Arch dam optimal seismic oscillation parameter identification method and device
By performing nonlinear dynamic time-course analysis and application of SHAP theory on high arch dams, combined with multiple machine learning models, the problem of missing unified indicators in the identification of optimal earthquake parameters of high arch dams is solved, and the impact on earthquake parameters is achieved quantitative evaluation reduces the inuniqueness of identification results and the uncertainty of seismic response prediction.
Patent Information
- Application Number
- CN202411173926.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-26
- Publication Date
- 2024-11-26
AI Technical Summary
The lack of unified indicators for evaluating the optimal earthquake parameters of high arch dams in the prior art, resulting in inunique identification results and it is difficult to effectively evaluate the degree of impact of earthquake parameters on arch dam damage.
By performing multiple nonlinear dynamic time-course analysis on pre-selected earthquake recording data, a damage index prediction model is constructed, and the SHAP theory is used to determine the intensity parameter with the greatest contribution as the optimal earthquake parameter, combined with multiple machine learning models (e.g. Support vector regression, gradient lifting tree, etc.) construct a damage index prediction model to quantitatively evaluate the impact of different parameters on arch dam damage.
The problem of inuniqueness of identification results in traditional methods is overcome, and a new idea is provided for the identification of optimal earthquake parameters of arch dams, which reduces the uncertainty of earthquake response prediction, and realizes a quantitative evaluation of the impact of earthquake parameters.
Smart Images

Figure CN119026424A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of arch dam seismic performance research, and in particular to an arch dam optimal seismic parameter identification method and device. Background Art
[0002] Most high arch dams are built in the southwest of my country, due to the large number of mountains and rivers, rich hydropower resources, and great natural drop. However, the geological structure in the southwest of my country is relatively unstable, earthquakes occur frequently, and the seismic resistance of high arch dam projects is very prominent. In recent years, the structural seismic analysis and design ideas under the framework of performance-based earthquake engineering (PBEE) have provided an effective way to solve the seismic safety evaluation of high arch dams. Identifying the optimal seismic motion parameters for high arch dams is one of the important contents of the PBEE framework, but such research is still in its infancy.
[0003] At present, the research on optimal seismic motion parameter identification mainly focuses on the fields of building structures and bridges. The core idea is to analyze the functional relationship between seismic motion parameters and the response of engineering structures, determine the parameters that are most suitable for establishing earthquake demand models, and thus reduce the uncertainty of structural response prediction. The evaluation criteria commonly used in existing studies to identify the optimal seismic motion parameters of structures are effectiveness, adequacy, proficiency, and practicality, and then determine the optimal seismic motion parameters for each evaluation criterion.
[0004] However, the optimal seismic parameters identified by different evaluation criteria are different, so there is a lack of a unified indicator for evaluating the optimal seismic parameters. Summary of the invention
[0005] The present invention provides an arch dam optimal seismic parameter identification method and device, which are used to solve the defect of the prior art that there is a lack of a unified index for evaluating the optimal seismic parameter, and realize an arch dam optimal seismic parameter identification method.
[0006] The present invention provides an optimal seismic parameter identification method for an arch dam, comprising: Using the pre-selected seismic recording data as input, performing multiple nonlinear dynamic time history analyses on the pre-constructed arch dam finite element, and obtaining an arch dam damage index set corresponding to the strength parameters of the seismic recording data, wherein the arch dam damage index set includes an arch dam damage index corresponding to each of the seismic recording data; Taking the intensity parameter of each piece of seismic record data as input, taking the arch dam damage index corresponding to each piece of seismic record data as output to construct a data set, and training to obtain a first damage index prediction model; Based on the SHAP theory, the intensity parameter with the greatest contribution in the first damage index prediction model is determined as the optimal scalar ground motion intensity parameter.
[0007] According to an optimal seismic parameter identification method for an arch dam provided by the present invention, the arch dam damage index is the arch dam damage volume ratio.
[0008] According to an optimal seismic parameter identification method for an arch dam provided by the present invention, the steps of taking the intensity parameter of each seismic record data as input, taking the arch dam damage index corresponding to each seismic record data as output to construct a data set, and training to obtain a first damage index prediction model specifically include: Taking the intensity parameter of each of the seismic record data as input, taking the arch dam damage index corresponding to each of the seismic record data as output to construct a data set, and training to obtain multiple different types of candidate prediction models; The candidate prediction model with the best prediction performance is determined as the first damage indicator prediction model.
