Dam safety monitoring data effective information extraction method and system based on adaptive time sequence decomposition

Abstract

The invention discloses a dam safety monitoring data effective information extraction method and system based on adaptive time sequence decomposition. The method comprises the following steps: firstly, obtaining a dam single-measuring-point monitoring time sequence; adopting a multi-population Jaya algorithm to adaptively optimize control parameters of variational mode decomposition (VMD); performing non-recursive decomposition on a data sequence by using the optimized VMD, and screening high-frequency and medium-and-low-frequency sub-modals through a correlation coefficient-energy entropy composite index; superposing the high-frequency sub-modes to form a high-frequency sequence, constructing a composite threshold in combination with kernel density estimation and a box graph method, and eliminating local singular values in the high-frequency sequence to generate a residual sequence; and finally, superposing and reconstructing the low and medium frequency sub-modes and the residual high-frequency sequence to realize effective information extraction of the monitoring data. According to the method, the feature extraction precision of dam monitoring data is remarkably improved through parameter adaptive optimization and composite criteria.

Description

Technical Field

[0001] The present invention relates to the field of water conservancy project safety monitoring and analysis, and in particular to a method for extracting effective information from dam safety monitoring data based on adaptive time series decomposition. Background Art

[0002] Dams are important water conservancy projects, providing benefits such as flood control, power generation, irrigation, and navigation. Dam safety monitoring and analysis provide timely and accurate insights into dam safety status and the timely detection of potential anomalies. However, due to a variety of accidental and non-accidental uncertainties (such as unreliable monitoring instruments, monitoring system failures, and human interference), dam safety monitoring data can be affected by noise and gross errors, which can interfere with subsequent dam safety monitoring, analysis, and early warning processes, leading to misjudgments. Therefore, effective information extraction from dam safety monitoring data is a crucial prerequisite for conducting dam safety monitoring, analysis, and early warning.

[0003] Dam safety monitoring data exhibits multi-scale evolution, strong environmental coupling, and non-stationary noise. Existing techniques for extracting effective information from dam safety monitoring data primarily focus on three approaches: statistical analysis, regression modeling, and signal decomposition. Traditional statistical analysis methods typically construct statistical indicators based on data distribution assumptions or data quantiles, making it difficult to simultaneously extract and filter local anomalies and noise characteristics from dam safety monitoring data. Regression modeling methods rely on a mapping relationship between environmental variables and dam safety monitoring effect quantities. When some environmental monitoring data is missing or the monitoring sequence is short, model performance degrades significantly, resulting in poor extraction of effective information from dam safety monitoring data.

[0004] In recent years, researchers have used signal decomposition methods (such as wavelet transforms, short-time Fourier transforms, and empirical mode decomposition algorithms) and time series feature decomposition and mining to reduce data noise and identify gross errors, thereby extracting effective information from the data. However, in engineering practice, these methods generally face challenges such as uncontrollable decomposition modal redundancy, difficulty in adaptive parameter selection, and inappropriate threshold setting. This results in significant residual noise or incomplete outlier removal, making them difficult to adapt to the requirements for extracting effective information from dam safety monitoring data. Summary of the Invention

[0005] Technical problem to be solved by the present invention: In view of the problems and shortcomings of the existing technology, the present invention provides a method for extracting effective information from dam safety monitoring data based on adaptive time series decomposition, which aims to use a variational mode decomposition algorithm with optimized parameters to perform adaptive non-recursive decomposition on the time series of dam single-measurement point safety monitoring data, and adopt a composite discriminant index to automatically identify and extract noise and gross error features in high-frequency sub-modes, thereby simultaneously achieving gross error elimination and noise reduction of the dam single-measurement point safety monitoring data, and improving the quality of effective information extraction from dam safety monitoring data.

