Dam deformation intelligent prediction method and system and storage medium
Through the DenseNet-BiGRU-SGWO model combined with dense connection convolutional network and bidirectional gated loop unit, the hyperparameter configuration is optimized, which solves the problem of insufficient deep learning application in dam deformation prediction, realizes high-precision deformation prediction, and improves the large-scale Reliability of dam safety monitoring.
Patent Information
- Application Number
- CN202411190679.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-27
- Publication Date
- 2024-12-06
AI Technical Summary
The existing dam deformation prediction methods are insufficient when using deep learning, and it is difficult to accurately predict the deformation trend of the dam. The prediction accuracy of the traditional methods is not high, making it difficult to meet the needs of dam safety monitoring.
The DenseNet-BiGRU-SGWO model is adopted, combining densely connected convolutional networks and bidirectional gated loop units, and optimizes hyperparameter configuration through the purified gray wolf intelligent algorithm to build a dam deformation intelligent prediction model for monitoring data training and prediction.
The accuracy and model performance of dam deformation prediction are significantly improved, and the advantages and disadvantages of the prediction effect are verified through multiple evaluation indicators, providing more reliable safety monitoring support.
Smart Images

Figure CN119089939A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of dam deformation monitoring, and in particular to a dam deformation intelligent prediction method, system and storage medium. Background Art
[0002] As an important basic livelihood facility in water conservancy projects, reservoir dams will seriously endanger the safety of life and property of people in the reservoir area once they fail. Dam displacement is the most intuitive and reliable monitoring indicator in monitoring data, which can reflect the comprehensive operation status of the dam. Fortunately, dam failure is a gradual process. If the abnormal state of the dam body can be discovered in advance, the disaster can be avoided and the dam body can be repaired in a timely manner. Therefore, it is an indispensable part of the safety monitoring of water conservancy projects to reasonably analyze the long-term monitoring data of the dam and carry out deformation prediction research. In engineering, dam deformation prediction and analysis are generally carried out based on physical analysis methods such as finite elements and data analysis methods such as intelligent algorithms. However, since the physical analysis method requires the simplification of materials and boundary conditions, the uncertainty is high, the calculation is time-consuming and labor-intensive, and it is difficult to efficiently and accurately predict the deformation trend of the dam; although the traditional machine learning method has a simple structure and clear principles, the prediction accuracy is not high and the fitting effect is poor; the deep learning method can better handle long-term time dependence, but in practice, there is a problem of insufficient application of deep learning for dam deformation prediction. Therefore, for those skilled in the art, how to build a deep learning model for dam deformation prediction is an urgent problem to be solved. Summary of the invention
[0003] The purpose of the present invention is to provide a dam deformation intelligent prediction method, system and storage medium to solve the problems raised in the background technology, realize dam deformation prediction analysis by establishing a DenseNet-BiGRU-SGWO model, and provide method support for dam safety monitoring.
[0004] To achieve the above object, the present invention provides the following solution: a method for intelligent prediction of dam deformation, the specific steps of which include the following:
[0005] Obtain monitoring data on dam horizontal displacement and environmental quantities;
[0006] Preprocessing the monitoring data to obtain first data and dividing the first data into a training set and a test set;
[0007] A dam deformation intelligent prediction model is constructed based on a densely connected convolutional network and a bidirectional gated recurrent unit, and the optimal hyperparameter configuration of the dam deformation intelligent prediction model is determined by a purified gray wolf intelligent algorithm;
[0008] The intelligent prediction model for dam deformation is trained using the training set to obtain a trained prediction model, and dam deformation prediction is performed based on the trained prediction model.
[0009] Preferably, the environmental quantities include upstream water level and daily average temperature.
[0010] Preferably, the specific steps of preprocessing the monitoring data are:
[0011] Performing interpolation processing on the monitoring data by using a cubic spline interpolation method;
[0012] Normalizing the interpolated monitoring data to obtain the first data;
[0013] Preferably, the calculation formula for performing the normalization process is:
[0014]
[0015] Among them, x represents the monitoring data obtained after interpolation, x norm represents the first data obtained after normalization, x max Indicates the maximum value of the input variable, x min Indicates the minimum value of the input variable.
[0016] Preferably, the first data is divided into a training set and a test set in a ratio of 8:2.
[0017] On the other hand, a dam deformation intelligent prediction system is provided, including a data acquisition module, a preprocessing module, a modeling module, and a prediction module; wherein:
[0018] The data acquisition module is used to acquire monitoring data of the dam's horizontal displacement and environmental quantities;
[0019] The preprocessing module is used to preprocess the monitoring data to obtain first data and divide the first data into a training set and a test set;
[0020] The modeling module is used to construct a dam deformation intelligent prediction model based on a densely connected convolutional network and a bidirectional gated recurrent unit, and determine the optimal hyperparameter configuration of the dam deformation intelligent prediction model through a purified gray wolf intelligent algorithm;
[0021] The prediction module is used to train the dam deformation intelligent prediction model using the training set to obtain a trained prediction model, and perform dam deformation prediction based on the trained prediction model.
