Small reservoir seepage data intelligent analysis method based on rainfall characteristics

Abstract

The invention discloses intelligent analysis of seepage data of a small reservoir based on rainfall characteristics, which comprises the following steps of: generating corresponding characteristics for seepage pressure data, rainfall capacity and reservoir water level through characteristic construction, and normalizing the data by adopting a Robust Scaler method; a long-short-term memory network model and a random forest model based on historical rainfall are constructed, and model parameters are adjusted and optimized in combination with Bayesian optimization and a grid search technology; systematically analyzing the real water level of the piezometric tube, filtering the potential influence of the water level of the reservoir on the water level of the piezometric tube, comparing the influence of the rainfall in the cumulative time period and the historical rainfall on the water level of the piezometric tube, and drawing a scatter diagram between the two by using a visual tool. The method is particularly suitable for intelligent analysis of seepage data of a small reservoir, and the intelligent level of reservoir safety monitoring and management can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent processing and diagnosis of seepage pressure data of reservoir dams, and specifically aims at the intelligent analysis of seepage data of small reservoirs based on rainfall characteristics. Background Art

[0002] As is well known, reservoir dams, as an important part of water conservancy infrastructure, play a key role in resisting heavy rainfall, regulating water volume and power generation. Their safe operation during rainfall is particularly important, which is directly related to the social and economic stability of the downstream area. With the increase in extreme weather events, the impact of rainfall on the seepage pressure of dams has become increasingly significant. Especially during heavy rainfall or continuous rainfall, the seepage pressure of the dam body often shows abnormal fluctuations due to the dynamic changes in the internal seepage state of the dam.

[0003] However, the existing technologies mainly rely on seepage pressure monitoring methods under static conditions. Currently, the calculation methods for seepage pressure thresholds under the influence of rainfall mainly adopt numerical integration methods and iterative calculation methods. However, these methods have obvious limitations in practical applications. The numerical integration method usually needs to simplify the rainfall process and is difficult to accurately capture the complex influence of dynamic changes such as rainfall intensity and frequency on the seepage pressure of the dam. The iterative calculation method has limited accuracy when dealing with complex terrain and non-linear material properties, and its response adjustment is not flexible enough under the condition of rainfall mutation. Since these methods cannot fully reflect the actual change law of seepage pressure under dynamic rainfall conditions, it is often difficult to provide accurate real-time seepage pressure prediction, reducing the reliability of dam monitoring and early warning.

[0004] In view of the above challenges, there is an urgent need to develop a method for processing and predicting seepage pressure data of small reservoirs based on rainfall characteristics to more accurately capture and analyze seepage pressure data under rainfall conditions, thereby improving the accuracy of monitoring and early warning capabilities. Summary of the Invention

[0005] The present invention adopts feature engineering, RobustScaler normalization, LSTM technology and random forest model to provide an efficient and accurate seepage pressure data processing solution to meet the safety monitoring requirements of reservoir dams under rainfall conditions and ensure the safe and stable operation of the dams.

[0006] The technical solution adopted by the present invention to solve its technical problems is:

[0007] Intelligent analysis of seepage data of small reservoirs based on rainfall characteristics includes the following steps:

[0008] Step S1: Generate corresponding features for seepage pressure data, rainfall amount and reservoir water level through feature construction, and perform normalization processing on these features using the RobustScaler method;

[0009] Step S2: Construct the input and output sequences of the LSTM model using historical cumulative rainfall data, determine the number of layers and the number of neurons in each layer of the model, and select the loss function and optimizer;

[0010] Step S3: Adjust the time window length, batch size, and learning rate, optimize the model parameters through the Bayesian network, and determine the LSTM model to predict the impact of historical rainfall on the piezometer level;

[0011] Step S4: Based on the cumulative rainfall during this period, construct the input features and target variables of the random forest model, select the number of trees and the maximum depth, and determine the loss function and evaluation criteria;

[0012] Step S5: Adjust the number of trees and the minimum sample split number, optimize the parameters through grid search, and determine the random forest model to predict the impact of cumulative period rainfall on the piezometer level;

[0013] Step S6: Conduct a systematic analysis of the true piezometer level, filter out the potential impact of the reservoir level on the piezometer level, compare the impacts of cumulative period rainfall and historical rainfall on the piezometer level, and use visualization tools to draw a scatter plot between the two.

