Machine Learning Algorithms Improve Accuracy in Numerical Simulation of Particle Flow Code

Strategic Insights -

Researchers at Liaocheng University in Shandong, People's Republic of China, have employed four machine learning algorithms to analyze the sensitivity of parameters in the numerical simulation of particle flow code (PFC2D). By using support vector machine (SVM), random forest (RF), gradient boosting decision tree (GBDT), and xtreme gradient boosting (XGBoost) algorithms, the researchers were able to increase the accuracy of the model's peak stress and elastic modulus. The study found that the RF algorithm outperformed other models in simulating the test set of mesoscopic parameters, with the highest trend evaluation index.

Key Takeaways:

  • The machine learning algorithms provide a better method for parameter calibration, aiding in a better understanding and prediction of micro-parameters for PFC2D rock models.
  • The research used four machine learning algorithms: support vector machine (SVM), random forest (RF), gradient boosting decision tree (GBDT), and xtreme gradient boosting (XGBoost).
  • The machine learning algorithms use 6 particle flow parameters encompassing 156 sets of data, as input variables, with the model's peak stress and elastic modulus (E) as output variables.
  • The parameters pb_coh and deform emod have the greatest positive impact on the model's peak stress and elastic modulus, respectively.
  • The RF algorithm outperforms other models in simulating the test set of mesoscopic parameters, with the highest trend evaluation index.
  • The research results indicate that machine learning algorithms provide a better method for parameter calibration than traditional methods.
  • The study was funded by the National Natural Science Foundation of China (NSFC), National Natural Science Foundation of China (NSFC), and Natural Science Foundation of Shandong Province.

Statistics:

  • The research used 156 sets of data as input variables for the machine learning algorithms.
  • Three performance evaluation metrics were used to assess the performance of the algorithms.
  • The RF algorithm outperformed other models with the highest trend evaluation index.
  • The parameters pb_coh and deform emod have the greatest positive impact on the model's peak stress and elastic modulus, respectively.

Sources:

  • Study On Micro-parameters of Parallel Bond Model Based On Machine Learning Algorithm. Computational Particle Mechanics, 2025.
  • NewsRx. Reports from Liaocheng University Highlight Recent Findings in Machine Learning (Study On Micro-parameters of Parallel Bond Model Based On Machine Learning Algorithm). Journal of Engineering. May 12, 2025; p 2691.