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 111
您当前的位置：首页 > 期刊文章 > 过刊浏览 > Volumes 108-119 (2025) > Volume 111
Research on particle motion behavior in gas-solid fluidized bed based on fluorescence tracing technology and machine learning
Yifan Wang a b c, Yangfan Xu a b c, Chenyang Zhou a b c, Ruikang Fan a b c, Ge Bai c, Yadong Zhang a b c, Yu Wang a b c, Yuanyuan Li a b c, Enhui Zhou a b c *
a Key Laboratory of Coal Processing and Efficient Utilization of Ministry of Education, China University of Mining and Technology, Xuzhou, 221116, China
b Jiangsu Key Laboratory for Clean Utilization of Carbon Resources, China University of Mining & Technology, Xuzhou, 221116, China
c School of Chemical Engineering & Technology, China University of Mining & Technology, Xuzhou, 221116, China
10.1016/j.partic.2026.01.031
Volume 111, April 2026, Pages 197-213
Received 8 December 2025, Revised 16 January 2026, Accepted 27 January 2026, Available online 7 February 2026, Version of Record 3 March 2026.
E-mail: zeh@cumt.edu.cn
Highlights

• A particle detection and tracking method based on machine learning.

• Fluorescent tracer particles were prepared to achieve clear imaging in fluidized beds.

• Movement behavior of tracer particles with different densities in fluidized beds.

• Influence of feeding heights on movement of tracer particles with different densities.


Abstract

Accurate quantification of particle motion in gas-solid separation fluidized beds is essential for optimizing separation efficiency and elucidating flow mechanisms. This study proposed a non-invasive particle tracking method that combines fluorescence tracing with a deep learning-based detection framework. Chemically synthesized spherical tracers were validated for spectral stability and mechanical integrity. A vision system utilizing YOLOv8n for detection, coupled with Kalman filtering and the Hungarian algorithm for trajectory association was developed. The results demonstrated that this system achieves high detection performance (mAP50≈0.91; Precision/Recall≥0.85) and effectively bridges missed detections by occlusion or motion blur. Analysis of the extracted kinematic parameters reveals a continuous transition in particle motion behavior governed by the competition between gravity and bubble-induced drag. High-density particles exhibited gravity-dominated unidirectional settling, whereas low-density particles showed bubble-dominated surface floating. Particles with densities near the bed density displayed large-scale circulation characterized by asymmetric slow settle-steep ascent patterns. Feeding height significantly influenced the settling depth of near-bed-density particles by altering initial kinetic energy, while its effect on high- and low-density conditions was negligible. This study provides a new option for analyzing particle dynamics in complex multiphase flows.

Graphical abstract
Keywords
Gas-solid separation fluidized bed; Fluorescent tracer; Detection-tracking vision system; Particle settling behavior; Particle dynamics
文章导读