[0006] In order to solve the above technical problems, the present invention adopts the following technical solutions:

[0007] The present invention first provides a method for extracting effective information from dam safety monitoring data based on adaptive time series decomposition, comprising the following steps:

[0008] Step S1, obtaining a time series of safety monitoring data of a single measuring point of the dam, wherein the safety monitoring data includes dam deformation, seepage pressure, seepage flow, and stress and strain data;

[0009] Step S2: Based on the time series of safety monitoring data of a single measuring point of the dam, a multi-population Jaya algorithm is used to adaptively optimize the variational mode decomposition control parameters;

[0010] Step S3, using a variational mode decomposition algorithm with optimized parameters to perform non-recursive decomposition on the time series of the dam single-point safety monitoring data;

[0011] Step S4: performing feature analysis on the sub-modal sequence generated by the decomposition, and adaptively screening high-frequency sub-modal and medium-low frequency sub-modal by using the composite discrimination index of correlation coefficient and energy entropy ratio;

[0012] Step S5: Superimpose the high-frequency sub-modes to obtain a high-frequency sequence, use the kernel density estimation method and the box plot method to formulate a composite threshold index for screening the high-frequency sequence, remove local singular values in the high-frequency sequence based on the composite threshold index, and generate a residual high-frequency sequence;

[0013] Step S6: Superimpose the medium and low frequency sub-modes and the residual high frequency sequence to obtain a reconstructed time series of the dam single measuring point safety monitoring data, thereby completing the extraction of effective information from the dam safety monitoring data.

[0014] To further improve the above technical solution, step S2 includes the following steps:

[0015] Step S21, setting the control parameter search space of variational mode decomposition, the control parameters include the decomposition mode number , penalty parameter and convergence tolerance , build control parameter set ;

[0016] Step S22: Setting the basic parameters of the multi-population Jaya algorithm, including the number of subpopulations and the maximum number of iterations ;

[0017] Step S23: Use variational modal decomposition to calculate the time series sub-modes of the dam single-point safety monitoring data, and arrange the decomposed sub-modes in order from low frequency to high frequency. The submode is denoted as , ;

[0018] extract signal sequence , calculate the Submodal envelope entropy function :

[0019]

[0020] Where, For the The envelope entropy of the layer sub-mode ( ); is the number of samples of the sub-modality.

[0021] Step S24: Minimizing the sum of the first two order envelope entropies is the goal, and constructing a second order envelope entropy loss function:

[0022] ,

[0023] Step S25: Based on the goal of minimizing the envelope entropy loss function in step S24, track the optimal solution in each candidate subpopulation, and use the multi-population Jaya algorithm to update the control parameters based on the optimal solution and the worst solution in all subpopulations, as follows:

[0024] ,

[0025] Where, For the During the iteration The first candidate subpopulation The value of a variable, For the In the iteration, all subpopulations The optimal value of a variable, For the In the iteration, all subpopulations The worst value of a variable; and There are two between A random number in the interval.

[0026] Step S26: Repeat steps S23-S25 until the number of iterations reaches , the optimal control parameter combination is obtained by taking the variational mode decomposition parameter corresponding to the minimum value of the second-order envelope entropy loss function when the iteration stops .

[0027] Furthermore, the step S4 includes the following steps:

[0028] Step S41, calculate the layer( ) sub-mode and the Pearson correlation coefficient of the dam single-point safety monitoring data time series before preprocessing :

[0029] ,

[0030] Where, is the time series of the dam single-point safety monitoring data before preprocessing, For the Layer sub-mode, and They are and The mean of is the number of samples of the sub-modality.

[0031] Step S42, calculate the layer( ) Energy entropy of the submode :

[0032] ,

[0033] Where, For the Layer sub-mode, is the number of samples of the sub-modality, is the decomposition mode number.

[0034] Step S43: repeat steps S41 and S42 to calculate the Pearson correlation coefficients of all submodes and the time series of the dam single-point safety monitoring data before preprocessing, as well as the energy entropy of all submodes;

[0035] Step S44: Calculate the energy entropy ratio of all sub-modes :

[0036] ,

[0037] Where, For the Layer sub-mode The energy entropy, is the decomposition mode number.

[0038] Step S45: Setting the correlation coefficient threshold Set the energy entropy ratio threshold to 0.2~0.3 is 0.1; comparison and The size relationship, comparison and The size relationship, while satisfying and No. If the layer submode is determined to be high frequency, it is a high frequency submode; otherwise, it is determined to be a medium or low frequency submode.

[0039] Furthermore, the step S5 includes the following steps:

[0040] Step S51: Superimpose the high-frequency sub-modes to obtain a high-frequency sequence :

[0041] ,

[0042] Where, is the number of high-frequency submodes, is the decomposition mode number.

