Reservoir bank landslide deformation prediction method and system based on water level fluctuation

Abstract

The invention provides a reservoir bank landslide deformation prediction method and system based on water level fluctuation, and relates to the technical field of landslide deformation prediction. The method comprises the following steps: acquiring monitoring data; preprocessing the monitoring data to obtain sample data; calculating a Pearson coefficient between reservoir water level change and landslide deformation; whether correlation exists between reservoir water level change and landslide deformation or not is judged; calculating an initial lag time length of influence of reservoir water level rise on landslide deformation; adjusting the initial lagging time length to obtain a dynamic lagging time length; based on the dynamic lag duration, establishing a time lag GM-BP deformation prediction model; inputting the sample data into a time-delay GM-BP deformation prediction model for training; acquiring real-time sample data; inputting real-time sample data into the trained time-delay GM-BP deformation prediction model, and outputting a final prediction result; and performing early warning according to the final prediction result. The accuracy of landslide deformation prediction can be effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of landslide deformation prediction, and in particular to a method and system for predicting reservoir bank landslide deformation based on water level fluctuation. Background Art

[0002] The reservoir bank landslide deformation prediction method based on water level fluctuation mainly studies the impact of reservoir water level changes on reservoir bank landslide deformation, especially the hysteresis effect caused by water level fluctuation. Reservoir bank landslide refers to the deformation and sliding of the soil on the reservoir slope due to factors such as water level fluctuation and soil saturation change. Water level fluctuation refers to the ups and downs of the reservoir water level caused by seasonal changes, precipitation, water storage and other factors.

[0003] In areas around reservoirs and hydropower stations, landslides pose a threat to infrastructure, the ecological environment, and human safety. Accurate deformation prediction can provide early warning for landslide disaster prevention and control, reducing loss of life and property.

[0004] However, existing methods for predicting reservoir bank landslide deformation fail to fully consider the dynamic adjustment and updating of real-time data and can only provide static prediction results. When considering the impact of water level changes on landslides, qualitative analysis is usually performed, lacking quantitative analysis, ignoring the changes in hysteresis effects, and making it difficult to provide accurate predictions and assessments. When making predictions, most methods use a single prediction method and cannot fully utilize the advantages of multiple models, resulting in limitations in practical applications and inaccurate prediction results. Summary of the Invention

[0005] In order to solve the technical problems that the existing reservoir bank landslide deformation prediction methods fail to fully consider the dynamic adjustment and update of real-time data and can only provide static prediction results, when considering the impact of water level changes on landslides, qualitative analysis is usually performed, quantitative analysis is lacking, and the changes in hysteresis effects are ignored, making it difficult to provide accurate predictions and evaluations. When making predictions, most of them adopt a single prediction method and cannot fully utilize the advantages of multiple models, resulting in limitations in practical applications and inaccurate prediction results, the present invention provides a reservoir bank landslide deformation prediction method and system based on water level fluctuations.

[0006] The technical solutions provided by the embodiments of the present invention are as follows: First aspect: An embodiment of the present invention provides a method for predicting reservoir bank landslide deformation based on water level fluctuation, comprising: S1: Acquire monitoring data, wherein the monitoring data includes reservoir water level fluctuation data, landslide deformation data and environmental data; S2: Preprocess the monitoring data, including abnormal data removal, missing data completion, and smoothing and noise reduction, to obtain sample data; S3: Calculate the Pearson coefficient between reservoir water level change and landslide deformation based on the sample data; S4: Based on the Pearson coefficient, determine whether there is a correlation between the reservoir water level change and the landslide deformation; if so, proceed to step S5, otherwise return to step S1; S5: Calculate the initial lag time of the impact of reservoir water level rise on landslide deformation; S6: According to the Pearson coefficient, the initial lag time is adjusted to obtain the dynamic lag time; S7: Based on the dynamic hysteresis time, a time-lag GM-BP deformation prediction model is established; S8: Input the sample data into the time-lag GM-BP deformation prediction model for training; S9: Get real-time sample data; S10: Input the real-time sample data into the trained time-delay GM-BP deformation prediction model and output the final prediction result; S11: Issue early warning based on the final prediction results.

[0007] Second aspect: An embodiment of the present invention provides a reservoir bank landslide deformation prediction system based on water level fluctuation, comprising: processor; The memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the reservoir bank landslide deformation prediction method based on water level fluctuation as described in the first aspect is implemented.

