Inversion method, device and equipment for pumped storage power station dam filling information correction

By using an inversion method based on a radial basis function model, and optimizing earth-rock dam parameters through orthogonal experiments and deep learning networks, the error problem in the finite element calculation of earth-rock dams is solved, the calculation accuracy is improved, and it is applicable to the design and operation and maintenance of pumped storage power stations.

CN119004598APending Publication Date: 2024-11-22CHINA ELECTRIC POWER RESEARCH INSTITUTE CO LTD +1
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202411034111.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-30
Publication Date
2024-11-22

AI Technical Summary

Technical Problem

Existing technologies have significant errors in finite element calculations of earth-rock dams, mainly because the mechanical constitutive model cannot fully reflect the mechanical properties of the actual rockfill, resulting in a large difference between the parameters after inversion analysis and the actual observed values.

Method used

A method for correcting and inverting dam filling information of pumped storage power stations based on radial basis function models is adopted. Highly sensitive constitutive parameters are determined by orthogonal experimental design. By combining finite element analysis and radial basis function deep learning network, the hierarchical parameters and constitutive parameter values ​​are optimized, and a parameter inversion model is established.

Benefits of technology

It improves the accuracy of finite element calculations for earth-rock dams, reduces the difference between parameters after inversion analysis and actual observations, and is applicable to the design of new pumped storage power stations and the operation and maintenance management of existing power stations.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119004598A_ABST
    Figure CN119004598A_ABST
Patent Text Reader

Abstract

The invention relates to an inversion method and device for pumped storage power station dam filling information correction. The method comprises the steps that construction parameters and layering information of a newly-built pumped storage power station dam are obtained; and inputting construction parameters and layering information of the newly-built pumped storage power station dam into the trained dam parameter inversion model to obtain layering parameters and high-sensitivity constitutive parameter values of the newly-built pumped storage power station dam. According to the method, the technical problem that large differences still exist between finite element calculation parameter values and actual observation values after earth and rockfill dam inversion analysis is solved. On the basis of conforming to actual engineering, an earth and rockfill dam static inversion analysis algorithm for dam filling parameter correction of the pumped storage power station is provided, and the problems are preliminarily solved. The new algorithm is suitable for design analysis of a newly-built pumped storage power station, operation and maintenance management of a built pumped storage power station and the like. The invention also relates to a device and a storage medium.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of hydropower systems, and in particular to an inversion method, device and equipment for correcting filling information of a pumped storage power station dam. Background Art

[0002] At present, when using the finite element method to analyze earth-rock dams, there are usually large errors. The errors may come from factors such as calculation models, boundary conditions, mesh division, material parameter values ​​and construction process simulation. The most important reason is that the mechanical constitutive model and its parameters used cannot fully reflect the mechanical properties of the actual rockfill. Earth-rock dams are composed of large volumes of rockfill, and the individual components of the rockfill are different, and the physical and mechanical properties vary greatly. In order to improve the finite element calculation accuracy of earth-rock dams, an inverse analysis method for earth-rock dams is proposed, that is, to inversely calculate relatively correct mechanical constitutive parameters from construction data and observation data.

[0003] With the inversion algorithm currently used, when the inverted parameters are used for static analysis, there is still a considerable error between the calculated results and the actual observed values ​​in both space and time. Summary of the invention

[0004] In order to solve the technical problem that after the inversion analysis of the earth-rock dam, there is still a large difference between the finite element calculated parameter values ​​and the actual observed values, the present invention provides an inversion method, device and equipment for correcting the filling information of the dam of a pumped storage power station.

[0005] In a first aspect, the present invention provides an inversion method for correcting dam filling information of a pumped storage power station, the method comprising:

[0006] Obtain construction parameters and layering information for new pumped storage power station dams;

[0007] The construction parameters and stratification information of the newly built pumped storage power station dam are input into the trained dam parameter inversion model to obtain the stratification parameters and highly sensitive constitutive parameter values ​​of the newly built pumped storage power station dam. The dam parameter inversion model is constructed based on the radial basis function model.

[0008] Based on the above technical solution, further, the dam parameter inversion model is trained by the following method:

[0009] Based on the orthogonal test method, the types of highly sensitive constitutive parameters in the filling parameters of the established pumped storage power station dam are determined;

[0010] The built pumped storage power station dam is divided into several areas according to the construction design plan, and each area is divided into multiple construction layers according to the characteristics of the rockfill material;

[0011] By using the highly sensitive constitutive parameter values ​​of the established pumped-storage power station dam, the triaxial parameter values ​​of each initial construction layer and the initial stratification information through finite element analysis, the stratification parameter values ​​and highly sensitive constitutive parameter values ​​of the established pumped-storage power station dam are obtained;

[0012] The dam parameter inversion model to be trained is trained using the established hierarchical parameter values ​​and highly sensitive constitutive parameter values ​​of the pumped storage power station dam to obtain the dam parameter inversion model.

