Internal temperature simulation system of gravimeter system

Abstract

The invention relates to the technical field of simulation monitoring, and particularly discloses a gravimeter system internal temperature simulation system, which comprises the steps of drawing a gravimeter system geometric model, determining material attributes, establishing a heat transfer physical field model, setting heat conduction and heat radiation parameters, and analyzing a gravimeter internal temperature gradient. Determining historical division groups of grids through a grid division module; the matching screening module is used for obtaining a matching parameter group; the correlation analysis module is used for acquiring a grid density value, performing correlation analysis on the grid density value and simulation calculation time, and constructing a simulation density group; the error analysis module is used for acquiring a simulation result error, performing fitting analysis and determining a subthreshold density data set; and finally, a density determination module determines the optimal grid density by using a genetic algorithm based on the subthreshold density data set and the simulation density set. According to the method, the internal temperature of the gravimeter system is effectively simulated, and support is provided for optimizing the grid division density of the gravimeter.

Description

Technical Field

[0001] The invention relates to the technical field of simulation monitoring, and in particular to an internal temperature simulation system of a gravimeter system. Background Art

[0002] In the field of gravity instruments, the temperature control device is an important device for relative gravimeters, and the temperature control effect can directly affect the accuracy of relative gravimeters. A change of 1°C in ambient temperature can cause a change of about 120,000 microgal in the reading of the gravity sensor, which is much larger than the design index of the relative gravimeter (5 microgal). According to theoretical calculations, a constant temperature device of 0.00001°C needs to be designed, but it is difficult to achieve in practice. In order to experiment with higher-precision temperature control, the temperature analysis of traditional gravimeter systems faces many challenges.

[0003] A Chinese patent application with publication number CN112784463A discloses a method for simulating the internal temperature of a relative gravimeter system, including: drawing a system geometric model according to the internal structure of the relative gravimeter system; assigning different material properties to each calculation domain in the system structure model to establish a heat transfer physical field model; analyzing the temperature gradient of the area in the heat transfer physical field model; and obtaining the temperature simulation result of the relative gravimeter system when the inner temperature control module works stably.

[0004] In the prior art, there is a lack of grid division in temperature monitoring processing, a lack of effective use of historical data, and difficulty in ensuring the rationality and consistency of grid division. This results in low calculation efficiency in subsequent simulation calculations, and is unable to reflect the physical properties of the complex internal structure of the gravimeter, and cannot provide a reliable basis for subsequent simulation optimization of the gravimeter.

[0005] In the existing technology, when analyzing the relationship between grid density, simulation calculation time, and simulation result error in the simulation analysis of the internal temperature of the gravimeter system, the existing technology can only perform simple qualitative analysis and cannot conduct accurate quantitative research. This makes it impossible for researchers to effectively control the simulation error from the perspective of monitoring grid division while ensuring calculation efficiency.

[0006] To this end, the present invention provides a gravimeter system internal temperature simulation system. Summary of the invention

[0007] The object of the present invention is to provide a temperature simulation system inside a gravimeter system to solve the above-mentioned problems.

[0008] The purpose of the present invention can be achieved through the following technical solutions: A gravimeter system internal temperature simulation system, comprising the following modules: Simulation construction module: used to construct a simulation model of the internal temperature of the gravimeter system and obtain the temperature gradient change field of the gravimeter system; Grid division module: used to grid the temperature gradient change field of the gravimeter system, obtain the grid density and division similarity value, and determine the historical division group of the grid; Matching and screening module: used to obtain the similarity values ​​of all grid divisions, perform cluster analysis to obtain grid parameter groups, compare the grid parameter groups with the historical division groups, and obtain matching parameter groups; Correlation analysis module: used to obtain the grid density value and the corresponding simulation calculation time of the matching parameter group, construct a standard grid density data group and a standard simulation time data group, and perform correlation analysis to obtain a simulation density group; Error analysis module: used to obtain the simulation result error, perform fitting analysis on the grid density and the simulation result error, and determine the subthreshold density data group; Density determination module: used to train the subthreshold density data group and the simulation density group using a genetic algorithm to determine the optimal grid density.

