井下铁矿石品位的测量与研究

发布时间：2018-03-19 13:12

本文选题：铁矿石　切入点：矿石品位　出处：《武汉理工大学》2015年硕士论文　论文类型：学位论文

【摘要】：随着我国经济的迅速发展,我国对矿产资源的消费需求也越来越大。近些年,采矿业发展迅速,这就对采矿技术提出了越来越高的要求。在开采矿产资源的过程当中,品位可以用来衡量矿石的品性和质量,品位是指矿石中有用组成部分的单位含量。矿石品位的高低直接反映了矿石品质的优劣程度。因此,找到一种合适的方法测量出矿石的品位,用来指导实际中的矿石的开采,将在很大程度上提高矿石的开采效率,增强了工厂的生产力和生产效益。本次课题研究的原始数据的采集来自于激光扫描仪的测量,利用MATLAB对数据进行处理。首先对扫描的数据进行三维数值模拟,得到多节车厢的三维网线图,再以一节车厢的扫描数据作为样本还原矿车表面的实际情况。其次,通过计算矿车表面小方块的面积占整个表面积的比例,还原小车表面的块矿比。对数据的处理只是课题的一部分,怎样通过数据去求得车厢内铁矿石的体积和松散系数也是亟待解决的一个问题。因为车厢内矿石是不规则的物体,所以,矿石的体积要用特殊的方法才能够计算。装载矿石的车厢是一个有规则长方体,这样可以很容易的计算出矿车的底面积,车厢内矿石的高度可以通过采集到的数据处理得到,再利用数值积分原理求得矿石的体积。求出了矿石的体积,进而就可以算出矿石的松散系数(松散系数是指土石料松动的体积与土石料未松动时的自然体积的比值。)。通过构造BP神经网络模型找出矿石的重量、体积及松散系数和矿石品位的某种对应关系。通过矿石的重量、体积和松散系数求出矿石的品位,进而与实验得出的品位相比较,矿石的块矿比用于松散系数的修正,从而缩小实验品位与计算品位的误差。构造BP神经网络模型,不断的去训练输入和输出的样本,得到最佳的网络权值矩阵,最终找到分析品位的计算方法,这种计算品位的方法将比实验方法更有效率,对于今后矿石品位的分析计量有着极其重要的作用。
[Abstract]:With the rapid development of China's economy, the consumption demand for mineral resources in China is also increasing. In recent years, the mining industry has developed rapidly, which puts forward higher and higher requirements for mining technology. In the process of mining mineral resources, Grade can be used to measure the quality and quality of ore. Grade refers to the unit content of useful components in ore. The grade of ore directly reflects the quality of ore. Finding a suitable way to measure the grade of the ore to guide the actual mining of the ore will greatly improve the mining efficiency of the ore. The raw data collected in this paper come from the measurement of laser scanner, and the data are processed by MATLAB. Firstly, the 3D numerical simulation of the scanned data is carried out. The three-dimensional net diagram of several cars is obtained, and then the scanning data of one car is taken as a sample to restore the actual situation of the car surface. Secondly, by calculating the proportion of the area of the small square on the surface of the car to the whole surface area, How to calculate the volume and loose coefficient of iron ore in the carriage through the data is also a problem to be solved urgently, because the ore in the car is an irregular object. Therefore, the volume of ore can only be calculated by a special method. The carriage containing the ore is a regular cuboid, so that the bottom area of the ore truck can be easily calculated. The height of the ore in the car can be obtained by processing the collected data, and the volume of the ore can be obtained by using the principle of numerical integration. Then the loose coefficient of ore can be calculated (loose coefficient is the ratio of the loose volume of soil and stone to the natural volume of soil and stone when it is not loose.) the weight of ore can be found by constructing a BP neural network model. According to the weight, volume and loose coefficient of the ore, the grade of the ore is calculated, and compared with the grade obtained from the experiment, the lump ore ratio of the ore is used to revise the loose coefficient. In order to reduce the error between experimental grade and calculation grade, the BP neural network model is constructed, the input and output samples are trained continuously, the optimal network weight matrix is obtained, and the calculating method of analysis grade is finally found. This method will be more efficient than the experimental method, and will play an extremely important role in the analysis and measurement of ore grade in the future.
【学位授予单位】：武汉理工大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TD861.1

【参考文献】