煤矿工作面先验约束加奇点模型CT重构方法研究
发布时间：2018-09-08 09:52
【摘要】：在利用CT层析成像方法重构回采工作面内地质构造等异常区分布时,受回采工作面客观条件影响,投影角度有限,只能采用一边发射信号对边接收的方法,导致投影数据不完备,系数矩阵高度稀疏,从而无法精确重构出工作面内部的地质构造及其分布.针对该问题,本文提出一种新的基于总变分正则化的先验约束加奇点模型重构算法,通过加入回采工作面两巷揭露的地质信息作为先验约束条件,同时对特定区域引入奇点模型以提高反演精度,锐化断层影响区域的边界.经过数值计算和回采工作面现场试验,证明该算法能显著提高重构图像的精度,锐化断层影响区域的范围,显著改善异常区的识别效果.
[Abstract]:When CT tomography is used to reconstruct the distribution of geological structure and other abnormal areas in the mining face, due to the influence of the objective conditions of the mining face and the limited projection angle, the method of transmitting signals and receiving the signals on the other side can only be adopted. Because of incomplete projection data and sparse coefficient matrix, the geological structure and its distribution inside the face can not be reconstructed accurately. In order to solve this problem, this paper presents a new algorithm for reconstruction of priori constraint plus singularity model based on total variation regularization, which takes geological information exposed in two roadways of mining face as a priori constraint condition. At the same time, the singular point model is introduced to the specific area to improve the inversion accuracy and sharpen the boundary of the area affected by fault. Through numerical calculation and field test of mining face, it is proved that the algorithm can significantly improve the accuracy of reconstructed image, sharpen the area affected by fault, and improve the recognition effect of abnormal area.
【作者单位】： 中国矿业大学矿业工程学院煤炭资源与安全开采国家重点实验室;
【基金】：国家自然科学基金青年基金项目(51004102)
【分类号】：TD166
本文编号：2230177
[Abstract]:When CT tomography is used to reconstruct the distribution of geological structure and other abnormal areas in the mining face, due to the influence of the objective conditions of the mining face and the limited projection angle, the method of transmitting signals and receiving the signals on the other side can only be adopted. Because of incomplete projection data and sparse coefficient matrix, the geological structure and its distribution inside the face can not be reconstructed accurately. In order to solve this problem, this paper presents a new algorithm for reconstruction of priori constraint plus singularity model based on total variation regularization, which takes geological information exposed in two roadways of mining face as a priori constraint condition. At the same time, the singular point model is introduced to the specific area to improve the inversion accuracy and sharpen the boundary of the area affected by fault. Through numerical calculation and field test of mining face, it is proved that the algorithm can significantly improve the accuracy of reconstructed image, sharpen the area affected by fault, and improve the recognition effect of abnormal area.
【作者单位】： 中国矿业大学矿业工程学院煤炭资源与安全开采国家重点实验室;
【基金】：国家自然科学基金青年基金项目(51004102)
【分类号】：TD166
【相似文献】
相关期刊论文 前1条
1 骆遥;姚长利;;长方体磁场及其梯度无解析奇点表达式理论研究[J];石油地球物理勘探;2007年06期
，本文编号：2230177
本文链接：https://www.wllwen.com/kejilunwen/kuangye/2230177.html
