Page 148 - Distributed model predictive control for plant-wide systems

122 Distributed Model Predictive Control for Plant-Wide Systems

The transfer functions of these three subsystems are, respectively,
4.05e −27s
G (s)=
1
50s + 1
5.72e −14s
G (s)=
2
60s + 1
7.20
G (s)=
3
19s + 1
[ ]
Both input and output constraints are −0.5 0.5 , and the sample period is 4 min. Let
the iterative condition be = 0.01, prediction condition P = 8, and control horizon M = 3.

0
y 1
–0.1 y 2
Outputs –0.2 y 3
–0.3
–0.4
0.4
u 1
0.2 u 2
u 3
Inputs 0
–0.2
–0.4
0 100 200 300 400 500 600
t/min
Figure 6.4 The outputs and inputs of each subprocess

0
Global optimization
–0.05 Nash optimization
Outputs –0.10
–0.15
–0.20
–0.25
0.4
Global optimization
0.3
Nash optimization
Inputs 0.2
0.1
0
–0.1
0 100 200 300 400 500 600
t/min
Figure 6.5 The outputs and inputs of the second subprocess