[0009] According to a method for identifying optimal seismic parameters of an arch dam provided by the present invention, the multiple different types of candidate prediction models include at least one or more of a support vector regression model, a gradient boosting tree model, an extreme gradient boosting tree model, a BP neural network model, a random forest model and an explanatory benchmark model.
[0010] According to an optimal seismic parameter identification method for an arch dam provided by the present invention, the step of determining the intensity parameter with the greatest contribution in the first damage index prediction model as the optimal scalar seismic intensity parameter based on the SHAP theory specifically includes: Based on the SHAP theory, calculating the marginal contribution value of the intensity parameter of each of the seismic record data to the output result of the first damage index prediction model; Taking a weighted average of all the marginal contribution values to obtain a SHAP value of each intensity parameter; The intensity parameter with the highest SHAP value is used as the optimal scalar ground motion intensity parameter.
[0011] According to the method for identifying the optimal seismic parameters of an arch dam provided by the present invention, after the step of determining the intensity parameter with the greatest contribution in the first damage index prediction model as the optimal scalar seismic intensity parameter, the method further includes: Combining multiple intensity parameters with the largest contributions to construct multiple vector intensity parameter combinations; Taking the vector intensity parameter combination corresponding to each of the seismic record data as input, taking the arch dam damage index corresponding to each of the seismic record data as output to construct a data set, and training to obtain a second damage index prediction model; Based on the SHAP theory, the vector intensity parameter combination with the greatest contribution in the second damage index prediction model is determined as the optimal vector seismic intensity parameter.
[0012] The present invention also provides an arch dam optimal earthquake motion parameter identification device, comprising: A determination module is used to use the pre-selected seismic recording data as input, perform multiple nonlinear dynamic time-history analyses on the pre-constructed arch dam finite element, and obtain an arch dam damage index set corresponding to the strength parameters of the seismic recording data, wherein the arch dam damage index set includes an arch dam damage index corresponding to each of the seismic recording data; A training module, used to take the intensity parameter of each of the seismic record data as input, take the arch dam damage index corresponding to each of the seismic record data as output to construct a data set, and train to obtain a first damage index prediction model; The identification module is used to determine the intensity parameter with the greatest contribution in the first damage indicator prediction model as the optimal scalar ground motion intensity parameter based on the SHAP theory.
[0013] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the method for identifying the optimal seismic parameters of an arch dam as described above is implemented.
[0014] The present invention also provides a non-transitory computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the method for identifying the optimal seismic parameters of an arch dam as described in any one of the above is implemented.
[0015] The present invention also provides a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the method for identifying optimal seismic parameters of an arch dam as described in any one of the above is implemented.
[0016] The method and device for identifying the optimal seismic motion parameters of an arch dam provided by the present invention can quantitatively evaluate the influence of different scalar earthquake intensity parameters on the arch dam damage based on the SHAP theory by training a first damage index prediction model for predicting the arch dam damage index based on the seismic motion scalar intensity parameters, thereby overcoming the problem of non-uniqueness of the identification results in traditional methods and providing a new idea for identifying the optimal seismic motion parameters of an arch dam. BRIEF DESCRIPTION OF THE DRAWINGS
[0017] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0018] Figure 1 It is a flow chart of the method for identifying the optimal seismic parameters of an arch dam provided by the present invention; Figure 2 It is a schematic diagram of moment magnitude-fault distance distribution of earthquake records selected in the arch dam optimal seismic parameter identification method provided by the present invention; Figure 3 It is a finite element model diagram of a high arch dam-reservoir water-foundation in the arch dam optimal seismic parameter identification method provided by the present invention; Figure 4 It is a data distribution diagram of the seismic intensity parameter set PGA and the corresponding arch dam damage index set in the arch dam optimal seismic parameter identification method provided by the present invention; Figure 5 The left figure in the figure is a comparison diagram of correlation coefficient results of different machine learning models in the arch dam optimal seismic parameter identification method provided by the present invention; Figure 5 The right figure in is a comparison chart of the mean absolute error results of different machine learning models in the arch dam optimal seismic parameter identification method provided by the present invention; Figure 6 It is the comparison result of the optimal machine learning model training set and test volume in the arch dam optimal seismic parameter identification method provided by the present invention; Figure 7 It is a summary diagram of the SHAP values of the scalar strength parameters in the arch dam optimal seismic parameter identification method provided by the present invention; Figure 8 It is the importance ranking of the SHAP values of the scalar strength parameters in the arch dam optimal seismic parameter identification method provided by the present invention; Fig. 9 It is a summary diagram of the SHAP values of the vector intensity parameters in the arch dam optimal seismic parameter identification method provided by the present invention; Fig.10 It is the importance ranking of the SHAP values of the vector intensity parameters in the arch dam optimal seismic parameter identification method provided by the present invention.