[0009] As a further technical solution of the present invention: the method for obtaining the historical division groups is: Obtain historical meshing data from historical simulation data of the simulation software, wherein the historical meshing data includes: mesh density, boundary conditions, and simulation parameters; The simulation parameters include: the characteristic length L of the grid, the simulation time step of the grid ; By formula: Get the heat flux q of the grid, quantify the boundary conditions of the grid, h is the convective heat exchange coefficient, is the temperature difference of the grid; By formula: Get the Fourier number Fo of the grid and quantify the simulation parameters of the grid. α is the thermal diffusion coefficient. The grid density, Fourier number, and heat flux are dimensionally processed, and the dimensionally processed results are weighted summed to obtain the partition similarity value; The partition similarity values ​​of different grids are obtained, and the grids whose partition similarity values ​​are within a preset similarity range are merged as a historical partition group.

[0010] As a further technical solution of the present invention: the matching parameter group is obtained in the following manner: From the gravimeter simulation model constructed by the simulation software, the similarity value of the division of the simulation model output grid is obtained, and the grid parameter model is constructed by the K-Means clustering analysis algorithm to obtain the grid parameter group; Combined with the historical division group, numerical analysis is performed on the division similarity values ​​of the grid parameter group to obtain the grid comparison group and overlap matching value; If the overlap matching value of each grid comparison group is higher than a preset overlap matching threshold, the grid parameter group in the grid comparison group is marked as a matching parameter group.

[0011] As a further technical solution of the present invention: the overlapping matching value and the grid comparison group are obtained by: Obtain the overlapping area of ​​the grid parameter group and multiple historical division groups in the gravimeter system; Select the historical division group with the largest overlapping area between the grid parameter group and multiple historical division groups, bind the grid parameter group and the historical division group with the largest overlapping area one by one, and obtain a grid comparison group; Get the grids that appear in both the grid parameter group and the historical division group in the grid comparison group and mark them as overlapping grids; The number of grids in the grid parameter group and the number of overlapping grids are obtained, and the number of overlapping grids is ratioed to the number of grids in the grid parameter group to obtain an overlapping grid ratio; Get the overlapping grid ratio of all grid comparison groups, marked as W i , i is the number of the grid comparison group, and the value range of i is [1,m]; Obtaining the division similarity values ​​of all overlapping grids in the grid comparison group, and summing up the division similarity values ​​of all overlapping grids to obtain the overlapping similarity value; The partition similarity values ​​within the grid parameter group are summed to obtain the parameter similarity value; The overlap similarity value is compared with the parameter similarity value to obtain the overlap similarity ratio, which is marked as C. i ; By formula: Get the overlapping match value Pz i , where a1 and a2 are preset proportional coefficients.

[0012] As a further technical solution of the present invention: the method for obtaining the simulation density group is: Obtain the grid density value of each matching parameter group in the gravimeter system, and obtain the simulation calculation time of different grid density values ​​in multiple historical division groups; The grid density values ​​of different matching parameter groups are used as the first data element to construct a grid density data group D. , where d j Represents the grid density value of the matching parameter group numbered j; The simulation calculation time corresponding to different matching parameter groups is used as the second data element to construct a simulation time data group T. , where t jrepresents the simulation calculation time of the matching parameter group numbered j; Standardize the grid density data group D and the simulation time data group T to construct a standard grid density data group Db and a standard simulation time data group Tb; Performing correlation analysis on the standard grid density data group Db and the standard simulation time data group Tb to determine whether the grid density corresponding to the standard grid density data group and the simulation time corresponding to the standard simulation time data group show a positive correlation; If there is a positive correlation, obtain the mesh density corresponding to the matching parameter group within the range value corresponding to the preset simulation calculation time; The mesh density corresponding to the matching parameter group within the range value corresponding to the simulation calculation time is integrated to obtain a simulation density group.

[0013] As a further technical solution of the present invention: the method for judging whether the grid density corresponding to the standard grid density data set and the simulation time corresponding to the standard simulation time data set present a positive correlation is as follows: The correlation coefficient r of the standard grid density data set Db and the standard simulation time data set Tb is calculated by the Pearson correlation coefficient; It is determined whether the correlation coefficient r of the standard grid density data group Db and the standard simulation time data group Tb is within a preset range. If so, it is considered that the grid density corresponding to the standard grid density data group and the simulation time corresponding to the standard simulation time data group are positively correlated.