[0013] Based on the above technical solution, further, the highly sensitive constitutive parameter types in the filling parameters of the established pumped storage power station dam are determined based on the orthogonal test method, specifically including:

[0014] Select the maximum vertical displacement V of the dam body and the maximum horizontal displacement H upstream u and the maximum horizontal displacement downstream H d , as the test index in the orthogonal test method;

[0015] Since the rockfill is a granular material, cohesion is not considered, and the rockfill is in a loaded state during the dam filling and water storage process, the unloading modulus base and unloading modulus index are not considered, and the internal friction angle of the material at unit atmospheric pressure in the Duncan-Zhang EB model is selected. The value of the internal friction angle of the material decreases with the atmospheric pressure Destruction ratio R f , tangent modulus base K, tangent modulus index n, bulk modulus coefficient K b and bulk modulus index m, a total of seven parameters were used as experimental factors;

[0016] Assuming that there is no interaction between the parameters in the model, an orthogonal table is selected to arrange the experiment according to the number of experimental factors and the experimental factors, and the experimental factors are randomly assigned to the last seven columns of the orthogonal table. After each element in the orthogonal table is replaced with the corresponding design parameter value according to its corresponding experimental factor and experimental level, the highly sensitive constitutive parameter types in the established pumped storage power station dam filling parameters are obtained.

[0017] Based on the above technical solution, further, the established pumped storage power station dam is divided into several areas according to the construction design plan, and each area is divided into multiple construction layers according to the characteristics of the rockfill material, specifically including:

[0018] According to the construction design plan, the established pumped storage power station dam is divided into two major areas: the main rockfill and the secondary rockfill;

[0019] According to the characteristics of rockfill materials and construction methods, the main rockfill and secondary rockfill are initially divided into several layers. The thickness of each layer is determined according to the actual project. The rockfill materials in the same layer meet the requirements of similar stone size, similar stone strength, and comparable water content and porosity.

[0020] Based on the above technical solution, further, the layered parameter values ​​and highly sensitive constitutive parameter values ​​of the established pumped-storage power station dam are obtained by finite element analysis based on the highly sensitive constitutive parameter values ​​of the established pumped-storage power station dam, the triaxial parameter values ​​of each initial construction layer, and the initial layered information, specifically including:

[0021] Using the established initial stratification information and highly sensitive constitutive parameter values ​​of the pumped storage power station dam, a stratification information and highly sensitive parameter matrix is ​​constructed;

[0022] The layered parameter values ​​and highly sensitive constitutive parameter values ​​of the established pumped storage power station dam are obtained by using the triaxial parameter values ​​of each construction layer, the layered information and the highly sensitive parameter matrix, and by utilizing finite element calculation.

[0023] In a second aspect, the present invention further provides an inversion device for correcting filling information of a pumped storage power station dam, the device comprising:

[0024] The first processing module is used to obtain the construction parameters and layering information of the newly built pumped storage power station dam;

[0025] The second processing module is used to input the construction parameters and stratification information of the newly built pumped-storage power station dam into the trained dam parameter inversion model to obtain the stratification parameters and highly sensitive constitutive parameter values ​​of the newly built pumped-storage power station dam. The dam parameter inversion model is constructed based on the radial basis function model.

[0026] Based on the above technical solution, further, it also includes a construction module for determining the types of highly sensitive constitutive parameters in the filling parameters of the established pumped storage power station dam based on the orthogonal test method;

[0027] The built pumped storage power station dam is divided into several areas according to the construction design plan, and each area is divided into multiple construction layers according to the characteristics of the rockfill material;

[0028] By using the highly sensitive constitutive parameter values ​​of the established pumped-storage power station dam, the triaxial parameter values ​​of each initial construction layer and the initial stratification information through finite element analysis, the stratification parameter values ​​and highly sensitive constitutive parameter values ​​of the established pumped-storage power station dam are obtained;

[0029] The dam parameter inversion model to be trained is trained using the established hierarchical parameter values ​​and highly sensitive constitutive parameter values ​​of the pumped storage power station dam to obtain the dam parameter inversion model.

[0030] Based on the above technical solution, further, the construction module is specifically used to select the maximum vertical displacement V of the dam body, the maximum horizontal displacement H upstream u and the maximum horizontal displacement downstream H d , as the test index in the orthogonal test method;

[0031] Since the rockfill is a granular material, cohesion is not considered, and the rockfill is in a loaded state during the dam filling and water storage process, the unloading modulus base and unloading modulus index are not considered, and the internal friction angle of the material at unit atmospheric pressure in the Duncan-Zhang EB model is selected. The value of the internal friction angle of the material decreases with the atmospheric pressure Destruction ratio R f , tangent modulus base K, tangent modulus index n, bulk modulus coefficient K b and bulk modulus index m, a total of seven parameters were used as experimental factors;

[0032] Assuming that there is no interaction between the parameters in the model, an orthogonal table is selected to arrange the experiment according to the number of experimental factors and the experimental factors, and the experimental factors are randomly assigned to the last seven columns of the orthogonal table. After each element in the orthogonal table is replaced with the corresponding design parameter value according to its corresponding experimental factor and experimental level, the highly sensitive constitutive parameter types in the established pumped storage power station dam filling parameters are obtained.

