Abstract
The present paper advocates the concept of a supervisory fuzzy controller. The proposed FLC algorithm employs a parallel architecture where an additional PD-type FLC can be employed either in conjunction with a conventional PID controller or a static fuzzy controller to improve system performance. The additional FLC implements a different rule base compared to a standard PD-type FLC, when employed alone. The effectiveness of the proposed scheme has been demonstrated by implementing it in simulation for two linear plant models of marine diesel engines and a nonlinear process plant. Its performance has been compared vis-à-vis a standard PD-type FLC, employed in simulation, for the same processes. The system performance is evaluated under both transient and steady state load disturbance conditions and also under noisy measurement environment.