[0014] As a further technical solution of the present invention: the subthreshold density data set is obtained by: Based on the different grid densities in the standard grid density data set Db, simulation training is performed inside the gravimeter system to obtain the simulation result errors under different grid densities; Obtain the simulation result errors under different grid density differences and construct a simulation error data set E; Performing standardization on the simulation error data group E to construct a standard simulation error data group Eb; Based on the standard grid density data group Db and the standard simulation error data group Eb, a scatter plot of grid density-simulation error is drawn in a two-dimensional rectangular coordinate system; The scatter plot of grid density-simulation error is fitted using the least squares method to obtain a fitting straight line, and it is determined whether the fitting effect of the fitting straight line meets expectations and whether the fitting straight line grows linearly; If the fitting effect of the fitting straight line meets expectations and the fitting straight line increases linearly, a preset simulation error threshold is obtained, and the simulation error threshold is used as a reference value; Based on the reference value, a reference line is drawn in the two-dimensional rectangular coordinate system where the scatter plot of mesh density-simulation error is located; Get the horizontal coordinate of the intersection of the reference line and the fitting line to get the grid density threshold; Based on the grid density threshold, the grid densities below the grid density threshold are integrated to obtain the subthreshold density data set.

[0015] As a further technical solution of the present invention: whether the fitting straight line meets expectations and whether the fitting straight line grows linearly is determined as follows: Based on the goodness of fit formula, obtain the goodness of fit R of the fitting line 2 ; If the goodness of fit R 2 If it is within the preset goodness range, the fitting effect is as expected; Get the positivity of the slope of the fitted line. If the slope of the fitted line is positive, it means that the fitted line shows a linear growth trend.

[0016] As a further technical solution of the present invention: the preferred method for determining the grid density is: Based on the subthreshold density data group and the simulation density group, a genetic algorithm is used to establish a fitness function and determine the optimal grid density; When the genetic algorithm reaches a preset number of iterations, the grid density at which the fitness function takes a minimum value is obtained as the preferred grid density.

[0017] As a further technical solution of the present invention: the fitness function is constructed as follows: Label each grid density in the subthreshold density data set as Y c , mark each mesh density in the simulation density group as Y g ; The fitness function of the genetic algorithm is ,in , are preset proportional coefficients, Y tar Optimal grid density for historical partitioning data.

[0018] Beneficial effects of the present invention: (1) Draw the geometric model of the gravimeter system, determine the material properties to establish the heat transfer physical field model, set the heat conduction and thermal radiation parameters, and analyze the temperature gradient inside the gravimeter; in the meshing module, use the finite element analysis method to mesh the interior of the gravimeter system, combine the historical meshing data, quantify the mesh boundary conditions and simulation parameters, and determine the historical meshing group, which can make full use of the historical data and make the new meshing more reasonable.

[0019] (2) Simulation training is performed based on different grid densities to obtain the simulation result error and construct a data group. A scatter plot is drawn and fitted for analysis to obtain the relationship between grid density and simulation error. The subthreshold density data group is then determined, which helps researchers understand the law between grid density and simulation error and provides a basis for improving simulation accuracy. Based on the subthreshold density data group and the simulation density group, a genetic algorithm is used to determine the optimal grid density. The grid density that minimizes the fitness function is found through iteration, which is conducive to integrating multiple factors, balancing computational efficiency and simulation accuracy, finding the most suitable grid density, and optimizing the performance of the internal temperature simulation of the gravimeter system. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The present invention will be further described below in conjunction with the accompanying drawings.

[0021] Figure 1 It is a module diagram of an internal temperature simulation system of a gravimeter system of the present invention; Figure 2 It is a flow chart of constructing an internal simulation model of a gravimeter system of the present invention; Figure 3 It is a flow chart of constructing a grid parameter model in a method for simulating internal temperature of a gravimeter system according to the present invention. DETAILED DESCRIPTION

