Self-expanding particle anti-seepage effect evaluation method based on multiple regression analysis

Abstract

The invention discloses a self-expanding particle anti-seepage effect evaluation method based on multivariate regression analysis, which can comprehensively consider the influence of a plurality of key factors such as permeability coefficient, flow and anti-seepage effective time on the self-expanding particle anti-seepage effect by using a multivariate regression analysis method, and is not limited to the effect of analyzing a single factor. The method has the main advantages that multiple interaction effects in a complex system can be comprehensively captured, and evaluation deviation caused by neglecting some variables is avoided. By simultaneously incorporating a plurality of factors for analysis, the multiple regression not only can reveal the independent influence of the factors on the anti-seepage effect, but also can identify the possible interaction or interdependence relationship among the factors.

Description

Technical Field

[0001] The invention belongs to the field of anti-seepage effect evaluation, and in particular relates to an anti-seepage effect evaluation method of self-expanding particles based on multivariate regression analysis. Background Art

[0002] In the construction of water conservancy projects, due to reasons such as construction design and construction quality, the project structure often faces the risk of instability and damage, especially seepage damage has become one of the main factors for the instability of water conservancy facilities such as dams. Seepage damage not only causes physical damage to the structure, but may also trigger a series of chain reactions, endangering surrounding towns, major transportation arteries and other key infrastructure. Therefore, early warning and effective control of seepage damage have become important issues in water conservancy projects.

[0003] The hidden dangers of dam leakage are particularly complex and difficult to predict. They often gradually emerge in the long-term accumulation and suddenly intensify under certain specific conditions, leading to serious consequences. In order to effectively deal with these hidden dangers, traditional leakage control measures such as concrete anti-seepage, high-pressure grouting and split grouting are widely used. These methods can indeed provide better anti-seepage effects by physically blocking and enhancing structural strength. However, these traditional technologies also have some obvious shortcomings. First, the construction period is usually long, and the construction process is complicated and demanding, which not only increases the cost of the project, but may also delay the progress of the entire project. Secondly, these methods may have adverse effects on the surrounding environment during implementation, such as interference with groundwater flow and damage to the ecosystem, which makes it difficult to meet the requirements of modern engineering for sustainable development.

[0004] In order to solve these problems, in recent years, self-expanding granular anti-seepage blanketing technology has gradually become a new anti-seepage measure that has attracted much attention. This technology combines the dual functions of anti-seepage and ecological restoration. By laying granular materials with self-expanding properties on the surface of the dam, it can automatically expand and fill the leakage channel when water comes into contact, thereby effectively preventing water from infiltrating. At the same time, self-expanding granular materials can also promote bioremediation, restore and maintain the local ecological environment. Therefore, this technology not only performs well in anti-seepage effects, but also has good environmental compatibility and ecological benefits, which is in line with the current concept of sustainable development.

[0005] However, although the self-expanding particle anti-seepage blanket technology has shown many advantages in practical applications, the evaluation method of its anti-seepage effect is still an issue that needs in-depth research. Traditional evaluation methods such as permeability coefficient method, flow method and porosity method usually only consider the influence of a single factor and cannot comprehensively and objectively evaluate the actual anti-seepage performance of self-expanding particles. These methods may perform well under certain specific conditions, but they often seem powerless when faced with complex and changeable actual environments. Since the anti-seepage effect is often affected by multiple factors, such as the distribution of granular materials, changes in environmental conditions, and the complexity of leakage channels, traditional single factor analysis is difficult to provide a comprehensive performance evaluation.

[0006] In view of this, the present invention is proposed. Summary of the invention

[0007] The technical problem to be solved by the present invention is to overcome the deficiencies of the prior art and provide a method for evaluating the anti-seepage effect of self-expanding particles based on multivariate regression analysis, thereby solving the problems raised in the above-mentioned background technology.

[0008] In order to solve the above technical problems, the basic concept of the technical solution adopted by the present invention is:

[0009] A method for evaluating the anti-seepage effect of self-expanding particles based on multivariate regression analysis comprises the following steps:

[0010] Step (1): determining relevant indicators for evaluating the anti-seepage effect of self-expanding particles;

[0011] Step (2): collecting test data of self-expanding particles and dividing the data into a training set and a validation set;

[0012] Step (3): Taking the permeability coefficient, flow rate and effective anti-seepage time as independent variables and the anti-seepage effect of the self-swelling particles as the dependent variable, a polynomial regression model is constructed;

[0013] Step (4): verifying the polynomial regression model using the remaining data and calculating the model's goodness of fit index;

[0014] Step (5): Optimize and adjust the parameters of the regression model according to the model verification results;

[0015] Step (6): Revalidate the optimized model and use the K-fold cross-validation method to further verify the robustness and generalization ability of the model;

[0016] Step (7): Use the optimized multivariate regression model to evaluate the anti-seepage effect of the self-expanding particles.