[0014] Further technology of the present invention:

[0015] Preferably, in step S2, the number of neurons in each layer is selected as 2 to 10 times the input features.

[0016] Preferably, step S3 specifically:

[0017] Step S31: Select the number of layers of the LSTM and the number of neurons in each layer, and determine the loss function and optimizer of the model to construct a preliminary model, and then set the time window length, batch size, and learning rate;

[0018] Step S32: Start model training using the training data, optimize the hyperparameters through the Bayesian network or other optimization methods, continuously adjust parameters such as the time window, batch size, and learning rate to minimize the loss function; after each iteration, check the change in the loss function. If the change in the loss function is less than the preset threshold, terminate the training loop, determine the final LSTM model to predict the impact of historical rainfall on the piezometer level, and obtain the final prediction result.

[0019] Preferably, the specific use of grid search for hyperparameter optimization in step S5:

[0020] Traverse possible parameter combinations through grid search, systematically evaluate the performance of the model under different parameters, and select the best combination of the number of trees, minimum sample split, and other hyperparameters; in this process, gradually adjust the number of trees and other hyperparameters through loops; when the number of trees reaches the preset upper limit, break out of the loop and stop further increasing the number of trees.

[0021] Preferably, in step S6, a scatter plot is drawn using a visualization tool, with the cumulative period rainfall and historical rainfall as independent variables and the piezometer level as the dependent variable.

[0022] Advantages of the present invention:

[0023] The RobustScaler method is used to normalize the data, effectively improving the model training efficiency; by constructing a long short-term memory network model and a random forest model based on historical rainfall, and combining Bayesian optimization and grid search techniques to optimize the model parameters, the prediction accuracy is significantly improved; after further analyzing and filtering out the influence of reservoir water level on the piezometer level, combined with rainfall data, the influence on the piezometer level is deeply studied; the invention is particularly applicable to the intelligent analysis of seepage data of small reservoirs, and can improve the intelligent level of reservoir safety monitoring and management. Brief Description of the Drawings

[0024] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0025] Figure 1 It is a schematic flow chart of the intelligent analysis research on seepage data of small reservoirs based on rainfall characteristics of the present invention. Detailed Embodiments

[0026] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will further elaborate on the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are some but not all of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0027] Combined with Figure 1 As shown, the present invention provides an intelligent analysis of seepage data of small reservoirs based on rainfall characteristics, including:

[0028] Step S1: Generate corresponding features for seepage pressure data, rainfall, and reservoir water level through feature construction, and normalize these features using the RobustScaler method;

[0029] Step S11: Divide the preprocessed data into a training set and a test set in proportion. Take 80% of the valid dataset as the training set and the remaining 20% as the test set, and then input the data into the prediction model.

[0030] Step S12: When constructing features for seepage pressure data, rainfall, and reservoir water level, generate various feature variables, including time series features (such as lag features, rolling window features, time period features, etc.), statistical features (mean, median, variance, etc.), cumulative features, and exponentially weighted moving average and other features.

[0031] Step S13: Apply the RobustScaler method to normalize the constructed feature variables, effectively reducing the impact of outliers on the data distribution, enabling the data to have good outlier resistance, and thus improving the robustness and accuracy of the model.

[0032] Step S2: Use historical cumulative rainfall data to construct the input and output sequences of the LSTM model, determine the number of layers of the model and the number of neurons in each layer, and select an appropriate loss function and optimizer;

[0033] Step S21: Use the LSTM model to capture the lag effect of past rainfall on the current piezometric level, that is

[0034]

[0035] where R t-1 , R t-2 , …, R t-k represent the rainfall in past periods, represents the lag effect of past rainfall on the current piezometric level. The LSTM processes the time series dependence through its internal memory units and gating mechanisms (such as forget gate, input gate, and output gate), as shown below:

[0036] Forget Gate: Determines which past input information needs to be "forgotten".

[0037] f t = σ(W f · [h t-1 , R t + b f )

[0038] where, f t is the output of the LSTM forget gate at time step t, ht-1 is the hidden state at the previous moment, W f is the weight matrix of the forget gate, R t is the input (rainfall) at the current moment, b f is the forget gate bias.

[0039] Input Gate: Controls which information needs to be updated into the memory cell.

[0040] i t = σ(W f · [h t-1 , R t + b i )

[0041] where i t is the output of the LSTM input gate at time step t, b i is the input gate bias.