[0033] Based on the above technical solution, further, the construction module is specifically used to divide the established pumped storage power station dam into two major areas of main rockfill and secondary rockfill according to the construction design plan;

[0034] According to the characteristics of rockfill materials and construction methods, the main rockfill and secondary rockfill are initially divided into several layers. The thickness of each layer is determined according to the actual project. The rockfill materials in the same layer meet the requirements of similar stone size, similar stone strength, and comparable water content and porosity.

[0035] In a third aspect, the present invention also provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, an inversion method for correcting the filling information of a pumped-storage power station dam described in any one of the above technical solutions is implemented.

[0036] In a fourth aspect, the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements an inversion method for correcting filling information of a pumped-storage power station dam as described in any one of the above technical solutions.

[0037] The present invention provides an inversion method for correcting the filling information of a pumped-storage power station dam, comprising obtaining the construction parameters and stratification information of a newly built pumped-storage power station dam; inputting the construction parameters and stratification information of the newly built pumped-storage power station dam into a trained dam parameter inversion model to obtain the stratification parameters and highly sensitive constitutive parameter values ​​of the newly built pumped-storage power station dam. The present invention solves the technical problem that after the inversion analysis of the earth-rock dam, there is still a large difference between the finite element calculated parameter values ​​and the actual observed values. On the basis of conforming to actual engineering, an earth-rock dam static inversion analysis algorithm for correcting the filling parameters of the pumped-storage power station dam is proposed, which preliminarily solves the above problems. The new algorithm is suitable for the design analysis of newly built pumped-storage power stations, the operation and maintenance management of existing pumped-storage power stations, etc. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] The accompanying drawings constituting a part of the present invention are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the accompanying drawings:

[0039] Figure 1 It is a flow chart of an inversion method for correcting dam filling information of a pumped storage power station provided by an embodiment of the present invention;

[0040] Figure 2 A schematic diagram of an orthogonal design provided by an embodiment of the present invention;

[0041] Figure 3 L provided in the embodiment of the present invention 18 (2x3 7 )Orthogonal representation;

[0042] Figure 4 A typical cross-sectional schematic diagram of a concrete face rockfill dam provided by an embodiment of the present invention;

[0043] Figure 5 A schematic diagram of the calculation of primary and secondary rockfill zones and layers for a concrete face rockfill dam provided by an embodiment of the present invention;

[0044] Figure 6 A schematic diagram of stress, strain and displacement calculation provided by an embodiment of the present invention;

[0045] Figure 7 The radial basis function deep learning network structure provided by the embodiment of the present invention is intended;

[0046] Figure 8 A schematic diagram of an inversion method for correcting filling information of a pumped storage power station dam provided by another embodiment of the present invention;

[0047] Fig. 9 A schematic diagram of a module of an inversion device for correcting filling information of a pumped-storage power station dam provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0048] The present invention will be described in detail below with reference to the accompanying drawings and in combination with embodiments. It should be noted that the embodiments and features in the embodiments of the present invention can be combined with each other without conflict.

[0049] The following detailed description is an exemplary description, which is intended to provide further detailed description of the present invention. Unless otherwise specified, all technical terms used in the present invention have the same meaning as those generally understood by those skilled in the art to which the present invention belongs. The terms used in the present invention are only for describing specific embodiments, and are not intended to limit the exemplary embodiments according to the present invention.

[0050] The following will be combined with the attached Figure 1 , an inversion method for correcting filling information of a pumped storage power station dam provided by an embodiment of the present invention is described, comprising the following steps:

[0051] 110. Obtain construction parameters and layering information of the newly built pumped storage power station dam;

[0052] 120. Input the construction parameters and stratification information of the newly built pumped-storage power station dam into the trained dam parameter inversion model to obtain the stratification parameters and highly sensitive constitutive parameter values ​​of the newly built pumped-storage power station dam, wherein the dam parameter inversion model is constructed based on the radial basis function model.

[0053] Based on the above embodiment, further, step 200 is included, including the following steps:

[0054] 210. Determine the types of highly sensitive constitutive parameters in the filling parameters of the established pumped storage power station dam based on the orthogonal test method;

[0055] 220. The built pumped storage power station dam is divided into several areas according to the construction design plan, and each area is divided into multiple construction layers according to the characteristics of the rockfill material;

[0056] 230. Obtaining layered parameter values ​​and highly sensitive constitutive parameter values ​​of the established pumped-storage power station dam through finite element analysis using the established highly sensitive constitutive parameter values ​​of the pumped-storage power station dam, the triaxial parameter values ​​of each initial construction layer, and the initial layered information;

[0057] Specifically, by comparing the actually monitored triaxial parameters, the stratified information, the highly sensitive constitutive parameter values ​​and the corresponding grid matrix are adjusted through the orthogonal test method, and trial calculations are continuously performed until the error between the stratified information, the highly sensitive constitutive parameter values ​​and the corresponding grid matrix is ​​less than the preset error value, thereby obtaining the stratified information and the highly sensitive constitutive parameter values.