[0043] Step S52: Calculate the high frequency sequence using kernel density estimation method The probability density function of , and get the probability density function The maximum value ; Through the probability density function Query to obtain probability density The corresponding minimum threshold of kernel density estimation and the maximum threshold of kernel density estimation , is a control coefficient between 0.01 and 0.1;

[0044] Step S53: Calculate high frequency sequence Upper quartile of , lower quartile and interquartile range , based on which the minimum threshold of the box plot is calculated and box plot maximum threshold :

[0045] ,

[0046] ,

[0047] Step S54: Filter high-frequency sequences based on the threshold index obtained in steps S42 and S43 The singular values in the high-frequency sequence Any value in ,when satisfy or or or , it is judged as a singular value and is removed.

[0048] Accordingly, a platform for extracting effective information from dam safety monitoring data based on adaptive time series decomposition is also provided, which is used to execute a method for extracting effective information from dam safety monitoring data based on adaptive time series decomposition of the present invention. The platform includes a reading module, a decomposition module, a screening module, a filtering module, and a reconstruction module.

[0049] The reading module is used to read and input the safety monitoring data of a single measuring point of the dam;

[0050] The decomposition module is used to perform an adaptive variational modal decomposition algorithm on the safety monitoring data of a single measuring point of the dam to obtain sub-modes of different frequency bands;

[0051] The screening module is used to perform adaptive screening on the sub-modes to obtain high-frequency sub-modes and medium- and low-frequency sub-modes;

[0052] The filtering module is used to superimpose the high-frequency sub-modes to obtain a high-frequency sequence, calculate the threshold index of the high-frequency sequence and remove the local singular values in the high-frequency sequence to obtain a residual high-frequency sequence;

[0053] The reconstruction module is used to superimpose the medium and low frequency sub-modes and the residual high frequency sequence to obtain the effective information extraction result of the safety monitoring data of a single measuring point of the dam.

[0054] In addition, the present invention provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the computer program is executed, the steps of the method of the present invention are implemented.

[0055] Finally, the present invention provides a computer-readable storage medium having a computer program stored thereon, which implements the steps of the method of the present invention when executed by a processor.

[0056] Compared with the existing technology, the method and platform for extracting effective information from dam safety monitoring data based on adaptive time series decomposition provided by the present invention have the following technical effects:

[0057] 1. The present invention realizes gross error identification and data noise reduction of single-point dam safety monitoring data by constructing a complete method system and platform; the present invention proposes adaptive decomposition of single-point dam safety monitoring data through variational modal decomposition, and adaptive optimization of variational modal decomposition parameters through a multi-population Jaya optimization algorithm and a second-order envelope entropy loss function, which overcomes the limitations of manual parameter selection, effectively avoids modal aliasing, and improves the robustness of monitoring data decomposition.

[0058] 2. The present invention proposes a method to adaptively distinguish the decomposed high-frequency sub-modes from the low-frequency sub-modes by using the correlation coefficient and energy entropy ratio review and discrimination indicators, thereby improving the recognition efficiency of the high-frequency sub-modes.

[0059] 3. The present invention proposes to use the kernel density estimation method and the box plot method to formulate a review and discrimination index to identify jump values, thereby improving the recognition accuracy of local jump values and gross errors in the dam safety monitoring data of a single measuring point.

[0060] 4. The present invention improves the gross error identification accuracy and noise reduction performance of dam safety monitoring data through the "data decomposition-modal identification-high-frequency filtering-data reconstruction" analysis concept, and has important engineering application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 This is a flow chart of the effective information extraction method for dam safety monitoring data based on adaptive time series decomposition.

[0062] Figure 2 This is the original dam crest deformation measurement point process line in Example 1.

[0063] Figure 3 This is the process line of the dam crest deformation measuring point obtained after effective information extraction in Example 1.

[0064] Figure 4 These are the sub-modal process lines of different frequencies obtained after time series decomposition in Example 1.

[0065] Figure 5 This is a schematic diagram of the structure of the effective information extraction platform for dam safety monitoring data based on adaptive time series decomposition in Example 2. DETAILED DESCRIPTION

[0066] The present invention is further illustrated below with reference to specific examples. It should be understood that these examples are only used to illustrate the present invention and are not used to limit the scope of the present invention. After reading the present invention, modifications of various equivalent forms of the present invention made by those skilled in the art all fall within the scope defined by the claims attached to this application.