[0022] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0023] Embodiment 1 like Figure 1 As shown, a gravimeter system internal temperature simulation system includes the following modules: Simulation construction module: used to construct a simulation model of the internal temperature of the gravimeter system and obtain the temperature gradient change field of the gravimeter system; like Figure 2 As shown in the figure, the construction method of the simulation model of the gravimeter system is as follows: S1. Draw the geometric model of the gravimeter system; In some embodiments, a three-dimensional finite element analysis software, ANSYS or ABAQUS, is used to perform a three-dimensional finite element analysis on the internal structure of the relative gravimeter system to draw a geometric model of the gravimeter system; S2, determine the material properties and establish a heat transfer physical field model; It should be noted that the corresponding material properties are assigned to the sealing materials and insulating materials existing in the gravimeter system; for the materials used, such as copper, PLA, FR4, and steel, the constant pressure heat capacity, density, and thermal conductivity are set as basic parameters; Establish a heat transfer physical field model including heat conduction, heat convection and heat radiation. For the existing air flow area, set the heat convection coefficient and boundary conditions to simulate the influence of natural or forced convection of air on the temperature field. S3, setting heat conduction and heat radiation parameters; Determine the anisotropy coefficient of heat conduction, as well as the emissivity and absorptivity of thermal radiation; It should be noted that when considering thermal radiation, it is necessary to accurately simulate the radiation heat transfer process between components and between components and the environment and set the simulation parameters; S4. Analyze the temperature gradient inside the gravimeter; Use the post-processing function of the simulation software to conduct an in-depth analysis of the calculated temperature field data, focusing on the temperature gradient distribution in the area where the inner temperature control module is located; By generating temperature distribution cloud maps and temperature change curves, the temperature change trends and gradient sizes at different locations inside the gravimeter can be intuitively displayed.

[0024] Based on the simulation model, the temperature gradient change field of the gravimeter system is constructed; It should be noted that the temperature gradient refers to the rate of change of temperature in space. In the gravimeter system, due to the different heating and heat dissipation conditions of different components, the temperature gradient will present different distribution states. By measuring the temperature inside the gravimeter system or building a temperature simulation model through simulation software, the temperature values ​​of each point in the gravimeter system can be obtained, and the temperature gradient change field can be obtained; The technical solution of this embodiment is: constructing a simulation model of the internal temperature of the gravimeter system, obtaining the temperature gradient change field of the gravimeter system, and providing data support for subsequent grid division.

[0025] Embodiment 2 like Figure 1 As shown, a gravimeter system internal temperature simulation system monitors the analysis grid of the temperature gradient in the gravimeter system, and also includes the following modules: Grid division module: used to grid the temperature gradient change field of the gravimeter system, obtain the grid density and division similarity value, and determine the historical division group of the grid; In some embodiments, the temperature gradient variation field of the gravimeter system is meshed using a finite element analysis method to obtain a density of the mesh, i.e., a mesh density; Obtain historical meshing data from the simulation model, wherein the historical meshing data includes: mesh density, boundary conditions, and simulation parameters; It should be noted that historical grid division data is obtained through simulation software; By formula: Obtain the heat flux q of the grid and quantify the grid boundary conditions. h is the convective heat exchange coefficient, h=5.26W / (m 2 / K), is the temperature difference of the grid; By formula: Get the Fourier number Fo of the grid and quantify the simulation parameters of the grid. α is the thermal diffusion coefficient, α=0.51. is the simulation time step of the grid, and L is the characteristic length of the grid; It should be noted that the simulation time step of the grid , the characteristic length L of the grid is obtained from the simulation model; The grid density, Fourier number, and heat flux are dimensionally processed, and the dimensionally processed results are weighted summed