[0019] Fig.11 It is a structural schematic diagram of the optimal seismic parameter identification device for an arch dam provided by the present invention; Fig.12 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION
[0020] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are 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.
[0021] Combine the following Figures 1 to 10 The optimal seismic parameter identification method for arch dams of the present invention is introduced. Figure 1 As shown, including: Step 101, using pre-selected earthquake record data as input, performing multiple nonlinear dynamic time history analyses on the pre-constructed arch dam finite element, and obtaining an arch dam damage index set corresponding to the strength parameters of the earthquake record data, wherein the arch dam damage index set includes an arch dam damage index corresponding to each earthquake record data; A plurality of earthquake record data are selected from an earthquake record database for identifying optimal earthquake parameters.
[0022] Optionally, the earthquake record database uses the NGA-WEST2 database.
[0023] Optionally, 500 earthquake records with magnitudes between 5 and 8 are screened in the NGA-WEST2 database, and their moment magnitude-fault distance distribution is as follows: Figure 2 shown.
[0024] On this basis, the earthquake intensity parameter and the value of the earthquake intensity parameter are determined. Among them, the earthquake intensity parameter is used to characterize the characteristics of the earthquake, and the value of each earthquake intensity parameter is calculated according to each earthquake record data.
[0025] Optionally, in this implementation, acceleration-related, velocity-related, displacement-related, duration-related and structural characteristic-related seismic intensity parameters are selected, and a total of 30 scalar emphasis parameters are selected as intensity parameters to be identified to form a seismic intensity parameter set, as shown in Table 1 below: Table 1. Information on selected earthquake intensity parameters
[0026] It can be understood that the earthquake intensity parameter set includes the values of intensity parameters corresponding to each of the selected 500 earthquake record data, wherein each earthquake record data corresponds to 30 intensity parameters.
[0027] Furthermore, a three-dimensional finite element model of a high arch dam is established in the finite element software Abaqus, which takes into account multiple nonlinear factors such as the dynamic nonlinearity of the dam material and the geometric nonlinearity of the dam section contact. Figure 3 shown.
[0028] In a specific implementation, the arch dam is a high arch dam, and a three-dimensional finite element model of the high arch dam with multiple nonlinear factors such as linearity, dam section contact geometric nonlinearity, etc. is as follows: Figure 3 As shown in the figure, the concrete plastic damage constitutive model is used to characterize the dynamic nonlinearity of the dam material, the contact boundary surface model is used to characterize the contact geometric nonlinearity of the dam section, and the viscoelastic artificial boundary is set to consider the radiation damping effect of the seismic wave. The finite element model parameter settings are shown in Table 2: Table 2. Material parameter values for arch dam-foundation partition
[0029] On this basis, 500 nonlinear dynamic time-history analyses were carried out on the constructed high arch dam finite element model using the determined seismic motion intensity parameter set as input. It can be understood that each nonlinear time-history analysis takes a seismic motion record in the seismic motion intensity parameter set as input to obtain the arch dam damage index corresponding to the seismic motion record.
[0030] Corresponding to the selected 500 earthquake record data, 500 arch dam damage indicators are obtained to form an arch dam damage indicator set.
[0031] The arch dam damage index is used to characterize the damage degree of the arch dam under the corresponding seismic recording data. Different types of arch dams can be selected as arch dam damage indicators according to different specific parameters.
[0032] Step 102, taking the intensity parameter of each of the seismic record data as input, taking the arch dam damage index corresponding to each of the seismic record data as output to construct a data set, and training to obtain a first damage index prediction model; Through the above method, the earthquake intensity parameters and the corresponding arch dam damage index set can be obtained.
[0033] On this basis, the first damage index prediction model is obtained by training the acquired earthquake intensity parameters and the corresponding arch dam damage index set.