[0008] The third aspect: An embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon. When the program is executed by a processor, the method for predicting reservoir bank landslide deformation based on water level fluctuation according to the first aspect is implemented.

[0009] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least: In an embodiment of the present invention, by calculating the Pearson coefficient between reservoir water level changes and landslide deformation, the correlation between the two can be accurately determined, avoiding the limitations of relying solely on assumptions. By dynamically adjusting the lag time, the limitation of fixed lag time in traditional methods is avoided, and the temporal relationship between water level changes and landslide deformation can be accurately captured, thereby improving the real-time and accuracy of the prediction. By establishing a time-lag GM-BP deformation prediction model, combining the time-lag GM model and the BP neural network, the advantages of both are fully utilized, which can not only handle the time lag problem in time series data, but also use neural networks for nonlinear prediction, which can effectively improve the accuracy of landslide deformation prediction and adapt to complex deformation evolution laws. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0011] Figure 1 A schematic flow chart of a method for predicting reservoir bank landslide deformation based on water level fluctuations provided by an embodiment of the present invention; Figure 2 A topological diagram of a neural network structure provided by an embodiment of the present invention; Figure 3 A schematic diagram of the structure of a time-delayed GM-BP model provided in an embodiment of the present invention; Figure 4 A schematic structural diagram of a reservoir bank landslide deformation prediction system based on water level fluctuations provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0012] The technical solution of the present invention is described below in conjunction with the accompanying drawings.

[0013] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as an "exemplary" in the present invention should not be interpreted as being preferred or advantageous over other embodiments or designs. Rather, the use of the word "exemplary" is intended to present concepts in a concrete manner. Furthermore, in the embodiments of the present invention, "and / or" can mean both or either of the two.

[0014] In the embodiments of the present invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, when the distinction is not emphasized, the meanings they convey are the same. The terms "of," "corresponding," and "corresponding" may sometimes be used interchangeably. It should be noted that, when the distinction is not emphasized, the meanings they convey are the same.

[0015] In the embodiments of the present invention, sometimes a subscript such as W1 may be written as a non-subscript such as W1. When the difference is not emphasized, the meanings to be expressed are the same.

[0016] In order to make the technical problems, technical solutions and advantages to be solved by the present invention clearer, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0017] Reference Manual Figure 1, which shows a flow chart of a method for predicting reservoir bank landslide deformation based on water level fluctuations provided by an embodiment of the present invention.

[0018] An embodiment of the present invention provides a method for predicting reservoir bank landslide deformation based on water level fluctuations, the method comprising: S1: Acquire monitoring data, wherein the monitoring data includes reservoir water level fluctuation data, landslide deformation data and environmental data.

[0019] Among them, reservoir water level fluctuation data refers to the change data of the water level in the reservoir, which is usually recorded regularly by the monitoring system around the reservoir to reflect the rise and fall of the water level. Landslide deformation data refers to deformation information related to landslides, including the displacement and deformation rate of the landslide area, which is usually obtained through sensors or ground surveys. Environmental data includes external environmental factors such as temperature, precipitation, wind speed, and soil moisture. These factors may affect the occurrence and evolution of landslides.

[0020] It should be noted that by obtaining multi-dimensional monitoring data, the impact of water level fluctuations on landslides can be analyzed more accurately, providing a comprehensive basis for subsequent analysis and prediction.

[0021] S2: Preprocess the monitoring data, including abnormal data removal, missing data completion, and smoothing and noise reduction, to obtain sample data.

[0022] Among them, abnormal data elimination refers to the removal of abnormal values ​​or erroneous data that do not conform to the normal range in the monitoring data. Some data may be lost or unavailable during the monitoring process. Missing data completion is to estimate these missing values ​​through algorithms to ensure the integrity of the data. Smoothing and noise reduction refers to the use of certain mathematical methods, such as sliding average or filtering algorithms, to reduce the noise in the monitoring data, improve data quality, and make it more in line with actual change trends.

[0023] It should be noted that by cleaning and repairing data, eliminating invalid information, ensuring data quality, and avoiding the impact of erroneous or missing data on subsequent analysis, the accuracy and reliability of the model can be improved.

[0024] S3: Based on the sample data, calculate the Pearson coefficient between reservoir water level changes and landslide deformation.

