Search

The interaction between granular material and rigid bodies or complex boundaries are common in geotechnical engineering and industry. A hybrid MPDEM algorithm is proposed to simulate these kinds of problems. It is combined with the ability for solving granular material in MPM and the advantages of contact detection in DEM, which can be applied to solve sliding and impacting behavior of rigid blocks. Furthermore, the coupling algorithm based on GPU can be adopted for large-scale simulation to improve the calculation efficiency which has powerful advantages in geologic hazards simulation with high efficiency. To validate the algorithm, the sliding friction and impact effects are conducted to check the contact detection and force calculation for interaction between granular and block or boundary. Besides, the simulation results of granular impacting on the blocks are in good agreement with the experiment, which represents the feasibility and ability of algorithm to solve the interaction with granular materials and rigid-blocks or boundary systems. Through efficiency analysis, it is concluded that the computational performance is higher with the increase in particle number, compared with the pure MPM contact method. Additionally, the Tesla V100 card is a better choice for large-scale (particle number more than 1 million) simulations. [ABSTRACT FROM AUTHOR]

Landslides cause huge damage to social economy and human beings every year. Landslide susceptibility mapping (LSM) occupies an important position in land use and risk management. This study is to investigate a hybrid model which makes full use of the advantage of supervised learning model (SLM) and unsupervised learning model (ULM). Firstly, ten continuous variables were used to develop a ULM which consisted of factor analysis (FA) and k-means cluster for a preliminary landslide susceptibility map. Secondly, 351 landslides with "1" label were collected and the same number of non-landslide samples with "0" label were selected from the very low susceptibility area in the preliminary map, constituting a new priori condition for a SLM, and thirteen factors were used for the modeling of gradient boosting decision tree (GBDT) which represented for SLM. Finally, the performance of different models was verified using related indexes. The results showed that the performance of the pretreated GBDT model was improved with sensitivity, specificity, accuracy and the area under the curve (AUC) values of 88.60%, 92.59%, 90.60% and 0.976, respectively. It can be concluded that a pretreated model with strong robustness can be constructed by increasing the purity of samples. [ABSTRACT FROM AUTHOR]