A First Course in FUZZY and NEURAL CONTROL

6.4. EXAMPLE: TEMPERATURE CONTROL 215

p2(i)=5*(exp(-time))*sin(w*time);

time=time+0.1;

end;

dp2=gradient(p2)

The plots in Figures 6.18 and 6.19 illustrate the generated test pattern vector
and its gradient.

Figure 6.18. Test pattern vector

Figure 6.19. Gradient of test pattern vector

We now simulate the trained networks to an arbitrary input and plot the re-
sulting variations in proportional and integral gain. These are shown in Figures
6.20 and 6.21.