Method for decomposing elastic deformation and plastic deformation of concrete dam based on wavelet transformation

Abstract

The invention belongs to the field of civil engineering and structural health monitoring, and provides a concrete dam elastic deformation and plastic deformation decomposition method based on wavelet transform in order to solve the problem of poor inversion precision caused by dam parameter analysis depending on the total deformation amount in the prior art, which can accurately distinguish different types of deformation and improve the accuracy of the concrete dam elastic deformation and plastic deformation decomposition. And scientific guidance is provided for refined simulation of the concrete dam state. The method comprises the following steps: acquiring sample data containing a concrete elastic deformation and plastic deformation mixed signal; cleaning the sample data; performing interpolation processing on the timestamps of the cleaned sample data; based on the sample data after interpolation processing, performing multi-scale decomposition by using wavelet transform; on the basis of a set frequency threshold value, wavelet coefficients obtained through multi-scale decomposition are screened, and an elastic deformation wavelet coefficient and a plastic deformation wavelet coefficient are separated out; and respectively reconstructing the separated elastic deformation wavelet coefficient and plastic deformation wavelet coefficient into complete time sequence data by utilizing a wavelet reconstruction technology.

Description

Technical Field

[0001] The invention belongs to the field of civil engineering and structural health monitoring, and in particular relates to a decomposition method of elastic deformation and plastic deformation of a concrete dam based on wavelet transform. Background Art

[0002] As a large-scale hydraulic structure, concrete dams are subject to multiple influences of the natural environment and operating conditions during their service, such as changes in reservoir water levels, temperature fluctuations, the weight of the dam body, and long-term loads of slope pressure. These external factors will cause changes in the stress and displacement of the dam body, resulting in structural deformation. The deformation of concrete dams mainly includes two categories: elastic deformation and plastic deformation. Elastic deformation is reversible and can be restored to its original state after being affected by external factors; while plastic deformation is irreversible, and the accumulated deformation will pose a potential threat to the stability of the dam structure during long-term service. Therefore, accurately distinguishing these two types of deformation and evaluating them separately is an important means to ensure the safe operation of the dam body, which helps to accurately reconstruct the elastic and plastic parameters, thereby further refining the simulation of the dam state.

[0003] Traditional monitoring technologies mostly rely on total deformation analysis, which makes it difficult to effectively distinguish between elastic and plastic deformations. This is because the deformation of concrete dams often contains multiple frequency components and is interfered by external noise, making the extraction and separation of deformation signals challenging. In addition, existing numerical analysis methods are prone to errors when faced with nonlinear deformations under complex working conditions, and are unable to make accurate judgments on the long-term cumulative effects of elastic and plastic deformations. Summary of the invention

[0004] The technical problem to be solved by the present invention is: in order to overcome the problem of poor inversion accuracy caused by relying on the total deformation to analyze dam parameters in traditional technology, a method for decomposing the elastic deformation and plastic deformation of concrete dams based on wavelet transform is proposed, which can accurately distinguish different types of deformation and provide scientific guidance for the refined simulation of the state of concrete dams.

[0005] The technical solution adopted by the present invention to solve the above technical problems is:

[0006] A method for decomposing elastic deformation and plastic deformation of a concrete dam based on wavelet transform comprises the following steps:

[0007] S1, obtaining sample data containing mixed signals of concrete elastic deformation and plastic deformation;

[0008] S2. Clean the acquired sample data;

[0009] S3, interpolating the timestamps of the cleaned sample data;

[0010] S4, based on the sample data after interpolation processing, multi-scale decomposition is performed using wavelet transform;

[0011] S5. Based on the set frequency threshold, the wavelet coefficients obtained by multi-scale decomposition are screened to separate the elastic deformation wavelet coefficients and the plastic deformation wavelet coefficients;

[0012] S6. Using wavelet reconstruction technology, the separated elastic deformation wavelet coefficients and plastic deformation wavelet coefficients are reconstructed into complete time series data respectively.