[0067] Example 1: Dam safety monitoring data can include any of the following: horizontal deformation, vertical deformation, cracks in dam joints, seepage pressure, seepage volume, and stress and strain. For example, in this example, horizontal deformation data from a measuring point on the crest of a large reservoir dam is selected. The data is monitored once daily, with a total of 1,158 sets of values, and the monitoring period is from 2020 to 2023. It should be understood that these horizontal deformation monitoring values are multiple dam deformation monitoring values obtained by monitoring instruments over a period of time.

[0068] Figure 1A flowchart of a method for extracting effective information from dam safety monitoring data based on adaptive time series decomposition is provided in Example 1 of the present invention. The method extracts effective information from dam monitoring data through the following steps, including steps S1 to S6:

[0069] Step S1: Get the time series of horizontal displacement monitoring data of a single measuring point on the dam crest. The original deformation process line of the dam crest is as follows: Figure 2 As shown;

[0070] Step S2: Based on the time series of deformation monitoring data of a single measuring point on the dam crest, a multi-swarm Jaya algorithm is used to adaptively optimize the variational modal decomposition control parameters, including steps S21 to S26:

[0071] Step S21, setting the control parameter search space of variational mode decomposition, the control parameters include the decomposition mode number , penalty parameter and convergence tolerance , decomposition modal number , penalty parameter and convergence tolerance The parameter optimization ranges are 、 and ;

[0072] Step S22: Setting the basic parameters of the multi-population Jaya algorithm, including the number of subpopulations and the maximum number of iterations , number of subpopulations and the maximum number of iterations Set to 1000 and 20 respectively;

[0073] Step S23: Use variational modal decomposition to calculate the time series sub-modes of the dam single-point safety monitoring data, and arrange the decomposed sub-modes in order from low frequency to high frequency. The submode is denoted as , , each sub-modal process line is as follows Figure 4 As shown;

[0074] extract signal sequence , calculate the Submodal envelope entropy function :

[0075]

[0076] Where, For the Envelope entropy of the layer submode; is the number of samples of the sub-modality.

[0077] Step S24: Minimizing the sum of the first two order envelope entropies is the goal, and constructing a second order envelope entropy loss function, specifically:

[0078] ,

[0079] Step S25: Based on the goal of minimizing the envelope entropy loss function in step S24, track the optimal solution in each candidate subpopulation, and use the multi-population Jaya algorithm to update the control parameters based on the optimal solution and the worst solution in all subpopulations. Specifically,

[0080] ,

[0081] Where, For the During the iteration The first candidate subpopulation The value of a variable, For the In the iteration, all subpopulations The optimal value of a variable, For the In the iteration, all subpopulations The worst value of a variable; and There are two between A random number in the interval.

[0082] Step S26: Repeat steps S23-S25 until the number of iterations reaches , the optimal control parameter combination is obtained by taking the variational mode decomposition parameter corresponding to the minimum value of the second-order envelope entropy loss function when the iteration stops .

[0083] Step S3: Using a parameter-optimized variational mode decomposition algorithm to perform non-recursive decomposition on the time series of the dam single-point safety monitoring data.

[0084] In steps S2 and S3, the variational mode decomposition algorithm used in this embodiment is used to decompose the nonlinear and non-stationary time series signal into a number of intrinsic mode functions, each of which has different center frequencies. The variational mode decomposition algorithm is non-recursive and flexible, and can directly extract the mode functions from the signal data without resetting the basis functions. The decomposition process of variational mode decomposition is a process for solving variational problems, and its implementation steps are as follows:

[0085] (1) The variational mode decomposition algorithm calculates the unilateral spectrum of each modal component through Hilbert transform, adds the correction exponential term to adjust the center frequency and modulates the single spectrum to the corresponding "baseband". Gaussian smoothness is used to estimate the bandwidth of the demodulated signal, and the dam safety monitoring data series is decomposed and transformed into a constrained variational problem:

[0086] ,

[0087] Where, represents the modal function; is the frequency of each mode center; is the modal number; Represents the original deformation monitoring sequence.

