Name | Last modified | Size | Description | |
---|---|---|---|---|
Parent Directory | - | |||
README.html | 2010-07-15 10:06 | 1.3K | ||
old/ | 2010-07-15 10:06 | - | ||
pse91.with_figures.pdf | 2010-07-15 10:06 | 1.2M | ||
Abstract.
By controllability (dynamic resilience) we generally mean the best
closed-loop performance achievable using any controller. Since the
controllability can not be altered by change of the control algorithm,
but only by design modifications, it follows that the term
controllability provides a link between process design and process
control. In this paper we focus on two aspects of controllability: The
plants sensitivity to disturbances and the limitations imposed by
interactions when using decentralized control. We use simple tools
such as the RGA, the PRGA (Performance RGA) and the closely related
Closed Loop Disturbance Gain (CLDG). For example, if k'th column
of the CLDG is large, then this indicates that disturbance k will be
difficult to reject. This may pinpoint the need for modifying the
process. The PRGA provides a measure of interaction which also
includes one-way coupling. In the paper we apply these measures to
distillation column control and fluid catalytic cracker (FCC) control.