[0034] The input of the first damage index prediction model is the earthquake intensity parameter, and the output is the arch dam damage index, which is used to predict the arch dam damage index of the corresponding arch dam through the input earthquake intensity parameter. It can be understood that the earthquake intensity parameter here is a scalar earthquake intensity parameter.
[0035] To this end, a data set is constructed by taking the intensity parameter corresponding to a piece of seismic record data obtained in the above steps as input and the arch dam damage index corresponding to the seismic record data as output. The data set of this embodiment includes the 500 pieces of data obtained.
[0036] Optionally, the first damage indicator prediction model is a machine learning model. An appropriate model can be selected from existing machine learning models, trained using a pre-built data set, and the trained model is used as the first damage indicator prediction model.
[0037] Step 103: Based on the SHAP theory, determine the intensity parameter with the greatest contribution in the first damage index prediction model as the optimal scalar ground motion intensity parameter.
[0038] SHAP (Shapley Additive Explanations, explainable machine learning) theory is based on the Shapley value in game theory and is used to explain the output of any machine learning model. The main idea of SHAP is to regard a predicted result as an additive combination of the contribution values of each feature, so that the contribution of each feature in the final output of the model is explainable.
[0039] In other words, the contribution of the 30 strength parameters in the input to the arch dam damage index is determined through the SHAP theory, and the strength parameter with the highest contribution is confirmed as the strength parameter with the greatest impact on the arch dam loss, which is the required optimal scalar seismic parameter.
[0040] The present invention trains a first damage index prediction model based on seismic scalar intensity parameters to predict arch dam damage indicators, thereby being able to quantitatively evaluate the influence of different scalar earthquake intensity parameters on arch dam damage based on the SHAP theory, overcoming the problem of non-uniqueness of identification results in traditional methods and providing a new idea for identifying the optimal seismic motion parameters for arch dams.
[0041] In the arch dam optimal seismic parameter identification method of the present invention, the arch dam damage index is the arch dam damage volume ratio.
[0042] Optionally, the arch dam damage index is the arch dam damage volume ratio, that is, the damage degree of the arch dam is characterized by the arch dam damage volume ratio.
[0043] On this basis, taking the strength parameter PGA (peak ground velocity) as an example, the corresponding damage volume ratio calculation formula is as follows: ; In the formula, and Respectively represent i Damage value of the dam and volume; represents the total number of dam units. The PGA and dtv data sample distribution is as follows Figure 4 shown.
[0044] The method of calculating the damage volume ratio corresponding to other strength parameters is similar to the above method, so it will not be repeated here.
[0045] In the arch dam optimal seismic parameter identification method of the present invention, the step of taking the intensity parameter of each seismic record data as input, taking the arch dam damage index corresponding to each seismic record data as output to construct a data set, and training to obtain a first damage index prediction model specifically includes: Taking the intensity parameter of each of the seismic record data as input, taking the arch dam damage index corresponding to each of the seismic record data as output to construct a data set, and training to obtain multiple different types of candidate prediction models; Furthermore, considering the applicability of different machine learning models to dams with different structures, different locations and different strength parameters, it is possible to train multiple candidate prediction models of different types and select the candidate prediction model with the best prediction effect for the current dam body and current strength parameters as the final first damage indicator model.
[0046] In a specific implementation, six different candidate prediction models are constructed. The candidate prediction model reflecting the nonlinear functional relationship between the earthquake intensity parameter IM and the arch dam damage index EDP constructed by the machine learning algorithm is expressed as follows: ; In the formula, The dimension is Arch dam damage volume ratio matrix; The dimension is 6 machine learning models; The dimension is Prediction bias of 6 machine learning models.
[0047] The candidate prediction model with the best prediction performance is determined as the first damage indicator prediction model.
[0048] Considering the applicability of different machine learning algorithms in predicting the damage of high arch dams, the correlation coefficient between the ground motion intensity parameter and the arch dam damage volume ratio is optionally used. R 2 and mean absolute error MAE To determine the candidate prediction model with the best prediction performance.
[0049] Optionally, the multiple different types of candidate prediction models may include at least one or more of a support vector regression model, a gradient boosting tree model, an extreme gradient boosting tree model, a BP neural network model, a random forest model, and an explanatory benchmark model.
