DEM simulation model automatic calibration method and system based on parallel computing

Abstract

The invention relates to the field of simulation model calibration, and provides a parallel computing-based DEM simulation model automatic calibration method, which comprises the steps of acquiring physical macroscopic features, generating multiple groups of initial simulation microscopic parameters, performing parallel simulation and data processing, acquiring simulation macroscopic features, and establishing an intelligent model between the simulation macroscopic features and the physical macroscopic features. Inputting the physical macroscopic features into the constructed intelligent model, obtaining corresponding simulation microscopic parameters as initial values, carrying out parallel simulation, data analysis and comparison on the initial values, obtaining the microscopic parameters corresponding to the simulation macroscopic features with the minimum error as initial values of the next round, and carrying out parallel simulation, data analysis and comparison on the initial values; performing parallel simulation, data analysis and comparison on the initial value circularly until the relative error between the simulation macroscopic feature and the physical macroscopic feature of the round is lower than a set threshold value or the correction frequency reaches a set upper limit, and completing calibration; based on parallel calculation, the calibration efficiency of simulation parameters can be greatly improved.

Description

Technical Field

[0001] The present invention relates to the field of simulation model calibration, and in particular to a DEM simulation model automatic calibration method based on parallel computing. Background Art

[0002] With the rapid development of my country's economy, rotary kilns, dryers, coating machines, and other rotary drum equipment are playing a vital role in the cement, metallurgy, and pharmaceutical industries. The core functions of these devices, such as mixing, drying, and coating materials, are crucial for improving product quality and production efficiency. However, due to the complex properties of granular materials, which combine both solid and fluid properties, the mechanism of material movement within the drum remains unclear.

[0003] With the rapid development of advanced computing technology, the Discrete Element Method (DEM) has become an important technical means to analyze the complex physical processes of these particulate materials. By utilizing DEM, the movement and interaction of a large number of particles can be simulated under specific conditions, thereby providing effective support for understanding and improving the material handling process in rotary drum equipment. DEM-based simulations allow researchers to deeply explore the impact of key operating conditions and material properties on mixing efficiency without relying on actual production experiments, which largely helps avoid the high costs and economic losses that may be caused by traditional experimental research.

[0004] To ensure the accuracy of DEM model simulation results, the model needs to be calibrated. This involves continuously adjusting the simulation input parameters so that the simulated macrofeatures match those obtained from physical experiments. However, current DEM model parameter calibration often relies on manual operation, a cumbersome and inefficient process that also makes accuracy difficult to guarantee. Our team proposed a data-driven parameter calibration method in their Chinese invention patent application, CN202310828095.5, entitled "A Method, Apparatus, Device, and Storage Medium for Calibrating Simulation Parameters of Granular Materials." However, this method simulates only one set of parameters at a time, obtains the simulated macrofeatures, and then matches them with the macrofeatures from the physical experiment. If the relative error is less than a set value, calibration is terminated. Otherwise, parameter adjustments are performed and simulation is repeated until the stopping condition is met. However, this method typically requires multiple rounds of parameter adjustment and simulation to complete DEM model parameter calibration, representing a serial parameter calibration method whose efficiency needs to be further improved.

[0005] In order to solve the problem of low efficiency of current DEM model calibration, a DEM simulation model automatic calibration method based on parallel computing is needed, which can greatly improve the calibration efficiency of simulation parameters. Summary of the Invention

[0006] The problem solved by the present invention is how to provide a DEM simulation model automatic calibration method based on parallel computing, which can greatly improve the calibration efficiency of simulation parameters.

