278
CHAPTER 7 THE ADMITTANCE
MODEL AND
NETWORK CALCULATIONS
When the coecient matrix
Y
bu
s
is symmetrical, which is almost always the
case, an important simplication results. As can be seen by inspection of Eq.
(7.68), when the rst column of L is divided by Yl1, we obtain the rst row of U;
when the second column of L is divided by Yi�\ we obtain the second row of U;
and so on for the other columns and rows of Eq. (7.68), provided �j
=
�'
i
'
Therefore, dividing the entries in each column of L by the principal diagonal
element in that column yields U
T
whenever
Y
"us is symmetrical. We can then
write
Y
II
Y
2I
y(l)
Y
I
I
22
L
=
uTn =
Y3
1
y
(l)
(7.75)
32
1
y
(
2
)
-
Y
II
y(l)
33
22
�
l
yO)
y(
2
)
4
2
4
3
1
y
(
3
)
Y
l
l
yell
y(
2
)
44
22
33
where diagonal matrix D contains the diagonal elements of L. Substituting in
Eq. (7.71) for L from Eq. (7.75), we obtain the nodal admittance equations in
the
f
orm
Y
V
=
uTnuv
=
I
b
u
s
(7.76)
Equation (7.76) may be solved for the unknown voltages V in three consecutive
steps as follows:
(7.77)
nv'
=
V"
(
7 .78)
= V'
(7.79)
These equations will be recognized as an extension of Eqs. (7.72). The interme
diate
result V" is rst found from Eq. (
7.77) by forward substitution. Next, each
entry in V' is calculated from Eq. (7.78) by dividing the corresponding element
of V" by the appropriate diagonal element of n. Finally, the solution V is
;obtained from Eq. (7.79) by back substitution as demonstrated in Example 7
.9.
Example 7.10. Using Eqs. (7.77) through (7.79), determine the solution vector V of
unknown voltages for the system and operating conditions of Example 7.9. ·