LSTM landslide displacement prediction method and system. Two different algorithms are used to predict the dataset and test set respectively. After obtaining the total displacement prediction results of the two different algorithms, the final landslide displacement prediction value is obtained by calculating the weight of each prediction value through the logistic regression algorithm. This effectively constructs the response relationship between landslide displacement and factors, which helps to improve the accuracy and stability of the prediction results. It demonstrates that the integrated model established by this method has the application potential in landslide displacement prediction in landslide-prone areas. Attached Figure Description
[0040] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:
[0041] Figure 1 The present invention provides an overall flowchart of a logistic regression-optimized SVR-LSTM landslide displacement prediction method and system.
[0042] Figure 2 This is a schematic diagram of landslide displacement prediction results in an embodiment of the logistic regression-optimized SVR-LSTM landslide displacement prediction method and system provided by the present invention. Detailed Implementation
[0043] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.
[0044] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.
[0045] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.
[0046] This invention is described in detail with reference to the schematic diagrams. When detailing the embodiments of this invention, for ease of explanation, the cross-sectional views illustrating the device structure may be partially enlarged, not adhering to the usual scale. Furthermore, the schematic diagrams are merely examples and should not be construed as limiting the scope of protection of this invention. In actual fabrication, the three-dimensional spatial dimensions of length, width, and depth should be included.
[0047] Furthermore, in the description of this invention, it should be noted that the terms "upper," "lower," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. These terms are used solely for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. In addition, the terms "first," "second," or "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.
[0048] Unless otherwise explicitly specified and limited, the terms "installation," "connection," and "joining" in this invention should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; similarly, they can refer to mechanical connections, electrical connections, or direct connections, or indirect connections through an intermediate medium, or internal connections between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.
[0049] Example 1
[0050] Reference Figure 1 As one embodiment of the present invention, a logistic regression-optimized SVR-LSTM landslide displacement prediction method is provided, comprising:
[0051] S1: Collect landslide displacement data from each monitoring point, and decompose the landslide displacement data into trend data and periodic data using the average moving average method. It should be noted that:
[0052] Landslide displacement data includes displacement, rainfall, and reservoir water level;
[0053] Furthermore, the landslide displacement data is decomposed into trend data and periodic data, including...
[0054] The average moving average method is used to smooth the fluctuations in landslide displacement data and extract the trend data. The periodic data is extracted based on the moving average period.
[0055] Specifically, the calculation of the moving average method includes,
[0056]
[0057] in, X represents the value of the trend term displacement at time step t. t This represents the cumulative displacement at time step t, where n represents the moving average period.
[0058] It should be noted that the single-point moving average method calculates a moving average of time series data over a certain period of time, which can better reflect the trend and changes of the time series.
[0059] S2: Construct a dataset based on landslide displacement data, and initially screen candidate input factors. Perform Pearson correlation coefficient analysis on the candidate input factors and periodic term data to obtain model input factors. It should be noted that:
[0060] Obtaining model input factors includes,
[0061] Pearson correlation coefficient analysis was performed on the candidate input factors and periodic term data. Several candidate input factors corresponding to the maximum values of Pearson correlation coefficients were selected as input factors in the periodic term displacement prediction model.
[0062] Specifically, the calculation of the Pearson correlation coefficient includes,
[0063]
[0064] Where X and Y represent two vectors of equal length, and N represents the number of vector elements;
[0065] It should be noted that the Pearson correlation coefficient method specifies that the range of values for the relationship between two continuous variables is [-1, 1], which is used to measure the magnitude and direction of the relationship between the two continuous variables. 1 indicates a positive correlation, 0 indicates no correlation, and -1 indicates a negative correlation.
[0066] S3: Support Vector Regression (SVR) and LSTM algorithms are used to predict trend and periodic data respectively, and the total displacement prediction results for each algorithm are obtained. It should be noted that:
[0067] Predicting trend and periodic data includes,
[0068] Support vector regression and LSTM algorithms were used to predict trend data and periodic data, respectively. A univariate prediction model was used for trend data and a multivariate prediction model was used for periodic data.
[0069] Based on the trend data, 55%–65% of the data is used as the training set for the univariate prediction model, and 10%–20% of the data is used as the validation set for the univariate prediction model to adjust the model’s hyperparameters. The optimal prediction model for trend data based on support vector regression and LSTM algorithms is obtained. Based on the optimal prediction model for trend data, the prediction results for trend data based on support vector regression and LSTM algorithms are obtained respectively.