[0007] To solve the above problems, the present invention provides a DEM simulation model automatic calibration method based on parallel computing, comprising: obtaining physical macro features, generating multiple sets of initial simulation micro parameters when the correction number count = 0, performing parallel simulation and data processing on the multiple sets of initial simulation micro parameters, obtaining simulation macro features corresponding to the initial simulation micro parameters, sending the multiple sets of simulation macro features and physical macro features to a comparison and optimization module, judging whether the relative error between the simulation macro features and the physical macro features is lower than a set threshold, and completing the calibration if so; if not, establishing an intelligent model between the simulation macro features and the physical macro features based on the simulation data during the parallel simulation of the initial simulation micro parameters;

[0008] When the number of corrections count = 1, the physical macro-features are input into the constructed intelligent model, and the corresponding simulation micro-parameters are obtained as initial values. Parallel simulation, data analysis and comparison of the initial values ​​are performed, and the micro-parameters corresponding to the simulation macro-features with the smallest error are obtained as the initial values ​​of the next round. Parallel simulation, data analysis and comparison of the initial values ​​are repeated until the relative error between the simulation macro-features and the physical macro-features of this round is lower than the set threshold or the number of corrections count reaches the set upper limit, and the calibration is completed.

[0009] Furthermore, the obtaining of physical macroscopic characteristics includes:

[0010] The images of the movement process of the granular material in the rotating drum are collected through the physical experiment module and sent to the data processing and analysis module for filtering and binarization. The repose angle data at different times are extracted, and the acquired repose angle data are fitted using a second-order model to obtain the dynamic repose angle as a physical macroscopic feature.

[0011] Furthermore, when the correction count=0, generating multiple sets of initial simulation micro parameters, and performing parallel simulation and data processing on the multiple sets of initial simulation micro parameters includes:

[0012] Use the parallel simulation parameter generation module to generate a parameter between 0 and 1 with a step size of Δμ b The input parameter μ b (1) ,…,μ b (k) ,μ b (k+1) ,…,μ b (n) , where μ b (1) =Δμ b,μ b (k+1) =μ b (k) +Δμ b (k>0), And create corresponding simulation folders in batches;

[0013] Use the DEM simulation module to perform parallel simulation and collect simulation data;

[0014] The simulation data is sent to the data processing and analysis module to obtain the simulation macro characteristics corresponding to each input parameter.

[0015] Furthermore, the establishment of an intelligent model between the simulated macro features and the physical macro features includes:

[0016] Establishing a simulation database between initial simulation microscopic parameters and their corresponding simulation macroscopic features;

[0017] Preprocess the data, remove outliers, and organize the data;

[0018] The least squares method is used as the core fitting algorithm and the polynomial function is used as the fitting model to construct an intelligent model between the simulation macro-features and the physical macro-features.

[0019] Furthermore, the elimination of abnormal values ​​includes using a moving average method to obtain a smoothed value of the simulation macro-features of the simulation output, and judging whether the distance between the simulation macro-features and the smoothed value in each group of simulation data is greater than a set value. If so, the group of simulation data pairs is removed, otherwise it is retained.

[0020] Furthermore, the data sorting includes rearranging the simulation data after outliers are removed, with the simulation macroscopic features as input and the simulation microscopic parameters as output.

[0021] Furthermore, the parallel simulation, data analysis and comparison of the initial values ​​include:

[0022] The initial value μ b0 As input, μ b0 Centered by Δμ b , generate equally spaced microscopic parameters to be simulated: μ b (1) ,…,μ b (i) ,…,μ b (k+1) ,…,μ b (n) , where: μ b (k+1) =μ b0 ,μ b (i)=μ b0 +(ik-1)Δμ b (i=1,...,n),n=2k+1;

[0023] Based on the microscopic parameters to be simulated, batch create corresponding simulation folders;

[0024] Using the microscopic parameters to be simulated, DEM simulation is performed in parallel to generate simulation data;

[0025] Input the simulation data into the data processing and analysis module to obtain the macro characteristics of each simulation in this round;

[0026] Determine whether the relative error between the simulated macro features and the physical macro features in this round is lower than the set threshold. If so, end the calibration. If not, select the micro parameters corresponding to the simulated macro features with the smallest error as the initial values ​​for the next round.

[0027] Furthermore, the parallel execution of DEM simulation includes:

[0028] Open n DEM simulation software interfaces, and run the microscopic parameters μ to be simulated in each interface. b (1) ,…,μ b (n) The corresponding DEM simulation code; and store the corresponding simulation data, images and other related parameters.