[0050] In this embodiment, the six machine learning models mentioned above are trained, and the results of using correlation coefficient and mean absolute error to evaluate their prediction performance are as follows: Figure 5 As shown, it can be seen that the gradient boosting tree model GBM has the best prediction performance. The comparison results of its training set and test set are as follows Figure 6 As shown, the optimal machine learning model is as follows: ; In the formula, It indicates the best fitting effect, that is, the machine learning model with the best predictive performance.
[0051] It is understandable that if the types and numbers of seismic intensity parameters are different or the arch dam damage indicators are different, the indicators for evaluating the prediction performance of the machine learning model can also be adaptively adjusted.
[0052] In the optimal seismic motion parameter identification method for an arch dam of the present invention, the step of determining the intensity parameter with the greatest contribution in the first damage index prediction model as the optimal scalar seismic motion intensity parameter based on the SHAP theory specifically includes: Based on the SHAP theory, calculating the marginal contribution value of the intensity parameter of each of the seismic record data to the output result of the first damage index prediction model; Taking a weighted average of all the marginal contribution values to obtain a SHAP value of each intensity parameter; The intensity parameter with the highest SHAP value is used as the optimal scalar ground motion intensity parameter.
[0053] Optionally, the marginal contribution of the intensity parameter of each earthquake record data to the arch dam damage index of the first damage index prediction model is calculated. MC j i , and the weighted average of all marginal contribution results is used to obtain the SHAP value, the result is as follows Figure 7 and Figure 8 As shown, its expression is as follows: ; In the formula, Indicates j SHAP value of each ground motion intensity parameter, Indicates j The first parameter of the ground motion intensity iThe strength parameter that determines the maximum SHAP value is S a (T1), that is, the optimal scalar ground motion intensity parameter is S a (T1).
[0054] In the method for identifying the optimal seismic motion parameters of an arch dam of the present invention, after the step of determining the intensity parameter with the greatest contribution in the first damage index prediction model as the optimal scalar seismic motion intensity parameter, the method further includes: Combining multiple intensity parameters with the largest contributions to construct multiple vector intensity parameter combinations; It can be understood that the single intensity parameter with the greatest impact on the arch dam damage index is obtained in the above method, which is the optimal scalar earthquake intensity parameter. The research conducted by the optimal scalar earthquake intensity parameter ignores the contribution of vector parameters to the structural response prediction. Therefore, it is also necessary to determine the optimal vector earthquake intensity parameter.
[0055] Specifically, through the above method, the SHAP values of the 30 determined intensity parameters are sorted, and the top ten intensity parameters of SHAP are selected, as shown in Table 3, and they are combined in pairs to form 45 vector intensity parameter combinations.
[0056] Table 3. Scalar intensity parameters with the top ten SHAP values
[0057] It can be understood that when the number of determined scalar intensity parameters is different, the number of intensity parameters selected when combining to obtain the vector intensity parameter combination can also be adjusted accordingly.
[0058] In other feasible implementations, other numbers of intensity parameters whose SHAP values are from large to small may be directly selected for combination.
[0059] Taking the vector intensity parameter combination corresponding to each of the seismic record data as input, taking the arch dam damage index corresponding to each of the seismic record data as output to construct a data set, and training to obtain a second damage index prediction model; Based on the SHAP theory, the vector intensity parameter combination with the greatest contribution in the second damage index prediction model is determined as the optimal vector seismic intensity parameter.
[0060] On this basis, that is, for each piece of seismic record data obtained, the corresponding vector intensity parameter combination is determined.
[0061] Therefore, the second damage indicator prediction model can be constructed in the same way as the first damage indicator prediction model. Specifically, the input is replaced by a scalar intensity parameter with a scalar intensity parameter combination, and the second damage indicator prediction model is retrained and determined.
[0062] Furthermore, for the second damage index prediction model, the method of determining the vector intensity parameter combination with the greatest contribution based on the SHAP theory is similar to the method of determining the optimal scalar ground motion intensity parameter, so it will not be repeated here.
[0063] For the 45 vector strength parameters constructed in this embodiment and the second damage index prediction model determined, the corresponding SHAP value results are calculated as follows: Fig. 9 and Fig.10 As shown in the figure, the optimal vector earthquake intensity parameter determined by this method is [S a (T1), S a (2T1)].