[0013] Furthermore, in step S1, the obtaining of sample data containing mixed signals of concrete elastic deformation and plastic deformation includes:

[0014] S11. Determine the key monitoring points of the concrete dam and monitor them, and simultaneously collect external environmental conditions;

[0015] S12. Systematically organize the collected monitoring data to obtain sample data of each monitoring point, where the sample data contains mixed signals of concrete elastic deformation and plastic deformation.

[0016] Furthermore, in step S2, the obtained sample data is cleaned, including:

[0017] S21, perform integrity check on the sample data of each monitoring point;

[0018] S22, performing denoising processing on the sample data;

[0019] S23. Correct or eliminate abnormal data in the sample data.

[0020] Furthermore, in step S23, the IQR (interquartile range) method is used to remove outliers:

[0021] Calculate the first quartile Q1 and the third quartile Q3 of the monitoring data;

[0022] Calculate the interquartile range IQR = Q3-Q1;

[0023] Set the upper and lower boundaries: lower boundary = Q1-1.5×IQR, upper boundary = Q3+1.5×IQR;

[0024] Data points outside the upper and lower boundaries are marked as outliers and removed.

[0025] Furthermore, in step S3, the interpolation processing of the timestamp of the cleaned sample data includes:

[0026] S31, checking the timestamps of the sample data one by one, and identifying uneven time intervals or missing data during the collection process;

[0027] S32. Perform linear interpolation on the missing data in the date axis according to the timestamp analysis results and data characteristics;

[0028] S33. By comparing the trend with the original data, verify whether the interpolated data is consistent with the actual situation in terms of trend and amplitude.

[0029] Furthermore, in step S4, the multi-scale decomposition is performed based on the sample data after the interpolation processing by using wavelet transform, including:

[0030] S41. Selecting a suitable wavelet basis function according to the frequency characteristics of the sample data;

[0031] S42. Use the selected wavelet basis function to perform multi-scale decomposition on the sample data.

[0032] Furthermore, in step S42, before performing multi-scale decomposition on the sample data using the selected wavelet basis function, the number of wavelet decomposition layers is adaptively adjusted according to the response characteristics of different monitoring points to water level changes.

[0033] Furthermore, in step S5, based on the set frequency threshold, the wavelet coefficients obtained by multi-scale decomposition are screened to separate the elastic deformation wavelet coefficients and the plastic deformation wavelet coefficients, including:

[0034] S51, combining the results of wavelet decomposition and the deformation characteristics of the dam body, determining the frequency ranges of elastic deformation and plastic deformation respectively;

[0035] S52, setting a frequency threshold according to the frequency range of deformation, and screening out effective elastic deformation wavelet coefficients and plastic deformation wavelet coefficients;

[0036] S53. Verify and analyze the screened elastic deformation wavelet coefficients and plastic deformation wavelet coefficients.

[0037] Furthermore, in step S6, the wavelet reconstruction technology is used to reconstruct the separated elastic deformation wavelet coefficients and plastic deformation wavelet coefficients into complete time series data, including:

[0038] S61, selecting the elastic deformation wavelet coefficients after screening, performing wavelet reconstruction, restoring their time domain signals, and generating a time series of the elastic deformation of the dam body;

[0039] S62. Select the screened plastic deformation wavelet coefficients, perform wavelet reconstruction, restore their time domain signals, and obtain the long-term cumulative curve of the dam body plastic deformation.

[0040] The beneficial effects of the present invention are:

[0041] (1) Based on the cleaning and interpolation of sample data, the accuracy, continuity and integrity of the data are ensured, thus providing support for subsequent more accurate wavelet decomposition.

[0042] (2) Based on wavelet transform technology, the monitoring data is decomposed into multiple scales, and elastic deformation and plastic deformation are effectively separated in different frequency ranges. Through the multi-scale analysis characteristics of wavelet transform, complex deformation signals can be simplified and extracted, which significantly improves the accuracy of deformation analysis.

[0043] (3) During the wavelet decomposition process, the number of wavelet decomposition layers is adaptively adjusted according to the response characteristics of different measuring points to water level changes, so as to more accurately decompose the elastic-plastic deformation of different measuring points.