[0029] Furthermore, when the calibration is completed, if the termination calibration condition is that the relative error between the simulated macro-features and the physical macro-features of this round is lower than the set threshold, the micro-parameter corresponding to the minimum relative error is selected as the final calibration value; if the termination calibration condition is that the number of corrections count reaches the set upper limit, the micro-parameter corresponding to the minimum relative error in all rounds is selected as the final calibration value.

[0030] The present invention also provides a DEM simulation model automatic calibration system based on parallel computing, which is used to execute the above-mentioned DEM simulation model automatic calibration method based on parallel computing, including:

[0031] Physical experiment module, used to obtain physical experiment data;

[0032] Data processing and analysis module, used to obtain physical macro characteristics based on physical experimental data, obtain simulation macro characteristics based on simulation data, and build intelligent models;

[0033] A parallel simulation parameter generation module is used to generate corresponding simulation microscopic parameters according to the number of corrections;

[0034] A DEM simulation module is used to perform parallel DEM simulation processing on multiple groups of simulation microscopic parameters generated by the parallel simulation parameter generation module;

[0035] and a comparison and selection module for judging the completion of calibration and outputting the final calibration value.

[0036] Compared with the prior art, the present invention has the following beneficial effects:

[0037] The present invention initially generates multiple groups of equally spaced simulation micro-parameters, performs parallel simulation, and establishes an intelligent model. At the same time, during subsequent calibration, multiple groups of parameters to be simulated are generated with the same step length, with the initial value as the center, and cyclic parallel simulation, data processing and comparison are performed to output the final calibration value. The above-mentioned parallel data processing method can effectively solve the problems of traditional calibration methods being cumbersome, complicated and time-consuming. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a schematic diagram of the overall principle structure of the present invention;

[0039] Figure 2 A table diagram of physical parameters of parallel simulation according to an embodiment of the present invention;

[0040] Figure 3 This is a table diagram of a simulation database when the number of corrections is 0 according to the first embodiment of the present invention;

[0041] Figure 4 A comparison table of the intelligent model simulation database and physical experiment results of the first embodiment of the present invention;

[0042] Figure 5 This is a table diagram of the intelligent model simulation database of the first embodiment of the present invention;

[0043] Figure 6 Schematic diagram of the fitting results of the intelligent model of the first embodiment of the present invention;

[0044] Figure 7 This is a table diagram of parallel simulation parameters when the number of corrections is 1 according to the first embodiment of the present invention;

[0045] Figure 8 This is a table diagram of simulation macro parameters when the number of corrections is 1 according to the first embodiment of the present invention;

[0046] Figure 9 This is a table comparing the parallel simulation results and the physical experiment results when the number of corrections is 1 according to the first embodiment of the present invention;

[0047] Figure 10 A table diagram of a simulation database according to a second embodiment of the present invention;

[0048] Figure 11 This is a table comparing the parallel simulation results and physical experiment results of the second embodiment of the present invention. DETAILED DESCRIPTION

[0049] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0050] In the description of the present invention, it should be noted that, unless otherwise expressly specified or limited, the terms "disposed," "installed," "connected," and "connected" should be understood broadly. For example, they may refer to fixed, detachable, or integral connections; mechanical connections; direct connections or indirect connections via an intermediate medium; and internal communication between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on specific circumstances.

[0051] Throughout this specification, references to the terms "embodiment," "one embodiment," and "an implementation" indicate that the specific features, structures, materials, or characteristics described in conjunction with that embodiment or implementation are included in at least one embodiment or implementation of the present invention. In this specification, schematic representations of these terms do not necessarily refer to the same embodiment or implementation. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or implementations.

[0052] like Figure 1 As shown, the present invention provides a DEM simulation model automatic calibration system based on parallel computing, which mainly includes five modules: physical experiment, data processing and analysis, parallel simulation parameter generation, DEM simulation, and comparison and selection. The main steps of its automatic calibration method are as follows:

[0053] Step 1: Use the physical experimental device to obtain the experimental data, images and other relevant parameters of the calibration object, and perform experimental data processing and analysis to obtain the macroscopic characteristics of the physical experiment fexp.