[0070] The above prediction method is used to obtain the prediction results of periodic terms based on the support vector regression algorithm and the LSTM algorithm.
[0071] S4: Using the model input factors and total displacement prediction results as candidate factors for the logistic regression classification algorithm, the weights of the support vector regression model and the LSTM model are calculated through the logistic regression classification algorithm. After processing the data, the total displacement prediction value optimized by logistic regression is obtained. It should be noted that:
[0072] The candidate factors for the logistic regression classification algorithm include:
[0073] The prediction results of the trend term data and the prediction results of the period term data based on the support vector regression algorithm and the LSTM algorithm are added together to obtain the total displacement prediction result based on the support vector regression algorithm and the LSTM algorithm.
[0074] The model input factors, the total displacement prediction results based on the support vector regression algorithm, the total displacement prediction results based on the LSTM algorithm, and the difference between the two total displacement prediction results are used as candidate factors for the logistic regression classification algorithm.
[0075] Furthermore, obtaining the total displacement prediction value includes,
[0076] Pearson correlation coefficient analysis was performed on the alternative factors and candidate input factors, and the factors with high correlation in the Pearson correlation coefficient analysis were selected as the input factors of the logistic regression classification algorithm model.
[0077] The first 60 input factors are used as the training set for the logistic regression classification algorithm model to obtain the optimal classification prediction model. The remaining 12 data are used as the prediction dataset. The weights of the support vector regression model and the LSTM model are calculated by the logistic regression classification algorithm. After processing the data, the total displacement prediction value optimized by logistic regression is obtained.
[0078] The displacement monitoring values and logistic regression-optimized total displacement prediction values in the prediction dataset are evaluated, and the root mean square error and mean absolute percentage error are calculated to achieve the prediction and evaluation of landslide displacement.
[0079] It should be noted that this invention provides a logistic regression-optimized SVR-LSTM landslide displacement prediction method and system. Two different algorithms are used to predict the dataset and test set respectively. After obtaining the total displacement prediction results of the two different algorithms, the final landslide displacement prediction value is obtained by calculating the weight of each prediction value through the logistic regression algorithm. This effectively constructs the response relationship between landslide displacement and factors, which helps to improve the accuracy and stability of the prediction results. It demonstrates that the integrated model established by this method has the application potential in landslide displacement prediction in landslide-prone areas.
[0080] The second aspect disclosed in this invention,
[0081] A logistic regression-optimized SVR-LSTM landslide displacement prediction system is provided, comprising:
[0082] The data acquisition unit is used to collect landslide displacement data from each monitoring point and uses the average moving average method to decompose the landslide displacement data into trend data and periodic data.
[0083] The data processing unit is used to construct a dataset based on landslide displacement data, preliminarily screen candidate input factors, perform Pearson correlation coefficient analysis on candidate input factors and periodic data to obtain model input factors, use support vector regression algorithm and LSTM algorithm to predict trend data and periodic data respectively, and obtain the total displacement prediction results corresponding to the two algorithms respectively.
[0084] The displacement prediction unit is used to take the model input factors and the total displacement prediction results as candidate factors for the LR classification algorithm. The weights of the support vector regression model and the LSTM model are calculated through the LR classification algorithm, and the data is processed to obtain the total displacement prediction value optimized by logistic regression.
[0085] The third aspect disclosed in this invention,
[0086] A device is provided, comprising:
[0087] processor;
[0088] Memory used to store processor-executable instructions;
[0089] The processor is configured to invoke instructions stored in memory to execute any of the aforementioned methods.
[0090] The fourth aspect disclosed in this invention,
[0091] A computer-readable storage medium is provided, having stored thereon computer program instructions, including:
[0092] A method by which computer program instructions are executed by a processor to implement any of the foregoing.
[0093] The present invention may be a method, apparatus, system and / or computer program product, and the computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for performing various aspects of the present invention.
