Chapter 11 Introduction to MIMO Design
© National Instruments Corporation 11-3 Xmath Interactive Control Design Module
The standard feedback system has two vector input signals, r and dact, and
three vector output signals, e, u, and y. It can therefore be described by the
3×2 block matrix that relates the three output vector signals to the two
input vector signals:
The entries of this block matrix, that is, the transfer functions from r and
dact to e, u, and y, have standard names and interpretations (which agree
with the standard SISO notation):
The sensitivity transfer function is denoted S and given by
S=(I+PC)–1. The sensitivity transfer function is the transfer function
from reference input r to the error signal e.
The closed-loop transfer function T is given by T=PC(I+PC)–1. T is
the transfer function from r to y. T can be expressed in several other
ways, for example:
The actuator effort transfer function C(I+PC)–1 is the transfer
function from r to u, and so is related to the actuator effort required.
For example, its step response matrix shows the closed-loop step
responses from each reference input signal to each actuator signal.
The transfer function from dact to e, P(I+CP)–1, is denoted Sact and
called the actuator-referred sensitivity transfer function. The
actuator-referred sensitivity transfer function determines the errors
generated by actuator-referred disturbances. It also can be expressed as
(I+PC)–1P. Notice that it is “complementary” to the transfer function
described just above, that is, C(I+PC)–1, in the sense that the two
transfer functions can be obtained from each other by swapping P
and C.
The transfer function from dact to u, CP(I+CP)–1, is called the
actuator-referred actuator effort transfer function. Notice that it is
related to the closed-loop transfer function by swapping P and C. It can
also be expressed as C(I+PC)–1P.
The transfer function from dact to y, (–P)(I+CP)–1, is denoted Tact and
called the actuator-referred closed-loop transfer function.
e
u
y
IPC+()
1PI CP+()
1
CI PC+()
1CP I CP+()
1
PC I PC+()
1PICP+()
1
=r
dact
TPCICP+()
1IPC+()
1PC I S===