[0017] Optionally, the step of collecting test data of the self-expanding particles in step (2) is:

[0018] The permeability coefficient of self-expanding particles was calculated by borehole water injection test;

[0019] The flow rate data is obtained through permeability tests, and the penetration time is controlled to obtain the test data of the effective anti-seepage time. The collected test data are integrated to obtain a data set, and the obtained data set is divided into a training set and a test set.

[0020] Optionally, in step (3), the permeability coefficient, flow rate and effective anti-seepage time are used as independent variables, and the anti-seepage effect of the self-swelling particles is used as the dependent variable, and the steps of constructing a polynomial regression model are as follows:

[0021] Let the independent variable be X, the dependent variable be Y, and the k independent variables that affect the dependent variable are X1, X2, ...X k If there is a certain correlation between y and x, a multiple regression model is established based on the dependent variable y and each influencing factor. Its expression is: Y = β0 + β1X1 + β2X2 + ... + β K X K +ε, where, β0, β1, β2,...β K is the regression parameter;

[0022] When there is only one independent variable, it is called univariate polynomial regression; if there are multiple independent variables, it is called multivariate polynomial regression. The mathematical expression of the polynomial regression model is: y = b o +b1X+b2X2+b3X3+…+b n X n +ε, where b o , b1, b2, b3, ... b n are the coefficients of the polynomial regression model, corresponding to the constant term X raised to the first power, X raised to the second power, and so on, up to X raised to the nth power. ε is the error term, which indicates the factors that cannot be fully explained by the model.

[0023] Optionally, in step (4), the polynomial regression model is verified by the remaining data, and the goodness-of-fit index of the model is calculated:

[0024] The remaining data were used to verify the initial regression model, using the formula Calculate the coefficient of determination R2 of the model to evaluate the model's fit and prediction accuracy, where y is the actual value, is the predicted value, is the mean of the actual values, and n is the number of samples.

[0025] Optionally, the step of optimizing and adjusting the parameters of the regression model according to the model validation results in step (5) is:

[0026] By formula Adjust the weights of the independent variables and add new explanatory variables to find the optimal combination of model parameters, where the squared error term Represents the error between the model prediction value and the actual value. By minimizing this error, the optimal combination of weights and explanatory variable coefficients can be found. (i) is the actual value of the i-th sample, x (i) is the independent variable vector of the i-th sample, is the inner product of the independent variable vector and the weight vector, representing the linear combination of the independent variables, Z (i) is the newly introduced explanatory variable, which is a combination of existing independent variables or external additional information. α is the weight coefficient of the new explanatory variable, which indicates the impact of the new variable on the output. is the L2 regularization term, which is used to control the complexity of the model. λ is the regularization strength coefficient, which controls the influence of regularization. is the L2 norm of the weight vector, that is, the sum of the squares of the weight vector.

[0027] Optionally, in step (6), the optimized model is revalidated and the K-fold cross validation method is used to further validate the robustness and generalization ability of the model:

[0028] After parameter optimization, the remaining data is used again to verify the optimized model and compare the model goodness of fit and prediction accuracy before and after optimization to evaluate the effect of parameter optimization;

[0029] If the model verification results are still not ideal, continue to adjust the parameters until they meet the requirements;

[0030] After the model validation results are satisfactory, the data is initially sampled and divided into K sub-samples. A single sub-sample is retained as data for the validation model, and the other K-1 samples are used for training and cross-validation repeated K times. Finally, each sub-sample is validated once and the results are averaged K times to obtain a single estimate. Different subsets are used for modeling and validation in each validation.

[0031] Optionally, after evaluating the anti-seepage effect, the following steps are also included:

[0032] Select the ARIMA model based on the data characteristics and use the training set to train the time series model;

[0033] Input new monitoring data into the trained time series model, dynamically update the model prediction results and generate short-term and long-term prediction values ​​of permeability coefficient, flow rate and effective time of anti-seepage based on the current time series data and model;

[0034] Based on the model prediction results, analyze possible future change trends and identify potential leakage risks;

[0035] Based on the trend prediction results, suggestions for adjusting anti-seepage measures are provided, and an early warning threshold is set. When the predicted permeability coefficient, flow rate or effective anti-seepage time exceeds the early warning threshold, the early warning mechanism is triggered.