[0058] 240. Use the established hierarchical parameter values ​​and highly sensitive constitutive parameter values ​​of the pumped-storage power station dam to train the dam parameter inversion model to be trained, so as to obtain the dam parameter inversion model.

[0059] Specifically, input parameters and output parameters are determined, wherein the input parameters are triaxial parameter values ​​measured for a number of earth-rock dams and construction parameters of each area, and the output parameters are stratification information and highly sensitive constitutive parameter values;

[0060] Establishing a radial basis function RBF model, and training the radial basis function RBF model using the input parameters and the output parameters to obtain a radial basis function RBF model that meets a preset accuracy value;

[0061] An intelligent algorithm is used in combination with the radial basis function RBF model to perform parameter optimization inversion to obtain the minimum value of the objective function;

[0062] When the intelligent algorithm runs to the maximum number of iterations, the dam parameter inversion model is obtained;

[0063] The radial basis function includes a Gaussian function and a multi-aperture function.

[0064] Based on the above embodiment, further, step 210 specifically includes:

[0065] Select the maximum vertical displacement V of the dam body and the maximum horizontal displacement H upstream u and the maximum horizontal displacement downstream H d , as the test index in the orthogonal test method;

[0066] Since the rockfill is a granular material, cohesion is not considered, and the rockfill is in a loaded state during the dam filling and water storage process, the unloading modulus base and unloading modulus index are not considered, and the internal friction angle of the material at unit atmospheric pressure in the Duncan-Zhang EB model is selected. The value of the internal friction angle of the material decreases with the atmospheric pressure Destruction ratio R f , tangent modulus base K, tangent modulus index n, bulk modulus coefficient K b and bulk modulus index m, a total of seven parameters were used as experimental factors;

[0067] Assuming that there is no interaction between the parameters in the model, an orthogonal table is selected to arrange the experiment according to the number of experimental factors and the experimental factors, and the experimental factors are randomly assigned to the last seven columns of the orthogonal table. After each element in the orthogonal table is replaced with the corresponding design parameter value according to its corresponding experimental factor and experimental level, the highly sensitive constitutive parameter types in the established pumped storage power station dam filling parameters are obtained.

[0068] Based on the above embodiment, further, step 220 specifically includes:

[0069] According to the construction design plan, the established pumped storage power station dam is divided into two major areas: the main rockfill and the secondary rockfill;

[0070] According to the characteristics of rockfill materials and construction methods, the main rockfill and secondary rockfill are initially divided into several layers. The thickness of each layer is determined according to the actual project. The rockfill materials in the same layer meet the requirements of similar stone size, similar stone strength, and comparable water content and porosity.

[0071] Based on the above embodiment, further, step 230 specifically includes:

[0072] Using the established initial stratification information and highly sensitive constitutive parameter values ​​of the pumped storage power station dam, a stratification information and highly sensitive parameter matrix is ​​constructed;

[0073] The layered parameter values ​​and highly sensitive constitutive parameter values ​​of the established pumped storage power station dam are obtained by using the triaxial parameter values ​​of each construction layer, the layered information and the highly sensitive parameter matrix, and by utilizing finite element calculation.

[0074] The present embodiment involves an inversion method for correcting the filling information of a pumped-storage power station dam, including obtaining the construction parameters and stratification information of a newly built pumped-storage power station dam; inputting the construction parameters and stratification information of the newly built pumped-storage power station dam into a trained dam parameter inversion model to obtain the stratification parameters and highly sensitive constitutive parameter values ​​of the newly built pumped-storage power station dam. The present invention solves the technical problem that after the inversion analysis of the earth-rock dam, there is still a large difference between the finite element calculated parameter values ​​and the actual observed values. On the basis of conforming to actual engineering, an algorithm for static inversion analysis of earth-rock dams with corrected filling parameters of pumped-storage power station dams is proposed, which preliminarily solves the above problems. The new algorithm is suitable for the design analysis of newly built pumped-storage power stations, the operation and maintenance management of existing pumped-storage power stations, etc.

[0075] It should be understood that in order to solve the problem that after the back analysis of earth-rock dam, the finite element calculation parameter values ​​are still quite different from the actual observation values, a static back analysis algorithm for earth-rock dam with modified filling parameters of pumped storage power station dam is proposed on the basis of conformity with actual engineering, which preliminarily solves the above problem. The new algorithm is suitable for the design analysis of new pumped storage power stations, the operation and maintenance management of existing pumped storage power stations, etc.

[0076] like Figure 8 As shown, a method for correcting the filling parameters of a pumped storage power station dam is as follows:

[0077] 1. Conduct sensitivity analysis on the constitutive parameters of earth-rock dam. Use orthogonal test to conduct sensitivity analysis of constitutive parameters and quantify the sensitivity of each parameter.

[0078] Orthogonal experimental design is another design method for studying multiple factors and multiple levels. It selects some representative points from the comprehensive test based on orthogonality for testing. These representative points have the characteristics of "even dispersion, neatness and comparability". Orthogonal experimental design is the main method of fractional factorial design. Orthogonal experimental design is an efficient, fast and economical experimental design method.

