基于边界跟踪和神经网络的煤岩界面识别方法研究  

Coal-rock interface recognition method based on boundary tracking algorithm and artificial neural network

作　　者：吴德忠 刘泉声 黄兴[2,3] 高峰 殷欣 WU De-zhong;UU Quan-sheng;HUANG Xing;GAO Feng;YIN Xin(Key Laboratory of Geotechnical and Structural Engineering Safety of Hubei Province,School of Civil Engineering,Wuhan University,Wuhan 430072,China;State Key Laboratory of Coal Resources in Western China,Xi'an University of Science and Technology,Xi'an 710054,China;State Key Laboratory of Geotechnical Engineering,Institute of Rock and Soil Mechanics,Chinese Academy of Sciences,Wuhan 430071,China)

机构地区：[1]武汉大学土木建筑工程学院岩土与结构工程安全湖北省重点实验室,湖北武汉430072 [2]西安科技大学西部煤炭绿色开发国家重点实验室,陕西西安710054 [3]国科学院武汉岩土力学研究所岩土力学与工程国家重点实验室,湖北武汉430071

出　　处：《煤炭工程》2021年第6期140-146,共7页Coal Engineering

基　　金：西部煤炭绿色安全开发国家重点实验室开放基金课题“煤矿长距离斜井和深部巷道TBM掘进围岩挤压变形卡机灾害预测控制方法”(SKLCRKF1915);国家自然科学基金资助项目“基于多算法改进融合的TBM掘进岩-机作用模型与智能决策方法”(52074258);陕西省自然科学基础研究计划联合基金项目“煤矿掘进装备多元信息感知与多算法融合智能化掘进技术”(2021JLM-06)。

摘　　要：为实现煤岩界面精准识别,采集了鄂尔多斯李家壕煤矿矿区巷道掘进面原始图像,提出了基于机器学习的分类算法和基于数字图像处理的边界提取算法,该算法为提取连续单像素宽度边界提供了良好的基础。据此提出一种基于边界跟踪算法和人工神经网络的煤岩界面识别方法,从而为巷道掘进机和采煤机滚筒空间位置的调整提供依据。采用文章提出的方法对从陕西神木榆家梁采煤工作面采集的原始图像进行验证,提取到的煤岩边界与真实的煤岩界面基本吻合,验证了该方法的有效性和可靠性。In order to accurately recognize coal-rock interface, the images of coal mine roadway heading surfaces are collected in a mining area in Ordos, and a pixel classification algorithm based on machine learning and a boundary extraction algorithm based on digital image processing were proposed, which provided a good foundation for extracting continuous single-pixel width boundaries. Based on this, a coal-rock interface recognition method based on boundary tracking algorithm and artificial neural network was proposed, which provided a basis for the adjustment of the cutting height position of the roadheader and shearer drum. The method proposed in this paper was used to verify the original images collected from the coal mining face in Yujialiang Coal Mine, Shenmu, Shaanxi, the extracted coal-rock boundary was basically consistent with the real coal-rock interface, the effectiveness and reliability of this method was proved.

关 键 词：煤岩识别 边界跟踪 神经网络 形态学处理 

分 类 号：TD802[矿业工程—矿山开采]