文献类型：期刊文献

英文题名：Modeling and simulation of vibrating ore drawing process based on visual perception and model predictive control

作者：Liu, Xu[1];Zhan, Kai[2];Zhu, Miao miao[3];Huang, Lingjuan[4];Jin, Feng[5];Liu, Jun[6];Wu, Zheng[7];Ma, Chao yang[8]

第一作者：Liu, Xu

通讯作者：Liu, X[1]

机构：[1]Univ Sci & Technol Beijing, Beijing Gen Res Inst Min & Met, Beijing Zhongse Xinda Technol Dev Co Ltd, Beijing, Peoples R China;[2]BGRIMM Technol Grp, Beijing, Peoples R China;[3]China Rongtong Secur Int Grp Co Ltd, Shenzhen, Peoples R China;[4]Beijing Union Univ, Teachers Coll, Beijing, Peoples R China;[5]Xidian Univ, Beijing Modern Res Inst Recycle Econ, Hangzhou Inst Technol, Xian, Peoples R China;[6]China Nonferrous Met Ind Technol Dev Co, Luoyang, Peoples R China;[7]Zhongse Asset Management Co Ltd, Beijing, Peoples R China;[8]BGRIMM Technol Grp, Beijing, Peoples R China

第一机构：Univ Sci & Technol Beijing, Beijing Gen Res Inst Min & Met, Beijing Zhongse Xinda Technol Dev Co Ltd, Beijing, Peoples R China

通讯机构：[1]corresponding author), Univ Sci & Technol Beijing, Beijing Gen Res Inst Min & Met, Beijing Zhongse Xinda Technol Dev Co Ltd, Beijing, Peoples R China.

年份：2026

卷号：42

期号：1

起止页码：163-183

外文期刊名：GOSPODARKA SUROWCAMI MINERALNYMI-MINERAL RESOURCES MANAGEMENT

收录：;WOS:【SCI-EXPANDED(收录号:WOS:001788321700007)】；

基金：This work was jointly supported by the National Key R&D Program of China (No. 2023YFC2907400) .

语种：英文

外文关键词：intelligent mining; vibrating ore drawing process; image detection; MPC

摘要：For modelling and control of the Vibrating Ore Drawing Process (VODP) in the under-mine rail transportation, a visual perception-based detection method for controlled variables in the drawing process is proposed and applied to the Model Predictive Control for achieving the adaptive draw of the chute. First, the method for estimating ore flow parameters is proposed based on a neural network visual perception method. The neural network-based target detection algorithm is constructed by the well-known DarkNet-53 structure, which is further optimized based on the YOLOv5-MINE structure. Second, the reference model of the VODP system is established by the system identification and data fitting approach. Then, based on this, we use model predictive control to control the system and give a stability analysis of the system with the input and output block diagram under the guidance of the prediction model. Finally, combined with advanced communication technology, simple simulation examples and practical industrial applications are given to illustrate the effectiveness and robustness of the proposed methodology. Field experiments conducted at an iron ore mine in China show that the application of the Visual Perception and Model Predictive Control system eliminates inefficiencies caused by human factors, resulting in a 6.8% increase in ore loading efficiency and a reduction in the need for operators by more than 50%. The proposed system provides a significant advancement in intelligent and unmanned mining operations, enhancing safety, efficiency, and resource utilization.