[0064] The present invention screens a large number of earthquake motions based on the NGA-WEST2 database, calculates a large number of commonly used earthquake motion intensity parameter values, ensures a wide distribution range of the selected earthquake motion magnitude and strong applicability of the intensity parameters, and at the same time, establishes an arch dam finite element model that considers multiple nonlinearities, thereby improving the accuracy of the calculation model. A variety of machine learning algorithms are used to construct a prediction model of earthquake damage and earthquake motion intensity parameters of high arch dams, evaluate the prediction performance of different models, determine the optimal prediction model, and avoid the instability and discreteness of the prediction results of a single model. The SHAP value is used to analyze the prediction model, and the scalar intensity parameters and vector intensity parameters that contribute most to the damage response of the arch dam under earthquake are identified, which effectively reduces the uncertainty of the earthquake response prediction of the high arch dam, and quantitatively evaluates the importance of different earthquake motion parameters to the damage of the arch dam, avoiding the problem of non-uniqueness of the identification results in the traditional method. The present invention can provide a new idea for the identification of the optimal earthquake motion parameters of the arch dam.
[0065] The optimal seismic parameter identification device for an arch dam provided by the present invention is described below. The optimal seismic parameter identification device for an arch dam described below and the optimal seismic parameter identification method for an arch dam described above can be referred to each other.
[0066] like Fig.11 As shown, the arch dam optimal earthquake motion parameter identification device includes a determination module 1101, a training module 1102 and an identification module 1103; The determination module 1101 is used to use the pre-selected seismic recording data as input, perform multiple nonlinear dynamic time-history analyses on the pre-constructed arch dam finite element, and obtain an arch dam damage index set corresponding to the strength parameters of the seismic recording data, wherein the arch dam damage index set includes an arch dam damage index corresponding to each of the seismic recording data; A plurality of earthquake record data are selected from an earthquake record database for identifying optimal earthquake parameters.
[0067] Optionally, the earthquake record database uses the NGA-WEST2 database.
[0068] Optionally, 500 earthquake records with magnitudes between 5 and 8 are screened in the NGA-WEST2 database, and their moment magnitude-fault distance distribution is as follows: Figure 2 shown.
[0069] On this basis, the earthquake intensity parameter and the value of the earthquake intensity parameter are determined. Among them, the earthquake intensity parameter is used to characterize the characteristics of the earthquake, and the value of each earthquake intensity parameter is calculated according to each earthquake record data.
[0070] Furthermore, a three-dimensional finite element model of a high arch dam is established in the finite element software Abaqus, which takes into account multiple nonlinear factors such as the dynamic nonlinearity of the dam material and the geometric nonlinearity of the dam section contact. Figure 3 shown.
[0071] On this basis, 500 nonlinear dynamic time-history analyses were carried out on the constructed high arch dam finite element model using the determined seismic motion intensity parameter set as input. It can be understood that each nonlinear time-history analysis takes a seismic motion record in the seismic motion intensity parameter set as input to obtain the arch dam damage index corresponding to the seismic motion record.
[0072] Corresponding to the selected 500 earthquake record data, 500 arch dam damage indicators are obtained to form an arch dam damage indicator set.
[0073] The arch dam damage index is used to characterize the damage degree of the arch dam under the corresponding seismic recording data. Different types of arch dams can be selected as arch dam damage indicators according to different specific parameters.
[0074] A training module 1102 is used to take the intensity parameter of each of the seismic record data as input, take the arch dam damage index corresponding to each of the seismic record data as output to construct a data set, and train to obtain a first damage index prediction model; Through the above method, the earthquake intensity parameters and the corresponding arch dam damage index set can be obtained.
[0075] On this basis, the first damage index prediction model is obtained by training the acquired earthquake intensity parameters and the corresponding arch dam damage index set.
[0076] The input of the first damage index prediction model is the earthquake intensity parameter, and the output is the arch dam damage index, which is used to predict the arch dam damage index of the corresponding arch dam through the input earthquake intensity parameter. It can be understood that the earthquake intensity parameter here is a scalar earthquake intensity parameter.
[0077] To this end, a data set is constructed by taking the intensity parameter corresponding to a piece of seismic record data obtained in the above steps as input and the arch dam damage index corresponding to the seismic record data as output. The data set of this embodiment includes the 500 pieces of data obtained.
