Title: Study of process dynamics in Floatex density separator based on computational intelligent vision

Authors: S.K. Mandal

Addresses: Materials Science and Technology Division, National Metallurgical Laboratory, Jamshedpur, Jharkhand 831 007, India

Abstract: This paper proposes a novel computational intelligent vision-based image processing and analysis techniques that can be used in Floatex density separator (FDS), a hindered settling classifier for recovery of quality products from minerals (iron, coal) in order to develop an optimum process control strategy suitable for manipulating complex variables characterised by relatively slow process dynamics. Image processing and analysis software, MATLAB 7.0 was used to process the high speed CCD camera image. The steps include calibration, contrast enhancement and segmentation. Image features are demonstrated and correlated with voidage, particle size distribution, bed height change and density mapping under specific feed rate, pulp density and teeter flow rate of the process. All these information are used for estimation of bed pressure, underflow density cut which will help in the development of accurate control system for reliable operation of FDS from a remote control station.

Keywords: computational intelligence; intelligent vision; computer vision; image processing; image analysis; contrast enhancement; calibration; Floatex density separator; FDS; product recovery; minerals; iron; coal; process control; process dynamics; calibration; contrast enhancement; image segmentation; bed pressure; underflow density cut; remote control.

DOI: 10.1504/IJCVR.2011.045268

International Journal of Computational Vision and Robotics, 2011 Vol.2 No.4, pp.314 - 322

Received: 08 May 2021
Accepted: 12 May 2021

Published online: 30 Jan 2012 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article