to obtain the partition similarity value; Obtain the standard deviation of the partition similarity value, and use the partition similarity value ±3 times the standard deviation as the preset similarity range; Obtaining the partition similarity values ​​of different grids, and merging the grids whose partition similarity values ​​are within a preset similarity range as a historical partition group; It should be noted that within the grid of the gravimeter system, there are multiple historical division groups; Matching and screening module: obtain the partition similarity values ​​of all grids of the simulation model, perform cluster analysis to obtain the grid parameter group, compare and screen the grid parameter group with the historical partition group to obtain the matching parameter group; From the simulation model, the similarity value of the grid division output by the simulation model is obtained, and the grid parameter model is constructed by using the K-Means clustering analysis algorithm to obtain the grid parameter group; Numerical analysis is performed on the grid similarity values ​​of the grid parameter groups to obtain the overlap matching values; like Figure 3 As shown in Figure 2, the grid parameter model is constructed as follows: S1, obtain the grid similarity value output by the simulation model and construct the grid vector X; Specifically, construct the grid vector X, , n is the total number of grids; S2, obtain the number of clusters K based on the elbow rule; Preferably, the number of clusters K=2; S3, based on the grid vector X, execute the K-Means algorithm; Randomly select K grids with similar partition values ​​as the cluster centers of the initial clusters; Get the partition similarity value x of each grid i, calculate the Euclidean distance from each grid to the cluster center, and assign it to the cluster cluster with the nearest cluster center according to the Euclidean distance of the grid; S4, repeat S3 until the cluster centers meet the convergence condition, and mark the clusters as grid parameter groups based on different clusters; It should be noted that the convergence condition is set by professionals in this field based on experience; Obtaining a grid parameter group and a historical division group in the gravimeter system, and if the grid parameter group and multiple historical division groups overlap in space, obtaining the overlapping area of ​​the grid parameter group and multiple historical division groups; Select the historical division group with the largest overlapping area between the grid parameter group and multiple historical division groups, bind the grid parameter group and the historical division group with the largest overlapping area one by one, and obtain a grid comparison group; Get the grids that appear in both the grid parameter group and the historical division group in the grid comparison group and mark them as overlapping grids; The number of grids in the grid parameter group and the number of overlapping grids are obtained, and the number of overlapping grids is processed by ratio processing with the number of grids in the grid parameter group to obtain the overlapping grid ratio; Get the overlapping grid ratio of all grid comparison groups, marked as W i , i is the number of the grid comparison group, and the value range of i is [1,m]; Obtaining the division similarity values ​​of all overlapping grids in the grid comparison group, and summing up the division similarity values ​​of all overlapping grids to obtain the overlapping similarity value; The partition similarity values ​​within the grid parameter group are summed to obtain the parameter similarity value; The overlap similarity value is compared with the parameter similarity value to obtain the overlap similarity ratio, which is marked as C. i ; By formula: Get the overlapping match value Pz i , where a1=0.264, a2=0.891; If the overlap match value Pz of each grid comparison group j If the overlap matching threshold is higher than the preset threshold, the grid parameter group in the grid comparison group is marked as a matching parameter group, otherwise, no processing is performed; It should be noted that the purpose of establishing the matching parameter group is to compare the partition similarity values ​​in the historical partition groups with the partition similarity values ​​in the current simulation data, and to screen out matching grids corresponding to the partition similarity values ​​in multiple historical partition groups; The technical solution of this embodiment is: gridding the gravimeter system, determining the historical division group of the grid, obtaining the division similarity values ​​of all grids of the simulation model, performing cluster analysis to obtain the grid parameter group, comparing and screening the grid parameter group with the historical division group, obtaining the matching parameter group, and improving the data basis for subsequent grid analysis.