[0088] (2) Introducing a quadratic penalty factor and Lagrange multipliers , so that the constrained variational formula is converted to an unconstrained variational formula as follows:

[0089] ,

[0090] Where, is the Lagrange multiplication operator; is the quadratic penalty factor.

[0091] (3) By alternating the direction of the multiplication operator, alternately update and solve 、 and , and get the optimal solution.

[0092] (4) Repeat step (3) until the iteration accuracy is met ,get Submodes. Iteration termination condition:

[0093] ,

[0094] Step S4: Perform feature analysis on the five groups of sub-modal sequences generated by decomposition, and adaptively screen high-frequency sub-modal and medium-low frequency sub-modal by using the composite discrimination index of correlation coefficient and energy entropy ratio, including steps S41 to S45:

[0095] Step S41, calculate the Pearson correlation coefficient between the layer submode and the time series of the dam single-point safety monitoring data before preprocessing :

[0096] ,

[0097] Where, is the time series of the dam single-point safety monitoring data before preprocessing, For the Layer sub-mode, and They are and The mean of is the number of samples of the sub-modality.

[0098] Step S42, calculate the Energy entropy of stratum modes :

[0099] ,

[0100] Where, For the Layer sub-mode.

[0101] Step S43: repeat steps S41 and S42 to calculate the Pearson correlation coefficients of all submodes and the time series of the dam single-point safety monitoring data before preprocessing, as well as the energy entropy of all submodes;

[0102] Step S44: Calculate the energy entropy ratio of all sub-modes :

[0103] ,

[0104] Where, For the Energy entropy of stratum modes , is the decomposition mode number.

[0105] Step S45: Setting the correlation coefficient threshold Set the energy entropy ratio threshold to 0.2~0.3 is 0.1; comparison and The size relationship, comparison and The size relationship, while satisfying and No. If the layer sub-mode is determined to be high frequency, it is a high frequency sub-mode; otherwise, it is determined to be a medium and low frequency sub-mode. According to the calculation, the first two groups of modes ( and ) are the medium and low frequency submodes, and the remaining three modes ( 、 and ) is the high frequency submode.

[0106] Step S5: high frequency sub-mode ( 、 and ) are superimposed to obtain a high-frequency sequence, a kernel density estimation method and a box plot method are used to formulate a threshold index for screening the high-frequency sequence, local singular values in the high-frequency sequence are eliminated based on the composite threshold index, and a residual high-frequency sequence is generated, including steps S51 to S54:

[0107] Step S51: Superimpose the high-frequency sub-modes to obtain a high-frequency sequence :

[0108] ,

[0109] Step S52: Calculate the high frequency sequence using kernel density estimation method The probability density function of , and get the probability density function The maximum value ; Through the probability density function Query to obtain probability density The corresponding minimum threshold of kernel density estimation and the maximum threshold of kernel density estimation In this embodiment Take it as 0.1;

[0110] Step S53: Calculate high frequency sequence Upper quartile of , lower quartile and interquartile range , based on which the minimum threshold of the box plot is calculated and box plot maximum threshold :

[0111] ,

[0112] ,

[0113] Step S54: Filter high-frequency sequences based on the threshold index obtained in steps S42 and S43 The singular values in the high-frequency sequence Any value in ,when satisfy or or or , it is judged as a singular value and is removed.

[0114] Step S6: Superimpose the low- and medium-frequency sub-modes and the residual high-frequency sequence to obtain a reconstructed time series of the dam single-point safety monitoring data. The reconstructed time series of the dam single-point safety monitoring data is as follows: Figure 3 shown.

[0115] Combine Figure 3 It can be seen that the dam horizontal deformation monitoring data after effective information extraction retains the change characteristics of the original data very well, and the local singular values and overall noise characteristics are filtered out.