[0036] Optionally, when the forecast results show an obvious risk trend, an early warning signal is automatically issued and the forecast results are sent to the staff via a communication signal.

[0037] After adopting the above technical solution, the present invention has the following beneficial effects compared with the prior art. Of course, any product implementing the present invention does not necessarily need to achieve all the advantages described below at the same time:

[0038] The present invention uses a multiple regression analysis method to comprehensively consider the influence of multiple key factors, such as permeability coefficient, flow rate and effective time of anti-seepage, on the anti-seepage effect of self-expanding particles, rather than being limited to analyzing the effect of a single factor. The main advantage of this method is that it can fully capture the multiple interaction effects in a complex system and avoid evaluation bias caused by ignoring certain variables. By incorporating multiple factors into the analysis at the same time, multiple regression can not only reveal the independent influence of these factors on the anti-seepage effect, but also identify the possible interactions or interdependencies between them.

[0039] The specific implementation modes of the present invention are further described in detail below with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The drawings described below are only some embodiments. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0041] In the figure:

[0042] Figure 1 This is a flow chart of the anti-seepage effect evaluation method.

[0043] It should be noted that these drawings and textual descriptions are not intended to limit the conceptual scope of the present invention in any way, but are intended to illustrate the concept of the present invention for those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0044] The present invention will now be described in further detail with reference to the accompanying drawings.

[0045] See also Figure 1 As shown, in this embodiment, a method for evaluating the anti-seepage effect of self-expanding particles based on multivariate regression analysis is provided, comprising the following steps:

[0046] Step (1): determining relevant indicators for evaluating the anti-seepage effect of self-expanding particles;

[0047] Step (2): collecting test data of self-expanding particles and dividing the data into a training set and a validation set;

[0048] Step (3): Taking the permeability coefficient, flow rate and effective anti-seepage time as independent variables and the anti-seepage effect of the self-expanding particles as the dependent variable, a polynomial regression model is constructed; by taking the permeability coefficient, flow rate and effective anti-seepage time as independent variables, the influence of these factors on the anti-seepage effect is systematically analyzed, thereby providing data-based decision support. This method helps to avoid subjective judgment and ensure the scientific nature of decision-making.

[0049] Step (4): Verify the polynomial regression model with the remaining data and calculate the goodness-of-fit index of the model; by selecting the polynomial regression model, the nonlinear relationship between the independent variable and the dependent variable can be captured, so that the model can better fit the data and provide more accurate prediction results.

[0050] Step (5): Optimize and adjust the parameters of the regression model according to the model verification results;

[0051] Step (6): Revalidate the optimized model and use the K-fold cross-validation method to further verify the robustness and generalization ability of the model; evaluate the robustness and generalization ability of the model through K-fold cross-validation to ensure that the model can maintain consistent performance on different data sets. This optimization process can effectively improve the prediction accuracy of the model. In addition, by using the K-fold cross-validation method, the robustness and generalization ability of the model are further verified to ensure that the model not only performs well on the training data, but also can adapt to new and unseen data.

[0052] Secondly, through parameter optimization and revalidation steps, we ensure that the model can be dynamically adjusted according to the validation results, thereby continuously improving the performance of the model. This dynamic optimization mechanism helps to cope with complex environments and changing conditions in practical applications.

[0053] Step (7): Use the optimized multivariate regression model to evaluate the anti-seepage effect of the self-expanding particles.

[0054] In this embodiment, the step of collecting the test data of the self-expanding particles in step (2) is:

[0055] The permeability coefficient of the self-expanding particles was calculated by the borehole water injection test, and its expression is: Among them, K is the permeability coefficient, which represents the permeability of the material, Q is the flow rate, L is the length of the permeation path, A is the cross-sectional area, and ΔH is the head difference.

[0056] The flow rate data is obtained through permeability test, and the permeability time is controlled to obtain the test data of the effective time of anti-seepage. The collected test data are integrated to obtain a data set, and the obtained data set is divided into a training set and a test set. Reasonable division of the data set into training set and test set helps to build a robust prediction model. The training set is used to build the model, while the test set is used to verify the generalization ability of the model to ensure that the model is not overfitted. By collecting data on permeability coefficient, flow rate and effective time of anti-seepage at the same time, a multi-parameter comprehensive analysis can be performed. This method can comprehensively evaluate the anti-seepage effect of self-expanding particles, thereby providing more reliable engineering guidance.