[0079] L9(3 4 ) orthogonal table as an example, L is the orthogonal table code, 9 means that the table has 9 rows, 3 means that each influencing factor has 3 level values, and 4 means that the orthogonal table can arrange 4 factors at most. Figure 1 For L9(3 4 ) for an intuitive explanation: the three-dimensional cube is divided into 27 intersections by three mutually perpendicular planes, and each dimension represents an influencing factor; each dimension is divided into three parts to represent the level of each factor, and all 27 intersections represent the total number of experiments required for three factors and three levels.

[0080] According to the orthogonal design concept, only 9 evenly distributed test points are needed. The specific positions are arranged as follows: each plane contains 9 intersection points, which can ensure that there are exactly 3 orthogonal test points, and there is 1 orthogonal test scheme distributed on each horizontal and vertical line. Figure 2 The following is a 9-marked experimental point scheme. Taking the triangle-marked point as an example, it represents A2B2C1, that is, factor A takes 2 levels, factor B takes 2 levels, and factor C takes 1 level.

[0081] The steps of orthogonal experimental design are as follows:

[0082] (1) Selection of test indicators. According to the selection principle of test indicators in parameter sensitivity analysis, considering that the settlement deformation and upstream and downstream horizontal deformation of the rockfill under the action of its own weight and water load have a greater impact on the dam, the maximum vertical displacement V of the dam body and the maximum upstream horizontal displacement H are selected. u and the maximum horizontal displacement downstream H d , as the main experimental indicator for parameter sensitivity analysis.

[0083] (2) Determine the test factors and factor levels. Rockfill is a granular material, and its cohesion is considered to be 0. In addition, the rockfill is in a loaded state during the dam filling and water storage process. The unloading modulus base and unloading modulus index in its model parameters are not involved in the calculation. Therefore, the unloading modulus base and unloading modulus index in the model are selected. R f ,K,n,K b A total of 7 parameters, including , were used for sensitivity analysis.

[0084] (3) Select the orthogonal array design experiment. Assuming that there is no interaction between the parameters in the model, according to the number of experimental factors and the number of factor levels, such as Figure 3 , select L 18 (2x3 7 ) The experiment was arranged by orthogonal table (the first column was set as an empty column), the experimental factors were randomly assigned to the last 7 columns of the orthogonal table, and each element in the orthogonal table was replaced with the corresponding design parameter value according to its corresponding factor and level, thus obtaining the parameter sensitivity of the rockfill model

[0085] By analyzing the finite element pre-calculation results, we know that the parameter K in the Duncan-Zhang EB model b , K has a great influence on the vertical displacement of the dam body, among which K b The sensitivity is the greatest, followed by parameter K, R f The impact on the upstream horizontal displacement of the dam body is significant. The sensitivity of K is the largest; the parameter value has little effect on the downstream displacement. b The sensitivity of other parameters is relatively small.

[0086] 2 According to the actual engineering construction steps, the earth-rock dam is divided into several areas according to the construction and design plan. According to the primary and secondary rockfill design, the earth-rock dam is first divided into two major areas, and then each area is preliminarily divided into several layers according to the construction thickness.

[0087] Figure 4-5 This is a typical earth-rock dam design profile. The typical dam body is divided into two parts: primary and secondary rockfill. The positions of the primary and secondary rockfill are not completely fixed. The secondary rockfill area is between the left and right limit surfaces. In China, it is generally preferred that the secondary rockfill is closer to the right limit surface, making the volume of the primary rockfill larger than that of the secondary rockfill.

[0088] Since the upstream coverage of the face plate cushion has a low impact on the finite element calculation results, a unified calculation parameter is used. The mechanical characteristic parameters of the rockfill body have a greater impact on the finite element calculation results, so it is processed in layers and zones. The processing principle of layers and zones shall be based on the actual engineering work.

[0089] First, according to the actual project, it is divided into two major areas: primary and secondary rockfill. Then, according to the characteristics of rockfill materials and construction methods, the primary and secondary rockfill bodies are initially divided into several layers. The thickness of each layer is determined according to the actual project. The rockfill materials in the same layer should meet the conditions of similar size, similar strength, moisture content, and equivalent porosity. After the stratification is determined, the different stratifications are grouped according to the finite element calculation process. Among them, the mechanical properties of the face plate and cushion materials are basically the same. Therefore, the unified mechanical parameters are used in the finite element calculation. Due to the large volume and long filling time of the primary and secondary rockfill bodies, there are often certain differences in the mechanical properties of the materials filled in various areas in the actual project. Due to the influence of construction factors, the differences in the mechanical properties of materials in different areas are further magnified. Therefore, the stratification method shown in the figure above is used to group the calculation parameters of different areas. The stratification rules are determined based on the characteristics of the filling materials, design methods, construction methods, monitoring data, etc.

[0090] 3 Using the finite element method, according to the stress, strain and deformation actually observed in each sub-region project, reasonable high-sensitivity constitutive parameter values ​​of each sub-region are inverted.