[0022] Finally, a computer storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the method for intelligent prediction of dam deformation are implemented.
[0023] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects: a DenseNet-BiGRU-SGWO model is constructed based on a densely connected convolutional network and a bidirectional gated recurrent unit as an intelligent prediction model for dam deformation, thereby realizing dam deformation prediction analysis; by combining multiple deep learning modules and introducing an optimization algorithm to optimize hyperparameters, the model performance is further improved. BRIEF DESCRIPTION OF THE DRAWINGS
[0024] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.
[0025] Figure 1 is a flow chart of the method of the present invention;
[0026] Figure 2 It is the long-term monitoring data diagram of horizontal displacement of the present invention;
[0027] Figure 3 It is the long-term monitoring data diagram of the environmental quantity of the present invention;
[0028] Figure 4 This is a structural diagram of the dam deformation intelligent prediction model of the present invention.
[0029] Figure 5 This is a prediction curve diagram of the IP1 measurement point predicted by the DenseNet-BiGRU-SGWO model of the present invention;
[0030] Figure 6 This is a prediction curve diagram of IP2 measurement points predicted by the DenseNet-BiGRU-SGWO model of the present invention;
[0031] Figure 7 This is a prediction curve diagram of the PL2 measurement point predicted by the DenseNet-BiGRU-SGWO model of the present invention;
[0032] Figure 8 This is a prediction curve diagram of IP3 measurement points predicted by using the DenseNet-BiGRU-SGWO model of the present invention. DETAILED DESCRIPTION
[0033] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0034] The purpose of this invention is to provide a method for intelligent prediction of dam deformation. Figure 1 As shown, the specific steps include the following:
[0035] S1. Obtain monitoring data of dam horizontal displacement and environmental quantities;
[0036] S2. Preprocess the monitoring data to obtain first data and divide the first data into a training set and a test set;
[0037] S3. Based on densely connected convolutional networks and bidirectional gated recurrent units, an intelligent prediction model for dam deformation is constructed, and the optimal hyperparameter configuration of the intelligent prediction model for dam deformation is determined by the purified gray wolf intelligent algorithm;
[0038] S4. Use the training set to train the dam deformation intelligent prediction model to obtain a trained prediction model, and perform dam deformation prediction based on the trained prediction model.
[0039] In step S1, according to the automatic safety monitoring system deployed on the dam body, the long-term monitoring data of the horizontal displacement of four vertical measuring points (IP1, IP2, PL2 and IP3) and environmental quantities (upstream water level, daily average temperature) are obtained.
[0040] Furthermore, the specific steps of preprocessing the monitoring data in step S2 are:
[0041] S21. Interpolate the monitoring data using the cubic spline interpolation method to fill in the missing values and obtain the interpolation results once a day, such as Figure 2 and Figure 3 As shown;
[0042] S22, normalizing the interpolated monitoring data to obtain first data;
[0043] The calculation formula for normalization is:
[0044]
[0045] Among them, x represents the monitoring data obtained after interpolation, x norm represents the first data obtained after normalization, x max Indicates the maximum value of the input variable, x minIndicates the minimum value of the input variable.
[0046] The first data is divided into a training set and a test set in a ratio of 8:2.
[0047] Further, such as Figure 4 As shown, in step S3, based on two deep learning neural network structures, densely connected convolutional network (DenseNet) and bidirectional gated recurrent unit (BiGRU), the excellent information extraction ability of the former and the powerful sequence data processing ability of the latter are fully utilized to build an intelligent prediction model for dam deformation.
[0048] The monitoring data is input into the densely connected convolutional network, and the spatial features in the monitoring data are captured through dense connections between layers to ensure that the model can fully utilize the spatial correlation of the monitoring data. The features extracted by the densely connected convolutional network are input into the bidirectional gated recurrent unit. Using its bidirectional structure, the data is processed simultaneously from both the forward and backward directions to fully grasp the dynamic changes in the time series. During the operation of the dam deformation intelligent prediction model, the purification gray wolf intelligent algorithm simulates the hunting behavior of gray wolf groups, performs global search and local optimization, and determines the optimal hyperparameter configuration of the dam deformation intelligent prediction model.
[0049] In order to verify the prediction effect of the dam deformation intelligent prediction model in the present invention, multiple models are used for comparative study, and the prediction performance of different models is evaluated based on the three evaluation indicators of R, MAE and RMSE.