[0057] In this embodiment, in step (3), the permeability coefficient, flow rate and anti-seepage effective time are used as independent variables, and the anti-seepage effect of the self-expanding particles is used as the dependent variable. The steps of constructing a polynomial regression model are:

[0058] Let the independent variable be X, the dependent variable be Y, and the k independent variables that affect the dependent variable are X1, X2, ...X k If there is a certain correlation between y and x, a multiple regression model is established based on the dependent variable y and each influencing factor. Its expression is: Y = β0 + β1X1 + β2X2 + ... + β K X K +ε, where, β0, β1, β2,...β K is the regression parameter; the introduction of the polynomial regression model provides a technical approach to deal with nonlinear relationships. In many practical applications, there may be complex nonlinear relationships between dependent variables and independent variables. The use of polynomial regression can effectively capture these relationships and avoid the errors that may be caused by linear models.

[0059] When there is only one independent variable, it is called univariate polynomial regression; if there are multiple independent variables, it is called multivariate polynomial regression. The mathematical expression of the polynomial regression model is: y = b o +b1X+b2X2+b3X3+...+b n Xn + ε, where b o , b1, b2, b3, ... b n are the coefficients of the polynomial regression model, corresponding to the constant term X raised to the first power, X raised to the second power, and so on, up to X raised to the nth power. ε is the error term, which represents the factors that cannot be fully explained by the model. By introducing the error term ε, we can identify the factors that cannot be explained by the model, and estimate the size and impact of these errors through statistical analysis.

[0060] Multiple regression and polynomial regression models allow for the consideration of multiple independent variables and their complex combinations, thereby improving the model's fitting ability and predictive performance. Technically, this enables the model to more accurately predict changes in dependent variables when processing multidimensional data, especially in high-dimensional space.

[0061] In this embodiment, the step (4) of verifying the polynomial regression model by using the remaining data and calculating the model's goodness of fit index is as follows:

[0062] The remaining data were used to verify the initial regression model, using the formula Calculate the coefficient of determination R of the model 2 Value, evaluate the model's fit and prediction accuracy, where y is the actual value, is the predicted value, is the average of the actual values, and n is the number of samples. 2 It is an important indicator to measure the degree of fit of the regression model. 2 , which can quantify the ability of the model to explain the changes of the independent variables on the dependent variable. 2 The closer the value is to 1, the better the model fits the data, which is of great significance for evaluating the performance of the model and can be used to estimate the model performance. 2 The calculation of R can identify the model fit. 2 A low value indicates that the model may not be sufficient to explain all the variation in the data. Based on this, technicians can adjust the complexity of the model to achieve the best fit and prediction effect. 2 The technician can get feedback on the model performance by calculating the R value. 2 If the value is not ideal, the model parameters can be adjusted or a different model form can be selected to form a dynamic model optimization mechanism.

[0063] In this embodiment, the step of optimizing and adjusting the parameters of the regression model according to the model verification result in step (5) is:

[0064] By formula Adjust the weights of the independent variables and add new explanatory variables to find the optimal combination of model parameters, where the squared error term Represents the error between the model prediction value and the actual value. By minimizing this error, the optimal combination of weights and explanatory variable coefficients can be found. (i) is the actual value of the i-th sample, x (i) is the independent variable vector of the i-th sample, is the inner product of the independent variable vector and the weight vector, representing the linear combination of the independent variables, Z (i)is the newly introduced explanatory variable, which is a combination of existing independent variables or external additional information. α is the weight coefficient of the new explanatory variable, which indicates the impact of the new variable on the output. is the L2 regularization term, which is used to control the complexity of the model. λ is the regularization strength coefficient, which controls the influence of regularization. is the L2 norm of the weight vector, that is, the sum of the squares of the weight vector. By optimizing the weights of the independent variables, the impact of each variable on the model output can be adjusted according to the actual situation of the data. This flexibility ensures that the model can accurately reflect the actual contribution of different independent variables and improve the prediction performance. Through global search by simulated annealing algorithm, the optimal solution can be found in high-dimensional parameter space. This is especially important for regression models with complex structures and can effectively improve the performance of the final model. By adding new explanatory variables, the expressive power of the model can be expanded, enabling it to capture more data features and potential patterns. New explanatory variables can be a combination of existing independent variables or additional external information, which enables the model to better adapt to complex data environments.