[0025] Among them, the Pearson coefficient (Pearson correlation coefficient) is a statistic used to measure the linear correlation between two variables. Its value range is -1 to 1. A value of 1 indicates a perfect positive correlation, a value of -1 indicates a perfect negative correlation, and a value of 0 indicates no linear relationship.

[0026] It should be noted that by calculating the Pearson coefficient between reservoir water level changes and landslide deformation, it is possible to clearly identify whether there is a significant linear correlation between the two variables, providing more accurate data support for deformation prediction.

[0027] In one possible implementation, the Pearson coefficient is calculated as follows:

[0028] in, r represents the Pearson coefficient, X i Indicates the i The reservoir water level change sequence of sample data, Y i Indicates the i The landslide deformation sequence of sample data, Representation sequence X and sequence Y The covariance of the corresponding variables, σ X and σ Y Represents the sequence X and sequence Y The standard deviation of and represent the average values ​​of the reservoir water level change sequence and the landslide deformation sequence, respectively. , n Indicates the total number of sample data.

[0029] S4: Based on the Pearson coefficient, determine whether there is a correlation between the reservoir water level change and the landslide deformation; if so, proceed to step S5, otherwise return to step S1.

[0030] Specifically, | r When |≥0.6, it indicates a strong correlation between the sequences. r When | is between 0.4 and 0.6, it indicates that there is a moderate degree of correlation between the sequences. r |< 0.4 indicates that the sequences are weakly correlated, and when r A positive value indicates that the two variables are positively correlated, whereas a negative value indicates that the two variables are negatively correlated.

[0031] It should be noted that by using the Pearson coefficient to determine the correlation between reservoir water level changes and landslide deformation, data samples with significant relationships can be effectively screened out, avoiding the interference of irrelevant data. If the correlation between the two is not strong, we can return in time and conduct new data collection or adjust the strategy. If the correlation exists, we can continue to conduct in-depth analysis and lay a solid foundation for the establishment of subsequent models.

[0032] S5: Calculate the initial lag time of the impact of reservoir water level rise on landslide deformation.

[0033] Among them, the initial lag time refers to the time delay between the water level change and the landslide deformation when the reservoir water level change affects the landslide deformation. That is, the direct impact of the water level change on the landslide deformation does not occur immediately, but takes a period of lag to manifest.

[0034] It should be noted that by calculating the initial lag time, the time delay relationship between water level changes and landslide deformation can be captured more accurately, which can reveal the impact mechanism of water level changes on the landslide evolution process, avoid ignoring the lag effect, and thus improve the accuracy of landslide deformation prediction.

[0035] In a possible implementation, S5 specifically includes: When the reservoir water level change and landslide deformation are linearly correlated, the cross-correlation function is used, and the delay time when the cross-correlation function reaches its maximum value is used as the initial lag time of the impact of the reservoir water level rise on the landslide deformation evolution.

[0036] The cross-correlation function is used to calculate the similarity between two time series, reflecting the temporal relationship between them. By calculating the correlation under different delays, the cross-correlation function can help identify the impact delay of one series on another.

[0037] The cross-correlation function is specifically:

[0038] in, R ( n ) represents the cross-correlation function, X ( k )express k The landslide deformation sequence at each moment, Y ( k )express k The reservoir water level change sequence at each moment, , N Indicates the total duration, n represents the delay time, when the cross-correlation function R ( n ) takes the maximum value, the corresponding delay time n This is the initial lag time.

[0039] When the reservoir water level change and landslide deformation are nonlinearly related, dynamic time warping is used to determine the initial lag time.

[0040] Specifically, Dynamic Time Warping (DTW) is an algorithm used to calculate the similarity between two time series. It is particularly suitable for situations where time series data has nonlinear changes. The basic idea of ​​DTW is to find an optimal alignment path by minimizing the "distance" between time series, so that the two time series are as close as possible to each other at different time points.

[0041] Using dynamic time warping, determining the initial lag duration specifically includes: Calculate the DTW distance between reservoir water level change and landslide deformation to obtain the optimal alignment path:

[0042]

[0043] in, X represents the landslide deformation sequence, Y represents the reservoir water level change sequence, D DTW (X, Y) represents the DTW distance between reservoir water level change and landslide deformation, P represents the optimal alignment path, x i Indicates the first i data points, y j Indicates the first j data points, ( i , j ) represents a point in the optimal alignment path.