[0054] Step 2: When the calibration times count = 0, use the parallel simulation parameter generation module to generate a value between 0 and 1 with a step size of Δμ b The input parameter μ b (1) ,…,μ b (k) ,μ b (k+1) ,…,μ b (n) , where μ b (1) =Δμ b ,μ b (k+1) =μ b (k) +Δμb (k>0), And create corresponding simulation folders in batches;

[0055] Then, the DEM simulation module is used to perform parallel simulation and collect simulation data; the data is sent to the data processing and analysis module to obtain the macroscopic characteristics corresponding to each simulation. s1 , f s2 ,…,f sn ; Combine the simulation macro features and the physical experimental macro features obtained in Step 1 exp The data is sent to the comparison and optimization module to determine whether the relative error e between the simulated macroscopic features and the physical features is lower than the set threshold E%. If it is lower than the set threshold, the calibration is terminated directly, and the microscopic parameter corresponding to the minimum relative error e is output as the calibration value. Otherwise, based on the data obtained from the simulation when count = 0, an intelligent model between the dynamic repose angle and the friction coefficient is established, and the calibration is continued in the order of Step 2 to Step 5.

[0056] If count=1, the macroscopic physical characteristics f exp Input into the constructed intelligent model to predict the corresponding initial value μ of the simulation microscopic parameter b0 ; Otherwise, continue.

[0057] The initial value μ b0 As input, it is sent to the parallel simulation parameter generation module. b0 Centered by Δμ b , generate equally spaced microscopic parameters to be simulated: μ b (1) ,…,μ b (i) ,…,μ b (k+1) ,…,μ b (n) , where: μ b (k+1) =μ b0 ,μ b (i) =μ b0 +(ik-1)Δμ b (i=1,...,n),n=2k+1;

[0058] Based on the generated simulation parameters, corresponding simulation folders are created in batches, which contain the main program codes required for simulation and the folder for saving simulation data.

[0059] Step 3: Use the simulation parameters generated in Step 2 to perform DEM simulation in parallel, that is, the calibration platform opens n DEM simulation software interfaces, and each interface runs μ b (1) ,…,μ b (n) The corresponding DEM simulation code; and store the corresponding simulation data, images and other related parameters.

[0060] Step 4: Input the simulation data generated in Step 3 into the data processing and analysis module to extract the macro features of each simulation.

[0061] Step 5: The simulation macro features f obtained in Step 4 s1 , f s2 ,…,f sn and the physical macroscopic characteristics f obtained in Step 1 exp Send it to the comparison and optimization module. Determine whether the relative error e between the simulated macro features and the physical macro features in the current round is lower than the set threshold E%. If the minimum relative error e is less than E%, the micro parameter corresponding to the minimum relative error e is selected as the final calibration value and the calibration process is terminated. If the minimum relative error e in the current round is still greater than E%, then determine whether the calibration round count reaches the set upper limit C. If so, the calibration is terminated and the micro parameter with the smallest error in all previous simulations is selected as the final calibration value; otherwise, the micro parameter corresponding to the minimum relative error e is used as the new input initial value μ b0 , run Step 2 to Step 5 again to continue calibration.

[0062] The above process continues iterating until one of the following stopping conditions is met: the relative error e falls below the set threshold E%, or the number of calibration rounds count reaches the set upper limit C. Finally, the microscopic parameters with the minimum error are output as the calibration result. Repeat the above steps until the stopping condition is met, completing the model calibration.

[0063] Further, two embodiments with different particle sizes are selected for detailed description, such as Figure 2 As shown, the length and diameter of the cylinder in the embodiment are 67 mm and 250 mm respectively, and the cylinder speed is 5.91 rpm; the particle material is acrylic spherical particles, and the calibrated microscopic parameter is the friction coefficient μ of the particle material. b , the macroscopic feature used is the dynamic repose angle f.

