Fuzzy control of mean arterial pressure in postsurgical patients with sodium nitroprusside infusion.

We developed a fuzzy control system to provide closed-loop control of mean arterial pressure (MAP) in postsurgical patients in a cardiac surgical intensive care unit setting by regulating sodium nitroprusside (SNP) infusion. The fuzzy controller, originally expert-system-based, was analytically converted to ten nonfuzzy control algorithms, which reduced execution time dramatically. The core of the control algorithms was a nonlinear proportional-integral (PI) controller whose proportional gain and integral gain adjusted continuously according to error and rate change of error of the process output. The gains became larger when process output was far from desired setpoint and smaller when process output was close to desired setpoint, resulting in more dynamic and stable control performance than the regular PI controller, especially when a linear process with time-delay or a nonlinear process was involved. The control algorithms, encoded in C programming language, were implemented to control MAP in patients. Preliminary clinical results showed that the average percentage of time in which MAP stayed between 90% and 110% of the MAP setpoint was 89.31%, with a standard deviation of 4.96%. These were calculated based on 12 patient trials, with total trial time of 95 and 13 min.