[0050] In this embodiment, four comparison models are selected: LSTM, SVR, ELM and MLP, and each model is trained using a training set to obtain the deformation prediction results of each model. Figure 5-8 The following is a prediction curve of the measurement points predicted by the DenseNet-BiGRU-SGWO model. The prediction performance of the DenseNet-BiGRU-SGWO model and the four comparison models are evaluated based on the R, MAE and RMSE evaluation indicators. The evaluation indicator calculation formula is as follows:
[0051]
[0052] Where n represents the total number of data samples, y i Represents the measured displacement deformation value, represents the predicted displacement deformation value, represents the mean value of the measured displacement deformation, Represents the mean of the predicted displacement deformation. These evaluation indicators are used to quantify the degree of fit and prediction accuracy of the model. The closer the R value is to 1R or the closer the MAE and RMSE are to 0, the better the model performance. The model evaluation results are shown in Table 1. It can be seen from Table 1 that the intelligent prediction model for dam deformation in the present invention has high prediction accuracy and good effect.
[0053]
[0054]
[0055] Table 1
[0056] On the other hand, a dam deformation intelligent prediction system is provided, including a data acquisition module, a preprocessing module, a modeling module, and a prediction module; wherein:
[0057] Data acquisition module, used to obtain monitoring data of dam horizontal displacement and environmental quantities;
[0058] A preprocessing module, used to preprocess the monitoring data to obtain first data and divide the first data into a training set and a test set;
[0059] A modeling module is used to build an intelligent prediction model for dam deformation based on a densely connected convolutional network and a bidirectional gated recurrent unit, and to determine the optimal hyperparameter configuration of the intelligent prediction model for dam deformation through the purified gray wolf intelligent algorithm;
[0060] The prediction module is used to train the dam deformation intelligent prediction model using the training set to obtain the trained prediction model, and perform dam deformation prediction based on the trained prediction model.
[0061] Finally, a computer storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the steps of a method for intelligently predicting dam deformation are implemented.
[0062] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of the present invention. At the same time, for those skilled in the art, according to the idea of the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.
Claims
1. A dam deformation intelligent prediction method, characterized in that: The specific steps include the following: Obtain monitoring data on dam horizontal displacement and environmental quantities; Preprocessing the monitoring data to obtain first data and dividing the first data into a training set and a test set; An intelligent prediction model for dam deformation is constructed based on densely connected convolutional networks and bidirectional gated recurrent units; The intelligent prediction model for dam deformation is trained using the training set to obtain a trained prediction model, and dam deformation prediction is performed based on the trained prediction model.
2. The intelligent prediction method for dam deformation according to claim 1 is characterized in that: The environmental quantities include upstream water level and average daily temperature.
3. The intelligent prediction method for dam deformation according to claim 1 is characterized in that: The specific steps of preprocessing the monitoring data are as follows: Performing interpolation processing on the monitoring data by using a cubic spline interpolation method; The interpolated monitoring data is normalized to obtain the first data.
4. The intelligent prediction method for dam deformation according to claim 3 is characterized in that: The calculation formula for performing the normalization process is: Among them, x represents the monitoring data obtained after interpolation, x norm represents the first data obtained after normalization, x max Indicates the maximum value of the input variable, x min Indicates the minimum value of the input variable.
5. The intelligent prediction method for dam deformation according to claim 1 is characterized in that: The first data is divided into a training set and a test set in a ratio of 8:
2.
6. A dam deformation intelligent prediction system, characterized in that: It includes data acquisition module, preprocessing module, modeling module and prediction module; among them, The data acquisition module is used to acquire monitoring data of the dam's horizontal displacement and environmental quantities; The preprocessing module is used to preprocess the monitoring data to obtain first data and divide the first data into a training set and a test set; The modeling module is used to construct a dam deformation intelligent prediction model based on a densely connected convolutional network and a bidirectional gated recurrent unit, and determine the optimal hyperparameter configuration of the dam deformation intelligent prediction model through a purified gray wolf intelligent algorithm; The prediction module is used to train the dam deformation intelligent prediction model using the training set to obtain a trained prediction model, and perform dam deformation prediction based on the trained prediction model.
7. A computer storage medium, characterized in that: The computer storage medium stores a computer program, and when the computer program is executed by the processor, the steps of the intelligent prediction method for dam deformation as described in any one of claims 1 to 5 are implemented.
Citation Information
Patent Citations
Efficient dam intelligent monitoring method and system
CN113807572A
Concrete dam deformation prediction method based on IPSO-Transform model
CN117540265A
Cited By
Dam deformation prediction method and device, storage medium and product
CN120492800A