[0044] Determine the lag time based on the optimal alignment path:

[0045] in, τ Indicates the initial lag time, i k Indicates the optimal path k Points in the landslide deformation sequence X The position in j k Indicates the optimal path k The water level change sequence of each point in the reservoir Y The position in , N Indicates the total number of alignment points on the optimal path.

[0046] It should be noted that by combining the cross-correlation function and dynamic time warping (DTW), the temporal impact of reservoir water level changes on landslide deformation can be accurately analyzed. The cross-correlation function is applicable to linear relationships and can quickly determine the lag period, while DTW is applicable to nonlinear relationships and can more accurately handle complex dynamic changes. Using these two methods can ensure the accurate calculation of the initial lag period, regardless of whether there is a linear relationship between water level changes and landslide deformation, thereby improving the model's early warning capability and adaptability, and ensuring the reliability of landslide deformation prediction.

[0047] S6: According to the Pearson coefficient, the initial lag time is adjusted to obtain the dynamic lag time.

[0048] The dynamic lag time is the lag time obtained by dynamically adjusting the initial lag time according to the change of the Pearson coefficient. By analyzing multiple time windows, the lag time in the model is adjusted to ensure a more accurate prediction of landslide deformation.

[0049] It should be noted that by dynamically adjusting the lag period using the Pearson coefficient, the complex time-lag effect of water level changes on landslide deformation can be effectively captured. The dynamic adjustment process avoids the limitation of static lag period that cannot adapt to actual changes, enhances the model's adaptability to nonlinear and time-varying data, and improves the accuracy and practicality of deformation prediction.

[0050] In a possible implementation, S6 specifically includes: S601: Setting a sliding window.

[0051] S602: Calculate the Pearson coefficient of each sliding window.

[0052] S603: According to each Pearson coefficient, the initial lag time is adjusted to obtain the dynamic lag time:

[0053]

[0054] Among them, Δ represents the increment operator, r w Indicates the w The Pearson coefficient of the sliding window, r w-1 Indicates the w -1 sliding window Pearson coefficient, Indicates the dynamic lag time. τ Indicates the initial lag time, δ Indicates the adjustment step size.

[0055] Reference Manual Figure 2 , which shows a topological diagram of a neural network structure provided by an embodiment of the present invention.

[0056] like Figure 2 , shows a typical feedforward neural network structure, which includes input layer, hidden layer and output layer. The input layer contains multiple neurons ( X 1, X 2,..., X n ), used to receive external input signals, the hidden layer contains multiple neurons, which are responsible for nonlinear processing of the input signals, and each neuron is connected by weight ( W 1, W 2,..., W n ) connects the input layer and the hidden layer, and the information is transmitted and activated by the function ( f ) is processed and passed to the next layer. Finally, the output layer ( Y 1, Y 2,..., Ym ) Output the processed results. The "forward transmission of information" and "backward propagation of error signals" in the figure represent the training process of the neural network, that is, the output is calculated through forward propagation, and then the weights are adjusted through back propagation to optimize the model.

[0057] Reference Manual Figure 3 , which shows a schematic structural diagram of the time-lag GM-BP model provided by an embodiment of the present invention.

[0058] S7: Based on the dynamic lag time, a time-lag GM-BP deformation prediction model is established.

[0059] It should be noted that the dynamic incorporation of changes in the dynamic lag time into the prediction model enables the model to adapt to the impact of water level changes on landslides. By combining the GM and BP models, linear and nonlinear factors can be effectively handled, thereby improving the accuracy and adaptability of the prediction results.

[0060] In a possible implementation, the time-delay GM-BP model includes a time-delay GM model and a BP model that are combined in series.

[0061] Among them, the Grey Model with Time Delay (GM) is a modeling method in grey system theory, mainly used to predict systems with uncertainty and small amounts of data. In the GM model with time delay, time delay refers to the time delay relationship between input (such as water level change) and output (such as landslide deformation). This model is suitable for processing dynamic systems with time lag. The BP model is a typical feedforward neural network, which is trained using the error back propagation algorithm. In the GM-BP model with time delay, the BP network is used to capture the nonlinear characteristics of the system and optimize it based on historical data to predict future deformation trends.

