Introduction - Advanced Control and Supervision of Mineral Processing Plants - page 13

A new set of sensors based on digital images has been widely applied in min-
eral processing. Owing to the multidimensional nature of the information provided
in the images, they pose new challenges for their integration with automation sys-
tems. Chapter four addresses these issues by first providing a brief description of the
following concepts and/or methods: (i) what a digital image is and what informa-
tion and/or features (color, textural, geometrical) can be extracted and/or quantified
from it; (ii) the multivariate statistical methods used in image analysis, such as prin-
cipal component analysis (PCA) and partial least squares regression, and finally (iii)
current image analysis techniques: multivariable image analysis (MIA), multivari-
ate image regression (MIR), and multi-resolution multivariate image analysis (MR-
MIA). After presenting various concepts and methods, a number of case studies and
applications to the mineral processing industry will be presented: (i) flotation froths:
concentrate grade prediction, froth health monitoring and froth appearance control
based on multivariate projection methods; and (ii) rock classification on conveyor
belts based on lithological compositions or grindability/hardness.
The fifth chapter deals with dynamic modeling, simulation and control of com-
minution circuits (crushing and grinding), the first stage of every mineral processing
plant. Short case studies highlight the use of various techniques in the control of such
circuits.
Since their first commercial application for mineral separation in the early1980s,
flotation columns have become a standard piece of equipment in mineral concentra-
tors particularly for cleaning operations. The sixth chapter presents and discusses the
most recent advances in dynamic modeling, instrumentation and automatic control
of flotation columns. It also examines how current industrial practice could benefit
from recent academic developments in these areas. A particular emphasis is placed
on the development of specific sensors for the continuous monitoring of process
operations and their regulation.
As a result of their success, many of these technological advances have been
marketed. Chapter 7 surveys these successes by reviewing commercial advanced
control systems, new sensors for grinding and flotation and system development
environments for mineral processing plants. It presents solutions based on expert
systems, fuzzy logic and neural networks currently available commercially. Tools
for linear and non-linear modeling, simulation, advanced control, monitoring, diag-
nosis and operator training are also reviewed. The chapter ends with a discussion of
benefits produced by advanced control in mineral processing.

References

[1] Hartman HL, Mutamsky JM (2002) Introductory Mining Engineering. John Wiley and Sons,

New Jersey

[2] Weiss NL (1985) SME Mineral Processing Handbook. Society for Mining Metallurgy &

Exploration, New York

[3] Lynch AJ (1979) Mineral Crushing and Grinding Circuits: Their Simulation, Design, and

Control (Developments in Mineral Processing Series, Vol.1). Elsevier Science Ltd., Amster-
dam