[0042] Output Gate: Determines which information is output from the memory cell to generate the current hidden state.

[0043] o t = σ(W f · [h t-1 , R t + b o )

[0044] where o t is the output of the LSTM output gate at time step t, b o is the output gate bias.

[0045] Step S22, Determine the structure and hyperparameter configuration of the LSTM model. Select an appropriate number of LSTM layers according to the complexity of the data. The number of neurons in each layer needs to be adjusted according to the task requirements, usually choosing 2 to 10 times the input features. Too many layers may lead to overfitting. Then, select an appropriate loss function to measure the deviation between the model prediction value and the actual value. Finally, select the Adam optimizer to improve the training efficiency and convergence speed with the advantages of adaptive learning rate and momentum. By reasonably configuring the model structure and hyperparameters, ensure that the LSTM model can effectively process historical rainfall data and make accurate predictions.

[0046] Step S3: Adjust the time window length, batch size, and learning rate, and optimize the model parameters through the Bayesian network to improve the prediction accuracy;

[0047] Step S31, Select the number of layers of the LSTM and the number of neurons in each layer, and determine the loss function and optimizer of the model to construct a preliminary model. Then set the key hyperparameters such as the time window length, batch size, and learning rate.

[0048] Step S32: Start model training using the training data. Optimize the hyperparameters through a Bayesian network or other optimization methods, and continuously adjust parameters such as the time window, batch size, and learning rate to minimize the loss function. After each iteration, check the change in the loss function. If the change in the loss function is less than the preset threshold, terminate the training loop, determine the final LSTM model to predict the impact of historical rainfall on the piezometer level, and obtain the final prediction result.

[0049] Step S4: Based on the cumulative rainfall during this period, construct the input features and target variables of the random forest model, select the number of trees and the maximum depth, and determine the loss function and evaluation criteria;

[0050] Step S41: Capture the impact of cumulative rainfall of different magnitudes on the piezometer level through a random forest. The regression prediction formula of the random forest can be expressed as:

[0051]

[0052] where f RF (R t ) is the predicted value of the random forest model at the current rainfall R t , representing the direct impact of the current rainfall on the piezometer level. T m (R t ) is the output of the m-th decision tree based on the current rainfall R t , and M is the number of decision trees in the forest. For each decision tree, each tree learns the relationship between rainfall and piezometer level and continuously bisects the data to reduce the prediction error in each leaf node. The model splits the data through rainfall R t , with the goal of minimizing the error between the residual ΔW i and the output of the tree.

[0053] Step S42: Select appropriate random forest model hyperparameters according to the data characteristics. First, determine the number of trees in the random forest. A larger number of trees usually improves the stability and accuracy of the model but increases the computational cost. Then, set the maximum depth of each tree to prevent overfitting and ensure the generalization ability of the model. Next, select the mean squared error as the loss function to measure the prediction accuracy of the model.

[0054] Step S5: Adjust the number of trees and the minimum sample split number, and optimize the parameters through grid search to optimize the model performance and improve the prediction accuracy;

[0055] Step S51: Adjust the number of trees and the minimum sample split number. First, according to the scale of the dataset and the task complexity, set the initial number of trees and the minimum sample split number. The number of trees determines the scale of the forest. A larger number of trees can improve the accuracy of the model, while the minimum sample split number controls the minimum number of samples at each node, avoiding overfitting and improving the generalization ability of the model. By adjusting these hyperparameters, find a suitable combination to enhance the stability and performance of the model.

[0056] Step S52: Use grid search for hyperparameter optimization. Through grid search, traverse possible parameter combinations, systematically evaluate the performance of the model under different parameters, and select the best combination of the number of trees, the minimum sample split number, and other hyperparameters. During this process, gradually adjust the number of trees and other hyperparameters through loops. When the number of trees reaches the preset upper limit, break out of the loop and stop further increasing the number of trees. This can avoid excessive trees increasing the computational burden and ensure the balance between model training efficiency and prediction accuracy.

[0057] Step S6: Conduct a systematic analysis of the actual piezometric tube water level to filter out the potential influence of the reservoir water level on the piezometric tube water level. On this basis, compare the influence of the rainfall during the period and the historical rainfall on the piezometric tube water level. Then, use a visualization tool (Matplotlib) to draw a scatter plot between the two.