[0044] (4) Based on the frequency screening mechanism, the noise and interference factors in the signal can be further eliminated, making the separated deformation information more accurate, which is suitable for the separation of frequently fluctuating elastic deformation and relatively stable plastic deformation in long-term monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 It is a flow chart of the decomposition method of elastic deformation and plastic deformation of concrete dam based on wavelet transform in the present invention;

[0046] Figure 2 It is a flow chart of adaptive wavelet layer decomposition in the present invention;

[0047] Figure 3 Schematic diagram of elastic-plastic displacement decomposition of a dam body in an embodiment of the present invention;

[0048] Figure 4 Schematic diagram of the elastoplastic displacement decomposition of the slope in an embodiment of the present invention. DETAILED DESCRIPTION

[0049] The present invention aims to overcome the problem of poor inversion accuracy caused by the traditional technology that relies on the total deformation to analyze the dam parameters, and provides a method for decomposing the elastic deformation and plastic deformation of concrete dams based on wavelet transform, which can accurately distinguish different types of deformation and provide scientific guidance for the refined simulation of the state of concrete dams. The core idea is that the wavelet transform technology can decompose the deformation signal into different frequency components due to its multi-scale decomposition ability, which is suitable for processing such complex and unstable deformation signals. The present invention corresponds the high-frequency part and the low-frequency part to the elastic deformation and plastic deformation through wavelet decomposition, which can achieve more accurate deformation analysis. At the same time, the combination of frequency screening mechanism and wavelet reconstruction technology can eliminate noise and external interference signals, and provide more reliable data support for the safety assessment of the dam body. This method can effectively improve the accuracy of dam deformation monitoring and the accuracy of assessment in practical applications.

[0050] In specific implementation, the implementation process of the method for decomposing elastic deformation and plastic deformation of concrete dam based on wavelet transform provided by the present invention can be found in Figure 1 , which comprises the following steps:

[0051] S1, obtaining sample data containing mixed signals of concrete elastic deformation and plastic deformation;

[0052] In this step, by monitoring the key information (such as water level, displacement, etc.) of multiple key monitoring points of the concrete dam, deformation data under different environmental conditions and operating states are obtained to ensure the integrity and diversity of the sample data. The sample data contains mixed signals of elastic and plastic deformation.

[0053] In an exemplary embodiment, the implementation process of this step is as follows:

[0054] S11. Determine the key monitoring points of the concrete dam, including the stress concentration area and displacement sensitive area of ​​the dam body, obtain data samples through long-term monitoring, and simultaneously collect external environmental conditions such as temperature, humidity, and load changes to ensure that the deformation data can fully reflect the response of the dam body under different environmental conditions.

[0055] S12. Systematically organize the collected monitoring data, including time, stress, displacement and other multi-dimensional information, to form a numerical sample library to facilitate subsequent data analysis.

[0056] S2. Clean the acquired sample data;

[0057] In this step, the collected deformation data is cleaned. This step ensures the quality of subsequent analysis data by removing outliers, missing data, and noise signals in the monitoring data. The data cleaning process includes noise filtering, anomaly detection, and processing to ensure the accuracy of the deformation data.

[0058] In an exemplary embodiment, the implementation process of this step is as follows:

[0059] S21. Check the integrity of the sample data of each monitoring point: Check the data of each monitoring point step by step, identify missing data and outliers, and ensure that the data used for subsequent analysis is complete. For missing values ​​in the data, linear interpolation is performed using adjacent data.

[0060] S22. De-noising the sample data: By performing wavelet transform on the data, extracting and deleting higher frequency coefficients, the high-frequency noise in the test data can be removed, the purity of the data can be improved, and the reliability of subsequent analysis can be ensured.

[0061] S23. Correct or eliminate abnormal data in sample data: Correct or eliminate abnormal data caused by human or biological factors to ensure the accuracy and reliability of the data and avoid misleading the analysis results. For example, the IQR (interquartile range) method can be used to remove outliers: calculate the first quartile Q1 and the third quartile Q3 of the monitoring data; calculate the interquartile range IQR = Q3-Q1; set upper and lower boundaries: lower boundary = Q1-1.5×IQR, upper boundary = Q3+1.5×IQR; mark the data points beyond the upper and lower boundaries as outliers and remove them.

