煤岩显微组分识别技术研究应用现状  

Research and Application of Maceral Identification Technology in Coal and Rock

作　　者：周雪飞 王清昭 Zhou Xuefei;Wang Qingzhao(China Coal Electric Co.,Ltd.,Beijing,101300;School of Chemistry and Environmental Engineering,China University of Mining and Technology(Beijing),Beijing,100083)

机构地区：[1]中煤电气有限公司,北京101300 [2]中国矿业大学(北京)化学与环境工程学院,北京100083

出　　处：《当代化工研究》2025年第3期134-136,共3页Modern Chemical Research

摘　　要：煤岩组分分析在煤精细加工中具有独特优势,广泛应用于科研和加工领域。研究煤的显微组分与其物理、化学性质之间的关系,有助于理解成因并推动合理利用。随着科技进步,针对煤的复杂性和显微组分的重要性,传统分析引入成熟算法,降低分析成本和人工工作量。新图像分割策略结合两级K-means算法和形状信息,提升了准确性和纯度,同时利用综合特征和随机森林识别显微成分。深度学习在煤岩图像处理中取得成功,U-Net网络有效实现显微组分分割,借助图像预处理和增强技术优化分割结果。这些研究为显微组分分析提供新的方法,为未来研究奠定基础。The analysis of coal rock components has unique advantages in coal refinement and is widely used in research and processing fields.Studying the relationship between microscopic components of coal and its physical and chemical properties helps to understand its formation and promote rational utilization.With advancements in technology,traditional analysis methods have adopted mature algorithms to address the complexity of coal and the importance of microscopic components,thereby reducing analysis costs and labor intensity.New image segmentation strategies combine two-level K-means algorithms with shape information to improve accuracy and purity,while utilizing integrated features and random forests to identify microscopic components.Deep learning has achieved success in processing coal rock images,with the U-Net network effectively enabling the segmentation of microscopic components,optimizing segmentation results through image preprocessing and enhancement techniques.These studies provide new methods for component analysis and lay a foundation for future research.

关 键 词：煤岩显微组分分析 图像分析 深度学习 智能识别 

分 类 号：TD67[矿业工程—矿山机电]