[0026] Embodiment 3 like Figure 1 As shown, a gravimeter system internal temperature simulation system also includes the following modules: Correlation analysis module: used to obtain the grid density values ​​and corresponding simulation calculation time of all matching parameter groups, construct the standard grid density data group Db and the standard simulation time data group Tb, and perform correlation analysis. If the grid density is related to the simulation calculation time, a simulation density group is constructed; Obtain the grid density value of each matching parameter group in the gravimeter system, and obtain the simulation calculation time of different grid density values ​​in multiple historical division groups; The grid density values ​​of different matching parameter groups are used as the first data element to construct the grid density data group D. , where d j Represents the grid density value of the matching parameter group numbered j; The simulation calculation time corresponding to different grid densities is used as the second data element to construct a simulation time data set T. , where t j represents the simulation calculation time of the matching parameter group numbered j; Standardize the grid density data group D and the simulation time data group T to construct a standard grid density data group Db and a standard simulation time data group Tb; For example, construct a standard grid density data set , standard simulation time data set ; Specifically, the method of standardizing the grid density data group D and the simulation time data group T is as follows: By formula: Get the normalized grid density value db of the matching parameter group numbered j j , and construct a standard grid density data set Db; By formula: Get the standardized simulation calculation time tb of the matching parameter group numbered j j , and construct a standard simulation time data group Db; Performing correlation analysis on the standard grid density data group Db and the standard simulation time data group Tb to determine whether the grid density corresponding to the standard grid density data group and the simulation time corresponding to the standard simulation time data group show a positive correlation; Based on the standard grid density data group Db and the standard simulation time data group Tb, the correlation coefficient r of the standard grid density data group Db and the standard simulation time data group Tb is calculated by the Pearson correlation coefficient; By formula: Obtain the correlation coefficient r of the standard grid density data group Db and the standard simulation time data group Tb; in, , , the value range of Pearson correlation coefficient r is [-1,1]; If the correlation coefficient r between the standard grid density data set Db and the standard simulation time data set Tb is in the range of (0,1], it is considered that the grid density corresponding to the standard grid density data set and the simulation time corresponding to the standard simulation time data set are positively correlated, that is, when the grid density increases, the simulation calculation time will also increase; If the correlation coefficient r between the standard grid density data set Db and the standard simulation time data set Tb is in the range of [-1,0), it is considered that the grid density corresponding to the standard grid density data set and the simulation time corresponding to the standard simulation time data set are negatively correlated, that is, when the grid density increases, the simulation calculation time will decrease; If the correlation coefficient between the standard grid density data set Db and the standard simulation time data set Tb is 0, it is considered that there is no correlation between the grid density corresponding to the standard grid density data set and the simulation time corresponding to the standard simulation time data set; If the grid density corresponding to the standard grid density data group and the simulation time corresponding to the standard simulation time data group show a positive correlation, the grid density corresponding to the preset simulation calculation time range value is obtained; Integrate the mesh density corresponding to the matching parameter group within the range value corresponding to the simulation calculation time to obtain a simulation density group; It should be noted that the simulation density group is established to analyze the relationship between grid density and simulation calculation time, and to provide data support for the subsequent determination of the optimal grid density; Error analysis module: used to obtain the grid density data set, the corresponding simulation result error, perform fitting analysis on the grid density and simulation result error, and determine the subthreshold density data set; Based on the different grid densities in the standard grid density data set Db, simulation training is performed inside the gravimeter system to obtain the simulation result errors under different grid densities; In some embodiments, the simulation result error is obtained by obtaining the experimental result and performing difference processing on the simulation result; Obtain the simulation result errors under different grid density differences and construct the simulation error data set E. , where e jIndicates the simulation result error numbered j, j and the standard grid density data set The number corresponds to; By formula: , standardize the simulation error data set E and construct the standard simulation error data set Eb, where eb j represents the simulation result error numbered j after normalization; Based on the standard grid density data group Db and the standard simulation error data group Eb, a scatter plot of grid density-simulation error is drawn in a two-dimensional rectangular coordinate system with the X axis as the grid density and the Y axis as the simulation error; The scatter plot of grid density-simulation error is fitted using the least squares method to obtain a fitting straight line, and it is determined whether the fitting effect of the fitting straight line meets expectations and whether the fitting straight line grows linearly; The equation of the fitted straight line is: , where k is the slope of the fitted line, is the fitting value, q is the intercept of the fitting line; By formula: Get the goodness of fit R of the fitted line 2 ; The goodness of fit R of the fitted line 2 Compare with the goodness range value to determine whether the fitting effect of the fitting straight line meets expectations; If the goodness of fit R 2 If it is within the goodness range, it is considered that the fitting effect of the fitting line meets expectations, otherwise, it does not meet expectations; Preferably, the goodness of fit range is [0.65, 1]; Get the positive or negative value of the slope k of the fitted line. If the slope k of the fitted line is positive, it means that the fitted line shows a linear growth trend. If it is negative, it means that the fitting line shows a linear decreasing trend; If the mesh density and the simulation error show a linear growth trend, a preset simulation error threshold is obtained, and the simulation error threshold is used as a reference value to draw a reference line in a two-dimensional rectangular coordinate system where a scatter plot of mesh density-simulation error is located; Get the horizontal coordinate of the intersection of the reference line and the fitting line to get the grid density threshold; Based on the grid density threshold, the grid density below the grid density threshold is taken as the data element to construct the subthreshold density data group; It should be noted that the subthreshold density data set is established to analyze the relationship between grid density and simulation error, and to provide data support for the subsequent determination of the optimal grid density; Density determination module: used to train the subthreshold density data group and the simulation density group using a genetic algorithm to determine the optimal grid density; Based on the subthreshold density data group and the simulation density group, a genetic algorithm is used to establish a fitness function and determine the optimal grid density; Label each grid density in the subthreshold density data set as Y c , the value range of c is [1,z]; Label each mesh density in the simulation density group as Y g , the value range of g is [1,p]; It should be noted that z and p are both smaller than m and n; The fitness function of the genetic algorithm is ,in , They are 0.86, 0.45, and Y tar Optimal grid density for historical partitioning data; When the genetic algorithm reaches a preset number of iterations, the grid density at which the fitness function takes the minimum value is obtained, which is the most preferred grid density; It should be noted that the number of iterations is set by the professional and technical personnel of the present invention based on experience; The technical solution of this embodiment is: obtain the grid density values ​​and the corresponding simulation calculation time of all matching parameter groups, construct a standard grid density data group Db and a standard simulation time data group Tb, and perform a correlation analysis. If the grid density is related to the simulation calculation time, construct a simulation density group, obtain the grid density data group and the corresponding simulation result error, perform a fitting analysis on the grid density and the simulation result error, determine the subthreshold density data group, and use a genetic algorithm to train the subthreshold density data group and the simulation density group to determine the preferred grid density, which is conducive to determining the preferred grid density and optimizing the grid division of the simulation model.