[0116] Example 2: See Figure 5 , is a schematic diagram of the structure of an effective information extraction platform for dam safety monitoring data based on adaptive time series decomposition provided by an embodiment of the present invention, including a reading module, a decomposition module, a screening module, a filtering module and a reconstruction module;

[0117] A reading module is used to read and input the safety monitoring data of a single measuring point of the dam, wherein the safety monitoring data of a single measuring point of the dam includes dam deformation, seepage pressure, seepage flow, and stress and strain data;

[0118] Decomposition module, used to perform adaptive variational mode decomposition algorithm on the safety monitoring data of a single measuring point of the dam, and obtain sub-modes of different frequency bands of the single measuring point dam safety monitoring sequence;

[0119] A screening module is used to adaptively screen the sub-modes, and adaptively screen the sub-modes decomposed by the decomposition module through a composite discrimination index composed of correlation coefficient and energy entropy ratio to obtain high-frequency sub-modes and medium- and low-frequency sub-modes;

[0120] The filtering module is used to superimpose high-frequency sub-modes to obtain a high-frequency sequence, review the discriminant index through the correlation coefficient and energy entropy ratio, and remove local singular values in the high-frequency sequence to obtain a residual high-frequency sequence;

[0121] The reconstruction module is used to superimpose the medium and low frequency sub-modes and the residual high frequency sequences to obtain the effective information extraction results of the safety monitoring data of a single measuring point of the dam.

[0122] Example 3: This embodiment proposes an electronic system, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method steps of the present invention.

[0123] Example 4: This example provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the method of the present invention are implemented, which will not be repeated here.

[0124] It should be noted that the processing flow of Examples 2-4 corresponds to the specific steps of the method provided in the embodiment of the present invention, and has the functional modules and beneficial effects corresponding to the execution method. For technical details not fully described in this embodiment, please refer to the method provided in the embodiment of the present invention.

[0125] The program code for implementing the methods of the present application can be written in any combination of one or more programming languages. Such program code can be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the program code is executed by the processor or controller, the functions / operations specified in the flow charts and / or block diagrams are implemented. The program code can be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0126] In the context of the present application, a machine-readable medium can be a tangible medium that can contain or store a program for use by an instruction execution system, device or equipment or used in combination with an instruction execution system, device or equipment. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared or semiconductor system, device or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium can include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0127] As described above, although the present invention has been shown and described with reference to specific preferred embodiments, it should not be construed as limiting the present invention itself. Various changes may be made to it in form and detail without departing from the spirit and scope of the present invention as defined in the appended claims.

Claims

1. A method for extracting effective information from dam safety monitoring data based on adaptive time series decomposition, characterized in that: The following steps are involved: Step S1, obtaining the time series of safety monitoring data of a single measuring point of the dam; Step S2: Based on the time series of safety monitoring data of a single measuring point of the dam, a multi-population Jaya algorithm is used to adaptively optimize the control parameters of the variational mode decomposition algorithm; Step S3, using a variational mode decomposition algorithm with optimized parameters to perform non-recursive decomposition on the time series of the dam single-point safety monitoring data; Step S4: performing feature analysis on the sub-modal sequence generated by the decomposition, and adaptively screening high-frequency sub-modal and medium-low frequency sub-modal by using the composite discrimination index of correlation coefficient and energy entropy ratio; Step S5: Superimpose the high-frequency sub-modes to obtain a high-frequency sequence, use the kernel density estimation method and the box plot method to formulate a composite threshold index for screening the high-frequency sequence, remove local singular values in the high-frequency sequence based on the composite threshold index, and generate a residual high-frequency sequence; Step S6: Superimpose the medium and low frequency sub-modes and the residual high frequency sequence to obtain a reconstructed time series of the dam single measuring point safety monitoring data, thereby completing the extraction of effective information from the dam safety monitoring data.

2. The method according to claim 1, characterized in that The safety monitoring data in step S1 include dam deformation, seepage pressure, seepage volume, and stress-strain data.

