Cumulative gangue mixing ratio prediction model for image-based in situ coal/gangue identification

IF 2.8 3区工程技术 Q1 MATHEMATICS, INTERDISCIPLINARY APPLICATIONS

Computational Particle Mechanics Pub Date : 2025-06-17 DOI:10.1007/s40571-025-01001-3

Jinwang Zhang, Jialin Zhao, Geng He, Xiaohang Wan, Melih Geniş, Haobo Zhang, Weijie Wei, Lianghui Li, Ahmet Özarslan, Dongliang Cheng, Jingzheng Wang

{"title":"Cumulative gangue mixing ratio prediction model for image-based in situ coal/gangue identification","authors":"Jinwang Zhang, Jialin Zhao, Geng He, Xiaohang Wan, Melih Geniş, Haobo Zhang, Weijie Wei, Lianghui Li, Ahmet Özarslan, Dongliang Cheng, Jingzheng Wang","doi":"10.1007/s40571-025-01001-3","DOIUrl":null,"url":null,"abstract":"<div><p>Image-based in situ coal/gangue identification has emerged as a pivotal tool for monitoring instantaneous gangue mixing ratios (IGMR) in fully mechanized top coal caving operations. However, intelligent coal caving control requires dynamic optimization based on the \"top coal recovery rate–cumulative gangue mixing ratio (CGMR)\" curve. This study establishes a predictive framework linking IGMR to CGMR through numerical simulations and machine learning. The authors proposed a particle swarm optimization–random forest (PSO–RF) hybrid model that outperforms conventional RF, achieving <i>R</i><sup>2</sup> values of 0.937 (advancing direction) and 0.962 (layout direction). Feature importance analysis reveals scraper speed, coal caving position, and sequential/interval caving strategies as dominant factors influencing CGMR. Physical experiments validate the model's robustness, demonstrating a 56% reduction in prediction error compared to baseline methods.</p></div>","PeriodicalId":524,"journal":{"name":"Computational Particle Mechanics","volume":"12 4","pages":"1913 - 1932"},"PeriodicalIF":2.8000,"publicationDate":"2025-06-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Computational Particle Mechanics","FirstCategoryId":"5","ListUrlMain":"https://link.springer.com/article/10.1007/s40571-025-01001-3","RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"MATHEMATICS, INTERDISCIPLINARY APPLICATIONS","Score":null,"Total":0}

引用次数: 0

Abstract

Image-based in situ coal/gangue identification has emerged as a pivotal tool for monitoring instantaneous gangue mixing ratios (IGMR) in fully mechanized top coal caving operations. However, intelligent coal caving control requires dynamic optimization based on the "top coal recovery rate–cumulative gangue mixing ratio (CGMR)" curve. This study establishes a predictive framework linking IGMR to CGMR through numerical simulations and machine learning. The authors proposed a particle swarm optimization–random forest (PSO–RF) hybrid model that outperforms conventional RF, achieving R² values of 0.937 (advancing direction) and 0.962 (layout direction). Feature importance analysis reveals scraper speed, coal caving position, and sequential/interval caving strategies as dominant factors influencing CGMR. Physical experiments validate the model's robustness, demonstrating a 56% reduction in prediction error compared to baseline methods.

Abstract Image

查看原文本刊更多论文

基于图像原位煤/矸石识别的累积矸石混合比预测模型

基于图像的煤/矸石原位识别已成为综放作业中监测瞬时矸石混合比（IGMR）的关键工具。而智能放煤控制需要基于“顶煤回收率-累计矸石掺量”曲线进行动态优化。本研究通过数值模拟和机器学习建立了连接IGMR和CGMR的预测框架。提出了一种粒子群优化-随机森林（PSO-RF）混合模型，其R2值分别为0.937（前进方向）和0.962（布局方向），优于传统的随机森林模型。特征重要性分析表明，刮板速度、放煤位置和顺序/间隔放煤策略是影响CGMR的主要因素。物理实验验证了模型的稳健性，与基线方法相比，预测误差降低了56%。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Computational Particle Mechanics Mathematics-Computational Mathematics

CiteScore

5.70

自引率

9.10%

发文量

期刊介绍： GENERAL OBJECTIVES: Computational Particle Mechanics (CPM) is a quarterly journal with the goal of publishing full-length original articles addressing the modeling and simulation of systems involving particles and particle methods. The goal is to enhance communication among researchers in the applied sciences who use "particles'''' in one form or another in their research. SPECIFIC OBJECTIVES: Particle-based materials and numerical methods have become wide-spread in the natural and applied sciences, engineering, biology. The term "particle methods/mechanics'''' has now come to imply several different things to researchers in the 21st century, including: (a) Particles as a physical unit in granular media, particulate ﬂows, plasmas, swarms, etc., (b) Particles representing material phases in continua at the meso-, micro-and nano-scale and (c) Particles as a discretization unit in continua and discontinua in numerical methods such as Discrete Element Methods (DEM), Particle Finite Element Methods (PFEM), Molecular Dynamics (MD), and Smoothed Particle Hydrodynamics (SPH), to name a few.