[0091] Using the finite element analysis method, the stress, strain and displacement values ​​at each location are calculated through the layered information and the highly sensitive constitutive parameter values. The n-layer information and the initial highly sensitive parameters corresponding to each layer are shown in Table 1. The n-layer information and the initial highly sensitive parameters corresponding to each layer are shown in Table 1.

[0092]

[0093] Table 1 Stratification information and high sensitivity parameters

[0094] like Figure 6 Through finite element calculation, the stress, strain and displacement of each layer can be analyzed. By comparing the actual monitoring values, the layer information and the highly sensitive constitutive parameter values ​​and the corresponding grid information are adjusted through orthogonal tests, and trial analysis is continuously performed until the error between the two is less than the allowable value, and the layer information and the highly sensitive constitutive parameter values ​​are derived.

[0095] 4. Figure 7 ,The radial basis function (RBF) deep learning network is used to analyze the construction parameters, stratification information, and constitutive information. By inputting the construction parameters actually observed in each sub-regional project, the relationship between the construction parameters and the reasonable constitutive parameter values ​​of each sub-region is established.

[0096] Radial basis function neural network, referred to as RBF neural network, is a feedforward neural network model based on radial basis function. It consists of input layer, hidden layer and output layer, in which the neurons in the hidden layer use radial basis function to process input data.

[0097] The steps of dam parameter inversion analysis based on 1# radial basis function model are as follows:

[0098] Step 1: Determine input and output parameters. The input parameters of the 1# radial basis function of the present invention are a large number of triaxial parameters measured from earth-rock dams; initial stratification information; and construction parameters of each sub-region (dry density, porosity, water content, etc.). The output parameters are highly sensitive constitutive parameter values.

[0099] Step 2: Establish an RBF proxy model. Use the training set and test set obtained in step 1 to establish an RBF proxy model with sufficient accuracy to replace the finite element calculation.

[0100] Step 3: Minimize the objective function value. Use intelligent algorithm combined with RBF proxy model to perform parameter optimization inversion to obtain the minimum value of the objective function.

[0101] Step 4: Output the results. When the intelligent algorithm runs to the maximum number of iterations, the inversion parameters and accuracy can be output.

[0102] Selection of radial basis function: Commonly used radial basis functions include Gaussian function, multi-aperture function, etc. The characteristics of these functions are that they have local response in the input space, that is, the closer the distance between the input data and its center, the larger the output value.

[0103] Training process: The weights and bias values ​​of the hidden layer neurons and the weights of the output layer are determined through the training algorithm. Commonly used training algorithms include the minimum mean square error (MSE) algorithm, the maximum likelihood estimation algorithm, etc. The advantage of the RBF neural network is that it can approximate functions of arbitrary complexity, and during the training process, only the parameters of the hidden layer neurons need to be adjusted, while the output layer parameters can be determined using simple methods such as linear regression.

[0104] The input parameters in the present invention are triaxial parameters and construction parameters (dry density, porosity, water content, etc.), and the output parameters are reasonable stratification information of earth-rock dam and highly sensitive constitutive parameter values.

[0105] The present invention proposes a new constitutive parameter inversion analysis method based on construction parameters and design construction scheme. First, a sensitivity analysis is performed on the constitutive parameters to obtain highly sensitive constitutive parameters, the stratification information is determined using the design scheme, and the initial Duncan-Zhang EB model parameters are determined through triaxial tests. The corresponding stress-strain values ​​are calculated using finite elements using highly sensitive constitutive parameters and stratification information, and compared with the actual monitored values. The parameters are adjusted using orthogonal tests, and trial calculations are continuously performed until the error between the two is less than the allowable value. The reasonable constitutive parameters of each sub-region and the radial basis function (RBF) deep learning network of construction parameters are used to complete the calculation from construction parameters (dry density, porosity, moisture content, etc.) to highly sensitive constitutive parameters of each stratification.

[0106] The present invention is applicable to numerical analysis of pumped storage power stations, including but not limited to early design and later operation and maintenance.

[0107] 1 Use orthogonal test to carry out sensitivity analysis of constitutive parameters.

[0108] 2 Based on the design and construction plan of earth-rock dam, the calculation method of earth-rock dam is divided into several areas.

[0109] 3 According to the actual observed stress, strain and deformation of each sub-regional project, the finite element and orthogonal analysis methods are used to invert the reasonable high-sensitivity constitutive parameter values ​​of each sub-region.

[0110] 4 Use the radial basis function (RBF) deep learning network to perform high-sensitivity analysis between layered high-sensitivity parameters and construction parameters and establish the connection between the two.

[0111] The following will be combined with the attached Fig. 9 , an inversion device for correcting dam filling information of a pumped storage power station provided by an embodiment of the present invention is described, comprising the following steps:

[0112] The first processing module is used to obtain the construction parameters and layering information of the newly built pumped storage power station dam;

[0113] The second processing module is used to input the construction parameters and stratification information of the newly built pumped-storage power station dam into the trained dam parameter inversion model to obtain the stratification parameters and highly sensitive constitutive parameter values ​​of the newly built pumped-storage power station dam. The dam parameter inversion model is constructed based on the radial basis function model.