[0062] S3, interpolating the timestamps of the cleaned sample data;

[0063] In this step, in order to ensure the consistency and continuity of the time series data, the timestamps of the deformed data are interpolated. Data interpolation can effectively supplement the missing time data caused by equipment failure or incomplete collection during the monitoring process, ensure the integrity of the time series, and avoid the influence of inconsistent time intervals on the accuracy of wavelet decomposition.

[0064] In an exemplary embodiment, the implementation process of this step is as follows:

[0065] S31. Check the timestamps of the sample data one by one to identify uneven time intervals or missing data during the collection process; provide a basis for subsequent interpolation processing; and use the average value instead of the monitoring value of the day for multiple monitoring data at the same time.

[0066] S32. Perform linear interpolation on the missing data in the date axis based on the timestamp analysis results and data characteristics; ensure that the interpolation result is smooth and consistent with the original data to generate a complete and continuous time series.

[0067] S33. By comparing the trend with the original data, verify whether the interpolated data is consistent with the actual situation in terms of trend and amplitude, so as to ensure the rationality of the interpolation processing.

[0068] S4, based on the sample data after interpolation processing, multi-scale decomposition is performed using wavelet transform;

[0069] In this step, a suitable discrete wavelet (such as db4 or sym5) is selected to decompose the monitoring data into sub-bands of different frequencies: the high-frequency part corresponds to elastic deformation, which is characterized by rapid generation and recovery with changes in the environment and load; the low-frequency part corresponds to plastic deformation, which is characterized by slow accumulation over a long period of time and irreversible deformation.

[0070] In an exemplary embodiment, the implementation process of this step is as follows:

[0071] S41. Select wavelet basis function: According to the frequency characteristics of the deformation data, select a suitable wavelet basis function (such as db4, sym5) to ensure that the wavelet transform can accurately decompose the elastic and plastic deformations and no new abnormal values ​​appear.

[0072] Generally speaking, the selection of wavelet basis functions has a great influence on coefficient extraction and reconstruction. For example, when using db1 and sym1 wavelets for reconstruction, the low-frequency band of the data, i.e., plastic deformation, will present multiple platform shapes instead of smooth curves, which is inconsistent with the development process of plastic deformation. Therefore, it is very important to select appropriate wavelet basis functions such as db4 and sym5 that can reconstruct smooth curves.

[0073] S42. Perform wavelet decomposition: perform multi-scale decomposition on the sample data using the selected wavelet basis function.

[0074] Here we assume that the plastic deformation of the dam body is only viscoplastic deformation. It can be expressed by the following formula:

[0075]

[0076] Among them, e T is the deformation caused by the variable applied load, e a is the deformation caused by permanent load. Both can cause elastic-plastic deformation to the dam body. T Usually affected by water level and temperature changes with annual and daily cycles, which can be recorded as e T =e T (T 1 ,T 2 ), since this part of the load has a large fluctuation range, if viscoplastic deformation occurs in the part with higher stress, it will have little effect on the overall viscoplasticity, while if viscoplastic deformation occurs in the part with lower stress, there will be a very obvious deformation difference within a few cycles. In fact, this kind of deformation is not observed, so it can be approximately considered that

[0077] Since the elastic deformation caused by permanent loads is basically unchanged, only creep parameters will affect this part of the monitoring results, so we only need to pay attention to The impact on the dam body. Therefore, the above formula can be further rewritten as:

[0078]

[0079] By observing the above formula, it can be considered that the elastic deformation generated by the dam is a fluctuating value with an initial value and changes with the period, while the viscoplastic deformation is a value that has little to do with the periodicity but grows all the time. The two can be distinguished by some means. If the deformation is regarded as a signal, the elastic deformation is a stable signal (a signal whose distribution parameters do not change with time, such as sint). According to the Duvat_Lions model, the viscoplastic strain rate is:

[0080]

[0081] The preprocessed deformation data is decomposed into multiple scales to extract different frequency components so that the high-frequency part reflects the transient elastic deformation and the low-frequency part reflects the long-term plastic deformation.