[0027] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the present invention.

Claims

1. A gravimeter system internal temperature simulation system, characterized in that: Includes the following modules: Simulation construction module: used to construct a simulation model of the internal temperature of the gravimeter system and obtain the temperature gradient change field of the gravimeter system; Grid division module: used to grid the temperature gradient change field of the gravimeter system, obtain the grid density and division similarity value, and determine the historical division group of the grid; Matching and screening module: used to obtain the similarity values ​​of all grid divisions, perform cluster analysis to obtain grid parameter groups, compare the grid parameter groups with the historical division groups, and obtain matching parameter groups; Correlation analysis module: used to obtain the grid density value and the corresponding simulation calculation time of the matching parameter group, construct a standard grid density data group and a standard simulation time data group, and perform correlation analysis to obtain a simulation density group; Error analysis module: used to obtain the simulation result error, perform fitting analysis on the grid density and the simulation result error, and determine the subthreshold density data group; Density determination module: used to train the subthreshold density data group and the simulation density group using a genetic algorithm to determine the optimal grid density.

2. The internal temperature simulation system of a gravimeter system according to claim 1, characterized in that: The method for obtaining the historical division group is as follows: Obtain historical meshing data from the simulation model, wherein the historical meshing data includes: mesh density, boundary conditions, and simulation parameters; The simulation parameters include: the characteristic length L of the grid, the simulation time step of the grid ; By formula: Get the heat flux q of the grid, quantify the boundary conditions of the grid, h is the convective heat exchange coefficient, is the temperature difference of the grid; By formula: Get the Fourier number Fo of the grid and quantify the simulation parameters of the grid. α is the thermal diffusion coefficient. The grid density, Fourier number, and heat flux are dimensionally processed, and the dimensionally processed results are weighted summed to obtain the partition similarity value; The partition similarity values ​​of different grids are obtained, and the grids whose partition similarity values ​​are within a preset similarity range are merged as a historical partition group.

3. The internal temperature simulation system of a gravimeter system according to claim 1, characterized in that: The matching parameter group is obtained in the following manner: From the simulation model, the grid division similarity value is obtained, and the grid parameter model is constructed through the K-Means clustering analysis algorithm to obtain the grid parameter group; Combined with the historical division group, numerical analysis is performed on the division similarity values ​​of the grid parameter group to obtain the grid comparison group and overlap matching value; If the overlap matching value of each grid comparison group is higher than a preset overlap matching threshold, the grid parameter group in the grid comparison group is marked as a matching parameter group.

4. The internal temperature simulation system of a gravimeter system according to claim 3, characterized in that: The overlapping matching value and the grid comparison group are obtained as follows: Obtain the overlapping area of ​​the grid parameter group and multiple historical division groups in the gravimeter system; Select the historical division group with the largest overlapping area between the grid parameter group and multiple historical division groups, bind the grid parameter group and the historical division group with the largest overlapping area one by one, and obtain a grid comparison group; Get the grids that appear in both the grid parameter group and the historical division group in the grid comparison group and mark them as overlapping grids; The number of grids in the grid parameter group and the number of overlapping grids are obtained, and the number of overlapping grids is ratioed to the number of grids in the grid parameter group to obtain an overlapping grid ratio; Get the overlapping grid ratio of all grid comparison groups, marked as W i , i is the number of the grid comparison group, and the value range of i is [1,m]; Obtaining the division similarity values ​​of all overlapping grids in the grid comparison group, and summing up the division similarity values ​​of all overlapping grids to obtain the overlapping similarity value; The partition similarity values ​​within the grid parameter group are summed to obtain the parameter similarity value; The overlap similarity value is compared with the parameter similarity value to obtain the overlap similarity ratio, which is marked as C. i ; By formula: Get the overlapping match value Pz i , where a1 and a2 are preset proportional coefficients.