[0064] Example 1:

[0065] Step 1: First, the image of the granular material movement process in the rotating drum is collected through the physical experiment module and sent to the data processing and analysis module for filtering, binarization, and other processing to extract the repose angle data at different times [10,11]. Then, the acquired repose angle data is fitted using the second-order model to obtain the dynamic repose angle f exp =15.4°.

[0066] Step 2: At this time, count = 0, use the parallel simulation parameter generation module to generate a step size of Δμ b =0.05, n=19, 19 groups of friction coefficients μ with equal spacing between 0 and 1 b and the corresponding simulation folders, such as Figure 3 As shown, the DEM simulation module is used for parallel simulation.

[0067] 1) Extraction of macroscopic features. The 19 sets of simulated images were processed and analyzed, and the steps were consistent with the method of extracting physical macroscopic features. The simulation input microscopic parameters (friction coefficient μ b ) and output macroscopic characteristics (dynamic repose angle f s ) between simulation databases, such as Figure 3 As shown;

[0068] The dynamic repose angle f obtained by calculation simulation s and the physical dynamic repose angle f exp The relative error e between them is calculated as follows:

[0069]

[0070] Some results can be found in Figure 4 As shown, the relative errors e were all less than the set threshold of 5%, and the minimum relative error was e=10.64%, which did not meet the condition for ending calibration.

[0071] 2) Data preprocessing

[0072] a) Eliminate outliers. First, use a 5-point moving average smoothing filter to process the dynamic repose angle of the simulation output when it tends to a stable state; then, determine whether the distance between the dynamic repose angle of the simulation output and the smoothing value is greater than the set value H (H = 1° in this example). If yes, remove the set of simulation data pairs to ensure the reliability of the modeling data; otherwise, retain it. For example, in this example Figure 3 The dynamic repose angle and smoothing value obtained from the 6th simulation were 29.92° and 24.512°, respectively, with a distance greater than 1° between the two values. The dynamic repose angle and smoothing value obtained from the 12th simulation were 27.84° and 29.144°, respectively, with a distance greater than 1° between the two values. Therefore, these two data sets were eliminated.

[0073] b) Data sorting. After removing outliers, the macroscopic features fs are used as input, and the microscopic parameters μ b As output, rearranged, it is used to build intelligent models, such as Figure 5 shown.

[0074] This embodiment uses the least squares method (LSM) as the core fitting algorithm and the polynomial function as the fitting model to construct the dynamic repose angle f s Friction coefficient with particles μ b The simulation database used for modeling is as follows: Figure 5 shown.

[0075] The polynomial model can have different orders. In order to ensure that the fitting results have high accuracy and avoid negative values, the polynomial order that performs best in these two aspects was finally selected.

[0076] The fitting results are as follows Figure 6 As shown, its mathematical expression is:

[0077] μ b =0.0000012656f s 5 -0.0000805962f s 4 +0.0017596225f s 3 -0.0156296032f s 2 +0.0617626772f s ;

[0078] Step 2: count = 1, the physical repose angle f collected by the physical experiment module exp =15.4° is used as the input of the intelligent model to obtain the initial prediction value of the micro friction coefficient. Use parallel simulation parameters to generate modules, As the center, Δμ b =0.02 as the step size, and use the method described in Step 3 to generate 5 simulation input friction coefficients μ b (n=5), such as Figure 7 As shown:

[0079] In order to realize the automatic execution of the algorithm, this paper builds a DEM simulation parameter generation module based on the LabVIEW platform. Combined with the set step size Δμ band n, automatically creates a corresponding number of simulation directories simulation_i (i=1,…,n,n=5). Each directory contains the input files and script codes required to run the DEM simulation, such as the main program main.dat and the subroutine function function.dat, and has the ability to run independently.

[0080] Step 3: The calibration platform will open five DEM simulation software interfaces at the same time and call the main program main.dat in the generated simulation directories simulation_1, simulation_2, ... for simulation respectively.

[0081] Step 4: Subsequently, the images and particle information obtained by simulation are transferred to the MATLAB environment through the built LabVIEW automation platform for further image processing and feature extraction operations, thereby realizing the recognition and extraction of the dynamic repose angle of the simulated macroscopic features.