[0082] The detected displacement signal is denoted as ψ j,k , then its wavelet transform is Thus, the wavelet coefficients, i.e., frequency information, are obtained.

[0083] In the process of wavelet decomposition, according to the response characteristics of different measuring points to water level changes, the number of wavelet decomposition layers is adaptively adjusted to more accurately decompose the elastic-plastic deformation of different measuring points. Based on this adaptive wavelet decomposition layer setting, the decomposition process is shown in Figure 2 First, the external factors that affect the displacement of the dam body due to the water level are decomposed. Generally, more than 10 decomposition layers are required, and the wavelet coefficients of each layer after decomposition are saved and recorded as W_w. Then, the monitoring displacement signal is decomposed layer by layer to extract the high-frequency part A1 and the low-frequency part A2. Initially, A1 is decomposed by wavelet once to obtain the first layer coefficient. Calculate the correlation coefficient between the current layer coefficient and the corresponding layer coefficient of B. If the correlation coefficient is greater than the set threshold, continue to decompose A1 until the correlation does not meet the requirements.

[0084] S5. Based on the set frequency threshold, the wavelet coefficients obtained by multi-scale decomposition are screened to separate the elastic deformation wavelet coefficients and the plastic deformation wavelet coefficients;

[0085] In this step, a frequency screening mechanism is introduced to screen the coefficients obtained after wavelet decomposition based on the set frequency threshold to eliminate interference factors. By screening the wavelet coefficients, irrelevant noise signals can be removed and the extracted elastic deformation and plastic deformation can be made purer. The screening mechanism can effectively improve the accuracy of deformation decomposition, especially in the case of mixed deformation signals, and can highlight the key deformation modes.

[0086] In an exemplary embodiment, the implementation process of this step is as follows:

[0087] S51. Determine the frequency range of deformation: Based on the results of wavelet decomposition and the deformation characteristics of the dam body, determine the frequency ranges of elastic deformation and plastic deformation respectively as the basis for screening wavelet coefficients.

[0088] S52, set frequency threshold: according to the deformation frequency range, set a reasonable frequency threshold, filter out effective wavelet coefficients, remove possible mixed noise signals and irrelevant deformation components, and for wavelet decomposition, each layer of decomposition corresponds to a different frequency range, so the frequency threshold can be set according to the decomposition effect of the selected specific wavelet basis function to effectively separate elastic deformation and plastic deformation. For example: when the db4 wavelet function is used as the basis function for processing, there is usually a good separation effect at the 7th to 9th layer; when the sym5 wavelet function is used as the basis function for processing, there is usually a good separation result at the 8th to 10th layer.

[0089] S53. Verify the screening results: Analyze the screened wavelet coefficients to ensure that the screened elastic deformation and plastic deformation coefficients can accurately reflect the actual deformation characteristics and laws.

[0090] S6. Using wavelet reconstruction technology, the separated elastic deformation wavelet coefficients and plastic deformation wavelet coefficients are reconstructed into complete time series data respectively;

[0091] In this step, the wavelet coefficients of separated elastic deformation and plastic deformation are reconstructed into complete time series data through wavelet reconstruction technology. In the reconstruction process, the trend curves of elastic deformation and plastic deformation can be obtained respectively, which are further used for health monitoring and safety assessment of the dam. The elastic deformation curve shows the changes in the short-term response of the dam body, while the plastic deformation curve reflects the long-term accumulated irreversible deformation.

[0092] In an exemplary embodiment, the implementation process of this step is as follows:

[0093] S61. Elastic deformation wavelet reconstruction: Select the screened elastic deformation wavelet coefficients, perform wavelet reconstruction, restore their time domain signals, and generate a time series of dam body elastic deformation.

[0094] S62. Plastic deformation wavelet reconstruction: Use the same method to reconstruct the screened plastic deformation wavelet coefficients to obtain the long-term cumulative curve of the dam body's plastic deformation, reflecting the irreversible deformation of the structure.