[0078] Optionally, the first damage indicator prediction model is a machine learning model. An appropriate model can be selected from existing machine learning models, trained using a pre-built data set, and the trained model is used as the first damage indicator prediction model.
[0079] The identification module 1103 is used to determine the intensity parameter with the greatest contribution in the first damage indicator prediction model as the optimal scalar ground motion intensity parameter based on the SHAP theory.
[0080] SHAP (Shapley Additive Explanations, explainable machine learning) theory is based on the Shapley value in game theory and is used to explain the output of any machine learning model. The main idea of SHAP is to regard a predicted result as an additive combination of the contribution values of each feature, so that the contribution of each feature in the final output of the model is explainable.
[0081] In other words, the contribution of the 30 strength parameters in the input to the arch dam damage index is determined through the SHAP theory, and the strength parameter with the highest contribution is confirmed as the strength parameter with the greatest impact on the arch dam loss, which is the required optimal scalar seismic parameter.
[0082] The present invention trains a first damage index prediction model based on seismic scalar intensity parameters to predict arch dam damage indicators, thereby being able to quantitatively evaluate the influence of different scalar earthquake intensity parameters on arch dam damage based on the SHAP theory, overcoming the problem of non-uniqueness of identification results in traditional methods and providing a new idea for identifying the optimal seismic motion parameters for arch dams.
[0083] Fig.12 An example of a physical structure diagram of an electronic device is shown in FIG. Fig.12As shown, the electronic device may include: a processor 1210, a communication interface 1220, a memory 1230 and a communication bus 1240, wherein the processor 1210, the communication interface 1220 and the memory 1230 communicate with each other through the communication bus 1240. The processor 1210 may call the logic instructions in the memory 1230 to execute the arch dam optimal seismic parameter identification method, the method comprising: taking the pre-selected seismic record data as input, performing multiple nonlinear dynamic time history analyses on the pre-constructed arch dam finite element, obtaining an arch dam damage index set corresponding to the intensity parameter of the seismic record data, the arch dam damage index set including the arch dam damage index corresponding to each seismic record data; taking the intensity parameter of each seismic record data as input, taking the arch dam damage index corresponding to each seismic record data as output to construct a data set, and training to obtain a first damage index prediction model; based on the SHAP theory, determining the intensity parameter with the greatest contribution in the first damage index prediction model as the optimal scalar seismic intensity parameter.
[0084] In addition, the logic instructions in the above-mentioned memory 1230 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.
[0085] On the other hand, the present invention also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the arch dam optimal seismic parameter identification method provided by the above-mentioned methods. The method includes: taking pre-selected seismic recording data as input, performing multiple nonlinear dynamic time-history analyses on a pre-constructed arch dam finite element, and obtaining an arch dam damage index set corresponding to the intensity parameters of the seismic recording data, wherein the arch dam damage index set includes an arch dam damage index corresponding to each of the seismic recording data; taking the intensity parameter of each of the seismic recording data as input, and taking the arch dam damage index corresponding to each of the seismic recording data as output to construct a data set, and training to obtain a first damage index prediction model; based on the SHAP theory, determining the intensity parameter with the greatest contribution in the first damage index prediction model as the optimal scalar seismic intensity parameter.
[0086] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the arch dam optimal seismic motion parameter identification method provided by the above-mentioned methods, the method comprising: taking pre-selected seismic motion recording data as input, performing multiple nonlinear dynamic time-history analyses on a pre-constructed arch dam finite element, and obtaining an arch dam damage index set corresponding to the intensity parameters of the seismic motion recording data, wherein the arch dam damage index set includes an arch dam damage index corresponding to each of the seismic motion recording data; taking the intensity parameter of each of the seismic motion recording data as input, and taking the arch dam damage index corresponding to each of the seismic motion recording data as output to construct a data set, and training to obtain a first damage index prediction model; based on the SHAP theory, determining the intensity parameter with the greatest contribution in the first damage index prediction model as the optimal scalar seismic motion intensity parameter.
[0087] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.
[0088] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.