3. The method according to claim 1, characterized in that Step S2 includes the following steps: Step S201: Set the control parameter search space of variational mode decomposition, the control parameters include the decomposition mode number , penalty parameters and convergence tolerance , build control parameter set ; Step S202: Set the basic parameters of the multi-population Jaya algorithm, including the number of subpopulations. and the maximum number of iterations ; Step S203: Use variational modal decomposition to calculate the time series sub-modes of the dam single-point safety monitoring data, and arrange the decomposed sub-modes in order from low frequency to high frequency. The submode is denoted as , ; extract The envelope signal sequence , calculate the Submodal envelope entropy function ; Step S204: Taking minimizing the sum of the first two order envelope entropies as the goal, construct a second order envelope entropy loss function, as follows: ， Where, is the optimal decomposition mode number, is the optimal penalty parameter, is the optimal convergence tolerance; Step S205: Based on the minimization of the second-order envelope entropy loss function in step S204, the optimal solution in each candidate subpopulation is tracked, and the control parameters are updated using the multi-population Jaya algorithm based on the optimal solution and the worst solution in all subpopulations, as follows: ， Where, For the During the iteration The first candidate subpopulation The value of a variable, For the In the iteration, all subpopulations The optimal value of a variable, For the In the iteration, all subpopulations The worst value of a variable; and There are two between Random numbers in the interval; Step S206: Repeat steps S203-S205 until the number of iterations reaches , the optimal control parameter combination is obtained by taking the variational mode decomposition parameter corresponding to the minimum value of the second-order envelope entropy loss function when the iteration stops .

4. The method according to claim 3, characterized in that Envelope entropy function The specific calculation method is as follows: ， Where, For the Envelope entropy of the layer submode; is the number of samples of the sub-modality.

5. The method according to claim 1, wherein The S4 comprises the following steps: Step S401, calculate the Pearson correlation coefficient between the layer submode and the time series of the dam single-point safety monitoring data before preprocessing , as follows: ， Where, is the time series of the dam single-point safety monitoring data before preprocessing, For the Layer sub-mode, and They are and The mean of is the number of samples of the sub-modality; Step S402, calculate the Energy entropy of stratum modes , as follows: ， Where, For the Layer sub-mode, is the decomposition mode number; Step S403: repeat steps S401 and S402 to calculate the Pearson correlation coefficients of all submodes and the time series of the dam single-point safety monitoring data before preprocessing, as well as the energy entropy of all submodes; Step S404: Calculate the energy entropy ratio of all sub-modes : ， Step S405: Set the correlation coefficient threshold , energy entropy ratio threshold ;Compare and The size relationship, comparison and The size relationship will satisfy both and No. If the layer submode is determined to be high frequency, it is a high frequency submode; otherwise, it is determined to be a medium or low frequency submode.

6. The method according to claim 5, characterized in that In step S405, the correlation coefficient threshold is set Set the energy entropy ratio threshold to 0.2~0.3 is 0.

1.

7. The method according to claim 1, characterized in that The S5 comprises the following steps: Step S501: Superimpose high-frequency sub-modes to obtain a high-frequency sequence : ， Where, is the number of high-frequency submodes, is the decomposition mode number; Step S502: Calculate the high frequency sequence using kernel density estimation method The probability density function of , and get the probability density function The maximum value ; Through the probability density function Query to obtain probability density The corresponding minimum threshold of kernel density estimation and the maximum threshold of kernel density estimation , is a control coefficient between 0.01 and 0.1; Step S503: Calculate high frequency sequence Upper quartile of , lower quartile and interquartile range , based on which the minimum threshold of the box plot is calculated and box plot maximum threshold : ， ； Step S504: Filter high-frequency sequences based on the threshold indicators obtained in steps S502 and S503 The singular values in the high-frequency sequence Any value in ,when satisfy or or or , it is judged as a singular value and is removed.

8. A platform for extracting effective information from dam safety monitoring data based on time series decomposition, used to execute the method according to any one of claims 1 to 7, characterized in that: include: Reading module, used to read and input the safety monitoring data of a single measuring point of the dam; Decomposition module, used to perform adaptive variational modal decomposition algorithm on the safety monitoring data of a single measuring point of the dam to obtain sub-modes in different frequency bands; The screening module is used to perform adaptive screening on sub-modes to obtain high-frequency sub-modes and medium- and low-frequency sub-modes; The filtering module is used to superimpose the high-frequency sub-modes to obtain a high-frequency sequence, calculate the threshold index of the high-frequency sequence and remove the local singular values in the high-frequency sequence to obtain the residual high-frequency sequence; The reconstruction module is used to superimpose and reconstruct the medium and low frequency sub-modes and the residual high frequency sequences to obtain the effective information extraction results of the safety monitoring data of a single measuring point of the dam.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the computer program is executed, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the program implements the steps of the method according to any one of claims 1 to 7.

Images