[0065] In this embodiment, in step (6), the optimized model is revalidated and the robustness and generalization ability of the model are further validated using the K-fold cross-validation method:

[0066] After parameter optimization, the remaining data is used to verify the optimized model again and compare the model goodness of fit and prediction accuracy before and after optimization to evaluate the effect of parameter optimization; by re-verifying the optimized model, the actual effect of the optimization measures can be evaluated. If the model performance is not ideal, continue to adjust the parameters. This iterative optimization process ensures that the model can gradually approach the optimal state and improve the overall prediction performance.

[0067] If the model verification results are still not ideal, continue to adjust the parameters until they meet the requirements;

[0068] After the model validation results are satisfied, the data is initially sampled and divided into K sub-samples. A single sub-sample is retained as the data for the validation model, and the other K-1 samples are used for training and cross-validation repeated K times. Finally, each sub-sample is validated once and the average of the results of K times is finally obtained to obtain a single estimate. Different subsets are used for modeling and validation in each validation. By dividing the data set into K sub-samples, K-1 sub-samples are used for training each time, and the remaining 1 sub-sample is used for validation. This method can maximize the use of data, reduce the deviation in model evaluation, and improve the reliability of the evaluation results. Through K-fold cross-validation, the model is trained and validated on different sub-samples, which can effectively prevent model overfitting, that is, the model overfits the training data and performs poorly on new data. This method improves the generalization ability of the model and makes its prediction performance on unseen data better.

[0069] In this embodiment, after evaluating the anti-seepage effect, the following steps are also included:

[0070] Select the ARIMA model based on the data characteristics and use the training set to train the time series model; select the ARIMA model based on the data characteristics to ensure that the time series model can effectively capture the trend, seasonality and random fluctuations in the data. This adaptability of model selection makes the prediction results more accurate and reliable.

[0071] New monitoring data is input into the trained time series model, the model prediction results are dynamically updated, and based on the current time series data and model, short-term and long-term prediction values ​​of permeability coefficient, flow rate and effective time of anti-seepage are generated; by dynamically inputting new monitoring data into the trained time series model, the model can update the prediction results in real time. This dynamic update mechanism enables the model to respond quickly to data changes and provide more timely and accurate predictions.

[0072] Based on the model prediction results, possible future change trends are analyzed to identify potential leakage risks; by analyzing the trends predicted by the model, possible future changes and potential leakage risks can be identified. This combination of prediction and analysis provides a strong basis for decision-making and helps prevent potential problems in advance.

[0073] According to the trend prediction results, we provide suggestions for adjusting anti-seepage measures and set warning thresholds. When the predicted permeability coefficient, flow rate or effective anti-seepage time exceeds the warning threshold, the warning mechanism is triggered. When the predicted key parameters reach a certain level, the warning mechanism is automatically triggered. This early warning can take action before problems occur, reducing losses and risks.

[0074] In this embodiment, when the prediction results show an obvious risk trend, an early warning signal is automatically issued, and the prediction results are sent to the staff through a communication signal. By sending early warning signals in a timely manner, the safety protection capability of the overall system is improved and the probability of accidents is reduced.

[0075] The present invention is not limited to the above-mentioned embodiments. Anyone should be aware that any structural changes made under the enlightenment of the present invention, and any technical solutions that are the same or similar to the present invention, fall within the protection scope of the present invention. The technology, shape, and structural parts not described in detail in the present invention are all well-known technologies.

Claims

1. A method for evaluating the anti-seepage effect of self-expanding particles based on multiple regression analysis, characterized in that: The following steps are involved: Step (1): determining relevant indicators for evaluating the anti-seepage effect of self-expanding particles; Step (2): collecting test data of self-expanding particles and dividing the data into a training set and a validation set; Step (3): constructing a polynomial regression model by taking the permeability coefficient, flow rate and anti-seepage effective time as independent variables and the anti-seepage effect of the self-swelling particles as the dependent variable; Step (4): verifying the polynomial regression model using the remaining data and calculating the model's goodness of fit index; Step (5): Optimize and adjust the parameters of the regression model according to the model verification results; Step (6): Revalidate the optimized model and use the K-fold cross-validation method to further verify the robustness and generalization ability of the model; Step (7): Use the optimized multivariate regression model to evaluate the anti-seepage effect of the self-expanding particles.

2. The method for evaluating the anti-seepage effect of self-expanding particles based on multiple regression analysis according to claim 1, characterized in that: The steps for collecting the test data of the self-expanding particles in step (2) are: The permeability coefficient of self-expanding particles was calculated by borehole water injection test; The flow rate data is obtained through permeability tests, and the penetration time is controlled to obtain the test data of the effective anti-seepage time. The collected test data are integrated to obtain a data set, and the obtained data set is divided into a training set and a test set.

