Multi-Objective Automated Machine Learning for Inversion of Mesoscopic Parameters in Discrete Element Contact Models
Accurate calibration of mesoscopic contact model parameters is essential for ensuring the reliability of Particle Flow Code in Three Dimensions (PFC3D) simulations in geotechnical engineering. Trial-and-error approaches are often used to determine the parameters of the contact model, but they are ti...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-07-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/15/15/8181 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849770914040774656 |
|---|---|
| author | Xu Ao Shengpeng Hao Yuyu Zhang Wenyu Xu |
| author_facet | Xu Ao Shengpeng Hao Yuyu Zhang Wenyu Xu |
| author_sort | Xu Ao |
| collection | DOAJ |
| description | Accurate calibration of mesoscopic contact model parameters is essential for ensuring the reliability of Particle Flow Code in Three Dimensions (PFC3D) simulations in geotechnical engineering. Trial-and-error approaches are often used to determine the parameters of the contact model, but they are time-consuming, labor-intensive, and offer no guarantee of parameter validity or simulation credibility. Although conventional machine learning techniques have been applied to invert the contact model parameters, they are hampered by the difficulty of selecting the optimal hyperparameters and, in some cases, insufficient data, which limits both the predictive accuracy and robustness. In this study, a total of 361 PFC3D uniaxial compression simulations using a linear parallel bond model with varied mesoscopic parameters were generated to capture a wide range of rock and geotechnical material behaviors. From each stress–strain curve, eight characteristic points were extracted as inputs to a multi-objective Automated Machine Learning (AutoML) model designed to invert three key mesoscopic parameters, i.e., the elastic modulus (<i>E</i>), stiffness ratio (<i>k<sub>s</sub></i>/<i>k<sub>n</sub></i>), and degraded elastic modulus (<i>E<sub>d</sub></i>). The developed AutoML model, comprising two hidden layers of 256 and 32 neurons with ReLU activation function, achieved coefficients of determination (<i>R</i><sup>2</sup>) of 0.992, 0.710, and 0.521 for <i>E</i>, <i>k<sub>s</sub></i>/<i>k<sub>n</sub></i>, and <i>E<sub>d</sub></i>, respectively, demonstrating acceptable predictive accuracy and generalizability. The multi-objective AutoML model was also applied to invert the parameters from three independent uniaxial compression tests on rock-like materials to validate its practical performance. The close match between the experimental and numerically simulated stress–strain curves confirmed the model’s reliability for mesoscopic parameter inversion in PFC3D. |
| format | Article |
| id | doaj-art-b8ed3ada01474e0b8e32bfd6ebc97e11 |
| institution | DOAJ |
| issn | 2076-3417 |
| language | English |
| publishDate | 2025-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj-art-b8ed3ada01474e0b8e32bfd6ebc97e112025-08-20T03:02:48ZengMDPI AGApplied Sciences2076-34172025-07-011515818110.3390/app15158181Multi-Objective Automated Machine Learning for Inversion of Mesoscopic Parameters in Discrete Element Contact ModelsXu Ao0Shengpeng Hao1Yuyu Zhang2Wenyu Xu3School of Resources and Safety Engineering, Chongqing University, Chongqing 400044, ChinaSchool of Resources and Safety Engineering, Chongqing University, Chongqing 400044, ChinaDepartment of Civil, Geological and Mining Engineering, Polytechnique Montréal, Montréal, QC H3T 1J4, CanadaSchool of Resources and Safety Engineering, Chongqing University, Chongqing 400044, ChinaAccurate calibration of mesoscopic contact model parameters is essential for ensuring the reliability of Particle Flow Code in Three Dimensions (PFC3D) simulations in geotechnical engineering. Trial-and-error approaches are often used to determine the parameters of the contact model, but they are time-consuming, labor-intensive, and offer no guarantee of parameter validity or simulation credibility. Although conventional machine learning techniques have been applied to invert the contact model parameters, they are hampered by the difficulty of selecting the optimal hyperparameters and, in some cases, insufficient data, which limits both the predictive accuracy and robustness. In this study, a total of 361 PFC3D uniaxial compression simulations using a linear parallel bond model with varied mesoscopic parameters were generated to capture a wide range of rock and geotechnical material behaviors. From each stress–strain curve, eight characteristic points were extracted as inputs to a multi-objective Automated Machine Learning (AutoML) model designed to invert three key mesoscopic parameters, i.e., the elastic modulus (<i>E</i>), stiffness ratio (<i>k<sub>s</sub></i>/<i>k<sub>n</sub></i>), and degraded elastic modulus (<i>E<sub>d</sub></i>). The developed AutoML model, comprising two hidden layers of 256 and 32 neurons with ReLU activation function, achieved coefficients of determination (<i>R</i><sup>2</sup>) of 0.992, 0.710, and 0.521 for <i>E</i>, <i>k<sub>s</sub></i>/<i>k<sub>n</sub></i>, and <i>E<sub>d</sub></i>, respectively, demonstrating acceptable predictive accuracy and generalizability. The multi-objective AutoML model was also applied to invert the parameters from three independent uniaxial compression tests on rock-like materials to validate its practical performance. The close match between the experimental and numerically simulated stress–strain curves confirmed the model’s reliability for mesoscopic parameter inversion in PFC3D.https://www.mdpi.com/2076-3417/15/15/8181mesoscopic parameter inversionPFC3D simulationsautomated machine learning (AutoML)uniaxial compression testslinear parallel bond contact model |
| spellingShingle | Xu Ao Shengpeng Hao Yuyu Zhang Wenyu Xu Multi-Objective Automated Machine Learning for Inversion of Mesoscopic Parameters in Discrete Element Contact Models Applied Sciences mesoscopic parameter inversion PFC3D simulations automated machine learning (AutoML) uniaxial compression tests linear parallel bond contact model |
| title | Multi-Objective Automated Machine Learning for Inversion of Mesoscopic Parameters in Discrete Element Contact Models |
| title_full | Multi-Objective Automated Machine Learning for Inversion of Mesoscopic Parameters in Discrete Element Contact Models |
| title_fullStr | Multi-Objective Automated Machine Learning for Inversion of Mesoscopic Parameters in Discrete Element Contact Models |
| title_full_unstemmed | Multi-Objective Automated Machine Learning for Inversion of Mesoscopic Parameters in Discrete Element Contact Models |
| title_short | Multi-Objective Automated Machine Learning for Inversion of Mesoscopic Parameters in Discrete Element Contact Models |
| title_sort | multi objective automated machine learning for inversion of mesoscopic parameters in discrete element contact models |
| topic | mesoscopic parameter inversion PFC3D simulations automated machine learning (AutoML) uniaxial compression tests linear parallel bond contact model |
| url | https://www.mdpi.com/2076-3417/15/15/8181 |
| work_keys_str_mv | AT xuao multiobjectiveautomatedmachinelearningforinversionofmesoscopicparametersindiscreteelementcontactmodels AT shengpenghao multiobjectiveautomatedmachinelearningforinversionofmesoscopicparametersindiscreteelementcontactmodels AT yuyuzhang multiobjectiveautomatedmachinelearningforinversionofmesoscopicparametersindiscreteelementcontactmodels AT wenyuxu multiobjectiveautomatedmachinelearningforinversionofmesoscopicparametersindiscreteelementcontactmodels |