[0114] Based on the above technical solution, further, it also includes a construction module for determining the types of highly sensitive constitutive parameters in the filling parameters of the established pumped storage power station dam based on the orthogonal test method;

[0115] The built pumped storage power station dam is divided into several areas according to the construction design plan, and each area is divided into multiple construction layers according to the characteristics of the rockfill material;

[0116] By using the highly sensitive constitutive parameter values ​​of the established pumped-storage power station dam, the triaxial parameter values ​​of each initial construction layer and the initial stratification information through finite element analysis, the stratification parameter values ​​and highly sensitive constitutive parameter values ​​of the established pumped-storage power station dam are obtained;

[0117] The dam parameter inversion model to be trained is trained using the established hierarchical parameter values ​​and highly sensitive constitutive parameter values ​​of the pumped storage power station dam to obtain the dam parameter inversion model.

[0118] Based on the above technical solution, further, the construction module is specifically used to select the maximum vertical displacement V of the dam body, the maximum horizontal displacement H upstream u and the maximum horizontal displacement downstream H d , as the test index in the orthogonal test method;

[0119] Since the rockfill is a granular material, cohesion is not considered, and the rockfill is in a loaded state during the dam filling and water storage process, the unloading modulus base and unloading modulus index are not considered, and the internal friction angle of the material at unit atmospheric pressure in the Duncan-Zhang EB model is selected. The value of the internal friction angle of the material decreases with the atmospheric pressure Destruction ratio R f , tangent modulus base K, tangent modulus index n, bulk modulus coefficient K b and bulk modulus index m, a total of seven parameters were used as experimental factors;

[0120] Assuming that there is no interaction between the parameters in the model, an orthogonal table is selected to arrange the experiment according to the number of experimental factors and the experimental factors, and the experimental factors are randomly assigned to the last seven columns of the orthogonal table. After each element in the orthogonal table is replaced with the corresponding design parameter value according to its corresponding experimental factor and experimental level, the highly sensitive constitutive parameter types in the established pumped storage power station dam filling parameters are obtained.

[0121] Based on the above technical solution, further, the construction module is specifically used to divide the established pumped storage power station dam into two major areas of main rockfill and secondary rockfill according to the construction design plan;

[0122] According to the characteristics of rockfill materials and construction methods, the main rockfill and secondary rockfill are initially divided into several layers. The thickness of each layer is determined according to the actual project. The rockfill materials in the same layer meet the requirements of similar stone size, similar stone strength, and comparable water content and porosity.

[0123] This embodiment involves an inversion device for correcting the filling information of a pumped-storage power station dam, which solves the technical problem that after the inversion analysis of the earth-rock dam, there is still a large difference between the finite element calculation parameter value and the actual observation value. On the basis of conforming to the actual project, a static inversion analysis algorithm for the earth-rock dam with corrected filling parameters of the pumped-storage power station dam is proposed, which preliminarily solves the above problem. The new algorithm is suitable for the design analysis of new pumped-storage power stations, the operation and maintenance management of existing pumped-storage power stations, etc.

[0124] The present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, an inversion method for correcting the filling information of a pumped-storage power station dam described in any one of the above technical solutions is implemented.

[0125] The present invention also provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, an inversion method for correcting filling information of a pumped-storage power station dam described in any one of the above technical solutions is implemented.

[0126] The above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, a person skilled in the art should understand that the specific implementation modes of the present invention can still be modified or replaced by equivalents, and any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be included in the scope of protection of the claims of the present invention.

Claims

1. An inversion method for correcting dam filling information of a pumped storage power station, characterized in that: The method comprises: Obtain construction parameters and layering information for new pumped storage power station dams; The construction parameters and stratification information of the newly built pumped storage power station dam are input into the trained dam parameter inversion model to obtain the stratification parameters and highly sensitive constitutive parameter values ​​of the newly built pumped storage power station dam. The dam parameter inversion model is constructed based on the radial basis function model.

2. The method according to claim 1, characterized in that The dam parameter inversion model is trained by the following method: Based on the orthogonal test method, the types of highly sensitive constitutive parameters in the filling parameters of the established pumped storage power station dam are determined; The built pumped storage power station dam is divided into several areas according to the construction design plan, and each area is divided into multiple construction layers according to the characteristics of the rockfill material; By using the highly sensitive constitutive parameter values ​​of the established pumped-storage power station dam, the triaxial parameter values ​​of each initial construction layer and the initial stratification information through finite element analysis, the stratification parameter values ​​and highly sensitive constitutive parameter values ​​of the established pumped-storage power station dam are obtained; The dam parameter inversion model to be trained is trained using the established hierarchical parameter values ​​and highly sensitive constitutive parameter values ​​of the pumped storage power station dam to obtain the dam parameter inversion model.