[0062] It should be noted that by combining the time-lag GM and BP models in series, the prediction ability of the grey system and the nonlinear modeling ability of the neural network can be comprehensively utilized to deal with complex landslide deformation problems. The time-lag GM model effectively handles linear and hysteresis characteristics, and the BP model can capture the nonlinear relationship in the data. Through this combination, not only the accuracy of the model is improved, but also the advantages of each model can be fully utilized, so as to better adapt to the dynamic changes of landslide deformation and improve the accuracy and reliability of prediction.

[0063] S8: Input the sample data into the time-lag GM-BP deformation prediction model for training.

[0064] It should be noted that by training the time-delay GM-BP model, it can learn and optimize parameters based on actual monitoring data, thereby improving the accuracy of landslide deformation prediction. By inputting sample data, the model can gradually identify the complex relationship between reservoir water level changes, landslide deformation and environmental factors, and then accurately predict future deformation trends. At the same time, the combination of time-delay GM and BP network can effectively handle time delays and nonlinear characteristics, further enhancing the robustness and adaptability of the model.

[0065] In a possible implementation, S8 specifically includes: S801: Input the sample data into the GM model, perform preliminary prediction, and obtain a preliminary predicted time series signal.

[0066] S802: Input the preliminary predicted time series signal into the BP neural network for training.

[0067] In a possible implementation, S801 specifically includes: S8011: Determine the time series of the time-lag related factors based on the sample data:

[0068] in, Indicates the i A time series of time-lagged correlation factors, Indicates the i The dynamic lag length of the time-lag correlation factor, n Indicates the total length of the time series.

[0069] S8012: Superimpose and discretize the time series to obtain the whitened differential equation:

[0070] in, represents the system development coefficient, express, b i Indicates the i The driving coefficient of the time-lag related factor, Indicates the driver item, Represents a time series generated by accumulation. , N represents the total number of time-lag related factors, t Indicates time.

[0071] S8013: Based on the whitened differential equation, the coefficient vector is calculated by the least squares method to obtain the approximate time response of the differential equation:

[0072] in, represents an approximate solution to the whitened differential equation.

[0073] Among them, the whitened differential equation refers to a differential equation that becomes independent by processing the correlation of the original data. It is often used to eliminate redundant information in the data and make the modeling process more efficient. The least squares method is a mathematical optimization method used to fit the model by minimizing the sum of squares of errors and is widely used in data fitting.

[0074] It should be noted that by optimizing the coefficients using the least squares method, the whitened differential equation can be accurately fitted, so that the time response of the system can better match the actual data, thereby improving the accuracy and stability of the prediction.

[0075] S8014: Based on the approximate time response formula of the differential equation, obtain the preliminary predicted time series signal:

[0076] in, Represents a preliminary forecast time series signal.

[0077] In a possible implementation, S802 specifically includes: S8021: Initialize the BP neural network and determine the weight value, hidden layer threshold, output layer threshold, learning rate and activation function of the BP neural network.

[0078] The specific activation function is:

[0079] in, f ( x ) represents the activation function, e Indicates natural Changshu, c Represents an arbitrary constant.

[0080] S8022: Input the preliminary predicted time series signal into the hidden layer of the BP neural network to obtain the hidden layer output:

[0081] in, H j Indicates the j The hidden layer output of neurons, f represents the hidden layer activation function, w ij Represents the input layer i neurons and the hidden layer j The connection weights between neurons, x i Represents the input layer i The input value of a neuron, a j represents the hidden layer j The threshold of a neuron, l Indicates the number of hidden layer nodes.

[0082] S8023: Pass the hidden layer output to the output layer of the BP neural network to obtain the predicted value:

[0083] in, O k Represents the output layer k The predicted value of a neuron, w jk represents the hidden layer j neurons in the output layer k The connection weights between neurons, b k Represents the output layer k The threshold of a neuron, m Represents the total number of neurons in the output layer.

[0084] S8024: Calculate the prediction error value based on the prediction result and the actual value:

[0085] in,e k Represents the output layer k The prediction error value of each neuron is Y k Represents the output layer k The actual value of the neuron, O k Represents the output layer k The predicted value of a neuron, , m Represents the total number of neurons in the output layer.

[0086] S8025: Adjust the weight value, hidden layer threshold and output layer threshold according to the network error value.

