A New Method for Debris Flow Risk Assessment and Prediction: Tri-Strategy Crested Porcupine Optimizer-Categorical Boosting-Shapley Additive Explanations DOI
Hao Yang, Tianlong Wang,

Nikita Igorevich Fomin

et al.

Published: Jan. 1, 2024

Language: Английский

The rheological intelligent constitutive model of debris flow: A new paradigm for integrating mechanics mechanisms with data-driven approaches by combining data mapping and deep learning DOI
Tianxing Ma, Hao Chen, Keying Zhang

et al.

Expert Systems with Applications, Journal Year: 2025, Volume and Issue: 269, P. 126405 - 126405

Published: Jan. 5, 2025

Language: Английский

Citations

5
Intelligent rockburst level prediction model based on swarm intelligence optimization and multi-strategy learner soft voting hybrid ensemble DOI Creative Commons
Qinghong Wang, Tianxing Ma, Shengqi Yang

et al.

Geomechanics and Geophysics for Geo-Energy and Geo-Resources, Journal Year: 2025, Volume and Issue: 11(1)

Published: Jan. 23, 2025

Language: Английский

Citations

2
Assessment of Peak Particle Velocity of Blast Vibration using Hybrid Soft Computing Approaches DOI Creative Commons
Haiping Yuan,

Yangyao Zou,

Hengzhe Li

et al.

Journal of Computational Design and Engineering, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 16, 2025

Abstract Blasting vibration is a major adverse effect in rock blasting excavation, and accurately predicting its peak particle velocity (PPV) vital for ensuring engineering safety risk management. This study proposes an innovative IHO-VMD-CatBoost model that integrates variational mode decomposition (VMD) the CatBoost algorithm, with hyperparameters globally optimized using improved hippocampus optimization algorithm (IHO). Compared to existing models, proposed method improves feature extraction from signals significantly enhances prediction accuracy, especially complex geological conditions. Using measured data open-pit mine blasting, extracts key features such as maximum section charge, total horizontal distance, achieving superior performance compared 13 traditional models. It reports root mean square error of 0.28 cm/s, absolute 0.17 index agreement 0.993, variance accounted value 97.28%, demonstrating high degree fit observed data, overall robustness PPV prediction. Additionally, analyses based on SHapley Additive Explanations framework provide insights into nonlinear relationships between factors like distance improving model's interpretability. The demonstrates robustness, stability, applicability various tests, confirming reliability scenarios, offering valuable solution safe mining design.

Language: Английский

Citations

1
Prediction of thermal conductivity of natural rock materials using LLE-transformer-lightGBM model for geothermal energy applications DOI
Yufan Wang, Tianxing Ma, Liangxu Shen

et al.

Energy Reports, Journal Year: 2025, Volume and Issue: 13, P. 2516 - 2530

Published: Feb. 10, 2025

Language: Английский

Citations

1
Modified Herschel–Bulkley–Papanastasiou model considering particle size distribution to debris flow rheological properties DOI
Tianxing Ma, Rui Luo, Liangxu Shen

et al.

Physics of Fluids, Journal Year: 2025, Volume and Issue: 37(2)

Published: Feb. 1, 2025

Debris flow, a typical non-Newtonian fluid, exhibits rheological behavior significantly influenced by particle size distribution. Traditional models often struggle with applicability and predictive accuracy in complex systems. This study proposes modified Herschel–Bulkley–Papanastasiou (HBP) model, incorporating distribution parameters to dynamically adjust yield stress shear viscosity, enhancing its describing debris flow under varying gradations. The model distinguishes the roles of fine coarse particles: particles reduce resistance through lubrication effects, while enhance viscosity via interlocking effects. To validate series experiments were conducted on 14 gradation conditions. Results showed HBP achieved fitting coefficients between 0.933 0.990, outperforming traditional demonstrating superior adaptability across different distributions. was further integrated into OpenFOAM framework for three-dimensional simulations flume experiment. These considered wall friction dynamic free surface changes. Comparative analysis physical revealed accurately captured behavior, dynamics, pressure field distributions channel bed In summary, overcomes limitations parameters, offering more precise versatile rheology. work provides robust theoretical numerical tool advancing prediction mitigation engineering applications.

Language: Английский

Citations

1
A novel method for predicting debris flow hazard: a multi-strategy fusion approach based on the light gradient boosting machine framework DOI
Tianlong Wang, Qi Ge, Tianxing Ma

et al.

Stochastic Environmental Research and Risk Assessment, Journal Year: 2025, Volume and Issue: unknown

Published: March 12, 2025

Language: Английский

Citations

1
Accurate estimation of suspended sediment concentration integrated remote sensing information and a novel stacking machine learning model DOI
Xiaotian Fang, Jiahua Zhang, Xiang Yu

et al.

Stochastic Environmental Research and Risk Assessment, Journal Year: 2025, Volume and Issue: unknown

Published: Feb. 8, 2025

Language: Английский

Citations

0
A novel social network search and LightGBM framework for accurate prediction of blast-induced peak particle velocity DOI
Tianxing Ma,

Cuigang Chen,

Liangxu Shen

et al.

Frontiers of Structural and Civil Engineering, Journal Year: 2025, Volume and Issue: unknown

Published: April 24, 2025

Language: Английский

Citations

0
Prediction of force chains for dense granular flows using machine learning approach DOI

Ching-Hung Cheng,

Cheng-Chuan Lin

Physics of Fluids, Journal Year: 2024, Volume and Issue: 36(8)

Published: Aug. 1, 2024

Force chain networks among particles play a crucial role in understanding and modeling dense granular flows, with widespread applications ranging from civil engineering structures to assessing geophysical hazards. However, experimental measurement of microscale interparticle contact forces flows is often impractical, especially for highly complex flow systems. On the other hand, discrete-based simulation approaches suffer extremely high computational costs. Thus, this study proposes an innovative machine-learning framework aimed at accurately predicting force using particle-scale bulk-scale features, novel topological parameters. A deep neural network was developed, achieving excellent accuracy 94.7%, recall 100%, precision 90.3%, f1-score 95% non-Bagnold type flow, where chains significantly affect characteristics. In addition, enrich future application proposed model, we introduce experimentally accessible feature set, demonstrating effective performance detecting chains. More importantly, our analysis importance Shapley additive explanations values facilitates informed decision-making when identifying real-world experiments. The architecture will be interest essential any proves exceedingly challenging.

Language: Английский

Citations

2
Short-term power load forecasting based on SKDR hybrid model DOI
Yongliang Yuan,

Qingkang Yang,

Jianji Ren

et al.

Electrical Engineering, Journal Year: 2024, Volume and Issue: unknown

Published: Nov. 13, 2024

Language: Английский

Citations

2