En
期刊文章
文章精选
当期目录
过刊浏览
高被引
高下载
特邀文章
专刊合集
待刊文章
Volume 115
您当前的位置：首页 > 期刊文章 > 过刊浏览 > Volumes 108-119 (2025) > Volume 115
DEM parameter calibration approach for cohesive ores based on PSO-BP neural network
Fangping Ye a b, Yanan Zhang a, Craig Wheeler b c, Bin Chen c, Chao Zhou a, Lei Nie a *
a Key Lab of Modern Manufacture Quality Engineering, Hubei University of Technology, Wuhan, 430068, China
b School of Engineering, The University of Newcastle, Callaghan, 2308, Australia
c TUNRA Bulk Solids, The University of Newcastle, Callaghan, 2308, Australia
10.1016/j.partic.2026.06.004
Volume 115, August 2026, Pages 309-320
Received 16 April 2026, Revised 26 May 2026, Accepted 3 June 2026, Available online 10 June 2026, Version of Record 19 June 2026.
E-mail: leinie@hbut.edu.cn
Highlights

• A PSO-BP based DEM calibration method is proposed for cohesive lateritic nickel ore.

• Dual PSO-BP model realizes fast macro-micro mapping and parameter inversion.

• Calibration errors are below 2%, and chute flow errors are within 5%.


Abstract

To enhance the calibration efficiency and accuracy of Discrete Element Method (DEM) parameters for cohesive bulk materials, a collaborative method integrating Particle Swarm Optimization (PSO) and Backpropagation (BP) neural networks is proposed. Key macroscopic indicators (steady-state shear stress, angle of repose) are obtained via Jenike shear and funnel tests across a 0-50% moisture range. Orthogonal experiments determine micro-parameters (e.g., static/rolling friction, surface energy) to build a macro-micro mapping database. The core of the PSO-BP dual-model lies in its collaborative mechanism: the forward BP model predicts macroscopic responses to replace time-consuming DEM simulations, while the PSO algorithm optimizes the inverse BP model to accurately infer optimal micro-parameters from experimental macro-indicators (steady-state shear stress, angle of repose). Validation shows low errors (1.14% for angle of repose, 1.63% for steady-state shear stress) and good chute flow velocity agreement. This method overcomes traditional limitations of arbitrariness and ignored parameter coupling, providing reliable support for DEM simulation and equipment design for cohesive bulk materials.

Graphical abstract
Keywords
Cohesive ores; DEM parameter calibration; PSO-BP neural network; Angle of repose; Shear test
文章导读