|
It is well known that most control problems can be adequately handled by the
Proportional-Integral-Derivative (PID) control strategy. In fact, many advanced
control algorithms and strategies are based on one form of PID or the other.
Moreover, most industrial process controls are handled by the standard PID
controller (Nagaraj et al., 2008) because of their simple structure and robustness
(Hugo, 2002) and the principles involved are learnt very easily.
Despite this popularity, the manual tuning of a PID controller is a very subjective
procedure which relies heavily on the knowledge and skill of the plant engineer or a
process operator (Rasmussen, 2002). Moreover, it is a tedious and time-consuming
task. The burden of manual tuning is compounded by the fact that most real life
processes contain tens of control loops that require separate tuning (Hugo, 2002).
This tedium may be due to the fact that more than 70% of industrial plants are
poorly tuned and potentially account for loss of revenue in terms of percentage of
defective products and energy utilization (Hugo, 2002; and Rasmussen, 2002).Furthermore, plant parameters are subject to change as operating conditions change
and as a result of aging, which then requires the re-tuning of the built-in process
controller(s).
In this work an online fast acting PID tuning scheme has been developed that
can quickly acquire (or re-adapt) the PID parameters during plant operation
automatically, and the proposed solution can be implemented on Peripheral Interface
Controller (PIC) hardware.
|
