Estimation of particle size in hydrocyclone underflow streams by use of Multivariate Image Analysis

被引：20

作者：

Aldrich, C. ^{[1
,2
]}

Uahengo, F. D. L. ^{[2
]}

Kistner, M. ^{[2
]}

机构：

[1] Curtin Univ Technol, Western Australian Sch Mines, Dept Min Engn & Met Engn, Perth, WA 6845, Australia

[2] Univ Stellenbosch, Dept Proc Engn, ZA-7602 Stellenbosch, South Africa

来源：

MINERALS ENGINEERING | 2015年 / 70卷

关键词：

Hydrocyclones; Particle size estimation; Multivariate Image Analysis; Discriminant analysis; TOMOGRAPHY; SEPARATION;

D O I：

10.1016/j.mineng.2014.08.018

中图分类号：

TQ [化学工业];

学科分类号：

0817 ;

摘要：

Photographic images were collected of the underflow slurry stream of a laboratory-scale hydrocyclone classifying Merensky, UG2 and Platreef platinum group metal ores. Textural descriptors of the images derived by means of a steerable pyramid algorithm could be used to predict the mean particle size of the underflow slurry streams. The model consisted of a linear discriminant classifier that first identified the underflow as comprising coarse, intermediate or fine particle flow. Use of the centroids or mean particle sizes of these three classes could explain on average approximately 81% of the variance in the mean particle sizes of the underflow streams. These measurements can be performed online, and could form the basis for more advanced control of hydrocyclones and mineral processing circuits in general. (C) 2014 Elsevier Ltd. All rights reserved.

引用

页码：14 / 19

页数：6

共 29 条

[1]

[Anonymous], 1995, PROCESS TOMOGRAPHY I

[2]

Bond J., 1999, P 1 WORLD C IND PROC, P102

[3] Particle size distribution soft-sensor for a grinding circuit [J].