[0087] Among them, the network error value refers to the difference between the predicted output of the neural network and the actual result. The error value reflects the accuracy of the model prediction and is often obtained by calculating the difference between the predicted result of the output layer and the actual result. The weight value is the parameter connecting each neuron and controls the transmission intensity of information. The hidden layer threshold (also known as bias) is the offset of the input signal of each neuron, which is usually used to adjust the activation function of the neuron and affect the output of the neuron. The output layer threshold is the bias term of the last layer of the neural network and determines how the final output is calculated.

[0088] It should be noted that by adjusting the model parameters (such as weights, hidden layer thresholds, and output layer thresholds) according to the network error value, the model learning process can be effectively optimized, the ability to fit complex data can be improved, the prediction error can be reduced, and the accuracy and robustness of the model can be enhanced.

[0089] S8026: Repeat steps S8022-S8025 until the maximum number of iterations is reached, and output the trained time-delay GM-BP model.

[0090] In a possible implementation, the formula for adjusting the weight value, hidden layer threshold, and output layer threshold is specifically:

[0091]

[0092]

[0093] in, η t Indicates time t The adaptive learning rate, η 0 represents the initial learning rate, v t express t The exponentially weighted average of the squared gradients at time instants, vt-1 express t The exponentially weighted average of the squared gradient at time -1, ε represents the smoothing parameter, β represents the decay rate, g t Indicates time t The parameter gradient of .

[0094] S9: Get real-time sample data.

[0095] S10: Input the real-time sample data into the trained time-delay GM-BP deformation prediction model and output the final prediction result.

[0096] It should be noted that by inputting real-time sample data into the trained time-lagged GM-BP deformation prediction model, the prediction results of landslide deformation can be obtained immediately, and dynamic monitoring and risk assessment can be carried out. This not only improves the timeliness and accuracy of the prediction, but also provides timely early warning information for practical applications, thereby reducing the disaster risk caused by landslides.

[0097] S11: Issue early warning based on the final prediction results.

[0098] It should be noted that by issuing early warnings based on the final prediction results, potential landslide risks can be identified in a timely manner, preventive measures can be taken in advance, the losses caused by landslide disasters can be reduced, and the efficiency and accuracy of disaster response can be improved.

[0099] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least: In an embodiment of the present invention, by calculating the Pearson coefficient between reservoir water level changes and landslide deformation, the correlation between the two can be accurately determined, avoiding the limitations of relying solely on assumptions. By dynamically adjusting the lag time, the limitation of fixed lag time in traditional methods is avoided, and the temporal relationship between water level changes and landslide deformation can be accurately captured, thereby improving the real-time and accuracy of the prediction. By establishing a time-lag GM-BP deformation prediction model, combining the time-lag GM model and the BP neural network, the advantages of both are fully utilized, which can not only handle the time lag problem in time series data, but also use neural networks for nonlinear prediction, which can effectively improve the accuracy of landslide deformation prediction and adapt to complex deformation evolution laws.

[0100] Reference Manual Figure 4 , showing a structural schematic diagram of a reservoir bank landslide deformation prediction system based on water level fluctuation provided by the present invention.

[0101] The present invention further provides a reservoir bank landslide deformation prediction system 20 based on water level fluctuation, which is applied to the above-mentioned reservoir bank landslide deformation prediction method based on water level fluctuation, and comprises: Processor 201.

[0102] The memory 202 stores computer-readable instructions. When the computer-readable instructions are executed by the processor 201 , the method for predicting reservoir bank landslide deformation based on water level fluctuation as described in the method embodiment is implemented.

[0103] The reservoir bank landslide deformation prediction system 20 based on water level fluctuation provided by the present invention can execute the above-mentioned reservoir bank landslide deformation prediction method based on water level fluctuation and achieve the same or similar technical effects. To avoid repetition, the present invention will not elaborate on them.

[0104] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least: In an embodiment of the present invention, by calculating the Pearson coefficient between reservoir water level changes and landslide deformation, the correlation between the two can be accurately determined, avoiding the limitations of relying solely on assumptions. By dynamically adjusting the lag time, the limitation of fixed lag time in traditional methods is avoided, and the temporal relationship between water level changes and landslide deformation can be accurately captured, thereby improving the real-time and accuracy of the prediction. By establishing a time-lag GM-BP deformation prediction model, combining the time-lag GM model and the BP neural network, the advantages of both are fully utilized, which can not only handle the time lag problem in time series data, but also use neural networks for nonlinear prediction, which can effectively improve the accuracy of landslide deformation prediction and adapt to complex deformation evolution laws.