[0095] Based on the above reconstructed deformation curve, the corresponding analysis can be performed:

[0096] The reconstructed elastic deformation curves are analyzed to observe the short-term dynamic response and recovery of the dam under external effects and to evaluate its elastic behavior.

[0097] Analyze the changing trend of the plastic deformation curve, evaluate the accumulation of irreversible deformation of the dam during long-term service, and provide a predictive basis for the health status of the dam.

[0098] Example

[0099] The method for decomposing elastic deformation and plastic deformation of a concrete dam based on wavelet transform provided in this embodiment includes the following steps:

[0100] S1. Collect and construct numerical samples including elastic and plastic deformation of concrete;

[0101] In this example, the displacement monitoring information of a dam is used as an example to obtain long-term displacement and water level information to ensure the integrity and diversity of the sample data. The sample data contains mixed signals of elastic and plastic deformation.

[0102] S2, cleaning the collected deformation data;

[0103] In this example, Z-score is used to identify outliers. The calculation formula is: Set the threshold to 3 and remove data that exceeds the threshold.

[0104] S3. To ensure the consistency and continuity of the time series data, interpolation processing is performed on the timestamps of the deformed data;

[0105] In this example, the displacement information of each day is collected unevenly. In order to maintain the consistency and continuity of the time series data, when there are more than two displacement information per day, the average value is used as the displacement information of the day, and the missing information of blank dates is filled by interpolation. This rule is used to remove duplicate values ​​and fill blank values.

[0106] S4, based on the processed data, multi-scale decomposition is performed using wavelet transform;

[0107] In this example, in the wavelet decomposition with water level fluctuation as the main waveform, in order to ensure that the decomposed and reconstructed data maintain a high similarity with the original data, the appropriate discrete wavelet bior6.8 is selected to decompose the monitoring data into sub-bands of different frequencies. The high-frequency part corresponds to elastic deformation, which is characterized by rapid generation and recovery with changes in the environment and load; the low-frequency part corresponds to plastic deformation, which is characterized by slow accumulation over a long period of time and irreversible deformation.

[0108] S5, introduce adaptive layer number screening mechanism;

[0109] In this example, the coefficients obtained after wavelet decomposition are screened based on the set frequency threshold to remove interference factors. By determining the number of decomposition layers to screen wavelet coefficients, irrelevant noise signals can be removed and the extracted elastic and plastic deformations can be made purer. The screening mechanism can effectively improve the accuracy of deformation decomposition, especially in the case of mixed deformation signals, and can highlight key deformation modes.

[0110] S6. Reconstruct the separated wavelet coefficients of elastic deformation and plastic deformation into complete time series data through wavelet reconstruction technology;

[0111] In this example, using wrcoef to reconstruct the signal can have a better tolerance for the signal length, which is conducive to reconstructing signals of different decomposition levels into sequences of uniform length, and using the filtered high-frequency signal as elastic displacement for reconstruction, and the low-frequency signal as plastic displacement. The specific display results of the dam displacement are as follows: Figure 3 As shown in the figure, elastic deformation is dominant, with little plastic deformation. The specific results of slope displacement are as follows: Figure 4 As shown, the plastic deformation is larger than the dam body deformation.

[0112] Finally, it should be noted that the above embodiments are only preferred implementations and are not intended to limit the present invention. It should be pointed out that for those skilled in the art, several modifications, equivalent replacements, improvements, etc. can be made without departing from the scope of the present invention and the scope of protection of the claims, and all of these should be included in the protection scope of the present invention.

Claims

1. A method for decomposing elastic and plastic deformation of concrete dams based on wavelet transform, characterized in that: The following steps are involved: S1, obtaining sample data containing mixed signals of concrete elastic deformation and plastic deformation; S2. Clean the acquired sample data; S3, interpolating the timestamps of the cleaned sample data; S4, based on the sample data after interpolation processing, multi-scale decomposition is performed using wavelet transform; S5. Based on the set frequency threshold, the wavelet coefficients obtained by multi-scale decomposition are screened to separate the elastic deformation wavelet coefficients and the plastic deformation wavelet coefficients; S6. Using wavelet reconstruction technology, the separated elastic deformation wavelet coefficients and plastic deformation wavelet coefficients are reconstructed into complete time series data respectively.