3. The method according to claim 2, characterized in that The highly sensitive constitutive parameter types in the filling parameters of the established pumped storage power station dam determined based on the orthogonal test method specifically include: Select the maximum vertical displacement V of the dam body and the maximum horizontal displacement H upstream u and the maximum horizontal displacement downstream H d , as the test index in the orthogonal test method; Since the rockfill is a granular material, cohesion is not considered, and the rockfill is in a loaded state during the dam filling and water storage process, the unloading modulus base and unloading modulus index are not considered. The internal friction angle φ0 of the material at unit atmospheric pressure, the value of the internal friction angle of the material decreasing with atmospheric pressure Δφ, and the damage ratio R in the Duncan-Zhang EB model are selected. f , tangent modulus base K, tangent modulus index n, bulk modulus coefficient K b and bulk modulus index m, a total of seven parameters were used as experimental factors; Assuming that there is no interaction between the parameters in the model, an orthogonal table is selected to arrange the experiment according to the number of experimental factors and the experimental factors, and the experimental factors are randomly assigned to the last seven columns of the orthogonal table. After each element in the orthogonal table is replaced with the corresponding design parameter value according to its corresponding experimental factor and experimental level, the highly sensitive constitutive parameter types in the established pumped storage power station dam filling parameters are obtained.

4. The method according to claim 2, characterized in that: The pumped storage power station dam that has been built is divided into several areas according to the construction design plan, and each area is divided into multiple construction layers according to the characteristics of the rockfill material, specifically including: According to the construction design plan, the established pumped storage power station dam is divided into two major areas: the main rockfill and the secondary rockfill; According to the characteristics of rockfill materials and construction methods, the main rockfill and secondary rockfill are initially divided into several layers. The thickness of each layer is determined according to the actual project. The rockfill materials in the same layer meet the requirements of similar stone size, similar stone strength, and comparable water content and porosity.

5. The method according to claim 2, characterized in that: The layered parameter values ​​and highly sensitive constitutive parameter values ​​of the established pumped-storage power station dam are obtained by finite element analysis based on the highly sensitive constitutive parameter values ​​of the established pumped-storage power station dam, the triaxial parameter values ​​of each initial construction layer, and the initial layered information, specifically including: Using the established initial stratification information and highly sensitive constitutive parameter values ​​of the pumped storage power station dam, a stratification information and highly sensitive parameter matrix is ​​constructed; The layered parameter values ​​and highly sensitive constitutive parameter values ​​of the established pumped storage power station dam are obtained by using the triaxial parameter values ​​of each construction layer, the layered information and the highly sensitive parameter matrix, and by utilizing finite element calculation.

6. An inversion device for correcting dam filling information of a pumped storage power station, characterized in that: The device comprises: The first processing module is used to obtain the construction parameters and layering information of the newly built pumped storage power station dam; The second processing module is used to input the construction parameters and stratification information of the newly built pumped-storage power station dam into the trained dam parameter inversion model to obtain the stratification parameters and highly sensitive constitutive parameter values ​​of the newly built pumped-storage power station dam. The dam parameter inversion model is constructed based on the radial basis function model.

7. The device according to claim 6, characterized in that Also included is a building module for determining highly sensitive constitutive parameter categories in the filling parameters of an established pumped storage power station dam based on an orthogonal test method; The built pumped storage power station dam is divided into several areas according to the construction design plan, and each area is divided into multiple construction layers according to the characteristics of the rockfill material; By using the highly sensitive constitutive parameter values ​​of the established pumped-storage power station dam, the triaxial parameter values ​​of each initial construction layer and the initial stratification information through finite element analysis, the stratification parameter values ​​and highly sensitive constitutive parameter values ​​of the established pumped-storage power station dam are obtained; The dam parameter inversion model to be trained is trained using the established hierarchical parameter values ​​and highly sensitive constitutive parameter values ​​of the pumped storage power station dam to obtain the dam parameter inversion model.

8. The device according to claim 7, characterized in that The construction module is specifically used to select the maximum vertical displacement V of the dam body and the maximum horizontal displacement H upstream. u and the maximum horizontal displacement downstream H d , as the test index in the orthogonal test method; Since the rockfill is a granular material, cohesion is not considered, and the rockfill is in a loaded state during the dam filling and water storage process, the unloading modulus base and unloading modulus index are not considered. The internal friction angle φ0 of the material at unit atmospheric pressure, the value of the internal friction angle of the material decreasing with atmospheric pressure Δφ, and the damage ratio R in the Duncan-Zhang EB model are selected. f , tangent modulus base K, tangent modulus index n, bulk modulus coefficient K b and bulk modulus index m, a total of seven parameters were used as experimental factors; Assuming that there is no interaction between the parameters in the model, an orthogonal table is selected to arrange the experiment according to the number of experimental factors and the experimental factors, and the experimental factors are randomly assigned to the last seven columns of the orthogonal table. After each element in the orthogonal table is replaced with the corresponding design parameter value according to its corresponding experimental factor and experimental level, the highly sensitive constitutive parameter types in the established pumped storage power station dam filling parameters are obtained.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the inversion method for correcting the dam filling information of the pumped storage power station according to any one of claims 1 to 5 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the inversion method for correcting the filling information of the pumped storage power station dam according to any one of claims 1 to 5 is implemented.

Citation Information

Cited By

  • Static force calculation method for prefabricated dam

    CN120145730A