[0094] Computer-readable storage media can be tangible devices capable of holding and storing instructions for use by an instruction execution device. Computer-readable storage media can be, for example—but not limited to—electrical storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any suitable combination thereof. More specific examples (a non-exhaustive list) of computer-readable storage media include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital multifunction disc (DVD), memory sticks, floppy disks, mechanical encoding devices, such as punch cards or recessed protrusions storing instructions thereon, and any suitable combination thereof. The computer-readable storage media used herein are not to be construed as transient signals themselves, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (e.g., light pulses through fiber optic cables), or electrical signals transmitted through wires.
[0095] Example 2
[0096] Reference Figure 2 This is the second embodiment of the present invention. Unlike the first embodiment, this embodiment provides a verification test of a logistic regression-optimized SVR-LSTM landslide displacement prediction method and system to verify and explain the technical effects used in this method.
[0097] This embodiment selects a landslide in Zigui County, Three Gorges Reservoir Area of the Yangtze River, and uses the method provided by the present invention to predict the landslide displacement.
[0098] First, landslide displacement-related data were collected from various monitoring points. In this embodiment, 84 sets of landslide displacement-related data were obtained from the landslide monitoring points, including the monthly rainfall, the cumulative rainfall over the past two months, the maximum daily rainfall in the current month, the number of rainy days in the current month, the average reservoir water level in the current month, the monthly reservoir water level change, the reservoir water level change over the past two months, the number of days the reservoir water level dropped in the current month, the cumulative drop in the current month, the number of days the reservoir water level rose in the current month, and the cumulative rise in the current month. The landslide displacement data was decomposed into trend data and displacement data using the moving average method. A dataset was constructed based on the landslide displacement data, and candidate input factors were initially screened. Pearson correlation coefficient analysis was performed on the candidate input factors and the periodic data to obtain the model input factors.
[0099] Secondly, the influencing factors are input into the original dataset, and the original dataset is decomposed into a fitting set and a prediction set according to a certain ratio. In this embodiment, 84 sets of data are input into the original dataset. The first 72 sets of data are used as the fitting set, and the remaining 12 sets of data are used as the prediction set. In the fitting set, 60 sets of data are used as the training set to train the model, and the remaining 12 sets of data are used as the validation set to adjust the hyperparameters of the model. After training, an optimal prediction model for the landslide trend term based on the SVR algorithm and the LSTM algorithm can be obtained, and the prediction results of the model are obtained. The prediction results of the periodic term and the prediction results of the trend term are added together to obtain the prediction results of the total landslide displacement based on the SVR algorithm and the LSTM algorithm, respectively.
[0100] Then, the model input factors, the predicted values of the SVR model, the predicted values of the LSTM model, and the difference between the predicted values of the SVR and LSTM models are used as candidate factors for the LR classification algorithm. The weights of the support vector regression model and the LSTM model are calculated through the LR classification algorithm, and the total displacement prediction value optimized by logistic regression is obtained after data processing.
[0101] Finally, the logistic regression-optimized SVR-LSTM model was compared with the SVR model and the LSTM model in terms of total displacement prediction results. The total prediction results of each model are as follows: Figure 2 As shown, the data comparison results are shown in Table 1;
[0102] Table 1: Data comparison results.
[0103] Model Total displacement RMSE (mm) SVR 34.58 LSTM 28.18 LR Optimization SVR-LSTM 26.48
[0104] based on Figure 2From Table 1, we can draw the following conclusions: ① Overall, the SVR model does not perform as well as the LSTM model, but the LSTM model does not perform well at all time steps; ② Compared with the SVR and LSTM models, the logistic regression-optimized SVR-LSTM model predicts the total displacement more closely to the measured value; ③ This example involves a logistic regression-optimized SVR-LSTM landslide displacement prediction method, which combines the advantages of both SVR and LSTM algorithms. It can effectively construct the response relationship between landslide displacement and factors, which helps to improve the accuracy and stability of the prediction results and has application potential in landslide displacement prediction in landslide-prone areas.
[0105] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.
