【作者】 黄仁东；

【导师】 古德生； 戴兴国；

【作者基本信息】 中南大学 ， 采矿工程， 2005， 博士

【摘要】 本文在全面检索、分析和评述国内外金属矿山隐患空区探测及隐患空区环境下工程安全控制研究已有成果和文献资料的基础上,采用理论分析、室内模拟实验和现场应用相结合的方法,系统地研究了金属矿山隐患空区声波层析成像识别及隐患空区环境下工程安全控制的理论与技术,并将研究成果应用于金属矿山隐患空区探测及其安全控制工作之中,取得了较好的效果。 论文首先论证了声波层析成像技术探测金属矿山隐患空区的技术可行性,给出了相应的理论依据。从地下声波波动方程出发,利用慧更斯-非涅耳原理、费马尔原理、克希霍夫绕射积分理论、反射透射原理探讨了声波在矿岩介质中的传播规律;从几何衰减、粘滞性吸收等几个方面研究了声波在矿岩介质中传播的衰减特性,分析并推导了特殊耗损参数——品质因子Q与吸收系数之间的关系。 研究了声波层析成像的技术与方法,提出了联合走时成像、振幅衰减成像的处理思路,并给出了综合层析成像的处理流程。 针对当前声波层析成像探测中信号处理不完善,多波叠加的信噪难以分离,初至时间难以确定的现状,论文提出了基于人工神经网络识别声波探测信号,确定初至时间的新思路,并首次以信号声音识别典型信号确定学习样本,建立了学习样本库;同时开发了一套与之配套的基于Dephi和Power builder开发的既能快速、准确处理数据和图像,又能方便、快捷管理数据的可视化自适应人工神经网络软件。 进行了金属矿山隐患空区声波层析成像探测的模拟实验,研究了空区形态特征、尺寸大小对探测结果的影响规律,获得了探测分辨率与隐患空区特征尺寸大小与波长的比值(D/λ_p)的关系,并分别以空气、乙醇、橡皮泥、甘油为充填介质进行模拟实验研究,获得了隐患空区中充填介质对探测图像的影响规律。 在现场探测试验方面,研究了一套适用于金属矿山隐患空区声波层析成像探测的实施技术和方法,获得了金属矿山隐患空区与层析成像显现特征之间的关系,及其图像识别与处理,平面定位及产状判定的综合技术。 最后论文结合厂坝铅锌矿群采空区环境下残矿的安全、高效回采及铜更多还原

【Abstract】 Based on the systematic study of the existed achievements and review of document, in combination with the theoretic analysis, simulation experiment and field application, the theories and technologies of exploration for hidden cavity and its safety control in metal mines have been studied in this paper. And some of achievements have been successfully applied in Changba Pb-Sn Mine and Xinqiao Pyrites Mine.The author suggests that the hidden cavity in metal mines can be explored with the Technology of Sonic Computerized Tomography. A case study is also provided to support it. From principles of wave theory and Huygens-F-renel, the law of sonic wave transmitting in the medium of ore-rock has been presented. And from 3D geometrical attenuation, stickiness absorption, heat exchange absorption, the attenuation character of sonic wave transmitting in the medium of ore-rock, and the special losing parameter-Q and its relationship with absorption coefficient has been presented.Owing to the status that signal is difficult to identify, difficult to separate from noise, and initial time is difficult to choose in sonic computerized tomography exploration, Tomgraphy arithmetic has been studied to improve interpretive precision. The technique, which combines travel-time tomography with amplitude-attenuation tomography and chooses initial time with Artificial Neural Network (ANN), is given. Meantime, a new method with signal sound to gain the standard sample was firstly introduced. Based on the Dephi and Power Builder, the software with self-adaptive modified feedforward and feedback BP Artificial Neural Network was developed. And its dealing process has been studied.Simulation experiments have been done to find the influence of cavity shape and size on the test results. The relationship between the resolution and the ratio (D/ λ )of cavity size to wave length has been determined.更多还原