【作者】 张晨；

【导师】 徐志强；

【作者基本信息】 中国矿业大学（北京） ， 检测技术与自动化装置， 2013， 博士

【摘要】 本文研究了现有煤和矸石的分选方法,在此基础上提出了利用激光三维扫描技术测量煤矸体积,使用称重装置称重,然后计算出密度,最后根据密度来识别煤和矸石的方法。本文所研究的方法,适用于对块煤和块矸的井下初选。本文详细研究了激光三维扫描技术的原理及应用,得出了光电识别体积测量装置的参数选取算法。详细研究了关于激光线提取的图像处理方法和算法,提出了适用于煤矸光电识别方法的激光线快速提取算法。设计了煤矸识别的实验装置。根据统计学中的中心极限定理,提出了煤矸的实际体积与测量体积的比值呈正态分布的统计规律,并进行了实验证实。在此规律基础上,建立了本方法用于煤矸识别的数据模型,给出了识别阈值的选取算法以及矸石识别率的算法。通过所设计的实验装置,对所提出理论及数学模型进行了实验验证,证明了所建立的数学模型及算法的正确性和可行性。更多还原

【Abstract】 This paper studies the exiting separation technology of coal and gangue and gives a new method to separate the gangue from coal according to the density which is calculated from the volume and weight of the ores. The method is used to separate lump gangue from lump coal in preliminary selection stage under ground.The volume is measured by using3D laser scanning technology and the weight is measured by using dynamic weighing technology. First this paper study in details about the3D laser scanning technology theory, then give the parameter selection formula for the photoelectric recognition device, which is used to measure the volume. The paper study in details about the algorithms of segmenting laser beam from the image, and then gives the algorithm of quickly segmenting laser beam, which is suitable for the coal and gangue photoelectric recognition devices. The paper gives the design of the photoelectric recognition devices and gets the distribution rule of voids between the measured volume and actual volume through a lot of experiments. A mathematical model is developed to separate the gangue from goal based on the distribution rule. The algorithm of the recognition threshold selecting and the algorithm of the percentage of the recognized gangue are given. By the experiment with the device designed by this paper, the theory and the mathematical model are verified.更多还原