Claims
1. A logistic regression-optimized SVR-LSTM landslide displacement prediction method, characterized in that, include: Collect landslide displacement data from each monitoring point, and use the average moving average method to decompose the landslide displacement data into trend data and periodic data; A dataset is constructed based on the landslide displacement data, and candidate input factors are initially screened. Pearson correlation coefficient analysis is performed on the candidate input factors and the periodic term data to obtain model input factors. The trend data and periodic data are predicted using support vector regression and LSTM algorithms respectively, and the total displacement prediction results corresponding to the two algorithms are obtained respectively. The model input factors and the total displacement prediction results are used as candidate factors for the logistic regression classification algorithm. The weights of the support vector regression model and the LSTM model are calculated by the logistic regression classification algorithm, and the data is processed to obtain the logistic regression optimized total displacement prediction value. The landslide displacement data is decomposed into trend data and periodic data, including: The fluctuations in the landslide displacement data are smoothed using the moving average method, and the trend data is extracted. The periodic data is extracted based on the moving average period. The calculation of the average moving average method includes, in, This indicates the displacement of the trend term at time step. t The value, Indicates the cumulative displacement at time step t The value, n Indicates the moving average period; The acquisition of the model input factors includes, Pearson correlation coefficient analysis was performed on the candidate input factors and the periodic term data, and several candidate input factors corresponding to the maximum values of the Pearson correlation coefficients were selected as input factors in the periodic term displacement prediction model. The calculation of the Pearson correlation coefficient includes, in, X and Y This represents two vectors of equal length. N Indicates the number of elements in the vector; The prediction of trend data and periodic data includes, The trend data and periodic data are predicted using support vector regression and LSTM algorithms, respectively. The trend data is predicted using a univariate prediction model, and the periodic data is predicted using a multivariate prediction model. Based on the trend data, 55% to 65% of the data is used as the training set for the univariate prediction model, and 10% to 20% of the data is used as the validation set for the univariate prediction model to adjust the model's hyperparameters. The optimal prediction model for trend data based on support vector regression and LSTM algorithms is obtained, and the prediction results for trend data based on support vector regression and LSTM algorithms are obtained based on the optimal prediction model for trend data. The above prediction method is used to obtain the prediction results of periodic terms based on the support vector regression algorithm and the LSTM algorithm; The candidate factors for the logistic regression classification algorithm include: The trend data prediction results and period data prediction results based on the support vector regression algorithm and LSTM algorithm are added together to obtain the total displacement prediction result based on the support vector regression algorithm and LSTM algorithm. The model input factors, the total displacement prediction results based on the support vector regression algorithm, the total displacement prediction results based on the LSTM algorithm, and the difference between the two total displacement prediction results are used as candidate factors for the logistic regression classification algorithm. The acquisition of the total displacement prediction value includes, Pearson correlation coefficient analysis was performed on the candidate factors and the candidate input factors, and the factors with high correlation in the Pearson correlation coefficient analysis were selected as the input factors of the logistic regression classification algorithm model. The first 60 sets of input factors are used as the training set for the logistic regression classification algorithm model to obtain the optimal classification prediction model. The remaining 12 sets of data are used as the prediction dataset. The weights of the support vector regression model and the LSTM model are calculated by the logistic regression classification algorithm. After processing the data, the total displacement prediction value optimized by logistic regression is obtained. The displacement monitoring values and logistic regression-optimized total displacement prediction values in the predicted dataset are evaluated, and the root mean square error and mean absolute percentage error are calculated to achieve the prediction and evaluation of landslide displacement.
2. A logistic regression-optimized SVR-LSTM landslide displacement prediction system, applying the logistic regression-optimized SVR-LSTM landslide displacement prediction method as described in claim 1, characterized in that, include: The data acquisition unit is used to collect landslide displacement data from each monitoring point and uses the average moving average method to decompose the landslide displacement data into trend data and periodic data. The data processing unit is used to construct a dataset based on the landslide displacement data, preliminarily screen candidate input factors, perform Pearson correlation coefficient analysis on the candidate input factors and the periodic data to obtain model input factors, use support vector regression algorithm and LSTM algorithm to predict the trend data and periodic data respectively, and obtain the total displacement prediction results corresponding to the two algorithms respectively. The displacement prediction unit is used to take the model input factors and the total displacement prediction results as candidate factors for the logistic regression classification algorithm, calculate the weights of the support vector regression model and the LSTM model through the logistic regression classification algorithm, and obtain the total displacement prediction value optimized by logistic regression after processing the data.
3. A device, characterized in that, The device includes, processor; Memory used to store processor-executable instructions; The processor is configured to invoke instructions stored in the memory to execute the method of claim 1.
4. A computer-readable storage medium having computer program instructions stored thereon, characterized in that, When the computer program instructions are executed by the processor, they implement the method of claim 1.