[0058] Step S61: Conduct a systematic analysis of the actual piezometric tube water level. First, filter out the potential influence of the reservoir water level on the piezometric tube water level through the SVR model. Through this process, ensure that the measured changes in the piezometric tube water level can accurately reflect the influence of factors such as rainfall, thereby improving the accuracy of the data and providing a cleaner dataset for subsequent analysis.

[0059] Step S62: After filtering out the influence of the reservoir water level, further analyze the influence of the rainfall during the period and the historical rainfall on the piezometric tube water level. Use a visualization tool to draw a scatter plot, taking the rainfall during the period and the historical rainfall as independent variables and the piezometric tube water level as the dependent variable to show their relationship. And through the analysis of the comparison chart between the piezometric tube water level after filtering out the reservoir water level, the cumulative rainfall, and the influence of the historical rainfall on the piezometric tube water level. The specific influence and accuracy of the rainfall on the change of the piezometric tube water level can be clearly seen, which helps to better understand the correlation between the two.

[0060] The above are only the preferred embodiments of the present invention and do not impose any form of limitation on the present invention. Any person skilled in the art can, without departing from the scope of the technical solution of the present invention, make many possible changes and modifications to the technical solution of the present invention by using the methods and technical contents disclosed above, or modify it into equivalent embodiments with equivalent changes. Therefore, any simple modification, equivalent replacement, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention still fall within the scope of protection of the technical solution of the present invention.

Claims

1. Intelligent analysis of seepage data of small reservoirs based on rainfall characteristics, characterized in that It includes the following steps: Step S1: Generate corresponding features for seepage pressure data, rainfall, and reservoir water level through feature construction, and normalize these features using the RobustScaler method; Step S2: Use historical cumulative rainfall data to construct the input and output sequences of the LSTM model, determine the number of layers and the number of neurons in each layer of the model, and select the loss function and optimizer; Step S3: Adjust the time window length, batch size, and learning rate, optimize the model parameters through the Bayesian network, and determine the LSTM model to predict the impact of historical rainfall on the piezometric tube water level; Step S4: Based on the cumulative rainfall during this period, construct the input features and target variables of the random forest model, select the number of trees and the maximum depth, and determine the loss function and evaluation criteria; Step S5: Adjust the number of trees and the minimum sample split number, optimize the parameters through grid search, and determine the random forest model to predict the impact of cumulative period rainfall on the piezometric tube water level; Step S6: Conduct a systematic analysis of the actual piezometric tube water level, filter out the potential impact of the reservoir water level on the piezometric tube water level, compare the impact of cumulative period rainfall and historical rainfall on the piezometric tube water level, and use visualization tools to draw a scatter plot between the two.

2. The intelligent analysis of the seepage data of the small reservoir based on rainfall characteristics according to claim 1, characterized in that In Step S2, the number of neurons in each layer is selected as 2 to 10 times the input features.

3. The intelligent analysis of the seepage data of the small reservoir based on rainfall characteristics according to claim 1, characterized in that, Specifically for Step S3: Step S31: Select the number of layers of LSTM and the number of neurons in each layer, and determine the loss function and optimizer of the model to construct a preliminary model, and then set the time window length, batch size, and learning rate; Step S32: Start model training using the training data, optimize the hyperparameters through the Bayesian network or other optimization methods, continuously adjust parameters such as the time window, batch size, and learning rate to minimize the loss function; after each iteration, check the change in the loss function. If the change in the loss function is less than the preset threshold, terminate the training loop, determine the final LSTM model to predict the impact of historical rainfall on the piezometric tube water level, and obtain the final prediction result.

4. The intelligent analysis of the seepage data of the small reservoir based on rainfall characteristics according to claim 1, characterized in that, Specifically for using grid search for hyperparameter optimization in Step S5: Traverse possible parameter combinations through grid search, systematically evaluate the performance of the model under different parameters, and select the best combination of the number of trees, minimum sample split number, and other hyperparameters; in this process, gradually adjust the number of trees and other hyperparameters through a loop; when the number of trees reaches the preset upper limit, jump out of the loop and stop further increasing the number of trees.

5. The intelligent analysis of the seepage data of the small reservoir based on rainfall characteristics according to claim 1 is characterized in that, In Step S6, use visualization tools to draw a scatter plot, taking the cumulative period rainfall and historical rainfall as independent variables and the piezometric tube water level as the dependent variable.

Images