1212 
MONITORING 
200 
180 
160 
140 
120 ■ 
100- . 
80 
60 
40 
SICK DOG 
AUTOMATIC CONTROL 
10 
TIME {minutes) 
Figure 8. — The response to command changes in blood 
pressure with phase I control. 
T 1 r 
5 6 7 
TIME (minutes) 
Figure 10. — The weaning of the vasopressor I. V. dur- 
ing the recovery period. 
creased. At five minutes the loop is closed. To be 
noted is how well the model's output tracks the 
blood pressure. Figures 8 and 9 are direct con- 
tinuations of the previous figure. Figure 10 is 
the weaning period where the controlled discon- 
tinuance of the drug is the point of interest. 
Note that the blood pressure set is changed 
from time to time and how the blood pressure 
responds. This is phase I control ; that is, we are 
using steady state control. This accounts for the 
rather slow change in blood pressure after a 
command change. 
Figures 11 through 14 are examples of similar 
maneuvers except now phase III control is being 
used. Note the rapidity of response to a given 
command change or other perturbation. 
FUTURE WORK 
We are currently attempting to implement 
phase IV control where optimal solutions will be 
determined for all constraints. In addition, we 
are working on asymetrical control. The vaso- 
pressors we have dealt with to date cayise quite 
180 
160 
140 
120 
100- 
SICK DOG 
AUTOMATIC CONTROL 
■ PRESSURE SET POINT 
6 7 
TIME (minutes) 
Figure 9. 
-The response to command changes in blood 
pressure with phase I control. 
3/2 5/71 SICK DOG 
ELAPSED TIME= 50 MINUTES 
AUTOMATIC CONTROL 
PHASE m 
200- 
180 
160 
140 
120 
100- 
80 
60 
40 
AVERAGE BLOOD PRESSURE 
/ 
PREDICTED BLOOD PRESSURE 
PRESSURE SET POINT 
WEIGHTS PLOTTED 
Figure 11. — An example of phase III control. 