[0082] The simulation images obtained from the five groups of simulations were processed and analyzed to obtain the corresponding simulation repose angle data as follows: Figure 8 shown.

[0083] Step 5: Calculate the dynamic repose angle f for each simulation s and the physical dynamic repose angle f exp (f exp =15.4°), such as Figure 9 shown.

[0084] according to Figure 9 It can be seen that when the simulated friction coefficients are 0.19 and 0.23 respectively, the calculated relative errors are all less than the set value of 5%, so the calibration process ends here. After comparison, the relative error is the smallest when the simulated friction coefficient is 0.23, so μ is determined. b =0.23 as the final calibration value.

[0085] Example 2:

[0086] Step 1: Using the physical experiment module and the data processing and analysis module, obtain the dynamic repose angle f of the granular material with a particle diameter of 4.8 mm in Example 2. exp =29.24°.

[0087] Step 2: At this time, count = 0, use the parallel simulation parameter generation module to generate parallel simulation parameters (the variable is the friction coefficient μ b ) and the corresponding simulation folder; and use the DEM simulation module for parallel simulation; then, send the simulation data to the data processing and analysis module to obtain the macro characteristics corresponding to each simulation (dynamic repose angle f s ), specifically Figure 10 shown.

[0088] Calculate the dynamic repose angle f for each set of simulations s and the physical dynamic repose angle f exp The relative error e, such as Figure 11 shown.

[0089] according to Figure 11 It can be seen that there are multiple sets of errors between the simulated dynamic repose angle and the physical dynamic repose angle that are less than the set value of 5%. Therefore, the calibration is completed, and the friction coefficient of 0.80 corresponding to the minimum relative error of 0.31% is output as the calibration value.

[0090] Although the present disclosure is disclosed as above, the protection scope of the present disclosure is not limited thereto. Those skilled in the art may make various changes and modifications without departing from the spirit and scope of the present disclosure, and these changes and modifications will fall within the protection scope of the present invention.

Claims

1. A DEM simulation model automatic calibration method based on parallel computing, characterized in that: include: Obtain physical macroscopic features. When the number of corrections count = 0, generate multiple sets of initial simulation microscopic parameters, perform parallel simulation and data processing on the multiple sets of initial simulation microscopic parameters, obtain simulation macroscopic features corresponding to the initial simulation microscopic parameters, send the multiple sets of simulation macroscopic features and the physical macroscopic features to a comparison and optimization module, and determine whether the relative error between the simulation macroscopic features and the physical macroscopic features is lower than a set threshold. If so, complete the calibration. If not, establish an intelligent model between the simulation macroscopic features and the physical macroscopic features based on the simulation data during the parallel simulation of the initial simulation microscopic parameters; When the number of corrections count = 1, the physical macro-features are input into the constructed intelligent model, and the corresponding simulation micro-parameters are obtained as initial values. Parallel simulation, data analysis and comparison of the initial values ​​are performed, and the micro-parameters corresponding to the simulation macro-features with the smallest error are obtained as the initial values ​​of the next round. Parallel simulation, data analysis and comparison of the initial values ​​are repeated until the relative error between the simulation macro-features and the physical macro-features of this round is lower than the set threshold or the number of corrections count reaches the set upper limit, and the calibration is completed.

2. The DEM simulation model automatic calibration method based on parallel computing according to claim 1 is characterized in that: The obtaining of physical macroscopic characteristics comprises: The images of the movement process of the granular material in the rotating drum are collected through the physical experiment module and sent to the data processing and analysis module for filtering and binarization. The repose angle data at different times are extracted, and the acquired repose angle data are fitted using a second-order model to obtain the dynamic repose angle as a physical macroscopic feature.

