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 108
您当前的位置：首页 > 期刊文章 > 过刊浏览 > Volumes 108-119 (2025) > Volume 108
Separation dynamic model of air dense medium fluidized bed for coal beneficiation and control
Gansu Zhang a b, Tianxin Li e, Hongyang Li c, Zhiqiang Li a b, Shuxian Su a b, Xuan Xu a b, Wei Dai d, Liang Dong a b d *
a International Joint Laboratory of Minerals Efficient Processing and Utilization, Ministry of Education, Xuzhou, 221116, China
b School of Chemical Engineering & Technology, China University of Mining & Technology, Xuzhou, 221116, China
c State Key Laboratory of Multimodal Artificial Intelligence Systems, Institute of Automation, Chinese Academy of Sciences, Beijing, 100190, China
d Artificial Intelligence Research Institute, China University of Mining & Technology, Xuzhou, 221116, China
e Ningxia Baofeng Energy Coal Preparation and Blending Technology Co., Ltd, Yinchuan, 751400, China
10.1016/j.partic.2025.11.019
Volume 108, January 2026, Pages 320-335
Received 8 October 2025, Revised 29 November 2025, Accepted 30 November 2025, Available online 6 December 2025, Version of Record 17 December 2025.
E-mail: dongl@cumt.edu.cn
Highlights

• The separation dynamics of ADMFB was modeled, containing dynamic and static information.

• A decoupling density model with self-balancing property was developed for control purposes.

• An ash consistency assumption for ash correction was proposed, with relative error around 5%.

• Model predictive control was implemented on ADMFB model, reducing control time by 46%.


Abstract

This study presents a comprehensive framework for the separation dynamic model of Air Dense Medium Fluidized Bed (ADMFB) to promote the intelligentization of coal preparation. The framework is a double-layer composite structure. In the outer layer, the static model of partition was extended from Dense Medium Cyclones to ADMFB. In the inner layer, the dynamic model of bed density was developed from the coupling model to the decoupling model with self-balancing characteristics. The model predicts product yield and ash content, which is validated against industrial separation data. The baseline performance is good but unstable, with maximum relative error 38.43 % and minimum relative error 4.48 % for coal ash content. To address this, an ash correction algorithm was innovatively put forward by introducing organic efficiency θ. The performance improvement is significant that coal ash content has maximum relative error 5.97 % and minimum relative error 0.48 %. Finally, Model Predictive Control (MPC) was implemented following the global linearization of the nonlinear bed density model, whose convergence speed to reach the tracking value is 46 % faster than constant control. Overall, this work provides a pathway for coal preparation plants to achieve more stable and intelligent production.

Graphical abstract
Keywords
Dry coal preparation; Separation dynamic model; Partition curve; Bed density regulation; Model predictive control
文章导读