[0105] It should be understood that the processor in the embodiments of the present invention may be a central processing unit (CPU), but may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc.

[0106] It should also be understood that the memory in the embodiments of the present invention may be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory may be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory may be random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic random access memory (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0107] The above embodiments can be implemented in whole or in part via software, hardware (e.g., circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. A computer program product comprises one or more computer instructions or computer programs. When the computer instructions or computer program are loaded or executed on a computer, the processes or functions according to the embodiments of the present invention are fully or partially generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. Computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired means (e.g., infrared, wireless, microwave, etc.). A computer-readable storage medium can be any available medium accessible by a computer or a data storage device such as a server or data center that contains a collection of one or more available media. Available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media. Semiconductor media can be solid-state drives.

[0108] It should be understood that the term "and / or" as used herein simply describes a relationship between associated objects, indicating that three possible relationships exist. For example, "A and / or B" can represent: A alone, A and B together, or B alone. A and B can be singular or plural. Furthermore, the character " / " as used herein generally indicates an "or" relationship between the associated objects, but it may also indicate an "and / or" relationship. For specific understanding, please refer to the context.

[0109] In this disclosure, "at least one" means one or more, and "plurality" means two or more. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, "at least one of a, b, or c" can mean: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or plural.

[0110] It should be understood that in various embodiments of the present invention, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0111] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present invention.

[0112] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described equipment, devices and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0113] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, and can be electrical, mechanical, or other forms.

[0114] Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0115] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0116] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or the portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the various embodiments of the method of the present invention. The aforementioned storage medium includes various media that can store program code, such as USB flash drives, mobile hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0117] An embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon. When the program is executed by a processor, the method for predicting reservoir bank landslide deformation based on water level fluctuation as described in the method embodiment is implemented.

[0118] The computer-readable storage medium provided by the present invention can realize the steps and effects of the reservoir bank landslide deformation prediction method based on water level fluctuation of the above method embodiment. To avoid repetition, the present invention will not elaborate on them.

[0119] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least: In an embodiment of the present invention, by calculating the Pearson coefficient between reservoir water level changes and landslide deformation, the correlation between the two can be accurately determined, avoiding the limitations of relying solely on assumptions. By dynamically adjusting the lag time, the limitation of fixed lag time in traditional methods is avoided, and the temporal relationship between water level changes and landslide deformation can be accurately captured, thereby improving the real-time and accuracy of the prediction. By establishing a time-lag GM-BP deformation prediction model, combining the time-lag GM model and the BP neural network, the advantages of both are fully utilized, which can not only handle the time lag problem in time series data, but also use neural networks for nonlinear prediction, which can effectively improve the accuracy of landslide deformation prediction and adapt to complex deformation evolution laws.

[0120] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

[0121] There are a few points to note: (1) The drawings of the embodiments of the present invention only relate to the structures related to the embodiments of the present invention. Other structures may refer to conventional designs.

[0122] (2) For the sake of clarity, the thickness of layers or regions in the drawings used to describe the embodiments of the present invention are exaggerated or reduced, that is, these drawings are not drawn to scale. It is understood that when an element such as a layer, film, region, or substrate is referred to as being "on" or "under" another element, the element may be "directly on" or "under" the other element or intervening elements may be present.

[0123] (3) In the absence of conflict, the embodiments of the present invention and the features therein may be combined with each other to form new embodiments.

[0124] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. The protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A method for predicting reservoir bank landslide deformation based on water level fluctuation, characterized in that: include: S1: Acquiring monitoring data, wherein the monitoring data includes reservoir water level fluctuation data, landslide deformation data, and environmental data; S2: performing preprocessing on the monitoring data, including abnormal data elimination, missing data filling, and smoothing and noise reduction, to obtain sample data; S3: Calculating the Pearson coefficient between reservoir water level change and landslide deformation based on the sample data; S4: judging whether there is a correlation between the reservoir water level change and the landslide deformation based on the Pearson coefficient; if so, proceeding to step S5, otherwise returning to step S1; S5: Calculating the initial lag time of the impact of the reservoir water level rise on the landslide deformation; S6: adjusting the initial lag time according to the Pearson coefficient to obtain a dynamic lag time; S7: establishing a time-lag GM-BP deformation prediction model based on the dynamic hysteresis duration; S8: inputting the sample data into the time-lag GM-BP deformation prediction model for training; S9: Get real-time sample data; S10: inputting the real-time sample data into the trained time-delay GM-BP deformation prediction model and outputting the final prediction result; S11: issuing an early warning based on the final prediction result.