5. The internal temperature simulation system of a gravimeter system according to claim 1, characterized in that: The method for obtaining the simulation density group is as follows: Obtain the grid density value of each matching parameter group in the gravimeter system, and obtain the simulation calculation time of different grid density values ​​in multiple historical division groups; The grid density values ​​of different matching parameter groups are used as the first data element to construct a grid density data group D. , where d j Represents the grid density value of the matching parameter group numbered j; The simulation calculation time corresponding to different matching parameter groups is used as the second data element to construct a simulation time data group T. , where t j represents the simulation calculation time of the matching parameter group numbered j; Standardize the grid density data group D and the simulation time data group T to construct a standard grid density data group Db and a standard simulation time data group Tb; Performing correlation analysis on the standard grid density data group Db and the standard simulation time data group Tb to determine whether the grid density corresponding to the standard grid density data group and the simulation time corresponding to the standard simulation time data group show a positive correlation; If there is a positive correlation, obtain the mesh density corresponding to the matching parameter group within the range value corresponding to the preset simulation calculation time; The mesh density corresponding to the matching parameter group within the range value corresponding to the simulation calculation time is integrated to obtain a simulation density group.

6. The internal temperature simulation system of a gravimeter system according to claim 5, characterized in that: The method for judging whether the grid density corresponding to the standard grid density data set and the simulation time corresponding to the standard simulation time data set present a positive correlation is as follows: The correlation coefficient r of the standard grid density data set Db and the standard simulation time data set Tb is calculated by the Pearson correlation coefficient; It is determined whether the correlation coefficient r of the standard grid density data group Db and the standard simulation time data group Tb is within a preset range. If so, it is considered that the grid density corresponding to the standard grid density data group and the simulation time corresponding to the standard simulation time data group are positively correlated.

7. The internal temperature simulation system of a gravimeter system according to claim 5, characterized in that: The subthreshold density data set is obtained in the following manner: Based on the different grid densities in the standard grid density data set Db, simulation training is performed inside the gravimeter system to obtain the simulation result errors under different grid densities; Obtain the simulation result errors under different grid density differences and construct a simulation error data set E; Performing standardization on the simulation error data group E to construct a standard simulation error data group Eb; Based on the standard grid density data group Db and the standard simulation error data group Eb, a scatter plot of grid density-simulation error is drawn in a two-dimensional rectangular coordinate system; The scatter plot of grid density-simulation error is fitted using the least squares method to obtain a fitting straight line, and it is determined whether the fitting effect of the fitting straight line meets expectations and whether the fitting straight line grows linearly; If the fitting effect of the fitting straight line meets expectations and the fitting straight line increases linearly, a preset simulation error threshold is obtained, and the simulation error threshold is used as a reference value; Based on the reference value, a reference line is drawn in the two-dimensional rectangular coordinate system where the scatter plot of mesh density-simulation error is located; Get the horizontal coordinate of the intersection of the reference line and the fitting line to get the grid density threshold; Based on the grid density threshold, the grid densities below the grid density threshold are integrated to obtain the subthreshold density data set.

8. The internal temperature simulation system of a gravimeter system according to claim 7, characterized in that: The method for determining whether the fitting straight line meets expectations and whether the fitting straight line grows linearly is as follows: Based on the goodness of fit formula, obtain the goodness of fit R of the fitting line 2 ; If the goodness of fit R 2 If it is within the preset goodness range, the fitting effect is as expected; Get the positivity of the slope of the fitted line. If the slope of the fitted line is positive, it means that the fitted line shows a linear growth trend.

9. The internal temperature simulation system of a gravimeter system according to claim 5, characterized in that: The preferred method for determining the grid density is: Based on the subthreshold density data group and the simulation density group, a genetic algorithm is used to establish a fitness function and determine the optimal grid density; When the genetic algorithm reaches a preset number of iterations, the grid density at which the fitness function takes a minimum value is obtained as the preferred grid density.

10. The internal temperature simulation system of a gravimeter system according to claim 9, characterized in that: The fitness function is constructed as follows: Label each grid density in the subthreshold density data set as Y c , mark each grid density in the simulation density group as Yg; The fitness function of the genetic algorithm is ,in , are the preset scale coefficients respectively, and Ytar is the optimal grid density of the historical partitioning data.

Images