[0089] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A method for identifying optimal seismic parameters of an arch dam, characterized in that: include: Using the pre-selected seismic recording data as input, performing multiple nonlinear dynamic time history analyses on the pre-constructed arch dam finite element, and obtaining an arch dam damage index set corresponding to the strength parameters of the seismic recording data, wherein the arch dam damage index set includes an arch dam damage index corresponding to each of the seismic recording data; Taking the intensity parameter of each piece of seismic record data as input, taking the arch dam damage index corresponding to each piece of seismic record data as output to construct a data set, and training to obtain a first damage index prediction model; Based on the SHAP theory, the intensity parameter with the greatest contribution in the first damage index prediction model is determined as the optimal scalar ground motion intensity parameter.
2. The arch dam optimal ground motion parameter identification method according to claim 1 is characterized in that: The arch dam damage index is the arch dam damage volume ratio.
3. The arch dam optimal ground motion parameter identification method according to claim 1, characterized in that: The step of taking the intensity parameter of each piece of seismic record data as input, taking the arch dam damage index corresponding to each piece of seismic record data as output to construct a data set, and training to obtain a first damage index prediction model specifically includes: Taking the intensity parameter of each of the seismic record data as input, taking the arch dam damage index corresponding to each of the seismic record data as output to construct a data set, and training to obtain multiple different types of candidate prediction models; The candidate prediction model with the best prediction performance is determined as the first damage indicator prediction model.
4. The arch dam optimal ground motion parameter identification method according to claim 3 is characterized in that: The multiple different types of candidate prediction models include at least one or more of a support vector regression model, a gradient boosting tree model, an extreme gradient boosting tree model, a BP neural network model, a random forest model, and an explanatory benchmark model.
5. The arch dam optimal seismic parameter identification method according to claim 1, characterized in that: The step of determining the intensity parameter with the greatest contribution in the first damage index prediction model as the optimal scalar ground motion intensity parameter based on the SHAP theory specifically includes: Based on the SHAP theory, calculating the marginal contribution value of the intensity parameter of each of the seismic record data to the output result of the first damage index prediction model; Taking a weighted average of all the marginal contribution values to obtain a SHAP value of each intensity parameter; The intensity parameter with the highest SHAP value is used as the optimal scalar ground motion intensity parameter.
6. The method for identifying optimal seismic parameters of an arch dam according to any one of claims 1 to 5, characterized in that: After the step of determining the intensity parameter with the greatest contribution in the first damage index prediction model as the optimal scalar ground motion intensity parameter, the method further includes: Combining multiple intensity parameters with the largest contributions to construct multiple vector intensity parameter combinations; Taking the vector intensity parameter combination corresponding to each of the seismic record data as input, taking the arch dam damage index corresponding to each of the seismic record data as output to construct a data set, and training to obtain a second damage index prediction model; Based on the SHAP theory, the vector intensity parameter combination with the greatest contribution in the second damage index prediction model is determined as the optimal vector seismic intensity parameter.
7. An optimal seismic parameter identification device for an arch dam, characterized in that: include: A determination module is used to use the pre-selected seismic recording data as input, perform multiple nonlinear dynamic time-history analyses on the pre-constructed arch dam finite element, and obtain an arch dam damage index set corresponding to the strength parameters of the seismic recording data, wherein the arch dam damage index set includes an arch dam damage index corresponding to each of the seismic recording data; A training module, used to take the intensity parameter of each of the seismic record data as input, take the arch dam damage index corresponding to each of the seismic record data as output to construct a data set, and train to obtain a first damage index prediction model; The identification module is used to determine the intensity parameter with the greatest contribution in the first damage indicator prediction model as the optimal scalar ground motion intensity parameter based on the SHAP theory.
8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the optimal seismic parameter identification method for an arch dam as claimed in any one of claims 1 to 6 is implemented.
9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for identifying optimal seismic parameters of an arch dam as claimed in any one of claims 1 to 6 is implemented.
10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method for identifying optimal seismic parameters of an arch dam as claimed in any one of claims 1 to 6 is implemented.
Citation Information
Patent Citations
Canonical correlation analysis-based multi-element seismic damage potential assessment method
CN108170906A
Air quality data standardization method and device, computer equipment and medium
CN116776095A
Machine learning and SHAP value-based interpretable creep fracture life prediction method
CN116842834A
Seismic oscillation intensity index optimization method for seismic response prediction
CN117251926A
Impact resistance protection design and disaster assessment method for reinforced concrete member
CN117421560A
Cited By
Earthquake influence area determination method and device
CN119917828A
Steel beam tube floor earthquake damage rapid prediction method, system, medium and equipment
CN120257457A