2. A method for decomposing elastic and plastic deformation of concrete dams based on wavelet transform as claimed in claim 1, characterized in that: In step S1, the step of obtaining sample data containing mixed signals of concrete elastic deformation and plastic deformation includes: S11. Determine the key monitoring points of the concrete dam and monitor them, and simultaneously collect external environmental conditions; S12. Systematically organize the collected monitoring data to obtain sample data of each monitoring point, where the sample data contains mixed signals of concrete elastic deformation and plastic deformation.

3. The method for decomposing elastic and plastic deformation of concrete dam based on wavelet transform according to claim 1, characterized in that: In step S2, the obtained sample data is cleaned, including: S21, perform integrity check on the sample data of each monitoring point; S22, performing denoising processing on the sample data; S23. Correct or eliminate abnormal data in the sample data.

4. A method for decomposing elastic and plastic deformation of concrete dams based on wavelet transform as claimed in claim 3, characterized in that: In step S23, the IQR method is used to remove outliers: Calculate the first quartile Q1 and the third quartile Q3 of the monitoring data; Calculate the interquartile range IQR = Q3-Q1; Set the upper and lower boundaries: lower boundary = Q1-1.5×IQR, upper boundary = Q3+1.5×IQR; Data points outside the upper and lower boundaries are marked as outliers and removed.

5. The method for decomposing elastic and plastic deformation of concrete dam based on wavelet transform according to claim 1, characterized in that: In step S3, the timestamp of the cleaned sample data is interpolated, including: S31, checking the timestamps of the sample data one by one, and identifying uneven time intervals or missing data during the collection process; S32. Perform linear interpolation on the missing data in the date axis according to the timestamp analysis results and data characteristics; S33. By comparing the trend with the original data, verify whether the interpolated data is consistent with the actual situation in terms of trend and amplitude.

6. A method for decomposing elastic and plastic deformation of concrete dams based on wavelet transform as claimed in claim 1, It is characterized in that In step S4, the multi-scale decomposition is performed based on the sample data after the interpolation processing by using wavelet transform, including: S41. Selecting a suitable wavelet basis function according to the frequency characteristics of the sample data; S42. Use the selected wavelet basis function to perform multi-scale decomposition on the sample data.

7. A method for decomposing elastic and plastic deformation of concrete dams based on wavelet transform as claimed in claim 6, characterized in that: In step S42, before using the selected wavelet basis function to perform multi-scale decomposition on the sample data, the number of wavelet decomposition layers is adaptively adjusted according to the response characteristics of different monitoring points to water level changes.

8. The method for decomposing elastic and plastic deformation of concrete dam based on wavelet transform as claimed in claim 1, characterized in that: In step S5, based on the set frequency threshold, the wavelet coefficients obtained by multi-scale decomposition are screened to separate the elastic deformation wavelet coefficients and the plastic deformation wavelet coefficients, including: S51, combining the results of wavelet decomposition and the deformation characteristics of the dam body, determining the frequency ranges of elastic deformation and plastic deformation respectively; S52, setting a frequency threshold according to the frequency range of deformation, and screening out effective elastic deformation wavelet coefficients and plastic deformation wavelet coefficients; S53. Verify and analyze the screened elastic deformation wavelet coefficients and plastic deformation wavelet coefficients.

9. The method for decomposing elastic and plastic deformation of concrete dam based on wavelet transform according to claim 1, characterized in that: In step S6, the wavelet reconstruction technology is used to reconstruct the separated elastic deformation wavelet coefficients and plastic deformation wavelet coefficients into complete time series data, including: S61, selecting the elastic deformation wavelet coefficients after screening, performing wavelet reconstruction, restoring their time domain signals, and generating a time series of the elastic deformation of the dam body; S62. Select the screened plastic deformation wavelet coefficients, perform wavelet reconstruction, restore their time domain signals, and obtain the long-term cumulative curve of the dam body plastic deformation.

Images