Introduction - Advanced Control and Supervision of Mineral Processing Plants - page 12

1.6 Synopsis
This topic is organized into seven chapters highlighting the central role played by
measuring systems, modeling and control theory in the design of effective supervi-
sory and control strategies for mineral processing plants.
This first chapter has provided a summary of the main processes involved in a
concentrator plant. Their main characteristics have been described along with the
external and internal variables affecting their operation. A set of basic and advanced
functions have been defined to address the challenges imposed by operational prob-
lems. Possible benefits obtained from their implementation have been brieflyde-
scribed. Finally, a historical review of developments in automation systems in min-
eral processing plants, emphasizing their technical and economical achievements,
has explored possible future developments. The following chapters present the main
techniques required to implement advanced functions in modern automation sys-
tems.
The second chapter is devoted to the process of data acquisition for the purposes
of process performance audits, diagnosis and supervision, advanced automatic con-
trol, and real-time optimization. This is very important, since control and super-
visory systems rely on data acquired by numerous sensors scattered around in a
very hostile environment. The consistence and availability of this information plays
a key role in building robust systems. Emphasis is put on the constraints of mass
and energy conservation, which are used as a foundation for the design of measure-
ment strategies, the upgrading of measured values by measurement-error filtering
techniques and the estimation of unmeasured process variables. Since key variables
in any mineral processing unit are usually flow rates and solid and/or metal con-
centrations, their reconciliation with the laws of mass conservation is central to
the discussed techniques. Tools are proposed for three different kinds of operat-
ing regimes: steady-state, stationary and dynamic. These reconciliation methods are
based on standard least squares and Kalman filtering techniques. Case studies in-
volving grinding, flotation and leaching processes illustrate the problems of sensor
placement, data reconciliation, fault detection and diagnosis. Strategies for coupling
data reconciliation with real-time optimization and automatic control techniques are
also proposed.
As described in previous chapters an adequate control of mineral processing
plants relies on the availability of on-line sensors. Owing to harsh environments
or operating conditions, sensors may not always be available; mathematical mod-
els can then be used to provide an estimate of the missing measurements. The third
chapter addresses the need for soft sensors and their importance from a technical and
economical point of view. It also presents state-of-the-art technology and possible
future developments. Different classes of soft sensors, soft sensor models and their
design methodologies are presented along with soft sensor management systems
( e.g. monitoring, automatic updating and selection) which attest to the robustness of
their operation. Several examples illustrate their implementation in industrial appli-
cations.