2. The method for predicting reservoir bank landslide deformation based on water level fluctuation according to claim 1, characterized in that: The Pearson coefficient is calculated as follows: in, r represents the Pearson coefficient, X i Indicates the i The reservoir water level change sequence of sample data, Y i Indicates the i The landslide deformation sequence of sample data, Representation sequence X and sequence Y The covariance of the corresponding variables, σ X and σ Y Represents the sequence X and sequence Y The standard deviation of and represent the average values ​​of the reservoir water level change sequence and the landslide deformation sequence, respectively. , n Indicates the total number of sample data.

3. The method for predicting reservoir bank landslide deformation based on water level fluctuation according to claim 1, characterized in that: The S5 is specifically: When the reservoir water level change and the landslide deformation are linearly correlated, a cross-correlation function is used, and the delay time when the cross-correlation function reaches a maximum value is used as the initial lag time of the impact of the reservoir water level rise on the landslide deformation evolution; When the reservoir water level change and the landslide deformation are nonlinearly correlated, dynamic time warping is used to determine the initial lag time.

4. The method for predicting reservoir bank landslide deformation based on water level fluctuation according to claim 1, characterized in that: The S6 specifically includes: S601: Setting a sliding window; S602: Calculate the Pearson coefficient of each sliding window; S603: Adjust the initial lag time according to each Pearson coefficient to obtain a dynamic lag time.

5. The method for predicting reservoir bank landslide deformation based on water level fluctuation according to claim 1, characterized in that: The time-delay GM-BP model includes a time-delay GM model and a BP model combined in series.

6. The method for predicting reservoir bank landslide deformation based on water level fluctuation according to claim 1, characterized in that: The S8 specifically includes: S801: Input the sample data into the GM model to perform preliminary prediction to obtain a preliminary predicted time series signal; S802: Input the preliminary predicted time series signal into the BP neural network for training.

7. The method for predicting reservoir bank landslide deformation based on water level fluctuation according to claim 6, characterized in that: The S801 specifically includes: S8011: Determine an initial time series of a time lag correlation factor based on the sample data; S8012: Superimposing and discretizing the time series to obtain a whitened differential equation; S8013: Based on the whitened differential equation, the coefficient vector is calculated using the least squares method to obtain the approximate time response of the differential equation; S8014: Obtain a preliminary predicted time series signal based on the approximate time response formula of the differential equation.

8. The method for predicting reservoir bank landslide deformation based on water level fluctuation according to claim 6, characterized in that: The S802 specifically includes: S8021: Initialize the BP neural network and determine the weight value, hidden layer threshold, output layer threshold, learning rate, and activation function of the BP neural network; S8022: Inputting the preliminary predicted time series signal into the hidden layer of the BP neural network to obtain a hidden layer output; S8023: passing the hidden layer output to the output layer of the BP neural network to obtain a predicted value; S8024: Calculating a prediction error value based on the prediction result and the actual value; S8025: Adjust the weight value, the hidden layer threshold, and the output layer threshold according to the network error value; S8026: Repeat steps S8022-S8025 until the maximum number of iterations is reached, and output the trained time-delay GM-BP model.

9. The method for predicting reservoir bank landslide deformation based on water level fluctuation according to claim 8, characterized in that: The formula for adjusting the weight value, the hidden layer threshold and the output layer threshold is specifically: in, η t Indicates time t The adaptive learning rate, η 0 represents the initial learning rate, v t express t The exponentially weighted average of the squared gradients at time instants, v t-1 express t The exponentially weighted average of the squared gradient at time -1, ε represents the smoothing parameter, β represents the decay rate, g t Indicates time t The parameter gradient of .

10. A reservoir bank landslide deformation prediction system based on water level fluctuation, characterized by: include: processor; A memory having computer-readable instructions stored thereon, wherein when the computer-readable instructions are executed by the processor, the method for predicting reservoir bank landslide deformation based on water level fluctuation according to any one of claims 1 to 9 is implemented.

Images