3. The DEM simulation model automatic calibration method based on parallel computing according to claim 1 is characterized in that: When the correction count is 0, generating multiple sets of initial simulation micro parameters, and performing parallel simulation and data processing on the multiple sets of initial simulation micro parameters includes: Use the parallel simulation parameter generation module to generate a parameter between 0 and 1 with a step size of Δμ b The input parameter μ b (1) ,…,μ b (k) ,μ b (k+1) ,…,μ b (n) , where μ b (1) =Δμ b ,μ b (k+1) =μ b (k) +Δμ b (k>0), And create corresponding simulation folders in batches; Use the DEM simulation module to perform parallel simulation and collect simulation data; The simulation data is sent to the data processing and analysis module to obtain the simulation macro characteristics corresponding to each input parameter.

4. The DEM simulation model automatic calibration method based on parallel computing according to claim 1 is characterized in that: The establishment of an intelligent model between the simulation macro-features and the physical macro-features includes: Establishing a simulation database between initial simulation microscopic parameters and their corresponding simulation macroscopic features; Preprocess the data, remove outliers, and organize the data; The least square method is used as the core fitting algorithm and the polynomial function is used as the fitting model to construct an intelligent model between the simulation macro-features and the physical macro-features.

5. The DEM simulation model automatic calibration method based on parallel computing according to claim 4 is characterized in that: The elimination of abnormal values ​​includes using a moving average method to obtain a smoothed value of the simulation macro-features of the simulation output, and judging whether the distance between the simulation macro-features and the smoothed value in each group of simulation data is greater than a set value. If so, the group of simulation data pairs is removed, otherwise it is retained.

6. The DEM simulation model automatic calibration method based on parallel computing according to claim 5 is characterized in that: The data sorting includes rearranging the simulation data after outliers are removed, with simulation macroscopic features as input and simulation microscopic parameters as output.

7. The DEM simulation model automatic calibration method based on parallel computing according to claim 1 is characterized in that: The parallel simulation, data analysis and comparison of the initial values ​​include: The initial value μ b0 As input, μ b0 Centered by Δμ b , generate equally spaced microscopic parameters to be simulated: μ b (1) ,…,μ b (i) ,…,μ b (k+1) ,…,μ b (n) , where: μ b (k+1) =μ b0 ,μ b (i) =μ b0 +(ik-1)Δμ b (i=1,...,n),n=2k+1; Based on the microscopic parameters to be simulated, batch create corresponding simulation folders; Using the microscopic parameters to be simulated, DEM simulation is performed in parallel to generate simulation data; Input the simulation data into the data processing and analysis module to obtain the macro characteristics of each simulation in this round; Determine whether the relative error between the simulated macro features and the physical macro features in this round is lower than the set threshold. If so, end the calibration. If not, select the micro parameters corresponding to the simulated macro features with the smallest error as the initial values ​​for the next round.

8. The DEM simulation model automatic calibration method based on parallel computing according to claim 7 is characterized in that: The parallel execution of DEM simulation includes: Open n DEM simulation software interfaces, and run the microscopic parameters μ to be simulated in each interface. b (1) ,…,μ b (n) The corresponding DEM simulation code; and store the corresponding simulation data, images and other related parameters.

9. The DEM simulation model automatic calibration method based on parallel computing according to claim 1 is characterized in that: When the calibration is completed, if the termination condition is that the relative error between the simulated macro features and the physical macro features of this round is lower than the set threshold, the micro parameter corresponding to the minimum relative error is selected as the final calibration value. If the termination condition is that the number of corrections count reaches the set upper limit, the micro parameter corresponding to the minimum relative error in all rounds is selected as the final calibration value.

10. A DEM simulation model automatic calibration system based on parallel computing, used to execute the DEM simulation model automatic calibration method based on parallel computing according to any one of claims 1 to 9, characterized in that: include: Physical experiment module, used to obtain physical experiment data; Data processing and analysis module, used to obtain physical macro characteristics based on physical experimental data, obtain simulation macro characteristics based on simulation data, and build intelligent models; A parallel simulation parameter generation module is used to generate corresponding simulation microscopic parameters according to the number of corrections; A DEM simulation module is used to perform parallel DEM simulation processing on multiple groups of simulation microscopic parameters generated by the parallel simulation parameter generation module; and a comparison and selection module for judging the completion of calibration and outputting the final calibration value.

Images