Hybrid self-organising fuzzy logic PID controllers for muscle relaxant anaesthesia

Hybrid PID forms of SOFLC were investigated in computer simulation for controlling muscle relaxation in the operating theatre with a view to reducing computational requirements and comparing their control performance. A 5/spl times/5 fuzzy rule base was used and an efficient fuzzy inferencing procedure identified which uses singletons for input and output fuzzy sets. The performance of a self-organising fuzzy PD+I controller was compared with a self-organising PI+I controller in computer simulations with regard to noise tolerance and tracking ability with a fixed setpoint and changing process gain. Both self-organising fuzzy logic controllers have similar control performance under the simulated condition. The SOFPD+I controller output is considerably more variable than the SOFPI+I controller under high noise conditions but gives a faster response to genuine process changes under low noise conditions.