3. The method for evaluating the anti-seepage effect of self-expanding particles based on multiple regression analysis according to claim 1, characterized in that: In step (3), the permeability coefficient, flow rate and anti-seepage effective time are used as independent variables, and the anti-seepage effect of the self-expanding particles is used as the dependent variable. The steps of constructing a polynomial regression model are as follows: Let the independent variable be X, the dependent variable be Y, and the k independent variables that affect the dependent variable are X1, X2, ...X k If there is a certain correlation between y and x, a multiple regression model is established based on the dependent variable y and each influencing factor. Its expression is: Y = β0 + β1X1 + β2X2 + ... + β K X K +ε, where, β0, β1, β2,...β K is the regression parameter; When there is only one independent variable, it is called univariate polynomial regression; if there are multiple independent variables, it is called multivariate polynomial regression. The mathematical expression of the polynomial regression model is: y = b o +b1X+b2X 2 +b3X 3 +…+b n X n +ε, where b o , b1, b2, b3, … b n are the coefficients of the polynomial regression model, corresponding to the constant term X raised to the first power, X raised to the second power, and so on, up to X raised to the nth power. ε is the error term, which indicates the factors that cannot be fully explained by the model.

4. The method for evaluating the anti-seepage effect of self-expanding particles based on multiple regression analysis according to claim 1, characterized in that: The step of verifying the polynomial regression model by using the remaining data in step (4) and calculating the goodness-of-fit index of the model: The remaining data were used to verify the initial regression model, using the formula Calculate the coefficient of determination R2 of the model to evaluate the model's fit and prediction accuracy, where y is the actual value, is the predicted value, is the mean of the actual values, and n is the number of samples.

5. The method for evaluating the anti-seepage effect of self-expanding particles based on multiple regression analysis according to claim 1, characterized in that: The step of optimizing and adjusting the parameters of the regression model according to the model verification results in step (5) is as follows: By formula Adjust the weights of the independent variables and add new explanatory variables to find the optimal combination of model parameters, where the squared error term Represents the error between the model prediction value and the actual value, y (i) is the actual value of the i-th sample, x (i) is the independent variable vector of the i-th sample, is the inner product of the independent variable vector and the weight vector, representing the linear combination of the independent variables, Z (i) is the newly introduced explanatory variable, α is the weight coefficient of the new explanatory variable, which indicates the impact of the new variable on the output. is the L2 regularization term, which is used to control the complexity of the model, and λ is the regularization strength coefficient, which controls the influence of regularization. is the L2 norm of the weight vector, that is, the sum of the squares of the weight vector.

6. The method for evaluating the anti-seepage effect of self-expanding particles based on multiple regression analysis according to claim 1, characterized in that: In step (6), the optimized model is revalidated and the K-fold cross-validation method is used to further verify the robustness and generalization ability of the model: After parameter optimization, the remaining data is used again to verify the optimized model and compare the model goodness of fit and prediction accuracy before and after optimization to evaluate the effect of parameter optimization; If the model verification results are still not ideal, continue to adjust the parameters until they meet the requirements; After the model validation results are satisfactory, the data is initially sampled and divided into K sub-samples. A single sub-sample is retained as data for the validation model, and the other K-1 samples are used for training and cross-validation repeated K times. Finally, each sub-sample is validated once and the results are averaged K times to obtain a single estimate. Different subsets are used for modeling and validation in each validation.

7. The method for evaluating the anti-seepage effect of self-expanding particles based on multiple regression analysis according to claim 1, characterized in that: After evaluating the anti-seepage effect, the following steps are also included: Select the ARIMA model based on the data characteristics and use the training set to train the time series model; Input new monitoring data into the trained time series model, dynamically update the model prediction results and generate short-term and long-term prediction values ​​of permeability coefficient, flow rate and effective time of anti-seepage based on the current time series data and model; Analyze future change trends and identify potential leakage risks based on model prediction results; Based on the trend prediction results, suggestions for adjusting anti-seepage measures are provided, and an early warning threshold is set. When the predicted permeability coefficient, flow rate or effective anti-seepage time exceeds the early warning threshold, the early warning mechanism is triggered.

8. The method for evaluating the anti-seepage effect of self-expanding particles based on multiple regression analysis according to claim 7, characterized in that: When the forecast results show a clear risk trend, an early warning signal is automatically issued and the forecast results are sent to the staff via communication signals.

Images